100382 VOLUME 3 DISEASE CONTROL PRIORITIES • THIRD EDITION Cancer DISEASE CONTROL PRIORITIES • THIRD EDITION Series Editors Dean T. Jamison Rachel Nugent Hellen Gelband Susan Horton Prabhat Jha Ramanan Laxminarayan Volumes in the Series Essential Surgery Reproductive, Maternal, Newborn, and Child Health Cancer Mental, Neurological, and Substance Use Disorders Cardiovascular, Respiratory, Renal, and Endocrine Disorders HIV/AIDS, STIs, Tuberculosis, and Malaria Injury Prevention and Environmental Health Child and Adolescent Development Disease Control Priorities: Improving Health and Reducing Poverty DISEASE CONTROL PRIORITIES Budgets constrain choices. Policy analysis helps decision makers achieve the greatest value from limited available resources. In 1993, the World Bank published Disease Control Priorities in Developing Countries (DCP1), an attempt to systematically assess the cost-effectiveness (value for money) of interventions that would address the major sources of disease burden in low- and middle-income countries. The World Bank’s 1993 World Development Report on health drew heavily on DCP1’s findings to conclude that specific interventions against noncommunicable diseases were cost-effective, even in environments in which substantial burdens of infection and undernutrition persisted. DCP2, published in 2006, updated and extended DCP1 in several aspects, including explicit consideration of the implications for health systems of expanded intervention coverage. One way that health systems expand intervention coverage is through selected platforms that deliver interventions that require similar logistics but deliver interventions from different packages of conceptually related interventions, for example, against cardiovascular disease. Platforms often provide a more natural unit for investment than do individual interventions. Analysis of the costs of packages and platforms—and of the health improvements they can generate in given epidemiological environments—can help to guide health system investments and development. DCP3 differs importantly from DCP1 and DCP2 by extending and consolidating the concepts of platforms and packages and by offering explicit consideration of the financial risk protection objective of health systems. In populations lacking access to health insurance or prepaid care, medical expenses that are high relative to income can be impoverishing. Where incomes are low, seemingly inexpensive medical procedures can have catastrophic financial effects. DCP3 offers an approach to explicitly include financial protection as well as the distribution across income groups of financial and health outcomes resulting from policies (for example, public finance) to increase intervention uptake. The task in all of the DCP volumes has been to combine the available science about interventions implemented in very specific locales and under very specific conditions with informed judgment to reach reasonable conclusions about the impact of intervention mixes in diverse environments. DCP3 ’s broad aim is to delineate essential intervention packages and their related delivery platforms to assist decision makers in allocating often tightly constrained budgets so that health system objectives are maximally achieved. DCP3 ’s nine volumes are being published in 2015 and 2016 in an environment in which serious discussion continues about quantifying the sustainable development goal (SDG) for health. DCP3 ’s analyses are well-placed to assist in choosing the means to attain the health SDG and assessing the related costs. Only when these volumes, and the analytic efforts on which they are based, are completed will we be able to explore SDG-related and other broad policy conclusions and generalizations. The final DCP3 volume will report those conclusions. Each individual volume will provide valuable, specific policy analyses on the full range of interventions, packages, and policies relevant to its health topic. More than 500 individuals and multiple institutions have contributed to DCP3. We convey our acknowledgments elsewhere in this volume. Here we express our particular gratitude to the Bill & Melinda Gates Foundation for its sustained financial support, to the InterAcademy Medical Panel (and its U.S. affiliate, the Institute of Medicine of the National Academy of Sciences), and to the External and Corporate Relations Publishing and Knowledge division of the World Bank. Each played a critical role in this effort. Dean T. Jamison Rachel Nugent Hellen Gelband Susan Horton Prabhat Jha Ramanan Laxminarayan VOLUME 3 DISEASE CONTROL PRIORITIES • THIRD EDITION Cancer EDITORS Hellen Gelband Prabhat Jha Rengaswamy Sankaranarayanan Susan Horton © 2015 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW, Washington, DC 20433 Telephone: 202-473-1000; Internet: www.worldbank.org Some rights reserved 1 2 3 4 18 17 16 15 This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) http://creativecommons.org /licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution—Please cite the work as follows: Gelband, H., P. Jha, R. Sankaranarayanan, and S. Horton, editors. 2015. Cancer. Disease Control Priorities, third edition, volume 3. Washington, DC: World Bank. doi:10.1596/978-1-4648-0349-9. License: Creative Commons Attribution CC BY 3.0 IGO Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by The World Bank. Third-party content—The World Bank does not necessarily own each component of the content contained within the work. The World Bank therefore does not warrant that the use of any third-party-owned individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to re-use a component of the work, it is your responsibility to determine whether permission is needed for that re-use and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. All queries on rights and licenses should be addressed to the Publishing and Knowledge Division, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org. Softcover Hardcover ISBN (paper): 978-1-4648-0349-9 ISBN: 978-1-4648-0350-5 ISBN (electronic): 978-1-4648-0369-7 DOI: 10.1596/978-1-4648-0349-9 DOI: 10.1596/978-1-4648-0350-5 Cover photo: © IAEA Imagebank/Dana Sacchetti/IAEA. Used with permission; further permission required for reuse. Cover and interior design: Debra Naylor, Naylor Design, Washington, DC Library of Congress Cataloging-in-Publication Data Cancer (Gelband) Cancer / editors, Hellen Gelband, Prabhat Jha, Rengaswamy Sankaranarayanan, Susan Horton. p. ; cm. — (Disease control priorities ; volume 3) Includes bibliographical references and index. ISBN 978-1-4648-0349-9 (alk. paper) — ISBN 978-1-4648-0350-5 (alk. paper) — ISBN 978-1-4648-0369-7 (electronics) I. Gelband, Hellen, editor. II. Jha, Prabhat, 1965- , editor. III. Sankaranarayanan, R. (Rengaswamy), 1952- , editor. IV. Horton, Susan, editor. V. World Bank, issuing body. VI. Title. VII. Series: Disease control priorities ; v. 3 [DNLM: 1. Neoplasms—economics. 2. Neoplasms—prevention & control. 3. Cost of Illness. 4. Developing Countries. 5. Health Services Research—economics. WA 395] RC262 362.19699’400681—dc23 2015019371 Contents Foreword xi Preface xiii Abbreviations xv 1. Summary 1 Hellen Gelband, Prabhat Jha, Rengaswamy Sankaranarayanan, Cindy L. Gauvreau, and Susan Horton PART 1 BURDEN 2. The Changing Global Burden of Cancer: Transitions in Human Development and Implications for Cancer Prevention and Control 23 Freddie Bray and Isabelle Soerjomataram PART 2 INTERVENTIONS 3. Breast Cancer 45 Benjamin O. Anderson, Joseph Lipscomb, Raul H. Murillo, and David B. Thomas 4. Cervical Cancer 69 Lynette Denny, Rolando Herrero, Carol Levin, and Jane J. Kim 5. Oral Cancer: Prevention, Early Detection, and Treatment 85 Rengaswamy Sankaranarayanan, Kunnambath Ramadas, Hemantha Amarasinghe, Sujha Subramanian, and Newell Johnson 6. Colorectal Cancer 101 Linda Rabeneck, Susan Horton, Ann G. Zauber, and Craig Earle 7. Treating Childhood Cancer in Low- and Middle-Income Countries 121 Sumit Gupta, Scott C. Howard, Stephen P. Hunger, Federico G. Antillon, Monika L. Metzger, Trijn Israels, Mhamed Harif, and Carlos Rodriguez-Galindo 8. Liver Cancer 147 Hellen Gelband, Chien-Jen Chen, Wendong Chen, Silvia Franceschi, Sir Andrew Hall, W. Thomas London, Katherine A. McGlynn, and Christopher P. Wild ix 9. Cancer Pain Relief 165 James Cleary, Hellen Gelband, and Judith Wagner PART 3 POLICY, CANCER SERVICES, AND RESEARCH 10. Global Hazards of Tobacco and the Benefits of Smoking Cessation and Tobacco Taxes 175 Prabhat Jha, Mary MacLennan, Frank. J. Chaloupka, Ayda Yurekli, Chintanie Ramasundarahettige, ´ksi, Samira Asma, and Prakash C. Gupta Krishna Palipudi, Witold Zaton 11. Cancer Services and the Comprehensive Cancer Center 195 Mary Gospodarowicz, Joann Trypuc, Anil D'Cruz, Jamal Khader, Sherif Omar, and Felicia Knaul 12. Screening for Cancer: Considerations for Low- and Middle-Income Countries 211 Terrence Sullivan, Richard Sullivan, and Ophira M. Ginsburg 13. Surgical Services for Cancer Care 223 Anna J. Dare, Benjamin O. Anderson, Richard Sullivan, C. S. Pramesh, Cheng-Har Yip, Andre Ilbawi, Isaac F. Adewole, Rajendra A. Badwe, and Cindy L. Gauvreau 14. Radiation Therapy for Cancer 239 David A. Jaffray and Mary K. Gospodarowicz 15. Need for National Commitments to Cancer Research to Guide Public Health Investment and Practice 249 Edward L. Trimble, Preetha Rajaraman, Ann Chao, Thomas Gross, Carol Levin, You-Lin Qiao, Timothy Rebbeck, Lisa Stevens, and Fang-hui Zhao PART 4 ECONOMICS 16. Cancer in Low- and Middle-Income Countries: An Economic Overview 263 Susan Horton and Cindy L. Gauvreau 17. Financing Cancer Care in Low-Resource Settings 281 Felicia Knaul, Susan Horton, Pooja Yerramilli, Hellen Gelband, and Rifat Atun 18. An Extended Cost-Effectiveness Analysis of Publicly Financed HPV Vaccination to Prevent Cervical Cancer in China 295 Carol Levin, Monisha Sharma, Zachary Olson, Stéphane Verguet, Ju-Fang Shi, Shao-Ming Wang, You-Lin Qiao, Dean T. Jamison, and Jane J. Kim DCP3 Series Acknowledgments 307 Volume and Series Editors 309 Contributors 311 Advisory Committee to the Editors 315 Reviewers 317 Index 319 x Contents Foreword When the biopsy results confirmed that I had oral with extensive data and probing analyses, both mor- cancer, I was 18 years old. If it sounded like a death sen- tality and suffering from cancer can be dramatically tence, there was reason for that thought. Survival rates reduced, even in the less affluent countries, through a from cancer were very low in those days, especially in combination of preventive measures (of which tobacco the poorer countries in the world (I was then a student control is the most well-known and frustratingly in Calcutta), and statistics offered very little reason for underused avenue), early diagnosis (distressingly low cheer. Now, at the age of 81, I can not only celebrate the for cancers in which early detection is not difficult to fact that I made it, with help from heavy-dose radiation, achieve and would make a major difference, such as but also that the battle against cancer in the world is oral, cervical, and breast cancer as well as the cancers increasingly being won. that afflict children), and of course early treatment However, the victory is not only partial, it is also (including well-established procedures as well as newly deeply uneven. With early diagnosis and effective treat- developed methods). ment, almost two-thirds of the people who get cancer The lesson that emerges from the well-aimed empirical in high-income countries now survive. In low- and analyses presented in this volume is not only that a major middle-income countries, only half of that propor- difference can be made in the incidence, management, tion—no more than one-third—make it. and elimination of cancer, even in the poorer countries This wonderfully illuminating book tells us about of the world, but that this can be done in cost-effective the state of the battle against cancer, but it also takes and affordable ways. Understanding and determination on the challenge of making lives better—and longer— are the deficiencies most in need of change. particularly in the poorer countries of the world. As the This is, ultimately, a cheerful book on a very grim chapters in this state-of-the-art book on cancer show, subject. It is also a hugely important invitation to action. Amartya Sen Thomas W. Lamont University Professor Harvard University Cambridge, Massachusetts Nobel Laureate, Economic Sciences 1998 xi Preface The burden of cancer in low- and middle-income control interventions that will have near-term and countries (LMICs) is large and growing. By contrast, long-range benefits. resources to control cancer in LMICs, either from Serious progress in cancer prevention and treatment domestic budgets or international aid, have not increased began about half a century ago in high-income countries. proportionately. Most populations in LMICs lack access The knowledge that has fueled progress is available imme- to effective cancer prevention, treatment, and palliation. diately for LMICs. In some cases, newer and better tech- This volume, Cancer, part of the 3rd edition of Disease nologies are now available: HPV testing can replace the Control Priorities, provides an up-to-date review of more resource- and infrastructure-intensive Pap smear the effectiveness, cost-effectiveness, cost, and feasibility for cervical cancer screening. Newer screening tests for of interventions for cancers that impose high disease colon cancer have similar advantages. Increasing national burdens in LMICs. incomes and broader national health coverage in middle- We propose an “essential package” of feasible inter- income countries, in particular, have already made a range ventions that countries can use in cancer planning, of services available to a wider swath of the population. knowing that some countries are well along in providing The pace needs to be accelerated and efforts can be broad- many of the elements. We recognize that the essential ened in low-income countries, where numbers of deaths cancer package may not be immediately feasible in from cancer are still relatively low, but increasing. low-income countries and only partially so in many Regarding tobacco—still the single most important middle-income countries. The package is not intended cancer-causing agent the world over—LMICs have the to limit cancer control to these measures, but we are knowledge to avert the epidemic that has now begun suggesting that these measures are likely to save large to subside in high-income countries. At the same numbers of lives at an affordable cost and should be pri- time, LMICs are underequipped to combat the tactics oritized by the public sector before large investments are of multinational tobacco companies. In a few cases, made in interventions that will have more limited effects. national treasuries profit from state-owned tobacco Local cancer patterns and resource availability may dic- companies. tate somewhat different priorities, and these should also Certain neglected areas are of special concern. guide national cancer planning. Progress is all but nonexistent in providing adequate Smoking cessation reduces the risks of developing pain control and palliative care, even in middle-income various cancers reasonably quickly, but other preven- countries. Limited progress has been made in cancer tive measures, such as vaccinations against cervical or registration and cause of death reporting. Very little liver cancer, will take longer to manifest full effects. progress is evident in documenting the costs and cost- Many types of cancer, which are not currently prevent- effectiveness of interventions in LMICs for even the able, will remain. Thus, the best approach to lowering highest-burden cancers. And very few clinical trials in the cancer burden is a system that promotes prevention cancer take place in LMICs. As a result, much of the as well as early detection and treatment. This volume evidence included in this volume is from high-income provides evidence that policy makers at all levels can countries, which we and our many co-authors have rein- use to support the immediate ramp-up of cancer terpreted as realistically as possible for LMICs. xiii We thank our dozens of co-authors for working Sir George Alleyne, Dr. Christopher Wild, and Sir Richard tirelessly, responding to several reviews, and producing Peto acted as special advisors for the volume, providing evidence that can be understood and acted on. We also guidance and wise counsel. give our thanks to the Cancer Surveillance Section of Cindy Gauvreau coordinated all aspects of the the International Agency for Research on Cancer for volume production, including chapter content and the custom maps and graphs in the volume and to the consistency. She vastly improved the quality of the National Cancer Institute, particularly the Center for volume that you see, and we are grateful for her Global Health, for supporting the work in many ways. many contributions. Many more individuals provided The Bill & Melinda Gates Foundation’s core support thoughtful comments, guidance, and encouragement; for DCP3, through the University of Washington, has we thank them all. made the whole enterprise possible. Others in the pro- The tide has been turned against cancer in high- cess also deserve our thanks: the Institute of Medicine income countries and can be in the rest of the world, for coordinating critical reviews and the World Bank armed with evidence and bolstered by political resolve. publishing staff for their wholehearted collaboration. This volume is intended to spur that effort. Hellen Gelband Prabhat Jha Rengaswamy Sankaranarayanan Susan Horton xiv Preface Abbreviations ADA American Dental Association ADR adenoma detection rate ALDH aldehyde dehydrogenase AML acute myeloid leukemia APL acute promyelocytic leukemia ASIR age-specific incidence rate BCS breast-conserving surgery BHGI Breast Health Global Initiative BL Burkitt lymphoma BMI body mass index BSE breast self-examination CBC complete blood count CBE clinical breast examination CEA cost-effectiveness analysis CI confidence interval CIN cervical intraepithelial neoplasia CISNET Cancer Intervention and Surveillance Modeling Network CME continuing medical education CMF cyclophosphamide, methotrexate, and 5-fluorouracil CRC colorectal cancer CT computed tomography CTC computed tomographic colonography CVG cost per vaccinated girl DALY disability-adjusted life year DCIS ductal carcinoma in situ ECEA extended cost-effectiveness analysis EDP early detection and prevention EPI Expanded Program for Immunization ER estrogen receptor FAC 5-fluorouracil, doxorubicin (®Adriamycin), and cyclophosphamide FAP familial adenomatous polyposis FCTC Framework Convention on Tobacco Control FIT fecal immunochemical test FS flexible sigmoidoscopy Gavi Gavi, the Vaccine Alliance GDP gross domestic product xv gFOBT guaiac fecal occult blood test GICR Global Initiative for Cancer Registry Development GNI gross national income GOPI Global Opioid Policy Initiative GTFRCC Global Task Force on Radiotherapy for Cancer Control HAU Hospice Africa Uganda HBsAg hepatitis B surface antigen HBV hepatitis B virus HCC hepatocellular carcinoma HCV hepatitis C virus HDI Human Development Index HDV hepatitis D virus Hib Haemophilus influenzae type B HICs high-income countries HIV human immunodeficiency virus HL Hodgkin lymphoma HPV human papillomavirus HR high-risk HSIL high grade squamous intraepithelial lesion IAEA International Atomic Energy Agency IAHPC International Association for Hospice and Palliative Care IARC International Agency for Research on Cancer ICD International Classification of Diseases ICER incremental cost-effectiveness ratio ICRCSN International Colorectal Cancer Screening Network IHC immunohistochemistry IMRT intensity modulated radiation therapy INCB International Narcotics Control Board IT information technology JCI Joint Commission International LEEP loop electrosurgical excision procedure LICs low-income countries LLETZ large loop excision of the transformation zone LMICs low- and middle-income countries LR low-risk LYS life-years saved MICs middle-income countries MISCAN microsimulation screening analysis MMG mammography MRI magnetic resonance imaging MRM modified radical mastectomy NAFD non-alcoholic fatty liver disease NCCN National Comprehensive Cancer Network NCD noncommunicable disease NCI National Cancer Institute NIAAA National Institute of Alcohol Abuse and Alcoholism NWTS National Wilms Tumor Study OECD Organisation for Economic Co-operation and Development OSMF oral submucous fibrosis PAF population attributable fraction PAHO Pan American Health Organization PBCR population-based cancer registry PET positron emission tomography xvi Abbreviations PODC Pediatric Oncology in Developing Countries PPP purchasing power parity PSA prostate-specific antigen QALY quality-adjusted life-year RCC regional cancer center RCT randomized controlled trial RT radiotherapy SEER Surveillance, Epidemiology, and End Results SES socioeconomic status SIL squamous intraepithelial lesion SLN sentinel lymph node SPS Seguro Popular de Salud SSP sessile serrated polyp TLS tumor lysis syndrome TNM tumor, nodes, metastasis TRM treatment-related mortality UCI Uganda Cancer Institute UHC universal health coverage UICC Union for International Cancer Control UMIC upper-middle-income country UNOP Unidad Nacional de Oncologia Pediátrica US ultrasound USMSTF U.S. Multi-Society Task Force on Colorectal Cancer/American Cancer Society USPSTF U.S. Preventive Services Task Force VAD vascular access device VIA visual inspection with acetic acid VIAM magnified visual inspection with acetic acid VLP virus-like particles VSL value of statistical life WBC white blood cell WHO World Health Organization WTO World Trade Organization YLL years of life lost Abbreviations xvii Chapter 1 Summary Hellen Gelband, Prabhat Jha, Rengaswamy Sankaranarayanan, Cindy L. Gauvreau, and Susan Horton INTRODUCTION The DCP3 package includes prevention strategies, but many cancers cannot be prevented to any great extent At the 2012 World Health Assembly, member states agreed by available methods. Some can be treated effectively to a goal of reducing rates of premature death from non- (breast and childhood cancers, for example), however, communicable diseases (NCDs) by 25 percent by 2025, and the availability of effective treatment bolsters public starting from a 2008 baseline (WHO 2011a, 2011b). The confidence in the overall program (Brown and others United Nations (UN) Sustainable Development Goals for 2006; Knaul and others 2011; Sloan and Gelband 2007). 2030, announced in September 2015, will include reducing Cancer control programs can mobilize broad political premature death from NCDs, of which cancer is a substan- support, as happened in Mexico with the addition of tial part (map 1.1). breast cancer and childhood cancer treatment into This chapter summarizes the analyses and conclu- expanded national health insurance coverage (Knaul sions of the 79 authors of this volume on cancer, Disease and others 2012). Control Priorities, 3rd edition (DCP3 Cancer), and ana- In high-income countries (HICs), most who lyzes interventions for effectiveness, cost-effectiveness, develop cancer survive, although survival depends affordability, and feasibility in low- and middle-income strongly on the type of cancer (table 1.1). In LMICs, countries (LMICs; see box 1.1 for key messages). The less than one-third survive, and in some the propor- intent is to help governments of LMICs commit to tion is much smaller (Ferlay and others 2015). The locally appropriate national cancer control strategies differences in survival are due partly to differences in that will include a range of cost-effective interventions, the patterns of cancer incidence; some types of can- customized to local epidemiological patterns and avail- cer that are common in many LMICs, such as lung, able funding, and to convey this commitment widely esophagus, stomach, and liver cancers, have a poor to their populations. Where affordable treatment can prognosis even in HICs (Bray and Soerjomataram be provided, conveying this to the public can motivate 2015, chapter 2 in this volume). The other major con- people to seek treatment when their cancer is at an ear- tributor to poor outcomes is that many fewer people lier, much more curable stage. Providing a package of come for treatment when their cancer is at an early, services that addresses a large part of the cancer burden curable stage than in HICs (Allemani and others 2015; will go a long way toward helping countries reach the Ferlay and others 2015). new NCD goals. DCP3 Cancer is one of nine planned The aim of DCP3 is to identify cost-effective, feasible, volumes in the DCP3 series (box 1.2). and affordable interventions that address significant Corresponding authors: Hellen Gelband, Center for Disease Dynamics, Economics & Policy, Gelband@cddep.org; and Prabhat Jha, Centre for Global Health Research, St. Michael’s Hospital and Dalla Lana School of Public Health, University of Toronto, Prabhat.Jha@utoronto.ca. 1 MAP 1.1 Cancer Mortality Before Age 70 Years, by World Bank Income Groupings, 2012 High-income Cancer deaths (thousand): Mainly tobacco-related 300 Mainly infection-related 130 Other 570 Total 1,000 Upper-middle-income Cancer deaths (thousand): Mainly tobacco-related 560 Mainly infection-related 540 Other 810 Total 1,910 Source: Based on WHO Global Health Estimates (WHO 2012) 2 Cancer IBRD 41651 | MAY 2015 Lower-middle-income Cancer deaths (thousand): Mainly tobacco-related 260 Mainly infection-related 260 Other 650 Total 1,170 Low-income Cancer deaths (thousand): Mainly tobacco-related 70 Mainly infection-related 90 Other 190 Total 350 Summary 3 Box 1.1 Key Messages Cancer is already a major cause of death in low- and Health Organization best buys, which are limited middle-income countries (LMICs), particularly in to interventions that are deliverable in primary care middle-income countries, and will increase as a per- settings. centage of deaths in all LMICs, driven by population The DCP3 essential package of cost-effective and aging and faster declines in other causes of death. feasible interventions would, if fully implemented, In most populations, helping current tobacco cost an additional $20 billion per year, or 3 percent users to quit and young people not to start smoking of total public spending on health in LMICs; 2.6 per- are the most urgent priorities in cancer prevention cent in upper-middle-income countries (UMICs); (and in the control of other noncommunicable and 5 percent in lower-middle-income countries; diseases), along with vaccination against hepatitis but 13 percent in low-income countries (LICs). In B and the human papillomavirus (HPV). Higher per capita terms, this would cost $5.70, $1.70, and tobacco taxes and accompanying interventions will $1.70 annually in UMICs, lower-middle-income, reduce cancer incidence and generate substantial and LICs, respectively. Such increases are potentially extra revenues for governments. feasible in all but the LICs, which would require external support. Other than tobacco- and virus-related cancers, however, most of the increase in cancer incidence is Cancer services that are considered appropriate not currently preventable, but many cases of cancer for a national cancer strategy should be covered can be effectively treated. Early breast cancer and through universal health coverage as soon as coun- cervical cancer are common, and often curable; pre- tries are able to do so. cancerous cervical lesions are even more curable. Global initiatives for cancer control in LMICs Childhood cancers are relatively rare, but some are needed to lower the costs of key inputs for the are highly curable. The interventions supported by essential package, including large-scale commodity the analyses in this Disease Control Priorities, 3rd purchases; to expand technical assistance; and to edition (DCP3 Cancer) go beyond current World promote cancer research. disease burdens in LMICs (box 1.3). Accordingly, we their affordability and feasibility, which differ mark- have examined the following: edly between low-, lower-middle-, and upper-middle- income countries. Even within the same income 1. The avoidable burden of premature death (defined as categories, countries may differ widely in epidemio- before age 70, which approximates current global life logical patterns and health systems, resulting in dif- expectancy) from cancer in LMICs (table 1.1) ferent country-specific essential packages. Hence, this 2. The main effective interventions for the prevention, is not intended to lead to a common cancer plan for early detection, treatment, and palliation of cancer, all LMICs, but to identify elements that will be appro- and their cost-effectiveness priate in many countries and spur discussion within 3. The costs and feasibility of developing health countries about rational cancer control planning and system infrastructure that could deliver progres- implementation. The result would be national cancer sively wider coverage of a set of cost-effective cancer plans that are tailored to local conditions but retain services. the characteristics of effectiveness, cost-effectiveness, feasibility, and affordability. Finally, we review some Using these inputs, we define an “essential package” ways in which global initiatives could help LMICs to of cost-effective interventions for cancer and discuss expand cancer control. 4 Cancer Box 1.2 From the Series Editors of Disease Control Priorities, 3rd Edition Budgets constrain choices. Policy analysis helps inexpensive medical procedures can have cata- decision makers achieve the greatest value from strophic financial effects. DCP3 offers an approach limited available resources. In 1993, the World Bank that explicitly includes financial protection as well as published Disease Control Priorities in Developing the distribution across income groups of the finan- Countries (DCP1), an attempt to assess systemat- cial and health outcomes resulting from policies (for ically the cost-effectiveness (value for money) of example, public finance) to increase intervention interventions that would address the major sources uptake (Verguet, Laxminarayan, and Jamison 2015). of disease burden in low- and middle-income coun- The task in all the volumes has been to combine the tries [Jamison and others 1993]). The World Bank’s available science about interventions implemented 1993 World Development Report on health drew in very specific locales and under very specific con- heavily on the findings in DCP1 to conclude that ditions with informed judgment to reach reasonable specific interventions against noncommunicable conclusions about the impacts of intervention mixes diseases were cost-effective, even in environments in in diverse environments. The broad aim of DCP3 is which substantial burdens of infection and under- to offer, for consideration and adaptation, essential nutrition persisted. intervention packages—such as the essential cancer DCP2, published in 2006, updated and extended package in this volume—and their related delivery DCP1 in several respects, including explicit con- platforms. This information will assist decision sideration of the implications for health systems makers in allocating budgets so that health system of expanded intervention coverage (Jamison and objectives are maximally achieved. others 2006). One way that health systems expand The nine volumes of DCP3 are being published intervention coverage is through selected platforms in 2015 and 2016 in an environment in which that deliver interventions that require similar logis- serious discussion continues about quantifying the tics but address heterogeneous health problems. sustainable development goal (SDG) for health (UN Platforms often provide a more natural unit for 2015). The analyses in DCP3 are well-placed to assist investment than do individual interventions, and in choosing the means to attain the health SDG and conventional health economics has offered little assessing the related costs. The final volume will understanding of how to make choices across plat- explore SDG-related and other broad policy con- forms. Analysis of the costs of packages and plat- clusions and generalizations, based on the analytic forms—and of the health improvements they can findings from the full set of volumes. Each indi- generate in given epidemiological environments— vidual volume will provide valuable, specific policy can help guide health system investments and analyses on the full range of interventions, packages, development. and policies relevant to its health topic. DCP3 differs substantively from DCP1 and Dean T. Jamison DCP2 by extending and consolidating the con- Rachel Nugent cepts of platforms and packages and by offering explicit consideration of the financial risk protection Hellen Gelband objective of health systems. In populations lacking Susan Horton access to health insurance or prepaid care, medical expenses that are high relative to income can be Prabhat Jha impoverishing. Where incomes are low, seemingly Ramanan Laxminarayan Summary 5 Table 1.1 Worldwide Cancer Deaths in 2012 at Ages 0–69 by Cancer Site and Country Income Grouping, and 5-Year Survival Rates in Low-, Middle-, and High-Income Countries Annual Deaths, age 0–69 years (thousands) by World Bank country 5-year survival (%), income group cancer registry data Lower- Upper- Low or Low middle middle High World middle High Population in billions income income income income (total) income income Cause of cancer and other deaths 0.8 2.4 2.3 1.2 6.7 5.5 1.2 Cancer, by site (ICD-10 C00-99) Lung, mouth, and esophagus 70 260 560 300 1,200 10 20 Liver 30 90 270 60 440 10 20 Breast 30 140 110 80 360 75 90 Stomach 20 80 210 50 360 20 40 Colon or rectum 20 80 120 100 310 50 60 Cervix 40 90 60 20 200 55 65 Ovary 8 30 30 30 100 25 40 Leukemia, age 0–14 years 3 10 10 2 30 65 90 age 15–69 years 10 40 60 30 140 30 50 Prostate 4 10 20 20 60 70 90 Other/unknown site 110 330 470 310 1,220 — — All cancers (% of all causes) 350 (6%) 1,170 (6%) 1,920 (22%) 1,000 (37%) 4,400 (14%) — — All noncommunicable diseases 1,660 6,300 5,950 2,200 16,070 — — Communicable/external causes 4,100 7,380 2,650 500 14,660 — — All causes 5,760 13,680 8,600 2,700 30,730 — — Sources: Population and mortality based on data from the UN Population Division (UNPD 2012) and WHO Global Health Estimates (WHO 2012). Estimated 5-year survival based on Allemani and others 2015. Note: Number of deaths above 10,000 are rounded to the nearest 10,000, so totals may differ. Estimated five-year survival rounded to the nearest 5 percent. — = Not applicable. and Bray and Soerjomataram 2015; Ferlay and others EVOLVING CANCER BURDEN 2015; WHO 2012). The WHO’s International Agency for Research on Cancer Worldwide, cancer death rates are slowly decreasing (IARC) estimates that in 2012 there were 14 million new (table 1.2). Between 2000 and 2010, age-standardized cases of cancer and 8 million deaths from cancer, more cancer death rates before age 70 years fell by about than half of them in people younger than age 70 years 1 percent per year, bolstered by worldwide declines in (table 1.1) (Ferlay and others 2015). Of the 4.4 million cervical cancer and stomach cancer, for reasons that are cancer deaths before age 70, 3.4 million were in LMICs; not fully understood. Male lung cancer rates decreased 1.9 million in UMICs, 1.2 million in lower-middle-in- in some countries, but in lower-middle-income coun- come countries, and 0.3 million in LICs. Two-thirds of tries, the death rates from tobacco-associated cancers the cancer deaths before age 70 years in LMICs were can- rose slightly. cers of the lung, mouth, or esophagus (0.9 million, many Absolute numbers of cancer deaths and cancer as a caused by tobacco), liver (0.4 million, many caused by proportion of all deaths will continue to rise because vaccine-preventable hepatitis B infection), stomach (0.3 of three factors: world population is increasing, par- million), breast (0.3 million), cervix (0.2 million, many ticularly in later middle age and old age; mortality caused by human papillomavirus [HPV] infection), and from diseases other than cancer is decreasing; and colon or rectum (0.2 million) (See table 1.1, figure 1.1; in some major populations the effects of tobacco are 6 Cancer Box 1.3 Methods The 79 authors of the 18 chapters in this volume economies were also sought. A recent review of the surveyed the published and gray literature to identify cost-effectiveness of cancer interventions for high- cost-effective interventions for the cancers studied. income countries (HICs) was also useful (Greenberg Cancer-specific incidence and mortality data are from and others 2010). The studies identified used various the International Agency for Research on Cancer’s outcome measures: life years saved, quality-adjusted GLOBOCAN (Ferlay and others 2015). Mortality life years (QALYs) gained, and disability-adjusted life data are from the World Health Organization’s Global years averted. Evidence from studies in LMICs was Health Estimates (WHO 2012), and demographic preferred, but rarely available. Evidence from HICs estimates are from the United Nations (UNPD 2012). was considered in all cases, and evidence from high- The analyses were stratified by World Bank income Asian economies was particularly important. country group classifications as defined by 2013 We adopted the scale used by the Commission on per capita gross national income: 34 low-income Macroeconomics and Health (2001) to define very countries (less than US$1,045), 50 lower-middle- cost-effective, cost-effective, and cost-ineffective as income countries (US$1,046 to US$4,125), and costing < 1, 1–3, and > 3 times per capita income per 55 upper-middle-income countries (US$4,126 to QALY (or other measure), respectively. (Commission US$12,745) (World Bank 2014a). on Macroeconomics and Health 2001). Cost-effectiveness estimates were compiled for The essential package includes interventions each cancer and each intervention. Systematic searches rated as very cost-effective and cost-effective and were conducted in PubMed for all interventions considered potentially affordable and feasible in covered in the volume, for studies in or including resource-constrained environments. Costs are low- and middle-income countries (LMICs), pub- expressed in 2012 prices. Costs are also expressed as lished in 2003–13 (Horton and Gauvreau 2015). For a percentage of national public spending on health, colorectal cancer, studies from high-income Asian estimated by the World Bank (World Bank 2014b). increasing (Jha 2009). Based on population growth Nkusu 2013; John and others 2011), because in many alone (at 2010 death rates age-standardized to the LMICs, cancer surgery, radiotherapy, and chemother- expected 2030 population), more than 6 million can- apy are paid for largely out of pocket. In Bangladesh cer deaths are expected in 2030 in people younger (Hamid, Ahsan, and Begum 2014) and Cameroon than age 70 years, and an equal number in people (Ilbawi, Einterz, and Nkusu 2013), for example, high age 70 years and older (table 1.3). Three-quarters of user fees increase the likelihood that patients will these future cancer deaths are expected to occur in not return at all for cancer surgery. Conversely, in middle-income countries (MICs). India, some standard types of cancer surgery (for example, mastectomy) are supposed to be provided at low, affordable cost in public hospitals; in China, the national health insurance scheme now offers stan- IMPACT ON HOUSEHOLDS dard types of cancer surgery at prices most people can Counter to common perceptions, cancer death rates afford. Nevertheless, even in China and India, cancer are often higher in lower-income groups than high- can still impose a major financial burden on families, income groups. In India, the age-standardized death especially in the lowest income groups, and in the case rate from cancer in middle age was twice that in illit- of India, cancer services are limited to certain large erate as in educated populations (Dikshit and others cities (Mallath and others 2014). An objective of DCP3 2012). As are other NCDs, cancer is an important is to evaluate interventions for their distributive effects, cause of catastrophic health expenditures that can push with particular emphasis on the effects on the poor and households into poverty (Hamid, Ahsan, and Begum on impoverishment at any economic level because of 2014; Hoang Lan and others 2013; Ilbawi, Einterz, and health care expenses. Summary 7 Figure 1.1 Incidence and Mortality of Selected Cancers before Age 70 For MICs that already have cancer treatment centers Years, Low- and Middle-Income Countries, 2012 and clinics, the DCP3 approach could be used to help evaluate additional interventions, now or in the future, 800 or to re-assess some current activities; in all LMICs it could help ensure due consideration of how interven- tions that are considered locally appropriate can achieve high population coverage. 600 The WHO has formulated a list of NCD best buys for New cases or deaths (000’s) LMICs, which were limited to services considered feasible at the primary care level. Those most relevant to cancer are three preventive measures: a set of tobacco control 400 interventions, hepatitis B vaccination to prevent liver cancer, and some form of screening and treatment for precancerous cervical lesions (WHO 2011b). The DCP3 Cancer essential package (table 1.4) adds HPV vaccination 200 (also included by the Commission on Investing in Health [Jamison and others 2013]) to prevent cervical cancer. The DCP3 also adds treatment of early-stage cervical cancer 0 (Denny and others 2015, chapter 4 in this volume); diag- Breast Lung Liver Cervix Stomach Colon- nosis and treatment for early breast cancer (Anderson and rectum others 2015, chapter 3 in this volume); diagnosis and treat- Cancer site ment for selected, highly curable childhood cancers (Gupta and others 2015, chapter 7 in this volume); and palliative Incidence Mortality care (Cleary, Gelband, and Wagner 2015, chapter 9 in this volume), including, at a minimum, opioid drugs for severe Source: Based on IARC GLOBOCAN data (Ferlay and others 2015). pain control. Treating early stage breast and cervical cancer includes quality surgery, which could also be available for many other early-stage resectable cancers. The package ESSENTIAL PACKAGES OF INTERVENTIONS is organized according to delivery platforms, classified as The DCP3 essential package of interventions for cancer national level policy, regulation, or community informa- is intended to be considered and modified as appropriate tion; primary health clinic or mobile outreach; first-level by governments. The specific interventions and the crite- hospital; or specialized cancer center. ria used to choose them (effectiveness, cost-effectiveness, The cost of the essential package is estimated for the feasibility, and affordability) are intended to help LMIC entire population, not restricted to age under 70 years. governments decide what to support and what not to We estimated the global and per capita costs of each support (Jamison 2015). intervention in the package separately for low-income, Table 1.2 Changes in Deaths from All Causes and Cancer, by Country Income Group, Ages 0–69, 2000–10 (Percent change in mortality rate) Change in % 2000–10 by World Bank country income group Lower- middle- Upper-middle- Cause of death Low-income income income High-income World All cancers –6 –2 –12 –13 –10 Lung, mouth, esophagus (mainly –6 +1 –11 –12 –9 tobacco-related) Cervix, liver, stomach (mainly –13 –2 –18 –24 –15 infection-related) All other cancers –4 –3 –9 –12 –8 All causes –21 –15 –23 –17 –19 Sources: Based on data from IARC GLOBOCAN (Ferlay and others 2015) and WHO Global Health Estimates (WHO 2012). 8 Cancer Table 1.3 Projected Deaths from All Causes and Cancer at Ages 0–69 Years, 2030 (at 2010 death rates, thousands) Lower- Low- middle- Upper-middle- High- Population/Cause of death income income income income World All Causes 8,620 18,110 11,600 2,960 41,290 Cancer 590 1,690 2,690 1,130 6,100 Lung, mouth, esophagus 130 390 820 350 1,690 Cervix, liver, and stomach 80 250 700 120 1,150 All other cancers 380 1,050 1,170 660 3,260 Sources: Based on data from UNPD 2012 and WHO Global Health Estimates (WHO 2012). Note: All deaths are rounded to nearest 10,000. All cancer deaths (in thousands) at ages 70+ would be 240, 800, 3,110, 2,450, and 6,600 in low-income countries, lower-middle-income countries, upper-middle-income countries, high-income countries, and worldwide, respectively. Table 1.4 Essential Cancer Intervention Packagea Cancer type/ Platform for intervention delivery Number of deaths, ages 0–69 years, 2012 Nationwide policies, regulation, Primary health clinic or First-level Specialized cancer (thousands) or community information mobile outreach hospitalb center/unitc All cancers Education on tobacco hazards, value of HPV and HBV vaccination, and importance of seeking early treatment for 3,230 common cancers Palliative care, including, at a minimum, opioids for pain reliefd Selected tobacco- Taxation; warning labels or plain Cessation advice and related cancers (oral, packaging; bans on public smoking, services, mostly without lung, and esophagus) advertising, and promotion; and pharmacological therapies 900 monitoring Breast cancer Treat early-stage cancer 280 with curative intente Cervical cancer School-based HPV vaccination Opportunisticf screening Treat pre- Treat early-stage cancer 180 (visual inspection or HPV DNA cancerous testing); treat precancerous lesions lesions Colorectal cancer Emergency Treat early-stage cancer 210 surgery for with curative intent obstruction Liver cancer Hepatitis B vaccination 380 (including birth dose) Childhood cancers Treat selected early-stage 80 g cancer with curative intent in pediatric cancer units/hospitals Note: Cancer totals are rounded to nearest 10,000. Education and basic palliative care are relevant for cancers at all ages. HBV = hepatitis B virus; HPV = human papillomavirus. a. Red type denotes emergency care. b. First-level hospitals are referred to as district hospitals in some countries. c. Some interventions may take place at first-level hospitals, by a specialized surgeon visiting once per month, for example. d. Palliative care should be available at all levels specified in the table and in the home. e. Early-stage cancer generally refers to stages I and II. f. Screening is opportunistic when a test is requested by a patient or offered by a practitioner to a patient attending for another reason. Organized screening is a well-defined process including formal invitations to participate, recalls, reminders, tracking results, ensuring follow-up, monitoring, and reporting program performance results. g. Including some solid tumors. Summary 9 lower-middle-income, and upper-middle-income tax increases in the 1990s to triple the price of cigarettes; countries. Most LMICs should be able to implement a by 2005, consumption had halved, but government reve- locally customized essential package that covers most of nues from tobacco had doubled (Van Walbeek 2005). In their population by 2030, given anticipated increases in France, lung cancer mortality among young adults fell public spending on health (Jamison and others 2013). shortly after the tax was raised. Brazil has also reduced The schedule of implementation will vary, however, smoking prevalence considerably (Monteiro and others as some interventions—in particular, higher tobacco 2007). Despite severe industry opposition, Mexico, and taxes and widespread pain palliation—can begin rap- very recently, India and the Philippines, have levied nota- idly in many countries (Foley and others 2006; Sloan ble increases in cigarette taxes, and in Mexico cigarette and Gelband 2007). By contrast, affordable availability sales have already started to decline (Jha and others 2015, of treatments that require considerable infrastructure chapter 10 in this volume; WHO 2013). The WHO’s development may take many years to achieve fully after Framework Convention on Tobacco Control, adopted a start is made. by more than 180 countries, is an important enabler of country action on tobacco (Jha 2015). Prevention Most countries (183 worldwide) now vaccinate infants Screening against hepatitis B, with global coverage estimated at The emphasis on diagnosing and treating cancers while 81 percent in 2013. This will prevent many liver can- they are still at an early stage might suggest the appro- cers some decades hence, but a birth dose, particularly priateness of many cancer screening programs (Sullivan, important in countries with high mother-to-child trans- Sullivan, and Ginsburg 2015, chapter 12 in this vol- mission, reached only 26 percent of newborns in 2011 ume), but population screening is expensive (even if (WHO 2011c). cost-effective, at least in some populations in HICs) Seventy-five countries (including HICs) have begun and requires considerable infrastructure. Only oppor- national HPV vaccination programs and others are tunistic cervical screening (with or without some added developing experience with the vaccine (Gavi 2013). outreach) meets the DCP3 criteria and is a suggested In addition, Gavi, the Vaccine Alliance, is supporting component of an essential package. Screening using pilot programs in several LICs in Sub-Saharan Africa. visual inspection with acetic acid (which makes abnor- The delivery cost of reaching adolescent girls with three mal tissue appear white) can detect precancerous lesions doses is the major barrier, as Gavi-subsidized vaccine that can be treated inexpensively (often during the same costs only US$0.20 to US$0.40 per dose, while pro- visit) to prevent cervical cancer from developing (Denny gram costs range between US$4 and US$13 per fully and others 2015; Goss and others 2013). When conve- immunized girl (Denny and others 2015; Gavi 2013). nient, rapid diagnostic tests for the main carcinogenic Hepatitis B and HPV vaccinations will have their main types of HPV infection become affordable and available effect on mortality during the second half of the cen- for use by fieldworkers, they could make such screening tury, when the cohorts of immunized children reach much more effective and reliable (Sankaranarayanan middle age. and others 2009). Two or three such screenings per Tobacco control, notably much higher tobacco excise lifetime, starting around age 35 years, at intervals of five taxes (which result in marked increases in adult cessa- to10 years, should reduce lifetime cervical cancer risk by tion), can have a more immediate effect: people who quit more than half (Goldie and others 2005). smoking before age 40 years avoid more than 90 percent The essential package does not include any type of the risk they would have incurred had they continued of screening for prostate or breast cancer. Both have to smoke (Jha and Peto 2014). This means a substantial attracted significant controversy in HICs, although saving of lives starting within 5–10 years after measures for different reasons. The most widespread means of are put in place. Higher cigarette taxes also discourage prostate cancer screening is through a blood test for youth initiation, which will prevent many deaths in the prostate-specific antigen (PSA, a protein produced at second half of the century. However, cessation remains elevated levels by cancerous prostate cells), with or uncommon in most LMICs, with adults quitting often without digital rectal examination. Although it is a as a result of cancer and other diseases, and not to avoid simple test, PSA is not supported by national programs them. Only 28 countries are undertaking comprehensive because it leads to overdiagnosis and overtreatment, tobacco control programs that include high taxes as a with many more men harmed by the side effects of major strategy (WHO 2013). There are already some overtreatment than are saved from prostate cancer. notable successes: France and South Africa used large The U.S. Preventive Services Task Force discourages 10 Cancer PSA testing (U.S. Preventive Services Task Force 2012). outcomes have been achieved in specialized childhood By contrast, screening mammography for breast cancer cancer centers and through national referral and man- is supported by most HICs as an expensive but moder- agement plans, particularly for acute lymphoblastic ately effective measure, although the optimal age range leukemia, Burkitt lymphoma, and Wilms tumor (Gupta for screening and screening frequency are still debated. and others 2014). Clinical breast examination might be a viable option in LMICs, but the effectiveness of this requires more research (Anderson and others 2015). Other common Palliative Care cancers with detectable precancerous stages are colorec- Many incurable cancers cause intractable pain. Opioid tal cancer (precancerous polyps) (Rabeneck and others medications can generally relieve this pain, greatly 2015, chapter 6 in this volume) and oral cancer (visible improving the quality of the last few weeks or months lesions) (Sankaranarayanan and others 2015, chapter 5 of life for patients and families. The simplest and least in this volume). Eventually, screening for more cancers expensive preparation, oral morphine, works for an may be added, but it is likely to be appropriate after estimated 90 percent of patients with severe terminal effective treatment is established. cancer pain (Foley and others 2006). It is also used by patients with HIV/AIDS and some other chronic condi- tions. Palliative care is widely available only in HICs, but Diagnosis and Treatment it could be made available in LMICs quite rapidly, even Accurate diagnosis is needed for cancer treatment, but before other types of treatment. Palliative care includes shortages of trained pathologists and other laboratory more than pain control and is relevant throughout the technologists and lack of facilities and supplies critically course of illness, but pain control is the core and the limit diagnostic capacity in many LMICs (Gospodarowicz greatest need is at the end of life. and others 2015, chapter 11 in this volume). In addition With appropriate organization and cooperation from to an initial diagnosis of cancer (often based on biopsy the government and the health care sector, opioids can specimens) that can help in assessing the need for major be provided even in rural areas, at home, at low cost. The surgery, diagnostic services can help determine treatment current reality is, however, that few people have access strategies after surgery. The status of tumors, nodes, and to effective pain medicines because of unnecessary and metastases has long been clinically useful, and other tests ill-conceived restrictions at the country level. In 2006 on the tumor itself can determine post-surgical man- (with only marginal progress since then), 66 percent of agement. In particular, breast cancer surgical specimens the world’s population lived in countries that had virtu- should undergo reliable testing to see if they carry the ally no consumption of opioids, 10 percent in countries estrogen receptor protein; if they do (that is, if the tumor with very low consumption, 3 percent in countries with is ER+), endocrine treatment will substantially reduce low consumption, and 4 percent in countries with mod- the risk of recurrence and death (box 1.4). erate consumption (Seya and others 2011). Treatment for early breast cancer and cervical cancer includes some or all of the following: surgery, radiother- apy, chemotherapy, and targeted (for example, endo- Local Priority Conditions crine) therapy, that is, all the basic components of cancer The essential package can be customized and aug- care (Anderson and others 2011; Knaul and others 2011). mented with locally appropriate and feasible interven- For early cervical cancer, surgery is the primary treat- tions. Examples include improved storage of grain and ment and radiotherapy is an adjunct. For whatever other foods to avoid fungal contamination that contrib- is considered complete treatment in a given country utes to high liver cancer rates in parts of Africa and Asia context, all components of care should be accessible by (Gelband and others 2015, chapter 8 in this volume; patients once treatment is started. Partial or incomplete Groopman, Kensler, and Wild 2008); opportunistic treatment can cause side effects, but with less chance of screening (especially of high-risk tobacco users) and clinical benefit. treatment for precancerous lesions and early-stage oral Childhood cancer is rare (accounting for 1 percent of cancer in India and other countries with high oral cancer cancer deaths in HICs), representing by far the smallest burdens (Dikshit and others 2012; Sankaranarayanan burden of the cancers targeted by the essential package. and others 2015); screening and treatment for colorectal Although they cannot be prevented, many common cancer in Argentina and Uruguay (Goss and others 2013; childhood cancers have high cure rates in HICs, making Rabeneck and others 2015); and elimination of liver them feasible targets (Gupta and others 2015). Cure flukes (with the drug praziquantel) to prevent bile duct rates in most LMICs are far lower, but reasonably good cancer in the limited areas where flukes are common, Summary 11 Box 1.4 Possible Strategies for Treating Early Breast Cancer in LMICs By definition, in early breast cancer (stage I or II), drugs, such as tamoxifen or, for post-menopausal all detectable disease can be removed surgically, but women, an aromatase inhibitor (AI) (Early Breast micrometastases may remain that, perhaps years later, Cancer Trialists’ Collaborative Group and others cause recurrence and death. Adjuvant treatments 2015), can be dispensed safely to outpatients and are may be given after surgery to reduce this risk. In high- available as relatively low-cost generics (although income countries, most women receiving appropri- even generic tamoxifen costs about US$15 per ate treatment for early breast cancer survive their year in India, and generic AIs currently cost about disease (Early Breast Cancer Trialists’ Collaborative US$50 per year). Chemotherapy also reduces recur- Group and others 2012). The success rate of breast- rence but is more toxic and requires more careful conserving surgery (lumpectomy) plus radiotherapy medical supervision to ensure safety and efficacy. to the conserved breast is about the same as for mas- New drugs, for example, trastuzumab, that target tectomy (removal of the entire breast, and perhaps other breast cancer receptors are not at present some local lymph nodes) and either can be offered, cost-effective in LMICs. if safe radiotherapy is available. The most basic Relatively simple regimens of generic cytotoxic surgical procedure for stage II breast cancer is some drugs (for example, four cycles of daunorubicin and form of mastectomy (Anderson and others 2015). cyclophosphamide with drug costs of about $200 In low- and middle-income countries (LMICs), for in India) should be practicable wherever surgery is women with early breast cancer, the first requirement practicable (Anderson and others 2015), and could is good quality, safe surgery. In low-income countries be offered to women who are otherwise in good (LICs), in particular, timely access to safe surgery is health but whose disease has already spread from the a major barrier. In middle-income countries (MICs), breast to the local lymph nodes (Early Breast Cancer where there is generally better population access to Trialists’ Collaborative Group and others 2012). surgical services, quality cancer surgery is the major More effective cytotoxic regimens (for example, with surgical concern, particularly adequate resection taxanes) would increase toxicity, drug costs, and of the tumor (Dare and others 2015). After techni- supervision costs. cally successful surgery, treatments can be based on Finally, global initiatives might well help to estrogen-receptor (ER) status, estimated recurrence lower the cost of cancer drugs and other com- risk, and general health (Anderson and others 2015). modities, and develop and disseminate standard- The ER status of surgically removed breast can- ized resource-appropriate treatment protocols, such cers can be determined (for about US$10, in India). as those developed by the Breast Health Global If the cancer is ER-positive, about five years of endo- Initiative. The successful global initiative to aid in crine drug therapy substantially reduces the 15-year the diagnosis and treatment of HIV/AIDS provides recurrence risk and is relatively nontoxic. Endocrine a model (Piot and Quinn 2013). or treatment of schistosomiasis to prevent bladder or power plants and folic acid) and claims of cure, even of intestinal cancer in parts of Asia, the Middle East and advanced cancers and even within the health care sys- North Africa, and Sub-Saharan Africa (IARC 1994). tem itself. A guidepost for the former category of claims Finally, occupational hazards should be monitored and is IARC’s Monographs on the Evaluation of Carcinogenic mitigated where necessary, for example, use of power Risks to Humans, which, since 1971, have evaluated tools on asbestos roofing or insulation, or heavy smoke more than 900 agents (http://monographs.iarc.fr/index. pollution in houses (IARC 2012). php). Treating advanced cancers—although a com- As important as what to include in a national can- mon practice in HICs—is expensive, often painful for cer package is what to exclude. Cancer is notorious for patients, and usually futile. Countries should carefully exaggerated claims of causation (for example, nuclear examine the resource requirements and likely success of 12 Cancer such treatments in deciding not only which cancers to irinotecan (a relatively new treatment for colon can- include in a package, but the appropriate interventions cer) are not currently considered cost-effective in the by stage. For advanced cancers with little possibility of United Kingdom (Tappenden and others 2007). cure, palliative care may be the best alternative. Excise taxes on tobacco (US$1–US$150 per disability-adjusted life year [DALY] averted) and hepa- titis B vaccination (less than US$100/DALY averted) are COST-EFFECTIVENESS OF INTERVENTIONS very cost-effective in all LMICs. Opportunistic cervical For most types of cancer, the cost-effectiveness liter- cancer screening and treatment of precancerous lesions ature for LMICs is slim (Horton and Gauvreau 2015, are likely also to be very cost-effective in all LMICs. The chapter 16 in this volume): nine studies were identified cost-effectiveness of HPV vaccination depends cru- for breast cancer, two (plus four from high-income cially on vaccine cost, but at US$15/per girl vaccinated, Asian countries) for colorectal cancer, one for liver HPV vaccination is likely to be very cost-effective in all cancer prevention, and none for pediatric cancer. LMICs. Some aspects of the treatment for early breast Seventeen studies were sourced from an expert search cancer are cost-effective wherever breast cancer surgery for cervical cancer, and a recent systematic review of has been performed (mainly MICs; less than US$150/ vaccines (Ozawa and others 2012) identified three DALY averted) (figure 1.2). studies for hepatitis B vaccination. A useful benchmark was to exclude from the essential package those inter- COSTS OF PACKAGES ventions that are not clearly cost-effective in HICs. Most new drug treatments for advanced cancer fall To provide per capita cost estimates for an essential pack- in this category, such as bevacizumab (a monoclonal age, we used available information on cost combined antibody) for metastatic breast cancer, which, at cur- with demographic information from three large, diverse rent prices, does not meet cost-effectiveness criteria countries (Brazil, India, and Nigeria) (expressed in 2012 in the United Kingdom (Rodgers and others 2011) U.S. dollars). Although Nigeria is a lower-middle-income and other HICs (Dedes and others 2009; Montero country, we used its demographic structure and lack and others 2012). Similarly, cetuximab (a monoclonal of existing facilities and human resources to represent antibody for metastatic colon and lung cancers) plus the scenario of LICs, mainly in Sub-Saharan Africa. Figure 1.2 Cost-Effectiveness of Selected Interventions Treat breast cancer LIC (1) Screen and treat breast cancer LIC (1) Screen and treat breast cancer MIC (2) a HPV vaccination @ $240+/girl (3) Treat colorectal cancer LIC (4) Non price interventions for tobacco (5) Treat breast cancer MIC (2) a HPV vaccination @$50/girl MIC (6) Hepatitis B vaccination LIC (7) 33% Price increase via tobacco tax (5) 1 10 100 1,000 10,000 Cost per DALY saved US $2012 Range Sources: (1) Zelle and others 2012; (2) Salomon and others 2012; (3) Insinga and others 2007; Praditsitthikorn and others 2011; Termrungruanglert and others 2012; (4) Ginsberg and others 2010; (5) Jha and others 2006; (6) Kawai and others 2012; Vanni and others 2012; (7) Prakash 2003; Griffiths, Hutton, and Das Dores Pascal 2005; Kim, Salomon, and Goldie 2007. Note: Studies used for calculations were from a systematic search, whose findings are available online (annex 16A). Cost-effectiveness has not been calculated for elements of the essential package for which relevant data were entirely lacking. DALY = disability-adjusted life year; HPV = human papillomavirus; LIC = low-income country; MIC = middle-income country; QALY = quality-adjusted life year. a. Based on a study reporting QALYs, not DALYs (the difference is small when interventions primarily reduce mortality). Summary 13 To account for training, pathology services, and other per capita in upper-middle-income countries, and system costs, we used a multiplier, equal to 50 percent of US$1.70 per capita in both lower-middle-income the intervention-based costs, drawn from similar costing countries and LICs (table 1.5). The annual cost of the studies for nutrition (Bhutta and others 2013) and health essential package of cancer services (table 1.6) in 2013 systems (Rao Seshadri and others 2015). However, we did dollars would be about US$13.8 billion, US$4.4 bil- not include the one-time investment costs for building lion, and US$1.4 billion in those groups of countries, hospitals, clinics, and other infrastructure that would respectively. There are obvious caveats on the preci- eventually be needed to support cancer and other clinical sion of the costs, including uncertainties of these costs services (Gospodarowicz and others 2015; Sloan and by 2030. Importantly, drug costs can fall substantially Gelband 2007). as drugs go off patent, and global initiatives could fur- The DCP3 essential package of cancer control inter- ther reduce the prices of key generic drugs and other ventions would cost roughly an additional US$5.70 commodities. Table 1.5 Approximate Per Capita Marginal Costs of the Essential Package for Low-Income, Lower-Middle-Income, and Upper-Middle-Income Countries (2012 U.S. dollars) Low- Lower-middle- Upper-middle- Intervention income Income income Comprehensive tobacco control measures 0.05 0.07 1.06 Palliative care and pain control 0.05 0.06 0.06 HBV vaccination 0.08 0.04 0.04 Promote early diagnosis and treat early-stage breast cancer 0.43 0.43 1.29 HPV vaccination 0.23 0.23 0.40 Screen and treat precancerous lesions and early-stage cervical cancer 0.26 0.29 0.87 Treat selected childhood cancers 0.03 0.03 0.09 Subtotal 1.13 1.15 3.81 Ancillary services (50% of subtotal) 0.57 0.58 1.91 TOTAL COSTS 1.70 1.73 5.72 Source: Based on online annex 1A and Horton and Gauvreau 2015, annex 16A. Note: HPV = human papillomavirus; HBV = hepatitis B virus. Table 1.6 Resource Requirements for the Essential Cancer Intervention Package for LMICs Expenditures Low-income Lower-middle income Upper-middle income Total LMICs Public spending on health as % GDP, 2.0 1.8 3.1 3.0 2013 Total public spending on health in 11 89 534 634 2013 (US$ billions) Required amount for cancer in 2013 1.4 4.4 13.8 19.6 (US$ billions) Cancer package as % of total public 13.0 4.9 2.6 3.1 spending on health in 2013a Note: GDP = gross domestic product; LMICs = low- and middle-income countries. a. Based on spending data from World Bank 2014b. 14 Cancer AFFORDABILITY AND DOMESTIC FINANCING most important single cancer prevention intervention at a practical level, and a tripling of the excise tax on OF ESSENTIAL CANCER SERVICES tobacco (thereby approximately doubling prices) could The total estimated annual cost of the essential pack- mobilize an additional US$100 billion worldwide in age of cancer interventions for all LMICs is about annual revenue (Jha and Peto 2014). For all LMICs, the $20 billion dollars (2013 U.S. dollars). A useful metric epidemiologic dividend that accrues from a decreased is the cost of the package as a proportion of current burden of infectious disease should generate savings that total public spending on health. This is 2.6 percent in can be spent on NCD control (Jamison and others 2011). UMICs, 5 percent in lower-middle-income countries, and 13 percent in LICs. By comparison, HICs devote 3–7 percent of their total health spending to cancer IMPLEMENTATION CHALLENGES FOR THE control (OECD 2013). Most LMICs allocate far less; cancer currently accounts for about 1 percent of health ESSENTIAL PACKAGE spending (public and private) in Brazil and India, and Within the essential package, some aspects of interven- 2 percent in China and Mexico (Goss and others 2013; tions may be implemented reasonably quickly, notably IARC 2014; Knaul and others 2011). tobacco control measures that involve taxation and Financing for cancer control will have to come regulation (Jha and others 2015) and policy changes to mainly from national health care budgets, particularly increase access to opioids (although establishing nation- in MICs, where rising incomes are enabling expansion wide programs and training a full cadre of providers may of public financing for health (Jamison and others 2013; take years) (Cleary, Gelband, and Wagner 2015). Some Knaul and others 2015, chapter 17 in this volume). interventions can be scaled to reasonably high coverage South Africa, for example, has assessed which interven- quickly with existing infrastructure, such as school-based tions it might include in an expanded national health HPV vaccination for adolescent girls, or hepatitis B vac- insurance package (Shisana and others 2006) and sim- cination for newborns. By contrast, other interventions ilar work is underway in India (Jha and Laxminarayan will need expanded clinical access, most notably surgical 2009; Rao Seshadri and others 2015). In LICs, it would treatment of early-stage breast cancer and cervical can- be inappropriate for governments to shift to spend- cer (Dare and others 2015, chapter 13 in this volume). ing 13 percent of their health care dollars on cancer. Increasing surgical capacity is relatively expensive but External assistance will be needed in those countries to feasible from an organizational perspective, especially establish an expansion path for cancer control. if existing district hospitals can be strengthened (Mock A clear principle to adopt is the eventual goal of and others 2015) (paired with central cancer clinical coverage for every person—even if coverage expands expertise). When high quality surgery becomes avail- gradually—but not coverage of everything (WHO able, early-stage, resectable tumors of various types, in 2000), since poorly conceived plans may provide cover- addition to breast and cervical lesions, can be removed. age of more expensive treatments for a few, while miss- Expanding chemotherapy treatment requires an exten- ing the opportunity to expand cost-effective population sive network of laboratories and follow-up, which in coverage. Public finance is not necessarily synonymous LICs and lower-middle-income countries is currently with public delivery (Musgrove 1996). Properly regu- feasible only in urban areas. Scaling up radiotherapy lated private hospitals, facilities, and providers can be requires large capital expenditures, and substantial atten- contracted to deliver cancer control interventions (Jha tion to clinical guidelines and treatment protocols as well and Laxminarayan 2009). as safety assurances (Jaffray and Gospodarowicz 2015, Several countries in Latin America are already chapter 14 in this volume). expanding their health insurance systems from coverage Particularly for LICs where minimal cancer services limited to occupational groups or selected vulnerable exist in the public sector, the needed expertise and groups, to more comprehensive coverage (Goss and resources for cancer treatment will require years of others 2013). However, for some lower-middle-income steady investment to expand physical and human infra- countries and most LICs, substantial increases in pub- structure. Elements that are missing or in short supply lic finance for health, paired with economic growth in LMICs (Bray and others 2014; Dikshit and others or external assistance, would be needed to adopt a full 2012; Gospodarowicz and others 2015) include trained package of interventions (Jamison and others 2013). professionals in oncology and relevant disciplines; Even those countries could benefit from considering appropriately-equipped facilities, including radiotherapy the future cancer burden, costs, and financing to project facilities, pathology services, and other laboratory testing a future cancer control plan. Higher tobacco taxes are the services (for example, estrogen-receptor testing for breast Summary 15 cancer tissue; box 1.3); supplies, including chemother- and others 2015). Upgrading hospitals to provide apy drugs; geographic access to facilities with affordable basic cancer surgical services (Mock and others 2015), cancer services, including surgery; public awareness of developing cancer referral networks, tracking service the availability and effectiveness of cancer control inter- performance, integrating the delivery of different types ventions; and cancer incidence and cause-of-death data. of services, and ensuring that financial flows accompany As more people are successfully treated and live for many services are also required. years, survivorship services (for example, rehabilitation, remedies for physical deficits caused by treatment, limit- ing the social stigma associated with having had cancer, GLOBAL INITIATIVES FOR CANCER CONTROL and follow-up for recurrence) will increase in impor- tance (Hewitt, Greenfield, and Stovall 2005), but costs for Only 1 percent of the US$30 billion development assis- survivorship are not included in our estimates. tance for health in 2010 was allocated to NCDs, only a The package emphasizes treatment for early-stage portion of which was for cancer (IHME 2012). NCD cervical and breast cancers (and similarly, early stages funding is likely to increase somewhat with increas- for other cancers included in specific country plans), ing global recognition of the importance of NCDs. because cure rates are substantially higher than they are However, it is unlikely that significant global funds will for more advanced cancers. Surgery is particularly be allocated to support national health systems to deal important, as surgery alone is curative for many early with cancer. As additional funding becomes available, cancers. Although locally appropriate cervix screening, we suggest three priorities for international support. which will identify many precancers and early cancers, is included in the package, there is no corresponding 1. Lower the costs to countries of key inputs for the screening intervention for breast cancer. Even without essential package and other cost-effective interven- screening, however, LMICs might be able to achieve tions, such as HPV and other vaccines, cancer drugs a somewhat earlier stage of presentation of common (including generics), screening tests (for example, cancers by making affordable treatment available and HPV tests), laboratory reagents and other test com- communicating this to people. Historical evidence from modities, surgery, radiotherapy machines, and other HICs, illustrated by stage-shifting for cervix cancer in relevant goods. The Global Fund for AIDS, TB and Sweden before organized screening began around 1960, Malaria; Gavi; the Clinton Health Access Initiative; supports this approach (Pontén and others 1995). and other international partnerships have devel- Cancer treatment can be organized through exist- oped mechanisms to reduce the price of a range of ing medical facilities (particularly district hospitals) or global commodities relevant to infectious disease through specialized centers, but the key is to ensure good control, utilizing economies of scale (relevant for links between facilities (Sloan and Gelband 2007), which purchases of drugs or radiotherapy machines), requires a centralized locus of control and the ability to subsidies for reputable and affordable medicines, adjust elements of the system that are not working to the advanced market commitments, and similar inno- benefit of patients (Gospodarowicz and others 2015). An vations (Piot and Quinn 2013). Similar efforts for example from childhood cancer illustrates this well. All cancer are possible, for example, as has been pro- children with cancer in Honduras (population 8 million) posed for radiotherapy by the Global Task Force are treated in two centers that collaborate and communi- on Radiotherapy for Cancer Control (Union for cate closely (Metzger and others 2003). In contrast, chil- International Cancer Control [UICC]). dren with cancer in Colombia (population 48 million) are 2. Expand technical assistance in cancer control. treated in more than 150 health care institutions of vary- International and regional networks exist for many ing size, with little to no communication between centers aspects of cancer care, such as treatment guide- (Gupta and others 2015). This adversely affects patient lines; networks on breast, cervical, and colorec- outcomes and costs. India, population 1.3 billion, faces tal cancer screening; childhood cancer treatment more challenging coordination of care, but is building a and research; and palliative care. Other support National Cancer Grid (Pramesh, Badwe, and Sinha 2014) modalities, for example, twinning institutions, have linking non-specialist hospitals to specialist cancer centers typically involved institutions in HICs and LICs and providing them with current treatment protocols. (North-South collaborations), but as in other areas, Building and improving a nation’s cancer control the opportunities should grow to add South-South capacity requires attention to the quality of services, collaborations. Within countries, peer-based and from pathology and diagnosis to surgery, chemother- professional standards of cancer care and reporting apy, radiotherapy, and palliative care (Gospodarowicz of outcomes and performance for various facilities 16 Cancer can improve quality of care (Peabody and others NOTES 2006; Varmus and Trimble 2011). Maps and figures in this chapter are based on incidence and 3. Support for research is a worthwhile investment for mortality estimates for ages 0–69, consistent with reporting in overseas developmental assistance. Research priori- all DCP3 volumes. The discussion of burden (including risk ties include tracking national cancer burdens, clinical factors) and interventions, however, includes all ages unless trials, and implementation science, including research otherwise noted. on delivery systems; cancer epidemiology and biol- The World Bank classifies countries according to four ogy; widely practicable, low-cost technologies; and income groupings. Income is measured using gross national economics (Trimble and others 2015, chapter 15 in income per capita, in U.S. dollars, converted from local cur- this volume). rency using the World Bank Atlas method. Classifications as of July 2014 are as follows: • Low-income countries = US$1,045 or less in 2013 BENEFITS OF EXPANDED CANCER CONTROL • Middle-income countries are subdivided: Despite substantial challenges in most LMICs, appre- • Lower-middle-income = US$1,046–US$4,125 ciable reductions in the cancer burden might well be • Upper-middle-income = US$4,126–US$12,745 • High-income countries = US$12,746 or more possible by 2030, with even greater reductions by 2050 and beyond (Norheim and others 2014), particularly through treating common cancers that are detected early, tobacco control that encourages widespread adult DCP3 CANCER AUTHOR GROUP smoking cessation, and vaccination against hepatitis B Issac Adewole, Hemantha Amarasinghe, Benjamin and HPV. O. Anderson, Federico G. Antillon, Samira Asma, Global cancer death rates at ages 0–69 years were Rifat Atun, Rajendra A. Badwe, Freddie Bray, Frank declining at about 10 percent per decade during 2000– J. Chaloupka, Ann Chao, Chien-Jen Chen, Wendong 10. If this were to continue, then from 2010 to 2030 Chen, James Cleary, Anna J. Dare, Anil D’Cruz, cancer death rates will fall by almost 20 percent. A Lynette Denny, Craig Earle, Silvia Franceschi, Cindy L. decrease of one-third in global cancer death rates by Gauvreau, Hellen Gelband, Ophira M. Ginsburg, Mary 2030, as proposed recently (Norheim and others 2014), K. Gospodarowicz, Thomas Gross, Prakash C. Gupta, would require faster progress in LMICs, particularly Sumit Gupta, Andrew Hall, Mhamed Harif, Rolando marked increases in tobacco cessation. The WHO esti- Herrero, Susan Horton, Scott C. Howard, Stephen P. mates that tobacco control, HPV and HBV vaccination, Hunger, Andre Ilbawi, Trijn Israels, David A. Jaffray, and opportunistic cervical cancer screening could avoid Dean T. Jamison, Prabhat Jha, Newell Johnson, Jamal about 6 percent of cancer deaths by 2030 (or about Khader, Jane J. Kim, Felicia Knaul, Carol Levin, Joseph 200,000 deaths before age 70 years annually). The DCP3 Lipscomb, W. Thomas London, Mary MacLennan, essential package could achieve greater reductions. If, as Katherine A. McGlynn, Monika L. Metzger, Raul expected, the availability of treatment shifts diagnoses Murillo, Zachary Olson, Sherif Omar, Krishna for common, treatable cancers to earlier stages, further Palipudi, C. S. Pramesh, You-Lin Qiao, Linda Rabeneck, lives could be saved. The benefits of pain relief are not Preetha Rajaraman, Kunnambath Ramadas, Chintanie measured in lives saved, but are important. Ramasundarahettige, Timothy Rebbeck, Carlos Finally, cancer control contributes to reduced inequal- Rodriguez-Galindo, Rengaswamy Sankaranarayanan, ity in health, providing relatively larger benefits to the Monisha Sharma, Ju-Fang Shi, Isabelle Soerjomataram, poor. In China, increased tobacco taxes and access to cer- Lisa Stevens, Sujha Subramanian, Richard Sullivan, vical cancer prevention, such as screening and HPV vac- Terrence Sullivan, David Thomas, Edward L. Trimble, cination, would disproportionately benefit those in the Joann Trypuc, Stéphane Verguet, Judith Wagner, Shao- lowest income quintile by reducing deaths and through Ming Wang, Christopher P. Wild, Pooja Yerramilli, better financial risk protection from catastrophic health Cheng-Har Yip, Ayda Yurekli, Witold Zaton ´ski, Ann G. expenditures (Levin and others 2015, chapter 18 in this Zauber, and Fang-Hui Zhao. volume; Verguet and others 2015). Cancer control is often approached with pessimism, but practicable, deliberate, cost-effective steps can enable REFERENCES many countries to reduce substantially by 2030 the Allemani, C., H. K. Weir, H. Carreira, R. Harewood, suffering and premature death from cancer, with much D. Spika, and others. 2015. “Global Surveillance of Cancer greater improvements by 2050. Survival 1995–2009: Analysis of Individual Data for Summary 17 25,676,887 Patients from 279 Population-Based Registries India: A Nationally Representative Survey.” The Lancet 379 in 67 Countries (CONCORD-2).” The Lancet 385 (9972): (9828): 1807–16. doi:10.1016/S0140-6736(12)60358-4. 977–1010. doi:10.1016/S0140-6736(14)62038-9. Early Breast Cancer Trialists’ Collaborative Group, Writing Anderson, B. O., E. Cazap, N. S. El Saghir, C. H. Yip, Committee, R. Peto, C. Davies, J. Godwin, R. Gray, H. M. Khaled, and others. 2011. “Optimisation of Breast and others. 2012. “Comparisons between Different Cancer Management in Low-Resource and Middle-Resource Polychemotherapy Regimens for Early Breast Cancer: Countries: Executive Summary of the Breast Health Global Meta-Analyses of Long-Term Outcome among 100,000 Initiative Consensus, 2010.” The Lancet Oncology (4): Women in 123 Randomised Trials.” The Lancet 379 (9814): 387–98. doi:10.1016/S1470-2045(11)70031-6. 432–44. doi:10.1016/S0140-6736(11)61625-5. Anderson, B. O., J. Lipscomb, R. H. Murillo, and D. R. Thomas. ———. Forthcoming. “Aromatase Inhibitors vs. Tamoxifen 2015. “Breast Cancer.” In Disease Control Priorities: Volume in Early Breast Cancer: Patient-Level Meta-Analysis of the 3, Cancer, edited by H. Gelband, P. Jha, R. Sankaranarayanan, Randomised Trials.” The Lancet. and S. Horton. 3rd ed. Washington, DC: World Bank. Ferlay, J., I. Soerjomataram, R. Dikshit, S. Eser, C. Mathers, and Bhutta, Z. A., J. K. Das, A. Rizvi, M. F. Gaffey, N. Walker, others. 2015. “Cancer Incidence and Mortality Worldwide: and others. 2013. “Evidence-Based Interventions for Sources, Methods and Major Patterns in GLOBOCAN Improvement of Maternal and Child Nutrition: What 2012.” International Journal of Cancer 136 (5): E359–86. Can Be Done and At What Cost?” The Lancet 382 (9890): doi:10.1002/Ijc.29210. 452–77. doi:10.1016/S0140-6736(13)60996-4. Foley, K. M., J. L. Wagner, D. E. Joranson, and H. Gelband. Bray, F., and I. Soerjomataram. 2015. “The Changing Global 2006. “Pain Control for People with Cancer and AIDS.” In Burden of Cancer.” In Disease Control Priorities: Volume 3, Disease Control Priorities in Developing Countries, edited Cancer, edited by H. Gelband, P. Jha, R. Sankaranarayanan, by D. T. Jamison, J. Breman, A. R. Measham, G. Alleyne, and S. Horton. 3rd ed. Washington, DC: World Bank. M. Claeson, and others. 2nd ed. New York: Oxford Bray, F., A. Znaor, P. Cueva, A. Korir, R. Swaminathan, and University Press and World Bank. others. 2014. “Planning and Developing Population-Based Gavi. 2013. “Millions of Girls in Developing Countries To Be Cancer Registration in Low- and Middle-Income Settings.” Protected against Cervical Cancer Thanks to New HPV Technical Publications, International Agency for Research Vaccine Deals.” Gavi, the Vaccine Alliance. on Cancer. Gelband, H., C.-J. Chen, W. Chen, S. Franceschi, A. Hall, Brown, M. L., S. J. Goldie, G. Draisma, J. Harford, and and others. 2015. “Liver Cancer.” In Disease Control J. Lipscomb. 2006. “Health Service Interventions for Cancer Priorities: Volume 3, Cancer, edited by H. Gelband, P. Jha, Control in Developing Countries.” In Disease Control R. Sankaranarayanan, and S. Horton. 3rd ed. Washington, Priorities in Developing Countries, edited by D. T. Jamison, DC: World Bank. J. Breman, A. R. Measham, G. Alleyne, M. Claeson, and Ginsberg, G. M., S. S. Lim, J. A. Lauer, B. P. Johns, and others. 2nd ed. New York: Oxford University Press and C. R. Sepulveda. 2010. “Prevention, Screening and World Bank. Treatment of Colorectal Cancer: A Global and Regional Cleary, J., H. Gelband, and J. Wagner. 2015. “Cancer Pain Generalized Cost Effectiveness Analysis.” Cost Effectiveness Relief.” In Disease Control Priorities: Volume 3, Cancer, and Resource Allocation 8: 2. doi:10.1186/1478-7547-8-2. edited by H. Gelband, P. Jha, R. Sankaranarayanan, and Goldie, S. J., L. Gaffikin, J. D. Goldhaber-Fiebert, A. Gordillo- S. Horton. 3rd ed. Washington, DC: World Bank. Tobar, C. Levin, and others. 2005. “Cost-Effectiveness of Commission on Macroeconomics and Health. 2001. Cervical-Cancer Screening in Five Developing Countries.” Macroeconomics and Health: Investing in Health for Economic New England Journal of Medicine 353 (20): 2158–68. Development. Geneva: World Health Organization. doi:10.1056/Nejmsa044278. Dare, A. J., B. O. Anderson, R. Sullivan, C. S. Pramesh, Gospodarowicz, M. K., J. Trypuc, A. D’Cruz, J. Khader, S. Omar, C.-H. Yip, and others. 2015. “Surgical Services for Cancer and others. 2015. “Cancer Services and the Comprehensive Care.” In Disease Control Priorities: Volume 3, Cancer, Cancer Care Center.” In Disease Control Priorities: Volume 3, edited by H. Gelband, P. Jha, R. Sankaranarayanan, and Cancer, edited by H. Gelband, P. Jha, R. Sankaranarayanan, S. Horton. 3rd ed. Washington, DC: World Bank. and S. Horton. 3rd ed. Washington, DC: World Bank. Dedes, K. J., K. Matter-Walstra, M. Schwenkglenks, Goss, Paul E., Brittany L. Lee, T. Badovinac-Crnjevic, B. C. Pestalozzi, D. Fink, and others. 2009. “Bevacizumab K. Strasser-Weippl, Y. Chavarri-Guerra, and others. 2013. in Combination with Paclitaxel for HER-2 Negative “Planning Cancer Control in Latin America and the Metastatic Breast Cancer: An Economic Evaluation.” Caribbean.” The Lancet Oncology 14 (5): 391–436. European Journal of Cancer 45 (8): 1397–406. doi:10.1016/J. Greenberg, D., C. Earle, C. H. Fang, A. Eldar-Lissai, and Ejca.2008.12.016. P. J. Neumann. 2010. “When Is Cancer Care Cost-Effective? Denny, L., R. Herrero, C. Levin, and J. J. Kim. 2015. “Cervical A Systematic Overview of Cost-Utility Analyses in Cancer.” In Disease Control Priorities: Volume 3, Cancer, Oncology.” Journal of the National Cancer Institute 102 (2): edited by H. Gelband, P. Jha, R. Sankaranarayanan, and 82–8. doi:10.1093/Jnci/Djp472. S. Horton. 3rd ed. Washington, DC: World Bank. Griffiths, U. K., G. Hutton, and E. Das Dores Pascoal. 2005. Dikshit, R., P. C. Gupta, C. Ramasundarahettige, V. Gajalakshmi, “The Cost-Effectiveness of Introducing Hepatitis B Vaccine L. Aleksandrowicz, and others. 2012. “Cancer Mortality in into Infant Immunization Services in Mozambique.” 18 Cancer Health Policy Planning 20 (1): 50–59. doi:10.1093/Heapol Cancer, edited by H. Gelband, P. Jha, R. Sankaranarayanan, /Czi006. and S. Horton. 3rd ed. Washington, DC: World Bank. Groopman, J. D., T. W. Kensler, and C. P. Wild. 2008. Jamison, D. T., J. G. Breman, A. R. Measham, G. Alleyne, “Protective Interventions to Prevent Aflatoxin-Induced M. Claeson, and others. 2006. Disease Control Priorities Carcinogenesis in Developing Countries.” Annual Review in Developing Countries. 2nd ed. Washington, DC: Oxford of Public Health 29: 187–203. http://www.ncbi.nlm.nih University Press and World Bank. .gov/pubmed/17914931. Jamison, D. T. 2015. “Disease Control Priorities, 3rd edition: Gupta, S., S. C. Howard, S. P. Hunger, F. G. Antillon, Improving Health and Reducing Poverty.” The Lancet M. L. Metzger, and others. 2015. “Childhood Cancers.” Feb 4. doi: 10.1016/S0140-6736(15)60097-6. doi: 10.1016/ In Disease Control Priorities: Volume 3, Cancer, edited by s0140-6736(15)60097-6 H. Gelband, P. Jha, R. Sankaranarayanan, and S. Horton. Jamison, D. T., P. Jha, V. Malhotra, and S. Verguet. 2011. 3rd ed. Washington, DC: World Bank. “The 20th Century Transformation of Human Health: Gupta, S., R. Rivera-Luna, R. C. Ribeiro, and S. C. Howard. Its Magnitude and Value.” In How Much Have Global 2014. “Pediatric Oncology as the Next Global Child Problems Cost The World? A Scorecard From 1900–2050, Health Priority: The Need for National Childhood Cancer edited by B. Lomborg. Cambridge, U.K:. Cambridge Strategies in Low- and Middle-Income Countries.” PloS University Press. Med 11 (6): E1001656. doi:10.1371/Journal.Pmed.1001656. Jamison, D. T., W. Mosley, A. R. Measham, and J. Bobadilla, eds. Hamid, S. A., S. M. Ahsan, and A. Begum. 2014. “Disease- 1993. Disease Control Priorities in Developing Countries. 1st Specific Impoverishment Impact of Out-of-Pocket ed. New York: Oxford University Press. Payments for Health Care: Evidence from Rural Bangladesh.” Jamison, D. T., L. H. Summers, G. Alleyne, K. J. Arrow, Applied Health Economics and Health Policy 12 (4): 421–33. S. Berkley, and others. 2013. “Global Health 2035: A World doi:10.1007/S40258-014-0100-2. Converging within a Generation.” The Lancet 382 (9908): Hewitt, M., S. Greenfield, and E. Stovall. 2005. From 1898–955. doi:10.1016/S0140-6736(13)62105-4. Cancer Patient to Cancer Survivor: Lost in Transition. Jha, P. 2009. “Avoidable Global Cancer Deaths and Total Deaths Washington, DC: Institute of Medicine and National from Smoking.” Nature Reviews Cancer 9 (9): 655–64. Research Council. doi:10.1038/Nrc2703. Hoang Lan, N., W. Laohasiriwong, J. F. Stewart, N. D. Tung, ———. 2015. “Deaths and Taxes: Stronger Global Tobacco and P. C. Coyte. 2013. “Cost of Treatment for Breast Control by 2025.” The Lancet 385 (9972): 918–20. Cancer in Central Vietnam.” Global Health Action 6: 18872. doi:10.1016/S0140-6736(15)60464-0. doi:10.3402/Gha.V6i0.18872. Jha, P., F. J. Chaloupka, J. Moore, V. Gajalakshmi, P. C. Gupta, Horton, S., and C. L. Gauvreau. 2015. “Cancer in Low- and and others. 2006. “Tobacco Addiction.” In Disease Control Middle-Income Countries: An Economic Overview.” In Priorities in Developing Countries, edited by D. T. Jamison, Disease Control Priorities: Volume 3, Cancer, edited by J. G. Breman, A. R. Measham, G. Alleyne, M. Claeson, and H. Gelband, P. Jha, R. Sankaranarayanan, and S. Horton. others. 2nd ed. Washington, DC: Oxford University Press 3rd ed. Washington, DC: World Bank. and World Bank. IARC (International Agency for Research on Cancer). 1994. Jha, P., and R. Laxminarayan. 2009. Choosing Health: An “Mongraphs on the Evaluation of Carcinogenic Risks Entitlement for All Indians. Toronto: Centre for Global to Humans: Volume 61, Schistosomes, Liver Flukes and Health Research. Helicobacter Pylori.” Lyon, IARC, France. Jha, P., M. Maclennan, A. Yurekli, C. Ramasundarahettige, ———. 2012. “Mongraphs on the Evaluation of Carcinogenic K. Palipudi, and others. 2015. “Global Hazards of Tobacco, Risks to Humans: Volume 100.” Lyon, IARC, France. Benefits of Cessation and of Taxation of Tobacco.” In ———. 2014. “World Cancer Report 2014.” In World Cancer Disease Control Priorities: Volume 3, Cancer, edited by Report, edited by B. W. Stewart and C. P. Wild. Geneva: IARC. H. Gelband, P. Jha, R. Sankaranarayanan, and S. Horton. IHME (Institute for Health Metrics and Evaluation). 2012. 3rd ed. Washington, DC: World Bank. Financing Global Health 2012: The End of the Golden Age? Jha, P., and R. Peto. 2014. “Global Effects of Smoking, of Seattle: IHME. Quitting, and of Taxing Tobacco.” New England Journal of Ilbawi, A. M., E. M. Einterz, and D. Nkusu. 2013. “Obstacles to Medicine 370 (1): 60–68. doi:10.1056/Nejmra1308383. Surgical Services in a Rural Cameroonian District Hospital.” John, R. M., H. Y. Sung, W. B. Max, and H. Ross. 2011. World Journal of Surgery 37 (6): 1208–15. doi:10.1007 “Counting 15 Million More Poor in India, Thanks to /S00268-013-1977-X. Tobacco.” Tobacco Control 20 (5): 349–52. doi:10.1136 Insinga, R. P., E. J. Dasbach, E. H. Elbasha, A. Puig, and /Tc.2010.040089. L. M. Reynales-Shigematsu. 2007. “Cost-Effectiveness of Kawai, K., G. T. De Araujo, M. Fonseca, M. Pillsbury, and Quadrivalent Human Papillomavirus (HPV) Vaccination P. K. Singhal. 2012. “Estimated Health and Economic Impact in Mexico: A Transmission Dynamic Model-Based of Quadrivalent HPV (Types 6/11/16/18) Vaccination Evaluation.” Vaccine 26 (1): 128–39. doi:10.1016/J in Brazil Using a Transmission Dynamic Model.” BMC .Vaccine.2007.10.056. Infectious Diseases 12: 250. doi:10.1186/1471-2334-12-250. Jaffray, D. A., and M. K. Gospodarowicz. 2015. “Radiation Kim, S. Y., J. A. Salomon, and S. J. Goldie. 2007. “Economic Therapy for Cancer.” In Disease Control Priorities: Volume 3, Evaluation of Hepatitis B Vaccination in Low-Income Summary 19 Countries: Using Cost-Effectiveness Affordability Curves.” Ozawa, S., A. Mirelman, M. L. Stack, D. G. Walker, and Bulletin of the World Health Organization 85 (11): 833–42. O. S. Levine. 2012. “Cost-Effectiveness and Economic Knaul, F. M., E. Gonzalez-Pier, O. Gomez-Dantes, D. Garcia- Benefits of Vaccines in Low- and Middle-Income Countries: Junco, H. Arreola-Ornelas, and others. 2012. “The Quest A Systematic Review.” Vaccine 31 (1): 96–108. doi:10.1016/J for Universal Health Coverage: Achieving Social Protection .Vaccine.2012.10.103. for All in Mexico.” The Lancet 380 (9849): 1259–79. Peabody, J. W., M. M. Taguiwalo, D. A. Robalino, and J. Frenk. doi:10.1016/S0140-6736(12)61068-X. 2006. “Improving the Quality of Care in Developing Knaul, F. M., J. R. Gralow, R. Atun, and A. Bhadalia, eds. Countries.” In Disease Control Priorities in Developing 2011. Closing the Cancer Divide: An Equity Imperative. Countries, edited by D. T. Jamison, J. Breman, A. R. Measham, Cambridge, MA: Harvard Global Equity Initiative and G. Alleyne, M. Claeson, and others. 2nd ed. New York: Harvard University Press. Oxford University Press and World Bank. Knaul, F., S. Horton, P. Yerramilli, H. Gelband, and R. Atun. Piot, P., and T. C. Quinn. 2013. “Response to the AIDS 2015. “Financing Cancer Care in Low-Resource Settings.” Pandemic: A Global Health Model.” New England Journal of In Disease Control Priorities: Volume 3, Cancer, edited by Medicine 368 (23): 2210–18. doi:10.1056/Nejmra1201533. H. Gelband, P. Jha, R. Sankaranarayanan, and S. Horton. Pontén, J., H.-O. Adami, R. Bergstrom, J. Dillner, L-G. Friberg, 3rd ed. Washington, DC: World Bank. and others. 1995. “Strategies for Global Control of Cervical Levin, C. E., M. Sharma, Z. Olson, S. Verguet, J. F. Shi, and Cancer.” International Journal of Cancer 60: 1–26. others. 2015. “An Extended Cost-Effectiveness Analysis of Praditsitthikorn, N., Y. Teerawattananon, S. Tantivess, Publicly Financed HPV Vaccination to Prevent Cervical S. Limwattananon, A. Riewpaiboon, and others. 2011. Cancer in China.” In Disease Control Priorities: Volume 3, “Economic Evaluation of Policy Options for Prevention and Cancer, edited by H. Gelband, P. Jha, R. Sankaranarayanan, Control of Cervical Cancer in Thailand.” Pharmacoeconomics and S. Horton. 3rd ed. Washington, DC: World Bank. 29 (9): 781–806. doi:10.2165/11586560-000000000-00000. Mallath, M. K., D. G. Taylor, R. A. Badwe, G. K. Rath, V. Shanta, Prakash, C. 2003. “Crucial Factors that Influence and others. 2014. “The Growing Burden of Cancer in India: Cost-Effectiveness of Universal Hepatitis B Immunization Epidemiology and Social Context.” The Lancet Oncology in India.” International Journal of Technology Assessment in 15 (6): E205–12. doi:10.1016/S1470-2045(14)70115-9. Health Care 19 (1): 28–40. Metzger, M. L., S. C. Howard, L. C. Fu, A. Pena, R. Stefan, and Pramesh, C., R. Badwe, and R. Sinha. 2014. “The National Cancer others. 2003. “Outcome of Childhood Acute Lymphoblastic Grid of India.” Indian Journal of Medical and Paediatric Leukaemia in Resource-Poor Countries.” The Lancet Oncology 35 (3): 226–27. doi:10.4103/0971-5851.142040. 362 (9385): 706–08. doi:10.1016/S0140-6736(03)14228-6. Rabeneck, L., S. Horton, A. G. Zauber, and C. Earle. 2015. Mock, C. N., P. Donkor, A. Gawande, D. T. Jamison, M. E. Kruk, “Colorectal Cancer.” In Disease Control Priorities: Volume 3, and others. 2015. “Essential Surgery: Key Messages from Cancer, edited by H. Gelband, P. Jha, R. Sankaranarayanan, Disease Control Priorities, (third edition).” The Lancet and S. Horton. 3rd ed. Washington, DC: World Bank. (Feb 4. pii: S0140-6736(15)60091-5.). doi:10.1016/ Rao Seshadri, S., P. Jha, P. Sati, C. Gauvreau, U. Ram, and S0140-6736(15)60091-5. others. 2015. Karnataka’s Roadmap to Improved Health: Montero, A. J., K. Avancha, S. Gluck, and G. Lopes. 2012. Cost Effective Solutions to Address Priority Diseases, Reduce “A Cost-Benefit Analysis of Bevacizumab in Combination Poverty and Increase Economic Growth. Report, Azim Premji with Paclitaxel in the First-Line Treatment of Patients University, Bangalore, India. with Metastatic Breast Cancer.” Breast Cancer Research and Rodgers, M., M. Soares, D. Epstein, H. Yang, D. Fox, and others. Treatment 132 (2): 747–51. doi:10.1007/S10549-011-1919-Y. 2011. “Bevacizumab in Combination with a Taxane for the Monteiro, C. A., T. M. Cavalcante, E. C. Moura, R. M. Claro, First-Line Treatment of HER2-Negative Metastatic Breast and C. L. Szwarcwald. 2007. “Population-Based Evidence Cancer.” Health Technology Assessment 15 (Suppl 1): 1–12. of a Strong Decline in the Prevalence of Smokers in Brazil doi:10.3310/Hta15suppl1/01. (1989–2003).” Bulletin of the World Health Organization Salomon, J. A., N. Carvalho, C. Gutierrez-Delgado, R. Orozco, 85 (7): 527–34. A. Mancuso, and others. 2012. “Intervention Strategies Musgrove, P. 1996. “Public and Private Roles in Health: Theory to Reduce the Burden of Non-Communicable Diseases and Financing Patterns.” Health, Nutrition and Population in Mexico: Cost Effectiveness Analysis.” BMJ 344: E355. Discussion Paper. World Bank, Washington, DC. doi:10.1136/Bmj.E355. Norheim, O. F., P. Jha, K. Admasu, T. Godal, R. J. Hum, and Sankaranarayanan, R., B. M. Nene, S. S. Shastri, K. Jayant, others. 2014. “Avoiding 40% of the Premature Deaths in R. Muwonge, and others. 2009. “HPV Screening for Cervical Each Country, 2010–30: Review of National Mortality Cancer in Rural India.” New England Journal of Medicine Trends to Help Quantify the UN Sustainable Development 360 (14): 1385–94. doi:10.1056/Nejmoa0808516. Goal for Health.” The Lancet 385 (9964): 239–52. doi:10.1016 Sankaranarayanan, R., K. Ramadas, H. Amarasinghe, /S0140-6736(14)61591-9. S. Subramanian, and N. Johnson. 2015. “Oral Cancer.” OECD (Organisation for Economic Co-Operation and In Disease Control Priorities: Volume 3, Cancer, edited by Development). 2013. Cancer Care: Assuring Quality to Improve H. Gelband, P. Jha, R. Sankaranarayanan, and S. Horton. Survival. Health Policy Studies, OECD. 3rd ed. Washington, DC: World Bank. 20 Cancer Seya, M. J., S. F. Gelders, O. U. Achara, B. Milani, and Van Walbeek, C. 2005. “Tobacco Control in South Africa.” W. K. Scholten. 2011. “A First Comparison between the Promotion and Education 12 (Suppl 4): 25–8. doi:10.1177/1 Consumption of and the Need for Opioid Analgesics at 0253823050120040107. Country, Regional, and Global Levels.” Journal of Pain & Vanni, T., P. Mendes Luz, A. Foss, M. Mesa-Frias, and Palliative Care Pharmacotherapy 25 (1): 6–18. doi:10.3109 R. Legood. 2012. “Economic Modelling Assessment of /15360288.2010.536307. the HPV Quadrivalent Vaccine in Brazil: A Dynamic Shisana, O., T. Rehle, J. Louw, N. Zungu-Dirwayi, P. Dana, Individual-Based Approach.” Vaccine 30 (32): 4866–71. and others. 2006. “Public Perceptions on National Health doi:10.1016/J.Vaccine.2012.04.087. Insurance: Moving towards Universal Health Coverage in Varmus, H., and E. L. Trimble. 2011. “Integrating Cancer South Africa.” South Africa Medical Journal 96 (9): 814–18. Control into Global Health.” Science Translational Medicine Sloan, F. A., and H. Gelband. 2007. Cancer Control Opportunities 3 (101): 101cm28. doi:10.1126/Scitranslmed.3002321. in Low- and Middle-Income Countries. Washington, DC: Verguet, S., C. L. Gauvreau, S. Mishra, M. Maclennan, National Academy Press. S. M. Murphy, and others. 2015. “The Consequences Sullivan, T., R. Sullivan, and O. M. Ginsburg. 2015. “Screening of Tobacco Tax on Household Health and Finances in for Cancer: Considerations for LMICs.” In Disease Control Rich and Poor Smokers in China: An Extended Cost- Priorities: Volume 3, Cancer, edited by H. Gelband, P. Jha, Effectiveness Analysis.” The Lancet Global Health 3 (4): R. Sankaranarayanan, and S. Horton. 3rd ed. Washington, E206–16. doi:10.1016/S2214-109x(15)70095-1. DC: World Bank. Verguet, S., R. Laxminarayan, and D. T. Jamison. 2015. Tappenden, P., R. Jones, S. Paisley, and C. Carroll. 2007. “Universal Public Finance of Tuberculosis Treatment in “Systematic Review and Economic Evaluation of India: An Extended Cost-Effectiveness Analysis.” Health Bevacizumab and Cetuximab for the Treatment of Economics 24 (3): 318–32. doi:10.1002/Hec.3019. Metastatic Colorectal Cancer.” Health Technology WHO (World Health Organization). 2000. The World Health Assessment 11 (12): 1–128: III–IV. Report 2000: Health Systems: Improving Performance. Termrungruanglert, W., P. Havanond, N. Khemapech, Geneva: WHO. S. Lertmaharit, S. Pongpanich, and others. 2012. “Model for ———. 2011a. Global Status Report on Non-Communicable Predicting the Burden and Cost of Treatment in Cervical Diseases 2010. Geneva: WHO. Cancer and HPV-Related Diseases in Thailand.” European ———. 2011b. Scaling Up Action against Non-Communicable Journal of Gynaecological Oncology 33 (4): 391–94. Diseases: How Much Will It Cost? Geneva: WHO. Trimble, E. L., P. Rajaraman, A. Chao, T. Gross, C. Levin, and ———. 2011c.“Final Meeting Report and Recommendations.” others. 2015. “Cancer Research: The Need for National Immunization Practices Advisory Committee, WHO, Commitment.” In Disease Control Priorities: Volume 3, 11–13 April. http://www.who.int/immunization/policy Cancer, edited by H. Gelband, P. Jha, R. Sankaranarayanan, /committees/ipac_2011_april_report.pdf?ua=1. and S. Horton. 3rd ed. Washington, DC: World Bank. ———. 2012. Global Health Estimates 2012. http://www.who UICC (Union for International Cancer Control). Global Task .int/healthinfo/global_burden_disease/en/. Force on Radiotherapy for Cancer Control. http://gtfrcc.org/. ———. 2013. WHO Report on the Global Tobacco Epidemic UN (United Nations). 2015. Sustainable Development Goals. 2013. Geneva: WHO. http://sustainabledevelopment.un.org/?menu=1300. World Bank. 2014a. Country and Lending Groups. http://data UNPD (United Nations Population Division). 2012. World .worldbank.org/about/country-and-lending-groups. Population Prospects: The 2012 Revision. New York: UNPD. ———. 2014b. World Bank Development Indicators 2014. Table U.S. Preventive Services Task Force. 2012. Prostate Cancer: 2.15, Health Systems. http://wdi.worldbank.org/table/2.15. Screening, May 2012: Final Recommendation Statement. Zelle, S. G., K. M. Nyarko, W. K. Bosu, M. Aikins, L. M. Niens, and http:// www.uspreventiveservicestaskforce.org/page others. 2012. “Costs, Effects and Cost-Effectiveness of Breast /document /recommendationstatementfinal/prostate Cancer Control in Ghana.” Tropical Medicine & International -cancer-screening. Health 17 (8): 1031–43.doi:10.1111/J.1365-3156.2012.03021.X. Summary 21 Chapter 2 The Changing Global Burden of Cancer: Transitions in Human Development and Implications for Cancer Prevention and Control Freddie Bray and Isabelle Soerjomataram INTRODUCTION common cancers, and the expected future scale of the disease by 2030. Changes in fertility and life expectancy are leading to We link geographical and temporal patterns of cancer major changes in the structure of the global popula- to corresponding levels of economic progress to pro- tion and, in turn, in the scale of the cancer problem vide an overview of the key characteristics of the global worldwide and at every resource level (WHO 2011b). In cancer transition. We use gross national income (GNI) addition to the increasing burden of cancer is a changing per capita as a national indicator of societal as well spectrum of common cancers that is in different regions as economic development (http://data.worldbank.org correlated with levels of human development (Bray and /news/new-country-classifications), and corresponding others 2012). The ongoing cancer transition includes rates of cancer incidence and mortality as markers of the a reduction in infection-related cancers (for example, extent of the global cancer transition. We draw attention stomach and cervical cancer) that is offset by increases to geographical variations and trends in cancer-specific in cancers linked to a Westernization of lifestyle (for rates according to differing economic profiles and in example, breast, prostate, and colorectal cancer). The each of the world’s regions. In addition, we provide transition also encompasses changes in risk behavior, a global, trends-based projection of the likely cancer including tobacco uptake, with a delayed but large burden in 2030, based on historical trends refined by impact on the burden from lung and other tobacco- incorporating an indicator of level of development. related cancers (Bray and others 2012). The cancer We also examine the number of potentially avoid- transition is not uniform, however: in Sub-Saharan able new cases and cancer deaths, assuming a reduc- Africa, recent increases in cervical cancer are observed tion in risk factors (Hanley 2001). Even today, tobacco in Uganda and Zimbabwe; in many countries, a resid- smoking is by far the most important risk factor for ual burden of cancers associated with infectious agents cancer (Lim and others 2012). Although the smoking accompanies the increasing burden of cancers associated habit is in decline in many high-income countries with economic transition (Parkin and others 2014). (HICs), tobacco consumption is still rising in many This chapter presents a global overview of the low- and middle-income countries (LMICs) (Thun cancer burden, patterns and profiles, recent trends in and others 2012). As part of the global socioeconomic Corresponding author: Freddie Bray, Section of Cancer Surveillance, International Agency for Research on Cancer, Lyon, France, brayf@iarc.fr. 23 transition, many countries presently classified as CANCER IN CONTEXT: COMPARISONS WITH low- or middle-income are increasingly adopting OTHER NONCOMMUNICABLE DISEASES BY Westernized diets and more sedentary and less phys- ically active lifestyles, leading to a rapid shift in the DEVELOPMENT LEVEL profile of common cancers in these populations (Bray Worldwide, an estimated 38 million deaths—almost and others 2012). In view of these developments, this two-thirds of the annual 56 million total deaths—are chapter also reviews the main causes of cancer, with an caused by NCDs, principally from cardiovascular disease, emphasis on the sources of disparities that contribute diabetes, chronic respiratory disease, and cancer (WHO to an increasingly greater proportional burden from 2011b). In 2012, nearly 80 percent of all NCD deaths cancer in LMICs, and the prospects for cancer control (28 million) occurred in LMICs; almost 30 percent of in different settings. deaths occurred before age 60 years in these countries. We conclude by pointing to how the high-level polit- Among NCDs, cancer is a leading cause of death ical commitment to reduce the rising burden of cancer (figure 2.1). and other noncommunicable diseases (NCDs) can This observation is consistent with the global socio- advance the measurement of cancer to inform cancer economic transitions. Between 1990 and 2010, there control action. There remains an overwhelming need was a 17 percent reduction in communicable mater- to improve the quality and coverage of population- nal and neonatal and nutritional deaths, offset by a based cancer registration in most LMICs, as an essential 30 percent increase in NCD-related deaths (Lozano component in planning and evaluating national cancer and others 2012). This increase was greater in LMICs, control activities. compared with HICs (39 percent versus 15 percent, This chapter uses the World Health Organization’s respectively). The shift is largely driven by population (WHO) geographical regions: Africa, the Americas, growth and improved longevity. Although the number South-East Asia, Europe, Eastern Mediterranean, and of NCD-related deaths increased from 27 million to Western Pacific. 34 million between 1990 and 2010, death rates declined by 19 percent (646 to 520 per 100,000) over the same period (Lozano and others 2012). The overall statistics conceal variations among Figure 2.1 Top 10 Causes of Death Worldwide, 2012 countries. Cancer is an even larger component of the (millions of deaths) NCD burden in HICs. In Japan, cancer represents over half of all NCD deaths combined; in approximately 40 countries, age-adjusted cancer mortality rates are Malignant neoplasms equal to or above those of cardiovascular disease in pre- Ischaemic heart mature mortality at ages under 70 years (WHO 2013). disease Stroke Lower respiratory KEY INDICATORS OF THE GLOBAL CANCER infections BURDEN HIV/AIDS Comparisons of incidence rates can be used to investi- Diarrheal diseases gate cancer risk factors, aid planning and prioritizing of cancer control resources, as well as facilitate monitoring Preterm birth complications and evaluating of the impact of specific primary preven- tion interventions. Mortality has often been considered Road injury the best means of evaluating overall success in reducing Chronic obstructive the cancer burden. An assessment of cancer-specific pulmonary disease mortality rates is particularly useful in evaluating the Cirrhosis of the liver effectiveness of secondary prevention, particularly where the goal is early detection of malignant tumors; it is also 0 1 2 3 4 useful in tertiary prevention in determining the impact Estimated number of deaths for men and of cancer management and treatment. Combined suc- women, ages 0–69 years (millions) cesses in cancer prevention, early detection, screening, and treatment have resulted in a reduction in over- Source: Global Health Observatory, WHO 2013. all cancer mortality rates in some more developed 24 Cancer countries, predominantly as a result of declines in the Incidence incidence and/or mortality from a number of specific Cancer incidence is defined as the frequency of occur- types of common cancer (Doll 1990, Karim-Kos and rence of new cases of cancer in a specific population others 2008). The broad spectrum of interventions and over a given period of time. It can be expressed as the their tendency to produce real or artifactual changes in absolute number of cases or as a rate per unit-time, with cancer incidence, mortality, or survival, however, lends new cancer cases the numerator and the corresponding support to combining analyses of all three measures. person-time at risk the denominator. Cancer is complex to monitor accurately because of PBCRs are the essential institutions enabling these its biological diversity; the underlying coding and classi- activities (Bray 2014; Parkin 2006). They collect and fication issues are considerably more intricate than the classify information on all new cases of cancer within other NCD categories. There has been a long history of a well-defined population and provide statistics on cancer registration in many areas of the world, however, occurrence for the purposes of assessing and controlling and a tradition of maintaining comparable accurate and the impact of cancer in the community. Registries are complete global cancer incidence data that spans half either national or regional in their coverage, with a high a century. The serial publication, Cancer Incidence in degree of completeness, accuracy, and comparability of Five Continents (CI5, http://ci5.iarc.fr), published by the the collected data essential for making reliable infer- International Agency for Research on Cancer (IARC) ences regarding geographical and temporal variations and the International Association of Cancer Registries, in the underlying rates. Although incidence trends are is regarded as the definitive source of local cancer data. unaffected by the impact of changes in treatment and The 10th volume of CI5, published in 2013, compiles consequently survival, changes in registration prac- incidence data from high-quality population-based tices, definitions of malignancy, and the International cancer registries (PBCRs) for 2003–07. Classification of Diseases can all impact recorded The IARC is also the key reference source for statistics incidence (Muir, Fraumeni, and Doll 1994). Inclusion on the global, regional, and national burden of cancer; in CI5 is a reliable marker of the quality of a given it publishes estimates of the core indicators for cancer registry’s data, in that the editorial process includes at the country level as part of the GLOBOCAN series numerous assessments of the quality of the submitted (http://globocan.iarc.fr). The most recent database con- data set. A substantial disparity exists in the availability tains estimates of the cancer incidence, mortality, prev- of high-quality cancer registration data between HICs alence, and disability-adjusted life years for all cancers and LMICs. While 99 percent of the North American combined for 2012, as well as for 27 cancer sites in population is covered in the 10th volume of CI5, only 184 countries and 20 regions (Ferlay and others 2013). 7 percent, 5 percent, and 2 percent of the respective The methods used to compile the GLOBOCAN South American, Asian, and African countries have estimates have evolved, but the underlying principle registries that were accepted for inclusion. remains the use of the best available observed data within a country to build up the global picture. Details are provided by Ferlay and others (2015). Estimation Mortality of cancer-specific incidence and mortality rates are Mortality provides a unique measure of the outcome dependent on local data sources, and the cancer or impact of cancer and is expressed as either the num- registry data from CI5 provide a key input into the ber of deaths occurring or a mortality rate per unit- compilation. Unfortunately, there remains a paucity time. Mortality is a product of the incidence and the of high-quality cancer incidence and mortality data in case-fatality of a given cancer, and fatality (1-survival) low-resource and medium-resource areas, and hence represents the probability that an individual with cancer the accuracy of the estimates from these regions is will die from it. Mortality rates then measure average risk generally lower; an alphanumeric scoring system has to the population of dying from a specific cancer. If case been introduced to provide a broad indication of fatality is constant across populations or in a specified the robustness of the estimation within each country population over time, geographic or temporal compar- (Ferlay and others 2015). isons of incidence can be inferred from mortality. It is The definitions, sources, and mode of estimation evident, however, most notably in HICs and for cancers of the key indicators used in this chapter are provided. where treatment and cancer management have markedly Counts and rates of cancer incidence and mortality improved, that incidence and mortality trends can be at are presented; rates per 100,000 were age-standardized considerable variance. using the world standard population (Doll, Payne, and Data derive from vital registration systems, where Waterhouse 1966). usually a clinical practitioner certifies the fact and cause The Changing Global Burden of Cancer: Transitions in Human Development and Implications for Cancer Prevention and Control 25 of death. The International Classification of Diseases Proportion of the Burden Attributable to Risk Factors provides the uniform system for nomenclature and The population attributable fraction (PAF) (Hanley coding and a suggested format for the death certif- 2001) is a standard measure used to quantify the pro- icate. Mortality data are affected by the accuracy of portion of disease burden attributable to a risk factor. the recorded cause of death and the completeness of It is calculated using information on the prevalence of a registration. Errors of death certification are well doc- risk factor in a population and the relative risk that the umented (Doll and Peto 1981); patients diagnosed with risk factor poses for development of disease, compared cancer, for example, may die of the disease without this with those where the risk factor is absent. The derived being written on the death certificate, and such inaccu- proportion is indicative of the reduction in cancer inci- racies are known to vary considerably between countries dence or mortality that would be expected if exposure and over time. to the risk factor were absent. For risk factors where The WHO mortality databank hosted at IARC zero exposure is inapplicable, such as for body weight, (http://www-dep.iarc.fr/WHOdb/WHOdb.htm) con- PAF is calculated by comparing population exposure to tains national cancer mortality data on more than reported exposures at the lowest level of cancer risk. 80 countries; for many, the data cover decades. The The results presented in this chapter are largely based greater availability by country and over time is confined on a systematic search of published and unpublished mainly to higher-income countries; at least 15 years of reviews from major databases, censuses, health and data are available for 78 countries (35 in Europe, 25 in nutrition surveys, and epidemiological studies (Lim and the Americas, 15 in Asia, 2 in Oceania, and 1 in Africa), others 2012). Wherever possible, the data are supple- but may explain its usage as a surrogate for incidence mented by national, regional, or global studies with a in geographic and temporal studies of cancer. In South focus on cancer (Agudo and others 2012; Boffetta and America, mortality data are more comprehensive than others 2006; Boffetta and others 2009; de Martel and incidence data, while both incidence and mortality others 2012; Ezzati and others 2005; Forman and others are limited in Asia and Africa. In such settings, where 2012; Goldie and others 2008; Goldstein and others deaths are often not attended by physicians, alternative 2005; McCormack and others 2012; McCormack and methods, such as verbal autopsy, may be used to comple- Schuz 2012; Renehan and others 2010). Because of the ment data collected from death certificates. known lag time between the inception of smoking and cancer development, lung cancer mortality rates were converted into smoking impact ratios and as a proxy for Income as a Proxy for Human Development exposure to smoking, thereby taking into account the The remainder of this section examines cancer profiles duration and intensity of the exposure. To calculate PAF with respect to the current burden, recent changes, linked to infection, the prevalence of infection among and the future burden at the global, regional, and cancer cases was used instead of the prevalence of infec- national levels. It is important to understand these pat- tion in the population, because of the unreliability of terns and trends in relation to the rapid increments in population surveys on infection prevalence (de Martel human development over recent decades. Tracking these and others 2012). As for effect sizes relating exposure to changes helps to clarify how demographic and epidemi- disease, wherever possible, the results were as reported in ologic transitions impact the overall burden and the meta-analysis or were derived from the most up-to-date changing distribution of common cancers in different epidemiological findings. populations. In keeping with previous DCP volumes, The quality of the estimation depends largely on we utilized national income level (GNI per capita) as a data availability, and a large gap remains in data avail- proxy for human development (UNDP 2009). ability and data quality in LMICs relative to HICs. The In examining aspects of the cancer burden according estimation of PAF also assumes that the association to resource levels, we used predefined categories of the between a risk factor and cancer is causal and that a distribution according to the World Bank classification reduction in exposure to a specific risk factor will lead of countries by income-group (see endnote for per to a decline in the incidence of the cancer in question capita income cut-offs). The estimated burden of cancer (Rockhill, Newman, and Weinberg 1998). To avoid an in 2030 is presented by gender for all cancers combined overestimation of the role of certain risk factors, we only for all levels of income on the basis of the rates in 2008, report results for factors that have been confirmed as assuming population growth and aging in 2030 occur having sufficient evidence in causing cancer (Cogliano according to the United Nations World Population and others 2011). Epidemiological studies that report Prospects medium-fertility variant. risk associations between exposure and cancer are prone 26 Cancer to several limitations. In estimating PAF, studies have Four cancers constitute nearly 50 percent of the tried to overcome this issue by exclusively using risk total cancer burden in HICs; for cancers of the prostate, estimates based on large meta-analyses that included lung, and colorectum, comparable numbers (between only high-quality studies and, wherever possible, only 350,000 and 400,000 new cases) of each type are esti- cohort studies. mated in 2012, with female breast cancer accounting for 550,000 cases (figure 2.3). Female breast cancer is also the most common form of cancer in LMICs, with GLOBAL CANCER BURDEN AND almost 770,000 estimated cases in 2012. In combina- DISTRIBUTION ACCORDING TO INCOME tion, eight cancers (breast, lung, liver, cervix, stomach, colorectum, esophagus, and leukemia) account for GROUP almost two-thirds of the estimated cancer burden in For 2012, GLOBOCAN estimates indicate that there were LMICs. It is noteworthy that several types of cancer, 9.1 million new cases of cancer and 4.4 million cancer more commonly associated with infection and poverty, deaths between the ages of 0 and 69 years (figure 2.2). less commonly diagnosed in HICs, are major cancers in Female breast cancer is the most frequently diagnosed LMICs; they include stomach and liver cancer—which cancer globally (1.3 million new cases, 14.7 percent of are more frequently diagnosed than colorectal or pros- all cancer cases), but it ranks third as the most common tate cancer in lower-income economies—as well as cause of cancer death (358,000 deaths, 8.2 percent of cancers of the cervix and esophagus. all cancer deaths). Lung cancer is the leading cause of In terms of risk of developing specific neoplasms death (767,000 deaths, 17.5 percent) and ranks second worldwide, the lifetime incidence of breast cancer is in incidence (1 million deaths, 10.8 percent). In terms highest (2.8 percent), with the lifetime risk of the of incidence, the most frequent cancer types thereafter next four most common cancers varying from 1 to are colorectal (744,000 cases, 174,000 deaths), prostate 1.5 percent. In HICs, however, the lifetime risk exceeds (556,000 cases, 56,000 deaths), stomach (539,000 cases, 2 percent for cancers of the breast (female), prostate, 242,000 deaths), and liver (503,000 cases, 336,000 deaths). lung, and colorectum. Figure 2.2 Global Burden of Cancer: Estimated Counts of Cancer Incidence and Mortality, by Region, 2012 a. Incidence b. Mortality Middle East and North Africa 3.5% Sub-Saharan Africa 5.9% Middle East and North Africa 3.8% Sub-Saharan Africa Europe and 7.9% Central Asia 23.9% South Asia Europe and 12.3% Central Asia Latin America 20.0% South Asia and the Caribbean 16.1% 7.3% Latin America and the Caribbean North America 7.9% 6.9% East Asia and Pacific North America East Asia and Pacific 38.1% 12.1% 34.4% 4,375,000 Deaths 9,070,000 New cases Source: Ferlay and others 2013. Note: Data are for men and women ages 0–69 years. The Changing Global Burden of Cancer: Transitions in Human Development and Implications for Cancer Prevention and Control 27 Figure 2.3 Global Burden of Cancer: Cancer Incidence and Mortality by Income Level, 2012 a. Incidence 800 600 0–69 years (thousands) New cases in 2012, 400 200 0 a us s id st ng m te h er om gu ac yro ea tu ta Liv er Lu om ha ph ec os ut Br Th op lor lym Pr St x, Es rvi Co kin Ce dg Ho n− No b. Mortality 0–69 years (thousands) 400 Deaths in 2012, 200 0 ng er t h m s us ia as m s gu ac tu em ste ea Liv er re Lu om ha ec Br ut nc ka sy op lor St Pa x, u us Es Le rvi Co rvo Ce ,ne ain Br Low and middle income High income Source: Ferlay and others 2013. Note: Values are age-standardized rates for the world population structure. In terms of mortality, the most frequent causes of mortality burden occurs in LMICs, particularly for cancer death in HICs are lung, colorectum, female those cancers more frequently diagnosed in LMICs, breast, and pancreas, which together account for including liver and stomach (figure 2.3). Together about 48 percent of the total cancer deaths in 2012. with lung, female breast, and colorectal cancer, these Relative to incidence, a greater proportion of the neoplasms constitute half of the total cancer mortality 28 Cancer burden in these areas. In terms of risk of death, the to high-income (figures 2.4 and 2.5); for colorectal lifetime cumulative mortality from lung cancer and cancer, the increases are more evident in registries female breast cancer ranks highest in HICs (1.8 transitioning from middle-income countries; those and 1.2 percent, respectively) and LMICs (1.3 and countries that have reached very high levels of income 1 percent, respectively). Colorectal cancer ranks third tend to have also attained very high incidence rates of in HICs, and liver and cervical cancer rank third and this cancer, and a recent peak in incidence is observed fourth, respectively, in LMICs. in several countries (figure 2.6). Nevertheless, the trends point to a Westernization effect in countries undergoing transition. Increasing average levels of DIVERSITY OF CANCER BY TYPE: economic progress serve as a proxy for a changing GEOGRAPHICAL VARIATIONS BY REGION population prevalence of reproductive, dietary, meta- The classification of cancer patterns by broad levels of bolic, and hormonal risk factors (toward levels more economic development draws attention to some of the akin to those more commonly observed in the West) key macroeconomic determinants of inequality and that correspondingly increase the risk of these cancers the variations in the risk of developing cancer globally. (Bray 2014). To underscore the diversity of cancer, it is important to Evidently, better diagnosis, earlier detection, and show the extent to which cancer-specific patterns vary. screening may have inflated the incidence burden in many Map 2.1, panel a, outlines the most common form of settings, particularly where precursor lesions are not the cancer among men in 184 countries in 2012. Worldwide, target of the intervention. This includes prostate and 10 types of cancer rank highest, with prostate cancer breast cancers where prostate-specific antigen testing and the most frequent cancer form in 68 countries, chiefly mammographic screening have impacted the increasing in HICs, but also in parts of Latin America and the incidence of the respective cancers in many populations Caribbean and in the southern regions of the African (Bleyer and Welch 2012; Bray and others 2010; Bray, continent. Lung cancer is the most common neoplasm McCarron, and Parkin 2004). Nevertheless, such increases in 42 countries, mainly in Eastern Europe, Northern are countered by another hallmark of cancer transition: Africa, and Asia, including China and Indonesia. Liver a reduction in infection-based neoplasms (Forman and cancer is the most common form in 27 countries, partic- others 2012; Vaccarella and others 2013). Most of the ularly in Western Africa and South-East Asia. registries record declines in cervical and stomach cancer Among men, the incidence and mortality patterns incidence (Bray and others 2012) (figures 2.7 and 2.8). vary most markedly by region in Africa (Jemal and There have been rapid increases in the incidence of others 2012), with liver cancer the most common cancer breast, prostate, and colorectal cancer documented in in 18 of the 47 Sub-Saharan African countries. Clusters countries in Sub-Saharan Africa with good quality inci- of very high incidence of Kaposi sarcoma are observed dence data; for example, rates of breast cancer have in 13 countries in Eastern Africa; prostate cancer is increased by 3.6 percent annually in Kampala, Uganda, the most frequently observed tumor in nine countries and by 4.9 percent in the black population of Harare, (Parkin and others 2014). In women—and in some Zimbabwe. There is little evidence of another hallmark contrast to the diversity of the highest ranking cancers of cancer transition—declining cervical cancer—in these in men—either breast cancer (143 countries) or cervical populations: rather, rates exhibit persistent increases in cancer (39 countries) ranks as the most frequent cancer incidence in recent years (Chokunonga and others 2013; diagnosed in almost all countries worldwide; exceptions Wabinga and others 2014). include the very high female incidence rates of thyroid The importance of lung cancer in the overall cancer in the Republic of Korea and liver cancer in global burden from cancer is a reflection of the prev- Mongolia (map 2.1, panel b). alence of tobacco smoking. The differences in lung cancer incidence by national income level reflect the historical and current variations of smoking habits GLOBAL CANCER TRANSITIONS: RECENT in the underlying populations. Lung cancer rates are decreasing in men in most countries, notably in TRENDS AND FUTURE BURDEN CIRCA 2030 HICs and in countries where the habit has long been The high-quality cancer registry data from CI5 (http:// established, but they are increasing in women (Jemal ci5.iarc.fr) illustrate temporal aspects of the global and others 2010). Such contrasting male and female cancer transition: increases in breast and prostate trends are consistent with the population-specific rise cancer are apparent in countries that have transitioned and fall in adult cigarette utilization over past decades from low-income to middle-income or middle-income (Lortet-Tieulent and others 2013). The correlation The Changing Global Burden of Cancer: Transitions in Human Development and Implications for Cancer Prevention and Control 29 Map 2.1 Most Common Cancers Diagnosed in Men and Women in 184 Countries, 2012, Ages 0–69 Years a. Most common cancers among men, 2012 Site (number of countries) Prostate (68) Lung (42) Liver (27) Kaposi Sarcoma (13) Colorectal (11) Stomach (11) Lip, Oral cavity (5) Non-Hodgkin Lymphoma (3) Leukemia (2) Esophagus (2) b. Most common cancers among women, 2012 Site (number of countries) Breast (143) Cervix Uterine (39) Liver (1) Thyroid (1) Source: Ferlay and others, 2013. between income and male smoking prevalence has China and Indonesia (Jha 2009). The future burden shown a curvilinear trend (Talley 2010) that mimics of lung cancer and other smoking-related cancers the smoking epidemic curve (Thun and others 2012). (IARC 2004), such as bladder, esophageal, and oral Smoking prevalence tends to be low in countries of cavity in different regions and countries, will largely low average income but rises as income increases, depend on gender-specific smoking patterns at the until it stabilizes and begins to decline as countries population level, including the duration of smoking, attain high mean income levels. extent of cessation, and types of tobacco smoked The prevalence of smoking in men has increased (Doll and others 2004; Pirie and others 2013; Thun sharply in many middle-income countries, including and others 2012). 30 Cancer Figure 2.4 Trends in Age-Standardized (World) Incidence Rates of Female Breast Cancer, by Income Level, 1980–2010 a. Low- and middle-income countries b. High-income countries 80 80 70 70 60 60 Age-standardized (world) incidence rate per 100,000, 0–69 years Age-standardized (world) incidence rate per 100,000, 0–69 years 50 50 40 40 30 30 25 25 20 20 15 15 13 13 1980 1990 2000 2010 1980 1990 2000 2010 Calendar year Calendar year Philippinesa Colombiaa Chinaa Australia United Kingdom, Israel England Costa Rica Indiaa Ugandaa Norway Czech Republic Japana Source: CI5plus, http://ci5.iarc.fr/CI5plus/Default.aspx. Note: Estimated from selected population-based cancer registries of consistently high quality (included in successive volumes of Cancer Incidence in Five Continents). Data for the economies in the graphs are for China (Hong Kong SAR, China; and Shanghai), Colombia (Cali), India (Chennai and Mumbai), Japan (Miyagi, Nagasaki, and Osaka), the Philippines (Manila), and Uganda (Kampala). a. Denotes rates based on an aggregate of one or more regional registries, as indicated. A recent study of incidence trends in 1988–2002 for to education. Although the HDI is highly correlated with the seven most frequently diagnosed cancers worldwide GNI, the latter reflects average national income, whereas (constituting more than 58 percent of the global bur- the former additionally provides some indication as to den) estimated the annual percentage change of trends how income is spent, at least in the areas of health and in age-adjusted incidence using data from 101 cancer education (UNDP 2009). registries (Bray and others 2012). The trends were strat- An estimated 21.6 million new cancer cases are pre- ified by medium, high, or very high levels of the Human dicted for 2030 (an increase of 53 percent from 2012), Development Index (HDI), a composite measure of life based solely on projected demographic changes and expectancy at birth, income as GNI per capita, and access unchanged cancer incidence rates (figure 2.9). If the The Changing Global Burden of Cancer: Transitions in Human Development and Implications for Cancer Prevention and Control 31 Figure 2.5 Trends in Age-Standardized (World) Incidence Rates of Prostate Cancer, by Income Level, 1980–2010 a. Low- and middle-income countries b. High-income countries 80 80 60 60 50 50 Age-standardized (world) incidence rate per 100,000, 0–69 years Age-standardized (world) incidence rate per 100,000, 0–69 years 30 30 20 20 15 15 10 10 5 5 2 2 1.5 1.5 1 1 1980 1990 2000 2010 1980 1990 2000 2010 Calendar year Calendar year a a Colombia Costa Rica Philippines Norway Australia United Kingdom, Ugandaa Indiaa Chinaa England Czech Republic Israel Japana Source: CI5plus, http://ci5.iarc.fr/CI5plus/Default.aspx. Note: Estimated from selected population-based cancer registries of consistently high quality (included in successive volumes of Cancer Incidence in Five Continents). Data for the economies in the graphs are for China (Hong Kong SAR, China; and Shanghai), Colombia (Cali), India (Chennai and Mumbai), Japan (Miyagi, Nagasaki, and Osaka), the Philippines (Manila), and Uganda (Kampala). a. Rates based on an aggregate of one or more regional registries, as indicated. cancer-specific incidence trends up to 2002 continue, an in the number of cancer patients through resource- absolute increase of 62 percent in incidence is expected, appropriate interventions. Such actions include pri- representing an overall increase of 1.2 million new cases mary prevention strategies to control lifestyle factors, per year in 2012–30. The increases in cancer incidence including tobacco avoidance and cessation foremost, are proportionally greatest in low HDI settings, with a a reduction in alcohol consumption and obesity, and predicted 77 percent increase in both genders. the promotion of increased levels of physical activity. Although such analyses appear to present a rather Vaccination programs for liver and cervical cancer and pessimistic view of the future burden of cancer, targeted early detection programs for breast and cervical cancer interventions can significantly reduce the projected rise are important. 32 Cancer Figure 2.6 Trends in Age-Standardized (World) Incidence Rates of Male Colorectal Cancer, by Income Level, 1980–2010 a. Low- and middle-income countries b. High-income countries 40 40 30 30 Age-standardized (world) incidence rate per 100,000, 0–69 years Age-standardized (world) incidence rate per 100,000, 0–69 years 25 25 20 20 15 15 10 10 7 7 5 5 4 4 1980 1990 2000 2010 1980 1990 2000 2010 Calendar year Calendar year a a a China Philippines Colombia Australia Czech Republic Norway Costa Rica Ugandaa Indiaa Israel United Kingdom, Japana England Source: CI5plus, http://ci5.iarc.fr/CI5plus/Default.aspx. Note: Estimated from selected population-based cancer registries of consistently high quality (included in successive volumes of Cancer Incidence in Five Continents). Data for the economies in the graphs are for China (Hong Kong SAR, China; and Shanghai), Colombia (Cali), India (Chennai and Mumbai), Japan (Miyagi, Nagasaki, and Osaka), the Philippines (Manila), and Uganda (Kampala). a. Rates based on an aggregate of one or more regional registries, as indicated. RISK FACTORS, INTERVENTIONS AND Smoking and dietary risk factors have particularly POLICIES, AND POTENTIAL IMPACTS ON THE important roles in potentially reducing the cancer bur- den in HICs and LMICs (Ezzati and others 2005; Jha BURDEN OF CANCER 2009; Lim and others 2012; WCRF and AICR 2007). Approximately 30 percent of all cancers worldwide For cancers with strong risk modifiers, trends in risk are considered preventable by modification of the factors are followed by trends of cancer incidence with predominant risk factors (Danaei and others 2005; a lag of 20–30 years (Doll and others 2004; Jha 2009). Martin-Moreno, Soerjomataram, and Magnusson 2008). Illustrative examples include the temporal patterns of The Changing Global Burden of Cancer: Transitions in Human Development and Implications for Cancer Prevention and Control 33 Figure 2.7 Trends in Age-Standardized (World) Incidence Rates of Cervical Cancer, by Income Level, 1980–2010 a. Low- and middle-income countries b. High-income countries 50 50 40 40 Age-standardized (world) incidence rate per 100,000, 0–69 years Age-standardized (world) incidence rate per 100,000, 0–69 years 30 30 25 25 20 20 15 15 10 10 7 7 6 6 5 5 4 4 3.5 3.5 1980 1990 2000 2010 1980 1990 2000 2010 Calendar year Calendar year Ugandaa Colombiaa Indiaa Czech Republic Norway Japana Philippinesa Costa Rica Chinaa United Kingdom, Australia Israel England Source: CI5plus, http://ci5.iarc.fr/CI5plus/Default.aspx. Note: Estimated from selected population-based cancer registries of consistently high quality (included in successive volumes of Cancer Incidence in Five Continents). Data for the economies in the graphs are for China (Hong Kong SAR, China; and Shanghai), Colombia (Cali), India (Chennai and Mumbai), Japan (Miyagi, Nagasaki, and Osaka), the Philippines (Manila), and Uganda (Kampala). a. Rates based on an aggregate of one or more regional registries, as indicated. cigarette smoking and lung cancer, and age at first child- Tobacco birth and breast cancer (Jha 2009; Soerjomataram and Tobacco smoking is the largest single avoidable cause others 2008). Unfortunately, for most risk factors, the of premature death. Worldwide, 70 percent of lung relation to specific cancer forms is weaker than these cancer deaths; 42 percent of cancers of the esophagus associations. In addition, risk factors often interact, mak- and oral cavity (Ezzati and others 2005; Lim and others ing it difficult to link directly the changing prevalence 2012); a significant proportion of cancers of the larynx, of a single risk factor to the observed trends of specific urinary bladder, and pancreas; a smaller proportion of cancer types. cancers of the kidney, stomach, and cervix; and myeloid 34 Cancer Figure 2.8 Trends in Age-Standardized (World) Incidence Rates of Men with Stomach Cancer, by Income Level, 1980–2010 a. Low- and middle-income countries b. High-income countries 50 50 40 40 Age-standardized (world) incidence rate per 100,000, 0–69 years Age-standardized (world) incidence rate per 100,000, 0–69 years 30 30 25 25 20 20 15 15 10 10 7 7 6 6 5 5 4 4 3 3 2.5 2.5 1980 1990 2000 2010 1980 1990 2000 2010 Calendar year Calendar year a a a Costa Rica China Colombia Japan Czech Republic United Kingdom, Philippinesa Indiaa Ugandaa England Norway Israel Australia Source: CI5plus, http://ci5.iarc.fr/CI5plus/Default.aspx. Note: Estimated from selected population-based cancer registries of consistently high quality (included in successive volumes of Cancer Incidence in Five Continents). Data for the economies in the graphs are for China (Hong Kong SAR, China; and Shanghai), Colombia (Cali), India (Chennai and Mumbai), Japan (Miyagi, Nagasaki, and Osaka), the Philippines (Manila), and Uganda (Kampala). a. Rates based on an aggregate of one or more regional registries, as indicated. leukemia are attributable to cigarette smoking (Agudo tobacco consumption are central to preventing deaths and others 2012). Between 1990 and 2010, mortal- from cancer and other diseases. ity rates from cancers related to smoking in higher Smoking prevalence has decreased for many decades resource settings decreased (Lim and others 2012), in HICs; in LMICs, it is stable or showing only lim- reflecting a mean decrease of smoking prevalence, espe- ited signs of a decline very recently. In China, for cially among men. However, in countries undergoing example, with one of the highest smoking rates in the transition, the proportion of cancer deaths attribut- world, prevalence among men dropped from 57 to able to smoking has increased from 12 to 14 percent 53 percent between 2002 and 2007 (Ng and others 2014). (Lim and others 2012), reflecting increasing cigarette This change may represent the first impact of increasing consumption (Ng and others 2014). Efforts to reduce tobacco control measures in the country. The WHO The Changing Global Burden of Cancer: Transitions in Human Development and Implications for Cancer Prevention and Control 35 Figure 2.9 Cancer Incidence Projected to 2030 with and without debatable whether the impact of tobacco on the cancer Incorporating Recent Changes in Rates by Gender, Income Level, burden in some LMICs will reach the level observed in and Age Group many HICs (McCormack and Boffetta 2011). a. Younger than 65 years Persistent Infections Low income The most recent estimates state that at least 16.1 percent Medium income or two million new cases of human malignancies Males worldwide are attributable to persistent infections with High income bacteria, viruses, or parasites (de Martel and others 2012). World This fraction is higher in LMICs (22.9 percent or 1.6 million new cases) than in HICs (7.4 percent or 410,000 Low income new cases). Helicobacter pylori, hepatitis B and C viruses, and human papilloma viruses (HPVs) are responsible Females Medium income for 1.9 million cases, mainly gastric, liver, and cervical High income cancers. The remaining 163,000 cases are attributed to World Epstein-Barr virus, human herpes virus type 8, human T-cell lymphotropic virus type 1, Opisthorchis viverrini, 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 Clonorchis sinensis, and Schistosoma haematobium. By Incidence (thousands) region, the attributable fraction shows marked varia- tion (figure 2.10), from 33.2 percent of all cancers in b. Age 65 years or older Sub-Saharan Africa to 3.3 percent in Australia and New Zealand. The proportion of cancers associated with Low income chronic infections ranges from 1 in 3 in Sub-Saharan Medium income Africa and 1 in 4 in China to 1 in 30 in Australia and New Males Zealand and 1 in 25 in North America. Awareness of the High income role of infectious agents in cancer has expanded rapidly, World leading to the formulation of new strategies, including vaccines against hepatitis B to prevent liver cancer, and Low income against high-risk strains of HPV to prevent cervical and other HPV-related cancer types (Beutels 2001; Harper Females Medium income and others 2004; Herrero and others 2013). High income Increasing coverage of hepatitis B virus vaccina- World tion has been related to a decrease in the prevalence of chronic hepatitis B virus throughout the world 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 (Goldstein and others 2005). For example, in Taiwan, Incidence (thousands) China, a 75 percent decrease in the incidence of hepa- Incidence 2012 Incidence 2030 (+ postulated tocellular carcinoma in children was reported since changes in trends) the initiation of routine infant hepatitis B vaccination Incidence 2030 (demographics) (Chang and others 1997). Gavi, the Vaccine Alliance, has subsidized vaccines in poor countries, leading to a Source: Bray and others 2012. rapid increase in coverage of hepatitis B vaccination, with a similar effort underway in implementing HPV MPOWER report on the global tobacco epidemic indi- vaccination (Kane and others 2012). Based on current cated that, of the 48 countries that had implemented at epidemiologic data in 72 countries, it is predicted that least one tobacco control measure since 2007, four-fifths we will observe a mean reduction in lifetime risk of were LMICs (WHO 2008). The impact of such mea- cervical cancer of between 31 percent (for example, sures will not be evident in lung cancer trends for some in Guinea and Senegal) and 60 percent (for example, decades, and increasing rates may continue to rise in in Ethiopia), assuming an increase in vaccination many LMICs (McCormack and Boffetta 2011). The prev- coverage to 70 percent (Goldie and others 2008). alence of women in LMICs who smoke has remained Differences in the results of vaccination stem from rather low, but it may be expected to rise as social and variation in risk factor prevalence and the background cultural prohibitions erode (Thun and others 2012). It is incidence of the cancer being targeted. Safe and 36 Cancer effective vaccines have proven highly important in Figure 2.10 Cancer and Infection: Estimated Attributable Fraction, by reducing infection-related cancers. Region or Country, 2008 Sub-Saharan Africa Alcohol China Excess alcohol consumption is an important risk factor Other Eastern Asia for several cancers, most notably, of the oral cavity, India Japan pharynx, larynx, and esophagus, as well as for cancers Other Oceania of the breast, liver, and colorectum (IARC 2010). Up to South America 4 percent of cancer cases worldwide are attributable to Other Central Asia North Africa and Western Asia alcohol intake (Boffetta and others 2006), with similar Europe distributions in HICs and LMICs (Rehm and Shield North America 2013). Global alcohol consumption has been stable in Australia and New Zealand recent decades, with a small decrease in countries where intake was historically high (for example, in Europe) and 0 5 10 15 20 25 30 35 a small increase in countries where intake has been rela- Attributable fraction (%) tively low (for example, in South-East Asia) (Shield and Source: de Martel and others 2012. others 2013). In HICs, a notable exception is an increase Note: Values are for all cancers combined, for all ages. in consumption in the Russian Federation between 1986 and 2000, which has subsequently stabilized. Among cancer deaths worldwide (387,000) would be attributable LMICs, Uganda has the heaviest alcohol intake level to dietary risk factors (Lim and others 2012). (10.9 liters of pure alcohol per capita, compared with Consumption of red meat and processed meat 11.3 in the Russian Federation) (WHO 2011a). This is increases the risk of colorectal cancer by an estimated in contrast to traditionally Muslim countries (for exam- 43 percent (WCRF and AICR 2007). Consumption of ple, the Arab Republic of Egypt and Indonesia), where red meat and processed meat is generally rising in LMICs alcohol consumption is very low (0.66 and 0.67 liters but is stable in HICs. Another major dietary risk factor of pure alcohol per capita, respectively) (WHO 2011a). is high salt intake, which increases the risk of stomach Sweden is one example of a successful government pol- cancer. Globally, salt intake has declined, and the asso- icy geared toward effectively reducing alcohol intake at ciated cancer burden has declined as a result (Brown the population level. After restricting beer sales based and others 2009); currently, 1.5 percent and 2 percent on alcohol content in 1977, consumption decreased by of all cancer deaths in HICs and LMICs, respectively, are approximately 15 percent (Makela, Tryggvesson, and attributed to high salt intake, compared with 2.2 percent Rossow 2002). Such an example points to the impor- and 2.7 percent, respectively, in 1990 (Lim and others tance of nationwide population interventions as an 2012). In the United Kingdom, 10 percent of new effective way to reduce alcohol consumption. stomach cancer cases diagnosed in 2010 may be attrib- utable to high salt intake (Parkin 2011). Although many of the underlying mechanisms are Diet not yet established, food and diet are clearly important Recent estimates report that about 15 percent of all determinants of cancer risk. Over the past century, better cancer deaths are related to unhealthy diets, including food preservation methods, reducing the need for salt, high intake of red meat, processed meat, and sodium, have probably been a key factor in the decline in stomach as well as low intake of fruits and vegetables (Lim and cancer incidence. The levels of red meat consumption others 2012). Yet, most research is still inconclusive about in recent decades may explain the low colorectal cancer specific dietary items (WCRF and AICR 2007). The most rates in southern Asia (WCRF and AICR 2007), as well recent comprehensive review of diet and cancer risks as the high rates in many western European countries confirms an elevated risk related to red and processed (Center, Jemal, and Ward 2009; Center and others 2009). meat but finds less evidence supporting the benefit of Evidence linking obesity and overweight to cancer is the consumption of fruits and vegetables in reducing firmer than the dietary evidence, indicating that 3 and risk (WCRF and AICR 2007). Variations in the strength 6 percent of cancer deaths in LMICs and HICs, respec- of the evidence in different studies over time make the tively, can be attributed to excess weight, as quantified by measurement of dietary intake problematic in epidemi- the body mass index (BMI) (Bergstrom and others 2001; ological studies; if only well-established risk factors are Lim and others 2012; Renehan and others 2010). BMI included in the estimation, approximately 5 percent of is already an important correlate of cancer incidence in The Changing Global Burden of Cancer: Transitions in Human Development and Implications for Cancer Prevention and Control 37 HICs, and it is likely to continue increasing in importance Many outdoor air pollutants may increase the risk of in some LMICs, as average BMI has increased in most neoplasms, most notably lung cancer (Raaschou-Nielsen Central and South American countries and also in south- and others 2013). Increasing air pollution in large cities ern Africa (Stevens and others 2012). BMI has remained in LMICs has caused concern about links to several low in Sub-Saharan Africa (except in southern Africa) chronic diseases, including cancer (Evans and others and South and Southeast Asia. High BMI has increased as 2013). Currently 3.6 percent of cancer deaths in LMICs a cause of cancer deaths more substantively in LMICs, a are attributed to ambient air pollution, compared with 54 percent increase over the past 30 years, compared with 2.4 percent in 1990 (Lim and others 2012). a 26 percent increase in HICs over the same period (Lim Protective measures in the workplace have led to and others 2012; Stevens and others 2012). the prevention of many cancers. A small proportion of cancer deaths are attributed to occupational risk factors (1.6 percent in LMICs and 1.3 percent in HICs). Despite Physical Inactivity the small numbers, there is cause for concern regarding Regular physical activity is associated with a reduc- occupational exposures in certain countries undergoing tion in the risk of developing colon cancer (Harriss development, including higher industrial exposures, as and others 2009) and breast cancer (Monninkhof well as a longer duration of exposure due to the common and others 2007). Many studies have suggested that practice of child labor (McCormack and Schuz 2012). regular physical activity additionally reduces the risk Asbestos is among the best known occupational expo- of cancers of the endometrium and prostate (WCRF sures and has been strongly linked to mesothelioma and and AICR 2007). Although this effect seems to be lung cancer (IARC 1987). Because of past exposure, the strongly linked to the impact of physical activity on current proportion of lung cancer attributable to asbestos body weight, the preventive effect of regular exercise is highest in the United States and the United Kingdom, for some cancers seems to act independently of weight ranging from 12 to 18 percent, but the future burden is control (Friedenreich and others 2006; Giovannucci likely to increase in China, India, and Russia, where cur- and others 1995; Pischon and others 2006). In 2010, rent exposure to and use of asbestos are at the highest global physical inactivity was estimated to be related levels worldwide (McCormack and others 2012). to 3.2 and 5.5 percent of cancer deaths in LMICs and Exposure to the ultraviolet component of sunlight HICs, respectively (Lim and others 2012). Motivating is the predominant environmental cause of skin cancer, people to increase physical activity or reduce their body affecting mainly fair skinned persons who live predom- weight has proven difficult; hence, there has been a inantly in HICs (de Vries and Coebergh 2004). A recent more recent emphasis on altering the built environment study predicted that protection from ultraviolet rays may (Diez Roux and others 2007; Giles-Corti and others reduce the incidence of melanoma by 13 to 21 percent in 2005). In Northern Ireland, a greenways urbanization various Western populations over a 40-year period up to project was considered very cost-effective at reducing 2050 (de Vries and others 2012). major chronic diseases, including breast and colon cancer (Dallat and others 2013). Hormonal and Reproductive Factors Hormonal and reproductive factors play an important Environmental Factors role in the etiology of several cancers among women, in Other environmental risk factors, such as household air particular, breast cancer. Established risk factors include pollution, may have a very minor role in causing cancer early age at menarche, older age at first birth, low par- in HICs, but these factors are much more important in ity, and late age at menopause (Beral, Bull, and Reeves LMICs, where they may account for about 2.4 percent 2005; Beral and others 2007; Beral and others 2011; of all cancer deaths (Evans and others 2013; Lim and Reeves and others 2012; Rosner and Colditz 1996). The others 2012). In China, the risk of lung cancer among current changes in these factors are expected to contrib- smokers is three times that of nonsmokers, a risk ratio ute to a 25 percent increase in breast cancer incidence that is much lower than in other countries (Lin and oth- (McCormack and Boffetta 2011). Reversing the current ers 2008), probably as a result of the common practice trend of these factors is not commonly considered fea- of cooking indoors with solid fuel—which significantly sible or desirable; hence, breast cancer prevention must increases lung cancer risk. Some improvements in cook- focus on other, more modifiable factors. These include ing conditions have already taken place and further reductions in BMI, physical inactivity, and alcohol con- improvements could potentially prevent a substantial sumption, and increasing the duration of breast-feeding number of lung cancer deaths in the future. (McCormack and Boffetta 2011). 38 Cancer Summary Vital registration (mortality data) and population- Quantifying the totality of preventable cancers based cancer registries (incidence and survival data) are through the effects of the interventions as described complementary, and having the availability of both to above is a complicated exercise, given imprecision a high standard is ideal. But for many LMICs, national in the estimates and risk factor interactions that are vital statistics are unavailable, partial, or inaccurate. not fully understood. One study has estimated that Where data are available, the coverage and complete- around 35 percent of cancer deaths are avoidable by ness of the mortality data are often poor. It is thus modifying the major cancer risk factors (Danaei and much easier at present to quantify premature cancer others 2005). mortality for higher resource countries than for those Of seven million deaths from cancer worldwide in at the lower end. Efforts to develop and improve vital 2001, 0.76 million preventable deaths were in HICs registration systems are being augmented by a number and 1.67 million were in LMICs. Primary prevention of verbal autopsy studies in countries and communities would most effectively target infection and tobacco, as where many deaths occur at home and without medical well as better nutrition and increased physical activity attention (for example, the Million Death Study in India (Jemal and others 2012; Jha 2009). In more developed [Dikshit and others 2012]). nations, interventions focusing on reducing smok- Population-based cancer registries describe the pres- ing, improving dietary habits, and avoiding excess ent scale and profile of cancer in the community and aid weight are likely to be most effective (Jha and others planning by assessing past and future trends. Knowing 2013; Murray and others 2013; Pirie and others 2013; the size and nature of the cancer problem serves as an Renehan and others 2010). important stimulus to implement and change policy. More effective early detection and better treatment However, although it is feasible to implement PBCRs will further reduce the number of deaths in LMICs even in the lowest income settings (Bray and others 2014), (Farmer and others 2010; Sankaranarayanan and others their availability is limited in most countries undergoing 2011). The large burden of cancer in these regions and development, with only one in five LMICs currently the great potential to prevent them call for rational and predisposed with the necessary population-based data balanced investment in a broad spectrum of cancer for cancer control action. This situation reflects a lack of control activities that includes primary prevention, advocacy for the value of registries historically and the screening, early detection, therapy, follow-up, and pal- subsequent lack of resources and prioritization of cancer liative care (Stewart 2012). In addition, surveillance of data for cancer control among the competing demands risk factor information and cancer data, particularly in on limited health care services (Wild 2014). LMICs, can improve global cancer risk assessment and Nevertheless, positive momentum is building inform national cancer control planning. through the recognition of the role of cancer and other NCDs in hampering human development, changing priorities within countries at the highest political level. PROSPECTS FOR ENHANCING CANCER Importantly and as part of the global monitoring frame- work, WHO member states are being asked to collect SURVEILLANCE SYSTEMS IN LMICS cancer incidence, by type of cancer per 100,000 population, Cancer is already a leading NCD at ages under 70 years one of 25 indicators to monitor progress toward the 25 worldwide and is the most common cause of death in by 25 target, thereby obliging countries to establish and more than 40 HICs. Global efforts are underway at a sustain population-based cancer registries. high political level to combat the rising cancer and NCD Clearly, in aligning registries to cancer control action, burden. The WHO Action Plan for the Global Strategy for technical guidance is needed to empower countries to the Prevention and Control of Noncommunicable Diseases improve their information systems so that they can be provides a road map for 2013–20 that includes a target used for their vital purpose. Several technical and fund- for member states of a global reduction of 25 percent in ing organizations are now working in a cooperative and premature mortality from the four main NCD groups by coordinated manner under the auspices of the Global 2025 (“25 by 25”) (WHO 2011b). Adequate surveillance Initiative for Cancer Registry Development (http://gicr mechanisms at the country level are needed to monitor .iarc.fr) to bring about the needed improvement in the progress. With a greater number of governments adopt- quantity and quality of data on cancer incidence through ing national cancer control plans, attention is increas- investment in PBCRs. There is a longstanding and ongo- ingly turning to population-based cancer surveillance ing commitment at IARC to support the development data as a central component in prioritizing and evaluat- of such registries worldwide, particularly where they ing interventions. are in limited supply and most needed, in most LMICs. The Changing Global Burden of Cancer: Transitions in Human Development and Implications for Cancer Prevention and Control 39 Map 2.2 Implementing the Global Initiative for Cancer Registry through Six IARC Regional Hubs for Cancer Registration Central and Western Asia and Northern Africa Izmir Cancer Registry, Ismir, Turkey Southern, Eastern and South-Eastern Asia Tata Memorial Centre, Mumbai, India Caribbean (Planned) Latin America Instituto Nacional de Cáncer Pacific Islands Buenos Aires, Argentina (Planned) Sub-Saharan Africa African Cancer Registry Network, Oxfordthods • Deliver standardized methods • Allow for flexibility in style to match local context and needs • Inspire innovation • Generate momentum to decrease timelines • Create linkages with partner initiatives to provide bigger impact Source: Global Initiative for Cancer Registry (GICR). Note: IARC = International Agency for Research on Cancer. The presence of six recently established IARC regional habits and the expanding tobacco markets in LMICs hubs for cancer registration (map 2.2) across LMIC may have major consequences on the future burden of regions is providing the necessary localized support and lung and other tobacco-related cancers in the decades training for cancer registries in the respective countries, that follow. as well as strengthening registry networking in each The global trends therefore place special empha- region to improve data standards and develop joint sis on the need for better cancer control in LMICs research activities. and the need for expanding coverage and quality of population-based cancer registration systems in each country. Cancer statistics should also serve as a cata- CONCLUSIONS lyst for further research on human inequality and the scale and profile of cancer from global and regional This chapter has examined the increasing global impor- perspectives, for better determination of how and why tance of cancer among the major NCDs, highlighted the macroeconomic determinants influence cancer indica- extent of the ongoing demographic and epidemiologic tors. It is imperative that public health clinicians and transitions, described the key risk factors involved, and cancer control specialists are alerted to the increasing highlighted the extent to which cancer deaths may be magnitude of cancer incidence and mortality world- avoided. wide. This analysis serves to highlight the need for Changing fertility patterns and increasing life expec- global action to reduce the increasing incidence and tancy are leading to major changes in the scale of mortality burden from cancer and alleviate suffering the cancer problem worldwide for populations at all among cancer patients. levels of human development. A cancer transition is clearly underway in countries transiting to high lev- els of income and development whereby reductions in infection-related cancers are offset by concomitant ACKNOWLEDGMENT increases in cancers linked to a “Westernization” of The authors thank Mathieu Laversanne, IARC, for statistical lifestyle, including “risky” behavior. Changes in tobacco analysis and generation of the graphics. 40 Cancer NOTES Bleyer, A., and H. G. Welch. 2012. “Effect of Three Decades of Screening Mammography on Breast-Cancer Incidence.” World Bank income classifications as of July 2014 are as follows, New England Journal of Medicine 367 (21): 1998–2005. based on estimates of gross national income (GNI) per capita doi:10.1056/Nejmoa1206809. for 2013: Boffetta, P., M. Hashibe, C. La Vecchia, W. Zatonski, and J. Rehm. 2006. “The Burden of Cancer Attributable to • Low-income countries (LICs): US$1,045 or less Alcohol Drinking.” International Journal of Cancer 119 (4): • Middle-income countries (MICs) are subdivided: 884–87. doi:10.1002/Ijc.21903. a) lower-middle-income: US$1,046–US$4,125 Boffetta, P., M. Tubiana, C. Hill, M. Boniol, A. Aurengo, and b) upper-middle-income (UMICs): US$4,126–US$12,745 others. 2009. “The Causes of Cancer in France.” Annals • High-income countries (HICs): US$12,746 or more of Oncology 20 (3): 550–55. doi:Mdn597 [Pii]10.1093 /Annonc/Mdn597. Except for figure 2.9 and estimates presented in the section Bray, F. 2014. “Transitions in Human Development and the Risk Factors, Interventions and Policies, and Potential Impacts Global Cancer Burden.” In World Cancer Report 2014, on the Burden of Cancer, the maps and figures in this chapter edited by C. P. Wild and B. A. Stewart. Lyon, France: are based on incidence and mortality estimates for ages 0 to 69, International Agency for Research on Cancer. consistent with reporting in all DCP3 volumes. Global cancer Bray, F., A. Jemal, N. Grey, J. Ferlay, and D. Forman. 2012. statistics are estimates for the year 2012 and have been pro- “Global Cancer Transitions According to the Human vided by the International Agency for Research on Cancer from Development Index (2008–2030): A Population-Based its GLOBOCAN 2012 database. Observable population-based Study.” The Lancet Oncology 13 (8): 790–801. doi:10.1016 data were derived from Cancer Incidence in Five Continents, /S1470-2045(12)70211-5. 10th edition and for trends over time from CI5 Plus (http:// Bray, F., J. Lortet-Tieulent, J. Ferlay, D. Forman, and A. Auvinen. ci5.iarc.fr/CI5plus/Default.aspx). The discussion of burden 2010. “Prostate Cancer Incidence and Mortality Trends in (including risk factors), however, includes all ages unless 37 European Countries: An Overview.” European Journal otherwise noted. Interventions also apply to all age groups, of Cancer 46 (17): 3040–52. doi:S0959-8049(10)00878-6 except where age ranges or cutoffs are specified. [Pii]10.1016/J.Ejca.2010.09.013. Bray, F., P. McCarron, and D. M. Parkin. 2004. “The Changing Global Patterns of Female Breast Cancer Incidence and Mortality.” Breast Cancer Research 6 (6): 229–39. doi:Bcr932 REFERENCES [Pii]10.1186/Bcr932. Agudo, A., C. Bonet, N. Travier, C. A. Gonzalez, P. Vineis, and Bray, F., A. Znaor, P. Cueva, A. Korir, R. Swaminathan, and others. 2012. “Impact of Cigarette Smoking on Cancer others. 2014. “Planning and Developing Population-Based Risk in the European Prospective Investigation into Cancer Registration in Low- and Middle-Income Settings.” Cancer and Nutrition Study.” Journal of Clinical Oncology Technical Publication 43, International Agency for Research 30 (36): 4550–57. doi:JCO.2011.41.0183 [Pii] 10.1200 on Cancer, Lyon, France. /JCO.2011.41.0183. Brown, I. J., I. Tzoulaki, V. Candeias, and P. Elliott. 2009. “Salt Beral, V., D. Bull, J. Green, and G. Reeves. 2007. “Ovarian Intakes around the World: Implications for Public Health.” Cancer and Hormone Replacement Therapy in the Million International Journal of Epidemiology 38 (3): 791–813. Women Study.” The Lancet 369 (9574): 1703–10. doi:S0140 doi:Dyp139 [Pii] 10.1093/Ije/Dyp139. -6736(07)60534-0 [Pii] 10.1016/S0140-6736(07)60534-0. Center, M. M., A. Jemal, R. A. Smith, and E. Ward. 2009. Beral, V., D. Bull, and G. Reeves. 2005. “Endometrial Cancer “Worldwide Variations in Colorectal Cancer.” CA: A and Hormone-Replacement Therapy in the Million Cancer Journal for Clinicians 59 (6): 366–78. doi:59/6/366 Women Study.” The Lancet 365 (9470): 1543–51. doi:S0140 [Pii]10.3322/Caac.20038. -6736(05)66455-0 [Pii] 10.1016/S0140-6736(05)66455-0. Center, M. M., A. Jemal, and E. Ward. 2009. “International Beral, V., G. Reeves, D. Bull, and J. Green. 2011. “Breast Cancer Trends in Colorectal Cancer Incidence Rates.” Cancer Risk in Relation to the Interval between Menopause and Epidemiology Biomarkers & Prevention 18 (6): 1688–94. Starting Hormone Therapy.” Journal of the National Cancer doi:18/6/1688 [Pii]10.1158/1055-9965.EPI-09-0090. Institute 103 (4): 296–305. doi:Djq527 [Pii] 10.1093/Jnci Chang, M. H., C. J. Chen, M. S. Lai, H. M. Hsu, T. C. Wu, /Djq527. and others. 1997. “Universal Hepatitis B Vaccination in Bergstrom, A., P. Pisani, V. Tenet, A. Wolk, and H. O. Adami. Taiwan and the Incidence of Hepatocellular Carcinoma 2001. “Overweight as an Avoidable Cause of Cancer in Children. Taiwan Childhood Hepatoma Study Group.” in Europe.” International Journal of Cancer 91 (3): New England Journal of Medicine 336 (26): 1855–59. 421–30. doi:10.1002/1097-0215(200002)9999:9999<::AID doi:10.1056/NEJM199706263362602. -IJC1053>3.0.CO;2-T [Pii]. Chokunonga, E., M. Z. Borok, Z. M. Chirenje, A. M. Nyakabau, Beutels, P. 2001. “Economic Evaluations of Hepatitis and D. M. Parkin. 2013. “Trends in the Incidence of Cancer B Immunization: A Global Review of Recent Studies in the Black Population of Harare, Zimbabwe 1991–2010.” (1994–2000).” Health Economics 10 (8): 751–74. doi:10.1002 International Journal of Cancer 133 (3): 721–29. doi:10.1002 /Hec.625 [Pii]. /Ijc.28063. The Changing Global Burden of Cancer: Transitions in Human Development and Implications for Cancer Prevention and Control 41 Cogliano, V. J., R. Baan, K. Straif, Y. Grosse, B. Lauby- Satellite Imagery.” Environmental Research 120: 33–42. Secretan, and others. 2011. “Preventable Exposures doi:S0013-9351(12)00240-X[Pii]10.1016/J.Envres.2012.08.005. Associated with Human Cancers.” Journal of the National Ezzati, M., S. J. Henley, A. D. Lopez, and M. J. Thun. 2005. “Role Cancer Institute 103 (24): 1827–39. doi:Djr483 [Pii] of Smoking in Global and Regional Cancer Epidemiology: 10.1093/Jnci/Djr483. Current Patterns and Data Needs.” International Journal of Dallat, M. A., I. Soerjomataram, R. F. Hunter, M. A. Tully, Cancer 116 (6): 963–71. doi:10.1002/Ijc.21100. K. J. Cairns, and others. 2013. “Urban Greenways Have Farmer, P., J. Frenk, F. M. Knaul, L. N. Shulman, G. Alleyne, and the Potential to Increase Physical Activity Levels Cost- others. 2010. “Expansion of Cancer Care and Control in Effectively.” European Journal of Public Health 24 (2): 190–95. Countries of Low and Middle Income: A Call to Action.” The doi:Ckt035[Pii]10.1093/Eurpub/Ckt035. Lancet 376 (9747): 1186–93. doi:S0140-6736(10)61152-X Danaei, G., S. Vander Hoorn, A. D. Lopez, C. J. Murray, [Pii] 10.1016/S0140-6736(10)61152-X. and M. Ezzati. 2005. “Causes of Cancer in the World: Ferlay, J., I. Soerjomataram, M. Ervik, R. Dikshit, S. Eser, Comparative Risk Assessment of Nine Behavioural and and others. 2013. GLOBOCAN 2012 v1.0. In Cancer Environmental Risk Factors.” The Lancet 366 (9499): Incidence and Mortality Worldwide: IARC CancerBase 1784–93. doi:S0140-6736(05)67725-2 [Pii]10.1016 No. 11 [Internet]. Lyon, France: International Agency for /S0140-6736(05)67725-2. Research on Cancer. de Martel, C., J. Ferlay, S. Franceschi, J. Vignat, F. Bray, and Ferlay, J., I. Soerjomataram, R. Dikshit, S. Eser, C. Mathers, and others. 2012. “Global Burden of Cancers Attributable to others. 2015. “Cancer Incidence and Mortality Worldwide: Infections in 2008: A Review and Synthetic Analysis.” The Sources, Methods and Major Patterns in GLOBOCAN Lancet Oncology 13 (6): 607–15. doi:S1470-2045(12)70137-7 2012.” International Journal of Cancer 136 (5): E359–86. [Pii]10.1016/S1470-2045(12)70137-7. doi:10.1002/Ijc.29210. de Vries, E., M. Arnold, E. Altsitsiadis, M. Trakatelli, B. Hinrichs, Forman, D., C. de Martel, C. J. Lacey, I. Soerjomataram, and others. 2012. “Potential Impact of Interventions Resulting J. Lortet-Tieulent, and others. 2012. “Global Burden of in Reduced Exposure to Ultraviolet (UV) Radiation (UVA Human Papillomavirus and Related Diseases.” Vaccine and UVB) on Skin Cancer Incidence in Four European 30 (Suppl. 5): F12–23. doi:S0264-410X(12)01080-8 [Pii] Countries, 2010–2050.” British Journal of Dermatology 167 10.1016/J.Vaccine.2012.07.055. (Suppl. 2): 53–62. doi:10.1111/J.1365-2133.2012.11087.X. Friedenreich, C., T. Norat, K. Steindorf, M. C. Boutron- de Vries, E., and J. W. Coebergh. 2004. “Cutaneous Malignant Ruault, T. Pischon, and others. 2006. “Physical Activity Melanoma in Europe.” European Journal of Cancer and Risk of Colon and Rectal Cancers: The European 40 (16): 2355–66. doi:S0959804904004678 [Pii]10.1016/J Prospective Investigation into Cancer and Nutrition.” .Ejca.2004.06.003. Cancer Epidemiological Biomarkers & Prevention 15 (12): Diez Roux, A. V., K. R. Evenson, A. P. McGinn, D. G. Brown, 2398–407. doi: 15/12/2398 [Pii] 10.1158/1055-9965 L. Moore, and others. 2007. “Availability of Recreational .EPI-06-0595. Resources and Physical Activity in Adults.” American Journal Giles-Corti, B., M. H. Broomhall, M. Knuiman, C. Collins, of Public Health 97 (3): 493–99. doi:AJPH.2006.087734 [Pii] K. Douglas, and others. 2005. “Increasing Walking: How 10.2105/AJPH.2006.087734. Important Is Distance to, Attractiveness, and Size of Public Dikshit, R., P. C. Gupta, C. Ramasundarahettige, V. Gajalakshmi, Open Space?” American Journal of Preventive Medicine 28 L. Aleksandrowicz, and others. 2012. “Cancer Mortality in (2 Suppl. 2): 169–76. doi:S0749-3797(04)00298-3 [Pii] India: A Nationally Representative Survey.” The Lancet 10.1016/J.Amepre.2004.10.018. 379 (9828): 1807–16. doi:10.1016/S0140-6736(12)60358-4. Giovannucci, E., A. Ascherio, E. B. Rimm, G. A. Colditz, Doll, R. 1990. “Are We Winning the Fight against Cancer? An M. J. Stampfer, and others. 1995. “Physical Activity, Obesity, Epidemiological Assessment. EACR-Muhlbock Memorial and Risk for Colon Cancer and Adenoma in Men.” Annals Lecture.” European Journal of Cancer 26 (4): 500–08. of Internal Medicine 122 (5): 327–34. Doll, R., P. Payne, and A. J. H. Waterhouse. 1966. Cancer Goldie, S. J., M. O’Shea, N. G. Campos, M. Diaz, S. Sweet, Incidence in Five Continents: A Technical Report. New York: and others. 2008. “Health and Economic Outcomes of Springer. HPV 16,18 Vaccination in 72 GAVI-Eligible Countries.” Doll, R., and R. Peto. 1981. “The Causes of Cancer: Quantitative Vaccine 26 (32): 4080–93. doi:S0264-410X(08)00493 Estimates of Avoidable Risks of Cancer in the United States -3[Pii]10.1016/J.Vaccine.2008.04.053. Today.” Journal of the National Cancer Institute 66 (6): Goldstein, S. T., F. Zhou, S. C. Hadler, B. P. Bell, E. E. Mast, and 1191–308. others. 2005. “A Mathematical Model to Estimate Global Doll, R., R. Peto, J. Boreham, and I. Sutherland. 2004. “Mortality Hepatitis B Disease Burden and Vaccination Impact.” in Relation to Smoking: 50 Years’ Observations on Male International Journal of Epidemiology 34 (6): 1329–39. British Doctors.” British Medical Journal 328 (7455): 1519. doi:Dyi206 [Pii]10.1093/Ije/Dyi206. doi: 10.1136/Bmj.38142.554479.AE Bmj.38142.554479 Hanley, J. A. 2001. “A Heuristic Approach to the Formulas for .AE [Pii]. Population Attributable Fraction.” Journal of Epidemiology Evans, J., A. Van Donkelaar, R. V. Martin, R. Burnett, and Community Health 55 (7): 508–14. D. G. Rainham, and others. 2013. “Estimates of Global Harper, D. M., E. L. Franco, C. Wheeler, D. G. Ferris, Mortality Attributable to Particulate Air Pollution Using D. Jenkins, and others. 2004. “Efficacy of a Bivalent L1 42 Cancer Virus-Like Particle Vaccine in Prevention of Infection Factors and Risk Factor Clusters in 21 Regions, 1990–2010: with Human Papillomavirus Types 16 and 18 in Young A Systematic Analysis for the Global Burden of Disease Women: A Randomised Controlled Trial.” The Lancet Study 2010.” The Lancet 380 (9859): 2224–60. doi:10.1016 364 (9447): 1757–65. doi:S0140673604173984 [Pii]10.1016 /S0140-6736(12)61766-8. /S0140-6736(04)17398-4. Lin, H. H., M. Murray, T. Cohen, C. Colijn, and M. Ezzati. Harriss, D. J., G. Atkinson, A. Batterham, K. George, 2008. “Effects of Smoking and Solid-Fuel Use on COPD, N. T. Cable, and others. 2009. “Lifestyle Factors and Lung Cancer, and Tuberculosis in China: A Time-Based, Colorectal Cancer Risk (2): A Systematic Review and Multiple Risk Factor, Modelling Study.” The Lancet 372 Meta-Analysis of Associations with Leisure-Time Physical (9648): 1473–83. doi:S0140-6736(08)61345-8 [Pii] 10.1016 Activity.” Colorectal Disease 11 (7): 689–701. doi:CDI1767 /S0140-6736(08)61345-8. [Pii] 10.1111/J.1463-1318.2009.01767.X. Lortet-Tieulent, J., E. Renteria, L. Sharp, E. Weiderpass, Herrero, R., W. Quint, A. Hildesheim, P. Gonzalez, L. Struijk, H. Comber, and others. 2013. “Convergence of Decreasing and others. 2013. “Reduced Prevalence of Oral Human Male and Increasing Female Incidence Rates in Major Papillomavirus (HPV) 4 Years after Bivalent HPV Tobacco-Related Cancers in Europe in 1988–2010.” Vaccination in A Randomized Clinical Trial in Costa Rica.” European Journal of Cancer. doi: S0959-8049(13)00952-0 Plos One 8 (7): E68329. doi:10.1371/Journal.Pone.0068329 [Pii] 10.1016/J.Ejca.2013.10.014. PONE-D-12-40404 [Pii]. Lozano, R., M. Naghavi, K. Foreman, S. Lim, K. Shibuya, and IARC (International Agency for Research on Cancer). 1987. others. 2012. “Global and Regional Mortality from 235 Overall Evaluations of Carcinogenicity: An Updating of Causes of Death for 20 Age Groups in 1990 and 2010: IARC Monographs 1–440. Monographs on the Evaluation of A Systematic Analysis for the Global Burden of Disease Carcinogenic Risks to Humans, IARC, Lyon, France. Study 2010.” The Lancet 380 (9859): 2095–128. doi:S0140 ———. 2004. “Tobacco Smoke and Involuntary Smoking.” -6736(12)61728-0 [Pii]10.1016/S0140-6736(12)61728-0. Monographs on the Evaluation of Carcinogenic Risks to Makela, P., K. Tryggvesson, and I. Rossow. 2002. “Who Drinks Humans, Volume 83, IARC, Lyon, France. More or Less When Policies Change? The Evidence from ———. 2010. “Alcohol Consumption and Ethyl Carbamate.” 50 Years of Nordic Studies.” In The Effects of Nordic Alcohol Monographs on the Evaluation of Carcinogenic Risks to Policies: Analyses of Changes in Control Systems, edited by Humans, IARC, Lyon, France. R. Room. Helsinki: Nordic Council for Alcohol and Drug Jemal, A., F. Bray, D. Forman, M. O’Brien, J. Ferlay, and Research. others. 2012. “Cancer Burden in Africa and Opportunities Martin-Moreno, J. M., I. Soerjomataram, and G. Magnusson. for Prevention.” Cancer 118 (18): 4372–84. doi:10.1002 2008. “Cancer Causes and Prevention: A Condensed /Cncr.27410. Appraisal in Europe in 2008.” European Journal of Cancer Jemal, A., M. M. Center, C. Desantis, and E. M. Ward. 44 (10): 1390–403. doi:S0959-8049(08)00090-7 [Pii] 2010. “Global Patterns of Cancer Incidence and Mortality 10.1016/J.Ejca.2008.02.002. Rates and Trends.” Cancer Epidemiology Biomarkers & McCormack, V. A., and P. Boffetta. 2011. “Today’s Lifestyles, Prevention 19 (8): 1893–907. doi:1055-9965.EPI-10-0437 Tomorrow’s Cancers: Trends in Lifestyle Risk Factors for [Pii] 10.1158/1055-9965.EPI-10-0437. Cancer in Low- and Middle-Income Countries.” Annals Jha, P. 2009. “Avoidable Global Cancer Deaths and Total Deaths of Oncology 22 (11): 2349–57. doi:Mdq763 [Pii] 10.1093 from Smoking.” Nature Reviews Cancer 9 (9): 655–64. /Annonc/Mdq763. doi:Nrc2703 [Pii] 10.1038/Nrc2703. McCormack, V., J. Peto, G. Byrnes, K. Straif, and P. Boffetta. Jha, P., C. Ramasundarahettige, V. Landsman, B. Rostron, 2012. “Estimating the Asbestos-Related Lung Cancer M. Thun, and others. 2013. “21st-Century Hazards of Burden from Mesothelioma Mortality.” British Journal Smoking and Benefits of Cessation in the United States.” of Cancer 106 (3): 575–84. doi:Bjc2011563 [Pii] 10.1038 New England Journal of Medicine 368 (4): 341–50. doi:0.1056 /Bjc.2011.563. /Nejmsa1211128. McCormack, V. A., and J. Schuz. 2012. “Africa’s Growing Kane, M. A., B. Serrano, S. De Sanjose, and S. Wittet. 2012. Cancer Burden: Environmental and Occupational “Implementation of Human Papillomavirus Immunization Contributions.” Cancer Epidemiology 36 (1): 1–7. doi:S1877 in the Developing World.” Vaccine 30 (Suppl. 5): -7821(11)00146-9 [Pii] 10.1016/J.Canep.2011.09.005. F192–200. doi:S0264-410X(12)00956-5 [Pii] 10.1016/J Monninkhof, E. M., S. G. Elias, F. A. Vlems, I. Van Der Tweel, .Vaccine.2012.06.075. A. J. Schuit, and others. 2007. “Physical Activity and Breast Karim-Kos, H. E., E. de Vries, I. Soerjomataram, V. Lemmens, Cancer: A Systematic Review.” Epidemiology 18 (1): 137–57. S. Siesling, and J. W. Coebergh. 2008. “Recent Trends of doi:10.1097/01.Ede.0000251167.75581.98. Cancer in Europe: A Combined Approach of Incidence, Muir, C. S., J. F. Fraumeni, Jr., and R. Doll. 1994. “The Survival and Mortality for 17 Cancer Sites Since the 1990s.” Interpretation of Time Trends.” Cancer Surveys 19–20: 5–21. European Journal of Cancer 44 (10): 1345–89. doi: 10.1016/j. Murray, C. J., J. Abraham, M. K. Ali, M. Alvarado, C. Atkinson, ejca.2007.12.015. and others. 2013. “The State of US Health, 1990–2010: Lim, S. S., T. Vos, A. D. Flaxman, G. Danaei, K. Shibuya, Burden of Diseases, Injuries, and Risk Factors.” Journal and others. 2012. “A Comparative Risk Assessment of of the American Medical Association 310 (6): 591–606. Burden of Disease and Injury Attributable to 67 Risk doi:1710486 [Pii] 10.1001/Jama.2013.13805. The Changing Global Burden of Cancer: Transitions in Human Development and Implications for Cancer Prevention and Control 43 Ng, M., M. K. Freeman, T. D. Fleming, M. Robinson, to the Comparative Risk Assessment for the 2010 Global L. Dwyer-Lindgren, and others. 2014. “Smoking Prevalence Burden of Disease Study.” Addiction 108 (5): 912–22. and Cigarette Consumption in 187 Countries, 1980–2012.” doi:10.1111/Add.12112. Journal of the American Medical Association 311 (2): 183–92. Soerjomataram, I., E. Pukkala, H. Brenner, and J. W. Coebergh. doi:10.1001/Jama.2013.284692. 2008. “On the Avoidability of Breast Cancer in Industrialized Parkin, D. M. 2006. “The Evolution of the Population-Based Societies: Older Mean Age at First Birth as an Indicator Cancer Registry.” Nature Reviews Cancer 6 (8): 603–12. of Excess Breast Cancer Risk.” Breast Cancer Research and doi:Nrc1948 [Pii] 10.1038/Nrc1948. Treatment 111 (2): 297–302. doi:10.1007/S10549-007-9778-2. ———. 2011. “Cancers Attributable to Dietary Factors in the Stevens, G. A., G. M. Singh, Y. Lu, G. Danaei, J. K. Lin, and UK in 2010. IV. Salt.” British Journal of Cancer 105 (Suppl. 2): others. 2012. “National, Regional, and Global Trends in S31–33. doi:Bjc2011480 [Pii] 10.1038/Bjc.2011.480. Adult Overweight and Obesity Prevalences.” Population Parkin, D. M., F. Bray, J. Ferlay, and A. Jemal. 2014. “Cancer in Health Metrics 10 (1): 22. doi:1478-7954-10-22 [Pii] Africa 2012.” Cancer Epidemilogy Biomarkers & Prevention 10.1186/1478-7954-10-22. 23 (6): 953–66. doi:10.1158/1055-9965.EPI-14-0281. Stewart, B. W. 2012. “Priorities for Cancer Prevention: Lifestyle Pirie, K., R. Peto, G. K. Reeves, J. Green, and V. Beral. 2013. “The Choices versus Unavoidable Exposures.” The Lancet 21st Century Hazards of Smoking and Benefits of Stopping: Oncology 13 (3): E126–33. doi:S1470-2045(11)70221-2 A Prospective Study of One Million Women in the UK.” The [Pii]10.1016/S1470-2045(11)70221-2. Lancet 381 (9861): 133–41. doi:S0140-6736(12)61720-6 Talley, M. B. 2010. Examining the Impact of Development, [Pii]10.1016/S0140-6736(12)61720-6. Tobacco Taxation, and Tobacco Prices on Global Adult Male Pischon, T., P. H. Lahmann, H. Boeing, C. Friedenreich, Smoking Prevalence. Altanta, GA: Georgia Institute of T. Norat, and others. 2006. “Body Size and Risk of Technology, Georgia State University. Colon and Rectal Cancer in the European Prospective Thun, M., R. Peto, J. Boreham, and A. D. Lopez. 2012. “Stages Investigation into Cancer and Nutrition (EPIC).” Journal of of the Cigarette Epidemic on Entering Its Second Century.” the National Cancer Institute 98 (13): 920–31. doi:98/13/920 Tobacco Control 21 (2): 96–101. doi:Tobaccocontrol-2011 [Pii]10.1093/Jnci/Djj246. -050294 [Pii] 10.1136/Tobaccocontrol-2011-050294. Raaschou-Nielsen, O., Z. J. Andersen, R. Beelen, E. Samoli, UNDP (United Nations Develpoment Programme). 2009. M. Stafoggia, and others. 2013. “Air Pollution and Lung Cancer Human Development Report 2009: Overcoming Barriers: Incidence in 17 European Cohorts: Prospective Analyses Human Mobility and Development. New York: UNDP. from the European Study of Cohorts for Air Pollution Effects Vaccarella, S., J. Lortet-Tieulent, M. Plummer, S. Franceschi, (ESCAPE).” The Lancet Oncology 14 (9): 813–22. doi:S1470 and F. Bray. 2013. “Worldwide Trends in Cervical Cancer -2045(13)70279-1 [Pii]10.1016/S1470-2045(13)70279-1. Incidence: Impact of Screening against Changes in Reeves, G. K., K. Pirie, J. Green, D. Bull, and V. Beral. 2012. Disease Risk Factors.” European Journal of Cancer 49 (15): “Comparison of the Effects of Genetic and Environmental 3262–73. doi:S0959-8049(13)00358-4 [Pii]10.1016/J.Ejca Risk Factors on In Situ and Invasive Ductal Breast Cancer.” .2013.04.024. International Journal of Cancer 131 (4): 930–37. doi:10.1002 Wabinga, H. R., P. M. Amulen, C. Okello, L. Mbus, and /Ijc.26460. D. M. Parkin. 2014. “Trends in the Incidence of Cancer Rehm, J., and K. D. Shield. 2013. “Global Alcohol-Attributable in Kampala, Uganda 1991–2010.” International Journal of Deaths from Cancer, Liver Cirrhosis, and Injury in 2010.” Cancer 135 (2): 432–9. doi: 10.1002/ijc.28661. http://www Alcohol Research 35 (2): 174–83. .ncbi.nlm.nih.gov/pubmed/24615279. Renehan, A. G., I. Soerjomataram, M. Tyson, M. Egger, WCRF (World Cancer Research Fund) and AICR (American M. Zwahlen, and others. 2010. “Incident Cancer Burden Institute for Cancer Research). 2007. Food, Nutrition, Attributable to Excess Body Mass Index in 30 European Physical Activity, and the Prevention of Cancer: A Global Countries.” International Journal of Cancer 126 (3): Perspective. Washington, DC: WCRF and AICR. 692–702. doi:10.1002/Ijc.24803. WHO (World Health Organization). 2008. Report on the Rockhill, B., B. Newman, and C. Weinberg. 1998. “Use and Global Tobacco Epidemic, 2008: The MPOWER Package. Misuse of Population Attributable Fractions.” American Geneva: WHO. Journal of Public Health 88 (1): 15–19. ———. 2011a. Global Status Report on Alcohol and Health. Rosner, B., and G. A. Colditz. 1996. “Nurses’ Health Study: Log- Geneva: WHO. Incidence Mathematical Model of Breast Cancer Incidence.” ———. 2011b. Global Status Report on Non-Communicable Journal of the National Cancer Institute 88 (6): 359–64. Diseases 2010. Geneva: WHO. Sankaranarayanan, R., R. Swaminathan, K. Jayant, and ———. 2013. “Global Health Observatory Data Repository.” H. Brenner. 2011. “An Overview of Cancer Survival in WHO, Geneva. http://apps.who.int/gho/data/node.main Africa, Asia, the Caribbean and Central America: The Case .a859?lang=en. for Investment in Cancer Health Services.” IARC Scientific Wild, C. P. 2014. “Foreword.” In Planning and Developing Publications 162: 257–91. Population-Based Cancer Registration in Low- and Shield, K. D., M. Rylett, G. Gmel, T. A. Kehoe-Chan, and Middle-Income Settings, edited by F. Bray, A. Znaor, J. Rehm. 2013. “Global Alcohol Exposure Estimates by P. Cueva, A. Korir, R. Swaminathan, and others. Lyon, Country, Territory and Region for 2005: A Contribution France: IARC. 44 Cancer Chapter 3 Breast Cancer Benjamin O. Anderson, Joseph Lipscomb, Raul H. Murillo, and David B. Thomas INTRODUCTION LMICs, because of changes in their childbearing patterns and their adoption of Western lifestyles (Parkin and Disparities in Global Breast Cancer Outcomes1 Fernandez 2006; Porter 2008). The number of young Breast cancer is the world’s most common cancer lives lost is even more disproportionate than the total among women, and it is the most likely reason that a number; in 2010, breast cancer killed 68,000 women ages woman will die from cancer (maps 3.1 and 3.2). Breast 15–49 years in LMICs, compared with 26,000 in this age cancer is becoming an increasingly urgent problem in range in HICs (Forouzanfar and others 2011). low- and middle-income countries (LMICs), where HICs have made tremendous progress in improv- incidence rates, historically low, have been rising by ing outcomes (figure 3.2). Mortality rates, which had as much as 5 percent per year (Bray and others 2013). been essentially unchanged in the United States for High-income countries (HICs) report the highest breast the five decades between 1930 and 1980, have dropped cancer incidence rates (figure 3.1), but these countries nearly 2 percent each year since 1990 (Jemal and oth- have also made the most progress in improving out- ers 2009). Similar reductions have occurred in other comes (Jemal and others 2002). In 2010, the majority HICs, such as Norway (Kalager and others 2010). The of the 425,000 global breast cancer deaths occurred in improvements are attributable to early detection by LMICs, and that percentage is expected to grow (Parkin screening, combined with timely and effective treatment and Fernandez 2006). (Weir and others 2003). Randomized trials of screening Breast cancer fatality rates are inversely correlated mammography in the 1970s and 1980s demonstrated with per capita gross domestic product (GDP) (Greenlee that early detection leads to stage shifting, improved and others 2000). Historically, breast cancer incidence survival, and reduced mortality (Chu, Smart, and had been low in LMICs, but these rates are rising dis- Tarone 1988). Endocrine therapy for estrogen receptor proportionately at the same time that mortality rates (ER)–positive cancers and cytotoxic chemotherapy for are continuing to rise or remain high (figure 3.2). The ER-negative cancers improve survival among early and aging of current global population means that nearly locally advanced breast cancers (Clarke 2006; Perloff and 50 percent more women will develop and die from others 1988). breast cancer in 2020 than in 2002. This estimate does Low survival rates in LMICs are largely attribut- not take into account the likely increases in age-specific able to late-stage presentation and limited diagnostic breast cancer incidence and mortality rates, especially and treatment capacities (Hisham and Yip 2003). In among recent birth cohorts and among urban women in India, 50–70 percent of cases are diagnosed with locally Corresponding author: Benjamin O. Anderson, MD, banderso@u.washington.edu 45 Map 3.1 Global Breast Cancer Incidence in Women in 2012 IBRD 41380 | FEBRUARY 2015 Breast Cancer Women 0–69 years ASR (World) Incidence Greater than 57.4 41.1–57.4 31.1–41.0 22.4–31.0 Less than 22.4 No data Source: Ferlay and others 2013. Note: Values are estimated ASR per 100,000 women. ASR = age-standardized rate. Map 3.2 Global Breast Cancer Mortality in Women in 2012 Breast Cancer Women 0–69 years ASR (World) Mortality Greater than 15.0 12.1–15.0 10.1–12.1 8.3–10.1 Less than 8.3 No data Source: Ferlay and others 2013. Note: Values are estimated ASR per 100,000 women. ASR = age-standardized rate. 46 Cancer Figure 3.1 Trends in Age-Standardized Incidence Rates in Women, Selected Countries, 1975–2010 a. b. c. 100 100 100 Age-standardized (world) incidence rate per 100,000, 0–69 years Age-standardized (world) incidence rate per 100,000, 0–69 years Age-standardized (world) incidence rate per 100,000, 0–69 years 80 80 80 60 60 60 40 40 40 20 20 20 0 0 0 80 10 80 90 00 00 00 10 10 80 90 90 75 85 95 75 75 95 85 05 05 85 05 95 19 20 19 19 20 20 20 20 20 19 19 19 19 19 19 19 19 19 19 20 20 19 20 19 Calendar year Calendar year Calendar year France Denmark Philippines Singapore United States Canada United Finland China Japan New Zealand Australia Kingdom India Thailand Colombia Costa Rica Spain Slovak Republic Sources: CI5 Plus (http://ci5.iarc.fr/CI5plus/Default.aspx); and Ferlay and others 2013. Note: Values are age-standardized rates of breast cancer incidence per 100,000 women, for the world population structure. advanced or metastatic disease (Chopra 2001); the same Risk Factors and Risk Reduction Strategies was true of 38 percent of European and 30 percent of Breast cancer risk increases with some factors that can- American cases in the early 1990s (Sant and others not be modified, such as age, genetic and familial risks, 2004). Accordingly, efforts to promote early detec- younger age at menarche, and older age at menopause; tion followed by appropriate treatment are essential some factors that are somewhat modifiable, including components of population-based breast cancer control delayed childbearing, avoidance of lactation, radiation strategies. exposure, and use of hormone replacement therapy; The two strategies of early detection and adjuvant and some that are more modifiable, including body systemic therapy are synergistic and mutually dependent mass index, sedentary lifestyle, and moderate to high for improving outcomes; early detection only works if it levels of alcohol use (McTiernan, Porter, and Potter can be followed by prompt therapy. Mathematical mod- 2008). Modifying these behaviors to the extent possible, eling suggests that between 28 percent and 65 percent although not proven in clinical trials to reduce risk, is of breast cancer mortality reduction can be attributed likely to be beneficial, can be good for general health to early detection; the balance is due to pharmacother- and noncommunicable disease prevention, and may apy (Berry and others 2005). The interdependence of be of interest to policy makers in LMICs. However, early detection and treatment underscores the essential because most of these factors elevate risk only mar- role of guidelines for administering this comprehensive ginally, even successful reduction of them may only strategy in limited-resource settings to shift morbidity have a small effect on overall risk. Some risk factors and mortality rates at the global level. are not amenable to change; others are associated with Breast Cancer 47 Figure 3.2 Trends in Age-Standardized Breast Cancer Mortality Rates in Women, Selected Countries, 1975–2010 a. b. c. 25 25 25 Age-standardized (world) mortality rate per 100,000, 0–69 years Age-standardized (world) mortality rate per 100,000, 0–69 years Age-standardized (world) mortality rate per 100,000, 0–69 years 20 20 20 15 15 15 10 10 10 5 5 5 0 0 0 90 80 90 90 85 95 00 10 05 80 75 80 85 95 00 10 85 95 00 10 05 75 05 75 19 19 19 19 19 19 20 20 20 19 19 19 19 19 20 20 19 19 20 20 20 19 20 19 Calendar year Calendar year Calendar year United Kingdom Denmark Israel Singapore New Zealand Canada France Finland Japan Republic of Korea United States Australia Slovak Spain Costa Rica Colombia Republic Sources: Ferlay and others 2013; WHO Mortality Database (http://www.who.int/healthinfo/statistics/mortality_rawdata/en/index.html). Note: Values are age-standardized rates of breast cancer mortality per 100,000 women, for the world population structure. desirable outcomes, such as the education of women, Breast Health Global Initiative which has major societal benefits even if it tends to Evidence-based, economically feasible, and culturally delay childbearing. appropriate guidelines that can be used in settings with limited resources to improve outcomes have been devel- oped by the Breast Health Global Initiative (BHGI), RESOURCE-STRATIFIED GUIDELINES an international health alliance established in 2002. Need for Cancer Care Guidelines Explicitly The BHGI has held five global summits addressing key Addressing Resource Limitations aspects of care: Early detection and comprehensive treatment together can improve outcomes. In HICs and upper-middle- • Health care disparities, Seattle, Washington, 2002 income countries, the guidelines for achieving these (Anderson and others 2003) goals are defined, updated, and disseminated (Morrow • Evidence-based resource allocation, Bethesda, and others 2002; Smith 2000; Theriault and others Maryland, 2005 (Anderson and others 2006) 2013). The World Health Organization (WHO) has • Guideline implementation, Budapest, 2007 (Anderson pointed out the limited utility of these guidelines in and others 2008) resource-constrained countries; they fail to include • Optimizing outcomes, Chicago, Illinois, 2010 implementation costs and provide no guidance as to (Anderson and others 2011) how treatment that is effective but less than optimal • Supportive care and quality of life, Vienna, 2012 (and less expensive) could be provided affordably for (Cardoso and others 2013; Cleary and others 2013; poorer populations (WHO 2002). Ganz and others 2013) 48 Cancer Modeled after the National Comprehensive Cancer where many women typically present with locally Network in the United States (Winn and Botnick 1997), advanced or metastatic tumors. Achieving stage shifting the BHGI applied an evidence-based consensus panel is likely to reduce mortality; even if this effect is small, process to build a framework defining resource prioritiza- the quality of life will be improved. Women would no tion for early detection (Yip and others 2008), diagnosis longer present with large, sometimes ulcerated masses (Shyyan and others 2008), treatment (Eniu and others that are painful, odiferous, ostracizing, and amenable 2008), and delivery systems (Eniu and others 2008) at four only to palliative treatment. Breast-preserving surgery levels of available resources: basic, limited, enhanced, and will also be possible in more cases, further reducing maximal (box 3.1). The framework is designed to facili- morbidity and enhancing the quality of life. tate strategy development and decision making by policy Early detection approaches include screening for makers and health care administrators initiating breast asymptomatic disease and early diagnosis of symptom- cancer control programs or reviewing existing services. atic disease (table 3.1). As a new screening program is Different resource levels may apply to different areas of implemented, more cancers will be detected initially, a country, because health care access and resources vary creating an apparent increase in disease incidence. As the with infrastructure and geography. The same methodol- screening program becomes established, the detection ogy has been applied to the development of guidelines rate will decline to a steady state. Once diagnosed, more for the management of hepatocellular carcinoma in Asia patients will require treatment, at a cost to the health (Poon and others 2009), non-small cell lung carcinoma care system. This increased cost may be partially offset by (Soo and others 2009), endometrial cancer (Tangjitgamol lower total treatment costs, because patients with earlier and others 2009), head and neck cancer (Wee and others stage disease require less therapy, but this model assumes 2009), and HER2/neu-positive breast cancer (Wong and that the patients would have been treated in both cases. others 2009). The cost and human resource requirements of increased demand for treatment must be factored into any deci- sion to establish a screening program. BHGI GUIDELINES Guidelines on Early Detection Mammographic Screening Shifting the stage distribution of disease to earlier stages The efficacy of mammographic screening was estab- is a necessary step to improving outcomes in LMICs, lished in trials in HICs that included monitoring Box 3.1 Definitions of Breast Health Global Initiative Resource Levels • Basic. Core resources or fundamental services • Maximal. High-level resources or services that that are necessary for any breast health care sys- may be used in some high-income countries and/ tem to function; basic-level services are typically or may be recommended by breast care guidelines applied in a single clinical interaction. that are not adapted to resource constraints. They • Limited. Second-tier resources or services that should be considered lower priority than those are intended to produce major improvements in resources or services listed in the basic, limited, or outcome and are attainable with limited financial enhanced categories, on the basis of cost and/or means and modest infrastructure; limited-level impracticality for broad use in resource-limited services may involve single or multiple clinical environments. To be useful, maximal-level interactions. resources typically depend on the existence and • Enhanced. Third-tier resources or services that functionality of all lower-level resources. are optional but important; enhanced-level Source: Anderson and others 2008. resources should produce further improvements Note: The stratification scheme implies incrementally increasing resource alloca- in outcome and increase the number and quality tion at the basic, limited, and enhanced levels. Maximal-level resources should not be targeted for implementation in LMICs, even though they may be used in of therapeutic options and patient choices. some higher-resource settings. Breast Cancer 49 Table 3.1 Early Detection Resource Allocation Level of available resources Early Detection Basic Limited Enhanced Maximal Public • Development of culturally • Culturally and • Regional awareness programs • National awareness education and sensitive, linguistically linguistically appropriate regarding breast health campaigns regarding breast awareness appropriate local education targeted outreach/ linked to general health and health using media programs for target populations education encouraging women’s health programs to teach value of early CBE for age groups at detection, breast cancer risk higher risk administered factors, and breast health at district/provincial awareness (education + level using health care self-examination) providers in the field Detection • Clinical history and CBE • Diagnostic breast US ± • Mammographic screening • Consider annual methods diagnostic mammography every 2 years in women ages mammographic screening in in women with positive 50–69a women ages 40 and older CBE • Consider mammographic • Other imaging technologies • Mammographic screening screening every 12–18 months as appropriate for high-risk of target groupa in women ages 40–49a groupsb Evaluation • Breast health awareness • Downsizing of • Downsizing and/or • Downsizing and/ goal regarding value of early symptomatic disease downstaging of asymptomatic or downstaging of detection in improving breast disease in women in highest asymptomatic disease in cancer outcome yield target groups women in all risk groups Source: Anderson and others 2008. Used with permission. Note: CBE = clinical breast examination; US = ultrasound; ± = with or without. a. Target group selection for mammographic screening should consider breast cancer demographics and resource constraints within the population. b. Magnetic resonance imaging is more sensitive than mammography in detecting tumors in asymptomatic women who have an inherited susceptibility to breast cancer. mammograms for image and interpretation quality. 40–49 years. The results were consistent across studies Mammography requires high-quality instrumentation in older women but were inconsistent for women in and specially trained radiologists, whose performance their forties. varies substantially, depending on training and level Reviewing the same trials, the U.S. Preventive Services of experience; screening efficacy is reduced if mam- Task Force concluded that the mortality benefit is mograms are of inferior quality, or if those who read more consistent among younger women than previously mammograms are not adequately trained and assessed described; 15 percent for ages 39–49; 14 percent for on an ongoing basis (Barlow and others 2004; Ichikawa ages 50–59; and 32 percent for ages 60–69 (Nelson and and others 2010). others 2009). The U.S. Preventive Services Task Force Ensuring the quality of imaging and interpretation withdrew a prior recommendation for routine screening is challenging in LMICs, due to the need to purchase for women ages 39–49 who are at average risk, based on machines, ensure ongoing quality control, and maintain the larger number of women who need to be screened screening registers. to save a life (1,904 for ages 39–49; 1,339 for ages 50–59; Many organizations and investigators, including the and 377 for ages 60–69) (table 3.2) (U.S. Preventive International Agency for Research on Cancer (IARC) Services Task Force 2009). (IARC 2008) and the U.S. Preventive Services Task Force, The positive predictive value of screening mam- have reviewed the evidence for the efficacy of screening mography increases with age. Premenopausal women with mammography (Nelson and others 2009). The tend to have denser breast tissue and a higher rate of evidence comes mainly from six trials, in which women benign lesions than postmenopausal women, result- were randomized to periodic mammographic screening ing in greater difficulty in detecting lesions and lower or no screening. Based on these trials, IARC estimated a sensitivity and specificity (that is, a higher rate of false- reduction in breast cancer of about 25 percent in women positives). Furthermore, the incidence of breast cancer ages 50–69 years and about 19 percent in women ages increases with age in most populations. 50 Cancer Table 3.2 Pooled Relative Risk for Breast Cancer smaller proportion more frequently. The availability Mortality from Mammographic Screening Trials for and cost of x-ray film may also limit the availability of Women Ages 39–74 Years screening in LMICs; digital mammography may pro- vide an alternative, if costs are reduced to affordable Relative risk for NNI to prevent levels. breast cancer one breast cancer Age Trials mortality (95% death (95% Because mammographic screening is expensive and (years) included credible interval) credible interval) requires considerable infrastructure, consideration has been given in LMICs to screening with clinical breast 39–49 8 0.85 (0.75–0.96) 1904 (929–6378) examination (CBE) and breast self-examination (BSE). 50–59 6 0.86 (0.75–0.99) 1339 (322–7455) 60–69 2 0.68 (0.54–0.87) 377 (230–1050) Clinical Breast Examination 70–74 1 1.12 (0.73–1.72) — CBE is a basic tool in clinical management of breast Source: Nelson and others 2009. cancer. However, no randomized trials to assess CBE Note: NNI = number needed to invite to screening; — = not available. alone as a screening tool have yet been completed, while the evidence from observational studies is inconsistent. A case-control study in Japan suggested that breast cancer deaths were reduced among asymp- Although it may be less cost effective to screen women tomatic women who underwent CBE screening (odds in their forties than women ages 50 and older, a relatively ratio = 0.56) (Kanemura and others 1999). A second high proportion of women with breast cancers in LMICs Japanese study reported significantly greater reduc- are ages 40–49, complicating the decision making about tions in the age-adjusted death rate from breast cancer the parameters of a screening program. Without screen- in areas with high rates of screening coverage (princi- ing, the largest age cohort of breast cancer patients will pally by physical examination), compared with areas be missed and many women will continue to present for where coverage was not established (Kuroishi and treatment with more advanced stage disease. others 2000). An early clinical trial conducted by the Some groups question whether the demonstrated Health Insurance Plan of Greater New York demon- mortality benefit of mammographic screening might be strated that a combination of mammography and outweighed by the risk of harm through false-positive CBE reduced the risk of breast cancer mortality; many studies or the potential for overtreatment of biologically of the tumors in the screened group were detected favorable early-stage disease (Baum 2013; Gotzsche and by CBE but not by mammography (Chu, Smart, and others 2012). Informed consent should be provided to Tarone 1988). However, more recent randomized women so that they have an appropriate understanding trials compared the combination of mammography of the potential risks as well as benefits of mammo- and CBE with mammography alone, and the two graphic screening (Thornton 2014). approaches appeared to have equivalent outcomes Standard practice in HICs is two-view mammog- (IARC 2008). As a result, IARC and the U.S. Preventive raphy, in which the breast is imaged in two planes Services Task Force concluded that the evidence for (MLO–medial/lateral oblique, CC–craniocaudal), the efficacy of CBE in reducing mortality from breast which allows a more accurate reading. However, in cancer is “inadequate” and “insufficient,” respectively older randomized trials, the quality of single-view (U.S. Preventive Services Task Force 2009). mammograms was nearly equivalent to that of two- Most of the evidence for the efficacy of CBE has come view mammograms. In LMICs with limited resources, from studies in HICs, where women typically present consideration can be given to single-view mammog- with relatively small tumors. CBE has been advocated for raphy, if doing so will increase the coverage and the LMICs for several reasons. CBE may be more efficacious screening interval can be extended beyond one or in LMICs, where women tend to present with larger two years. In women ages 50–69, screening every tumors; CBE is less expensive than mammography 33 months was as efficacious as screening every 18–24 because it can be performed by trained health workers months; in women ages 40–49, the reduction in who are not physicians and requires less equipment than mortality was inversely related to the screening inter- mammography. A recent study of 1,179 screened women val (Breast Screening Frequency Trial Group 2002). in Jakarta, Indonesia, compared the use of mammogra- When allocating limited resources for mammographic phy and CBE in a previously unscreened population and screening, it is more cost effective to screen a higher identified 14 breast cancers. Of the 14 cancers, 13 were proportion of women less frequently than to screen a detected by CBE (Kardinah and others 2014). Breast Cancer 51 If CBE is used for screening breast cancer in LMICs, to a control group that received no such education and it should be done in such a way that its effectiveness no formal breast cancer screening. Mortality from breast can be evaluated. A randomized trial in the Philippines cancer was unchanged by BSE instruction in these trials. was unsuccessful because too few women with positive Both the IARC working group and the U.S. Preventive findings on CBE consented to further diagnostic tests to Services Task Force concluded that the efficacy of BSE determine if the findings reflected a malignant or benign is unproven (U.S. Preventive Services Task Force 2009). finding (Pisani and others 2006). A randomized trial However, it has been questioned if the negative findings of CBE and visual inspection of the cervix by specially of these BSE trials are relevant to LMICs, where women trained women with a tenth-grade education is under- commonly present with large (> 4 cm) cancers at initial way in Mumbai, India. Preliminary results show more diagnosis. In the Shanghai trial, women in the control breast cancers are being detected at early stages (stages 0, group were not taught BSE but nonetheless were largely I, or II) in the screening group (62 percent) than in the successful in finding cancers when they were still small, control group (44 percent), but breast cancer mortality that is, where 45 percent of the cancers were found as results are not yet available (Mittra and others 2010). in situ or T1 invasive cancers measuring less than 2 cm In addition to assessments of its efficacy—sensitivity, (Thomas and others 2002). These favorable findings shifted stage distribution, and reduction in breast cancer among the untrained control women from Shanghai mortality—all such efforts to evaluate the usefulness stand in stark contrast to regions of India, where of CBE should include a measure of its specificity. If 76 percent of women present with locally advanced or large numbers of lesions detected by CBE are found metastatic (stage III or IV) disease at initial presentation not to be cancerous on further evaluation, this puts a (Chopra 2001). In this latter setting, the actual benefit of heavy burden on local diagnostic facilities. In addition, BSE training could potentially be much greater. if many women undergo unnecessary breast biopsies, It remains unknown whether BSE could reduce this may not be acceptable, either to the women targeted mortality from breast cancer in populations in LMICs. for screening or to policy makers who allocate scarce It is not unreasonable to advocate that BSE be used as a resources. screening tool in these settings, either alone or in combi- CBE accuracy depends on the skill of providers, the nation with CBE. No new trials of BSE alone have been definition of proper techniques, and the type of training undertaken, but BSE instruction has been included in received. Strategies for providing routine feedback to some of the studies of CBE. The IARC working group health care providers about the accuracy of their exami- recommended randomized trials of BSE in conjunction nations as determined on final diagnostic work-up are with mammography. In LMICs where mammographic integral to successful CBE programs. screening cannot be provided at least every two years, it may be particularly useful to teach BSE. Breast Self-Examination Any introduction of BSE should be accompanied by The aim of BSE is to detect asymptomatic breast condi- evaluation of its efficacy, including quantification of the tions and should be distinguished from programs that benign lesions that must be evaluated. In both of the promote early treatment of symptomatic breast cancer. completed BSE trials, many more benign breast lesions BSE is the systematic search performed regularly by the were detected in the groups that received BSE instruc- women themselves for a lump or other change in the tion than in the control groups. breast that is suggestive of cancer. In formal BSE train- ing, a woman receives instruction in the four elements Breast Awareness Education of the examination: visual inspection of the breasts in a Programs to promote early diagnosis and treatment of mirror to look for asymmetry and dimpling; palpation symptomatic breast cancer are not screening programs, in both the standing and lying positions with the arm because they are not designed to detect asymptomatic above the head, using a circular motion with the pads of lesions. Their purpose is to encourage women who have the three middle fingers, with systematic coverage of the symptoms suggestive of breast cancer to seek medical entire breast and axilla; squeezing of the nipple to detect care. Women can be educated to detect suspicious discharge; and monthly BSE practice. changes in their breasts and empowered to overcome Most evidence for the efficacy of BSE comes from two social barriers that might prevent them from seeking randomized trials from Saint Petersburg (Semiglazov care. Breast self-awareness programs should not be initi- and others 1999) and Shanghai (Thomas and others ated unless adequate diagnostic and treatment facilities 2002). In both studies, women were randomized to are available. And the programs should be established in either an intervention group that received instruction in such a way that they can be evaluated to determine their BSE and periodic reminders to practice the procedure or effectiveness. 52 Cancer Guidelines on Diagnosis Clinical Evaluation Diagnosis is a critical and often overlooked aspect of A patient’s history of general health and of factors spe- breast cancer management. Two key components of cific to breast disease provides important information diagnosis are confirmation of a cancer diagnosis based for clinical assessment of breast disease and comor- on clinical evaluation and tissue sampling, and testing bid conditions that might influence therapy choices. with the imaging and tumor markers needed for treat- Complete physical examination performed in conjunc- ment planning (table 3.3). tion with CBE provides guidance on the extent of Table 3.3 Diagnosis Resource Allocation Level of available resources Diagnosis Basic Limited Enhanced Maximal Clinical • History • US-guided FNAB • Image guided breast • Physical examination of sonographically sampling suspicious axillary nodes • Preoperative needle • CBE • SLN biopsy with blue dyea localization under mammo • Tissue sampling for cancer and/or US guidance diagnosis (cytologic or histologic) prior to initiation • SLN biopsy using of treatment radiotracera Imaging and See footnote b • Diagnostic breast US • Diagnostic mammography • PET scan, MIBI lab tests • Plain chest and skeletal • Specimen radiography scan, breast MRI, radiography BRCA 1/2 testing • Bone scan, CT scan • Liver US • Mammographic • Cardiac function double reading • Blood chemistry profileb monitoring • CBCb Pathology • Pathology diagnosis obtained • Determination of ER • Measurement of HER2/ • IHC staining of for every breast lesion by any status by IHCc neu overexpression or sentinel nodes for available sampling procedure • Determination of margin gene amplificationd cytokeratin to detect • Pathology report containing status, DCIS content, • Determination of PR micrometastases appropriate diagnostic presence of LVI status by IHC • Pathology double and prognostic/predictive • Frozen section or touch reading information to include tumor prep SLN analysisd • Gene profiling tests size, lymph node status, histologic type, and tumor grade • Process to establish hormone receptor status possibly including empiric assessment of response to therapyc • Determination and reporting of TNM stage Source: Anderson and others 2008. Used with permission. Note: BRCA1/2 = breast cancer genes 1 and 2; CBC = complete blood count; CBE = clinical breast examination; CT = computed tomography; DCIS = ductal carcinoma in situ; ER = estrogen receptor; FNAB = fine-needle aspiration biopsy; HER2 = human epidermal growth factor receptor 2; IHC = immunohistochemistry; LVI = lymphovascular invasion; mammo = mammography; MIBI = methoxy-isobutyl-isonitrile; MRI = magnetic resonance imaging; PET = positron emission tomography; PR = progesterone receptor; SLN = sentinel lymph node; TNM = malignant tumor system; US = ultrasound. a. The use of SLN biopsy requires clinical and laboratory validation of the SLN technique. b. Systemic chemotherapy requires blood chemistry profile and CBC testing for safety. When chemotherapy is available at the basic level, these tests also should be provided. c. ER testing by IHC is preferred for establishing hormone receptor status and is cost effective when tamoxifen is available. When tamoxifen is available at the basic level, IHC testing of ER status should be provided. d. If the costs associated with trastuzumab were substantially lower, trastuzumab would be used as a limited-level therapy. In this case, measurement of HER2/neu overexpression and/or gene amplification would need to be available at the limited level to properly select patients for this highly effective but expensive HER2/neu-targeted biological therapy. Breast Cancer 53 disease, the presence of metastatic disease, and the ability HER2/neu oncogene testing provides information to tolerate more aggressive therapeutic regimens. on the relative aggressiveness of the cancer (HER2/ neu-positive cancers are more aggressive), as well as on Diagnostic Imaging the likely drug sensitivity of the cancer (Yoo and others Breast imaging, initially with ultrasound and at higher 2012). However, the most effective drug for HER2/ resource levels with diagnostic mammography, improves neu-positive cancers is trastuzumab, which is unafford- preoperative diagnostic assessment and permits image- able in most regions. Less expensive therapies are under guided needle sampling of suspicious lesions. Imaging investigation (Pinto and others 2013). also provides important information about the extent of disease, which helps determine whether breast conserva- tion (lumpectomy followed by radiation therapy) is an Guidelines on Treatment option or mastectomy is required. Ultrasound is partic- Surgery ularly valuable as an adjunct to CBE in providing detail The modified radical mastectomy is the mainstay on the size and extent of masses and thickenings, which of treatment of local and regional (nodal) disease helps to distinguish benign cysts from solid lesions and at the basic level of breast health care for early-stage characterizes the shape and growth pattern of lesions. (table 3.4) and late-stage (table 3.5) disease (Anderson Diagnostic mammography, while helpful for breast and others 2008). The operation is not technically dif- conservation therapy, is not mandatory in LMICs when ficult, although surgeons must be trained to remove the these resources are lacking. However, where screening breast and dissect axillary nodes properly (Thorat and mammography is common and where nonpalpable, others 2008). noninvasive cancers are often diagnosed, diagnostic mammography is critical for determining the extent Radiation Therapy of disease and properly selecting patients for breast At increasing resource levels, the availability of radiation conservation surgery versus mastectomy (Theriault and therapy allows for consideration of breast-conserving others 2013). therapy, postmastectomy chest wall radiation, and palliation of painful or symptomatic metastases (see Tissue Sampling tables 3.4 and 3.5). Although radiation therapy requires Needle biopsy is preferred to surgical excision for ini- significant infrastructure and can be cost limiting in tial diagnosis of the most suspicious breast lesions; if improving treatment, the establishment of a radiation resource limitations preclude this option, then surgical facility can be an important first step in creating an excision is necessary. Mastectomy should never be used oncology center of excellence in an LMIC (Bese and as a method of tissue diagnosis. Whether the tissue others 2008). is obtained by needle sample or surgical excision, the sample must be processed and then evaluated by a Systemic Pharmacotherapy pathologist to determine whether the disease is malig- Although surgery and radiation address local dis- nant or benign, and invasive or noninvasive (Shyyan and ease in the breast and regional disease in the lymph others 2008). node beds, systemic therapy addresses microscopic disease elsewhere that can become metastases. When Tumor Markers patients die from breast cancer, the cause is wide- Once a cancer diagnosis is made, additional testing pro- spread metastatic disease. It is pharmacotherapy that vides information on which to base pharmacotherapy ultimately improves breast cancer survival rates, since choices. For example, tamoxifen and aromatase inhib- this is the only treatment directed at systemic dis- itors are affordable generic oral medications that are ease. Pharmacotherapies for breast cancer consist of quite effective in the management of ER-positive cancers endocrine (hormonal) therapy, cytotoxic chemother- with relatively manageable side effects, but these agents apy, and biological targeted (antibody) therapies (see are relatively ineffective against ER-negative cancers tables 3.4 and 3.5). (Howell and others 1998). The availability of ER testing is critical to proper selection of cancer therapy when • Endocrine therapy requires relatively few special- endocrine therapies are available. Standard testing is ized resources, but it requires knowledge of hor- based on immunohistochemical (IHC) methods, where mone receptor status to identify the patients most quality assessment of testing methodology is important likely to benefit. For ER-positive cancers, tamoxifen to avoid false-negative results (Hammond and others and aromatase inhibitors are oral drugs taken daily 2010; Masood and others 2008). for five years or more that can be dispensed from 54 Cancer Table 3.4 Treatment Resource Allocation for Stage I and Stage II Disease Level of available resources Treatment Basic Limited Enhanced Maximal a b Stage I Local-regional Surgery Modified radical Breast-conserving surgery SLN biopsy using radiotracer treatment mastectomy SLN biopsy with blue dyeb Breast reconstruction surgery Radiation therapy Breast-conserving whole- breast irradiation as part of breast-conserving therapya Systemic Chemotherapy Classic CMFc, AC, EC, or Taxanes Growth factors treatment FACc Dose-dense chemotherapy Endocrine therapy Oophorectomy in Aromatase inhibitors LH-RH premenopausal women agonists Tamoxifend Biological therapy See footnote e Trastuzumab for treating HER2/neu-positive diseasee Stage II Local-regional Surgery Modified radical Breast-conserving surgerya SLN biopsy using radiotracerb treatment mastectomy SLN biopsy with blue dyeb Breast reconstruction surgery Radiation therapy See footnote f Postmastectomy irradiation Breast-conserving whole- of chest wall and regional breast irradiation as part of nodes for high-risk casesf breast-conserving therapya Systemic Chemotherapy Classic CMFc, Taxanes Growth factors treatment AC, EC, or FACc Dose-dense chemotherapy Endocrine therapy Oophorectomy in Aromatase inhibitors LH-RH premenopausal women agonists Tamoxifend Biological therapy See footnote e Trastuzumab for treating HER2/neu-positive diseasee Source: Anderson and others 2008. Used with permission. Note: AC = doxorubicin and cyclophosphamide; CMF = cyclophosphamide, methotrexate, and 5-fluorouracil; EC = epirubicin and cyclophosphamide; FAC = 5-fluorouracil, doxorubicin, and cyclophosphamide; HER2/neu = human epidermal growth factor receptor 2; LH-RH = luteinizing hormone-releasing hormone; SLN = sentinel lymph node. a. Breast-conserving surgery can be provided as a limited-level resource but requires radiation therapy. If breast-conserving radiation is unavailable, then patients should be transferred to a higher-level facility for postlumpectomy radiation. b. The use of SLN biopsy requires clinical and laboratory validation of the SLN technique. c. Systemic chemotherapy requires blood chemistry profile and complete blood count testing for safety. When chemotherapy is available at the basic level, these tests also should be provided. d. ER testing by IHC is preferred for establishing hormone receptor status and is cost effective when tamoxifen is available. When tamoxifen is available at the basic level, then IHC testing of ER status also should be provided. e. If the costs associated with trastuzumab were substantially lower, trastuzumab would be used as a limited-level resource. In this case, measurement of HER2/neu overexpression and/or gene amplification would need to be available at the limited level to select patients properly for this highly effective but expensive HER2/neu-targeted biological therapy. f. Chest wall and regional lymph node irradiation substantially decreases the risk of postmastectomy local recurrence. If available, it should be used as a basic-level resource. pharmacies without special infrastructure and are testing accuracy. Alternative simplified techniques considered very safe. Endocrine therapy could be for ER testing are of significant interest but remain given to all breast cancer patients, but it would experimental. be a waste of resources since it is only effective • Systemic cytotoxic chemotherapy is effective in against ER-positive cancers. IHC methods involve most biologic subtypes of breast cancer. It is partic- special tissue-staining techniques with labeling ularly important in the management of ER-negative antibodies, which requires pathology laboratory cancers but is resource intensive. Chemotherapy infrastructure; quality control is quite important to has significant side effects that must be managed Breast Cancer 55 Table 3.5 Treatment Resource Allocation for Locally Advanced and Metastatic Disease Level of available resources Treatment Basic Limited Enhanced Maximal Locally Local-regional Surgery Modified radical Breast-conserving surgery advanced treatment mastectomy Breast reconstruction surgery Radiation therapy See footnote a Postmastectomy Breast-conserving whole- irradiation of chest breast irradiation as part wall and regional of breast-conserving nodesa therapy Systemic treatment Chemotherapy Preoperative Taxanes Growth factors (Adjuvant and chemotherapy with AC, Dose-dense neoadjuvant) EC, FAC, or CMFb chemotherapy Endocrine therapy Oophorectomy in Aromatase inhibitors premenopausal women LH-RH agonists Tamoxifenc Biological therapy See footnote d Trastuzumab for treating HER2/neu-positive diseased Metastasic and Local-regional Surgery Total mastectomy recurrent treatment for ipsilateral breast tumor recurrence after breast-conserving surgery Radiation therapy Palliative radiation therapy Systemic treatment Chemotherapy Classic CMFb Sequential single agent or Bevacizumab Anthracycline combination chemotherapy: monotherapy or in Trastuzumab combinationb Lapatinib Endocrine therapy Oophorectomy in Aromatase inhibitors Fulvestrant premenopausal women Tamoxifenc Biological therapy Nonopioid and opioid Bisphosphonates Growth analgesics and factors symptom management Source: Anderson and others 2008. Used with permission. Note: Treatment resource allocation table for locally advanced, metastatic (stage IV), and recurrent breast cancer. AC = doxorubicin and cyclophosphamide; CMF = cyclophosphamide, methotrexate, and 5-fluorouracil; EC = epirubicin and cyclophosphamide; FAC = 5-fluorouracil, doxorubicin, and cyclophosphamide; HER2/neu = human epidermal growth factor receptor 2; LH-RH = luteinizing hormone-releasing hormone. a. Chest wall and regional lymph node irradiation substantially decreases the risk of postmastectomy local recurrence. If available, it should be used as a basic-level resource. b. Systemic chemotherapy requires blood chemistry profile and complete blood count testing for safety. When chemotherapy is available at the basic level, these tests should be provided. c. Estrogen receptor testing by immunohistochemistry (IHC) is preferred for establishing hormone receptor status and is cost effective when tamoxifen is available. When tamoxifen is available at the basic level, then IHC testing of estrogen receptor status should be provided. d. If the costs associated with trastuzumab were substantially lower, trastuzumab would be used at a limited level. In this case, measurement of HER2/neu overexpression and/or gene amplification would need to be available at the limited level to properly select patients for this highly effective but expensive HER2/neu-targeted biological therapy. effectively. Correct drug selection is based on the sufficiently high to provide optimal effects on the extent or stage of the cancer and on tumor markers cancer but as low as possible to minimize adverse that can predict likely drug sensitivity. Proper drug events. Proper management of these agents is criti- dosing is important and must be individualized to cal; they must be handled under sterile conditions, the patient’s body mass index; the dosage should be they must be properly and safely administered, and 56 Cancer health care workers should not be directly exposed to unreported, or if certification of the cause of death is these agents. not accurately recorded. In the absence of adequate • Biological targeted therapies use monoclonal anti- information on the population to be screened, reviews bodies to control disease. HER2 neu-targeted ther- of death certificates can be useful. The number of deaths apy with trastuzumab is very effective in tumors certified as due to breast cancer during specific years can that overexpress the HER2/neu oncogene, but cost be obtained to provide a rough estimate of the number largely prevents the use of this treatment in LMICs of annual breast cancer deaths in the population. In (Eniu and others 2008); a standard one-year course addition, the proportion of all deaths due to breast can- of treatment in the United States is approximately cer can be calculated and compared with the proportion US$100,000. It remains unclear whether generic of deaths from other causes. If breast cancer is a small forms of trastuzumab will be available in the future. problem relative to other preventable causes of death, then a screening program may not be warranted, but if breast cancer deaths constitute a relatively high propor- IMPLEMENTATION OF BHGI GUIDELINES: tion of preventable deaths, this information can help to AN EARLY DETECTION MODEL justify the costs of screening. Before an LMIC initiates a breast cancer control pro- A review of hospital records can be useful in assessing gram or evaluates existing programs, careful assessment the magnitude of the problem. If a single hospital serves of the local situation is needed. This assessment consists all cancer patients in a defined population, then the of three parts: hospital records can be reviewed to estimate the number of cases to be expected annually. A record review can • Breast cancer problem in the population provide an indication of the importance of breast cancer • Existing infrastructure that will be utilized for the relative to other cancers in the population and relative to program other reasons for hospitalization. If admissions for breast • Social and cultural barriers to women’s participation cancer constitute a relatively high proportion of all pre- in the program ventable causes of admission, or a high proportion of all admissions for cancer, then a screening program may be justifiable; if breast cancer is a rare cause of hospitaliza- Assessing the Breast Cancer Problem tion, then it may not warrant high prioritization. A realistic estimate of the number of women with breast The presence of social or financial barriers that keep cancer in the population in which screening is proposed women from accessing services will limit the effective- is an essential part of the planning process; the lower the ness of a screening program; it will be important to prevalence, the higher the number of women who have consider initiatives to overcome such barriers. to be screened to detect cases and prevent deaths. Assessing the magnitude of the problem before imple- Estimating the frequency of breast cancer in LMICs menting a screening program also includes determining can be a challenge and may require using less than ideal the disease stage distribution at diagnosis. This infor- methods. The ideal situation is where the number of mation would be found in population-based cancer women in the target screening population is known registries or hospital-based registries. In the absence of and a reliable population-based cancer registry covers a cancer registries, the records in clinics, hospitals, and substantial portion of the population. These elements, pathology laboratories can be reviewed. If a high propor- in conjunction with information from countries with tion of breast cancers are diagnosed at advanced stages, well-developed breast cancer screening programs, would then a screening program, or an educational program support a reasonable estimate of expected impact. In to encourage earlier diagnosis of symptomatic breast countries with poorer enumeration of the target pop- cancers, could have a substantial impact on the burden. ulation or cancer registration, more extrapolation is Conversely, if a high proportion of breast cancers are involved; estimates of breast cancer burden and expected already being diagnosed at early stages, then screening benefit become less reliable but still useful. For exam- programs based on BSE, and probably also on CBE, are ple, incidence rates for populations that may be similar unlikely to have a large impact on mortality or morbidity. to populations without registries are found in Cancer Incidence in Five Continents (Curado and others 2007). Accurate mortality rates can serve as good measures Assessing the Infrastructure for Screening, of the extent of the breast cancer problem. However, Diagnosis, and Treatment mortality rates can be misleading if the population Mammography, the only screening method of proven size is not accurately enumerated, if many deaths are efficacy, serves as an example for assessing infrastructure. Breast Cancer 57 For a mammographic screening program to result in of cancer fatalism. Muslim women were half as likely earlier detection of breast cancers in the population in as Christian women to participate in CBE screening. LMICs and elsewhere, six elements need to function Women were more likely to perform BSE if they were adequately: more highly educated, resided in cities, were Christian, were less religious, and had a first-degree relative with • Means to recruit enough women in the target pop- breast cancer. These results suggest that participation ulation to have a meaningful impact on the breast in screening might be improved by recruiting religious cancer burden leaders as spokespersons for early detection and by staff- • Facilities to ensure high-quality mammograms ing screening clinics with women physicians and nurses • Sufficient number of radiologists who can properly sensitive to the needs of conservative Muslim women interpret mammograms in a timely manner who must remain covered in public. • Means to recontact women with suspicious findings Women who are correctly diagnosed and properly and ensure that they come to facilities for further treated for early-stage breast cancer can survive the evaluation in a timely fashion disease and can organize breast cancer survivor groups, • Adequate diagnostic facilities and trained patholo- such as Reach for Recovery (figure 3.3). Such groups can gists to provide timely and accurate tissue diagnoses play a vital role in educating the public about the value • Sufficient facilities and personnel to provide timely of early detection and in providing newly diagnosed and appropriate treatment women with practical and emotional support (Ashbury and others 1998). Survivor groups can organize into Regardless of the screening modality used—BSE, political advocacy groups that have a real and positive CBE, or screening mammography—gaps in this system impact on health care policy or national cancer research at any level must be identified and addressed before a agendas (Schmidt 2009; Visco 2007). program of early detection is established. Identifying Target Groups Assessing Social and Cultural Barriers Identifying a target group for screening in LMICs Women in LMICs may be unaware of breast cancer, or should be based on the burden of disease in the popula- they may have misconceptions about its nature or cur- tion, the potential benefit from screening, and available ability or have fatalistic attitudes toward diseases in gen- resources (Humphrey and others 2002). Other than the eral (Yip and others 2008). Under such circumstances, small subset of women at very high risk of developing programs to enhance public awareness of breast cancer breast cancer due to genetic predisposition, it is very dif- and to teach that breast cancer outcomes are improved ficult to predict which women are destined to develop through early detection are critical to improving par- breast cancer. Although women with BRCA (breast ticipation in early detection programs, regardless of the cancer gene) mutations generally have a strong family selected methods for early detection. history of breast and/or ovarian cancer, family history Cultural barriers to participation need to be identi- fied and strategies developed to overcome them. These barriers may include the attitudes of women as well as Figure 3.3 Synergistic Relationship between Public their husbands; in some cultures, women must obtain Participation and Health Care Delivery in Downstaging their husbands’ permission to seek medical services. Breast Cancer and Improving Outcomes Efforts to empower women and educate men may be required for programs to succeed. Cultural and social Public Participation Health care delivery barriers are highly specific to different countries, reli- gions, and ethnic groups and cannot be comprehensively Awareness Early detection reviewed here. However, an example illustrates how they may be addressed. In a survey in the Palestinian Authority (Azaiza and others 2010), women were more Advocacy Diagnosis likely to undergo screening mammography if they were less religious, if they described fewer personal barriers to examinations, and if they indicated a lower degree of cancer fatalism. Survivorship Treatment Women who consented to CBE had a higher per- ceived effectiveness of CBE and described lower levels Source: Harford and others 2011. Used with permission. 58 Cancer is not a particularly good tool for selecting women for at each of BHGI’s four defined resource levels: Basic, screening, since approximately 80 percent of breast can- Limited, Enhanced, and Maximal. By contrast, WHO’s cers occur in women who lack a known family history Choosing Interventions That Are Cost-Effective of breast cancer. As a result, the only two risk factors (WHO-CHOICE) framework provides a somewhat dif- used for determining candidacy for screening are gen- ferent and not entirely comparable set of breast cancer der and age. guidelines (Murray and Lopez 1996; Tan-Torres Edejer In Western countries, breast cancer incidence and others 2003). WHO-CHOICE was established increases sharply with age until the usual age at meno- as an initiative to provide evidence to policy makers pause and then increases more slowly. In LMICs, inci- who must decide on the interventions and programs dence increases with age until menopause but then that maximize health outcomes for given available either continues to increase less steeply with age than resources, reporting on the costs and effects of a wide in Western countries, levels off, or decreases with age range of health interventions but without direct align- (Freedman and others 2006). This phenomenon is the ment with current stage-based treatment strategies for result of an aging and growing population (Chia and specific diseases. others 2005; Wong, Cowling, and others 2007); over Whether grounded in the BHGI or WHO-CHOICE time, the age-specific incidence curves for LMICs are framework, the development of analytical mod- expected to more closely approximate those of Western els to identify clinically effective and cost-effective countries (Yip and others 2008). approaches for improving breast health requires link- In LMICs where incidence rates in women ages 50 ing resources to interventions and patient outcomes years and older are beginning to increase with age, the (Brown and others 2006; Gold and others 1996). The prevalence of the disease in the population may have value-for-money question can be summarized as fol- approached a level at which establishing screening pro- lows: given the assumed level of resources available grams will be cost effective. within a given geopolitical area, what administratively and financially feasible set of interventions has the greatest favorable impact on health outcomes? This Considering Coverage and Impact section reviews and assesses progress to date in iden- Coverage of the target population is also important. If tifying cost-effective interventions for breast cancer screening is efficacious in reducing mortality from breast in LMICs. cancer, but only a small proportion of the women in the target population receives the service, then the impact of Appraisal of the Literature the screening program on mortality in the population will be minimal. The following simple formula illus- As the recent review by Zelle and Baltussen (2013) well trates the impact (Thomas and others 2013): illustrates, substantial variations exist in the precise purpose, scope, methodology, assumptions, and tech- Impact = Efficacy × Coverage nical quality of published cost-effectiveness analyses on breast cancer interventions in LMICs. The follow- If we assume that mammograms reduce mortality ing section focuses on a selected set of studies that from breast cancer by 25 percent, and if 40 percent of yields findings relevant to components of the BHGI the women in a target population are screened, then the guidelines while paying close attention to important screening program would be expected to reduce mortal- cost-effectiveness analysis data and methods issues. ity in the target population by 10 percent (0.25 × 0.40 Nine of the studies were among 23 selected for detailed = 0.10). Mammographic screening programs in LMICs manuscript-quality evaluation by Zelle and Baltussen should be designed in such a way that the proportion of and were generally among the highest rated, according women in the target population who are screened can be to the authors’ scoring scheme. The studies illustrate maximized. how cost-effectiveness analyses can be carried out in limited-resource settings. Salient features and recommendations of these key studies are summarized in the online annex 3A to COST-EFFECTIVE INTERVENTIONS IN LMICs this chapter, tables 3A.1 through 3A.3. We have fur- The BHGI resource-adapted guidelines provide a ther mapped the study contexts to BHGI resource sound foundation for creating intervention packages levels and highlighted features following methodological for early detection (table 3.1), diagnosis (table 3.3), best-practice guidelines. Here, we focus on the cost- and treatment by disease stage (tables 3.4 and 3.5) effectiveness results. Our conclusions about whether an Breast Cancer 59 intervention is cost effective were guided throughout by was cost effective; Lee and others concluded that sub- the following WHO recommendations (WHO 2001): stituting docetaxel for fluorouracil in another, more complex regimen was very cost effective. A China- • If the intervention’s applicable incremental cost- based analysis (Bai and others 2012) concluded that effectiveness ratio (ICER) is less than or equal to the radiation therapy following breast-conserving surgery, nation’s or region’s GDP per capita, the intervention compared with no radiation, was very cost effective is “very cost effective.” under a wide range of assumptions. • If the ICER is between one and three times the GDP per capita, it is “cost effective.” Combination Screening-Treatment Interventions • If the ICER exceeds three times the GDP per capita, The four cost-effectiveness analyses summarized in the intervention is “not cost effective.” annex table 3A.3 are all based on the WHO-CHOICE framework (Tan-Torres Edejer and others 2003); each examines alternative intervention packages involving Early Detection screening and treatment for breast cancer from the per- Annual CBEs are very cost effective in Vietnam (Nguyen spective of a particular nation or world region. Three of and others 2013); biennial mammography screening is the papers (Ginsberg and others 2012; Groot and others cost effective in Hong Kong SAR, China (Wong, Kuntz, 2006; Salomon and others 2012) investigate roughly the and others 2007). An analysis examining combinations same six options: treat only stage I, only stage II, only of both screening modalities applied to different age stage III, or only stage IV disease; treat all stages; or treat groups in India concluded that several alternative CBE all stages, plus some variant of a breast screening and/or detection strategies were very cost effective, while the educational program (the “extensive program” option). most efficient among various mammography screening Zelle and others (2012) evaluated a total of 17 options, strategies analyzed (biennial screening for women ages including these six and others that differed largely on 40–60) was not cost effective (Okonkwo and others whether screening was by CBE or mammography and 2008). An analysis in the Republic of Korea determined by the age range for screening. that the most cost-efficient mammography strategy Some modeling assumptions of WHO-CHOICE are (given a prior decision to adopt this modality) is not in alignment with standard stage-based treatment screening every three years for women ages 45–65 (Lee, protocols. For example, two model options assume that Jeong, and others 2009). early-stage breast cancer is not treated with chemother- apy when, in practice, most stage II and some stage I Treatment breast cancers do warrant chemotherapy on the basis In analyses of hormonal adjuvant therapies post-surgery, of high-level, prospective randomized clinical trials Fonseca and others (2009) found that the aromatase (Theriault and others 2013). This modeling assumption inhibitor anastrozole was cost effective compared with could lead to incorrect cost-effectiveness conclusions, tamoxifen in a Brazil-based cost-effectiveness analysis. since chemotherapy is among the most expensive of In Korea, Yang and others (2010) calculated that tamox- the required multimodality treatments. WHO-CHOICE ifen was very cost effective compared with no hormonal also assumes that lumpectomy and sentinel lymph node therapy for hormone receptor-positive patients under a biopsy is the surgery for Stage I and II when, in practice, variety of assumptions; it was very cost effective or cost breast cancers in LMICs are often too large for breast- effective for hormone receptor-negative patients in only conserving surgery, and sentinel lymph node biopsy is a subset of cases (for example, when the patient was often an unavailable technique in these settings such that stage III and under age 50). A self-described preliminary complete axillary node dissection is routinely performed cost-effectiveness analysis that excluded certain cost for all invasive cancers. categories found that a combination of oophorectomy There is a general convergence in the recommenda- and tamoxifen was very cost effective compared with tions, notwithstanding the diversity in geopolitical “observation” in a Vietnam-based analysis (Love and setting for these cost-effectiveness analyses. In the others 2002). Asian and Sub-Saharan African regions analyzed, Groot The cost-effectiveness of alternative combination and others (2006) concluded that the extensive program chemotherapy regimens post-surgery was investigated was very cost effective. For Southeast Asia and Sub- for China (Liubao and others 2009) and Korea (Lee, Saharan Africa, Ginsberg and others (2012) found that Jee, and others 2009). Liubao and others found that variants of the extensive program (differing by the substituting docetaxel for doxorubicin in a treatment assumed fraction of the female population covered) package that otherwise included cyclophosphamide were all cost effective. Focusing on Brazil, Salomon and 60 Cancer others (2012) calculated that the option of treating all can be evaluated during their early years (McCannon, four stages of breast cancer was very cost effective, while Berwick, and Massoud 2007). Methods that have been the extensive program that considered various screening employed for program evaluation include observational options was cost effective. For Ghana, Zelle and others studies and randomized trials. Two types of observa- (2012) found that a variant of the extensive program tional studies are comparisons of screening modalities defined to include biennial CBE screening for women and assessments of temporal trends in stage of disease. ages 40–69 was cost effective. Randomized trials may be clinic- or population-based. Implications for Choosing Cost-Effective Breast Comparison of Modalities Cancer Interventions A recent study in Indonesia that compared the use Although we agree with Zelle and Baltussen (2013) of screening mammography and CBE in a previously that many published analyses of breast cancer inter- unscreened population found similar efficacy for breast ventions in LMICs suffer from serious data or meth- cancer detection (Kardinah and others 2014). Midwives ods limitations, noteworthy exceptions exist, including, and trained lay health workers were trained to per- for the most part, the studies included in annex 3A, form CBE; volunteers recruited women to come to tables 3A.1 through 3A.3. the clinics for screening. Among the 1,179 previously Yet even these conceptually strong studies reveal unscreened women, 289 had a suspicious finding on another important limitation: the range of intervention CBE and/or mammogram (24.5 percent) and required topics examined represents only a fraction of the impor- further work-up: 167 had an abnormal CBE and 191 tant questions in prevention and control. This issue had an abnormal mammogram. After work-up and becomes evident by comparing the interventions evalu- tissue sampling, 14 breast cancers (1.2 percent) were ated in annex 3A, tables 3A.1 through 3A.3, with the range diagnosed in this unscreened population. Of the 14 of BHGI-recommended interventions across the cancer cancers, 13 were detected by CBE. Mammography only continuum (tables 3.1, 3.3, 3.4, and 3.5). Although there identified one additional cancer not found by CBE. are several well-executed cost-effectiveness analyses on These findings suggest that when starting a screening early detection and screening (annex 3A, table 3A.1), we program in a previously unscreened population, most found no analyses of comparable scope and quality inves- of the prevalent cancers will be found by CBE; mam- tigating alternative breast cancer diagnostic techniques mography adds few additional cancer cases in the initial and procedures in LMICs. In the treatment domain, we screening phase. The study also demonstrates that a included a handful of excellent cost-effectiveness analyses large fraction of women (14 percent in this study) will examining alternative chemotherapy regimens, hor- require diagnostic evaluation beyond CBE. Screening monal therapy strategies, or radiation following breast programs based on CBE will require significant diag- cancer surgery (annex 3A, table 3A.2); there are no cost- nostic infrastructure based on additional imaging and effectiveness analyses examining alternative multimodal tissue sampling. adjuvant treatment strategies in LMICs. The cost-effective analyses focused on combina- Temporal Trends in Stage of Disease tion screening-treatment interventions (annex 3A, table 3A.3), while limited, do suggest that breast cancer In Malaysia, almost all diagnosed cancers are treated in a single referral hospital in the State of Sarawak, with a intervention packages consistent with the BHGI guidelines population of approximately two million. By reviewing could be evaluated through these analyses and tailored the medical records of all women with breast cancer at to the resource level that best characterizes the region, that hospital before and after an early detection initia- nation, or subnational arena for application. tive, it was possible to assess the impact of the program on breast cancer in the population (Devi, Tang, and Corbex 2007). The intervention consisted of the follow- FIELD STUDIES ing elements: The summarized evidence for the efficacy of mam- mography, CBE, and BSE is based largely on studies • Training community nurses who worked in rural in HICs and upper-middle-income countries, and the clinics to perform CBE and teach BSE results may not be directly applicable to LMICs. We • Circulating pamphlets and posters to motivate recommend strongly that early detection programs in women to go to their nearest clinic at the earliest LMICs be designed in advance in such a way that they signs of a breast problem Breast Cancer 61 • Instructing community nurses to hold health educa- Clinic-Based Cluster Randomized Trial tion talks and discussion groups on early diagnosis The National Cancer Institute of Colombia (Instituto during monthly visits to villages, to teach BSE and Nacional de Cancerología, or INC; 2006) adapted the perform CBE BHGI guidelines for MICs to develop a pilot screening • Strengthening the system for referring women with program in Bogotá. The INC guidelines recommend signs and symptoms of breast cancer to first-level screening with annual clinical CBE and mammography hospitals for diagnosis every two years for women ages 50–69. Based on these guidelines, the INC designed a pilot study to evaluate The proportion of breast cancers that were diag- opportunistic screening as a programmatic approach nosed at late stages (stage III or IV) was 77 percent in to improve early detection in the country (Murillo 1993 before the program began and 37 percent in 1998 and others 2008). Opportunistic screening is defined as after the program began. Since these statistics are for the systematic offer of CBE and mammography for all nearly all women in the population who were treated for women ages 50–69 who visit health centers on their own, breast cancer, regardless of whether they participated regardless of motivation. It implies that screening is in the program, they reflect the impact of the program clinic-based with no outreach outside the health centers. on the population and suggest that the program had a The primary objectives of the study are to evaluate the positive impact. effect of opportunistic screening on population cover- age, to determine the impact of opportunistic screening on clinical stage at diagnosis, and to identify the basic Population-Based Randomized Trial requirements for implementing opportunistic screening In a cluster randomized trial of CBE and cervical within the Colombian health system. cancer screening by visual inspection of the cervix A cluster randomized trial was undertaken with after acetic acid application (VIA) by lay women in health centers as the units of randomization. The Mumbai (Mittra and others 2010), 20 informal settle- screening consisted of recruitment in the clinics and ments (slums) were randomly allocated to screening or follow-up by health care assistants (auxiliary nurses, control groups (10 slums in each group) and women backed up by registered nurses), CBE performed by ages 35–64 years in each slum were considered eligible general practitioners, and mammography by radiolo- for the trial; more than 75,000 women were eligible gists and radiology technicians. Mammography quality for each arm of the study. Women with a tenth-grade control comprises examination and adjustment of education were trained to perform CBE and VIA; these mammography machines before starting screening, trained workers then invited women in the screening quality control of mammography films, and evalu- arm of the trial for screening; women in the control ation of mammography reading according to inter- arm received no screening. Three of four rounds of national standards. CBE quality control was done by screening at two-year intervals were completed between breast surgeons who periodically visit health centers 1998 and 2005. to evaluate general practitioners’ practice of CBE and The preliminary results showed that more breast differences between the diagnoses of the surgeons cancers were detected in the screening arm than in the and general practitioners are recorded. In the control control arm (125 versus 87 cases). The proportion of can- group, women who would be eligible for screening cers detected at early stages was higher in the screened arm if they had been in the intervention group are given than the control arm (62 percent versus 44 percent, stage 0, general information about breast cancer but are not I, or II disease); this is a difference that achieved statistical offered screening. difference by the third cycle of screening (p = 0.004). These Women attending the health centers regularly results indicate that CBE performed by specially trained were assigned to opportunistic screening or no women may be efficacious in reducing mortality from intervention, according to the random allocation of breast cancer in the slums of India; the results clearly indi- their clinic. Because the Colombian health system is cate that continuation of the trial is warranted to provide insurance-oriented, randomization was stratified by direct evidence for a reduction in breast cancer mortality health insurance company to control for the effect by CBE. The analyses of the data from this trial include of administrative factors on access to screening and all women and all breast cancers in the screening group, diagnosis. After the enrollment of approximately whether or not the women were actually screened. The 12,000 women (about 6,000 per arm), 88.9 percent results indicate the impact of screening as it was actually and 99.8 percent, respectively, in the opportunistic implemented in the target population. screening branch who were offered screening had a 62 Cancer mammogram and a CBE, compared with 11.7 percent Clinical Breast Examination as an Early and 5 percent, respectively, of the control women. The Detection Tool preliminary results show a threefold greater rate of Screening mammography is unfeasible in most of the breast cancer detection in the screened group than in world, but CBE is a practical option in many LMICs to the control group (15 and 5 cases, respectively) and provide at least basic breast cancer treatment. Evidence a higher proportion of cases at an early stage at diag- suggests that in LMICs with rising breast cancer nosis (13/15 = 86.6 percent versus 3/5 = 60 percent). incidence rates, CBE will help curb the rise in mortality. Furthermore, the enrollment rate (2.9 patients per What remains unknown is whether CBE, combined center per day) does not overburden the general practi- with highly effective treatment, can further curb breast tioners or other clinic staff. At completion of follow-up, cancer mortality and lower breast cancer mortality at it should be possible to compare the stage of disease at the population level. The increasing use of CBE provides diagnosis in the women screened compared with the opportunities for research on health care delivery in control arm. It will also be possible to compare the limited resource settings. stage distribution at diagnosis between the two groups, providing a measure of the impact of the screening program in the population of women served by the Diagnosis, Treatment, and Patient Triage clinics participating in the trial. Fine-needle aspiration and core needle biopsy tissue sampling techniques are necessary for diagnosing pal- pable lumps and distinguishing between benign and CONCLUSIONS malignant lesions. However, the key systems ques- tions go beyond selection of a tissue sampling tech- Population-Based Screening Mammography nology. The cost of needles, availability of pathology Breast cancer screening remains a major area of con- services, and patient selection to determine which troversy. In HICs, the debate persists about whether patients should be brought to second-level or third- mammographic screening leads to increased detection level care centers require systematic assessment of exist- of cancers that would not become significant threats to ing resources in a health care delivery system. Before an a woman’s life in her natural lifetime. Treatment of these early detection strategy is implemented, an assessment cancers would, by definition, constitute overtreatment. must be made of available resources, missing tools, Arguing against this premise is the fact that in nearly all and geographic distribution of the patient population, countries where age-adjusted breast cancer mortality is as well as social and cultural issues that could affect decreasing, screening mammography has been estab- patient participation. lished, for example, Australia, Denmark, Italy, Spain, Once a situation-appropriate early detection, diag- and the United States. In contrast, Japan has not imple- nosis, and patient triage strategy is devised, economic mented screening mammography; despite adequate evaluation becomes relevant and important. The evolu- treatment standards, breast cancer mortality has pla- tion of breast health systems will require pilot projects to teaued and stabilized, but it has not yet decreased. These determine what systems can work and, in parallel, it will observations are consistent with Cancer Intervention require new analyses to assess economic impact. It will and Surveillance Modeling Network (CISNET) model- be important to proceed in a stepwise, systematic fash- ing, which demonstrates synergy between screening and ion, documenting outcomes, so that successful models adjuvant treatment in reducing breast cancer mortality can be adapted and adopted in other settings. at the population level (Berry and others 2005); the modeling suggests that screening mammography will continue to be important in breast cancer control for the Economic Analyses and the Future of Breast Health foreseeable future. Care in LMICs The effectiveness of screening mammography could The BHGI resource-stratified guideline approach can be improved if it were possible to identify patients who be used in priority-setting analyses in LMICs. Cost- did not need immediate treatment, even though a lesion effectiveness analyses can identify interventions yielding is found. Although the debate continues regarding the the greatest gain in health (for example, life years gained age at which screening mammography should begin— and disability-adjusted life years averted) per dollar age 40 years or age 50 years—new biological research spent, from which practical strategies can be built. We on cancer progression may identify a patient subgroup recognize, however, the limitations in comparing and warranting a wait-and-watch strategy. extrapolating results from one country to another and Breast Cancer 63 among groups within the same population, depending includes all ages unless otherwise noted. Interventions also on the methodology, input data, and assumptions in the apply to all age groups, except where age ranges or cutoffs studies. A core principle in developing BHGI guidelines are specified. has been to recognize that one size does not fit all. The WHO-CHOICE framework has brought consis- tency, but these models are highly simplified compared REFERENCES with the breast cancer models used in HICs. In each case, Anderson, B. O., S. Braun, R. W. Carlson, J. R. Gralow, and decisions must be made: M. D. Lagios. 2003. “Overview of Breast Health Care Guidelines for Countries with Limited Resources.” The • Determining which analytical models to use (WHO- Breast Journal 9 (Suppl. 2): S42–50. CHOICE, CISNET, other micro-simulation models) Anderson, B. O., E. Cazap, N. S. El Saghir, C. H. Yip, • Finding appropriate data that are as faithful as possi- H. M. Khaled, and others. 2011. “Optimisation of Breast ble to the place of application Cancer Management in Low-Resource and Middle- • Sorting out at the policy level how to work in concert Resource Countries: Executive Summary of the Breast with national and local health agencies so the mod- Health Global Initiative Consensus, 2010.” The Lancet eling is not only academically sound but practical for Oncology 12 (4): 387–98. doi:S1470-2045(11)70031-6 [pii]; 10.1016/S1470-2045(11)70031-6. local decision making Anderson, B. O., R. Shyyan, A. Eniu, R. A. Smith, C. H. Yip, and others. 2006. “Breast Cancer in Limited-Resource The resource-stratified approach, although enor- Countries: An Overview of the Breast Health Global mously challenging, identifies key building blocks for Initiative 2005 Guidelines.” The Breast Journal 12 (Suppl. 1): cost-effectiveness analysis models that will translate S3–15. doi:10.1111/j.1075-122X.2006.00199.x. resource-conditional breast health guidelines into prior- Anderson, B. O., C. H. Yip, R. A. Smith, R. Shyyan, S. F. Sener, itized intervention strategies. and others. 2008. “Guideline Implementation for Breast Healthcare in Low-Income and Middle-Income Countries: Overview of the Breast Health Global Initiative Global NOTES Summit 2007.” Cancer 113 (Suppl. 8): 2221–43. doi:10.1002 /cncr.23844. The annex in this chapter is as follows. It is available at http:// Ashbury, F. D., C. Cameron, S. L. Mercer, M. Fitch, and dcp-3.org/cancer: E. Nielsen. 1998. “One-on-One Peer Support and Quality of Life for Breast Cancer Patients.” Patient Education and • Annex 3A: Summary of Salient Features and Counseling 35 (2): 89–100. doi:S0738399198000354 [pii]. Recommendations of the Cost-effectiveness Studies Azaiza, F., M. Cohen, M. Awad, and F. Daoud. 2010. “Factors Relevant to Breast Cancer Screening, Treatment, and Associated with Low Screening for Breast Cancer in Control in Low- and Middle-Income Countries. the Palestinian Authority: Relations of Availability, Environmental Barriers, and Cancer-Related Fatalism.” World Bank income classifications as of July 2014 are as Cancer 116 (19): 4646–55. doi:10.1002/cncr.25378. follows, based on estimates of gross national income (GNI) Bai, Y., M. Ye, H. Cao, X. Ma, Y. Xu, and others. 2012. “Economic per capita for 2013: Evaluation of Radiotherapy for Early Breast Cancer after Breast-Conserving Surgery in a Health Resource-Limited • Low-income countries (LICs): US$1,045 or less Setting.” Breast Cancer Research and Treatment 136 (2): • Middle-income countries (MICs) are subdivided: 547–57. doi:10.1007/s10549-012-2268-1. • Lower-middle-income: US$1,046–US$4,125 Barlow, W. E., C. Chi, P. A. Carney, S. H. Taplin, C. D’Orsi, • Upper-middle-income (UMICs): US$4,126–US$12,745 and others. 2004. “Accuracy of Screening Mammography • High-income countries (HICs): US$12,746 or more Interpretation by Characteristics of Radiologists.” Journal of the National Cancer Institute 96 (24): 1840–50. doi:10.1093 1. Maps and figures in this chapter are based on inci- /jnci/djh333. dence and mortality estimates for ages 0 to 69, consis- Baum, M. 2013. “Harms from Breast Cancer Screening tent with reporting in all DCP3 volumes. Global cancer Outweigh Benefits If Death Caused by Treatment Is statistics are estimates for the year 2012 and have been Included.” British Medical Journal 346: f385. doi:10.1136 provided by the International Agency for Research on /bmj.f385. Cancer from its GLOBOCAN 2012 database. Observable Berry, D. A., K. A. Cronin, S. K. Plevritis, D. G. Fryback, population-based data were derived from Cancer Incidence L. Clarke, and others. 2005. “Effect of Screening and in Five Continents, 10th edition, and for trends over time Adjuvant Therapy on Mortality from Breast Cancer.” from CI5 Plus (http://ci5.iarc.fr/CI5plus/Default.aspx). The New England Journal of Medicine 353 (17): 1784–92. The discussion of burden (including risk factors), however, doi:353/17/1784 [pii]; 10.1056/NEJMoa050518. 64 Cancer Bese, N. S., A. Munshi, A. Budrukkar, A. Elzawawy, and Eniu, A., R. W. Carlson, N. S. El Saghir, J. Bines, N. S. Bese, C. A. Perez. 2008. “Breast Radiation Therapy Guideline and others. 2008. “Guideline Implementation for Breast Implementation in Low- and Middle-Income Countries.” Healthcare in Low- and Middle-Income Countries: Cancer 113 (Suppl. 8): 2305–14. Treatment Resource Allocation.” Cancer 113 (Suppl. 8): Bray, F., J. S. Ren, E. Masuyer, and J. Ferlay. 2013. “Global 2269–81. doi:10.1002 /cncr.23843. Estimates of Cancer Prevalence for 27 Sites in the Adult Evans, D. B., T. T. Edejer, T. Adam, and S. S. Lim. 2005. “Methods Population in 2008.” International Journal of Cancer 132 (5): to Assess the Costs and Health Effects of Interventions 1133–45. doi:10.1002/ijc.27711. for Improving Health in Developing Countries.” British Breast Screening Frequency Trial Group. 2002. “The Frequency Medical Journal 331 (7525): 1137–40. doi:10.1136 of Breast Cancer Screening: Results from the UKCCCR /bmj.331.7525.1137. Randomised Trial. United Kingdom Co-Ordinating Ferlay, J., I. Soerjomataram, M. Ervik, R. Dikshit, S. Eser, and Committee on Cancer Research.” European Journal of others. 2013. “GLOBOCAN 2012 v1.0, Cancer Incidence Cancer 38 (11): 1458–64. and Mortality Worldwide.” IARC CancerBase No. 11 Brown, M. L., S. I. Goldie, G. Draisma, J. Harford, and [Internet]. International Agency for Research on Cancer. J. Lipscomb. 2006. “Health Service Interventions for http://globocan.iarc.fr. Cancer Control in Developing Countries.” In Disease Fonseca, M., G. T. Araujo, and E. D. Saad. 2009. “Cost- Control Priorities in Developing Countries, 2nd edition, Effectiveness of Anastrozole, in Comparison with Tamoxifen, edited by D. T. Jamison, J. G. Breman, A. R. Measham, in the Adjuvant Treatment of Early Breast Cancer in Brazil.” G. Alleyne, M. Claeson, D. B. Evans, P. Jha, A. Mills, and Revista da Associação Médica Brasileira 55 (4): 410–15. P. Musgrove, 569–89. Washington, DC: World Bank and Forouzanfar, M. H., K. J. Foreman, A. M. Delossantos, Oxford University Press. R. Lozano, A. D. Lopez, and others. 2011. “Breast and Cardoso, F., N. Bese, S. R. Distelhorst, J. L. Bevilacqua, Cervical Cancer in 187 Countries between 1980 and O. Ginsburg, and others. 2013. “Supportive Care During 2010: A Systematic Analysis.” The Lancet 378 (9801): Treatment for Breast Cancer: Resource Allocations in 1461–84. doi:S0140-6736(11)61351-2 [pii]; 10.1016/ Low- and Middle-Income Countries: A Breast Health S0140-6736(11)61351-2. Global Initiative 2013 Consensus Statement.” Breast 22 (5): Freedman, L. S., B. K. Edwards, L. A. G. Ries, and 593–605. doi:10.1016/j.breast.2013.07.050. J. L. Young. 2006. Cancer Incidence in Four Member Chia, K. S., M. Reilly, C. S. Tan, J. Lee, Y. Pawitan, and Countries (Cyprus, Egypt, Israel, and Jordan) of the Middle others. 2005. “Profound Changes in Breast Cancer East Cancer Consortium (MECC) Compared with US SEER. Incidence May Reflect Changes into a Westernized Lifestyle: NIH Pub. No. 06-5873. Bethesda, MD: National Cancer A Comparative Population-Based Study in Singapore and Institute. Sweden.” International Journal of Cancer 113 (2): 302–06. Ganz, P. A., C. H. Yip, J. R. Gralow, S. R. Distelhorst, K. S. Albain, Chopra, R. 2001. “The Indian Scene.” Journal of Clinical and others. 2013. “Supportive Care after Curative Treatment Oncology 19 (Suppl. 18): 106S–111S. for Breast Cancer (Survivorship Care): Resource Allocations Chu, K. C., C. R. Smart, and R. E. Tarone. 1988. “Analysis of in Low- and Middle-Income Countries: A Breast Health Breast Cancer Mortality and Stage Distribution by Age Global Initiative 2013 Consensus Statement.” Breast 22 (5): for the Health Insurance Plan Clinical Trial.” Journal of the 606–15. doi:10.1016/j.breast.2013.07.049. National Cancer Institute 80 (14): 1125–32. Ginsberg, G. M., J. A. Lauer, S. Zelle, S. Baeten, and R. Baltussen. Clarke, M. 2006. “Meta-Analyses of Adjuvant Therapies 2012. “Cost Effectiveness of Strategies to Combat Breast, for Women with Early Breast Cancer: The Early Breast Cervical, and Colorectal Cancer in Sub-Saharan Africa and Cancer Trialists’ Collaborative Group Overview.” Annals of South East Asia: Mathematical Modelling Study.” British Oncology 17 (Suppl. 10): x59–62. Medical Journal 344: e614. doi:10.1136/bmj.e614. Cleary, J., H. Ddungu, S. R. Distelhorst, C. Ripamonti, Gold, M. R., J. E. Siegel, L. B. Russell, and M. C. Weinstein. G. M. Rodin, and others. 2013. “Supportive and Palliative 1996. Cost-Effectiveness in Health and Medicine, 1st edition. Care for Metastatic Breast Cancer: Resource Allocations New York: Oxford University Press. in Low- and Middle-Income Countries: A Breast Health Gotzsche, P. C., K. J. Jorgensen, P. H. Zahl, and J. Maehlen. Global Initiative 2013 Consensus Statement.” Breast 22 (5): 2012. “Why Mammography Screening Has Not Lived up to 616–27. doi:10.1016/j.breast.2013.07.052. Expectations from the Randomised Trials.” Cancer Causes Curado, M. P., B. Edwards, H. R. Shin, J. H. Storm, J. Ferlay, and Control 23 (1): 15–21. doi:10.1007/s10552-011-9867-8. others. 2007. “Cancer Incidence in Five Continents.” Vol. 9, Greenlee, R. T., T. Murray, S. Bolden, and P. A. Wingo. IARC Scientific Publication 160, IARC Press, Lyon. 2000. “Cancer Statistics, 2000.” CA: A Cancer Journal for Devi, B. C., T. S. Tang, and M. Corbex. 2007. “Reducing Clinicians 50 (1): 7–33. by Half the Percentage of Late-Stage Presentation for Groot, M. T., R. Baltussen, C. A. Uyl-de Groot, B. O. Anderson, Breast and Cervix Cancer over 4 Years: A Pilot Study of and G. N. Hortobagyi. 2006. “Costs and Health Effects of Clinical Downstaging in Sarawak, Malaysia.” Annals of Breast Cancer Interventions in Epidemiologically Different Oncology 18 (7): 1172–76. doi:mdm105 [pii]; 10.1093 Regions of Africa, North America, and Asia.” The Breast /annonc/mdm105. Journal 12 (Suppl. 1): S81–90. Breast Cancer 65 Hammond, M. E., D. F. Hayes, M. Dowsett, D. C. Allred, Kuroishi, T., K. Hirose, T. Suzuki, and S. Tominaga. 2000. K. L. Hagerty, and others. 2010. “American Society of “Effectiveness of Mass Screening for Breast Cancer in Clinical Oncology/College of American Pathologists Japan.” Breast Cancer 7 (1): 1–8. Guideline Recommendations for Immunohistochemical Lee, S. G., Y. G. Jee, H. C. Chung, S. B. Kim, J. Ro, and others. Testing of Estrogen and Progesterone Receptors in Breast 2009. “Cost-Effectiveness Analysis of Adjuvant Therapy Cancer.” Archives of Pathology & Laboratory Medicine 134 for Node Positive Breast Cancer in Korea: Docetaxel, (6): 907–22. doi:10.1043/1543-2165-134.6.907. Doxorubicin and Cyclophosphamide (TAC) Versus Harford, J. B., I. V. Otero, B. O. Anderson, E. Cazap, Fluorouracil, Doxorubicin and Cyclophosphamide (FAC).” W. J. Gradishar, and others. 2011. “Problem Solving for Breast Cancer Research and Treatment 114 (3): 589–95. Breast Health Care Delivery in Low and Middle Resource doi:10.1007/s10549-008-0035-0. Countries (LMCs): Consensus Statement from the Breast Lee, S. Y., S. H. Jeong, Y. N. Kim, J. Kim, D. R. Kang, Health Global Initiative.” Breast 20 (Suppl. 2): S20–29. and others. 2009. “Cost-Effective Mammography Hisham, A. N., and C. H. Yip. 2003. “Spectrum of Breast Screening in Korea: High Incidence of Breast Cancer Cancer in Malaysian Women: Overview.” World Journal of in Young Women.” Cancer Science 100 (6): 1105–11. Surgery 27 (8): 921–23. doi:10.1111/j.1349-7006.2009.01147.x. Howell, A., E. Anderson, R. Blamey, R. B. Clarke, J. M. Dixon, Liubao, P., W. Xiaomin, T. Chongqing, J. Kamon, C. Gannong, and others. 1998. “The Primary Use of Endocrine Therapies.” and others. 2009. “Cost-Effectiveness Analysis of Adjuvant Recent Results in Cancer Research 152: 227–44. Therapy for Operable Breast Cancer from a Chinese Humphrey, L. L., M. Helfand, B. K. Chan, and S. H. Woolf. 2002. Perspective: Doxorubicin Plus Cyclophosphamide versus “Breast Cancer Screening: A Summary of the Evidence for Docetaxel Plus Cyclophosphamide.” Pharmacoeconomics the U.S. Preventive Services Task Force.” Annals of Internal 27 (10): 873–86. Medicine 137 (5 Part 1): 347–60. Love, R. R., N. B. Duc, D. C. Allred, N. C. Binh, N. V. Dinh, IARC (International Agency for Research on Cancer). 2008. and others. 2002. “Oophorectomy and Tamoxifen Adjuvant “Screening for Breast Cancer.” In World Cancer Report, Therapy in Premenopausal Vietnamese and Chinese edited by P. Boyle and B. Levin, 296–301. Lyon: IARC. Women with Operable Breast Cancer.” Journal of Clinical Ichikawa, L. E., W. E. Barlow, M. L. Anderson, S. H. Taplin, Oncology 20 (10): 2559–66. B. M. Geller, and others. 2010. “Time Trends in Radiologists’ Masood, S., L. Vass, J. A. Ibarra, Jr., B. M. Ljung, H. Stalsberg, and Interpretive Performance at Screening Mammography others. 2008. “Breast Pathology Guideline Implementation from the Community-Based Breast Cancer Surveillance in Low- and Middle-Income Countries.” Cancer 113 (Suppl. Consortium, 1996–2004.” Radiology 256 (1): 74–82. 8): 2297–304. doi:10.1002/cncr.23833. doi:10.1148/radiol.10091881. McCannon, C. J., D. M. Berwick, and M. R. Massoud. 2007. Instituto Nacional de Cancerología. 2006. Recomendaciones “The Science of Large-Scale Change in Global Health.” para la detección temprana de cáncer de mama en Colombia. Journal of the American Medical Association 298 (16): Guía de práctica clínica No.1. Bogotá: Instituto Nacional de 1937–39. Cancerología. McTiernan, A., P. Porter, and J. D. Potter. 2008. “Breast Cancer Jemal, A., R. Siegel, E. Ward, Y. Hao, J. Xu, and others. 2009. Prevention in Countries with Diverse Resources.” Cancer “Cancer Statistics, 2009.” CA: A Cancer Journal for Clinicians 113 (Suppl. 8): 2325–30. doi:10.1002/cncr.23829. 59 (4): 225–49. doi:caac.20006 [pii];10.3322/caac.20006. Mittra, I., G. A. Mishra, S. Singh, S. Aranke, P. Notani, and Jemal, A., A. Thomas, T. Murray, and M. Thun. 2002. “Cancer others. 2010. “A Cluster Randomized, Controlled Trial of Statistics, 2002.” CA: A Cancer Journal for Clinicians 52 (1): Breast and Cervix Cancer Screening in Mumbai, India: 23–47. Methodology and Interim Results after Three Rounds of Johannesson, M., and M. C. Weinstein. 1993. “On the Decision Screening.” International Journal of Cancer 126 (4): 976–84. Rules of Cost-Effectiveness Analysis.” Journal of Health doi:10.1002/ijc.24840. Economics 12 (4): 459–67. Morrow, M., E. A. Strom, L. W. Bassett, D. D. Dershaw, B. Fowble, Kalager, M., M. Zelen, F. Langmark, and H. O. Adami. 2010. and others. 2002. “Standard for Breast Conservation “Effect of Screening Mammography on Breast-Cancer Therapy in the Management of Invasive Breast Carcinoma.” Mortality in Norway.” The New England Journal of Medicine CA: A Cancer Journal for Clinicians 52 (5): 277–300. 363 (13): 1203–10. doi:10.1056/NEJMoa1000727. Murillo, R., S. Díaz, O. Sánchez, F. Perry, M. Piñeros, and Kanemura, S., I. Tsuji, N. Ohuchi, H. Takei, T. Yokoe, and others. 2008. “Pilot Implementation of Breast Cancer Early others. 1999. “A Case Control Study on the Effectiveness of Detection Programs in Colombia.” Breast Care 3: 29–32. Breast Cancer Screening by Clinical Breast Examination in Murray, C. J. L., and A. D. Lopez. 1996. “The Global Burden Japan.” Japanese Journal of Cancer Research 90 (6): 607–13. of Disease: A Comprehensive Assessment of Mortality and Kardinah, D., B. O. Anderson, C. Duggan, I. A. Ali, and Disability from Diseases, Injuries, and Risk Factors in 1990 D. B. Thomas. 2014. “Short Report: Limited Effectiveness and Projected to 2020.” In Global Burden of Disease and of Screening Mammography in Addition to Clinical Breast Injury. Cambridge, MA: Harvard School of Public Health. Examination by Trained Nurse Midwives in Rural Jakarta, Nelson, H. D., K. Tyne, A. Naik, C. Bougatsos, B. K. Chan, Indonesia.” International Journal of Cancer 134 (5): 1250–55. and others. 2009. “Screening for Breast Cancer: An doi:10.1002/ijc.28442. Update for the U.S. Preventive Services Task Force.” 66 Cancer Annals of Internal Medicine 151 (10): 727–37, W237–42. Smith, R. A. 2000. “Breast Cancer Screening among Women doi:10.7326/0003-4819-151-10-200911170-00009. Younger Than Age 50: A Current Assessment of the Issues.” Nguyen, L. H., W. Laohasiriwong, J. F. Stewart, P. Wright, CA: A Cancer Journal for Clinicians 50 (5): 312–36. Y. T. B. Nguyen, and others. 2013. “Cost-Effectiveness Soo, R. A., B. O. Anderson, B. C. Cho, C. H. Yang, M. Liao, and Analysis of a Screening Program for Breast Cancer in others. 2009. “First-Line Systemic Treatment of Advanced Vietnam.” Value in Health Regional Issues 2(1): 21–28. Stage Non-Small-Cell Lung Cancer in Asia: Consensus Okonkwo, Q. L., G. Draisma, A. der Kinderen, M. L. Brown, Statement from the Asian Oncology Summit 2009.” and H. J. de Koning. 2008. “Breast Cancer Screening Policies The Lancet Oncology 10 (11): 1102–10. doi:10.1016 in Developing Countries: A Cost-Effectiveness Analysis for /S1470-2045(09)70238-4. India.” Journal of the National Cancer Institute 100 (18): Tan-Torres Edejer, T., R. Baltussen, T. Adam, R. Hutubessy, 1290–300. doi:10.1093/jnci/djn292. A. Acharya, and others. 2003. Making Choices in Health: Parkin, D. M., and L. M. Fernandez. 2006. “Use of Statistics WHO Guide to Cost-Effectiveness Analysis. Geneva: WHO. to Assess the Global Burden of Breast Cancer.” The Breast Tangjitgamol, S., B. O. Anderson, H. T. See, C. Lertbutsayanukul, Journal 12 (Suppl. 1): S70–80. N. Sirisabya, and others. 2009. “Management of Endometrial Perloff, M., G. J. Lesnick, A. Korzun, F. Chu, J. F. Holland, Cancer in Asia: Consensus Statement from the Asian and others. 1988. “Combination Chemotherapy with Oncology Summit 2009.” The Lancet Oncology 10 (11): Mastectomy or Radiotherapy for Stage III Breast Carcinoma: 1119–27. doi:10.1016/S1470-2045(09)70290-6. A Cancer and Leukemia Group B Study.” Journal of Clinical Theriault, R. L., R. W. Carlson, C. Allred, B. O. Anderson, Oncology 6 (2): 261–69. H. J. Burstein, and others. 2013. “Breast Cancer, Version Pinto, A. C., F. Ades, E. de Azambuja, and M. Piccart-Gebhart. 3.2013: Featured Updates to the NCCN Guidelines.” Journal 2013. “Trastuzumab for Patients with HER2 Positive of the National Comprehensive Cancer Network 11 (7): Breast Cancer: Delivery, Duration and Combination 753–60; quiz 761. Therapies.” Breast 22 (Suppl. 2): S152–55. doi:10.1016/j Thomas, D. B., D. L. Gao, R. M. Ray, W. W. Wang, C. J. Allison, .breast.2013.07.029. and others. 2002. “Randomized Trial of Breast Self- Pisani, P., D. M. Parkin, C. Ngelangel, D. Esteban, L. Gibson, Examination in Shanghai: Final Results.” Journal of the and others. 2006. “Outcome of Screening by Clinical National Cancer Institute 94 (19): 1445–57. Examination of the Breast in a Trial in the Philippines.” Thomas, D. B., R. H. Murillo, Kardinah, and B. O. Anderson. International Journal of Cancer 118 (1): 149–54. 2013. “Breast Cancer Early Detection and Clinical Poon, D., B. O. Anderson, L. T. Chen, K. Tanaka, Guidelines.” In Cancer Epidemiology: Low and Middle W. Y. Lau, and others. 2009. “Management of Hepatocellular Income Countries and Special Populations, edited by Carcinoma in Asia: Consensus Statement from the Asian A. S. Soliman, D. Schottenfeld, and P. Boffetta, 378–95. New Oncology Summit 2009.” The Lancet Oncology 10 (11): York: Oxford University Press. 1111–18. doi:S1470-2045(09)70241-4 [pii]; 10.1016 Thorat, M. A., A. Rangole, M. S. Nadkarni, V. Parmar, and /S1470-2045(09)70241-4. R. A. Badwe. 2008. “Revision Surgery for Breast Cancer: Porter, P. 2008. “‘Westernizing’ Women’s Risks? Breast Cancer Single-Institution Experience.” Cancer 113 (Suppl. 8): in Lower-Income Countries.” The New England Journal of 2347–52. doi:10.1002/cncr.23839. Medicine 358 (3): 213–16. Thornton, H. 2014. “Bringing an End to Mandatory Breast Salomon, J. A., N. Carvalho, C. Gutierrez-Delgado, R. Orozco, Cancer Screening in Uruguay.” British Medical Journal 348: A. Mancuso, and others. 2012. “Intervention Strategies g390. doi:10.1136/bmj.g390. to Reduce the Burden of Non-Communicable Diseases U.S. Preventive Services Task Force. 2009. “Screening for in Mexico: Cost Effectiveness Analysis.” British Medical Breast Cancer: U.S. Preventive Services Task Force Journal 344: e355. doi:10.1136/bmj.e355. Recommendation Statement.” Annals of Internal Sant, M., C. Allemani, F. Berrino, M. P. Coleman, T. Aareleid, Medicine 151 (10): 716–26, W-236. doi:151/10/716 [pii]; and others. 2004. “Breast Carcinoma Survival in Europe 10.1059/0003-4819-151-10-200911170-00008. and the United States.” Cancer 100 (4): 715–22. Visco, F. 2007. “The National Breast Cancer Coalition: Setting Schmidt, C. 2009. “Komen/ASCO Program Aims to Swell the Standard for Advocate Collaboration in Clinical Trials.” Ranks of Minority Oncologists.” Journal of the National Cancer Treatment and Research 132: 143–56. Cancer Institute 101 (4): 224–25, 227. Wee, J. T., B. O. Anderson, J. Corry, A. D’Cruz, K. C. Soo, Semiglazov, V. F., V. M. Moiseyenko, A. G. Manikhas, and others. 2009. “Management of the Neck after S. A. Protsenko, R. S. Kharikova, and others. 1999. “Role Chemoradiotherapy for Head and Neck Cancers in Asia: of Breast Self-Examination in Early Detection of Breast Consensus Statement from the Asian Oncology Summit Cancer: Russia/WHO Prospective Randomized Trial in St. 2009.” The Lancet Oncology 10 (11): 1086–92. doi:S1470 Petersburg.” Cancer Strategy 1: 145–51. -2045(09)70266-9 [pii]; 10.1016/S1470-2045(09)70266-9. Shyyan, R., S. F. Sener, B. O. Anderson, L. M. Garrote, Weir, H. K., M. J. Thun, B. F. Hankey, L. A. Ries, H. L. Howe, and G. N. Hortobagyi, and others. 2008. “Guideline others. 2003. “Annual Report to the Nation on the Status of Implementation for Breast Healthcare in Low- and Middle- Cancer, 1975–2000, Featuring the Uses of Surveillance Data Income Countries: Diagnosis Resource Allocation.” Cancer for Cancer Prevention and Control.” Journal of the National 113 (Suppl. 8): 2257–68. doi:10.1002/cncr.23840. Cancer Institute 95 (17): 1276–99. Breast Cancer 67 Winn, R. J., and W. Z. Botnick. 1997. “The NCCN Guideline Postoperative Adjuvant Tamoxifen Therapy for Korean Program: A Conceptual Framework.” Oncology (Williston Women with Breast Cancer: Retrospective Cohort Park) 11 (11A): 25–32. Study of the Korean Breast Cancer Society Database.” WHO (World Health Organization). 2001. Macroeconomics Clinical Therapeutics 32 (6): 1122–38. doi:10.1016/j and Health: Investing in Health for Economic Development. .clinthera.2010.05.013. Geneva: WHO. Yip, C. H., R. A. Smith, B. O. Anderson, A. B. Miller, D. B. Thomas, ———. 2002. “Executive Summary.” In National Cancer and others. 2008. “Guideline Implementation for Breast Control Programmes: Policies and Managerial Guidelines, Healthcare in Low- and Middle-Income Countries: Early 1–24. Geneva: WHO. Detection Resource Allocation.” Cancer 113 (Suppl. 8): Wong, N. S., B. O. Anderson, K. S. Khoo, P. T. Ang, C. H. Yip, 2244–56. doi:10.1002/cncr.23842. and others. 2009. “Management of HER2-Positive Breast Yoo, C., J. H. Ahn, K. H. Jung, S. B. Kim, H. H. Kim, and others. Cancer in Asia: Consensus Statement from the Asian 2012. “Impact of Immunohistochemistry-Based Molecular Oncology Summit 2009.” The Lancet Oncology 10 (11): Subtype on Chemosensitivity and Survival in Patients with 1077–85. doi:10.1016/S1470-2045(09)70230-X. Breast Cancer Following Neoadjuvant Chemotherapy.” Wong, I. O., B. J. Cowling, C. M. Schooling, and G. M. Leung. Journal of Breast Cancer 15 (2): 203–10. doi:10.4048 2007. “Age-Period-Cohort Projections of Breast Cancer /jbc.2012.15.2.203. Incidence in a Rapidly Transitioning Chinese Population.” Zelle, S. G., and R. M. Baltussen. 2013. “Economic Analyses International Journal of Cancer 121 (7): 1556–63. of Breast Cancer Control in Low- and Middle-Income Wong, I. O., K. M. Kuntz, B. J. Cowling, C. L. Lam, and Countries: A Systematic Review.” Systematic Reviews 2: 20. G. M. Leung. 2007. “Cost Effectiveness of Mammography doi:10.1186/2046-4053-2-20. Screening for Chinese Women.” Cancer 110 (4): 885–95. Zelle, S. G., K. M. Nyarko, W. K. Bosu, M. Aikins, L. M. Niens, and doi:10.1002/cncr.22848. others. 2012. “Costs, Effects and Cost-Effectiveness of Breast Yang, J. J., S. K. Park, L. Y. Cho, W. Han, B. Park, and Cancer Control in Ghana.” Tropical Medicine & International others. 2010. “Cost-Effectiveness Analysis of 5 Years of Health 17 (8): 1031–43.doi:10.1111/j.1365-3156.2012.03021.x. 68 Cancer Cervical Cancer Chapter 4 Lynette Denny, Rolando Herrero, Carol Levin, and Jane J. Kim INTRODUCTION BURDEN OF CERVICAL CANCER1 Cervical cancer, a largely preventable disease, is one of Global Burden of Disease the most common cancers found in women living in Cervical cancer, caused by HPV, is the third leading low- and middle-income countries (LMICs). A striking malignancy among women in the world, after breast reduction in the incidence of and mortality from cervi- cancer and colorectal cancer, with an estimated 527,624 cal cancer occurred in the past century in those countries new cases and 265,653 deaths in 2012 (Ferlay and others that were able to establish successful national screening 2013). Incidence and mortality rates have been declining programs. These programs relied on cytology-based in most areas of the world in the past 30 years, at a world- Papanicolaou smears to identify cervical cancer precur- wide rate of about 1.6 percent per year (Forouzanfar sors that can be removed before progressing to invasive and others 2011). This decline is a result of increased cancer. Prevention of up to 91 percent of all invasive access to health services, reductions in some risk factors cervical cancers has been achieved in countries able to (such as fertility rates), improvements in treatment, and implement widespread cytology-based screening. successful cytology-based screening programs. However, However, these programs are expensive and require more than 80 percent of cases and 88 percent of deaths robust and well-funded health care systems. Few LMICs occur in LMICs. Cervical cancer is still the leading have initiated or sustained cytology-based cervical can- cancer in women in many LMICs; some areas report cer prevention programs, and these countries experience recent increases in rates, including several economies in very high incidence and mortality rates. The unequal Europe and Central Asia (Arbyn and others 2011). burden of cervical cancer is an example of the impact A striking characteristic of cervical cancer is its varia- of unequal access to health care. Fortunately, alternative tion by country, with a generally strong inverse correla- strategies to prevent cervical cancer have been investi- tion between the level of development and the incidence gated and extensively evaluated in these settings. The and mortality. Survival once the disease has developed recent introduction of two commercially available vac- is also much better in richer than in poorer countries. cines against human papillomavirus (HPV) has offered Figure 4.1 shows trends of incidence and mortality in the possibility of primary prevention of cervical cancer. selected countries. This chapter focuses on these innovations. Corresponding author: Lynette Denny, MBChB (UCT), MMED (O&G), PhD, FCOG (SA), lynette.denny@uct.ac.za 69 Figure 4.1 Trends of Age-Standardized Rates of Cervical Cancer Incidence and Mortality in Selected Countries a. Low- and lower-middle-income countries b. High- and upper-middle-income countries 50 50 40 40 30 30 20 20 Age-standardized (world) rate per 100,000, 0–69 years Age-standardized (world) rate per 100,000, 0–69 years 15 15 10 10 7 7 5 5 3 3 2 2 1.5 1.5 1 1 1980 1990 2000 2010 1980 1990 2000 2010 Calendar year Calendar year Incidence Mortality Uganda India Philippines Brazil United States: United States: Black White Thailand Philippines China Spain Brazil China United States: United States: White Black Spain Source: CI5plus (http://ci5.iarc.fr/CI5plus/Default.aspx) and WHO Mortality Database (http://www.who.int/healthinfo/statistics/mortality_rawdata/en/index.html). Note: Data for the economies in the graphs are for Uganda (Kampala), Thailand (Chiang Mai), Philippines (Manila), India (Chennai and Mumbai), Brazil (Goiâna), Spain (Granada, Murcia, Navarra, and Tarragona), China (Hong Kong SAR, China, and Shanghai), and the United States (Surveillance, Epidemiology, and End Results Program). All available data for these economies are shown. 70 Cancer Regional Burden of Disease the rates of incidence and mortality between national Despite the declining global rates, the number of new income groupings. cases and deaths has increased constantly by about Even within HICs, the highest incidence and 0.5 percent per year because of population aging. mortality rates are among the poorest or most mar- With no new intervention, the increase will continue, ginalized women. For example, in the United States, particularly in LMICs where the life expectancy of where the average rates are low and cervical cancer women is improving. For example, in Latin America has consistently declined in recent decades, strong and the Caribbean, the estimated number of new disparities still exist by race and socioeconomic status cases is likely to increase by 75 percent between 2002 (Singh 2012), reflecting the variability in accessibility and 2025 if the incidence rates remain at 2002 levels, of services. The other notable characteristic of cer- because of population growth and aging alone (Parkin vical cancer is that it affects relatively young women and others 2008). who often have many children and are frequently The disease is strongly influenced by cultural and sole providers. The median age at death for women religious practices that govern sexual behavior and with cervical cancer is 54 years; the burden of disease transmission of HPV. Sub-Saharan Africa has the high- among women under age 40 years is high compared est estimated rates of cervical cancer; in Guinea, Malawi, with other cancers, because of the large numbers of and Zambia, the age-standardized incidence rate is over women in these age groups in LMICs and the fact that 50 per 100,000 (Arbyn and others 2011). In contrast, in cervical cancer rates begin to rise at younger ages than countries in the Middle East and North Africa, such as other cancers. Algeria, the Arab Republic of Egypt, Libya, Sudan, and Because cervical cancer affects relatively young Tunisia, where sexual behaviors are more conservative, women, it ranks highest among cancers according the recorded incidence rates are below 10 per 100,000 to a disability-adjusted life years (DALYs) metric. women. In high-income countries (HICs), rates are even In a recent study, DALYs caused by cervical cancer lower, at about 5 per 100,000 women. ranged from 84 per 100,000 women in areas with a In Latin America and the Caribbean, Guyana, very high Human Development Index (HDI)2 to 595 Honduras, Jamaica, and Nicaragua have rates around 40 per 100,000 in areas with a low HDI (Soerjomataram per 100,000. In Asia, the highest rates are in Bangladesh, and others 2012). Breast cancer DALYs ranged from a Cambodia, India, and Nepal. high of 566 age-adjusted DALYs per 100,000 in pop- Map 4.1 shows the incidence of cervical cancer by ulations with a very high HDI to 387 in those with a country in 2012; figure 4.2 shows the contrast between low HDI. Map 4.1 Age-Standardized Cervical Cancer Incidence Rates, 2012 IBRD 41381 | FEBRUARY 2015 Cervcan Cancer in Women, 0–69 years ASR (World) Indcidence Greater than 20 18.3–20 12.8–18.3 7.4–12.8 Less than 7.4 No data Source: Ferlay and others 2013. Note: ASR = age-standardized rate. Cervical Cancer 71 Figure 4.2 Age-Standardized Cervical Cancer Incidence and Mortality Rates per 100,000 Women, by World Bank Income Group Low-income Lower-middle-income (except India) Upper-middle-income (except China) High-income China India 0 10 20 30 Age-standardized rate (world) per 100,000, 0–69 years Incidence Mortality Source: Ferlay and others 2013. NATURAL HISTORY OF CERVICAL CANCER (n = 8,977) of the cases were positive for HPV DNA (de Sanjose and others 2010). HPV types 16, 18, and 45 were The natural history has been studied extensively, the three most common types in each histologic form and persistent infection of the cervix with certain of cervical cancer (squamous cell, adenocarcinoma, and high-risk types of HPV has been well established as adenosquamous carcinoma), accounting for 61 percent, a necessary cause of cervical cancer (Walboomers 10 percent, and 6 percent, respectively. and others 1999). HPV is a very common sexually Good evidence suggests that HPV infection pre- transmitted infection, usually acquired soon after ini- cedes the development of cervical cancer by decades tiation of sexual activity. Most HPV infections clear and that persistent infection with HPV is necessary spontaneously within one to two years; those that for the development and progression of precancerous persist, particularly high-risk types of HPV (includ- lesions of the cervix, either to higher grades of precan- ing HPV 16 and 18), may progress to cervical cancer cerous disease or to cancer. Cervical cancer progresses precursors, and ultimately to invasive cervical cancer. slowly from a preinvasive state to invasive cervical High-risk types of HPV are identified in nearly all cancer, a process that can take 10–30 years (Wright cancers of the cervix, and the relative risk of cervical and Kurman 1994). cancer associated with persistent, ongoing infection However, HPV infections are very common, par- with high-risk types of HPV is higher than the risk of ticularly among young women (Herrero and others lung cancer associated with smoking. HPV 16 and 18 2005), where the majority of infections are likely to are responsible for about 70 percent of cases world- regress spontaneously as a result of activation of the wide (http://www.iarc.fr). There is little geographic immune system. variation in the predominant HPV types associated with cervical cancer. A study that evaluated HPV infection in 10,575 Cervical Cytology Classification and Terminology histologically confirmed cases of invasive cancer from In 1988, the Bethesda classification of cytology was 38 countries in Asia, Europe, Latin America and the adopted and has been revised several times (National Caribbean, North America, Oceania, and Sub-Saharan Cancer Workshop 1989). The latest consensus guidelines Africa over a 60-year period found that 85 percent for the management of abnormal cytology in the United 72 Cancer States were published in 2013 (Massad and others 2013) are better suited to low-resource settings. Depending and can be accessed at http://www.asccp.org. mainly on the target age group and frequency of screening, these tools may be effective in reducing cervical cancer rates. The new interventions include Cervical Cancer and Infection with Human the following: Immunodeficiency Virus Women infected with human immunodeficiency virus • Screening with visual inspection with acetic acid (HIV) have an increased risk of being infected with HPV (VIA) and are at increased risk for cervical cancer. Studies have • Screening with HPV DNA testing consistently shown higher prevalence of HPV infection, • Treatment with ablative techniques (cryotherapy and more persistent infections with HPV, greater infections cold coagulation) with multiple types of HPV, and higher prevalence • Treatment using excisional techniques, called loop of cervical cancer precursors in HIV-infected women electrocautery excision procedure (LEEP), also (Ellerbrock and others 2000; Harris and others 2005; known as large loop excision of the transformational Palefsky and others 1999). The Rwandan Women’s zone (LLETZ), and cone biopsy. Interassociation Study and Assessment is an observa- tional prospective cohort study of 710 HIV-positive and Impact of Cervical Cytology-Based Screening 226 HIV-negative Rwandan women enrolled in 2005 Programs (Singh and others 2009). The prevalence of HPV was significantly higher in the HIV-positive group overall Cytology-based cervical cancer screening, which began and in each 10-year age group. Forty-six percent of in the early 1960s in the Scandinavian countries, was HIV-positive women had high-risk types of HPV and not evaluated in randomized trials to assess the impact 35 percent were infected with multiple types, both of of screening on cervical cancer incidence or mortality. which were associated with a higher risk of abnormal The marked reduction in cervical cancer incidence and cytological findings. mortality after cytology-based screening programs were The association with HIV is important because initiated in a variety of LMICs was interpreted as strong integrating cervical cancer prevention strategies with nonexperimental support for organized cervical cancer chronic care for HIV-positive women is essential to screening programs. maximizing the health-giving benefit of antiretroviral The International Agency for Research on Cancer therapy. In many countries in Sub-Saharan Africa, (IARC) conducted a comprehensive analysis of data antiretroviral therapy is free, but cervical cancer screen- from several of the largest screening programs in the ing and treatment are not. world in 1986; the analysis showed that well-organized, cytology-based screening programs were effective in reducing cervical cancer incidence and mortality SECONDARY PREVENTION OF CERVICAL (Hakama 1986). In the Nordic countries, following the introduction of nationwide screening in the 1960s, CANCER THROUGH SCREENING mortality rates from cervical cancer fell between 84 and Historically, cervical cancer screening, also known as 11 percent, respectively, corresponding to the country secondary prevention of cervical cancer, was based on with the shortest screening interval and widest age examining cells collected from the surface of the cervix range (Iceland) and to the country with only 5 percent by Pap smear (cytology), followed by colposcopy for population coverage by an organized screening pro- women with abnormal smears and histological assess- gram (Norway) (Laara, Day, and Hakama 1987). ment, followed by surgical treatment for histologically Further, the age-specific trends indicated that the tar- proven cancer precursors. This approach resulted in get age range of a screening program was a more impor- dramatic reductions in cervical cancer incidence and tant determinant of risk reduction than the frequency mortality in health systems that were robust enough to of screening within that age range. This finding was support relatively complex screening programs effec- in agreement with the estimates of the IARC Working tively. However, very few LMICs have been able to Group on Cervical Cancer Screening that for interscreen initiate or sustain cytology-based screening programs intervals of up to five years, the protective effect of because of lack of adequate resources or health care or organized screening exceeded 80 percent throughout the laboratory infrastructure. targeted age group (IARC Working Group on Cervical For screening and treatment of precancerous Cancer Screening 1986a, 1986b). It is clear that the extent lesions, several new tools have been developed that to which screening programs have succeeded or failed Cervical Cancer 73 to decrease the incidence of and mortality from cervical Alternative Approaches to Cytology for Cervical cancer is largely a function of three factors: Cancer Screening Visual Inspection with Acetic Acid • The extent of screening coverage of the population VIA involves applying a 3–5 percent acetic acid solution at risk to the cervix and then examining it with the naked eye • The target age of women screened using a bright light source. No expensive equipment or • The reliability of cytology services in the program. supplies are needed, and screening takes less than five minutes. A well-defined aceto-white area close to the Gakidou, Nordhagen, and Obermeyer (2008) evalu- transformation zone indicates a positive test. ated screening programs in 57 countries and found that VIA is inexpensive and simple and can be carried out the levels of effective screening coverage using cytology by primary care staff. Most important, VIA provides an vary widely across countries, from over 80 percent immediate result that can be used to decide on treat- in Austria and Luxembourg to less than 1 percent in ment, usually with cryotherapy, which requires training Bangladesh, Ethiopia, and Myanmar. Many women in but no surgery or anesthetic. low-income countries (LICs) had never had a pelvic It is difficult to recommend VIA unconditionally, examination. This proportion of women is largest in however, because its sensitivity and specificity are lower Bangladesh, Ethiopia, and Malawi, where more than than those of other screening methods (table 4.1). VIA 90 percent of women report never having had a pelvic sensitivity and specificity are variable, because they are examination, compared with 9 percent of women liv- highly dependent on the training and skill of the staff ing in the richest global wealth decile. Although crude carrying out the examinations. The accuracy of the test coverage rates are high for women in the richest wealth decreases with the increasing age of the women screened. deciles, effective coverage rates are overall low, with rates In cross-sectional studies, the sensitivity and specificity of around 60 percent and less than 10 percent in the of VIA compared favorably with cytology in detecting poorest countries. high-grade cervical cancer precursor lesions and cervi- Screening efforts have failed to produce the expected cal cancer. Sensitivity has varied from 49 to 96 percent, reductions in cervical cancer mortality in many places, and specificity has varied from 49 to 98 percent (Denny, even when large numbers of Pap smears were performed, Quinn, and Sankaranarayanan 2006). However, many of because the wrong women have been screened (for these studies suffer from verification bias, where the true example, younger women attending antenatal clinics), status of disease in test-negative women is unknown. coverage of the most at-risk population was too low Sauvaget and others (2011) performed a meta-analysis (that is, women ages 35–64 years), the quality of cervical of 26 studies of VIA with confirmatory testing, using smears was poor (Irwin, Oberle, and Rosero-Bixby 1991; high-grade squamous intraepithelial lesions (HSIL) as Lazcano-Ponce and others 1994; Sankaranarayanan and the disease threshold. Sauvaget and others (2011) report a Pisani 1997), and follow-up of screen-positive women was sensitivity of 80 percent specificity (range 79–82 percent) incomplete. In all cases, funds were spent for little gain. and 92 percent specificity (range 91–92 percent) for Table 4.1 Performance and Characteristics of Screening Methods Screening test Sensitivity Specificity Characteristics Conventional cytology Moderate High (91–96%) Adequate health care infrastructure required; laboratory based; stringent (44–78%) training and quality control required HPV DNA testing High (66–100%) Moderate Laboratory based; high throughput; objective, reproducible, and robust; (61–96%) currently expensive Visual inspection methods • VIA Moderate Low (49–86%) Low technology; low cost (67–79%) • VIAM Moderate Low (86–87%) Linkage to immediate treatment possible; suitable for low-resource settings (62–73%) Colposcopy Low (44–77%) Low (85–90%) Expensive; inappropriate for low-resource settings Source: Ranges of sensitivity and specificity adapted from Cuzick and others 2008. Note: HPV = human papillomavirus; VIA = visual inspection with acetic acid; VIAM = magnified visual inspection with acetic acid. 74 Cancer VIA, with a positive predictive value of 10 percent. HPV Testing They conclude that in very low-resource settings where Highly sensitive and reproducible laboratory techniques the infrastructure for laboratory-based testing is not to detect oncogenic HPV and cervical cancer have been available, VIA is a reasonable alternative to cytology. developed and are being used or considered in place However, in more recent randomized studies, VIA has of cervical cytology for primary screening, in addition performed less well. to other potential uses (Cuzick and others 2008). The Despite its limitations, the possibility of immediate cervix is sampled with a brush, which is inserted into diagnosis and treatment makes VIA the only possible the endocervix and then removed and placed in a tube alternative in many low-resource settings. One poten- containing special transport media. The U.S. Food and tial use of VIA that would have a significant impact is Drug Administration has approved five of the many tests following an HPV test, for HPV-positive women only, available for routine laboratory service: to make treatment decisions. The utility of VIA in this context is promising but yet to be proven. • Hybrid Capture 2 detects 13 oncogenic types of HPV (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, and 68). Case Study of Upscaling VIA • Cervista HPV HR detects 14 HPV types (16, 18, 31, From 2005 through 2009, the World Health Organization 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, and 68). (WHO) sponsored a VIA demonstration project in six • Cervista HPV 16/18 detects only HPV 16 and 18. Sub-Saharan African countries: Madagascar, Malawi, • Aptima (transcription–mediated amplification test) Nigeria, Tanzania, Uganda, and Zambia (WHO 2012). detects RNA from 14 HPV types (16, 18, 31, 33, 35, In all, 19,579 women were screened with VIA. Of these, 39, 45, 51, 52, 56, 58, 59, 66, and 68). 1,980 were VIA-positive (11.5 percent); cancer was • Cobas 4800 (real-time polymerase chain reaction suspected in 326 (1.7 percent). Of the VIA-positive women, [PCR]–based test) detects 14 HPV types (16, 18, 31, 1,737 were eligible for cryotherapy (87.7 percent); of these, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, and 68). 1,058 (60.9 percent) were treated, 601 (34.6 percent) were lost to follow-up, and 78 women were not treated. Of the Other tests that use PCR technology are being used in women treated, 243 (39.1 percent) were treated during the many clinical studies. same visit as the screening. HPV testing is an excellent alternative to cytology for No information was available for 230 of the 326 cervical cancer screening (Arbyn and others 2012). In women in whom cancer was suspected (70.5 percent); meta-analyses of cross-sectional studies, the sensitivity of the 96 women investigated, cancer was confirmed of the Hybrid Capture 2 (HC2) DNA test, the most in 79, but no staging information was recorded; 77 of the commonly used test, was 90 percent to detect CIN2+ and women were treated, mostly with radiation. 95 percent to detect CIN3+, with more heterogeneity in This is an interesting study of “real world” VIA studies from LMICs. Compared with cytology, the sen- screening, with all of the difficulties of any screening sitivity of HC2 is 23–46 percent higher on average, and program, even with a test as simple as VIA. These the specificity is 3–8 percent lower (note we are using difficulties range from achieving adequate coverage; the terminology as reported by the authors, hence the to losing to follow-up the large number of women switch between cervical intraepithelial neoplasia (CIN) needing treatment (only 60 percent of eligible women and squamous intraepithelial lesion (SIL) terminology). were treated); to treating women on the same day as Another advantage of HPV testing is the possibility screening (“screen and treat”), which occurred for less of linking screening to treatment without colposcopy than 40 percent of the women. The failure to refer over or prior histological sampling, particularly once either 70 percent of women with suspicious lesions for further simplified or point-of-care HPV tests are developed. evaluation—possibly because cervical biopsy is not a A randomized screening trial to evaluate safety investi- free service in any of these countries and most women gated the acceptability and efficacy of screening women could not afford to pay—is disturbing. The greatest and treating those with positive tests without colpos- utility of VIA in countries that cannot afford any copy and histological sampling (Denny and others alternative is to establish the necessary infrastructure 2010). A total of 6,555 previously unscreened women, to provide health care services to older women. Once ages 35–65 years, were tested for high-risk types of VIA becomes successfully implemented, it should be HPV using HC2 (Qiagen, Gaithersburg, MD, United relatively easy to introduce more sensitive methods of States) and VIA, performed by nurses in primary care screening into the system. In many LMICs, establishing settings. This study found that the HPV screen-and- a sustainable and appropriate infrastructure is most treat arm was associated with a 3.7-fold reduction in the likely the priority. cumulative detection of CIN2 or greater by 36 months; Cervical Cancer 75 VIA was associated with a 1.5-fold reduction. For every progresses to cervical cancer precursors and cervical 100 women screened, the HPV and screen-and-treat cancer. HPV testing identifies women at risk, but not strategy averted 4.1 cases of CIN2 and greater compared those HPV-positive women who are most likely to with VIA-and-treat strategies, which averted 1.8 cases. have or to develop in the near future significant disease A further advantage of HPV testing is that speci- requiring treatment. The challenge is to triage these mens can be obtained by self-collection, with almost women by further testing with visual methods, cytology, complete preservation of the sensitivity and specificity molecular biomarkers, or a combination of techniques. of the screening method. Self-collection, which can be Among the visual methods, colposcopy with sub- done at home, is accepted by women and could signifi- sequent biopsy and treatment of visible lesions is the cantly increase participation in screening, particularly usual procedure in cytology-based programs. However, by women who are reluctant to undergo a gynecological this method requires highly specialized training and examination or who live in remote areas. relatively costly equipment. More importantly, the col- Another landmark study was a cluster randomized poscopic impression, colposcopically guided biopsy, trial of villages and centers where 131,746 women ages and histologic diagnosis are poorly reproducible and 30–59 years were recruited and randomly assigned to have important limitations to the point of reducing one of four groups: HPV testing; cytologic testing; VIA; the potential of highly sensitive screening tests. The or the standard of care, which involved no organized or current practice of selecting the most worrisome lesion opportunistic screening (Sankaranarayanan and others for biopsy misses up to one-third of prevalent small 2009). The incidence rate of cervical cancer stage 2 HSIL lesions. The collection of multiple biopsies from or higher and death rates from cervical cancer were aceto-white lesions can increase the sensitivity of colpos- significantly higher in the cytologic, VIA, and control copy (Pretorius and others 2011). groups compared with the HPV testing group. Further, Cytology of HPV-positive women is under strong the age-standardized incidence rate (ASIR) of invasive consideration as a triage method in screening pro- cancer among women who had negative test results on grams, given the high specificity of cytology and ample cytological or VIA testing was more than four times the expertise and infrastructure existing in some areas. rate among HPV-negative women. This method has the advantage of being highly specific, The high negative predictive value of HPV testing but it suffers from limited sensitivity. Sensitivity of (nearly 100 percent) allows the extension of the screen- cytology is influenced by many factors and is complex, ing interval, with consequent savings that can offset but used as a triage test for women already identified the possibly higher cost of the test compared with as high risk, cytology may suffice. The reduction in the cytology. Screening with HPV testing under age 30 is number of cytology tests required and the restriction not recommended, as HPV infection in this group of to HPV-positive women may improve the quality of women is common, and most infections are likely to cytology by reducing the workload and the number of be transient with a low likelihood of developing into negative slides. cancer. Screening younger women will add to the costs Using DNA biomarkers, limiting further follow-up of the program and may result in significant over- to women infected with HPV 16 and 18, which are treatment that may be associated with reproductive responsible for about 70 percent of cervical cancer and morbidity, in addition to significant emotional and precursors, can reduce the number of women referred social problems. to colposcopy while maintaining adequate sensitivity The HPV test is already in use for primary screen- (Castle and others 2011). Overexpression of certain ing in several countries, although in the United States, oncoproteins is a marker for increased risk of progres- primary HPV testing has been recommended only in sion to cervical cancer and may be a better predictor combination with cytology in primary screening or for of cancer risk than HPV DNA testing alone, although triage of cytologic abnormalities. A recent study includ- this is yet to be confirmed (Dockter and others 2009). ing more than 300,000 women in the United States One biomarker under intensive study is p16ink4a, which concluded that HPV testing without cytology might be is overexpressed in cancerous and precancerous cervical sufficiently sensitive for primary screening (Katki and cells. In a meta-analysis of studies using several detec- others 2011). tion methods, the proportion of smears overexpressing p16ink4a increases with the severity of cytological abnor- Triage of Positive HPV Tests malities (12 percent of normals and 89 percent of HSIL) Even among women over age 30 years, most HPV and histological abnormalities (2 percent of normals infections regress; only a minority of women develop and 82 percent of CIN3) (Sahasrabuddhe, Luhn, and persistent infection with high-risk types of HPV that Wentzensen 2011). A rapid test for the E6 oncoproteins 76 Cancer of HPV types 16, 18, and 45 is undergoing clinical trials coverage will most likely be the most clinically effective (Schweizer and others 2010). and cost-effective strategy for cervical cancer prevention. Public Health Challenges to Implementing HPV PRIMARY PREVENTION OF CERVICAL Vaccination CANCER: HPV VACCINES From the point of view of developing countries, intro- Vaccines that prevent infection with certain types of ducing the HPV vaccine poses many challenges. The HPV are a major breakthrough in preventing cervical most obvious is cost. The current price of both bivalent cancer. Monovalent (against HPV 16), bivalent (against and quadrivalent vaccines is high, although the costs HPV 16 and 18; Cervarix, GlaxoSmithKline Biologicals, have decreased considerably as a result of initiatives Rixensart, Belgium), and quadrivalent (against HPV 6, to enable implementation of HPV vaccination in low- 11, 16, and 18; Gardasil, Merck and Co., Inc., West Point, resource settings. However, cost is only one aspect. Pennsylvania) vaccines have been tested in randomized Unlike the development of a platform for vaccinating placebo-controlled trials and shown to be safe, immu- infants and children against a range of diseases (the nogenic, and highly efficacious at preventing HPV infec- Extended Program for Immunization [EPI]), few LMICs tion for up to eight years after vaccination. The bivalent have established pubescent/adolescent health platforms and quadrivalent vaccines are delivered by intramuscu- or school health systems from which to vaccinate young lar injection at zero, one, and six months, with the first girls and possibly boys. The infrastructure will have to dose between the ages of 9 and 13 years. be created; for this to happen, a great deal of political will must be generated. Studies supporting the efficacy of HPV vaccines involve adolescents, so they are effective Efficacy of HPV Vaccines in that age group; however, no completed studies have Evidence from well-conducted, randomized, placebo- included infants, so it would be premature to consider controlled trials demonstrates that these vaccines pre- adding an HPV vaccine to infant EPI. Several studies vent both persistent cervical infection with the types including young children are ongoing. included in the vaccines in women not previously In addition to the need to create a new infrastruc- exposed to HPV infection, as well as preinvasive lesions ture, both vaccines require a cold chain and thus a of the anogenital tract associated with the types present reliable source of electricity, which is absent in many in the vaccines in males and females. In addition, the LMICs, particularly in Sub-Saharan Africa. The need for quadrivalent vaccine prevents genital warts caused by three injections and follow-up poses its own challenges, types 6 and 11 (both associated with benign disease) in as does the necessity for intramuscular injection, which males and females (The Future II Study Group 2007; requires skill and medical waste disposal. However, Harper and others 2006; Koutsky and others 2002; Mao recent data indicate that the immunogenicity and effi- and others 2006; Roteli-Martins and others 2012; Villa cacy of two doses of the vaccine may be comparable to and others 2005). three doses, a promising development that could sim- Bivalent and quadrivalent vaccines appear to offer plify the logistics and reduce the cost of HPV vaccina- full protection against types 16 and 18, which together tion programs. Furthermore, the vaccine is administered cause an estimated 70 percent or more of cervical can- to young girls to prevent a disease that will manifest cers worldwide, and a slightly lower fraction of cervical itself only after 30 years or more. Developing a national cancer precursors. Some evidence suggests that the strategy will require those familiar with vaccination, immune response to vaccination against types 16 and including pediatricians and public health officials, to 18 also provides some cross-protection against types 45 communicate with those who work in the adult oncol- and 31, which are important in the etiology of cervical ogy field; these two worlds rarely intersect. cancer, thereby increasing the projected protection from A new pubescent or adolescent health platform vaccination to 75–80 percent. could be used beyond HPV vaccination. Such a However, both vaccines are prophylactic and should platform would provide an excellent opportunity to be administered to individuals prior to infection. HPV offer a range of services to young people, including is the most common sexually transmitted infection in booster vaccinations against hepatitis B and tetanus; the world. Ideally, the vaccine should be administered to possibly an anti-HIV vaccination in the future; anti- girls and possibly boys prior to the onset of sexual activ- helminthic medication; nutritional assessment; and ity, the age of which varies considerably by country and education about drug, tobacco, and alcohol use and culture. Vaccination of girls ages 9–12 years with high pregnancy prevention and sexuality. Cervical Cancer 77 Case Studies of HPV Vaccine Implementation Whether countries introduce the vaccine into the Rwanda, a country of 11 million people, introduced public health sector will be determined by several factors: an HPV vaccination program in partnership with Merck, the manufacturer of the quadrivalent vaccine, • Burden of HPV-associated disease in the country in 2010. Merck guaranteed three years of vaccinations • Ability to convince politicians and health officials, at no cost and concessional prices for future doses. In particularly those who work with children and vac- April 2011, 93,888 Rwandan girls in primary grade cination, that it is worthwhile to invest in vaccinating 5 received their first dose of the HPV vaccine, which children to prevent a disease of adulthood represented 95 percent coverage of all Rwandan girls • Creation of the appropriate infrastructure for the in the first round, followed by 94 percent in the second administration of the vaccine and 93 percent in the third (Binagwaho and others • Cost 2012). The success of this program is attributed to the school-based vaccination and community involve- TREATMENT OF CERVICAL CANCER ment in identifying girls absent from or not enrolled in school. As a result of screening, particularly at long intervals, On World Cancer Day 2013, Gavi, the Vaccine some more advanced cancers will be detected, and Alliance, announced that it would provide support some women will come for treatment because of symp- for the rollout of HPV vaccination in eight developing toms, commonly abnormal vaginal bleeding (postcoital, countries: Ghana, Kenya, the Lao People’s Democratic irregular, or postmenopausal), offensive vaginal dis- Republic, Madagascar, Malawi, Niger, Sierra Leone, charge, pelvic pain, dysuria, or symptoms of local or and Tanzania; the price has since been established at advanced metastatic disease. US$4.50 per dose (http://www.gavi.org). Further, Gavi As for all cancers, treatment of cervical cancer is plans to have one million girls vaccinated by introducing determined by the stage of the disease at presentation. the HPV vaccine in 20 countries by 2015 and hopes to Cervical cancer is staged clinically, for example, through reach 30 million by 2020 by introducing the vaccine in a pelvirectal examination combined with some basic 40 countries. tests as part of the metastatic work-up. Most institutions Ladner and others (2012) report on the Gardasil Access rely on the International Federation of Gynecology and Program, managed by Axios Healthcare Development, Obstetrics (FIGO) 2009 staging. which received a large donation of the quadrivalent Treatment options for most stage 1 cancers favor vaccine from Merck. Participating projects received free surgery alone and usually are curative. For women with vaccine and were responsible for the costs related to later stage 1, stage 2, and early stage 3 cancers, primary the importation, transportation, storage, and distribu- treatment is chemotherapy and radiotherapy, with cura- tion of the vaccine, as well as the costs of community tive intent but lower success rates than for earlier stages. outreach, program management, and data collection. For stage 4 disease, treatment is usually palliative and Eight programs were implemented in seven countries: may involve chemotherapy, radiotherapy, and surgery, Bhutan, Bolivia, Cambodia, Cameroon, Haiti, Lesotho, although few women in LMICs are likely to have access and Nepal. The eight programs targeted 87,380 girls, to these services. of whom 76,983 (88 percent) received three doses of the vaccine. Three vaccine delivery models were used: COST-EFFECTIVENESS ANALYSIS health facility–based, school-based, and mixed (health facility– and school-based). The mixed model resulted in Model-Based Cost-Effectiveness Analysis the best coverage (96.6 percent); the school-based model In addition to the strong evidence of the clinical was intermediate (88.6 percent); and the health facility effectiveness of primary and secondary prevention of model was the least effective (79.9 percent). The esti- cervical cancer worldwide, a critical factor in decision mated coverage was 94.9 percent for the five programs making, particularly in resource-poor settings, is the that targeted girls ages 9–13 years, and 80.0 percent for financial impact and cost-effectiveness of alternative the three programs that vaccinated girls outside that age strategies. Most economic evaluations of cervical range. cancer prevention approaches have utilized mathe- These data, which show high coverage in low- matical models to project the long-term public health resource settings, are encouraging. They suggest that and economic impacts of prevention strategies in with sufficient political will, the implementation of HPV different populations. State-of-the-art methods, as well vaccination in low-resource settings should be possible as the limitations of modeling, have been discussed in the near future. extensively in published review papers (Brisson, Van 78 Cancer de Velde, and Boily 2009; Canfell and others 2012; Kim, Measures of Cost-Effectiveness Brisson, and others 2008). Not surprisingly, HPV vaccination was cost-effective in more countries as the cost of the vaccine decreased. Consistently across the regional and country-specific HPV Vaccination studies cited, the results have suggested that for a cost per The economic evaluations of HPV vaccination have vaccinated girl (CVG) of US$50 or less, HPV vaccination focused primarily on vaccination of preadolescent girls of preadolescent girls was good value for money in most prior to sexual initiation; only a handful of evaluations of the countries evaluated. In countries with a relatively have addressed HPV vaccination of other targeted lower disease burden and/or lower per capita gross groups, such as preadolescent boys or older women domestic product, the vaccine cost threshold at which (Tsu and Murray 2011; Tsu, Murray, and Franceschi HPV vaccination was cost-effective was lower, at US$10 2012). Several regional reports published as part of an or US$25 CVG. One study, published by the manufac- HPV monograph series have projected health benefits turers of the quadrivalent HPV vaccine, included strate- (for example, cancer risk reduction and life expectancy, gies of vaccinating males and females up to age 24 years adjusted or unadjusted for disability or quality of life) in Mexico; the study found that the most cost-effective and economic outcomes of HPV vaccination of pread- strategy was vaccinating 12-year-old girls alone (Insinga olescent girls in all countries in the following regions: and others 2007). Generally, the factors with the greatest influence • East Asia and Pacific (25 countries) (Goldie, Diaz, on the cost-effectiveness results were the vaccine cost Kim, and others 2008) and discount rate, which reflect the time preference for • Europe and Central Asia (28 countries) (Berkhof and health benefits and costs and are important to capture, others 2013) given the long time horizon between vaccine expendi- • Latin America and the Caribbean (33 countries) ture and expected cancer benefits. Vaccine efficacy and (Goldie, Diaz, Constenla, and others 2008) the length of vaccine protection—and the requirement • Middle East and North Africa (20 countries) (Kim, for booster doses—also influence the results, with the Campos, and others 2013) cost-effectiveness profile diminishing greatly, assuming • Sub-Saharan Africa (48 countries) (Kim, Sharma, protection lasts only 10–20 years and/or requires at least and others 2013) one booster dose. Variations in cancer incidence moder- ately influenced the cost-effectiveness ratios. A related analysis evaluated HPV vaccination in In interpreting cost-effectiveness results, a critical 72 countries eligible for support from Gavi (Goldie, distinction must be made between value for resources O’Shea, and others 2008). A handful of country-specific and affordability. Affordability will be a critical deter- analyses in these regions and economies have also been minant for success in preventing cervical cancer in conducted, including Brazil (Goldie and others 2007; LMICs with high cervical cancer incidence (Natunen Vanni and others 2012); China (Canfell and others and others 2013). Despite the high value that HPV 2011); India (Diaz and others 2008); Malaysia (Aljunid vaccination can provide at US$25–US$50 per vac- and others 2010; Ezat and Aljunid 2010); Mexico cinated girl, the immediate financial expenditures (Reynales-Shigematsu, Rodrigues, and Lazcano-Ponce required for adoption of HPV vaccination at this cost 2009); Taiwan, China (Demarteau and others 2012); and may not be attainable in many countries. For example, Thailand (Sharma and others 2011). the financial requirements for vaccinating five birth The overwhelming majority of these studies has cohorts over five years at 70 percent coverage in all of concluded that HPV vaccination of preadolescent girls Sub-Saharan Africa will range from US$110.0 million has the potential to reduce substantially the morbidity (US$0.55 per dose) to US$2.8 billion (US$19.50 per and mortality associated with cervical cancer, under dose) (Kim, Sharma, and others 2013). At least one assumptions of sustained, high vaccine efficacy and HPV vaccine manufacturer has offered a price as low reasonable uptake. For example, when assuming vac- as US$5 per dose to Gavi, undoubtedly diminishing cination coverage of 70 percent and complete, lifelong the financial barrier to accessing HPV vaccines. Study protection against HPV 16/18 cervical cancer, HPV results suggest that the upfront financial investments in vaccination was estimated to avert more than 670,000 HPV vaccination may be offset by downstream savings cervical cancer cases in Sub-Saharan Africa alone over in costs of cancer care averted at such a low vaccine the lifetimes of women in five consecutive birth cohorts price. Careful planning to ensure the sustainability of vaccinated as young adolescents (Kim, Sharma, and an HPV vaccination program will be as important as others 2013). the decision to implement it. Cervical Cancer 79 Recent publications on the incremental program costs associated with screening, the authors conclude that of introducing and scaling up HPV vaccination suggest HPV DNA testing and VIA, requiring one or two clinic that integrating HPV vaccination into existing immuniza- visits, two to three times per lifetime, at age 35 years, tion services is feasible but will likely incur an additional are attractive alternatives to traditional three-visit, financial burden to countries above the cost of the vaccine cytology-based testing programs. In a more recent study, (Hutubessy and others 2012; Levin and others 2013). Levin and others (2010) find that increased coverage The support by Gavi not only to fund HPV vaccines levels of cervical cancer screening using rapid HPV tests directly, but also to develop country HPV vaccination were cost-effective for a two-visit strategy for screening programs through demonstration projects, will be instru- and treatment of precancerous lesions in China. mental in creating sustainable programs; to date, at least The most influential factors in determining the rel- 14 countries have applied for demonstration projects. ative value of different screening strategies include the assumptions regarding loss to follow-up between clinic Future Directions for Cost-Effectiveness visits, the clinical performance of the screening test (that Analysis of HPV Vaccines is, the sensitivity/specificity), and the relative costs of Head-to-head comparisons of the bivalent and quadriva- the test. Patient time spent receiving interventions and lent vaccines are lacking for LICs, but such comparisons traveling to clinics are also found to be influential, given may be more relevant in the future. Economic evaluations the long distances to the clinics, lack of paved roads, and in HICs that have introduced the HPV vaccine have been limited public transportation in some settings. Treatment conducted comparing the bivalent and quadrivalent vac- of precancerous lesions can range from inexpensive cines. To date, the results have been conflicting; three stud- cryotherapy to more complex and costly LEEP, cold ies find that the quadrivalent vaccine is more cost-effective knife conization, and simple hysterectomies. However, than the bivalent vaccine (Dee and Howell 2010; Jit and these costs are rarely the main drivers affecting the cost- others 2011; Lee and others 2011). In contrast, two studies effectiveness of cervical cancer screening strategies. (Demarteau and others 2012; Ezat and Aljunid 2010) find Remaining challenges include improving the accept- that the cost-savings from reducing more cases of cervical ability and accessibility of these services among previously cancer (bivalent vaccine) outweigh the cost-savings from unscreened women. Even with the strong momentum reducing cases of genital warts (quadrivalent vaccine). toward introducing HPV vaccination programs, invest- Current studies have not yet explored the potential added ing in expanded quality screening and treatment services benefits from broad-coverage HPV vaccines that target and increasing demand for these services among older additional oncogenic HPV types and are anticipated to be women remain critical, given that screening rates are very available in the near future; these second-generation vac- low in Asia, Latin America and the Caribbean, and Sub- cines are expected to yield even greater cancer reductions Saharan Africa, irrespective of income, and that these and are likely further to impact optimal screening, but the women are not the target group for HPV vaccination. efficacy and costs are unknown. Combined Vaccination and Screening Cervical Cancer Screening Increasingly, analyses are considering the potential syn- Studies evaluating screening strategies alone have pri- ergies between preadolescent HPV vaccination followed marily assessed screening tests (for example, cytology, by screening in adulthood. The majority of recent HPV DNA tests, and VIA), frequencies (for exam- studies are set in upper-middle-income countries, such ple, one to three times per lifetime, at 3- to 10-year as Malaysia (Aljunid and others 2010; Ezat and Aljunid intervals), and ages at screening (for example, 30–50 2010), Mexico (Insinga and others 2007; Reynales- years). In a seminal study, Goldie and others (2005) Shigematsu, Rodrigues, and Lazcano-Ponce 2009), Peru assess the cost-effectiveness of screening strategies in (Goldie and others 2012), South Africa (Sinanovic and five LMICs with heterogeneous epidemiologic, demo- others 2009), and Thailand (Praditsitthikorn and others graphic, and economic profiles. They find that strategies 2011; Sharma and others 2011; Termrungruanglert and that required the fewest visits—and thereby minimized others 2012). Only a handful of such studies are in LICs loss to follow-up—were consistently the most cost- and lower-middle-income countries, such as countries effective. The reduction in lifetime cervical cancer risk in Eastern Africa (Campos and others 2012), India was 25–36 percent, with only one screening per lifetime; (Diaz and others 2008), and Vietnam (Kim, Kobus, and 47–52 percent, with two screenings per lifetime; and others 2008). 57–60 percent, with three screenings per lifetime. Taking The findings suggest an opportunity to improve on into consideration the direct medical and patient costs cervical cancer prevention by following preadolescent 80 Cancer HPV vaccination with screening (HPV DNA testing) and VIA. Both tests have their advantages and dis- of women one to three times per lifetime, starting at advantages, but the development of a highly repro- about age 40 years. For example, in Thailand, pread- ducible, reliable, and accurate point-of-care HPV olescent HPV vaccination combined with screening DNA test (or an alternative test yet to be developed in older women reduced the risk of cervical cancer by but fulfilling these criteria) will enable women to over 50 percent (Sharma and others 2011). In Mexico, be screened and treated in one visit and without the HPV vaccination combined with cytology screening need for colposcopy and laboratory infrastructure. every three years reduced cancer incidence and mor- HPV DNA testing has shown very promising results; tality by 75 percent (Reynales-Shigematsu, Rodrigues, however, issues of specificity, overtreatment, and and Lazcano-Ponce 2009). Similar to findings in studies effective triage still need to be resolved. that explored the cost-effectiveness of HPV vaccination alone, this set of literature generally finds that adding Screening and vaccinating either separately or HPV vaccination for preadolescent girls to existing or together are shown to be highly cost-effective public modified screening programs has the potential to be a health interventions. cost-effective strategy, with the vaccine price being a key factor in determining cost-effectiveness. Despite finding that HPV vaccination is cost-effective, these studies reit- NOTES erate the concern over affordability (Canfell and others The World Bank classifies countries according to four income 2011; Praditsitthikorn and others 2011; Sharma and groupings. Income is measured using gross national income others 2011). per capita, in U.S. dollars, converted from local currency using the World Bank Atlas method. Classifications as of July 2014 are as follows: Conclusion of Cost-Effectiveness Analysis • Low-income countries = US$1,045 or less in 2013 The findings from recent cost-effectiveness analyses • Middle-income countries are subdivided: clearly indicate that there are promising opportunities • Lower-middle-income = US$1,046–US$4,125 to prevent cervical cancer in different world settings. • Upper-middle-income = US$4,126–US$12,745 HPV vaccination for preadolescent girls and screening • High-income countries = US$12,746 or more of adult women, even only three times per lifetime, can avert a significant proportion of cervical cancer cases in 1. The map and figures in this chapter are based on incidence a cost-effective manner. In addition to many other crit- and mortality estimates for ages 0–69 years, consistent ical inputs to health decisions, such as political will and with reporting in all DCP3 volumes. Cancer statistics cultural acceptability, evidence on the cost-effectiveness are estimates for 2012 and have been provided by the and affordability of HPV vaccination and screening International Agency for Research on Cancer from its from rigorous model-based analyses can help to inform GLOBOCAN 2012 database. Observed population-based decision makers and stakeholders in their deliberations cancer incidence rates were derived from Cancer Incidence in Five Continents, 10th edition, and for trends over time on how best to prevent cervical cancer worldwide. from CI5plus (http://ci5.iarc.fr/CI5plus/Default.aspx). The discussion of burden (including risk factors), however, includes all ages unless otherwise noted. Interventions also CONCLUSIONS apply to all age groups, except where age ranges or cutoffs are specified. Cervical cancer remains one of the most common cancers 2. HDI is a composite of three dimensions of human devel- among women living in LMICs, yet it is a preventable opment: a long and healthy life (life expectancy at birth), and treatable cancer. Resource-constrained countries access to knowledge (adult literacy and enrollment at have been unable to initiate or sustain cytology-based different educational levels), and standard of living (gross cervical cancer screening programs because of weak domestic product adjusted for purchasing power parity). health care infrastructure and prohibitive cost. There are two new avenues for cervical cancer prevention: REFERENCES • Primary prevention through prophylactic vaccina- Aljunid, S., A. Zafar, S. Saperi, and M. Amrizal. 2010. tion against the most common HPV types causally “Burden of Disease Associated with Cervical Cancer in associated with cervical cancer. Malaysia and Potential Costs and Consequences of HPV • Use of alternative screening tests and strategies for Vaccination.” Asian Pacific Journal of Cancer Prevention cervical cancer prevention, namely, HPV DNA testing 11 (6): 1551–59. Cervical Cancer 81 Arbyn, M., X. Castellsague, S. de Sanjose, L. Bruni, M. Saraiya, Prevention: Long-Term Results of a Randomized Screening and others. 2011. “Worldwide Burden of Cervical Cancer in Trial.” Journal of the National Cancer Institute 102 (20): 1557–67. 2008.” Annals of Oncology 22 (12): 2675–86. Denny, L., M. Quinn, and R. Sankaranarayanan. 2006. Arbyn, M., G. Ronco, A. Anttila, C. J. Meijer, M. Poljak, and “Screening for Cervical Cancer in Developing Countries.” others. 2012. “Evidence Regarding Human Papillomavirus Vaccine 24 (Suppl. 3): S3/71–77. Testing in Secondary Prevention of Cervical Cancer.” Diaz, M., J. J. Kim, G. Albero, S. de Sanjose, G. Clifford, and Vaccine 30 (Suppl. 5): F88–99. others. 2008. “Health and Economic Impact of HPV 16 and Berkhof, J., J. A. Bogaards, E. Demirel, M. Diaz, M. Sharma, 18 Vaccination and Cervical Cancer Screening in India.” and others. 2013. “Cost-Effectiveness of Cervical Cancer British Journal of Cancer 99 (20): 230–38. Prevention in Central and Eastern Europe and Central Dockter, J., A. Schroder, C. Hill, L. Guzenski, J. Monsonego, and Asia.” Vaccine 31 (S7): H71–79. others. 2009. “Clinical Performance of the APTIMA HPV Binagwaho, A., C. M. Wagner, M. Gatera, C. Karema, C. T. Nutt, Assay for the Detection of High-Risk HPV and High-Grade and others. 2012. “Achieving High Coverage in Rwanda’s Cervical Lesions.” Journal of Clinical Virology 45 (Suppl. 1): National Human Papillomavirus Vaccination Programme.” S55–61. Bulletin of the World Health Organization 90 (8): 623–28. Ellerbrock, T. V., M. A. Chiasson, T. J. Bush, X. W. Sun, Brisson, M., N. Van de Velde, and M. C. Boily. 2009. “Economic D. Sawo, and others. 2000. “Incidence of Cervical Squamous Evaluation of Human Papillomavirus Vaccination in Intraepithelial Lesions in HIV-Infected Women.” Journal of Developed Countries.” Public Health Genomics 12 (5–6): the American Medical Association 283 (8): 1031–37. 343–51. Ezat, S. W., and S. Aljunid. 2010. “Comparative Cost- Campos, N. G., J. J. Kim, P. E. Castle, J. Ortendahl, M. O’Shea, Effectiveness of HPV Vaccines in the Prevention of and others. 2012. “Health and Economic Impact of HPV Cervical Cancer in Malaysia.” Asian Pacific Journal of Cancer 16/18 Vaccination and Cervical Cancer Screening in Prevention 11 (4): 943–51. Eastern Africa.” International Journal of Cancer 130 (11): Ferlay, J., I. Soerjomataram, M. Ervik, R. Dikshit, S. Eser, and 2672–84. others. 2013. GLOBOCAN 2012 v1.0, Cancer Incidence Canfell K., H. Chesson, S. L. Kulasingam, J. Berkhof, M. Diaz, and Mortality Worldwide: IARC CancerBase No. 11. and J. J. Kim. 2012. “Modelling Preventative Strategies International Agency for Research on Cancer, Lyon, France. against Human Papillomavirus-Related Disease in http://globocan.iarc.fr. Developing Countries.” Vaccine 30 (Suppl. 5): F157–67. Forouzanfar, M. H., K. J. Foreman, A. M. Delossantos, Canfell, K., J. F. Shi, J. B. Lew, R. Walker, F. H. Zhao, and R. Lozano, A. D. Lopez, and others. 2011. “Breast and others. 2011. “Prevention of Cervical Cancer in Rural Cervical Cancer in 187 Countries between 1980 and 2010: China: Evaluation of HPV Vaccination and Primary HPV A Systematic Analysis.” The Lancet 378 (9801): 1461–84. Screening Strategies.” Vaccine 29 (13): 2487–94. Gakidou, E., S. Nordhagen, and Z. Obermeyer. 2008. “Coverage Castle, P. E., M. H. Stoler, T. C. Wright, Jr., A. Sharma, T. L. Wright, of Cervical Cancer Screening in 57 Countries: Low Average and others. 2011. “Performance of Carcinogenic Human Levels and Large Inequalities.” PLoS Medicine 5 (6): e132. Papillomavirus (HPV) Testing and HPV16 or HPV18 Goldie, S. J., M. Diaz, D. Constenla, N. Alvis, J. K. Andrus, and Genotyping for Cervical Cancer Screening of Women Aged others. 2008. “Mathematical Models of Cervical Cancer 25 Years and Older: A Subanalysis of the ATHENA Study.” Prevention in Latin America and the Caribbean.” Vaccine The Lancet Oncology 12 (9): 880–90. 26 (Suppl. 11): L59–72. Cuzick, J., M. Arbyn, R. Sankaranarayanan, V. Tsu, G. Ronco, Goldie, S. J., M. Diaz, S. Y. Kim, C. E. Levin, H. V. Minh, and others. 2008. “Overview of HPV-Based and Other and J. J. Kim. 2008. “Mathematical Models of Cervical Novel Options for Cervical Cancer Screening in Developed Cancer Prevention in the Asia Pacific Region.” Vaccine 26 and Developing Countries.” Vaccine 206: k29–41. (Suppl. 12): M17–29. de Sanjose, S., G. V. Quint, L. Alemany, D. T. Geraets, Goldie, S. J., L. Gaffikin, J. D. Goldhaber-Fiebert, A. Gordillo- J. E. Klaustermeier, and others. 2010. “Human Papillomavirus Tobar, C. Levin, and others. 2005. “Cost-Effectiveness of Genotype Attribution in Invasive Cervical Cancer: Cervical Cancer Screening in Five Developing Countries.” A Retrospective Cross-Sectional Worldwide Study.” The New England Journal of Medicine 353 (20): 2158–68. Lancet Oncology 11 (11): 1048–56. Goldie, S. J., J. J. Kim, K. E. Kobus, J. Goldhaber-Fiebert, Dee, A., and F. A. Howell. 2010. “A Cost-Utility Analysis of J. A. Salomon, and others. 2007. “Cost-Effectiveness of HPV Adding a Bivalent or Quadrivalent HPV Vaccine to the Irish Vaccination in Brazil.” Vaccine 25 (33): 6257–70. Cervical Screening Program.” European Journal of Public Goldie, S. J., C. Levin, N. R. Mosqueira-Lovon, J. Ortendahl, Health 20 (2): 213–29. J. J. Kim, and others. 2012. “Health and Economic Impact Demarteau, N., C. H. Tang, H. C. Chen, C. J. Chen, and of HPV 16 and 18 Vaccination of Pre-adolescent Girls G. Van Kriekinge. 2012. “Cost-Effectiveness Analysis of and Cervical Cancer Screening of Adult Women in Peru.” the Bivalent Compared with the Quadrivalent Human Revista Panamericana de Salud Pública 32 (6): 426–34. Papillomavirus Vaccines in Taiwan.” Value Health 15 (5): Goldie, S. J., M. K. O’Shea, N. G. Campos, M. Diaz, S. J. Sweet, 622–31. and others. 2008. “Health and Economic Outcomes of HPV Denny, L., L. Kuhn, C. C. Hu, W. Y. Tsai, and T. C. Wright, 16, 18 Vaccination in 72 GAVI-Eligible Countries.” Vaccine Jr. 2010. “Human Papillomavirus–Based Cervical Cancer 26 (32): 4080–93. 82 Cancer Hakama, M. 1986. “Cervical Cancer: Risk Groups for Cervical Cancer Prevention in Sub-Saharan Africa.” Vaccine Screening.” IARC Scientific Publications 76: 213–19. 31 (S5): G60–72. Harper, D. M., E. L. Franco, C. M. Wheeler, A. B. Moscicki, Kim, J. J., K. E. Kobus, M. Diaz, V. Van Minh, and S. J. Goldie. B. Romanaowski, and others. 2006. “Sustained Efficacy up 2008. “Exploring the Cost-Effectiveness of HPV Vaccination to 4–5 Years of a Bivalent L1 Virus-Like Particle Vaccine in Vietnam: Insights for Evidence-Based Cervical Cancer against Human Papillomavirus Types 16 and 18: Follow-Up Prevention Policy.” Vaccine 26 (32): 4015–24. from a Randomised Control Trial.” The Lancet 367 (9518): Kim, J. J., M. Sharma, M. O’Shea, S. Sweet, M. Diaz, and 1247–55. others. 2013. “Model-Based Impact and Cost-Effectiveness Harris, T. G., R. D. Burk, J. M. Palesky, L. S. Massad, of Cervical Cancer Prevention in the Middle East and J. Y. Bang, and others. 2005. “Incidence of Cervical Northern Africa.” Vaccine 31 (S6): G65–77. Squamous Intraepithelial Lesions Associated with HIV Koutsky, L. A., K. A. Ault, C. M. Wheeler, D. R. Brown, Serostatus, CD4 Cell Counts, and Human Papillomavirus E. Barr, and others. 2002. “A Controlled Trial of a Human Test Results.” Journal of the American Medical Association Papillomavirus Type 16 Vaccine.” New England Journal of 293 (12): 1471–76. Medicine 372 (2): 1645–51. Herrero, R., P. E. Castle, M. Schiffman, M. C. Bratti, Laara, E., N. E. Day, and M. Hakama. 1987. “Trends in A. Hildesheim, and others. 2005. “Epidemiologic Profile Mortality from Cervical Cancer in the Nordic Countries: of Type-Specific Human Papillomavirus Infection and Association with Organised Screening Programs.” The Cervical Neoplasia in Guanacaste, Costa Rica.” Journal of Lancet 1 (8544): 1247–49. Infectious Diseases 191 (11): 1796–807. Ladner J., M. Besson, R. Hampshire, L. Tapert, M. Chirenje, Hutubessy, R., A. Levin, S. Wang, W. Morgan, M. Ally, and and others. 2012. “Assessment of Eight HPV Vaccination others. 2012. “A Case Study Using the United Republic of Programs Implemented in Lowest Income Countries.” Tanzania: Costing Nationwide HPV Vaccine Delivery Using BMC Public Health 12: 370. the WHO Cervical Cancer Prevention and Control Costing Lazcano-Ponce, E., A. de Ruiz, L. Lopez-Carillo, M. Vazquez- Tool.” BMC Medicine 10: 136. Manriquez, and M. Hernandez-Avila. 1994. “Quality IARC (International Agency for Research on Cancer) Working Control Study on Negative Gynecological Cytology in Group on Cervical Cancer Screening. 1986a. “Summary Mexico.” Diagnostic Cytopathology 10 (1): 10–14. Chapter.” In Screening for Cancer of the Uterine Cervix, Lee, V. J., S. K. Tay, Y. L. Teoh, and M. Y. Tok. 2011. “Cost- edited by M. Hakama, A. B. Miller, and N. E. Day, 133–42. Effectiveness of Different Human Papillomavirus Vaccines Lyon: IARC. in Singapore.” BMC Public Health 11: 203. ———. 1986b. “Screening for Squamous Cervical Cancer: Levin, C. E., J. Sellors, J. F. Shi, L. Ma, Y. L. Qiao, and others. Duration of Low Risk after Negative Results of Cervical 2010. “Cost-Effectiveness Analysis of Cervical Cancer Cytology and Its Implication for Screening Programmes.” Prevention Based on a Rapid Human Papillomavirus British Medical Journal 293 (6548): 659–64. Screening Test in a High-Risk Region of China.” Insinga, R. P., E. J. Dasbach, E. H. Elbasha, A. Puig, and International Journal of Cancer 127 (6): 1404–11. L. M. Reynales-Shigematsu. 2007. “Cost-Effectiveness of Levin, C. E., H. Van Minh, J. Odaga, S. S. Rout, D. N. Ngoc, Quadrivalent Human Papillomavirus (HPV) Vaccination and others. 2013. “Delivery Cost of Human Papillomavirus in Mexico: A Transmission Dynamic Model-Based Vaccination of Young Adolescent Girls in Peru, Uganda Evaluation.” Vaccine 26 (1): 128–39. and Viet Nam.” Bulletin of the World Health Organization Irwin, K. L., M. W. Oberle, and L. Rosero-Bixby. 1991. 91 (8): 585–92. “Screening Practices for Cervical and Breast Cancer Mao, C., L. A. Koutsky, K. A. Ault, C. M. Wheeler, D. R. Brown, in Costa Rica.” Bulletin of the Pan American Health and others. 2006. “Efficacy of Human Papillomavirus-16 Organization 25 (1): 16–26. Vaccine to Prevent Cervical Intraepithelial Neoplasia: Jit, M., R. Chapman, O. Hughes, and Y. H. Choi. 2011. A Randomized Controlled Trial.” Obstetrics and Gynecology “Comparing Bivalent and Quadrivalent Human 107 (1): 18–27. Papillomavirus Vaccines: Economic Evaluation Based on Massad, L. S., M. H. Einstein, W. K. Huh, H. A. Katki, Transmission Model.” British Medical Journal 343: d5775. W. K. Kinney, and others. 2013. “Updated Consensus doi:10.1136/bmj.d5775. Guidelines for Management of Abnormal Cervical Cancer Katki, H. A., W. K. Kinney, B. Fetterman, T. Lorey, N. E. Poitras, Screening Tests and Cancer Precursors.” Journal of Lower and others. 2011. “Cervical Cancer Risk for Women Genital Tract Disease 17 (5): S1–S27. Undergoing Concurrent Testing for Human Papillomavirus National Cancer Workshop. 1989. “The 1988 Bethesda System and Cervical Cytology: A Population-Based Study in for Reporting Cervical/Vaginal Cytologic Diagnosis.” Routine Clinical Practice.” The Lancet Oncology 12 (7): Journal of the American Medical Association 262 (7): 663–72. 931–34. Kim, J. J., M. Brisson, J. Edmunds, and S. J. Goldie. 2008. Natunen, K., T. A. Lehtinen, S. Torvinen, and M. Lehtinent. “Modeling Cervical Cancer Prevention in Developed 2013. “Cost-Effectiveness of HPV-Vaccination in Medium Countries.” Vaccine 26 (11): K76–86. or Low Income Countries with High Cervical Cancer Kim, J. J., N. G. Campos, M. O’Shea, M. Diaz, and I. Mutyaba. Incidence—A Systematic Review.” Journal of Vaccines and 2013. “Model-Based Impact and Cost-Effectiveness of Vaccination 4 (1): 2–10. Cervical Cancer 83 Palefsky, J., H. Minkoff, L. Kalish, A. Levine, H. S. Sacks, Sinanovic, E., J. Moodley, M. A. Barone, S. Mall, S. Cleary, and and others. 1999. “Human Papillomavirus Infection and others. 2009. “The Potential Cost-Effectiveness of Adding Cervical Cytology in HIV-Infected and HIV-Uninfected a Human Papillomavirus Vaccine to the Cervical Cancer Rwandan Women.” Journal of Infectious Diseases 12: Screening Program in South Africa.” Vaccine 27 (44): 1851–61. 6196–202. Parkin, D. M., M. Almonte, L. Bruni, G. Clifford, M. P. Curado, Singh, G. K. 2012. “Rural-Urban Trends and Patterns in and others. 2008. “Burden and Trends of Type-Specific Cervical Cancer Mortality, Incidence, Stage, and Survival Human Papillomavirus Infections and Related Diseases in the United States, 1950–2008.” Journal of Community in the Latin America and Caribbean Region.” Vaccine 26 Health 37: 217–23. (Suppl. 11): L1–15. Singh, D., K. Anastos, D. Hoover, R. Burk, Q. Shi, and others. Praditsitthikorn, N., Y. Teerawattananon, S. Tantivess, 2009. “Human Papillomavirus Infection and Cervical S. Limwattananon, A. Riewpaiboon, and others. Cytology in HIV-Infected and HIV-Uninfected Rwandan 2011. “Economic Evaluation of Policy Options for Women.” Journal of Infectious Diseases 199 (12): 1851–61. Prevention and Control of Cervical Cancer in Thailand.” Soerjomataram, I., J. Lortet-Tieulent, D. M. Parkin, J. Ferlay, Pharmacoeconomics 29 (9): 781–806. C. Mathers, and others. 2012. “Global Burden of Cancer in Pretorius, R. G., J. L. Belinson, R. J. Burchette, S. Hu, X. Zhang, 2008: Systematic Analysis of Disability-Adjusted Life-Years and others. 2011. “Regardless of Skill, Performing More in 12 World Regions.” The Lancet 380: 1840–50. Biopsies Increases the Sensitivity of Colposcopy.” Journal of Termrungruanglert, W., P. Havanond, N. Khemapech, Lower Genital Tract Disease 15 (3): 180–88. S. Lertmaharit, S. Pongpanich, and others. 2012. “Model for Reynales-Shigematsu, L. M., E. R. Rodrigues, and E. Lazcano- Predicting the Burden and Cost of Treatment in Cervical Ponce. 2009. “Cost-Effectiveness Analysis of a Quadrivalent Cancer and HPV-Related Diseases in Thailand.” European Human Papilloma Virus Vaccine in Mexico.” Archives of Journal of Gynaecological Oncology 33 (4): 391–94. Medical Research 40 (6): 503–13. The Future II Study Group. 2007. “Quadrivalent Vaccine Roteli-Martins, C., P. Naud, P. De Borba, J. Teixeira, N. De against Human Papillomavirus to Prevent High-Grade Carvalho, and others. 2012. “Sustained Immunogenicity and Cervical Lesions.” New England Journal of Medicine 356: Efficacy of the HPV-16/18 AS04-Adjuvanted Vaccine: Up to 1915–27. 8.4 Years of Follow-Up.” Human Vaccine Immunotherapy Tsu, V., and M. Murray. 2011. “Limited Benefit of HPV 8 (3): 390–97. Vaccination for Sexually Active Women in Developing Sahasrabuddhe, V., V. P. Luhn, and N. Wentzensen. 2011. Countries.” Vaccine 29 (50): 9290–91. “Human Papillomavirus and Cervical Cancer: Biomarkers Tsu, V., M. Murray, and S. Franceschi. 2012. “Human for Improved Prevention Efforts.” Future Microbiology Papillomavirus Vaccination in Low-Resource Countries: 6 (9): 1083–98. Lack of Evidence to Support Vaccinating Sexually Active Sankaranarayanan, R., B. M. Nene, S. S. Shastri, K. Jayant, Women.” British Journal of Cancer 107 (9): 1445–50. R. Muwonge, and others. 2009. “HPV Screening for Cervical Vanni, T., P. Luz, A. Mendes, M. Foss, M. Mesa-Frias, and Cancer in Rural India.” New England Journal of Medicine others. 2012. “Economic Modelling Assessment of the HPV 360 (14): 1385–94. Quadrivalent Vaccine in Brazil: A Dynamic Individual- Sankaranarayanan, R., and P. Pisani. 1997. “Prevention Based Approach.” Vaccine 30 (32): 4866–71. Measures in the Third World: Are They Practical?” In New Villa, L. L., R. L. Costa, C. A. Petta, R. P. Andrade, K. A. Ault, Developments in Cervical Cancer Screening and Prevention, and others. 2005. “Prophylactic Quadrivalent Human edited by E. Franco and J. Monsonego, 70–83. Oxford: Papillomavirus (Types 6, 11, 16, and 18) L1 Virus-Like Blackwell Science Ltd. Particle Vaccine in Young Women: A Randomised Double- Sauvaget, C., J. M. Fayette, R. Muwonge, R. Wesley, and Blind Placebo-Controlled Multicentre Phase II Efficacy R. Sankaranarayanan. 2011. “Accuracy of Visual Inspection Trial.” The Lancet Oncology 6 (5): 271–78. with Acetic Acid for Cervical Cancer Screening.” International Walboomers, J. M., M. V. Jacobs, M. M. Manos, F. X. Bosch, Journal of Gynecology and Obstetrics 113 (1): 14–24. J. A. Kummer, and others. 1999. “Human Papillomavirus Is Schweizer, J., P. S. Lu, C. W. Mahoney, M. Berard-Bergery, a Necessary Cause of Invasive Cervical Cancer Worldwide.” M. Ho, and others. 2010. “Feasibility Study of a Human Journal of Pathology 189: 12–19. Papillomavirus E6 Oncoprotein Test for Diagnosis of WHO (World Health Organization). 2012. “Prevention of Cervical Precancer and Cancer.” Journal of Clinical Cervical Cancer through Screening and Using Visual Microbiology 48 (12): 4646–48. Inspection with Acetic Acid (VIA) and Treatment with Sharma, M., J. Ortendahl, E. van der Ham, S. Sy, and J. J. Kim. Cryotherapy.” WHO, Geneva. 2011. “Cost-Effectiveness of Human Papillomavirus Wright, T. C., and R. J. Kurman. 1994. “A Critical Review of Vaccination and Cervical Cancer Screening in Thailand.” the Morphologic Classification Systems of Preinvasive British Journal of Obstetrics and Gynaecology 119 (2): Lesions of the Cervix: The Scientific Basis for Shifting the 166–76. Paradigm.” Papillomavirus Report 5: 175–82. 84 Cancer Chapter 5 Oral Cancer: Prevention, Early Detection, and Treatment Rengaswamy Sankaranarayanan, Kunnambath Ramadas, Hemantha Amarasinghe, Sujha Subramanian, and Newell Johnson INTRODUCTION in India, East Asia, Eastern Europe, and parts of South America (Forman and others 2013), where organized Oral cancer is the 11th most common cancer in the prevention and early detection efforts are lacking. This world, accounting for an estimated 300,000 new cases chapter discusses the epidemiology, prevention, early and 145,000 deaths in 2012 and 702,000 prevalent cases detection, and treatment of oral cancers, as well as the over a period of five years (old and new cases) (tables 5.1 cost-effectiveness of interventions. and 5.2) (Bray and others 2013; Ferlay and others 2013). For this chapter, oral cancers include cancers of the mucosal lip, tongue, gum, floor of the mouth, palate, and ORAL CANCER: INCIDENCE, MORTALITY, AND mouth, corresponding to the International Classification SURVIVAL of Diseases, 10th revision [ICD-10], codes C00, C02, C03, C04, C05, and C06, respectively. Two-thirds of Incidence and Mortality1 the global incidence of oral cancer occurs in low- and Oral cancer incidence and mortality are high in India; middle-income countries (LMICs); half of those cases Papua New Guinea; and Taiwan, China, where chew- are in South Asia. India alone accounts for one-fifth of ing of betel quids with tobacco or without tobacco or all oral cancer cases and one-fourth of all oral cancer areca nut chewing is common, as well as in Eastern deaths (Ferlay and others 2013). Europe, France, and parts of South America (Brazil Tobacco use, in any form, and excessive alcohol use and Uruguay), where tobacco smoking and alcohol are the major risk factors for oral cancer. With dietary consumption are high. The age-standardized inci- deficiencies, these factors cause more than 90 percent dence rates for men are, on average, twice as high as of oral cancers. Preventing tobacco and alcohol use and those for women (tables 5.1 and 5.2). Incidence rates increasing the consumption of fruits and vegetables can do not follow a particular pattern from low- to high- potentially prevent the vast majority of oral cancers income countries (HICs), when countries are grouped (Sankaranarayanan and others 2013). When primary into wealth strata (figure 5.1). In selected countries prevention fails, early detection through screening and where some reliable cancer registries exist, India is relatively inexpensive treatment can avert most deaths. highest and Belarus is lowest, with incidence rates However, oral cancer continues to be a major cancer varying by more than five times in men and women. Corresponding author: Rengaswamy Sankaranarayanan, MD, International Agency for Research on Cancer, SankarR@iarc.fr 85 Table 5.1 Oral Cancer in Men (All Ages): Global Incidence, Mortality, and Prevalence, World Health Organization Geographic Classification, 2012 Incidence Mortality Prevalence Population Number ASR (W) Number ASR (W) Number Five-year World 198,975 5.5 97,919 2.7 198,267 467,157 More developed regions 68,042 7 23,380 2.3 67,978 195,233 Less developed regions 130,933 5 74,539 2.8 130,289 271,924 WHO Africa region 8,009 3.4 5,026 2.2 7,763 18,446 WHO Americas region 31,898 5.9 8,532 1.5 31,805 94,953 WHO East Mediterranean region 11,601 5.1 6,185 2.8 11,533 27,236 WHO Europe region 45,567 7.1 18,621 2.8 45,499 118,151 WHO South-East Asia region 70,816 8.9 45,247 5.7 70,667 122,976 WHO Western Pacific region 31,013 2.7 14,292 1.2 30,929 85,233 Africa 10,230 3.3 6,083 2.1 9,961 23,560 Latin America and Caribbean 12,988 4.6 5,244 1.9 12,918 32,424 Asia 111,994 5.2 65,045 3 111,683 230,389 Europe 42,573 7.5 17,598 3 42,539 111,347 Oceania 2,280 9.6 661 2.7 2,279 6,908 Source: Incidence/mortality data: Ferlay and others 2013. Prevalence data: Bray and others 2013. Note: ASR (W) = age-standardized incidence rate per 100,000 population, for the world population structure; WHO = World Health Organization. Table 5.2 Oral Cancer in Women (All Ages): Global Incidence, Mortality, and Prevalence, World Health Organization Geographic Classification, 2012 Incidence Mortality Prevalence Population Number ASR (W) Number ASR (W) Number Five-year World 101,398 2.5 47,409 1.2 100,784 234,992 More developed regions 32,781 2.6 9,908 0.6 32,683 93,180 Less developed regions 68,617 2.5 37,501 1.4 68,101 141,812 WHO Africa region 5,475 2 3,504 1.4 5,349 12,766 WHO Americas region 17,302 2.6 4,271 0.6 17,204 48,526 WHO East Mediterranean region 9,080 4.1 4,812 2.2 8,993 21,570 WHO Europe region 20,366 2.4 6,556 0.7 20,305 51,933 WHO South-East Asia region 32,648 3.9 20,487 2.5 32,482 58,034 WHO Western Pacific region 16,511 1.3 7,776 0.6 16,435 42,123 Africa 7,046 2 4,258 1.3 6,892 16,409 Latin America and Caribbean 7,645 2.2 2,381 0.7 7,586 17,813 Asia 56,856 2.5 32,363 1.4 56,549 117,362 Europe 18,843 2.5 6,033 0.7 18,789 48,653 Oceania 1,351 5.3 484 1.9 1,350 4,042 Sources: Incidence/mortality data: Ferlay and others 2013. Prevalence data: Bray and others 2013. Note: ASR (W) = age-standardized incidence rate per 100,000 population, for the world population structure; WHO = World Health Organization. 86 Cancer The estimated age-standardized incidence rates of oral However, a recent increase in cancers at the base of the cancer also vary among countries in different regions tongue,possiblydrivenbythehumanpapillomavirus(HPV), (maps 5.1 and 5.2). has been observed in white men in the United States The buccal (cheek) mucosa is the most common site (Saba and others 2011). for oral cancer in South and Southeast Asia; in all other Oral cancer incidence and mortality rates have been regions, the tongue is the most common site (Forman and declining steadily in most European countries over others 2013). Regional variations in incidence and the site the past two decades; until recently, rates had been of occurrence relate to the major causes, which are alcohol and smoking in Western countries, and betel quid and tobacco chewing in South and Southeast Asia (Lambert Figure 5.1 Age-Standardized Incidence and Mortality Rates of Oral and others 2011). Oral cancer mortality rates range Cancer, by World Bank Income Classification, 2012 between 1 and 15 per 100,000 persons in different regions; mortality rates exceed 10 per 100,000 in Eastern European Men Women countries, such as the Czech Republic, Hungary, and the Low-income Slovak Republic (Ferlay and others 2013). Oral cancer mortality rates are influenced by oral cancer incidence, Lower-middle-income (except India) access to treatment, and variations in site distribution. Upper-middle-income The observed trends in incidence and mortality (except China) among men and women are closely correlated with High-income the patterns and trends in tobacco and alcohol use. An increasing trend in incidence has been reported China in Karachi, Pakistan (Bhurgri and others 2006), and India in Taiwan, China (Tseng 2013), caused by increases in tobacco and areca nut chewing and alcohol drinking. 20 10 0 10 20 Oral cancer incidence and mortality rates have been Age-standardized rate (w) per 100,000, 0–69 years steadily declining over the past two decades because of Incidence Mortality declining smoking prevalence and alcohol consumption in the United States (Brown, Check, and Devesa 2011). Source: Ferlay and others 2013. Map 5.1 Age-Standardized Incidence Rates of Oral Cancer in Men, 2012 IBRD 41382 | FEBRUARY 2015 Lip, Oral Cavity Cancer in Men, 0–69 years ASR (World) Indcidence Greater than 108 76–108 46–76 31–46 Less than 31 No data Source: Ferlay and others 2013. Note: ASR = Age-Standardized Rate. Oral Cancer: Prevention, Early Detection, and Treatment 87 Map 5.2 Age-Standardized Incidence Rates of Oral Cancer in Women, 2012 Lip, Oral Cavity Cancer in Women, 0–69 years ASR (World) Indcidence Greater than 38 28–38 2–28 1.4–2 Fewer than 1.4 No data Source: Ferlay and others 2013. Note: ASR = Age-Standardized Rate. increasing in some Central European countries, includ- ORAL CANCER: RISK FACTORS AND ing Hungary and the Slovak Republic, reflecting changes PREVENTION in alcohol and tobacco consumption (Bonifazi and oth- ers 2011). Oral cancer mortality has declined steadily in The major causes of oral cancer worldwide remain France since reaching a peak in the early 1990s, and the tobacco in its many different forms, heavy consump- decline correlates with the reduction in per capita alco- tion of alcohol, and, increasingly, infection with certain hol consumption. Incidence and mortality have been types of HPV. Although the relative contribution of risk stable in the Nordic countries, the Russian Federation, factors varies from population to population, oral can- and the United Kingdom. Mortality rates have been cer is predominantly a disease of poor people (Johnson steadily declining in Australia and Hong Kong SAR, and others 2011). Prevention of this devastating disease China, but increasing in Japan and the Republic of Korea can come from fundamental changes in socioeconomic (Yako-Suketomo and Matsuda 2010). status, as well as from actions to reduce the demand, production, marketing, and use of tobacco products and alcohol (Johnson and others 2011). A healthy diet, Survival good oral and sexual hygiene, and awareness of the In the United States, five-year survival improved by signs and symptoms of disease are important. Success more than 11 percentage points between 1992 and 2006 depends on political will, intersectoral action, and (Pulte and Brenner 2010) and is now approximately culturally sensitive public health messages dissemi- 65 percent (Howlader and others 2010; Ries and oth- nated through educational campaigns and mass media ers 2008). In Europe, it is approximately 50 percent initiatives. (Sant and others 2009). In India, five-year survival is less than 35 percent; in China, the Republic of Korea, Pakistan, Singapore, and Thailand, it ranges between Smokeless and Smoking Tobacco Use 32 and 54 percent (Sankaranarayanan and others 2010; Smokeless tobacco in the form of betel quid, oral snuff, Sankaranarayanan and Swaminathan 2011). Overall, the and betel quid substitutes (locally called guktha, nass, five-year survival for early, localized cancers exceeds 80 naswar, khaini, mawa, mishri, and gudakhu) increases percent and falls to less than 20 percent when regional the risk of oral precancerous lesions and oral cancer lymph nodes are involved. between 2-fold and 15-fold (Gupta and others 2011; 88 Cancer Gupta, Ariyawardana, and Johnson 2013; IARC 2004b, Areca Nut Chewing 2007; Javed and others 2010; Johnson and others Areca nut or betel nut, because it is often wrapped in 2011; Somatunga and others 2012). In most areas, betel leaf, is now regarded as a type 1 carcinogen (IARC betel quid consists of tobacco, areca nut, slaked lime, 2004b, 2007). It is chewed raw, dried, or roasted, or as catechu, and several condiments, wrapped in a betel part of betel quid, by millions of people in Asia; its use leaf. In recent years, small, attractive, and inexpensive is spreading across the Pacific, as well as in emigrant sachets of betel quid substitutes containing a flavored Asian communities worldwide. Cheap, prepackaged and sweetened dry mixture of areca nut, catechu, and areca nut products, such as pan masala, are of recent slaked lime with tobacco (gutkha) or without tobacco concern, especially among youth. The inclusion of (pan masala), often claiming to be safer products, have tobacco in the betel quid adds considerably to the car- become widely available and are increasingly used by cinogenicity (Amarasinghe and others 2010; Johnson young people, particularly in India. These products and others 2011). have been strongly implicated in oral submucous fibrosis (OSMF), which places individuals at high risk for malignancy. Alcohol Use More than 50 percent of oral cancers in India, Epidemiological studies indicate that drinking alcoholic Sudan, and the Republic of South Sudan, and about beverages increases the risk of oral cancer twofold to six- 4 percent of oral cancers in the United States, are fold and is an independent risk factor (IARC 2010), with attributable to smokeless tobacco products. Smokeless risk increasing with quantity consumed. The risk varies tobacco use among young people is increasing in South by population and individual and subsite within the oral Asia, with the marketing of conveniently packaged cavity (Radoi and others 2013). The combined use of products made from areca nut and tobacco; as a con- alcohol and tobacco has a multiplicative effect on oral sequence, oral precancerous conditions in young adults cancer risk. The various pathways by which alcohol may have increased significantly (Gupta and others 2011; exert carcinogenic influence include topical exposure Sinha and others 2011). leading to a direct effect on cell membranes, altered cell Consistent evidence from many studies indicates permeability, variation in enzymes that metabolize alco- that tobacco smoking in any form increases the risk hol, and/or systemic effects, such as nutritional deficiency, of oral cancer by twofold to tenfold in men and immunological deficiency, and disturbed liver function. women (IARC 2004a). Risk increases substantially with A recent review failed to identify an association between duration and frequency of tobacco use; risk among the use of mouthwash containing alcohol and oral cancer former smokers is consistently lower than among risk, or any significant trend in risk with increasing daily current smokers, and there is a trend of decreasing use of mouthwash (Gandini and others 2012). risk with increasing number of years since quitting. Use of smokeless tobacco and alcohol in combination with tobacco smoking greatly increases the risk of Poor Nutrition oral cancer. The biological plausibility is provided by the identification of several carcinogens in tobacco, High consumption of fruits and vegetables is associated the most abundant and strongest being tobacco-spe- with a reduction of 40–50 percent in the risk of oral cific N-nitrosamines, such as N-nitrosonornicotine cancer (Lucenteforte and others 2009; Pavia and others and 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone 2006; World Cancer Research Fund/American Institute (IARC 2007). These are formed by N-nitrosation of for Cancer Research 2007). In HICs, selected aspects of nicotine, the major alkaloid responsible for addiction diet—such as lack of vegetables and fruits—may account to tobacco. for 15–20 percent of oral cancers; this proportion is The fact that more than 80 percent of oral cancers can likely to be higher in LMICs. Chemoprevention studies be attributed to tobacco and/or alcohol consumption have not established a preventive effect of retinoid and justifies regular oral examinations targeting tobacco and carotenoid dietary supplements (Chainani-Wu, Epstein, alcohol users, as well as prevention efforts focusing on and Touger-Decker 2011; Wrangle and Khuri 2007). tobacco and alcohol control (Radoi and others 2013). The World Health Organization Framework Convention on Tobacco Control, an evidence-based international Other Risk Factors treaty, aims to reduce the demand for tobacco globally by Genetic Factors price, tax, and non-price measures. (See chapter 10 for a Most carcinogens are metabolized through the full discussion of tobacco control.) cytochrome p450 system in the liver. If this system is Oral Cancer: Prevention, Early Detection, and Treatment 89 defective by virtue of inheriting a particular form of Very early preclinical invasive cancers (early-stage the gene (a polymorphism), the risk of many cancers is cancers without symptoms) present as painless small enhanced. This risk is particularly important with oral ulcers, nodular lesions, or growths. These changes can and other head and neck cancers, although the relative be easily seen and are clinically detectable through risks are modest at 1.5 or lower (that is, less than a dou- careful visual inspection and palpation of the oral bling of risk) (Lu, Yu, and Du 2011). mucosa. Early, localized oral cancers—less than four Polymorphisms in alcohol-metabolizing enzymes centimeters—that have not spread to the regional lymph also contribute to the risk. Individuals with the fast- nodes can be effectively treated and cured with surgery metabolizing version (allele) of alcohol dehydrogenase or radiotherapy alone, with no functional or cosmetic (ADH3[1-1]) have a greater risk of developing oral can- defects, resulting in five-year survival rates exceeding cer in the presence of alcoholic beverage consumption 80 percent. than those with the slow-metabolizing forms; this higher Leukoplakia is a white plaque that may be catego- risk re-enforces the role of acetaldehyde as the carcino- rized clinically as homogeneous or nonhomogeneous. gen involved (Harty and others 1997). Homogeneous lesions are thin, flat, uniform, smooth, and white. Nonhomogeneous lesions may have a white Mate Drinking and red appearance or tiny, white, pinhead-size raised Mate, a leaf infusion that is commonly drunk many nodules on a reddish background or a proliferative, times a day in parts of South America—usually very warty appearance. Erythroplakia presents as a red patch hot—appears to enhance the risk of oral cancer by a with smooth or granular surface that cannot be charac- small amount (Deneo-Pellegrini and others 2012). terized clinically or pathologically as any other definable disease (Warnakulasuriya, Johnson, and Van Der Waal Viruses 2007). Erythroplakia has a higher probability than leu- Recent evidence suggests that HPV infection may be koplakia to harbor occult invasive cancer and to undergo an independent risk factor for cancer of the base of malignant transformation. the tongue, tonsils, and elsewhere in the oropharynx. Oral lichen planus may present as interlacing white HPV may modulate the process of carcinogenesis in lines (known as Wickham’s striae) with a reddish border, some tobacco- and alcohol-induced oral and oro- or as a mix of reddish and ulcerated areas. pharyngeal cancers, and it may act as the primary OSMF, mostly restricted to people of Indian sub- oncogenic agent for inducing carcinogenesis among continent origin and in certain Pacific islands such as nonsmokers (Johnson and others 2011; Prabhu and Mariana Islands, presents with a burning sensation, Wilson 2013). Growing evidence suggests that such blanching of the oral mucosa, and intolerance to spicy oropharyngeal infections can be sexually transmitted food. Stiffening and atrophy of the oral and pharyngeal (Heck and others 2010). mucosa occurs as the disease progresses, leading to reduced mouth opening and difficulty in swallowing Chronic Trauma and speaking. It now seems clear that chronic trauma, from sharp Palatal lesions are seen in populations who smoke teeth, restorations, or dentures, contributes to oral can- with the lighted end of the tobacco product inside the cer risk, although this higher risk commonly occurs only mouth, known as reverse smoking, resulting in white or in the presence of the other local risk factors (Piemonte, mixed reddish-white lesions of the palate. Lazos, and Brunotto 2010). A higher risk of malignant transformation may be associated with the following factors: female gender, lesions of long duration, large precancerous lesions, ORAL CANCER: NATURAL HISTORY precancerous lesions in nonusers of tobacco, tongue Oral cancer has a long preclinical phase that consists and floor of mouth lesions, nonhomogeneous lesions, of well-documented precancerous lesions. The pre- and lesions showing epithelial dysplasia and aneuploidy cancerous lesions include homogeneous leukoplakia, (Hsue and others 2007; Napier and Speight 2008). nonhomogeneous leukoplakia, verrucous leukopla- However, it is impossible to predict with certainty which kia, erythroplakia, OSMF, lichen planus, and chronic precancerous lesion will become malignant during traumatic ulcers. The estimated annual frequency follow-up in patients. The malignant transformation of of malignant transformation of oral precancerous precancerous lesions can be prevented by interventions, lesions ranges from 0.13 percent to 2.2 percent such as avoiding exposure to tobacco use and alcohol (Amagasa, Yamashiro, and Uzawa 2011; Napier and drinking, and in selected instances, by excision of the Speight 2008). lesions. 90 Cancer ORAL CANCER SCREENING: ACCURACY, users following three rounds of oral visual screening EFFICACY, AND POTENTIAL HARMS has been demonstrated in a cluster-randomized con- trolled trial in India (Sankaranarayanan and others Although an affordable, acceptable, easy to use, accu- 2005; Sankaranarayanan and others 2013). A 15-year rate, and effective screening test for oral cancer is follow-up found sustained reduction in oral cancer available in high-risk countries, a decision to introduce mortality, with larger reductions in those adhering population-based screening should take into account to repeated screening rounds; there was a 38 percent the level of health service development and available reduction in oral cancer incidence (95 percent confi- resources to meet the increased treatment demand that dence interval [CI] 8–59 percent), and an 81 percent screening generates. The target population for oral can- reduction in oral cancer mortality (95 percent CI cer screening consists of those age 30 years and older 69–89 percent) in tobacco and/or alcohol users who who use tobacco and/or alcohol. were screened four times (Sankaranarayanan and others Visual screening of the oral cavity has been widely 2013). The studies (Sankaranarayanan and others 2005; evaluated for its feasibility, safety, acceptability, accuracy Sankaranarayanan and others 2013) were the basis for to detect oral precancerous lesions and cancer, and effi- the conclusions of the recent Cochrane Collaboration cacy and cost-effectiveness in reducing oral cancer mor- Review (Brocklehurst and others 2013) and an American tality (Johnson and others 2011; Sankaranarayanan and Dental Association (ADA) expert panel review on others 2005; Sankaranarayanan and others 2013). Visual population-based oral cancer screening (Rethman screening involves systematic visual and physical exami- and others 2010). The ADA review recommended that nation of the intraoral mucosa under bright light for clinicians look for signs of precancerous lesions or signs of oral potentially malignant disorders (OPMDs), early-stage cancers while performing routine visual as well as early oral cancer, followed by careful inspec- and tactile screening in all subjects, particularly in tion and digital palpation of the neck for any enlarged those who use tobacco or alcohol or both; the panel lymph nodes. It is a provider-dependent, subjective test; also concluded that the life-saving benefits for subjects accordingly, its performance in detecting lesions varies with treatable lesions were more important than the among providers. Comprehensive knowledge of the oral potential harms incurred by those with benign or non- anatomy, the natural history of oral carcinogenesis, and progressive lesions (Rethman and others 2010). The the clinico-pathological features of the OPMDs and pre- Cochrane Review (Walsh and others 2013) concluded clinical cancers are important prerequisites for efficient that evidence suggests that a visual examination as part providers of oral visual screening. of a population-based screening program reduces the The potential harms of oral visual screening may mortality rate of oral cancer in high-risk individuals; in include additional diagnostic investigations, such as addition, it could result in diagnoses of oral cancer at incisional or excisional biopsy; anxiety associated with an earlier stage of disease and improvement in survival false-positive screening tests; detection and treatment of rates across the population as a whole (Brocklehurst and biologically insignificant conditions that may have no others 2013). impact on oral cancer incidence; and false reassurance The U.S. Preventive Services Task Force released a from false-negative tests. draft Recommendation Statement, which stated that for adults age 18 years or older seen in primary care settings, the current evidence is insufficient to assess the balance Visual Screening by Health Care Personnel of benefits and harms of screening for oral cancer in A variety of health care personnel—including dentists, asymptomatic adults. However, this statement overlooks general practitioners, oncologists, surgeons, nurses, the benefits of early detection of oral cancers among and auxiliary health workers—may provide oral visual users of tobacco or alcohol or both, as well as other screening after training (Ramadas and others 2008). benign conditions whose early detection may improve Sensitivity ranges from 40 percent to 93 percent, oral health. Discouraging oral visual examination in and specificity ranges from 50 percent to 99 percent primary care is clearly not in the interests of oral cancer for detecting precancerous lesions and early asymp- control and improving oral health (Edwards 2013). tomatic oral cancers (Downer and others 2004; Mathew and others 1997; Mehta and others 1986; Warnakulasuriya and others 1984; Warnakulasuriya Self-Examination and Other Screening Methods and Nanayakkara 1991). Although mouth self-examination using a mirror has A significant reduction of 34 percent in oral cancer been evaluated as a screening test in some studies mortality among a high-risk group of tobacco or alcohol (Elango and others 2011; Mathew and others 1995; Scott Oral Cancer: Prevention, Early Detection, and Treatment 91 and others 2010), whether it could lead to reductions in assessment of the upper aerodigestive tract is necessary oral cancer mortality is not known. There is insufficient because patients with oral cancer have a high risk of evidence to recommend the routine use of other oral cancers developing in other head and neck sites and in screening tests, such as toluidine blue staining, chemilu- the lungs. minescence, tissue fluorescence imaging, tissue fluores- Once a diagnosis of oral cancer is confirmed, stag- cent spectroscopy, and salivary analysis and cytology for ing assessment is completed and treatment is planned. primary screening of oral cancer (Johnson and others The Union for International Cancer Control Tumor, 2011; Patton, Epstein, and Kerr 2008; Richards 2010; Su Nodes, Metastasis (TNM) staging system is widely used and others 2010). for staging oral cancer (Patel and Shah 2005; Sobin Despite the high risk of oral cancer in the Indian and Wittekind 2002) (table 5.3): T indicates the size subcontinent, no national or regional screening pro- and extent of spread of the primary tumor, N indicates grams exist in the region. The only large-scale, ongoing, the extent spread to the regional lymph nodes in the national oral cancer screening programs are in Cuba and neck, and M indicates the spread to distant organs. Taiwan, China. The TNM categorization is further grouped into stages 0 through IV, which denote increasing severity of disease • The Cuban program has been in existence since 1984. and decreasing survival. An evaluation conducted in 1994 indicated that 12–26 Oral cancer staging involves assessing the clinical percent of the target population has been screened extent of disease through physical examination, biop- annually, but less than 30 percent of screen-positive sies, and imaging investigations, including X-rays of individuals complied with referrals (Fernandez and the mandible, maxillary sinuses, and chest; comput- others 1995). The program was reorganized in 1996, erized tomography (CT) scans; magnetic resonance with the target age raised from 15 years to 35 years, imaging (MRI); and positron emission tomography screening intervals increased from one to three years, (PET) imaging, depending on what resources are avail- and the referral system revamped. No further formal able. Advanced imaging techniques such as CT, MRI, evaluation has been conducted, but there has been no and PET may be useful in more accurately evaluating reduction in oral cancer incidence or mortality rates local spread, such as invasion of muscles, bone, and in Cuba over the past three decades. The outcomes cartilage, and lymph node metastases, as well as in from the Cuban program emphasize that screen- planning treatment, but these investigations are seldom ing programs without efficient organization and feasible in LMICs. resources are not an effective use of limited resources. • Oral cancer screening was initiated in Taiwan, China, in 2004, targeting those age 18 years and older who ORAL CANCER: MANAGEMENT were smokers or betel nut chewers; the target popula- Oral cancer is predominantly a loco-regional disease tion for oral cancer screening was revised in 2010 to that tends to infiltrate adjacent bone and soft tissues cover smokers or chewers age 30 years and older. The and spreads to the regional lymph nodes in the neck. screening program has led to almost half of the oral Distant metastasis is uncommon at the time of diagnosis. cancers diagnosed in stages I and II, with a declining A thorough inspection and palpation of the oral cavity trend in oral cancer mortality rates. and examination of the neck is mandatory. CT and MRI imaging are widely used to assess the extent of involve- ment of adjacent structures, such as bones and soft tis- ORAL CANCER: EARLY CLINICAL DIAGNOSIS sues. Surgery and radiotherapy are the main treatment AND STAGING modalities. Given the skills, expertise, and infrastructure required for staging and treatment with minimal physi- Primary care dental and general practitioners should cal, functional, and cosmetic morbidity, oral cancer treat- play a major role in referring patients to cancer treatment ment is usually provided in specialized cancer hospitals, facilities for early diagnosis and treatment. Improving such as comprehensive cancer centers, or in hospitals at the skills of these primary care doctors is essential to the highest level of health services, third-level centers. improving prospects for early diagnosis, particularly among patients who use tobacco or alcohol in any form. Routine biopsy in those clinically presenting with fea- Treatment of Early-Stage Oral Cancer (Stages I and II) tures of precancerous lesions may lead to early diagnosis Surgery and radiotherapy are widely used for the of underlying invasive oral cancer. In addition to his- treatment of early oral cancer, either as single modalities tory, physical examination, and biopsy, a simultaneous or in combination. The choice of modality depends on 92 Cancer Table 5.3 Clinical Staging of Oral Cancer, Treatment Modalities, and Prognosis, by Clinical Stage Composite Five-year stage Extent of disease TNM category Treatment options survival (percent) 0 Cancer is limited to the epithelium (carcinoma TisN0M0 Limited surgical excision ~100 in-situ) (Tis) and has not spread to deeper layers and nearby organs, regional (neck) lymph nodes (N0), or distant organs (M0) I Primary tumor measures 2 cm or less (T1) and has T1N0M0 Radical surgery or radical > 90 not spread to regional organs, regional (neck) lymph radiotherapy nodes (N0), or distant organs (M0) II Primary tumor is larger than 2 cm and smaller than T2N0M0 Radical surgery or radical > 70 4 cm (T2) and has not spread to regional organs, radiotherapy; in selected regional (neck) lymph nodes (N0), or distant organs cases, combination (M0) therapy III Primary tumor measures > 4 cm (T3) and has not T3N0M0 Combined modality 30–40 spread to neck nodes (N0) or distant organs (M0); or T1 to T3, N1, M0 treatment with surgery 20–25 tumor is any size (T1 to T3) and has spread to one and/or radiotherapy and/ lymph node measuring 3 cm or less on the same side or chemotherapy of the of the neck (N1) as the primary tumor and the cancer has not spread to distant organs (M0) IV Tumor involves nearby structures, including the T4, N0, or N1, M0 Multimodality 5–10 mandible, tongue muscles, maxillary sinus, and skin Any T, N2, or N3 M0 management treatment <5 (T4); or tumor is any size but involves one lymph with surgery and/ node measuring 3–6 cm on the same side of the Any T, any N, M1 or radiotherapy and/ neck (N2a) or one lymph node measuring no more or chemotherapy for than 6 cm on the opposite side of the neck (N2b), or cancers without distant two or more lymph nodes no more than 6 cm on any metastases; palliative side of the neck (N2c); or lymph node involvement radiotherapy and/or measuring more than 6 cm (N3); or distant chemotherapy and pain/ metastases (M1) symptom relief measures Source: Staging: Sobin and Wittekind 2002. Note: cm = centimeter; TNM = Tumor, Node, Metastasis. the location of the tumor, cosmetic and functional out- for small tumors (Fujita and others 1999; Marsiglia comes, age of the patient, associated illnesses, patient’s and others 2002; Wendt and others 1990). Deep infil- preference, and the availability of expertise. trative cancers have a high propensity to spread to Most early-stage oral cancers can be locally excised regional lymph nodes; therefore, brachytherapy alone, or treated with radiotherapy, with no or minimal func- which does not treat regional nodes adequately, is tional and physical morbidity. Elective neck dissection not recommended. Newer techniques, such as three- to remove lymph nodes may be considered in selected dimensional conformal radiotherapy and intensity cases, such as patients with stage I tongue cancer and modulated radiotherapy, can minimize the side effects stage II cancers at other oral sites, who may be at high of radiotherapy by delivering the radiation dose to the risk of microscopic but not clinically evident involve- tumor more precisely and accurately while avoiding ment of the neck nodes (N0) (El-Naaj and others 2011; healthy surrounding tissues. However, these treatments Hicks, Jr., and others 1997; Vijayakumar and others require advanced equipment and are more expensive 2011; Woolgar 2006; Zwetyenga and others 2003). than conventional radiotherapy. External beam radiotherapy and brachytherapy— using radioactive sources implanted in the tumor— either alone or in combination, is an alternative to Treatment of Locally Advanced Tumors of the Oral surgery for early-stage oral cancers. Excellent outcomes Cavity (Stages III and IV) have been demonstrated following brachytherapy alone Locally advanced tumors are aggressive, and loco- or in combination with external beam radiotherapy regional treatment failure rates are high. A combined Oral Cancer: Prevention, Early Detection, and Treatment 93 modality approach integrating surgery, radiotherapy Complications of Surgery with or without chemotherapy, and planned and exe- The common complications of oral surgery are infec- cuted by a multidisciplinary team is always preferred. tion, collection of blood (hematoma), skin necrosis, Appropriate importance should be given to factors flap failure, and wound breakdown. Resorption of such as functional and cosmetic outcomes and the bone, osteomyelitis, and salivary fistula can also occur. available expertise. Surgery followed by postoperative Complications are more frequent when neck dissection radiotherapy is the preferred modality for patients with is part of the surgery. Fatal hemorrhage can occur if the deep infiltrative tumors and those with bone infiltration carotid artery is exposed in the wound. Resection of the (Lundahl and others 1998). Postoperative concurrent structures can interfere with cosmetic appearance and chemo-radiation has been found to be superior to radio- functions such as speech, swallowing, and airway. These therapy alone in those with surgical margins showing complications can be minimized through reconstructive cancerous changes indicating incomplete excision of surgery and by good prosthetic rehabilitation. the tumor (Bernier and others 2004; Cooper and others 2004). The use of chemotherapy prior to surgery may eliminate the need to remove the mandible—a major Posttreatment Follow-Up benefit—although it does not confer a survival benefit Patients with oral cancer are at risk for developing (Licitra and others 2003). loco-regional recurrences and second malignancies. Primary radiotherapy, with or without chemother- After completion of the treatment, patients should be apy, is a reasonable option for locally advanced tumors followed up at regular intervals to detect any signs of without bone involvement, especially for patients who recurrence. Patients should be encouraged to give up have inoperable disease, who are medically unfit for sur- tobacco and alcohol and know the signs and symptoms gery, or who are likely to have unacceptable functional of recurrence. and cosmetic outcomes with surgery. Incorporating chemotherapy with surgery or radiotherapy is useful in younger patients who are in good general condition, Prognosis increasing survival by about 5 percentage points at five years (Blanchard and others 2011). Lymph node involvement and tumor size are the most important prognostic factors. Data for the United States for 1975–2007 report a five-year survival for all stages of Side Effects of Radiotherapy oral cancer of 60.9 percent, 82.5 percent for early-stage Side effects may occur during or immediately following disease, and 54.7 percent for locally advanced oral radiotherapy—acute reactions—or months to years cavity cancer (Ries and others 2008). The reported five- after treatment. Acute reactions are self-limiting and year overall survival rates for oral cancer for all stages generally resolve within two to three weeks. These combined from populations in LMICs such as China, reactions are caused by the inflammation of tissues Cuba, India, Pakistan, and Thailand ranged from 26 to within the radiotherapy treatment field. Alteration of 45 percent; for stages I and II, the survival rates ranged taste, pain, difficulty in eating, mucosal ulceration of from 36 to 83 percent. The inferior survival rates in the oral cavity, bacterial and fungal infections, increased LMICs versus HICs reflect disparities in the availabil- thickness of saliva, discoloration and peeling of the ity, accessibility, and affordability of diagnostic and overlying skin, loss of hair within the field of treat- treatment services (Sankaranarayanan and others 2010; ment, and edema of the skin are the major side effects. Sankaranarayanan and Swaminathan 2011). Maintenance of good oral hygiene, frequent cleaning of the oral cavity with soda-saline solution, analgesics, and control of infection are recommended for conservative ECONOMICS OF PREVENTING AND management of these side effects. Good hydration, a SCREENING FOR ORAL CANCERS IN LMICS high-calorie diet, and avoidance of spicy and hot food are recommended. Cost-Effectiveness Assessments Late effects of radiation are related to dose per frac- Only a few cost-effectiveness studies of oral cancer tion, total dose, and the type and volume of the tissue screening focus on LMICs; therefore, we include a irradiated. Late effects include loss of hair within the broader range of studies, including some from HICs. irradiated area, dry mouth (xerostomia), thickening of Although these studies may not be directly relevant to the skin, dental caries, and, rarely, necrosis of the man- resource-limited settings, they provide valuable insights dible or maxillary bone. into the potential cost-effectiveness of interventions. 94 Cancer Primary Prevention reflects the costs and effectiveness likely to be experi- Interventions targeted at reducing or eliminating tobacco enced in LMICs. In general, screening was at ages 35 or and alcohol use should be considered for implementation 40 years and older; three of the four studies included when shown to be cost-effective. All the interventions both high-risk and average-risk individuals. All of the presented are cost-effective, even for LMICs. In the case studies presented incremental cost-effectiveness, com- of tobacco cessation, increasing the price of tobacco prod- pared with the scenario of no screening. A variety of ucts is the most cost-effective approach, with incremental interventions was assessed, using invitation and oppor- cost-effectiveness ratios ranging from US$4 to US$34 tunistic screening; visual inspection was performed by per disability-adjusted life year. Alcohol control inter- specialists (oral cancer surgeons), dentists, or trained ventions tend to have higher cost-effectiveness ratios; health care workers. advertising bans and reduced access range from US$367 The results indicate that screening is cost-effective to US$1,307; combination strategies (including price even in LMICs. The study from India provides evidence increases, reduced access, and advertisement bans) range that oral cancer screening by visual inspection costs less from US$601 to US$1,704. (Interventions to decrease than US$6 per person in a screening program; this has tobacco use are covered in more detail in chapter 10.) an incremental cost-effectiveness ratio of US$835 per life year saved. The most cost-effective and affordable Screening option in the limited-resource setting is to offer oral Table 5.4 summarizes findings from the relevant cost- cancer screening to high-risk individuals, for exam- effectiveness studies. Among the four studies of the ple, tobacco and alcohol users. The incremental cost- cost-effectiveness of oral cancer screening, three—all set effectiveness ratio for screening high-risk individuals in HICs—used decision analytic modeling; the other, in southern India is US$156 per life year saved. There the only one from a resource-constrained environment, is wide variation in the incremental cost-effectiveness used data from a randomized clinical trial in India. Only reported across the studies, probably because of factors the Indian study (Subramanian and others 2009) directly such as the underlying prevalence of disease and the Table 5.4 Oral Cancer Screening Cost-Effectiveness Studies Methodology/ Interventions/tests Cost-effectiveness Study Country Setting/population cost data compared assessment Van Der Meij, Netherlands Individuals with OLP Decision analytic model: Screening by oral US$53,430 per ELS or Bezemer, and all relevant clinical specialist or dentist US$2,137 per QALY for Van Der Waal costs versus no screening screening by specialist, 2002 compared with no screening: lower cost- effectiveness ratio for screening by dentists Speight and United Kingdom Screening programs Decision analytic model: No screening compared Opportunistic screening others 2006 for individuals ages all relevant invitation with invitation and of high-risk individuals 40 years and older in and clinical costs opportunistic screening ages 40–60 years primary care settings most cost-effective: US$19,000 per QALY Subramanian India 13 clusters randomized Cost-effectiveness Visual inspection by US$835 per LYS for all and others in Kerala; those ages assessment, including trained health care individuals; US$156 2009 35 years or older were program and clinical workers compared with per LYS for high-risk eligible for the study costs usual care individuals Dedhia and United States Community-based Markov model: clinical Oral exam (visual A budget of US$3,363 others 2011 screening for high-risk costs included but no inspection and manual per person over men (ages 40 years or program costs palpation) by trained a 40-year cycle older, tobacco and/or health care workers, for screening is alcohol users) compared with no cost-effective screening Note: ELS = equivalent life saved; LYS = life year saved; OLP = oral lichen planus; QALY = quality-adjusted life year. Oral Cancer: Prevention, Early Detection, and Treatment 95 local cost of cancer treatment. The cost of care related and optimum treatment. As this chapter has demon- to screening, diagnosis, and treatment can differ sub- strated, however, these interventions are affordable and stantially, even among countries classified as LMICs. cost-effective. Accordingly, it is essential to systematically assess costs at the country or even local level to analyze the cost- effectiveness and resources required to implement oral NOTES cancer screening. The World Bank classifies countries according to four income groupings. Income is measured using gross national income Future Research Needs per capita, in U.S. dollars, converted from local currency using the World Bank Atlas method. Classifications as of July 2014 Primary prevention, especially smoking cessation, and are as follows: secondary prevention, focused on high-risk individ- uals, are likely to be cost-effective and affordable in • Low-income countries = US$1,045 or less in 2013 LMICs. Additional studies are required to assess the cost- • Middle-income countries are subdivided: effectiveness and budget implications of visual screening • Lower-middle-income = US$1,046–US$4,125 for oral cancers in LMICs. These studies should focus • Upper-middle-income = US$4,126–US$12,745 on the screening delivery structure to identify the most • High-income countries = US$12,746 or more cost-effective approach to provide oral cancer screening to high-risk individuals. 1. Maps and figures in this chapter are based on incidence When cancer screening policies are implemented, and mortality estimates for ages 0–69, consistent with the success of the program will depend on participa- reporting in all DCP3 volumes. Global cancer statistics are estimates for the year 2012 and have been provided by tion by the target population. Even when screening and the International Agency for Research on Cancer from its follow-up care are free of charge, patients may not be GLOBOCAN 2012 database. Observable population-based able to afford to lose a day’s wages to attend screening data were derived from Cancer Incidence in Five Continents, clinics or travel to health centers to receive follow-up 10th edition and for trends over time from CI5 Plus (http:// diagnostic testing or treatments. The indirect costs ci5.iarc.fr/CI5plus/Default.aspx). The discussion of burden borne by the patients may be particularly challenging (including risk factors), however, includes all ages unless among those in the lower socioeconomic strata. These otherwise noted. Interventions also apply to all age groups, are the very individuals likely to be at higher risk for except where age ranges or cutoffs are specified. developing oral cancers; it is, therefore, vital that identi- fying approaches to encourage and sustain participation among this potentially hard-to-reach, high-risk popula- REFERENCES tion be given high priority. Amagasa T., M. Yamashiro, and N. Uzawa. 2011. “Oral Premalignant Lesions: From a Clinical Perspective.” International Journal of Clinical Oncology 16 (1): 5–14. CONCLUSION Amarasinghe, H. K., N. W. Johnson, R. Lalloo, M. Kumaraarachchi, and S. Warnakulasuriya. 2010. “Derivation and Validation A multifaceted approach that integrates health educa- of a Risk-Factor Model for Detection of Oral Potentially tion, tobacco and alcohol control, early detection, and Malignant Disorders in Populations with High Prevalence.” early treatment is needed to reduce the burden of this British Journal of Cancer 103 (3): 303–09. eminently preventable cancer. How to accomplish this Bernier J., C. Domenge, M. Ozsahin, K. Matuszewska, is known; astonishingly, it has not been applied in most J. L. Lefebvre, and others. 2004. “Postoperative Irradiation countries, and not at all in the high-burden countries. with or Without Concomitant Chemotherapy for Locally Improving awareness among the general public and Advanced Head and Neck Cancer.” New England Journal of primary care practitioners, investing in health services Medicine 350 (19): 1945–52. to provide screening and early diagnosis services for Bhurgri Y., A. Bhurgri, A. Usman, S. Pervez, N. Kayani, and others. 2006. “Epidemiological Review of Head and tobacco and alcohol users, and providing adequate treat- Neck Cancers in Karachi.” Asian Pacific Journal of Cancer ment for those diagnosed with invasive cancer are crit- Prevention 7 (2): 195–200. ically important oral cancer control measures. Imaging, Blanchard P., B. Baujat, V. Holostenco, A. Bourredjem, C. Baey, histopathology, cancer surgery and radiotherapy infra- and others. 2011. “Meta-Analysis of Chemotherapy in structure and services, trained professionals, and the Head and Neck Cancer (MACH-NC): A Comprehensive availability of chemotherapeutic agents are inadequate Analysis by Tumour Site.” Radiotherapy and Oncology in many LMICs, seriously compromising early detection 100 (1): 33–40. 96 Cancer Bonifazi M., M. Malvezzi, P. Bertuccio, V. Edefonti, W. Garavello, the Oral Cancer Control Program in Cuba.” Epidemiology and others. 2011. “Age-Period-Cohort Analysis of Oral 6 (4): 428–31. Cancer Mortality in Europe: The End of an Epidemic?” Forman, D., F. Bray, D. H. Brewster, C. Gombe Mbalawa, Oral Oncology 47 (5): 400–07. B. Kohler, and others. 2013. Cancer Incidence in Five Bray, F., J. S. Ren, E. Masuyer, and J. Ferlay. 2013. “Estimates Continents. Vol. 10 (Electronic Version). Lyon, France: of Global Cancer Prevalence for 27 Sites in the Adult International Agency for Cancer. http://ci5.iarc.fr. Population in 2008.” International Journal of Cancer Fujita, M., Y. Hirokawa, K. Kashiwado, Y. Akagi, K. Kashimoto, 132 (5): 1133–45. and others. 1999. “Interstitial Brachytherapy for Stage I and Brocklehurst, P., O. Kujan, L. A. O’Malley, G. Ogden, S. Shepherd, II Squamous Cell Carcinoma of the Oral Tongue: Factors and A. M Glenny. 2013. “Screening Programmes for the Influencing Local Control and Soft Tissue Complications.” Early Detection and Prevention of Oral Cancer.” Cochrane International Journal of Radiation Oncology Biology Physics Database of Systematic Reviews 11: CD004150. 44 (4): 767–75. Brown L. M., D. P. Check, and S. S. Devesa. 2011. “Oropharyngeal Gandini, S., E. Negri, P. Boffetta, V. C. La, and P. Boyle. 2012. Cancer Incidence Trends: Diminishing Racial Disparities.” “Mouthwash and Oral Cancer Risk Quantitative Meta- Cancer Causes and Control 22 (5): 753–63. Analysis of Epidemiologic Studies.” Annals of Agricultural Chainani-Wu, N., J. Epstein, and R. Touger-Decker. 2011. and Environmental Medicine 19 (2): 173–80. “Diet and Prevention of Oral Cancer: Strategies for Clinical Gupta, B., A. Ariyawardana, and N. W. Johnson. 2013. “Oral Practice.” Journal of the American Dental Association Cancer in India Continues in Epidemic Proportions: 142 (2): 166–69. Evidence Base and Policy Initiatives.” International Dental Cooper, J. S., T. F. Pajak, A. A. Forastiere, J. Jacobs, B. H. Campbell, Journal 63 (1): 12–25. and others. 2004. “Postoperative Concurrent Radiotherapy Gupta, P. C., C. S. Ray, D. N. Sinha, and P. K. Singh. 2011. and Chemotherapy for High-Risk Squamous-Cell “Smokeless Tobacco: A Major Public Health Problem in Carcinoma of the Head and Neck.” New England Journal of the SEA Region: A Review.” Indian Journal of Public Health Medicine 350 (19): 1937–44. 55 (3): 199–209. Dedhia, R. C., K. J. Smith, J. T. Johnson, and M. Roberts. 2011. Harty, L. C., N. E. Caporaso, R. B. Hayes, D. M. Winn, “The Cost-Effectiveness of Community-Based Screening E. Bravo-Otero, and others. 1997. “Alcohol Dehydrogenase 3 for Oral Cancer in High-Risk Males in the United States: Genotype and Risk of Oral Cavity and Pharyngeal Cancers.” A Markov Decision Analysis Approach.” Laryngoscope Journal of the National Cancer Institute 89 (22): 1698–705. 121 (5): 952–60. Heck, J. E., J. Berthiller, S. Vaccarella, D. M. Winn, E. M. Smith, Deneo-Pellegrini, H., E. De Stefani, P. Boffetta, A. L. Ronco, and others. 2010. “Sexual Behaviours and the Risk of Head G. Acosta, and others. 2012. “Mate Consumption and Risk and Neck Cancers: A Pooled Analysis in the International of Oral Cancer: Case-Control Study in Uruguay.” Head & Head and Neck Cancer Epidemiology (INHANCE) Neck 35 (8): 1091–95. Consortium.” International Journal of Epidemiology 39 (1): Downer, M. C., D. R. Moles, S. Palmer, and P. M. Speight. 166–81. 2004. “A Systematic Review of Test Performance in Hicks, W. L., Jr., T. R. Loree, R. I. Garcia, S. Maamoun, Screening for Oral Cancer and Precancer.” Oral Oncology D. Marshall, and others. 1997. “Squamous Cell Carcinoma 40 (3): 264–73. of the Floor of Mouth: A 20-Year Review.” Head & Neck Edwards, C. P. 2013. “Oral Cancer Screening for Asymptomatic 19 (5): 400–05. Adults: Do the United States Preventive Services Task Force Hirsch, J. M., M. Wallstrom, A. P. Carlsson, and L. Sand. Draft Guidelines Miss the Proverbial Forest for the Trees?” 2012. “Oral Cancer in Swedish Snuff Dippers.” Anticancer Oral Surgery, Oral Medicine, Oral Pathology and Oral Research 32 (8): 3327–30. Radiology 116 (2): 131–34. Howlader, N., L. A. Ries, A. B. Mariotto, M. E. Reichman, Elango, K. J., N. Anandkrishnan, A. Suresh, S. K. Iyer, J. Ruhl, and others. 2010. “Improved Estimates of Cancer- S. K. Ramaiyer, and others. 2011. “Mouth Self-Examination Specific Survival Rates from Population-Based Data.” to Improve Oral Cancer Awareness and Early Detection in a Journal of the National Cancer Institute 102 (20): 1584–98. High-Risk Population.” Oral Oncology 47 (7): 620–24. Hsue, S. S., W. C. Wang, C. H. Chen, C. C. Lin, Y. K. Chen, El-Naaj, I. A., Y. Leiser, M. Shveis, E. Sabo, and M. Peled. 2011. and L. M. Lin. 2007. “Malignant Transformation in 1458 “Incidence of Oral Cancer Occult Metastasis and Survival Patients with Potentially Malignant Oral Mucosal Disorders: of T1-T2N0 Oral Cancer Patients.” Journal of Oral & A Follow-Up Study Based in a Taiwanese Hospital.” Journal Maxillofacial Surgery 69 (10): 2674–79. of Oral Pathology & Medicine 36 (1): 25–29. Ferlay, J., I. Soerjomataram, M. Ervik, R. Dikshit, S. Eser, and IARC (International Agency for Research on Cancer). 2004a. others. 2013. GLOBOCAN 2012 V1.0, Cancer Incidence and IARC Monographs on the Evaluation of Carcinogenic Risks Mortality Worldwide: IARC Cancerbase No. 11 [Internet]. to Humans, Volume 83: Tobacco Smoke and Involuntary Lyon, France: International Agency for Research on Cancer. Smoking. Lyon, France: IARC. http://globocan.iarc.fr. ———. 2004b. IARC Monographs on the Evaluation of Fernandez, G. L., R. Sankaranarayanan, J. J. Lence Anta, Carcinogenic Risks to Humans, Volume 85: Betel-Quid and S. A. Rodriguez, and D. M. Parkin. 1995. “An Evaluation of Areca-Nut Derived Nitrosamines. Lyon, France: IARC. Oral Cancer: Prevention, Early Detection, and Treatment 97 ———. 2007. IARC Monographs on the Evaluation of Overview of the Literature.” Journal of Oral Pathology & Carcinogenic Risks to Humans, Volume 89: Smokeless Tobacco Medicine 37 (1): 1–10. and Some Tobacco-Specific N-Nitrosamines. Lyon, France: Patel, S. G., and J. P. Shah. 2005. “TNM Staging of Cancers IARC. of the Head and Neck: Striving for Uniformity among ———. 2010. IARC Monographs on the Evaluation of Diversity.” CA: A Cancer Journal for Clinicians 55 (4): Carcinogenic Risks to Humans, Volume 96: Alcohol 242–58. Consumption and Ethyl Carbamate. Lyon, France: IARC. Patton, L. L., J. B. Epstein, and A. R. Kerr. 2008. “Adjunctive Javed, F., M. Chotai, A. Mehmood, and K. Almas. 2010. “Oral Techniques for Oral Cancer Examination and Lesion Mucosal Disorders Associated with Habitual Gutka Usage: Diagnosis: A Systematic Review of the Literature.” Journal A Review.” Oral Surgery, Oral Medicine, Oral Pathology, Oral of the American Dental Association 139 (7): 896–905. Radiology, and Endodontics 109 (6): 857–64. Pavia, M., C. Pileggi, C. G. Nobile, and I. F. Angelillo. 2006. Johnson, N. W., S. Warnakulasuriya, P. C. Gupta, E. Dimba, “Association between Fruit and Vegetable Consumption M. Chindia, and others. 2011. “Global Oral Health and Oral Cancer: A Meta-Analysis of Observational Inequalities in Incidence and Outcomes for Oral Cancer: Studies.” American Journal of Clinical Nutrition 83 (5): Causes and Solutions.” Advances in Dental Research 23 (2): 1126–34. 237–46. Piemonte, E. D., J. P. Lazos, and M. Brunotto. 2010. “Relationship Lambert, R., C. Sauvaget, C. M. De Camargo, and between Chronic Trauma of the Oral Mucosa, Oral R. Sankaranarayanan. 2011. “Epidemiology of Cancer from Potentially Malignant Disorders and Oral Cancer.” Journal the Oral Cavity and Oropharynx.” European Journal of of Oral Pathology & Medicine 39 (7): 513–17. Gastroenterology & Hepatology 23 (8): 633–41. Prabhu, S., and D. Wilson. 2013. “Human Papillomavirus and Licitra L., C. Grandi, M. Guzzo, L. Mariani, V. S. Lo, and others. Oral Disease: Emerging Evidence: A Review.” Australian 2003. “Primary Chemotherapy in Resectable Oral Cavity Dental Journal 58 (1): 2–10. Squamous Cell Cancer: A Randomized Controlled Trial.” Pulte, D., and H. Brenner. 2010. “Changes in Survival in Head Journal of Clinical Oncology 21 (2): 327–33. and Neck Cancers in the Late 20th and Early 21st Century: Lu, D., X. Yu, and Y. Du. 2011. “Meta-Analyses of the Effect of A Period Analysis.” Oncologist 15 (9): 994–1001. Cytochrome P450 2E1 Gene Polymorphism on the Risk of Radoi, L., S. Paget-Bailly, D. Cyr, A. Papadopoulos, F. Guida, Head and Neck Cancer.” Molecular Biology Reports 38 (4): and others. 2013. “Tobacco Smoking, Alcohol Drinking 2409–16. and Risk of Oral Cavity Cancer by Subsite: Results of a Lucenteforte, E., W. Garavello, C. Bosetti, and V. C. La. French Population-Based Case-Control Study, the ICARE 2009.”Dietary Factors and Oral and Pharyngeal Cancer Study.” European Journal of Cancer Prevention 22 (3): Risk.” Oral Oncology 45 (6): 461–67. 268–76. Lundahl, R. E., R. L. Foote, J. A. Bonner, V. J. Suman, J. E. Lewis, Ramadas, K., E. Lucas, G. Thomas, B. Mathew, A. Balan, and and others. 1998. “Combined Neck Dissection and others. 2008. A Digital Manual for the Early Diagnosis of Postoperative Radiation Therapy in the Management of the Oral Neoplasia. Lyon, France: International Agency for High-Risk Neck: A Matched-Pair Analysis.” International Research on Cancer. http://screening.iarc.fr/atlasoral Journal of Radiation Oncology Biology Physics 40 (3): .php. 529–34. Rethman, M. P., W. Carpenter, E. E. Cohen, J. Epstein, Marsiglia, H., C. Haie-Meder, G. Sasso, G. Mamelle, and C. A. Evans, and others. 2010. “Evidence-Based Clinical A. Gerbaulet. 2002. “Brachytherapy for T1-T2 Floor-of-the- Recommendations Regarding Screening for Oral Squamous Mouth Cancers: The Gustave-Roussy Institute Experience.” Cell Carcinomas.” Journal of the American Dental Association International Journal of Radiation Oncology Biology Physics 141 (5): 509–20. 52 (5): 1257–63. Richards, D. 2010. “Does Toluidine Blue Detect More Oral Mathew, B., R. Sankaranarayanan, K. B. Sunilkumar, Cancer?” Evidence-Based Dental Practice 11 (4): 104–05. B. Kuruvila, P. Pisani, and others. 1997. “Reproducibility Ries, L. A. G., D. Melbert, M. Krapcho, D. G. Stinchcomb, and Validity of Oral Visual Inspection by Trained Health N. Howlader, and others. 2008. Cancer Statistics Review, Workers in the Detection of Oral Precancer and Cancer.” 1975–2005. Bethesda, MD: National Cancer Institute. British Journal of Cancer 76 (3): 390–94. http://seer.cancer.gov/csr/1975-2005. Mathew, B., R. Sankaranarayanan, R. Wesley, and M. K. Nair. Saba, N. F., M. Goodman, K. Ward, C. Flowers, S. Ramalingam, 1995. “Evaluation of Mouth Self-Examination in the and others. 2011. “Gender and Ethnic Disparities in Control of Oral Cancer.” British Journal of Cancer 71 (2): Incidence and Survival of Squamous Cell Carcinoma of the 397–99. Oral Tongue, Base of Tongue, and Tonsils: A Surveillance, Mehta, F. S., P. C. Gupta, R. B. Bhonsle, P. R. Murti, D. K. Daftary, Epidemiology and End Results Program-Based Analysis.” and J. J. Pindborg. 1986. “Detection of Oral Cancer Using Oncology 81 (1): 12–20. Basic Health Workers in an Area of High Oral Cancer Sankaranarayanan, R., K. Ramadas, S. Thara, R. Muwonge, Incidence in India.” Cancer Detection and Prevention G. Thomas, and others. 2013. “Long Term Effect of Visual 9 (3–4): 219–25. Screening on Oral Cancer Incidence and Mortality in a Napier, S. S., and P. M. Speight. 2008. “Natural History of Randomized Trial in Kerala, India.” Oral Oncology 49 (4): Potentially Malignant Oral Lesions and Conditions: An 314–21. 98 Cancer Sankaranarayanan, R., K. Ramadas, G. Thomas, R. Muwonge, Development of Cancer in Patients with Oral Lichen S. Thara, and others. 2005. “Effect of Screening on Oral Planus.” Community Dentistry and Oral Epidemiology Cancer Mortality in Kerala, India: A Cluster-Randomised 30 (5): 342–51. Controlled Trial.” The Lancet 365 (9475): 1927–33. Vijayakumar, M., R. Burrah, K. S. Sabitha, H. Nadimul, and Sankaranarayanan, R., and R. Swaminathan. 2011. Cancer B. C. Rajani. 2011. “To Operate or Not to Operate: N0 Neck Survival in Africa, Asia, the Caribbean and Central America. in Early Cancer of the Tongue? A Prospective Study.” Indian Scientific Publications No. 162. Lyon, France: International Journal of Surgical Oncology 2 (3): 172–75. Agency for Research on Cancer. Walsh, T, J. L. Liu, P. Brocklehurst, A. M. Glenny, M. Lingen, and Sankaranarayanan, R., R. Swaminathan, H. Brenner, K. Chen, others. 2013. “Clinical Assessment to Screen for Oral Cavity K. S. Chia, and others. 2010. “Cancer Survival in Africa, Cancer and Potentially Malignant Disorders in Apparently Asia, and Central America: A Population-Based Study.” The Healthy Adults.” Cochrane Database of Systematic Reviews Lancet Oncology 11 (2): 165–73. 11: CD010173. Sant, M., C. Allemani, M. Santaquilani, A. Knijn, F. Marchesi, Warnakulasuriya, K. A., A. N. Ekanayake, S. Sivayoham, and others. 2009. “EUROCARE-4. Survival of Cancer J. Stjernsward, J. J. Pindborg, and others. 1984. “Utilization Patients Diagnosed in 1995–1999.Results and Commentary.” of Primary Health Care Workers for Early Detection of Oral European Journal of Cancer 45 (6): 931–91. Cancer and Precancer Cases in Sri Lanka.” Bulletin of the Satyanarayana, L., and S. Asthana. 2008. “Life Time Risk for World Health Organization 62 (2): 243–50. Development of Ten Major Cancers in India and Its Trends Warnakulasuriya, S., N. W. Johnson, and I. Van Der Waal. 2007. over the Years 1982 to 2000.” Indian Journal of Medical “Nomenclature and Classification of Potentially Malignant Sciences 62 (2): 35–44. Disorders of the Oral Mucosa.” Journal of Oral Pathology & Scott, S. E., K. Rizvi, E. A. Grunfeld, and M. Mcgurk. 2010. “Pilot Medicine 36 (10): 575–80. Study to Estimate the Accuracy of Mouth Self-Examination Warnakulasuriya, K. A., and B. G. Nanayakkara. 1991. in an At-Risk Group.” Head & Neck 32 (10): 1393–401. “Reproducibility of an Oral Cancer and Precancer Detection Sinha, D. N., K. M. Palipudi, I. Rolle, S. Asma, and Program Using a Primary Health Care Model in Sri Lanka.” S. Rinchen. 2011. “Tobacco Use among Youth and Adults Cancer Detection and Prevention 15 (5): 331–34. in Member Countries of South-East Asia Region: Review Wendt, C. D., L. J. Peters, L. Delclos, K. K. Ang, W. H. Morrison, of Findings from Surveys under the Global Tobacco and others. 1990. “Primary Radiotherapy in the Treatment Surveillance System.” Indian Journal of Public Health 55 (3): of Stage I and II Oral Tongue Cancers: Importance of the 169–76. Proportion of Therapy Delivered with Interstitial Therapy.” Sobin, L. H., and C. Wittekind. 2002. UICC TNM Classification International Journal of Radiation Oncology Biology Physics of Malignant Tumours. Geneva: Union for International 18 (6): 1287–92. Cancer Control. Woolgar, J. A. 2006. “Histopathological Prognosticators in Somatunga, L. C., D. N. Sinha, P. Sumanasekera, K. Galapatti, Oral and Oropharyngeal Squamous Cell Carcinoma.” Oral S. Rinchen, and others. 2012. “Smokeless Tobacco Use in Oncology 42 (3): 229–39. Sri Lanka.” Indian Journal of Cancer 49 (4): 357–63. World Cancer Research Fund and AICR (American Institute Speight, P. M., S. Palmer, D. R. Moles, M. C. Downer, for Cancer Research). 2007. Food, Nutrition, Physical D. H. Smith, and others. 2006. “The Cost-Effectiveness Activity, and the Prevention of Cancer: A Global Perspective. of Screening for Oral Cancer in Primary Care.” Health Washington, DC: AICR. Technology Assessment 10 (14):1–144, Iii-Iv. Wrangle, J. M., and F. R. Khuri. 2007. “Chemoprevention of Su, W. W., A. M. Yen, S. Y. Chiu, and T. H. Chen. 2010. Squamous Cell Carcinoma of the Head and Neck.” Current “A Community-Based RCT for Oral Cancer Screening with Opinion in Oncology 19 (3): 180–87. Toluidine Blue.” Journal of Dental Research 89 (9): 933–37. Yako-Suketomo, H., and T. Matsuda. 2010. “Comparison Subramanian, S., R. Sankaranarayanan, B. Bapat, T. Somanathan, of Time Trends in Lip, Oral Cavity and Pharynx Cancer G. Thomas, and others. 2009. “Cost-Effectiveness of Oral Mortality (1990–2006) between Countries Based on the Cancer Screening: Results from a Cluster Randomized WHO Mortality Database.” Japanese Journal of Clinical Controlled Trial in India.” Bulletin of the World Health Oncology 40 (11): 1118–19. Organization 87 (3): 200–06. Zwetyenga, N., C. Majoufre-Lefebvre, F. Siberchicot, Tseng, C. H. 2013. “Oral Cancer in Taiwan: Is Diabetes a Risk H. Demeaux, and J. Pinsolle. 2003. “[Squamous-cell Factor?” Clinical Oral Investigations 17 (5): 1357–64. carcinoma of the tongue: treatment results and prognosis].” Van Der Meij, E. H., P. D. Bezemer, and I. Van Der Waal. Revue de stomatologie et de chirurgie maxillo-faciale 104 (1): 2002. “Cost-Effectiveness of Screening for the Possible 10–17. Oral Cancer: Prevention, Early Detection, and Treatment 99 Chapter 6 Colorectal Cancer Linda Rabeneck, Susan Horton, Ann G. Zauber, and Craig Earle INTRODUCTION of countries to develop interventions increases with income, suggesting a progression in policy options as Adenocarcinoma of the colon and rectum (colorectal country income increases. cancer, CRC) is the third most common cancer, the The focus of this chapter is on those who are at aver- fourth most common cause of cancer death, and the age risk for CRC. In our discussion of policy options, second most common cancer in terms of the number of we use a slightly different typology than income for individuals living with cancer five years after diagnosis resource availability, following chapter 3 in this volume worldwide. An estimated 1,361,000 people are diagnosed (Anderson and others 2015). The resources available with CRC annually; approximately 694,000 people die at a health facility can be described as basic, limited, from CRC annually; and 3,544,000 individuals are living enhanced, and maximal. The basic level corresponds with CRC (Ferlay and others 2013). approximately to the situation in low-income countries Randomized controlled trials (RCT) have shown (LICs), the limited level to the situation in rural areas that screening is associated with a reduction in CRC of lower-middle-income countries and upper-middle- mortality; in several high-income countries (HICs), income countries, the enhanced level to the situation in organized, population-based screening programs have urban areas of lower-middle-income and upper-middle- been introduced, starting in 2006. Some screening tests income countries, and the maximal level to the situation detect cancer at an early stage when treatment is less in HICs. We provide suggestions for appropriate screen- arduous and more often results in cure. Other screening ing and treatment strategies that correspond to these tests have the ability to detect adenomas as well as cancer. resource levels for policy makers to consider. Screening provides the opportunity to identify and remove adenomas and thereby to prevent the develop- ment of the disease (Lieberman and others 2012). BURDEN AND EPIDEMIOLOGY1 In general, the burden of disease, as measured by incidence and mortality rates, tracks the World CRC is the third most common cancer in men (746,000 Bank grouping of countries into low-, lower-middle, cases, 10.0 percent of all cancers in men worldwide); it upper-middle, and high-income: the lowest-income is the second most common cancer in women (614,000 countries have the lowest burden of disease. The ability cases, 9.2 percent of all cancers in women worldwide). to intervene to introduce screening and offer access to CRC incidence rates vary approximately tenfold in high-quality treatment is a function of resource avail- both genders worldwide and are higher in men than ability, which is associated with income level. The ability in women; the overall sex ratio of the age-standardized Corresponding author: Linda Rabeneck, MD, MPH, FRCPC, Cancer Care Ontario and University of Toronto, linda.rabeneck@cancercare.on.ca. 101 rates is 1.44:1 (Ferlay and others 2013). CRC incidence with the introduction of screening. CRC incidence rates and mortality rates vary widely across regions, with have stabilized or are declining in many HICs. Initially, mortality highest among men in HICs (map 6.1). the stabilization or decline may have been caused by The distribution of incidence in men and women and declines in some risks, such as smoking; more recently, mortality in women may be seen in online annex 6A the change is likely to be caused by increased screening. (maps 6A.1, 6A.2, and 6A.3). Approximately 55 percent Data from the Surveillance, Epidemiology, and End of persons diagnosed with CRC are in HICs. Australia, Results (SEER) Program for the state of Connecticut, Canada, New Zealand, the United States, and Western United States, from 1940 to 2009, is one of the lon- Europe have the highest estimated incidence rates, gest consistent time-series available. Incidence rates while incidence rates are intermediate in Latin America increased rapidly until the 1980s and then declined. and the Caribbean. The lowest rates are in Sub-Saharan (See online annex 6A, figure 6A.1). The peak in the Africa, with the exception of southern Africa, and 1980s represents the introduction of screening (primar- South Asia. ily with fecal occult blood tests) and is consistent with CRC mortality rates are lower in women than in an initial increase in incidence with screening because men. However, compared with incidence rates, variabil- of detection of early-stage and preclinical disease. The ity in mortality rates worldwide is less (sixfold in men, decline post-1985 likely represents the impact of screen- fourfold in women). Estimated CRC mortality rates are ing, as well as a decrease in risk factors such as smoking. highest in Eastern and Central Europe (20.3 per 100,000 The inverted U-shaped curve is more pronounced for for men and 11.7 per 100,000 for women); Western men than for women (figure 6A.1). CRC incidence rates Africa has the lowest estimated mortality rates (3.5 and are now declining to the lowest level since the 1940s 3.0, respectively). (Edwards and others 2010). Trends for the United States as a whole are similar to those for Connecticut, but data are not available as far Temporal Trends in Incidence and Mortality back as 1940 in a continuous series. A declining trend Temporal trends in CRC incidence and mortality in a in CRC mortality is also seen in other HICs, including population reflect changes in the prevalence of risk fac- Australia, Denmark, and Japan. Incidence has not yet tors in the population, coupled with changes associated begun to decline in these three countries, likely because Map 6.1 Global Colorectal Cancer Mortality in Men, 2012 Source: Based on data from Ferlay and others 2013. Note: ASR = Age-Standardized Rate. 102 Cancer a “bulge” in reported cases occurs as CRC is detected at rates are increasing. Data from LICs are sparse, because of increasingly earlier stages. One would eventually expect the limited availability and coverage of cancer registries. to see incidence rates decline as in the United States, once a steady state is reached in screening. The increase in CRC incidence and mortality in Incidence and Mortality by Income Group middle-income countries (MICs), such as Brazil, China, We classified the age-adjusted incidence and mortality the Philippines, and Thailand (figures 6.1 and 6.2), is rates for CRC by World Bank groupings of countries into occurring prior to the onset of organized screening. Even LICs, lower-middle-income countries, upper-middle- in lower-middle-income countries such as India, incidence income countries, and HICs (figure 6.3). The CRC Figure 6.1 Trends in Age-Standardized Incidence and Mortality Rates of Colorectal Cancer in Men, Selected Countries, 1980–2010 a. Low- and lower-middle-income b. High- and upper-middle-income 30 30 20 20 Age-standardized rate (world) per 100,000, 0–69 years Age-standardized rate (world) per 100,000, 0–69 years 15 15 10 10 7 7 5 5 3 3 2 2 1980 1990 2000 2010 1980 1990 2000 2010 Calendar year Calendar year Incidence Mortality Philippines Thailand Uganda United States: Blacka United States: Black Indiaa Philippines United States: Whitea United States: White Chinaa China Spaina Spain Brazila Brazil Source: CI5 Plus (http://ci5.iarc.fr/CI5plus/Default.aspx) and WHO Mortality Database (http://www.who.int/healthinfo/statistics/mortality_rawdata/en/index.html). Note: Mortality data are not available for all economies shown. Incidence estimated from selected population-based cancer registries of consistently high quality (included in successive volumes of Cancer Incidence in Five Continents). Incidence data for the economies in the graphs are for Brazil (Goiania), China (Hong Kong SAR and Shanghai), India (Chennai and Mumbai), Philippines (Manila), Spain (Granada, Murcia, Navarra, and Tarragona), Thailand (Chiang Mai), Uganda (Kampala), United States (SEER). a. Denotes rates based on an aggregate of one or more regional registries, as indicated. Colorectal Cancer 103 Figure 6.2 Trends in Age-Standardized Incidence and Mortality of Colorectal Cancer Rates in Women, Selected Countries, 1980–2010 a. Low- and lower-middle-income b. High- and upper-middle-income 25 25 20 20 15 15 Age-standardized rate (world) per 100,000, 0–69 years Age-standardized rate (world) per 100,000, 0–69 years 10 10 7 7 5 5 3 3 2 2 1.5 1.5 1980 1990 2000 2010 1980 1990 2000 2010 Calendar year Calendar year Incidence Mortality Philippinesa Thailanda Indiaa United States: Blacka United States: Black Ugandaa Philippines United States: Whitea United States: White Chinaa China Spaina Spain Brazila Brazil Source: CI5 Plus (http://ci5.iarc.fr/CI5plus/Default.aspx) and WHO Mortality Database (http://www.who.int/healthinfo/statistics/mortality_rawdata/en/index.html). Note: Mortality data are not available for all economies shown. Incidence estimated from selected population-based cancer registries of consistently high quality (included in successive volumes of Cancer Incidence in Five Continents). Incidence data for the economies in the graphs are for Brazil (Goiania), China (Hong Kong SAR and Shanghai), India (Chennai and Mumbai), Philippines (Manila), Spain (Granada, Murcia, Navarra, and Tarragona), Thailand (Chiang Mai), Uganda (Kampala), United States (SEER). a. Denotes rates based on an aggregate of one or more regional registries, as indicated. incidence and mortality rates increase with increasing countries had the lowest CRC incidence and mortality country income, as indicated by the World Bank income rates, approximately 67 percent of men and 68 percent groupings. of women who develop CRC die from the disease. This We derived the mortality-to-incidence ratio as an is in strong contrast to the experience in HICs, where approximation of the CRC-specific mortality rate using the incidence and mortality rates are much higher, but the data from figure 6.3. The mortality-to-incidence only approximately 31 percent of men and 30 percent ratio roughly represents the percentage of people with of women with CRC die of this cancer. The corre- CRC who die of this disease. Although the low-income sponding figures for lower-middle-income countries 104 Cancer are 60 percent of men and 58 percent of women; for Figure 6.3 Age-Standardized Incidence and Mortality Rates of Colorectal upper-middle-income countries, they are 49 percent of Cancer by World Bank Income Classification, 2012 men and 47 percent of women. These results indicate that Men Women better survival is associated with higher country income. For China, the mortality-to-incidence ratio is 37 percent Low-income for men and 36 percent for women who develop CRC, Lower-middle-income whereas for India, the mortality-to-incidence ratio is 70 (except India) percent for men and 68 percent for women, indicating a Upper-middle-income very high case fatality rate from CRC in India. (except China) In summary, the current burden of disease parallels High-income the World Bank income groupings, with the lowest- China income countries having the lowest burden of disease and poorest disease-specific survival. India 30 20 10 0 10 20 30 Risk Factors Age-standardized rate (world) per 100,000, 0–69 years Age, sex, and family history are independent risk factors Incidence Mortality for CRC. The incidence of CRC increases with age, with Source: Ferlay and others 2013. approximately 7 percent of cases occurring in those younger than 50 years. The risk is somewhat higher among men than women. About 75 percent of new cases of CRC occur in those with no known predisposing fac- INTERVENTIONS tors. Those at increased risk because of a family history of CRC but without an identifiable genetic syndrome Screening account for 15–20 percent of cases. HNPCC (Lynch CRC Screening Tests syndrome) accounts for about 5 percent of cases, and Guaiac Fecal Occult Blood Test (gFOBT) gFOBT is a familial adenomatous polyposis (FAP) accounts for stool test that indirectly detects blood in the stool that about 1 percent (Winawer and others 1997). HNPCC may be caused by bleeding from CRC. A positive test is and FAP are genetic polyposis syndromes. Those with not specific for the presence of human blood; however, it a family history of CRC in a parent, sibling, or child are may reflect blood from ingested animal meats, for exam- at a twofold increased risk of the disease (Butterworth, ple. gFOBT is supported by evidence from RCTs with Higgins, and Pharoah 2006; Johnson and others 2013). long-term follow-up and CRC mortality as the outcome. In terms of modifiable risk factors, epidemiological RCTs of periodic (annual or biennial) gFOBT show a evidence supports roles for diet, lifestyle, and medica- reduction in mortality from CRC of 13 to 33 percent, tions (Chan and Giovannucci 2010). In general, diets with up to 50 percent compliance with periodic gFOBT high in red meat are associated with an increased risk. (Hardcastle and others 1996; Kronborg and others 1996; For red meat, a recent meta-analysis reported a relative Mandel and others 1993; Towler and others 1998). risk of 1.13 for consumption of five versus no serv- Individuals with a positive gFOBT must be followed up ings per week (Johnson and others 2013). In addition, by colonoscopy. smoking, obesity, and a sedentary lifestyle are asso- ciated with an increased risk. For smoking, a relative Fecal Immunochemical Test (FIT) FIT is a stool test risk of 1.26 for a 30-pack per year smoker versus a that uses an antibody against human globin, the pro- non-smoker, and for obesity, a relative risk of 1.10 for tein part of hemoglobin. A positive FIT is specific for body mass index greater than 30 versus 22 kg/m2 were human blood. There are no large-scale RCTs that have recently reported (Johnson and others 2013). Calcium evaluated FIT with long-term follow-up and CRC mor- supplements may be associated with a modest reduction tality as the outcome, although two RCTs are underway. in risk. Aspirin and nonsteroidal anti-inflammatory FIT is supported by RCTs of FIT versus gFOBT (van drugs and postmenopausal hormone therapy among Rossum and others 2008) and FIT versus colonoscopy women are inversely associated with CRC risk, although (Quintero and others 2012). These are cross-sectional the magnitudes of these effects are uncertain. The rise RCTs with cancer detection and advanced adenoma as in CRC incidence rates in low- and middle-income the outcomes. The RCT of FIT compared with gFOBT countries (LMICs) is largely attributed to the adoption shows that the use of FIT is associated with higher of Western diets and sedentary lifestyles. adherence/compliance rates than the use of gFOBT, and Colorectal Cancer 105 also that FIT is superior to gFOBT in detection rates and colonoscopy is being compared with usual care. In the positive predictive values for adenomas and cancer (van COLONPREV trial in Spain and the CONFIRM trial Rossum and others 2008). The RCT of FIT compared by the U.S. Department of Veterans Affairs, colonos- with colonoscopy showed higher participation in the copy is being compared with FIT. FIT group than in the colonoscopy group, and that the numbers of subjects in whom CRC was detected were Fecal DNA A cross-sectional study comparing the similar in the two groups, but a greater number of sub- performance of a stool DNA prototype versus gFOBT jects with adenomas were identified in the colonoscopy versus colonoscopy reported that the stool DNA test group (Quintero and others 2012). detected a greater proportion of CRCs and CRCs plus adenomas with high-grade dysplasia than the gFOBT, Flexible Sigmoidoscopy (FS) FS is an endoscopic pro- 51.6 percent vs. 12.9 percent, and 40.8 percent vs. cedure in which a flexible fiberoptic instrument is used 14.1 percent, respectively (Imperiale and others to examine the rectum and lower (distal) colon, unlike 2004). However, the majority of neoplastic lesions colonoscopy, which examines the rectum and total was not detected by either test. Since then, stool (upper and lower) colon. Cancers and precancerous DNA technology has continued to evolve (Ahlquist lesions, such as adenomas observed in this area, can be and others 2012). A large-scale, cross-sectional study removed or biopsied. Large-scale RCTs of FS, coupled (DeeP-C) of a next-generation stool DNA test has with colonoscopy for those who test positive, have shown recently been published (Imperiale and others 2014) reductions in CRC incidence (Atkin and others 2010; and indicates 92.3 percent sensitivity for the detection Schoen and others 2012; Segnan and others 2011) and of CRC and 42.4 percent sensitivity for the detection CRC mortality (Atkin and others 2010; Schoen and of advanced adenomas. others 2012) over a 10-year period. A meta-analysis of the results from the published RCTs of FS screening reported, Computed Tomographic Colonography (CTC) CTC in an intention-to-treat analysis, an 18 percent reduction is a computerized tomography examination of the in relative risk of CRC incidence and a 28 percent reduc- abdomen and pelvis in which imaging information, tion in CRC mortality (Elmunzer and others 2012). when processed with special imaging software, pro- vides images of the colon and rectum. Images are also Colonoscopy Colonoscopy is an examination of the produced of structures outside the colon (extracolonic entire colon and rectum with a flexible fiberop- findings). The technique requires bowel preparation tic endoscope. Colonoscopy detects asymptomatic and colonic insufflation, but not conscious sedation, cancers, and precancerous lesions can be removed. as is generally required for colonoscopy. CTC does Evidence from the National Polyp Study, analyzed as not permit biopsy or polyp removal; colonoscopy an observational cohort study, indicates that colonos- is required. The two largest reports evaluating CTC copy with polypectomy is associated with a reduc- in asymptomatic persons are cross-sectional stud- tion in CRC incidence and mortality (Winawer and ies that compared CTC and colonoscopy for the others 1993; Zauber and others 2012). No published detection of adenomas (Johnson and others 2008; evidence from RCTs has evaluated screening colonos- Pickhardt and others 2003). Taken together, these copy with long-term follow-up and CRC mortal- studies found that CTC was comparable to colonos- ity as the outcome. However, indirect evidence to copy for detecting adenomas ≥ 10 mm (pooled sen- support screening colonoscopy comes from RCTs sitivity was 92 percent), but fell short in detecting of gFOBT, in which persons with a positive gFOBT smaller adenomas (6–9 mm). underwent colonoscopy. In these trials, colonoscopy with polypectomy was responsible for the mortality Role of Colonoscopy in Diagnosis and Surveillance reduction associated with gFOBT screening. Further Having adequate colonoscopy resources is a key aspect indirect evidence comes from RCTs of FS screening, in implementing CRC screening, because of the role of which demonstrate a reduction in CRC mortality. colonoscopy in diagnosis and surveillance. When a less By extrapolation, it would be expected that since invasive screening test—such as gFOBT, FIT, or FS—is colonoscopy evaluates the entire colon, screening used, colonoscopy is required to investigate those with colonoscopy would be associated with a reduction in a positive (abnormal) screen, and it is the final common CRC mortality that might exceed that observed for pathway to establish a diagnosis. Colonoscopy is also screening FS. Several large-scale RCTs of colonoscopy recommended for surveillance, depending on the find- are underway, with CRC mortality as the primary ings at the initial colonoscopy, which are used to stratify outcome. In the NordICC trial in Europe, screening risk for subsequent colorectal neoplasia. 106 Cancer Guidelines for Colonoscopic Surveillance In 2012, What are the possible reasons for postcolonoscopy the U.S. Multi-Society Task Force on CRC/American CRCs? Cancer Society (USMSTF) published updated surveil- lance guidelines that take into account the serrated • First, the lesion may not have been seen because the neoplasia pathway (Lieberman and others 2012). cecum was not reached during the procedure (reach- Based on the findings at the baseline colonoscopy, ing the cecum is more challenging technically than the recommended surveillance intervals are as fol- reaching the distal colon); the bowel preparation was lows: 10 years for those with no polyps or small not adequate and the mucosa was not fully visualized (< 10 mm) hyperplastic polyps in the rectum or (the upper or proximal colon is more difficult to sigmoid; 5–10 years for those with 1 or 2 small clean); or the technique was inadequate, that is, the (< 10 mm) tubular adenomas; three years for those lesion was simply not seen. with three to 10 tubular adenomas, one or more • Second, a prior polypectomy may have been incom- tubular adenomas ≥ 10 mm, one or more villous ade- plete. Polypectomy is the key to reducing CRC nomas, or an adenoma with high-grade dysplasia; and incidence following colonoscopy. less than three years for those with more than 10 ade- • Third, a rapidly progressing CRC may not have been nomas. When serrated lesions are detected at the present at the initial colonoscopy, which may have baseline colonoscopy, the recommended surveillance been truly negative. intervals are as follows: five years for those with sessile serrated polyps (SSPs) < 10 mm with no dysplasia, What all of this means is that colonoscopy fails three years for SSPs ≥ 10 mm or SSPs with dysplasia or to detect a small but clinically important percentage a traditional serrated adenoma, and one year for those of lesions, and this lack of effectiveness is more pro- with serrated polyposis syndrome. nounced in the right colon, making meticulous tech- nique paramount. Colonoscopy Effectiveness in Usual Practice Large-scale, population-based studies have shown that CRC Screening Guidelines lesions can be missed at colonoscopy, including ade- International Agency for Research on Cancer In nomas and cancers (Baxter and others 2009; Bressler 2010, the International Agency for Research on Cancer and others 2007). It is often not possible to be cer- (IARC), an agency of the World Health Organization tain that a CRC that was not detected at the time of (WHO), published a landmark document, European colonoscopy, but subsequently diagnosed, was a missed Guidelines for Quality Assurance in Colorectal Cancer cancer. The alternative explanation is that the “new” Screening and Diagnosis (Segnan, Patnick, and von cancer was not present at the time of the colonoscopy, Karsa 2010). The recommendations are based on a but arose and grew rapidly following the procedure. comprehensive and systematic review of the scientific These new or missed cancers have been referred to as evidence and are intended for organized screening postcolonoscopy CRCs (Rabeneck and Paszat 2010) or programs (see below). The guidelines recommended interval cancers. annual or biennial screening with gFOBT, FIT screen- Evidence from population-based, case-control stud- ing at an interval not to exceed three years and at a ies shows that colonoscopy is associated with a reduc- minimum to include screening for those ages 60–64 tion in overall CRC mortality; however, the magnitude years, and FS screening at an interval not less than of the effect is less in the proximal colon (Baxter and 10 years, with the best age range for screening between others 2009; Brenner and others 2011). This means ages 55 and 64 years but not exceeding age 74 years. that in usual practice, colonoscopy is less effective in Colonoscopy was not recommended for CRC screening the proximal colon. This finding is attributed to sub- in the European Union. optimal colonoscopy quality or differences in tumor biology between those cancers that arise in the proximal U.S. Preventive Services Task Force (USPSTF) USPSTF compared with the distal colon. These findings have recommends routine screening for average risk persons given rise to renewed emphasis on the critical impor- ages 50–75 years, no routine screening for persons ages tance of colonoscopy quality and increased attention to 76–85 years, and no screening for persons older than age CRC carcinogenesis. In particular, the recognition that 85 years (USPSTF 2008). USPSTF recommends annual lesions arising in the serrated neoplasia pathway have screening with high-sensitivity gFOBT, or FS every five been underdetected at colonoscopy may explain, in part, years with high-sensitivity gFOBT every three years, the lesser effectiveness of colonoscopy in the proximal or colonoscopy every 10 years. USPSTF did not assess (right) colon. barium enema because of lack of evidence and declining Colorectal Cancer 107 use, and it concluded that the evidence was insufficient Performance of Organized Screening Programs In to assess the benefits and harms of CTC and fecal DNA general, a period of at least 10 years is required for CRC screening. to plan, pilot, and implement an organized CRC screening program (von Karsa and others 2010). U.S. Multi-Society Task Force/American Cancer The European Guidelines define key performance Society (USMSTF) USMSTF defines two categories of indicators for CRC screening, including participation, screening test (Levin and others 2008). In the first cat- follow-up colonoscopy among those with a positive egory are tests that primarily detect cancers. Tests that gFOBT, retention rates, cancer detection rates, and are recommended in this category for screening persons CRC mortality (von Karsa and others 2010). High per- at average risk (age 50 years and older, no symptoms, formance in all of these measures is required in CRC no family history of the disease) are the following: screening programs. A few programs have published annual high-sensitivity gFOBT, annual FIT, or stool early results for participation. For example, Ontario DNA (interval uncertain). In the second category are launched its provincewide, organized screening pro- tests that detect cancers and adenomas. Recommended gram, ColonCancerCheck, in 2008. The program is tests in this category for screening persons at average based on gFOBT for those at average risk and colonos- risk are: FS every five years, colonoscopy every 10 years, copy for those at increased risk, defined by a family double-contrast barium enema every five years, or CTC history of one or more first-degree relatives with the every five years. disease. The target population—men and women ages 50–74 years—is 3.4 million. Prior to launch of the pro- Organized Versus Opportunistic Screening gram, gFOBT participation in 2003−04 was 15 percent; Screening is not simply a test; it is a process. Chapter 12, this in 2010–11, gFOBT participation was 29.8 percent volume, discusses organized and opportunistic screening (Rabeneck and others 2014). in detail. Compared with opportunistic screening, orga- Some high-income Asian countries have begun to nized screening focuses much greater attention on the implement organized screening: Japan, the Republic quality of the screening process, including follow-up of of Korea, and Singapore. Currently, organized, participants (Miles and others 2004). Thus, a key advan- population-based screening programs do not exist tage of organized screening is that it provides greater in the majority of LMICs. As MICs develop orga- protection against the possible harms of screening. Poor nized cancer screening programs, CRC screening follow-up of those who test positive may occur, for is under consideration. There are or have been example, when those with a positive (abnormal) FIT fail pilot studies of CRC screening conducted in sev- to undergo colonoscopy, the recommended next step in eral upper-middle-income countries and economies the screening process (Miles and others 2004). using gFOBT, including Thailand (Lampang prov- ince: http://www.iarc.fr /en/staffdirectory/display- Organized CRC Screening Worldwide The staff.php?id=10114); Shanghai is embarking on a International Colorectal Cancer Screening Network pilot study, as is Argentina (Manzur 2013). These (ICRCSN) is a consortium of organized initiatives pilot studies are important to lay the groundwork for delivering screening to their populations. In 2008, national programs (Goss and others 2013). ICRCSN conducted a survey of full or pilot programs Other countries offer opportunistic screening, largely that fulfill at least four of the IARC criteria for an orga- restricted to the population covered by work-based nized screening program (Benson and others 2012). health insurance. As a high-income economy, Taiwan, At that time, 43 organized screening programs were China, offers free screening under the national health identified, of which 35 programs had been collecting insurance program (Ng and Wong 2013). Formal sector data for at least 12 months and were eligible for the sur- employees and/or government employees in much of vey. Of the 35 programs from 24 countries, 26 were full Latin America and the Caribbean are covered for cancer programs and nine were pilot programs. The majority screening. However, most MICs do not have organized of the programs were in Europe, with a few from North screening programs. America, South America, and the Western Pacific. The majority (28) used stool-based tests as their primary International Efforts to Advance CRC Screening: screening test: 16 used gFOBT, nine used the FIT, and IARC The vision of the Early Detection and Prevention three used both tests. Wide variations were observed in (EDP) Section of the IARC is to serve as the major global the ability of the jurisdictions to report on performance resource for high-quality scientific and evidence-based indicators, such as the participation rate, the gFOBT or information on cancer prevention and early detection FIT positivity rate, and the cancer detection rate. interventions. The EDP Section evaluates and reports 108 Cancer on interventions for early detection and prevention. to CRC screening and sharing of best practices across The findings guide the development of public health the world. policy, with a particular focus on cancer control in The International Digestive Cancer Alliance pro- LMICs. The EDP Section’s work catalyzes the implemen- motes the screening, early detection, and primary tation of CRC prevention and early detection programs prevention of digestive cancers worldwide, includ- that follow the principles of organized screening as ing development of practice guidelines (http://www outlined, to the extent feasible. .worldgastroenterology.org/assets/downloads/en/pdf Experience from the European Union shows that /guidelines/06_colorectal_cancer_screening.pdf). The a minimum period of 10 years is required to estab- International Digestive Cancer Alliance recommended lish a population-based cancer screening program, staging the approach to screening, with respect to the with any impact taking even longer (Lee and others choice of screening test, to the resources available in a 2013). Examples of early detection and prevention work given country (Winawer 2007). include the European Guidelines for Quality Assurance in Colorectal Cancer Screening and Diagnosis (Segnan, Patnick, and von Karsa 2010). In addition, a network Diagnosis of reference and training centers (European School In HICs, persons with CRC can present in several ways: of Screening Management) has been created that is developing and piloting training courses for plan- • First, the cancer can be detected as a result of ning, implementation, quality assurance, and evaluation screening. When gFOBT, FIT, or FS is used as the of population-based cancer screening programs. The initial screening test, colonoscopy is undertaken as European School of Screening Management is intended a diagnostic test to evaluate those with an abnormal to serve as a platform to connect and facilitate collabora- screening test. During the colonoscopy, polyps are tion among relevant personnel from HICs and LMICs. removed and masses or other suspicious lesions are Further, the EDP Section provides scientific and techni- either removed or biopsied to establish a pathologi- cal support to upper-middle-income countries, such as cal diagnosis. Albania and Belarus, to assist them in moving forward • Second, the cancer can be detected when an individ- with population-based cancer screening programs. ual undergoes colonoscopy to evaluate large bowel symptoms, such as rectal bleeding, anemia, or a Other International Organizations and Networks change in bowel habits. Promoting Screening The International Cancer • Third, some individuals may present as an emer- Screening Network is a consortium of countries that gency, such as a large bowel obstruction, in which have active population-based cancer screening programs case the cancer may be diagnosed at surgery without and active efforts to evaluate and improve the processes prior diagnostic evaluation. and outcomes from cancer screening in practice. These programs can be national or subnational in scope, and established or pilot-based. Administered by the Applied Staging Research Program of the U.S. National Cancer Institute, When the cancer is diagnosed in nonemergency pre- the consortium includes 33 countries and holds biennial sentations, staging and complete visualization of the meetings; specific activities are moved forward through colon and rectum with colonoscopy are undertaken. working groups. Participation in the International Complete colonoscopy is also undertaken for the Cancer Screening Network is open to any country that purpose of detecting synchronous cancers (present in has initiated a population-based screening program. 3–5 percent of cases); if not done prior to definitive The World Endoscopy Organization is a federation treatment, it should be done within 6–12 months. of national digestive endoscopy societies. Its mission Colonoscopy will also detect synchronous premalig- includes the advancement of digestive endoscopy for the nant adenomas, which can be removed to reduce the diagnosis and treatment of gastrointestinal diseases in risk of subsequent cancer. Barium enema, a radiolog- underserved areas. The World Endoscopy Organization ical test, was used to diagnose CRC prior to the wide- has an active CRC Screening Committee that holds spread availability of colonoscopy in HICs and may annual meetings in the East Asia and Pacific region, still be relevant in LMICs. Europe, and the United States, in conjunction with the Clinical staging to determine the extent of dis- major regional scientific meetings of gastroenterology ease focuses on imaging the liver and lungs, the two societies. The meetings provide a forum to facilitate the most common sites of spread. In HICs, computerized presentation and discussion of new knowledge related tomography is used to detect these distant metastases Colorectal Cancer 109 (Leufkens and others 2011). Chest X-ray and abdominal • Total meso-rectal excision, the meticulous, sharp ultrasound are less expensive alternative tests that can be dissection of perirectal tissues with removal of the used in lower-income settings. primary tumor and lymph nodes all in one piece, has Pathologic stage I refers to CRCs that are confined been shown to decrease local relapse rates. to the surface of the bowel. Stage II means the cancer • To avoid a stoma, transanal excision of small, early- has invaded through the muscle layer of the bowel wall. stage distal tumors with good prognostic features Stage III cancers involve the local lymph nodes. These can be considered. (Good is defined here as T1N0, stages are usually determined by examining the tumor < 3 cm, < 30 percent circumference, not poorly dif- pathologically after surgery. Stage IV cancers have spread ferentiated, and no lymphovascular or perivascular (metastasized) into other organs. invasion.) Radiation Treatment The availability of radiation therapy is most relevant Surgery for Colon Cancer for cancers of the rectum, as local recurrence is much The cornerstone of treatment is surgical resection. For more common than in colon cancer, because of the early-stage cancers, surgery alone may cure the disease. inability to obtain wide margins and the lack of a serosal For colon cancer, the preferred procedure is a hemicolec- barrier. Radiation therapy has improved local control tomy (resection of either the right or the left colon) with for persons with stages II and III rectal cancer (Hoffe, wide (> 5 cm) margins of normal colon. This procedure Shridhar, and Biajioli 2010). Evidence suggests that can typically be performed by a general surgeon. Where compared with postoperative radiation, preoperative available, minimally invasive (laparoscopically-assisted) radiation is associated with improved surgical outcomes techniques have produced similar long-term results and disease-free survival (Sebag-Montefiore and others compared with an open procedure, but with shorter 2009). This decision depends on determining the stage hospital stays and increased speed of recovery. Achieving of cancer preoperatively, which requires diagnostic ser- these benefits requires an experienced surgeon and vices such as magnetic resonance imaging or specialized specialized instruments, however; even in HICs, the endorectal ultrasound capability. Where these are not cost-effectiveness of this approach has been questioned. available, postoperative delivery of radiation and che- A minimum of 12 lymph nodes in the surgical specimen motherapy still provides important benefits. In settings is required for adequate staging and is associated with with access to radiation but difficulty obtaining or better outcomes than a lesser dissection. In patients delivering chemotherapy, or where travel requirements presenting with stage IV colon cancer where cure is not preclude the 5.5 weeks of daily long-course radiotherapy, possible, if the primary (that is, the site of the original short-course radiotherapy may be a preferred option cancer) is not associated with symptoms and the met- (Içli and others 2010). astatic disease (that is, the sites where the disease has spread) is anticipated to be controlled with chemother- Chemotherapy apy, the primary tumor does not necessarily need to Evidence-based practice guidelines recommend six be resected. months of adjuvant chemotherapy following surgery for persons with stage III colon cancer (Benson and Surgery for Rectal Cancer others 2000) and stages II and III rectal cancer. FOLFOX Surgery for rectal cancer is much more complex. High- (FOLinic acid [leucovorin], Fluorouracil, OXaliplatin) volume, specialized surgeons and centers have been is the preferred regimen. If chemoradiotherapy is given associated with better outcomes: less likely to need an for rectal cancer, only four months of chemotherapy are ostomy bag, lower rates of local recurrence, better overall required. In addition to the cost of the drugs, however, survival. systemic chemotherapy always requires the ability to monitor blood counts for safety and may require venous • Mid-to-upper rectal tumors can be resected with a access devices. low anterior resection, which leaves the rectal sphinc- ter intact, thereby avoiding colostomy. Management of Metastatic Colorectal Cancer • Lower-lying tumors, that is, those within 2–3 cm of the Metastatic CRC is treated the same way, regardless of anal sphincter or levator muscles, require an abdom- whether it started in the colon or rectum. Although inal perineal resection and creation of a permanent metastatic disease is generally incurable, it is increasingly stoma requiring colostomy. Surgeon skill often deter- recognized that, where possible, in 10–20 percent of mines how low a low anterior resection can be done. patients, aggressive resection of liver and lung metastases 110 Cancer may lead to cure 20–30 percent of the time. Such surgery a fairly recent publication, Greenberg and others requires highly specialized training and centers, even in (2010), was used. This is a systematic review of the cost- HICs. Alternatives to surgical resection, such as radiofre- effectiveness literature for various cancers, primarily quency ablation and stereotactic body radiotherapy, can for industrialized countries, and has the advantage provide long-term control in these situations, but sur- that all the costs have been standardized to those of a gical resection is preferred where feasible. Perioperative common year. For additional analyses, comparing the chemotherapy with FOLFOX has been shown to improve cost-effectiveness of CRC screening and treatment with disease-free survival in this setting. For the majority of screening and treatment for other cancers in LMICs, see patients with metastatic CRC, however, treatment is pal- chapter 16 in this volume (Horton and Gavreau 2015). liative, with an expected median survival with surgery alone of 6–12 months. Screening International Partnerships for CRC Care Cost-effectiveness studies for CRC screening in the As with CRC screening, international partnership United States, several European countries, and high- arrangements can support diagnosis and treatment in income Asian countries and economies (Hong Kong LMICs. The American Society of Clinical Oncology SAR, China; Korea; Singapore; and Taiwan, China) gen- (http://www.asco.org) established an International erally conclude that screening is cost-effective compared Affairs Committee in 2007 with a goal of reducing dispar- with no screening. The cost-effectiveness of several ities in cancer care and maximizing chances of survival screening options can be considered. Guidelines gen- through the global exchange of oncologic knowledge. The erally recommend that screening should begin at age National Comprehensive Cancer Network (http://www 50 years (except for those with strong family history of .nccn.org) provides translations of many of its guide- CRC), but it can stop at different ages (such as 70, 75, lines into other languages, such as Chinese, Japanese, 80, or 85). Tests can be undertaken in combination, such and Spanish, and has published local adaptations of as FS combined with a sensitive gFOBT. The use of an guidelines, for example, for countries in the Middle East efficiency frontier for each individual country can help to and North Africa. One National Comprehensive Cancer identify the most appropriate screening strategy given the Network institution, the MD Anderson Cancer Center budget constraints (see figure 6.4 for the United States). at the University of Texas, lists 28 sister institutions, as well as affiliates in at least 18 countries, 10 of these Studies in the United States in LMICs (http://www.mdanderson.org/education-and Cost-effectiveness studies in the United States generally -research/education-and-training/schools-and-programs have used one of a few large cancer microsimulation /global-academic-programs/sister-institutions/index models. Comparative studies (for example, Pignone .html). Similar partnerships exist with other major and others 2002 and Pignone, Russell, and Wagner cancer centers. 2005) suggest that results are sensitive to the param- eters used, particularly cost. The models also entail different assumptions about disease progression that COST-EFFECTIVENESS OF CRC SCREENING also affect relative test performance (Pignone, Russell, and Wagner 2005). Since the review by Pignone and AND TREATMENT others (2002), three microsimulation models for CRC A systematic search of the literature on CRC screening have become part of the U.S. National Cancer Institute and treatment was undertaken using PubMed from 2004 consortium for Cancer Intervention and Surveillance to 2013 to identify relevant articles for LMICs, as well Modeling Network (CISNET), a consortium of National as selected high-income Asian countries and economies Cancer Institute-sponsored investigators that use statis- that can help serve as regional models, principally Hong tical modeling to improve the understanding of cancer Kong SAR, China; the Republic of Korea; Singapore; control interventions in prevention, screening, and treat- and Taiwan, China. The parameters were medical sub- ment, and their effects on population trends in incidence ject heading terms (colorectal neoplasms, colonic neo- and mortality. These models can be used to guide public plasms, rectal neoplasms, CRC, colon cancer, or rectal health research and priorities. cancer); colonoscopy or sigmoidoscopy; and multiple Figure 6.4 shows the results for cost-effectiveness terms related to economic evaluation, cost, cost analysis, analysis from one CISNET model, the Microsimulation and cost-effectiveness. Screening Analysis model from Erasmus University This search was supplemented by a nonsystematic Medical Center, for eight CRC screening strategies search using the Internet for certain LMICs. For HICs, beginning at age 50 years. Figure 6.4 is a modification of Colorectal Cancer 111 Figure 6.4 Discounted Costs and Discounted Life Years Gained for Eight to be on the efficiency frontier. In this example, the Colorectal Cancer Screening Strategies and the Efficient Frontier lowest-cost option of those options on the frontier is Costs without screening the less sensitive gFOBT (Hemoccult II). However, if 120 the budget available for testing increases, then a more sensitive gFOBT (Hemoccult Sensa) or FIT is the next best choice for cost relative to life years gained with Discounted life years gained per 1,000, 50 year olds 100 screening. The next choice would be FS with some type of fecal occult blood test (FOBT). Colonoscopy lies just below the efficient frontier, with the rankings for life 80 years gained relative to costs similar to those of FS with gFOBT or FIT. The rankings using two other models (the SimCRC model from University of Minnesota 60 and Massachusetts General Hospital, and the CRC- SPIN model from Group Health Research Institute; see http://cisnet.cancer.gov/profiles/) were broadly similar, 40 although the absolute cost per life year gained differed between models. All of the strategies had average costs per life year gained (compared with no screening) well 20 within the threshold considered cost-effective in the United States (below US$50,000 per life year gained). Taken together, the results from the U.S. analyses 0 $2,000 $2,500 $3,000 $3,500 using simulation models suggest the following: Discounted total costs (thousands) per 1,000, 50 year olds • Screening is cost-effective; since compliance is not Efficient HII HS FIT FS 100 percent, encouraging screening by any of a small frontier group of strategies will save life years. FS + HII FS + HS FS + FIT COL • Where the total budget available is limited, the Source: Zauber 2010, based on Knudsen and others 2010. lowest-cost testing strategy involves gFOBT. Note: COL = colonoscopy; FIT = fecal immunochemical test; FS = flexible sigmoidoscopy; HII = Hemoccult • The differences in the cost-effectiveness of some II (fecal occult blood test, FOBT); HS = Hemoccult Sensa (FOBT). This analysis assumes 100 percent strategies are modest and susceptible to variation in adherence with each strategy. assumptions. • Knowledge of details of disease progression is limited, a larger analysis as given by Knudsen and others (2010), for example, how untreated adenomas progress to which assumed that subjects were 100 percent adherent cancer or how this progression varies by individual to screening tests and subsequent diagnostic or treat- characteristics such as age, gender, and family history. ment for those with positive tests or cancer diagnosis. Different assumptions regarding disease progression Gains in life years are plotted on the y-axis and total affect the results of the simulation model. costs are plotted on the x-axis. Each possible interven- tion strategy is represented by a point. The higher the The evidence base for different screening methods point is, the more effective is the screening strategy; the continues to evolve. FS and colonoscopy can be per- further to the right the point is, the more expensive is formed by appropriately trained nonphysicians, includ- the screening strategy (Marks 2002). ing nurses (Maule 1994; Wallace and others 1999). Cost-effectiveness analysis does not select which Newer generations of chemotherapy agents increase strategy is economically preferred overall, but only life expectancy modestly, but they considerably increase which strategy is the most effective in terms of life years treatment costs; accordingly, most screening methods gained for a given level of desired (or possible) expen- (although not colonoscopy by a narrow margin) remain diture. The eight CRC screening strategies are ranked in cost-saving (Lansdorp-Vogelaar and others 2009). order of the procedure, with the least life years gained relative to no screening (Knudsen and others 2010). The European Studies strategies are then compared incrementally by order- The European studies identified by the systematic ing of the life years gained relative to the costs of each review by Pignone and others (2002) (two studies of screening strategy. Those strategies that have the most gFOBT and one of FS, in Denmark, Norway, and the life years gained for a given level of cost are considered United Kingdom) all supported the conclusion that 112 Cancer screening is cost-effective compared with no screening, • Relative costs of procedures such as colonoscopy/FS and the cost per life year saved was lower than in the and colonoscopy/gFOBT vary by country, which is United States because of the higher overall medical likely to affect rankings. costs in the United States (Gyrd-Hansen, Søgaard, and • For the Republic of Korea, for example, it has been Kronborg 1998; Norum 1998; Whynes and others 1998). argued that colonoscopy reimbursement rates are artificially low and do not reflect cost (Park, Yun, and Studies in High-Income Economies in Asia Kwon 2005). This is likely to affect the relative rank- Table 6.1 summarizes cost-effectiveness results for four ing of strategies for that country. studies of Asian economies, with the standardized model for the United States from Pignone, Russell, and These four Asian economies do not all have published Wagner (2005) for comparison. The Asian economies cost-effectiveness thresholds used for public decision are Hong Kong SAR, China; the Republic of Korea; making. However, from the WHO Commission on Singapore; and Taiwan, China. Observations from these Health (WHO 2001), health interventions costing up results include the following: to three times per capita gross domestic product per disability-adjusted life year saved should be considered. • Screening in the four Asian economies is cost- By this criterion, all the methods of CRC screening con- effective, although the simpler models used in some sidered here (gFOBT, FS, and colonoscopy) would be studies may underestimate costs. acceptable in these four economies. Another study for Table 6.1 Selected Costs and Cost-Effectiveness of Screening, Per Capita GDP, and Colorectal Cancer Incidence in Four High-Income Asian Economies Compared with the United States Hong Kong United States Korea, Rep. Taiwan, China SAR, China Singapore Item cost or value (US$) 2005a 2004b 2004c 2003d 2004e gFOBT 10 1.91 0.60 4 5.59 Colonoscopy (diagnostic) 625 43.80 66.20 450 368.72 Colonoscopy 900 — — 830 446.93 (polyp removal) Sigmoidoscopy 200 22.18 35.30 244 134.08 Treat CRC, local 24,000 4,291.67 3,117.60 16,552 11,173.18 Treat CRC, regional 31,000 — 7,705.90 27,321 19,553.07 Treat CRC, distant 40,000 8,583.33 7,647.10 71,751 — Colon perforation 24,000 2,500 1,617.60 10,790 4,863.69 Cost-effectiveness versus no gFOBT 9,676 COL (5 years) 1,142; gFOBT 70 gFOBT 6,222 gFOBT 91 screening COL 21,000 others dominated FS 594 FS 8,044 FS 190 US$/LYS (median of 5 models, COL 407 COL 7,211 COL 225 standardized assumptions) Per capita GDPf 46,760 19,028 17,461 31,426 34,466 g CRC incidence /100,000 34.1 men 46.9 men 40.2 men 50.1 (crude, men 34.1 men (age-standardized 25.0 women 25.6 women 29.7 women and women 25.0 women to world population) (Taiwan, China) combined) Note: COL = colonoscopy; CRC = colorectal cancer; FS = flexible sigmoidoscopy; GDP = gross domestic product; gFOBT = guaiac fecal occult blood test; LYS = life years saved; — = not available. a. Pignone, Russell, and Wagner 2005. b. Park, Yun, and Kwon 2005. c. Wu and others 2006. d. Tsoi and others 2008. e. Wong, Leong, and Leong 2004. f. World Bank 2013, except Taiwan, China, which is http://www.indexmundi.com. g. Bray and others 2013, except Hong Kong SAR, China, which is Tsoi and others 2008. Colorectal Cancer 113 Hong Kong SAR, China, for women only (Woo, Kim, be cost-effective in LMICs. This is likely to be a better and Leung 2007), concluded that CRC screening had guide if relative costs and “standard care”—the alter- higher costs than for men per disability-adjusted life year native to which a particular treatment is compared— saved and would not be cost-effective (CRC incidence are reasonably similar. rates in women are lower than in men). Studies in LMICs RECOMMENDATIONS Two global cost-effectiveness models report estimates Country income level does not have to dictate the avail- of cost-effectiveness of interventions for various ability of screening, diagnosis, and treatment. Countries world regions. Ginsberg and others (2010) conclude can help overcome resource constraints by accessing that expanding treatment in low-income countries is technical assistance from the IARC and international a higher priority than screening. Ginsberg and others networks, or from partnerships with cancer centers or (2012) come to a similar conclusion when focusing cancer agencies in other countries. Local champions on Southeast Asia and Sub-Saharan Africa, and they are essential for moving CRC screening and treatment conclude that screening colonoscopy is cost-effective forward as a priority. in Sub-Saharan Africa. However, the feasibility of Table 6.2 summarizes the authors’ recommendations implementing this approach in light of resource avail- on how screening and diagnosis for CRC might be ability and health system infrastructure limitations implemented in four different resource environments: was not addressed. LICs, rural areas of MICs, urban areas of MICs, and A systematic search did not identify other studies of HICs. These correspond approximately to the basic, the cost-effectiveness of screening in LMICs. One other limited, enhanced, and maximal resource environments study was found from an unsystematic search using for a similar exercise undertaken for breast cancer by the Internet for the Islamic Republic of Iran (Barouni the Global Breast Health Initiative and discussed in and others 2012). From the results presented, it is pos- chapter 3. Recommendations for the treatment of CRC sible to conclude that colonoscopy every 10 years is in these resource environments are found in online cost-effective, but not very cost-effective, in the Islamic annex 6B (tables 6B.1 through 6B.5). These recommen- Republic of Iran, but that gFOBT screening is not dations provide initial guidance only and need to be cost-effective. There are problems with the calculation validated by a larger international expert group. of the incremental cost-effectiveness ratios in this study, but enough information is given to permit the reader to recalculate them. Low-Income Countries We anticipate that CRC screening would be equally In LICs, the incidence of CRC is relatively low; other cost-effective in urban areas of other upper-middle- diseases—including other cancers—are a higher income countries where incidence rates approach levels priority for screening and treatment. Laying the similar to those in HICs (30 or more per 100,000 in foundation for cancer screening and treatment is men, age-standardized rates). Lambert, Sauvaget, and important. This process includes investing in public Sankaranarayanan (2009, 255) conclude that population health and primary health care where screening is screening for CRC is not the highest priority in most initiated, in hospital systems, and in a cancer regis- LMICs, but that it deserves to be developed “in limited try. Investments in health require medical personnel, regions of large emerging countries where there is a as well as good systems for monitoring and evalua- shift to Western lifestyle and an aging population,” and tion and quality control. Smaller countries lacking they point to Mumbai; Hong Kong SAR, China; and São specific resources, such as radiation facilities and Paulo as examples. specialized laboratories, may need to rely on other countries in the same region. Even in LICs, surgery for colon cancer at a good Treatment district hospital is possible to save lives and improve the No literature on the cost-effectiveness of treatment quality of remaining life. If colonoscopy is unavailable for CRC was found for LMICs. Chapter 16 summa- as a diagnostic tool, barium enema may be an option. rizes the evidence of cost-effectiveness of treatment Radiation therapy is available only in limited volumes, if in HICs, making the assumption that treatments that at all, and laboratory services required for chemotherapy are “very cost-effective” in HICs are the first candi- are not likely to be available. Pain management for late- dates for consideration in MICs, while treatments stage cancers is an ethical imperative, since the ability to that are “not cost-effective” in HICs are unlikely to treat effectively is extremely limited. 114 Cancer Table 6.2 Proposed Strategies for Colorectal Cancer Screening and Diagnosis, by Country Resource Level Level of resources General Detection and diagnosis Basic • Build capacity: human, physical (for • Barium enema if colonoscopy not available; in emergency situations, may be example, radiation capacity), cancer diagnosed at surgery registry Limited • Establish capacity for colonoscopy • Opportunistic screening for those covered by health insurance (needed for diagnosis) • Diagnostic colonoscopy (or barium enema) for those with symptoms • Engage in partnership arrangements with cancer centers to build capacity • Establish national guidelines • Build quality assurance for lab testing Enhanced • Join international screening networks • Establish organized screening in high-incidence cities/regions starting at age • Provide support to less-well-resourced 50 years in persons at average risk: use annual or biennial sensitive gFOBT countries in region or FIT; FS (see text for discussion of interval); or colonoscopy every 10 years • Considerable infrastructure is required to support organized screening, including invitations, recalls, reminders, tracking screening test results, ensuring follow-up of those with an abnormal screening test, etc. Maximal • National (or jurisdiction-wide) organized screening: starting at age 50 years in persons at average risk: use annual or biennial sensitive gFOBT or FIT; FS (see text for discussion of interval); or colonoscopy every 10 years; in those at increased risk because of family history, consider colonoscopy Note: Since no international consensus-setting exercise has occurred, this categorization represents a basis for further discussion and work and not a definitive analysis. The basic resource level is assumed to correspond to low-income countries (limited or no access to radiation and likely insufficient support for blood chemistry to undertake chemotherapy). The limited resource level corresponds to rural areas of middle-income countries, where distances to radiation and chemotherapy resources make their use in treatment difficult. In urban areas of middle-income countries (enhanced level), radiation therapy is available, as are many chemotherapy drugs no longer under patent. The maximal level corresponds to resource availability in high-income countries. See chapter 16 in this volume for more detailed discussion of resource levels. The recommendations are cumulative: any intervention that is feasible at a lower resource level is also an option at higher resource levels. Blank cells indicate that no additional options of a particular type of treatment are available at the particular resource level considered. FIT = fecal immunochemical test; FS = flexible sigmoidoscopy; gFOBT = guaiac fecal occult blood test. Middle-Income Countries 100,000 may need to begin planning (for example, In MICs, there is an increase (more pronounced countries such as Cuba, Lebanon, and Malaysia). Data initially in urban areas) in CRC incidence and the are not available using GLOBOCAN for individual ability to intervene. Opportunistic screening increases cities, but similar CRC incidence thresholds could be for those covered by health insurance. Those coun- used to consider when to begin to take action on CRC tries that have already begun organized screening for screening. other cancers (including countries in Latin America For CRC screening, gFOBT is inexpensive; however, and the Caribbean and upper-middle-income Asia; additional investments are needed to implement all the see chapter 12 in this volume [Sullivan, Sullivan, and components of organized screening. MICs initiating Ginsburg 2015]) may decide to implement screening organized CRC screening may be advised to use FIT initially as a pilot study in selected urban regions. Other rather than gFOBT; doing so may become more attrac- regions that are beginning organized screening for other tive if a larger demand for such tests results in a decrease cancers may decide to incorporate screening as well, for in the unit costs of the kits. example, by developing programs in cities in Asia. MICs also have more resources for treatment and Priority countries are those where age-standardized can extend this to a larger proportion of the population. CRC incidence rates in men are more than 30 per 100,000 As cancers are detected earlier, the goal of treatment (for example, Hungary, Serbia, and other countries in shifts from palliation to cure. MICs can be active par- Eastern Europe), in addition to those economies with ticipants in international networks and local centers of existing pilot programs, such as Argentina. Countries excellence and can provide support for other countries where CRC incidence rates in men approach 20 per in their region. Colorectal Cancer 115 High-Income Countries GLOBOCAN 2012 database. Observable population-based data were derived from Cancer Incidence in Five Continents, In HICs, cost-effectiveness considerations suggest that 10th edition and for trends over time from CI5 Plus (http:// FIT, FS accompanied by a sensitive gFOBT or FIT, or ci5.iarc.fr/CI5plus/Default.aspx). The discussion of bur- colonoscopy are options for screening. The Republic den, including risk factors, however, includes all ages unless of Korea and Singapore, which started their screen- otherwise noted. Interventions also apply to all age groups, ing programs more recently than European countries, except where age ranges or cutoffs are specified. have opted to use FIT. Adherence to screening varies; although each test has its advocates, the best test is the one that gets done, and done well. A wider range of REFERENCES treatment options is feasible in these countries, which typically have higher cost-effectiveness thresholds. Ahlquist, D. A., H. Zou, M. Domanico, D. W. Mahoney, T. C. Yab, and others. 2012. “Next-Generation Stool DNA Test Accurately Detects Colorectal Cancer and Large CONCLUSIONS Adenomas.” Gastroenterology 142: 248–56. Anderson, B. O., J. Lipscomb, R. H. Murillo, and D. B. Thomas. The burden of CRC is increasing worldwide. There is “Breast Cancer.” In Disease Control Priorities (third edi- an unparalleled opportunity for prevention and early tion): Volume 3, Cancer, edited by H. Gelband, P. Jha, detection of CRC, based on our knowledge of colorectal R. Sankaranarayanan, and S. Horton. Washington, DC: carcinogenesis. The implementation of effective screen- World Bank. ing tests, both stool-based (gFOBT, FIT) and endoscopic Atkin, W. S., R. Edwards, I. Kralj-Hans, K. Wooldrage, A. R. Hart, and others. 2010. “Once-Only Flexible Sigmoidoscopy (FS, colonoscopy), coupled with advances in treatment Screening in Prevention of Colorectal Cancer: A Multicentre (colonoscopic polypectomy, surgery, radiation therapy, Randomised Controlled Trial.” The Lancet 375 (9726): and chemotherapy) are cost-effective approaches. 1624–33. Since screening is a process, it is most effective when Barouni, M., M. H. Larizadeh, A. Sabermahani, and H. Ghaderi. delivered within an organized program, requiring infra- 2012. “Markov’s Modeling for Screening Strategies for structure and resources to ensure the benefits while Colorectal Cancer.” Asia Pacific Journal of Cancer Prevention minimizing the harms. CRC screening and treatment 13: 5125–29. are becoming priorities in an increasing number of Baxter, N. N., M. A. Goldwasser, L. F. Paszat, R. Saskin, countries, as health resources are enhanced and changes D. R. Urbach, and others. 2009. “Association of Colonoscopy in lifestyles and risk factors lead to a rise in incidence. and Death from Colorectal Cancer: A Population-Based, Research on the cost-effectiveness of options in these Case-Control Study.” Annals of Internal Medicine 150: 1–8. Benson, V. S., W. S. Atkin, J. Green, M. R. Nadel, J. Patnick, countries is needed, ideally adapting available, well- and others. 2012. “Toward Standardizing and Reporting constructed models to these environments. The devel- Colorectal Cancer Screening Indicators on an International opment of regional CRC screening guidelines would be Level: The International Colorectal Cancer Screening helpful; the resource-based recommendations outlined Network.” International Journal of Cancer 130 (12): in this chapter may be useful in that process. 2961–73. Benson, A. B., M. A. Choti, A. M. Cohen, J. H. Doroshow, C. Fuchs, and others. 2000. “NCCN Practice Guidelines for NOTES Colorectal Cancer.” Oncology (Williston Park) 14: 203–12. Bray, F., J. S. Ren, E. Masuyer, and J. Ferlay. 2013. “Estimates World Bank Income Classifications as of July 2014 are as of Global Cancer Prevalence for 27 Sites in the Adult follows based on estimates of gross national income (GNI) per Population in 2008.” International Journal of Cancer capita for 2013: 132: 1133–45. Brenner, H., J. Change-Claude, C. M. Seiler, A. Rickert, and • Low-income countries (LICs) = US$1,045 or less M. Hoffmeister. 2011. “Protection from Colorectal Cancer • Middle-income countries (MICs) are subdivided: after Colonoscopy: A Population-Based, Case-Control a) lower-middle-income = US$1,046–US$4,125 Study.” Annals of Internal Medicine 154: 22–30. b) upper-middle-income (UMICs) = US$4,126–US$12,745 Bressler, B., L. F. Paszat, Z. Chen, D. M. Rothwell, C. Vinden, • High-income countries (HICs) = US$12,746 or more. and others. 2007. “Rates of New or Missed Colorectal Cancers after Colonoscopy and Their Risk Factors: A 1. Maps and figures in this chapter are based on incidence Population Based Analysis.” Gastroenterology 132: 96–102. and mortality estimates for ages 0–69, consistent with Butterworth, A. S., J. P. T. Higgins, and P. Pharoah. 2006. reporting in all DCP3 volumes. Global cancer statistics “Relative and Absolute Risk of Colorectal Cancer for are estimates for 2012 and have been provided by the Individuals with a Family History: A Meta-Analysis.” International Agency for Research on Cancer from its European Journal of Cancer 42: 216–27. 116 Cancer Chan, A. T., and E. L. Giovannucci. 2010. “Primary Prevention Imperiale, T. F., D. F. Ransohoff, S. H. Itzkowitz, B. A. Turnbull, of Colorectal Cancer.” Gastroenterology 138 (6): 2029–43. and M. E. Ross. 2004. “Fecal DNA versus Fecal Occult Edwards, B. K., E. Ward, B. A. Kohler, C. Eheman, A. G. Zauber, Blood for Colorectal-Cancer Screening in an Average-Risk and others. 2010. “Annual Report to the Nation on the Status Population.” New England Journal of Medicine 351 (26): of Cancer, 1975–2006, Featuring Colorectal Cancer Trends 2704–14. and Impact of Interventions (Risk Factors, Screening, and Johnson, C. D., M. H. Chen, A. Y. Tolendano, J. P. Heiken, Treatment) to Reduce Future Rates.” Cancer 116 (3): 544–73. A. Dachman, and others. 2008. “Accuracy of CT Elmunzer, B. J., R. A. Hayward, P. S. Schoenfeld, S. D. Saini, Colonography for Detection of Large Adenomas and A. Deshpande, and others. 2012. “Effect of Flexible Cancers.” New England Journal of Medicine 359: 1207–17. Sigmoidoscopy-Based Screening on Incidence and Johnson, C. M., C. Wei, J. E. Ensor, D. J. Smolenski, C. I. Amos, Mortality of Colorectal Cancer: A Systematic Review and and others. 2013. “Meta-Analysis of Colorectal Cancer Risk Meta-Analysis of Randomized Controlled Trials.” PLOS Factors.” Cancer Causes and Control 24: 1207–22. Medicine 9 (12): 1–9. Knudsen, A. B., I. Lansdorp-Vogelaar, C. M. Rutter, J. E. Savarino, Ferlay J., I. Soerjomataram, M. Ervik, R. Dikshit, S. Eser, and M. Van Ballegooijen, and others. 2010. “Cost-Effectiveness others. 2013. GLOBOCAN 2012 V1.0, Cancer Incidence and of Computed Tomographic Colonography Screening for Mortality Worldwide: IARC Cancerbase No. 11 [Internet]. Colorectal Cancer in the Medicare Population.” Journal of Lyon, France: International Agency for Research on Cancer. the National Cancer Institute 102: 1238–52. http://globocan.iarc.fr. Kronborg, O., C. Fenger, J. Olsen, O. D. Jorgensen, and Ginsberg, G. M., J. A. Lauer, S. Zelle, S. Baeten, and R. Baltussen. O. Sondergaard. 1996. “Randomised Study of Screening 2012. “Cost Effectiveness of Strategies to Combat Breast, for Colorectal Cancer with Faecal-Occult-Blood Test.” The Cervical, and Colorectal Cancer in Sub-Saharan Africa and Lancet 348: 1467–71. South East Asia: Mathematical Modeling Study.” British Lambert, R., C. Sauvaget, and R. Sankaranarayanan. 2009. Medical Journal 344: E614. “Mass Screening for Colorectal Cancer Is Not Justified Ginsberg, G. M., S. S. Lim, J. A. Lauer, B. P. Johns, and in Most Developing Countries.” International Journal of C. R. Sepulveda. 2010. “Prevention, Screening and Circulation 125: 253–56. Treatment of Colorectal Cancer: A Global and Regional Lansdorp-Vogelaar, I., M. Van Ballegooijen, A. G. Zauber, Generalized Cost-Effectiveness Analysis.” Cost Effectiveness J. D. F. Habbemaand, and E. J. Kuipers. 2009. “Effect and Resource Allocation 8: 2–17. of Rising Chemotherapy Costs on the Cost Savings of Goss, P. E., B. L. Lee, R. Bidovinac-Cmjevic, K. Strasser-Weippl, Colorectal Cancer Screening.” Journal of the National Y. Chavani-Guerra, and others. 2013. “Planning Cancer Cancer Institute 101: 1412–22. Control in Latin America and the Caribbean.” The Lancet Lee, S. J., W. J. Boscardin, I. Stijacic-Cenzer, J. Conell-Price, Oncology 14: 391–436. S. O’Brien, and others. 2013. “Time Lag to Benefit Greenberg, D., C. Earle, C. H. Fang, A. Eldar-Lissai, and after Screening for Breast and Colorectal Cancer: Meta- P. J. Neumann. 2010. “When Is Cancer Care Cost-Effective? A Analysis of Survival Data from the United States, Sweden, Systematic Overview of Cost-Utility Analyses in Oncology.” United Kingdom, and Denmark.” British Medical Journal Journal of the National Cancer Institute 102: 82–88. 346: E8441. Gyrd-Hansen, D., J. Søgaard, and O. Kronborg. 1998. Leufkens, A. M., M. A. Van Den Bosch, M. S. Van Leeuwen, and “Colorectal Cancer Screening: Efficiency and Effectiveness.” P. D. Siersema. 2011. “Diagnostic Accuracy of Computed Health Economics 7: 9–20. Tomography for Colon Cancer Staging: A Systematic Hardcastle, J. D., J. O. Chamberlain, M. H. Robinson, Review.” Scandinavian Journal of Gastroenterology 46 (7–8): S. M. Moss, S. S. Amar, and others. 1996. “Randomised 887–94. doi:10.3109/00365521.2011.574732. Controlled Trial of Faecal-Occult-Blood Screening for Levin, B., D. A. Lieberman, B. McFarland, K. S. Andrews, Colorectal Cancer.” The Lancet 348: 1472–77. D. Brooks, and others. 2008. “Screening and Surveillance for Hoffe, S. E., R. Shridhar, and M. C. Biajioli. 2010. “Radiation the Early Detection of Colorectal Cancer and Adenomatous Therapy for Rectal Cancer: Current Status and Future Polyps, 2008: A Joint Guideline from the American Directions.” Cancer Control 17: 25–34. Cancer Society, the U.S. Multi-Society Task Force on IARC (International Agency for Research on Cancer). 2010. Colorectal Cancer, and the American College of Radiology.” Handbook of Cancer Prevention, vol. 10, Cervix Cancer Gastroenterology 134 (5): 1570–95. Screening. Lyon, France: IARC Press. Lieberman, D. A., D. K. Rex, S. J. Winawer, F. M. Giardiello, Içli, F., H. Akbulut, S. Bazarbashi, M. A. Kuzu, M. K. Mallath, D. A. Johnson, and others. 2012. “United States Multi- and others. 2010. “Modification and Implementation of Society Task Force on Colorectal Cancer Guidelines for NCCN Guidelines on Colon Cancer in the Middle East and Colonoscopy Surveillance after Screening and Polypectomy: North Africa Region.” Journal of the National Comprehensive A Consensus Update by the US Multi-Society Task Force Cancer Network 8 (Suppl. 3): S22–25. on Colorectal Cancer.” Gastroenterology 143 (3): 844–57. Imperiale, T. F., D. F. Ransohoff, S. H. Itzkowitz, T. R. Levin, doi:10.1053/J.Gastro.2012.06.001. P. Lavin, and others. 2014. “Multitarget Stool DNA Testing Mandel, J. S., J. H. Bond, T. R. Church, D. C. Snover, for Colorectal-Cancer Screening.” New England Journal of G. M. Bradley, and others. 1993. “Reducing Mortality from Medicine 370 (14): 1287–97. Colorectal Cancer by Screening for Fecal Occult Blood. Colorectal Cancer 117 Minnesota Colon Cancer Control Study.” New England Randomized Controlled Trial—SCORE.” Journal of the Journal of Medicine 328: 1365–71. National Cancer Institute 103 (17): 1310–22. Manzur, J. 2013. “Argentinean Perspectives on Cancer Control Segnan, N., J. Patnick, and L. von Karsa, eds. 2010. European in Latin America.” The Lancet Oncology 14: 387–88. Guidelines for Quality Assurance in Colorectal Cancer Marks, D. H. 2002. “Visualizing Cost-Effectiveness Analysis.” Screening and Diagnosis. Luxembourg: Publications Office Journal of the American Medical Association 287: 2428–29. of the European Union. Maule, W. F. 1994. “Screening for Colorectal Cancer by Nurse Sullivan, T., R. Sullivan, and O. M. Ginsburg. “Screening Endoscopists.” New England Journal of Medicine 330: for Cancer: Considerations in Low- and Middle-Income 183–87. Countries.” In Disease Control Priorities (third edi- Miles, A., J. Cockburn, R. A. Smith, and J. Wardle. 2004. tion): Volume 3, Cancer, edited by H. Gelband, P. Jha, “A Perspective from Countries Using Organized Screening R. Sankaranarayanan, and S. Horton. Washington, DC: Programs.” Cancer 101: 1201–13. World Bank. Ng, S. C., and S. H. Wong. 2013. “Colorectal Cancer Screening Towler, B., L. Irwig, P. Glasziou, J. Kewenter, D. Weller, and in Asia.” British Medical Bulletin 105: 29–42. others. 1998. “A Systematic Review of the Effects of Screening Norum, J. 1998. “Prevention of Colorectal Cancer: A Cost- for Colorectal Cancer Using the Faecal Occult Blood Test, Effectiveness Approach to a Screening Model Employing Hemoccult.” British Medical Journal 317 (7158): 559–65. Sigmoidoscopy.” Annals of Oncology 9: 613–18. Tsoi, K. K. F., S. S. M. Ng, M. C. M. Leung, and J. J. Y. Sung. Park, S. M., Y. H. Yun, and S. Kwon. 2005. “Feasible Economic 2008. “Cost-Effectiveness Analysis on Screening for Strategies to Improve Screening Compliance for Colorectal Colorectal Neoplasm and Management of Colorectal Cancer in Korea.” World Journal of Gastroenterology 11 (11): Cancer in Asia.” Alimentary Pharmacology and Therapeutics 1587–93. 28: 353–63. Pickhardt, P. J., J. R. Choi, I. Hwang, J. L. Butler, M. A. Puckett, USPSTF (U.S. Preventive Services Task Force). 2008. “Screening and others. 2003. “Computed Tomographic Virtual for Colorectal Cancer: U.S. Preventive Services Task Force Colonoscopy to Screen for Colorectal Neoplasia in Recommendation Statement.” Annals of Internal Medicine Asymptomatic Adults.” New England Journal of Medicine 149: 627–37. 349: 2191–200. van Rossum, L. G., A. F. Van Rijn, R. J. Laheij, M. G. Van Pignone, M., L. Russell, and J. Wagner, eds. 2005. Economic Oijen, P. Fockens, and others. 2008. “Random Comparison Models of Colorectal Cancer Screening in Average-Risk Adults. of Guaiac and Immunochemical Fecal Occult Blood Washington, DC: National Academies Press. Tests for Colorectal Cancer in a Screening Population.” Pignone, M., S. Saha, R. Hoerger, and J. Mandelblatt. 2002. Gastroenterology 135: 82–90. “Cost-Effectiveness Analyses of Colorectal Cancer Screening: von Karsa, L., T. A. Lignini, J. Patnick, R. Lambert, and A Systematic Review for the U.S. Preventive Services Task C. Sauvaget. 2010. “The Dimensions of the CRC Problem.” Force.” Annals of Internal Medicine 137 (2): 96–104. Best Practice & Research Clinical Gastroenterology Quintero, E., A. Castells, L. Bujanda, J. Cubiella, D. Salas, and 24: 381–96. others. 2012. “Colonoscopy versus Fecal Immunochemical Wallace, M. B., J. A. Kemp, F. Meyer, K. Horton, A. Reffel, and Testing in Colorectal-Cancer Screening.” New England others. 1999. “Screening for Colorectal Cancer with Flexible Journal of Medicine 366: 697–706. Sigmoidoscopy by Nonphysician Endoscopists.” American Rabeneck, L., and L. F. Paszat. 2010. “Circumstances in Which Journal of Medicine 107: 214–28. Colonoscopy Misses Cancer.” Frontline Gastroenterology WHO (World Health Organization). 2001. Macroeconomics 1: 52–58. and Health: Investing in Health for Economic Development. Rabeneck, L., J. M. Tinmouth, L. F. Paszat, N. N. Geneva: WHO. Baxter, L. D. Marrett, and others. 2014. “Ontario’s Whynes, D. K., A. R. Neilson, A. R. Walker, and J. D. Hardcastle. Coloncancercheck: Results from Canada’s First Province- 1998. “Faecal Occult Blood Screening for Colorectal Cancer: Wide Colorectal Cancer Screening Program.” Cancer, Is It Cost-Effective?” Health Economics 7: 21–29. Epidemiology, Biomarkers, and Prevention 23: 508–15. Winawer, S. J. 2007. “Colorectal Cancer Screening.” Best Practice Schoen, R. E., P. F. Pinsky, J. L. Weissfeld, L. A. Yokochi, & Research Clinical Gastroenterology 21 (6): 1031–48. T. Church, and others. 2012. “Colorectal-Cancer Incidence Winawer, S. J., R. H. Fletcher, L. Miller, F. Godlee, M. H. Stolar, and Mortality with Screening Flexible Sigmoidoscopy.” New and others. 1997. “Colorectal Cancer Screening: England Journal of Medicine 366: 2345–57. Clinical Guidelines and Rationale.” Gastroenterology Sebag-Montefiore, D., R. J. Stephens, R. Steele, J. Monson, 112: 594–642. R. Grieve, and others. 2009. “Preoperative Radiotherapy ver- Winawer, S. J., A. G. Zauber, M. N. Ho, M. J. O’Brien, sus Selective Postoperative Chemoradiotherapy in Patients L. S. Gottlieb, and others. 1993. “Prevention of Colorectal with Rectal Cancer (MRC CR07 and NCIC-CTG C016): A Cancer by Colonoscopic Polypectomy.” New England Multicentre, Randomised Trial.” The Lancet 373: 811–20. Journal of Medicine 329 (27): 1977–81. Segnan, N., P. Armaroli, L. Bonelli, M. Risio, S. Sciallero, and Wong, S. S., A. P. K. Leong, and T. Z. Leong. 2004. “Cost- others. 2011. “Once-Only Sigmoidoscopy in Colorectal Effectiveness Analysis of Colorectal Cancer Screening Cancer Screening: Follow-Up Findings of the Italian Strategies in Singapore: A Dynamic Decision Analytic 118 Cancer Approach.” Studies in Health Technology Information of Colorectal Cancer Screening with Stool DNA Testing 107 (Pt 1): 104–10. in Intermediate-Incidence Countries.” BMC Cancer Woo, P. P. S., J. J. Kim, and G. M. Leung. 2007. “What Is the 6: 136–48. Most Cost-Effective Population-Based Cancer Screening Zauber, A. G. 2010. “Cost-Effectiveness of Colonoscopy.” Program for Chinese Women?” Journal of Clinical Oncology Gastroenterology Clinics of North America 20: 751–70. 25: 617–24. Zauber, A. G., S. J. Winawer, M. J. O’Brien, I. Lansdorp-Vogelaar, World Bank. 2013. World Databank. http://databank M. Van Ballegooijen, and others. 2012. “Colonoscopic .worldbank.org/data/home.aspx. Polypectomy and Long-Term Prevention of Colorectal- Wu, G. H.-M., Y.-M. Wang, A. M.-F. Yen, J.-M. Wong, Cancer Deaths.” New England Journal of Medicine H.-C. Lai, and others. 2006. “Cost-Effectiveness Analysis 366: 687–96. Colorectal Cancer 119 Chapter 7 Treating Childhood Cancer in Low- and Middle-Income Countries Sumit Gupta, Scott C. Howard, Stephen P. Hunger, Federico G. Antillon, Monika L. Metzger, Trijn Israels, Mhamed Harif, and Carlos Rodriguez-Galindo BURDEN OF CHILDHOOD CANCER IN LMICs care workers must appropriately refer patients to third-level centers capable of recognizing and diagnos- In high-income countries (HICs), the annual inci- ing pediatric malignancies and then entering data into dence of childhood cancer is approximately 140 per cancer registries. Breaks in the chain of events may occur 1 million children younger than age 15 years, although at any step. estimates vary between and within countries (Parkin A comparison of leukemia and non-leukemia cancer and others 1998). Incidence rates from low- and incidence rates is instructive. Pediatric leukemia may middle-income country (LMIC) registries are generally present with a variety of nonspecific symptoms, such significantly lower, as annual rates per 1 million children as fever, anemia, malaise, or hemorrhage; many of of 45.6 in Namibia and 64.4 in India, respectively, illus- the symptoms are also associated with infections. Most trate (Parkin and others 1998). Some of this variation non-leukemia cancers present with enlarging masses may relate to differences in environmental exposures more easily recognizable as malignant. Accordingly, or to biologic susceptibility. However, deficiencies in the magnitude of underdiagnosis would be expected to diagnosis and registration likely contribute significantly be greater in leukemia than in non-leukemia cancers; to differences in the reported incidence of cancer, both registry data bear this out. In the most recent global overall and of particular subtypes, such as acute leuke- compilation of pediatric cancer data, leukemia inci- mias (Howard and others 2008). dence in low-income countries (LICs) averaged 16.4 Incidence data from high-quality cancer registries per million children, far lower than the incidence rate with complete population coverage are rare in LMICs. In of 36.5 in middle-income countries (MICs) and 40.9 in 2006, only 8 percent of people in Asia and 11 percent in HICs (figure 7.2) (Howard and others 2008). The non- Sub-Saharan Africa were covered by population-based leukemia cancer incidence was broadly similar in all cancer registries; when only high-quality registries are income groups: 85 in LICs, 70 in MICs, and 89 in HICs considered, these rates are 4 percent and 1 percent, (Howard and others 2008). The underdiagnosis of child- respectively (Ferlay and others 2010). hood brain tumors is likely even greater; many regions Multiple steps are required for children with cancer report few or no incident cases of pediatric central to be included in a registry (figure 7.1). Caregivers must nervous system malignancies (Parkin and others 1998). seek medical attention for symptoms. Primary health Corresponding author: Sumit Gupta, Toronto Hospital for Sick Children, Toronto, Canada, sumit.gupta@sickkids.ca 121 Figure 7.1 Links in the Chain of Childhood Cancer Diagnosis and Registration with Potential Barriers in Low- and Middle-Income Countries Clinical diagnosis Histologic Recognition of Arrival to Arrival to Registration of cancer diagnosis possible cancer primary care tertiary care in a database (e.g., detectable of cancer mass) Education/beliefs Education/beliefs Trained Education/beliefs Personnel Personnel Symptoms/signs Symptoms/signs personnel Symptoms/signs (surgery, (data management) Distance Clinical Distance oncology, Supportive Transportation infrastructure Transportation pathology) administration SES Funding SES Pathology lab Hardware/Software Family support Family support Funding Funding Source: © John Wiley and Sons. Reproduced, with permission, from Howard and others 2008; further permission required for reuse. Note: SES = socioeconomic status. Figure 7.2 Reported Incidence Rate of Childhood Leukemia and Its Association with 2005 Gross National Income, Selected Economies 60 Costa Rica 55 Ecuador Hong Kong SAR, China 50 Philippines Singapore Canada Finland Denmark 45 United Arab Emirates Australia Sweden Norway Uruguay Italy Colombia Spain United States Switzerland 40 China Croatia France New Zealand United Kingdom Pakistan Korea, Rep. Algeria Hungary Netherlands Iceland 35 Peru Portugal Estonia Slovenia Japan Vietnam Poland Slovak Republic Germany Incidence Bulgaria Kuwait 30 Brazil 25 Thailand Israel South Africa 20 Zimbabwe India 15 Uganda 10 Nigeria Papua New Guinea 5 Namibia Mali Malawi 0 10,000 20,000 30,000 40,000 50,000 60,000 Mean annual per capita gross national income Source: © John Wiley and Sons. Reproduced, with permission, from Howard and others 2008; further permission required for reuse. Note: Incidence rates are per 1 million children per year. Underdiagnosis and underregistration are not Comparing Indian cancer registries, the male-to-female uniform across all segments of the population. In ratio in acute lymphoblastic leukemia (ALL) incidence Jordan and Honduras, higher leukemia incidence rates ranged from 1.7 per million in Mumbai to 2.6 in Delhi, are reported in urban compared with rural districts compared with 1.3 in Canada during the same time (Al Sheyyab and others 2003; Metzger and others 2003). period (Parkin and others 1998). At least in some cases, 122 Cancer underdiagnosis may affect girls and rural children Ineffectiveness of Prevention and Screening disproportionately. Most pediatric malignancies are not caused by modifiable In addition, not only is childhood cancer severely risk factors, and public health campaigns would have underrepresented in LMIC cancer registration; only a limited impact on decreasing the incidence, although proportion of the children who are registered receive impact on delayed presentation is possible. Similarly, appropriate treatment. From a survey of health care population-based screening programs have not been workers in 10 LMICs, including Bangladesh, the shown to affect cancer mortality in children (Schilling Philippines, Tanzania, and Vietnam, 15–37 percent of and others 2002). Decreasing childhood cancer mortal- the expected patients were seen (Ribeiro and others ity rates requires early and accurate diagnosis followed 2008). Including children missed by registries would by effective treatment. lower this percentage even further. Thus, the approximately 175,000 children diagnosed with cancer globally every year are likely to represent Achievability of Cure a significant underrepresentation of the worldwide In HICs, over 80 percent of children with cancer are incidence. Expansion of current cancer registries, cured of their disease (Ellison, Pogany, and Mery 2007; improvement in diagnosis and registration, and novel Pui and others 2012; Smith and others 2010). Although methodologies are needed to establish the true pediatric cure rates in LMICs are much lower, there are many cancer burden (Ferlay and others 2010; Magrath and examples of successful treatment with less intensive others 2013). The International Agency for Research regimens that can nevertheless cure a significant por- on Cancer is assembling an updated volume of the tion of patients in LMICs. Burkitt lymphoma (BL), the International Incidence of Childhood Cancer, drawn most common childhood malignancy in many parts of predominantly from registry data. Comparisons with Sub-Saharan Africa, is cured in 90 percent of cases in previous editions will allow an assessment of progress. HICs, using intensive regimens and intense and costly supportive care (Patte and others 2007; Woessmann and others 2005). However, up to 50 percent of Sub-Saharan WHY TREAT CHILDHOOD CANCER IN African children with BL are curable with only three to LMICs? six doses of single-agent cyclophosphamide and intrath- Epidemiologic Transition ecal therapy (Harif and others 2008). In most HICs, cancer represents the leading cause of non-accidental death in children older than age Spillover Effect from Pediatric to Adult Oncology one year (Ellison, Pogany, and Mery 2007; Siegel, Naishadham, and Jemal 2013). Although infection In societies in which cancer may be seen as a death accounted for 64 percent of global deaths in the first sentence, pediatric oncology offers the opportunity to five years of life in 2010 (Liu and others 2012), major demonstrate high cure rates in a manageable number shifts in the magnitude and causes of childhood mor- of patients through the establishment of a defined and tality have occurred in many LMICs, especially in feasible cancer infrastructure. Such success can serve MICs. In Brazil, mortality in children younger than as powerful encouragement to governments and policy age five years decreased from 129 per 1,000 live births makers to create and expand programs targeting adults in 1970 to 59 per 1,000 in 1990, and to 19 per 1,000 in with cancer, in addition to ensuring that children with 2010; cancer now leads the causes of non-accidental cancer are not neglected in the face of far greater num- death in that country. Worldwide, 106 countries wit- bers of adult patients. nessed accelerated declines in childhood mortality from 1990 to 2011; about 80 percent of the decline was from infectious disease control (Lozano and PLATFORMS FOR CHILDHOOD CANCER others 2011). Consequently, noncommunicable causes TREATMENT DELIVERY represent a greater proportion than before (Liu and others 2012; Patton and others 2012). Indeed, while Dedicated Centers 3.2 percent of deaths among children ages 5–14 years Childhood cancer treatment requires specialized in LICs are estimated to be caused by cancer, the diagnostic and therapeutic capabilities, as well as the equivalent figures for LICs and upper-middle-income ability to manage potential complications. Expensive, countries are 6.0 percent and 18.6 percent, respectively high-technology equipment is not required, however. (Magrath and others 2013). Although volume-outcome relationships have not been Treating Childhood Cancer in Low- and Middle-Income Countries 123 convincingly demonstrated in pediatric oncology, the It is worth highlighting the importance of stable drug dominant paradigm is to manage care through a lim- supplies. Shortages of essential chemotherapy agents ited number of treatment centers in which resources have been shown to impact pediatric survival, even in and expertise are concentrated. Satellite centers can HICs (Metzger, Billett, and Link 2012). In LMICs, the deliver some treatment, decreasing the burden on impact of inconsistent chemotherapy availability is likely families, providing rapid management of complications to be even greater. and, in LMICs, decreasing abandonment of treatment In many LMICs, childhood cancer services are (Metzger and others 2003; Pediatric Oncology Group of delivered through cancer hospitals serving primarily Ontario 2012). adult populations. In these instances, appropriately sized Tables 7.1 and 7.2 list the personnel and infra- pediatric equipment and specific pediatric expertise are structural requirements for an ideal LMIC center still required. Even when these requirements are met, delivering pediatric cancer care; however, many insti- the neglect of pediatric populations in the face of large tutions in LMICs deliver curative treatment in the volumes of adult patients may still adversely impact the absence of many of these elements (Harif and others quality of childhood cancer care. 2008; Madani and others 2006; Pedrosa and others 2000). Such treatment must be adapted to local capa- bilities. For example, centers without an intensive Twinning Programs care unit or ventilators will not be able to deliver as “Twinning” is currently the most effective model for sus- intensive chemotherapy as ones with these resources, tained improvement in childhood cancer care in LMICs. but they will nonetheless be able to cure a portion Twinning programs foster interactions between hospitals of children. in LMICs and established cancer treatment centers, Table 7.1 Examples of Essential Personnel for Ideal Pediatric Cancer Care in Low- and Middle-Income Countriesa Personnel Requirements Medical doctors Individuals who have received training or have experience managing pediatric oncology patients are essential to lead the unit and coordinate all other personnel needed to achieve cure. In many centers, pediatricians, adult hematologists, adult oncologists, or surgeons with some degree of extra training or experience may fill this role. Training and fellowship programs now exist in several LMICs. Surgeons Surgery is necessary for the diagnosis and treatment of many pediatric malignancies, such as Wilms tumor. However, some cancers are curable without surgical intervention. Radiation oncologists Radiation therapy is used for a variety of pediatric malignancies in HICs, such as Hodgkin lymphoma, Wilms tumor, and sarcomas. However, in some cases, substituting additional chemotherapy or surgery can result in cure (Mauz-Korholz and others 2010; Nachman and others 2002). Pathologists Correct diagnosis is the foundation of cancer care, and a professional who has experience in the diagnosis of pediatric malignancies and who is connected with disease-specific pathology experts for difficult cases is ideal. Nursing Strong nursing support with additional training in safe chemotherapy administration is needed. Expertise in the recognition and management of complications related to either the malignancy or treatment is desirable. An open line of communication between nursing and medical colleagues is crucial. Models for training nurses in pediatric oncology in LMICs have been described (Day and others 2011; Day and others 2012). Pharmacists Dedicated pharmacists are needed to prepare chemotherapy and to facilitate the safe preparation, handling, and disposal of chemotherapeutic medications. Social workers Addressing the emotional, social, financial, and spiritual needs of children and families facilitates adherence to treatment, improves quality of life, and reduces the risk of abandonment. Dieticians or nutritionists Nutritional support is particularly important in LMICs where malnutrition at diagnosis or during treatment is prevalent (Israels and others 2009; Sala and others 2005; Viana and others 2001). Note: HICs = high-income countries; LMICs = low- and middle-income countries. a. This list is not meant to be exhaustive. Other personnel, including infectious disease specialists and intensive care physicians, play crucial roles but may not be available in many resource-constrained settings. All the elements listed are desirable, but a proportion of children will still be cured in their absence. 124 Cancer Table 7.2 Infrastructure Needed to Deliver Ideal Pediatric Cancer Care in Low- and Middle-Income Countriesa Infrastructure Requirements Inpatient and outpatient beds Sufficient inpatient and outpatient beds are required, preferably designated for pediatric oncology patients. A hand hygiene program, isolation capabilities, and other infection control methods are desirable. Laboratory and pathology Basic hematologic, biochemical, microbiologic, and pathologic laboratory services capable of timely turnaround services are desirable. Although advanced diagnostic modalities, such as flow cytometry and cytogenetics, are available in HICs, their absence does not preclude the establishment of a pediatric oncology center (Hunger, Sung, and Howard 2009). Diagnostic imaging Basic imaging capabilities are necessary. While advanced modalities—such as computerized tomography and magnetic resonance imaging—are ideal, basic modalities, such as plain radiographs and ultrasonography are sufficient to begin treating childhood cancer (Madani and others 2006; Marjerrison and others 2012). Chemotherapy and supportive Reliable supplies of selected chemotherapeutic agents and supportive care medications, such as care medications antimicrobials, antiemetics, and analgesics, are crucial. The World Health Organization Model List of Essential Medications for Children provides a starting point for specific medications (WHO 2013). Blood product availability Treatment protocols may cause bone marrow suppression, necessitating the timely and reliable delivery of safe blood products. However, this is not the case for all chemotherapies; treatment for several malignancies requires minimal transfusion support. Psychosocial support Abandonment of therapy is a significant cause of treatment failure in many LMICs. The provision of financial support in case of inability to pay for medical care, and of transport and accommodation when necessary, decreases the risk of abandonment and must be considered an essential part of oncology care in LMICs. Surgical facilities Surgery is necessary for diagnosis and treatment of many pediatric malignancies, for example, Wilms tumor. Many cancers are curable without surgical intervention. Radiation facilities Radiation therapy is used for a variety of pediatric malignancies in HICs, for example, Hodgkin lymphoma, Wilms tumor, and sarcomas. However, in some cases, substituting additional chemotherapy or surgery can result in cure (Mauz-Korholz and others 2010; Nachman and others 2002). Note: HICs = high-income countries; LMICs = low- and middle-income countries. a. This list is not meant to be exhaustive. While all of the elements listed are desirable, a proportion of children can still be cured in their absence. with the goal of improving survival rates among children adapting treatment regimens, improving supportive with cancer (Ribeiro and Pui 2005). Twinning allows a care, and reducing treatment abandonment. bidirectional exchange and combines disease-specific Despite the success of the twinning paradigm in multidisciplinary expertise with local knowledge and improving individual pediatric cancer units, improve- capabilities. ments must be translated into national childhood cancer Twinning programs can involve the flow of financial strategies to have the greatest impact. Most LMICs lack resources, although the presence of committed indi- policies to ensure good pediatric oncology care, and viduals on both sides predicts success better than the many have no national cancer plan, let alone one tar- availability of funding. Interactive online tools such geting the unique needs of children. Notable exceptions as Cure4Kids (http://www.Cure4Kids.org) facilitate include Seguro Popular in Mexico, which includes an communication between participating centers (St. Jude accreditation process for hospitals treating children Children’s Research Hospital 2012). In some cases, with cancer, and reimbursement for care provided twinning programs have been associated with rapid by qualifying institutions. Since this program began, increases in cure rates (annex map 7A.1). The Pediatric abandonment of treatment has fallen from 52 percent Oncology in Developing Countries (PODC) commit- to 5 percent (Rivera-Luna and others 2012), although tee of the International Society of Pediatric Oncology access to care and the survival of treated patients varies (SIOP) has created a forum for interested people widely among accredited pediatric cancer units (Perez- from centers in HICs and LMICs to develop twinning Cuevas and others 2013). Current efforts in China to programs. Indeed, the 12 working groups of PODC are build comprehensive health insurance programs that exclusively dedicated to improving care for children cover childhood cancer treatment hold great promise with cancer in LMICs by fostering twinning programs, but are in their infancy. Treating Childhood Cancer in Low- and Middle-Income Countries 125 GENERAL PRINCIPLES OF TREATMENT supportive care to prevent and manage treatment compli- cations. The same high intensity chemotherapy delivered Importance of Locally Adapted Treatment Protocols at two centers, one with 24-hour availability of intensive Although not true for all cancers, increasing the inten- care and the other without, will result in higher TRM sity of treatment has increased cure rates (Matthay rates in the latter. In HICs, advances in supportive care and others 1999; Womer and others 2012; Woods and have allowed the delivery of ever higher intensity treat- others 1996). Different childhood cancers require dif- ments. Even in this context, however, the ideal balance ferent treatment intensities for maximum cure rates; has at times been difficult to find; intensifying treatment for example, the chemotherapy for Wilms tumor is far for AML initially resulted in high TRM rates in Europe less intense than for acute myeloid leukemia (AML). and North America, which later decreased as cancer One of the great achievements of pediatric oncology units developed the new level of supportive care required in recent decades is the refinement of risk stratification (Creutzig and others 2004; Lange and others 2008). systems, allowing for an assessment of the aggressive- In many LMIC centers, supportive care capabilities ness of a particular child’s cancer and for treatment lag behind those in HICs. Transposing treatment proto- intensity to be matched to disease risk, thereby reducing cols designed for HIC levels of supportive care to LMIC both undertreatment and overtreatment (Crawford, centers is therefore almost certain to cause high levels of MacDonald, and Packer 2007; Maris 2010; Metzger and TRM (Gupta and others 2009; Gupta and others 2011). Dome 2005; Pui, Robison, and Look 2008). The possibility of doing more harm than good is sig- Avoiding overtreatment is crucial in LMICs, since nificant. An important example is described in box 7.1, it carries with it an increased risk of treatment-related where decreasing treatment intensity actually led to mortality (TRM), defined as death from complications of higher cure rates. Questions to ask when trying to deter- treatment, as opposed to the disease itself (Creutzig and mine the supportive care capabilities of an individual others 2004; Ethier and others 2011; Gupta and others institution include the following: 2009; Gupta and others 2011; Prucker and others 2009). At some point, any benefit in disease control of intensify- • Are 24-hour nursing and medical coverage available ing treatment will be outweighed by an increase in TRM. for inpatients? Finding the balance point for each malignancy at each • How quickly can antibiotics be ordered, received, and pediatric cancer center is key to optimizing therapy and given to patients when urgent treatment is necessary? curing the maximum number of children possible. • How quickly can a blood transfusion be ordered, This ideal balance point depends on the malignancy in received, and given to patients when urgent treatment question, as well as a particular center’s ability to provide is necessary? Box 7.1 Acute Promyelocytic Leukemia: Cost and Treatment Intensity Acute promyelocytic leukemia (APL) is a subtype of acute myeloid leukemia, with cure rates of about 80 percent in high-income countries. In Guangzhou, China, Luo and others (2009b) treated 30 children with APL between 1999 and 2008. Before September 2004, children were treated on an intensive protocol including high-dose cytarabine and high cumulative doses of anthracycline. After September 2004, children were treated with a far less intensive protocol with fewer chemotherapy cycles, lower anthracycline doses, and no cytarabine. The total cost of therapy was lower, decreasing the financial burden on parents. With the first protocol, of 16 children, six abandoned therapy and seven developed bacterial sepsis, one of whom died. With the less intensive protocol, none of the 14 children studied abandoned therapy, and there was only one episode of sepsis, with no resultant infectious deaths. The three-year, event-free survival was 37.5 percent with the more intense protocol, and 79.6 percent with the less intensive treatment. Although the number of patients is small, this example illustrates an important principle: increased intensity and cost of treatment can do more harm than good. Sources: Ortega and others 2005; Testi and others 2005; Luo and others 2009b. 126 Cancer • Are basic radiographic, microbiologic, and hemato- others 2009a; Mostert and others 2006). Interestingly, logic diagnostic tests available? even in the context of a treatment program in which che- • Is intensive care, including ventilator and inotropic motherapy, supportive care, lodging, and transport were support, available? provided at no cost to families, families of low socioeco- • What is the prevalence of malnutrition in the popu- nomic status were still at higher risk of abandonment lation? What programs are available in the pediatric (Bonilla and others 2009). Various efforts in LMICs cancer unit to address malnutrition? have decreased abandonment rates, including providing • Are families able to reach medical attention quickly in financial support, adapting treatment protocols based on case of a treatment complication? a family’s financial resources, providing parental educa- • Where do outpatients go when emergencies develop tion, and establishing a social work program (box 7.2) after hours? Who treats them there? Are pediatric oncol- (Bonilla and others 2009; Howard and others 2004; Luo ogy professionals involved in their care after hours? and others 2008; Mostert and others 2010). Thus, just as some level of basic supportive care Further consequences stem from the principle that capacity is necessary to treat children with cancer, basic increased intensity and cost of treatment can do more educational and aid programs aimed at preventing harm than good. Many diagnostic modalities are utilized abandonment are also imperative. to classify the extent of disease, including stage and risk group, of particular patients. For example, in ALL, the most common childhood cancer in many countries, flow Outcome Evaluation cytometry and cytogenetics help to identify high-risk sub- Although it is possible to theorize as to what pro- groups, such as T-cell or hypodiploid ALL (Pui, Robison, tocol modifications are best suited to a particular and Look 2008). Children with these high-risk subgroups LMIC institution, there is no substitute for the actual are treated with higher intensity protocols. In a center in which higher intensity therapy leads to unacceptable TRM rates, spending limited resources on developing these diagnostic modalities is difficult to justify. However, mak- ing a correct diagnosis (such as distinguishing between Box 7.2 myeloid and lymphoblastic leukemia) is often life-saving and cost-effective (Howard and others 2005). Examples of Successful Efforts to Decrease the Abandonment of Therapy in Children Abandonment of Therapy with Cancer Abandonment is defined as the “failure to start or com- • In Guatemala City, Guatemala, through the establishment plete [potentially] curative treatment” (Mostert and others of a psychosocial team including both social workers and 2011, 719). The phenomenon of abandonment, virtually psychologists whose aim is to support families throughout unknown in HICs, is a significant problem in LMICs; in the cancer experience, abandonment has decreased from some contexts, it constitutes the most common cause of 42 to 2 percent (F. Antillon, personal communication). treatment failure (Arora, Eden, and Pizer 2007). The impor- • In Recife, Brazil, through the provision of lodging, tance of this issue led SIOP to establish the Abandonment social work, transportation, and food subsidies, and the of Treatment Working Group (Mostert and others 2011). establishment of a parent group, a fundraising foun- A systematic review of pediatric acute lymphoblastic leu- dation, and a patient tracking system, abandonment kemia in LMICs found that abandonment rates ranged among children with acute lymphoblastic leukemia from 3 percent to an astonishing 74 percent (Gupta and (ALL) decreased from 16 to 1 percent from 1980 to 2002 others 2013). None of 83 published reports of abandon- (Howard and others 2004). ment were from LICs, so the review likely underestimates • In Yogyakarta, Indonesia, after the introduction of a the global incidence of abandonment. parental education program, upfront treatment refusal Many reasons for abandonment have been cited, for children with ALL decreased from 14 to 2 percent including a lack of financial resources, poor disease among poor parents (Mostert and others 2010). comprehension, cultural factors, belief in alternative medicines, fear of treatment toxicity, inadequate care on the part of health care workers, and decreased aware- Sources: F. Antillon, personal communication; Howard and others 2004; Mostert and others 2010. ness of aid programs (Bonilla and others 2009; Howard and others 2004; Kulkarni and Marwaha 2010; Luo and Treating Childhood Cancer in Low- and Middle-Income Countries 127 monitoring of treatment outcomes. Collection of basic more complex chemotherapy (for example, ALL) and data on patient demographics, disease characteristics, multimodality treatment (for example, Wilms tumor) and treatment outcomes, including cause of death, and could eventually advance to treatment of sarcomas, allows for evaluation of a specific treatment proto- brain tumors, and diseases that require high levels of col, as well as the design of future interventions. For supportive care (for example, AML, high-risk neurob- example, it is not enough to know that children with lastoma). Table 7.3 lists characteristics of 13 of the most ALL in an individual center have a mortality rate of common childhood cancers; this information should be 50 percent, without evaluating the causes of death. considered before deciding which malignancies to treat If the predominant cause of death was TRM, then and which resources to develop in a specific setting. For appropriate interventions would include the strength- each type of cancer, the elements required for successful ening of supportive care, perhaps accompanied by treatment may differ based on stage and risk group. For de-intensification of treatment. However, if the pre- example, while intensive chemotherapy, surgery, radia- dominant cause was relapse, increasing treatment tion, and autologous stem cell transplantation cure only intensity may be appropriate. Outcome monitoring a minority of advanced-stage neuroblastoma in older allows for the gradual evolution of treatment strat- children, surgery alone may cure localized and biologi- egies in a safe and efficient manner and cure of the cally favorable neuroblastoma in a younger child. maximum number of children possible at each stage The subsequent sections discuss five childhood (Hunger, Sung, and Howard 2009). cancers often targeted by LMIC centers because of Health care workers in many LMICs lack the time their high potential cure rates with relatively low to collect, review, and analyze outcome information. intensity treatment regimens. In addition, these five In most settings, a dedicated data manager with suf- cancers collectively account for a significant portion ficient training, infrastructure, and support is needed of pediatric malignancies: ALL, HL, Wilms tumor, BL, to ensure accurate and timely data entry. It is worth and retinoblastoma. Each section outlines aspects of emphasizing that the collection and analysis of these diagnosis and treatment and how both may be adapted data are neither academic research nor a luxury. Indeed, to local resource constraints. outcome monitoring is essential to improving the care and outcomes at any pediatric cancer center, whether Acute Lymphoblastic Leukemia in LMICs or HICs. However, quality improvement efforts in LMICs often mean the difference between life Stephen P. Hunger1 and Federico G. Antillon2 1 and death, whereas those in HICs affect more subtle Children’s Hospital Colorado and the Department of outcomes. Pediatrics, University of Colorado School of Medicine, Aurora, CO, United States. 2 Unidad Nacional de Oncología Pediátrica, Guatemala City, Guatemala. TREATMENT OF SPECIFIC CANCERS The ideal malignancy targeted for treatment in LMICs ALL, a cancer of white blood cells (WBC), is the most would be one that accounts for a significant proportion common childhood cancer, accounting for 25 percent of the local cancer burden and that is curable with either of cancers among those younger than 15 years of age, simple surgery or short-course chemotherapy alone. The and 20 percent of those that occur before 20 years treatment of this ideal target would involve minimal acute of age (Ries and others 1999). ALL is universally toxicity and few chronic late effects—survivorship issues fatal without effective therapy. In North America and specific to LMIC children are unstudied. Of course, no Western Europe, five-year survival rates have steadily single malignancy perfectly fits this profile. Which malig- improved, from below 10 percent in the 1960s to over nancies should be treated in a particular LMIC center 90 percent today (Hunger and others 2012; Moricke depends on the local incidence, the available treatment and others 2010; Pui and others 2009; Silverman and modalities, the institutional level of supportive care possi- others 2010). However, most children who develop ALL ble, and theoretically attainable cure rates. do not reside in these countries. China and India are A center that is only beginning to treat childhood predicted to have four to five times as many pediatric cancer could start with malignancies for which cure is ALL cases as the United States; Indonesia, Nigeria, and possible with relatively simple and low-intensity che- Pakistan are predicted to have about the same number motherapy, such as BL or Hodgkin lymphoma (HL). of cases as the United States (online annex table 7A.1). A center that has achieved significant cure rates in Thus, it is critical to consider how pediatric ALL can these cancers could then address malignancies requiring be cured in countries that have very different income 128 Cancer Table 7.3 Characteristics of Childhood Cancers to Consider When Determining Which Malignancies Are Appropriate for Treatment in a Particular Resource-Constrained Setting Approximate Approximate treatment Supportive Late HIC cure ratea duration care level Chemotherapy Surgery Radiation effects/ Cancer (percent) (months) required necessary? necessary? necessary? disability ALL 90 24–40 ++ Yes No No + AML 60 5–7 ++++ Yes No No ++ Hodgkin lymphoma 90 2–8 ++ Yes No Nob ++ Burkitt lymphoma 90 6–8 +++c Yes No No + Medulloblastoma 75 8–10 ++ Yes Yes Yes ++++ Neuroblastoma 65 8–10 +++ Yes Yes Yes +++ b Wilms tumor 90 4–8 + Yes Yes No ++ b Rhabdomyosarcoma 70 8–12 ++ Yes Yes No ++ Osteosarcoma 70 8–12 ++ Yes Yes No ++ b Ewing sarcoma 75 8–12 ++ Yes Yes No ++ d e f Retinoblastoma 95 0–3 + No Yes No ++ e Testicular cancer 90 0–3 + No Yes No − g Hepatoblastoma 85 4–6 + Yes Yes No + Note: The scale is from not very significant (−) to very significant (++++). ALL = acute lymphoblastic leukemia; AML = acute myeloid leukemia; HIC = high-income country. a. Unless otherwise specified, HIC cure rates are taken from Surveillance, Epidemiology, and End Results Program registry data (Smith and others 2010). b. Radiation is indicated in select cases. c . While HIC regimens for Burkitt lymphoma require significant supportive care, lower intensity regimens requiring minimal supportive care can also be used. d. Dimaras and others 2012. e. Chemotherapy is required for advanced cases, although localized cases may be cured without it. f . Local control methods, including cryotherapy and laser therapy, are often used instead of surgery in HICs, but these are unavailable in many low- and middle-income countries. g. Perilongo and others 2009; Zsiros and others 2010. structures and health care systems than those in North and cytogenetic or molecular genetic studies, to define America and Western Europe. sentinel abnormalities, many of which have important prognostic implications. However, these tests are often Diagnosis of ALL not available in LMICs. A major prognostic factor is Children with ALL are commonly brought to medical the rapidity of response to single-agent or multiagent attention for symptoms caused by ineffective produc- therapy, which can be measured in a simple and inex- tion of normal blood cells because of replacement of pensive manner by peripheral blood or bone marrow the bone marrow by leukemia, including pallor, bleed- morphology, or in a complicated and expensive man- ing, fever, infections, and bone pain. They may also ner using advanced flow cytometry and/or molecular have leukemic involvement of other organs, includ- genetic techniques. ing liver, spleen, mediastinum, central nervous system, and testicles. General Concepts of Pediatric ALL Treatment ALL is diagnosed based on review of peripheral blood Contemporary treatment for ALL consists of com- cell counts and a bone marrow aspirate/biopsy, tests that plex combination chemotherapy regimens that can be performed at most medical facilities. Simple fac- last 2.5–3 years, with six to eight months of rela- tors predictive of outcome include age (younger is better, tively intensive therapy, followed by 1.5–2 years of except for infants less than one year) and initial WBC low-intensity maintenance therapy, during which most count (lower is better). More sophisticated and often children can resume normal activities and attend school. very expensive diagnostic tests readily available in HICs Chemotherapy drugs included in these regimens have include immunophenotyping, to determine cell lineage, been widely available for decades; most are relatively Treating Childhood Cancer in Low- and Middle-Income Countries 129 inexpensive, with the exception of asparaginase prepara- collaborations with pediatric cancer programs in Monza tions, which are extremely expensive (Masera and others and Milan, Italy, and St. Jude Children’s Research 2004). Radiation therapy to the brain was a critical Hospital in the United States. AHOPCA now conducts component of early effective ALL regimens, but the use its own non-randomized clinical trials. In Guatemala, of cranial irradiation has been greatly reduced in most ALL survival rates now range from 50 percent (high-risk) contemporary HIC regimens (Pui and Howard 2008). to 90 percent (low-risk) for different patient subgroups Although treatment of pediatric ALL is associated (Antillon-Klussmann and others 2010). This strategy with significant risk of short- and long-term side effects, is possible in countries with reasonably well-developed most children cured of ALL will lead healthy and pro- health care systems, with infant mortality rates less than ductive lives. Cure rates are much lower for children with 40–50 per 1,000 live births serving as a good surrogate ALL that relapses, with the chance of cure related to site marker (online annex, table 7A.1). of relapse, ALL genetic features, and time between initial However, high rates of ALL TRM can be a major diagnosis and relapse (Nguyen and others 2008). problem (Gupta and others 2011). Regimens that are Because most children with ALL live in LMICs, delivered safely with TRM rates less than 5 percent efforts have been made to improve treatment available in North America and Western Europe can be associ- in those countries through partnerships with centers in ated with TRM rates 5–10 times higher in LMICs; the HICs (Masera and others 1998). This twinning has led problem is much worse in countries with less devel- to major improvements in ALL survival in LMICs, often oped health care systems, reflected by infant mortality through adoption of intact or modified HIC treatment rates more than 50 per 1,000 live births. High rates regimens (Howard and others 2004; Veerman, Sutaryo, of TRM severely compromise cure rates and can be a and Sumadiono 2005). Critical to these successes has major impediment to program development in LMICs. been the transfer of knowledge regarding treatment Treatment of relapsed ALL has a very low chance of regimens, supportive care, and emotional and psycho- success in LMICs. social support. Abandonment of care is a major issue One way to address these problems is through the use in LMICs because of economic and social pressures on of graduated intensity regimens, whereby centers first parents and cultural beliefs that a child has been healed implement less intensive regimens similar to those used (Sitaresmi and others 2010; Wang and others 2011). in North America and Western Europe in the 1970s and Innovative programs have been developed to support 1980s, and increase treatment intensity only when they patients and families and greatly reduce abandonment; establish these therapies to be safe and effective in their a Guatemalan program reduced abandonment rates local settings (Hunger, Sung, and Howard 2009). This from 42 percent to less than 2 percent (unpublished strategy is attractive because it starts with regimens that observations, Rivas and Antillon). Successful imple- are less costly, less toxic, and do not require sophisticated mentation and improvement of therapies also requires diagnostic tests, but that can cure about 50 percent of close tracking of patient characteristics and outcomes, children with ALL if TRM can be kept low and abandon- necessitating access to databases and data management ment can be minimized. personnel (Ayoub and others 2007). An example from the pediatric cancer program in Santo Domingo, the Dominican Republic, shows Specifics of Pediatric ALL Treatment the potential benefit of this strategy. In 2005–07, The development of large cooperative treatment a relatively intensive HIC-type treatment regimen groups that conduct clinical trials, which often include was followed for 91 children with ALL; however, it 70 percent or more of children with ALL in a given was associated with excessive TRM. Following this country (Hunger and others 2012), has been critical to experience, a less intensive regimen was used to treat improvements in survival for pediatric ALL in HICs. 101 patients diagnosed in 2008–10. The less intensive This development has resulted in near-universal access treatment improved 24-month overall survival from to effective treatments in most HICs (limited in some 40 to 70 percent, accompanied by a decrease in TRM cases because of country-specific differences in health from 29 of 91 cases in the early period to 8 of 101 in care financing) and the widespread availability of knowl- the later period (Hunger and others 2011). edge about the specifics of effective treatment regimens. Twinning has provided outstanding examples of very Costs of Pediatric ALL Treatment effective transfer of knowledge and adoption of con- Pediatric ALL treatment in North America and Western temporary treatment regimens in LMICs, such as the Europe is widely recognized to be very expensive and Central American Association of Pediatric Hematology highly cost effective. A report from the Dutch Childhood Oncology (AHOPCA), largely developed through Oncology Group showed mean total costs for treating 130 Cancer pediatric ALL to be US$115,858–US$163,350 per case, ages in LICs, and it often occurs before adolescence. with highly favorable costs per life year saved of Furthermore, in LMICs, HL is most often Epstein-Barr US$1,962–US$2,655 (van Litsenburg and others 2011). virus-positive and of mixed cellular histology (Siddiqui However, effective treatments can be implemented for and others 2006). Disease stage and bulk, as well as much lower costs. Luo and others reported in 2008 that the presence of “B-symptoms” (fevers, drenching night a reduced intensity, low-cost protocol that obtained a sweats, or greater than 10 percent weight loss in the past four-year event-free survival rate of 72.8 percent could six months) are established prognostic factors. Other be implemented in Guangzhou, China, for a total hospi- potential prognostic factors include the erythrocyte tal cost of US$4,300 per case; the range is from US$3,100 sedimentation rate and low hemoglobin and albumin to US$6,800 (Luo and others 2008). More intensive levels, although these may be less reliable indicators in regimens obtained slightly better results and could be children suffering from chronic malnutrition or para- implemented for US$9,900–US$12,500, similar to the sitic infections. average cost of US$11,000 per patient reported from Shanghai, China (Liu and others 2009). Diagnosis of HL An excisional lymph node biopsy is recommended, Summary as fine-needle aspirates are often inadequate for diag- ALL is the most common pediatric cancer. Five-year nosis. This is, in fact, the only surgical procedure survival rates exceed 90 percent in HICs. Through routinely required in the treatment of HL. Pathology twinning, centers in LMICs with infant mortality rates is basic; the diagnosis can be confirmed with a sim- less than 40–50 per 1,000 live births have attained ple hematoxylin and eosin stain without the need for cure rates of about 70 percent. Outcomes for relapsed immunohistochemistry. ALL are much worse, stressing the need for effective therapy at initial diagnosis. Graduated intensity regi- Staging and Treatment Options mens have the promise to decrease TRM and improve In HICs, the ideal initial evaluation of children for HL survival, and they may be particularly effective in includes computed tomography of the neck, chest, LMICs with infant mortality rates greater than 50 per abdomen, and pelvis, accompanied by FDG-positron 1,000 live births. emission tomography. Staging and the presence of B-symptoms allow risk stratification with therapy tailored according to risk of relapse and adapted based Hodgkin Lymphoma on disease response after two cycles of chemotherapy. Monika L. Metzger1 Risk-stratified, response-adapted therapy offers the 1 Division of Leukemia/Lymphoma, St. Jude Children’s potential to maximize cure and minimize toxicity Research Hospital, Memphis, TN, United States. (Hodgson, Hudson, and Constine 2007). In LMICs with limited availability of diagnostic In HICs, over 80 percent of children with HL survive imaging, a thorough physical examination for determi- long-term. In LMICs, survival has been lower because nation of all pathologic peripheral adenopathy, chest of lack of adequate staging, drug shortages, inadequate radiograph for extent of mediastinal involvement, and access to radiotherapy, delays in therapy, and social hard- ultrasonography for intra-abdominal adenopathy can ship leading to abandonment of therapy. Most children be sufficient for staging. Bone marrow biopsy is not with HL in LMICs present to medical attention with recommended for most patients, since it is expensive, advanced-stage disease and a long history of symptoms. painful, and rarely affects risk classification or therapy Despite these obstacles, many LMIC patients can still be (Hines-Thomas and others 2010). In some cases, a pos- cured with basic chemotherapy, with or without consol- itive bone marrow biopsy may actually harm the patient idative radiotherapy. HL is curable, diagnosable without by leading to the false perception that bone marrow expensive technology, and constitutes an important por- involvement is incurable or that consolidative radiation tion of children with cancer. therapy is not indicated. In cases of limited staging evaluation, the treat- Epidemiology and Prognostic Factors ment approach must account for incomplete ascer- Childhood HL rarely presents before five years of age tainment of affected areas. Accordingly, more weight in HICs; however, in LMICs it can be seen in children must be placed on effective chemotherapy and less on as young as age one year. In HICs, HL has a bimodal local control with radiotherapy, which would not be age distribution in early adulthood and after the age of applied to disease sites undetected by incomplete stag- 50 years. The age distribution is shifted toward younger ing evaluations. Furthermore, radiation therapy is often Treating Childhood Cancer in Low- and Middle-Income Countries 131 unavailable, inconsistently available, or too toxic when Costs of HL Treatment given by radiation oncologists without pediatric exper- The bulk of the cost of HL therapy is due to pathologic tise. Risk stratification in many LMICs should also be evaluation, radiation therapy, and diagnostic imaging broader, similar to early HIC chemotherapy-only trials. studies; chemotherapy and supportive care constitute a Table 7.4 provides examples of chemotherapy-only and far smaller portion. In a study evaluating the cost of ther- combined modality treatment regimens used success- apy in Sub-Saharan Africa for a child with stage II dis- fully in LMICs. ease and followed for two years, the total cost was more During HL treatment, the minimum necessary sup- than US$6,500 in a continent where the annual gross portive care consists of antibiotics and antiemetics, domestic product (GDP) per inhabitant is usually less blood products are rarely needed, and therapy can be than US$2,000 (Stefan and Stones 2009). However, these administered in the outpatient setting without the need costs can be significantly reduced by carefully choosing for growth factors. the minimal necessary diagnostic imaging techniques Table 7.4 Treatment Results of Pediatric Hodgkin Lymphoma Trials in Low- and Middle-Income Countries Outcome % (years) Event-free Disease-free Overall Chemotherapy Stagea Number of patients survival survival survival Chemotherapy-only regimens Grupo Argentino de Tratamiento de Leucemia Agudab CVPP x 3 IA, IIA 10 86 (7) — — CVPP x 6 IB, IIB 16 87 (7) — — c Nicaragua COPP x 6 I, IIA 14 100 (3) — 100 (3) COPP-ABV x 8–10 IIB, III, IV 34 75 (3) — — Chennai, Indiad COPP/ABV x 6 I–IIA 10 89 (5) — — COPP/ABV x 6 IIB–IVB 43 90 (5) — — e New Delhi, India COPP x 6 All stages 34 — 80 (5) — Ugandaf MOPP x 6 I–IIIA 38 — 75 (5) — IIIB–IV 10 — 60 (5) — Combined modality trials New Delhi, Indiag 4 ABVD + 25–40 Gy IFRT I–IIA 79 — 91 (5) — 6–8 ABVD + 25–40 Gy to bulky disease sites IIB, III, IV 183 — 73 (5) — Note: ABVD = doxorubicin (Adriamycin), bleomycin, vinblastine, dacarbazine; COPP = cyclophosphamide, vincristine (Oncovin), procarbazine, prednisone; CVPP = cyclophosphamide, vincristine, procarbazine, prednisone; IFRT = involved-field radiation therapy; MOPP = mechlorethamine (Mustargen), vincristine (Oncovin), procarbazine, and prednisone; — = no information available. a. Stage I represents involvement of a single lymph node region or extralymphatic site. Stage II represents involvement of two or more lymph nodes on the same side of the diaphragm. Stage III represents involvement of lymph node regions on both sides of the diaphragm. Stage IV represents involvement of extralymphatic organs (for example, lung). B represents the presence of B symptoms (fever, night sweats, weight loss), while A represents the absence of B symptoms. b. Sackmann-Muriel and others 1997. c. Baez and others 1997. d. Sripada and others 1995. e. Chandra and others 2008. f. Olweny and others 1978. g. Ganesan and others 2011. 132 Cancer required for staging and chemotherapy regimens that Both strategies result in similar long-term survival for will permit the omission of radiotherapy. The most HIC patients (Graf, Tournade, and de Kraker 2000; important cost to avoid is that of relapse. Green 2004). Preoperative chemotherapy, however, reduces surgical complications, such as tumor rupture, and downstages the tumor at surgery, thereby allowing Wilms Tumor for lower intensity, postoperative chemotherapy and Trijn Israels1 reducing the need for radiotherapy. This is a sensible 1 Department of Pediatric Hematology/Oncology, VU strategy for many LMIC patients, who often present University, Amsterdam, The Netherlands. with large tumors in settings where supportive care is limited and radiotherapy may not be available. Wilms tumor is relatively common, accounting for Survival rates in LMICs are lower than in HICs, rang- 5–7 percent of all childhood cancers (Stiller and Parkin ing from 11 percent to 81 percent (Abuidris and others 1996). In many settings, Wilms tumor is the most com- 2008; Israels 2012; Israels and others 2012; Moreira and mon malignant abdominal tumor. As treatment pro- others 2012; Wilde and others 2010). Known challenges grams for pediatric oncology are developed, Wilms tumor are late presentation with advanced disease, malnutrition, should be one of the first tumors targeted because of its abandonment of treatment, and poor facilities for spe- frequency and curability. Treatment also requires the cific cancer treatment and supportive care (Abuidris and development of multidisciplinary capacities that may others 2008; Harif and others 2005; Moreira and others benefit other children and programs across the hospital. 2012). Capacity building, earlier presentation, a multidis- Great progress has been made in the treatment ciplinary approach, social support, improved supportive of children with Wilms tumor in recent decades. care, and treatment adapted to local circumstances are key The survival rates in HICs now exceed 85 percent. to improving results (Hadley 2010; Hadley, Rouma, and Multidisciplinary treatment combines surgery and Saad-Eldin 2012; Israels and others 2012). chemotherapy, with radiotherapy in a selected group of patients (Graf, Tournade, and de Kraker 2000; Treatment Settings Green 2004). Two treatment strategies have been The facilities and resources available for the care of chil- used for Wilms tumor worldwide. The first operates dren with Wilms tumor vary among centers, but they on tumors upfront, as practiced by the Children’s can be defined using the following settings (table 7.5): Oncology Group in North America, followed by chemotherapy; the second starts with preopera- • Setting 1 is one in which the minimal requirements tive chemotherapy, as practiced in Europe (SIOP). for treatment with curative intent are available. Table 7.5 Classification of Different Settings Providing Care for Children with Wilms Tumora Setting Medical facilities Specialists Drugs Supportive care Diagnostic facilities 0 Pain medication Physical exam 1. Minimal Pediatric ward Surgeon Vincristine Antibiotics Full blood count requirements for (Pediatrician) Nurse Actinomycin Whole blood Chest x-ray curative intent (Doxorubicin) Morphine Ultrasonography Social support 2. Intermediate Pediatric Pathologist Doxorubicin All blood products CT scan oncology ward Pediatric surgeon Cyclophosphamide Central venous access Radiotherapy Pediatric oncologist Etoposide Ifosfamide Pathology Radiation oncologist Carboplatin Multidisciplinary care Oncology nurse 3. State of the art Intensive care unit Pediatric pathologist Mechanical Special stains Pediatric radiation ventilation Immunohistochemistry oncologist Pharmacist Hemodialysis Cytogenetics (oncology) Intensivist Pressure support Note: CT = computed tomography. a. Facilities and resources mentioned are in addition to those associated with lower settings. In setting 2, mentioned facilities may or may not be available. Treating Childhood Cancer in Low- and Middle-Income Countries 133 • Setting 3 is one where all state-of-the-art facilities are for children is often unavailable in developing coun- available. tries. The recent National Wilms Tumor Study and • Setting 2 is in between. SIOP studies have shown that omitting or decreasing radiation therapy may not compromise cure rates, Diagnosis but these studies have not been done in children with The diagnosis of Wilms tumor can be made with rea- very advanced disease or large tumors. Studies from sonable certainty based on history, physical examination, Morocco and Nicaragua have demonstrated that cure and ultrasonography of the abdomen. The typical pre- can be achieved in some patients with advanced disease sentation of a child with Wilms tumor in low-income without radiotherapy (Baez and others 2002; Madani settings is that of a malnourished young child with a and others 2006). Higher cure rates in these populations large abdominal or flank mass, who is relatively well may, however, require radiotherapy. without acute pain or severe general malaise, but with Table 7.6 shows some elements of the therapy used hematuria and hypertension (Green 2004; Israels 2012). and the results from selected countries with limited Ultrasonography of the abdomen is extremely useful to resources. More detailed treatment recommendations confirm the diagnosis (De Campo 1986; Hartman and can be found in a recently published SIOP guideline Sanders 1982; Lowe and others 2000). An x-ray should developed for use in LMICs (Israels and others 2013). be done to detect chest metastases. In HICs, pathology is useful to confirm the diag- Cost of Wilms Treatment nosis and, in addition to stage, to help risk stratify To date, cost analyses related to the treatment of children children and determine postoperative chemotherapy. with Wilms tumor in LMICs have not been reported. In many LMICs, however, the availability of patholo- Although of relatively long duration (six months to gists with pediatric expertise is limited and pathology two years), treatment is of relatively low intensity and results often are available too late to effect clinical does not involve expensive chemotherapeutic agents. decision making. Other challenges include the appro- The costs of surgery are likely to be high. Social support priate processing of specimens and the availability of enabling parents to complete treatment is very likely to special stains and immunohistochemistry, although be cost-effective in LMICs. central pathology review or telepathology may be helpful (Vujanic and others 2009). Fortunately, a diag- nosis can often be made with some certainty based on Burkitt Lymphoma clinical findings and ultrasonography. Postoperative Mhamed Harif 1 chemotherapy can be based on surgical staging, only 1 Unité Hématologie et Oncologie Pédiatrique, Centre if needed. d’Oncologie et Hématologie, Centre Hospitalier A diagnostic biopsy before preoperative chemother- Mohammed VI, Marrakech, Morocco. apy is not standard practice in current SIOP Wilms protocols; it is only recommended in LMICs when there BL is a mature B-cell neoplasm that arises in lymphoid is serious doubt about the diagnosis (Vujanic and others tissue, commonly in the jaw or abdomen. Described first 2003). Such biopsies may result in bleeding, infection, or in 1957 by Denis Burkitt in Uganda, it remains the most tumor spillage with consequent upstaging. common pediatric cancer in malaria-endemic regions of Sub-Saharan Africa (Burkitt 1958; Lewis and others 2012). Treatment of Wilms Tumor BL invariably arises from chromosomal translocations in Preoperative chemotherapy should be used for chil- which an oncogene (c-myc) is juxtaposed with genes dren with Wilms tumor in LMICs, even in cases of encoding immunoglobulins. These translocations lead to small, seemingly easily resectable tumors (Lemerle and an overexpression of monoclonal surface immunoglobu- others 1983). Preoperative chemotherapy reduces surgi- lins in malignant cells, which is important for diagnosing cal complications, downstages the tumor, and allows for and distinguishing it from other lymphoid cancers. less intense postoperative chemotherapy and the poten- Although more than 90 percent of children with BL tial avoidance of radiotherapy (Graf, Tournade, and de in HICs can be cured, doing so requires timely, accurate Kraker 2000). Reliable and continuous access to the che- diagnosis and risk-directed treatment with high intensity motherapeutic drugs such as vincristine, actinomycin D, chemotherapy and well-developed supportive care (Patte and doxorubicin is essential. and others 2007). In many LMICs with limited support- Radiotherapy is used in patients with advanced- ive care, delivery of such therapy causes excessive toxic stage or unfavorable histology disease in centers with death; adapted regimens are necessary to cure as many advanced capabilities. Unfortunately, safe radiotherapy patients as possible (Hesseling, Israels, and others 2012). 134 Cancer Table 7.6 Reported Outcomes of Patients with Wilms Tumor Treated in Low- and Middle-Income Countries Numbers of Event-free survival Overall survival Country or region Settinga patients Chemotherapy Radiotherapy % (years) % (years) Sudanb 1 37 Generally postoperative, based No 11c — on NWTS-5, 37 percent received preoperative chemotherapy based on specific indications Malawid 1 84 Preoperative and postoperative, No 46 — modified from SIOP protocols Egypt, Arab Rep.e 2 62 Postoperative Yes 58 (4) 70 (4) f Central America 2 374 Postoperative, based on NWTS-4 Yes 59 (3) 74 (3) g Morocco 2 86 Preoperative and postoperative, based Yes 77 (5) 79 (5) on SIOP protocols South Africah 2 (−3) 188 Preoperative and postoperative, based Yes 75 (5) 81 (5) on SIOP protocols Turkeyi 2 (−3) 327 Preoperative and postoperative, based Yes 56 (10) 61 (10) on SIOP protocols Note: NWTS = National Wilms Tumor Study; SIOP = International Society of Pediatric Oncology; — = not available. a. Setting 1 is one in which the minimal requirements for treatment with curative intent are available. Setting 3 is one where all state-of-the-art facilities are available; Setting 2 is in between. b. Abuidris and others 2008. c. 89 percent of children in this study abandoned therapy prior to the completion of therapy. d. Israels and others 2012. e. Abd El-Aal, Habib, and Mishrif 2005. f. Ortiz and others 2012. g. Madani and others 2006. h. Davidson and others 2006. i. Kutluk and others 2006. Nevertheless, in even the most resource-constrained need for a biopsy. A fine-needle aspiration may be suf- environment, a simplified protocol for patients with BL ficient in patients whose clinical features are consistent can cure 50 percent (Hesseling and others 2009). Indeed, with BL (Razack and others 2011). When possible, the treatment of BL is likely to be highly cost-effective in all presence of mature B-cell markers (for example, CD20, settings (Bhakta and others 2012). immunoglobulin) and proliferative markers, such as Ki67, should be verified to differentiate BL from other Diagnosis small, round, blue cell tumors. Suspected BL is a medical emergency. BL is the fastest In cases in which the diagnosis is very likely and growing human malignancy, in some cases doubling pathologic confirmation will be delayed, chemother- its volume every 24 hours. The risks of tumor lysis syn- apy with cyclophosphamide, vincristine, and predni- drome (TLS)—a collection of metabolic derangements sone (COP) may be initiated empirically in potentially caused by the rapid turnover of malignant cells, disease life-threatening situations. These agents have low tox- progression, nutritional deterioration, and concomitant icity and are active for most lymphomas. The benefits infection—make diagnosis and therapy critical. Indeed, of prompt therapy initiation greatly outweigh the risks, any child from an endemic region presenting with as delayed therapy can lead to metabolic complications massive facial swelling or an abdominal mass requires such as TLS that can be rapidly fatal. immediate physical and laboratory evaluation for any of these complications. Staging Evaluations and Risk Stratification Biopsy of the suspected tumor is recommended Staging evaluations in HICs includes a detailed physical for diagnosis, but extensive surgery is contraindicated. examination to document peripheral adenopathy and The top priority must always be to make a diagnosis testicular involvement; computed tomography imag- in the fastest, least invasive way possible and to initiate ing of the neck, chest, abdomen, and pelvis to define therapy rapidly. In rare cases, BL cells may be seen in the all sites of adenopathy; and the evaluation of cere- peripheral blood, as in Burkitt leukemia, obviating the brospinal fluid, bone marrow aspirates, and biopsies. Treating Childhood Cancer in Low- and Middle-Income Countries 135 Ideally, lumbar punctures are delayed until a diagnosis short-duration therapy with vincristine, cyclophos- is made, so that intrathecal therapy can be administered phamide, doxorubicin, cytarabine, high-dose metho- at the time of the diagnostic puncture. The Murphy trexate, and intrathecal agents. Duration and intensity (St. Jude) staging system is most commonly used to vary according to risk group, but overall the therapy classify the extent of disease (Murphy 1978). In LMICs, produces a 90 percent cure rate (Patte and others 2007). a physical examination, chest radiograph, abdomi- However, this treatment approach in settings with lim- nal ultrasound, bone marrow aspiration, and lumbar ited supportive care exposes patients to high rates of puncture may provide sufficient staging information mortality and abandonment. (Marjerrison and others 2012). In LMICs and even in very poor settings, it has Disease risk assignment, and thus treatment intensity, been shown that at least 50 percent of children with BL is determined mainly by disease stage. Lactate dehydro- and up to 70 percent of children with localized stage genase level indicates disease activity and affects risk I or stage II disease can be cured with intravenous or group assignment in some, but not all, HIC protocols. oral cyclophosphamide in combination with intrathecal Inadequate response to treatment, defined in HICs as methotrexate (Harif and others 2008; Hesseling and less than 20 percent reduction in tumor size after the ini- others 2009; Traore and others 2011). Treatment with tial chemotherapy cycle or residual cancer after the first simplified regimens is feasible everywhere and should intense blocks of therapy, require intensification of ther- always be attempted (table 7.7). apy. Different definitions of inadequate response have In all cases, optimizing supportive care includes the been used in resource-constrained settings (Hesseling, prevention and treatment of TLS, infection, and vomiting. Israels, and others 2012). In either case, the dimensions TLS is the most common cause of early death in patients of all masses must be documented at presentation. with BL (Howard, Jones, and Pui 2011). Aggressively hydrating (three liters/m2/day), frequently monitoring Treatment urine output and serum chemistry values, and controlling The optimal treatment regimen for a particular patient uric acid with rasburicase (where available) or allopurinol depends on disease stage, as well as the environment can prevent acute kidney injury in most cases. Nutritional of care. Families with high socioeconomic status, good support and the prompt diagnosis and treatment of febrile transportation, and proximity to a pediatric cancer unit neutropenia and mucositis are the mainstays of support- with excellent infrastructure and supportive care can be ive care after the first week. Family education, written treated on an HIC regimen, including intensive- and care pathways, and creative nutritional supplements can Table 7.7 Selected Cohorts and Outcomes of Children with Burkitt Lymphoma Treated in Low- and Middle-Income Countries with Locally Adapted Protocols of Lower Intensity Study Countries Subgroups Number of patients Outcome (percent) Hesseling, Njume, Cameroon Stages I and II 18 EFS 94 and others 2012 Stage III, clinical remission, or 58 EFS 76 residual abdominal < 30 mL Stage IV, no clinical remission, or 45 EFS 40 residual abdominal mass > 30 mL Ngoma and Tanzania, Kenya, Nigeria All stages 326 EFS 52; OS 62a others 2012 Traore and Burkina Faso, Cameroon, Côte Stage I 19 EFS 44 others 2011 d’Ivoire, Madagascar, Mali, Senegal Stage II 23 EFS 49 Stage III 128 EFS 30 Stage IV 6 EFS 17 Note: EFS = event-free survival; OS = overall survival. a. No significant differences according to stage. 136 Cancer produce remarkable results, even in LMICs (Gavidia and (6–10 per million) in India, Sub-Saharan Africa, and others 2012; Israels and others 2009). among children of Native American descent in North Relapses are usually seen during the first six months America (Stiller and Parkin 1996). Whether this vari- and are rare after one year. Follow-up after one year ation is because of ethnic or socioeconomic factors is focuses on identifying late toxicities and assisting with unknown, although an environmental role has been reintegration into society. In LMICs, recruiting sur- suggested (de Camargo and others 2011; Fajardo- vivors to improve community awareness of pediatric Gutierrez and others 2007). An estimated 8,000 children cancer care and the possibility of cure is essential. develop retinoblastoma each year worldwide. This More detailed treatment recommendations can be burden is unequally distributed, with the majority found in a published SIOP guideline developed for use of children living in LMICs; these settings witness in LMICs (Hesseling, Israels, and others 2012). 90 percent of metastatic cases and virtually all cases of abandonment (Chantada and others 2011). Costs of BL Treatment As in other pediatric malignancies, data on the cost- Prevention and Early Detection effectiveness of treatment are rare. Given that a small As with virtually all childhood cancers, retinoblastoma number of doses of cyclophosphomide, a relatively inex- is not amenable to primary prevention. However, pensive drug, can cure a significant portion of children, identification of the hereditary forms and proper the treatment of BL is likely to be highly cost-effective. counseling of these patients and their families is key to A paper using data from Malawi demonstrated that limiting the incidence and burden of retinoblastoma using the World Health Organization (WHO) defini- in those relatives. tion, treatment costs under US$14,243 per case would be The successful management of retinoblastoma considered very cost-effective (Bhakta and others 2012). depends on the ability to detect the disease while it is still Actual estimated costs of treatment per case, at US$50, intraocular. Disease stage correlates with delay in diagno- were far lower, although this figure only accounted for sis; growth and invasion occur in sequence, with extension the costs of chemotherapy and is likely an underestimate. beyond the retina occurring only once the tumor has reached large intraocular dimensions. In HICs, retinoblas- toma typically presents while still intraocular; in LMICs, Retinoblastoma 60–90 percent of children present with extraocular tumor. Carlos Rodriguez-Galindo1 Poverty, limited health care access, poor education, and 1 Dana-Farber/Children’s Hospital Cancer Center, other aspects of low socioeconomic status are factors in Harvard Medical School, Boston, MA, United States. delayed diagnosis and underdiagnosis in LMICs. The true magnitude of the problem is difficult to ascertain, given Retinoblastoma is the most frequent neoplasm of the eye the paucity of population-based cancer registries. in childhood, representing 2.5–4 percent of all pediatric Conversely, retinoblastoma educational and public cancers and 11 percent of cancers in the first year of life. awareness campaigns have been shown to increase refer- Retinoblastoma presents in two distinct clinical forms. rals, decrease rates of advanced disease, and improve outcomes in LMICs (Leander and others 2007; Rodriguez- • Bilateral or multifocal (25 percent of cases) is heredi- Galindo and others 2008). Also critical is the ability of the tary, characterized by the presence of germline muta- first health care contact to identify the problem and make tions of the RB1 gene. Multifocal retinoblastoma may the appropriate referrals. A lack of knowledge on the be inherited from an affected survivor or be the result part of frontline health care workers has been shown to of a new germline mutation. be a significant barrier, highlighting the importance of • Unilateral retinoblastoma (75 percent) is almost always targeting educational initiatives to primary health care nonhereditary. Retinoblastoma is a cancer of the very providers (Leal-Leal and others 2011). young; two-thirds of the cases are diagnosed before age two years, and 90 percent of the cases are diagnosed Diagnosis and Staging before age five years (Ries and others 1999). The diagnosis of intraocular retinoblastoma does not require pathologic confirmation. An examination under Epidemiology anesthesia with a maximally dilated pupil and scleral The incidence of retinoblastoma in the United States and indentation is required to examine the entire retina. Europe is 2–5 per million children (approximately one Additional imaging studies, including bi-dimensional in 14,000–18,000 live births). However, the incidence ultrasound, computerized tomography, and magnetic is not consistent around the world, appearing higher resonance imaging, are desirable but not necessary to Treating Childhood Cancer in Low- and Middle-Income Countries 137 evaluate extraocular extension and to differentiate retin- Patients presenting with metastatic disease are not oblastoma from other causes of leukocoria. curable with standard therapies in any setting; patients The staging of retinoblastoma reflects the sequential without central nervous system spread may benefit from nature of its progression, beginning with extension into intensive chemotherapy and consolidation with high- the ocular coats (choroids and sclera) and optic nerve. dose chemotherapy and autologous stem cell rescue Loco-regional dissemination occurs by direct extension (Dunkel and others 2010; Rodriguez-Galindo and others into the orbital contents and pre-auricular lymph nodes. 2007b). In children in LMICs presenting with advanced Extraorbital disease manifests as both intracranial dis- extraocular retinoblastoma, measures to decrease suffer- semination and hematogenous metastases to bones, ing and improve quality of life may be most appropriate. bone marrow, and liver. Patients are accordingly staged Low-dose oral chemotherapy and radiation therapy may as having intraocular, orbital, or extraorbital disease result in temporary symptom control. (Chantada and others 2006). More detailed treatment recommendations can be For patients with intraocular retinoblastoma, dedi- found in a published SIOP guideline developed for use cated staging of the eye is performed to guide treatment in LMICs (Chantada and others 2013). modalities. This classification system is based on tumor size and location within the eye, as well as the extent of Costs of Retinoblastoma Treatment tumor seeding within the vitreous cavity and subretinal Little is known about the cost-effectiveness of retin- space, all of which must be documented on the initial oblastoma treatment, but measures targeting early exam under anesthesia. An evaluation for the presence diagnosis are likely key. Failures in public awareness of metastatic disease (bone scintigraphy, bone marrow and deficiencies in education among frontline health aspirates and biopsies, lumbar puncture) should be care providers represent major barriers in early diag- considered in patients presenting with intraocular retin- nosis and result in the high incidence of metastatic oblastoma with specific high-risk features (Rodriguez- disease and mortality rates in LMICs (Chantada and Galindo and others 2007a). others 2011). In LMICs, children with retinoblastoma are usually diagnosed with advanced intraocular dis- Treatment ease; by the time leukocoria is obvious, the tumor may The treatment goal is to save life and preserve vision; fill more than 50 percent of the globe, complicating accordingly, treatment is individualized according to the ocular salvage. Delayed diagnosis remains an issue in unilaterality or bilaterality of the disease, potential for HICs and LMICs, although with consequences on a dif- vision, and disease stage. In HICs, more than 90 percent ferent scale. As retinoblastoma is a cancer of the infant of children present with intraocular disease; clinical and and young child, initiatives targeting early recognition research programs aim to improve ocular salvage and during standard health supervision visits and immu- preserve vision. Although surgical removal of the eye nizations should facilitate diagnosis, decrease disease (enucleation) is commonly performed for patients with and treatment burdens and costs, and increase survival advanced intraocular unilateral disease, more conserva- (Rodriguez-Galindo 2011). tive approaches are followed for children with bilateral and early unilateral disease. Modalities include systemic or intra-arterial chemotherapy, as well as intensive focal COST-EFFECTIVENESS OF TREATING treatments, such as laser thermotherapy and cryother- apy (Gobin and others 2011; Rodriguez-Galindo and CHILDHOOD CANCER others 2007b). Financial objections are often raised to the treatment Orbital radiation therapy is used when the preceding of childhood cancer in resource-constrained settings; methods fail. For patients undergoing upfront enucle- policy makers and lay persons may assume that any ation, chemotherapy is only used in the presence of high- such treatment is prohibitively expensive. However, this risk features, which in HICs occurs in 20–25 percent of assumption is often unsupported. cases (Rodriguez-Galindo and others 2007b). In general, Indeed, preliminary evidence suggests that treating the outcome for children with retinoblastoma in HICs childhood cancer may be highly cost-effective. Standard is excellent, with survival rates in excess of 95 percent. WHO methodology defines cost-effectiveness as the Many of the modalities discussed require state-of-the-art ratio of the cost required to avert one disability-adjusted equipment and expertise that are unavailable in most life year to the annual per capita GDP of the area LMIC settings. Thus, for LMIC patients presenting with (WHO 2003). Ratios of 3:1 are considered cost-effective, orbital disease, the use of chemotherapy, enucleation, and while ratios of 1:1 are considered very cost-effective. radiation therapy may offer the best chances of cure. Bhakta and others found that the amount that 138 Cancer could be spent on a single case and still remain under Discussions of cost and cost-effectiveness in pediatric the very cost-effective threshold was US$257,000 oncology should consider three additional factors. for ALL in Brazil and US$14,243 for BL in Malawi (Bhakta and others 2012). Although treatments costing • First, adapted treatment regimens of lower intensity these theoretical thresholds may still be unachievable can cure a significant proportion of children, with for many LMICs, Bhakta and others also found that further increases in intensity delivering real, but these cancers could be treated for a fraction of the diminishing, gains. This observation suggests that threshold values: US$16,400 and less than US$50, in most LMICs, an initial modest commitment of respectively. Table 7.8 and figure 7.3 illustrate funds to childhood cancer will result in a dramatic cost-effective thresholds for several malignancies in increase in survival, although further improvements various countries and compare them with actual will require significant additional resources. costs, when available. These figures, however, do not • Second, traditional cost-effective models assume account for the initial expenditures associated with a finite resource pool; funding one intervention developing new pediatric oncology treatment centers, requires cutting another. This zero-sum assump- such as the initial training of personnel or acquisition tion may not be applicable to childhood cancer. of infrastructure. Further data on theoretical cost- In multiple LMICs, largely through the efforts effectiveness thresholds and real costs are needed to aid of nongovernmental organizations, private funds LMIC policy makers. that otherwise may have remained outside the Table 7.8 Comparison of Cost-Effectiveness Thresholds among Common Childhood Cancers, Selected Countries Threshold Brazil Malawi El Salvador El Salvador China Brazil United States Brazil Morocco Type of pediatric ALL BL SR-ALL HR-ALL ALL ALL ALL BL Wilms cancer Source Howard Hesseling Bonilla and Bonilla and Tang and Brandalise Pui and others Sandlund Madani and and others and others others 2010 others 2010 others and others 2009 and others others 2006 2004 2009 2008 2010 1997 Event-free 5-year 1-year 5-year 5-year 5-year 5-year 5-year 5-year 5-year survival definition Percentage 1 — — — 48.3 — — — — abandoning treatmenta Percentage 63 48 56.3 48.6 38.5 83.6 85.6 39 56.0 event-free survivalb Gross domestic $11,900 $900 $7,600 $7,600 $8,500 $11,900 $49,000 $11,900 $5,100 product per capita Life expectancy 72.79 52.31 73.69 73.69 74.84 72.79 78.49 72.79 76.11 Age at diagnosis 5.4 6.9 4.6 7.1 4.7 5.3 5.3 5.5 3 Upper limit 257,075 14,243 147,756 129,037 58,620 344,385 1,454,695 167,146 100,285 of very cost- effective (US$ per patient) Upper limit of 771,225 42,729 443,268 387,112 175,859 1,033,156 4,364,086 501,438 300,855 cost-effective (US$ per patient) Source: Bhakta and others 2012. Note: ALL = acute lymphoblastic leukemia; BL = Burkitt lymphoma; HR-ALL = high-risk ALL; SR-ALL = standard-risk acute lymphoblastic leukemia; — = not available. a. When no abandonment percentage is listed, the authors included abandonment as an event when calculating event-free survival. b. In all studies cited, relapse and abandonment were included as events when calculating event-free survival. Treating Childhood Cancer in Low- and Middle-Income Countries 139 Figure 7.3 Cost-Effective Thresholds Compared with Actual society may well hold resonance for governments, Costs in Selected Pediatric Malignancies policy makers, health care workers, and the general public. 10,000,000 1,000,000 CONCLUSIONS AND FUTURE DIRECTIONS 100,000 Although the advances in pediatric oncology in HICs have not been fully realized in most LMICs, significant US$ 10,000 progress has been achieved in some pediatric cancer units. The challenge remains to extend this progress to 1,000 all cancer centers in LMICs and to close the survival gap. The following steps are key prerequisites: 100 • The development of national childhood cancer strat- 10 egies is needed to move beyond the twinning para- -ri L L or Br L Br Bra L US e) LL LL ilm BL digm and to increase cure rates for entire populations. Ch ALL igh AL AL AL il B if um il A il A ec wi r h isk A ina az sk st (R z ala Lobbying of governments by clinicians and parent -r lva ard M W Sa and groups is required, as are strengthening links between co az do t oc rs childhood cancer advocates in HICs and LMICs. or do M lva El • To better inform governments and health officials, Sa El Cost-effective threshold (3:1) Drug costs onlya further research into the cost and cost-effectiveness of treatment is necessary. Without such data, the Very cost-effective threshold (1:1) Actual costb misconception of childhood cancer treatment as Gross domestic product per capita unaffordable will persist. • The outcomes of children with cancer should be Source: © BMJ Publishing Group Ltd. Reproduced, with permission, from Bhakta and monitored by individual treatment centers using data others 2012; further permission required for reuse. Note: ALL = acute lymphoblastic leukemia; BL = Burkitt lymphoma; US$= U.S. dollars. entry systems. These data should be used continually a. Costs only include chemotherapy and supportive care medications, such as to evaluate and modify the local implementation of antibiotics and antipyretics. therapeutic interventions. Governments can encour- b. Includes total costs for the entire treatment. Not included are the costs of lost economic productivity, associated infrastructure and personnel costs, or indirect costs age this process through national childhood cancer to parents, such as transportation, accommodation, and food. strategies that include high-quality pediatric registries. • Further research is needed into how to effectively treat various different childhood malignancies in set- tings of different resource constraints. Studies iden- health system have instead been allocated to pedi- tifying how to prevent common causes of treatment atric oncology centers. The success of Unidad failure in LMICs should be conducted. Nacional de Oncología Pediátrica (UNOP) in • The formation of cooperative groups of LMIC cen- Guatemala provides an example of how multiple ters should be encouraged as forums for protocol sectors can be mobilized, creating a positive-sum evaluation and advocacy; AHOPCA, the French- scenario. An initial outlay of funds to UNOP African Pediatric Oncology Group, and the Brazilian through a twinning program was leveraged into Childhood Cooperative Group for ALL Treatment are additional resources from government and private three excellent examples. Collaborations with HIC donors. The creation of an independent fundrais- cooperative groups may aid this process. ing organization (Fundación Ayúdame a Vivir, http://ayuvi.org.gt) was essential to this outcome. Pediatric oncology treatment can create a cohort Figure 7A.1 in the online annex illustrates the of cancer survivors in LMICs while building cancer results of this process. management capacity and galvanizing cancer advocacy • Finally, determining whether resources should be efforts more generally. Closing the pediatric oncology allocated to the treatment of childhood cancer may survival gap will help not only the more than 150,000 be more complex than simple analyses of cost and children in LMICs who develop cancer every year; it will cost-effectiveness. Arguments pertaining to justice, also have long-lasting benefits for the societies to which equity, and the non-monetary value of children to they belong. 140 Cancer NOTE Bonilla, M., S. Gupta, R. Vasquez, S. L. Fuentes, G. deReyes, and others. 2010. “Predictors of Outcome and Methodological World Bank income classifications as of July 2014 are as follows, Issues in Children with Acute Lymphoblastic Leukemia in based on estimates of gross national income per capita for 2013: El Salvador.” European Journal of Cancer 46 (18): 3280–86. Bonilla, M., N. Rossell, C. Salaverria, S. Gupta, R. Barr, and • Low-income countries: US$1,045 or less others. 2009. “Prevalence and Predictors of Abandonment • Middle-income countries: of Therapy among Children with Cancer in El Salvador.” • Lower-middle-income: US$1,046–US$4,125 International Journal of Cancer 125 (9): 2144–46. • Upper-middle-income: US$4,126–US$12,745 Brandalise, S. R., V. R. Pinheiro, S. S. Aguiar, E. I. Matsuda, • High-income countries: US$12,746 or more R. Otubo, and others. 2010. “Benefits of the Intermittent Use of 6-Mercaptopurine and Methotrexate in Maintenance Treatment for Low-Risk Acute Lymphoblastic Leukemia in REFERENCES Children: Randomized Trial from the Brazilian Childhood Abd El-Aal, H. H., E. E. Habib, and M. M. Mishrif. 2005. Cooperative Group: Protocol ALL-99.” Journal of Clinical “Wilms’ Tumor: The Experience of the Pediatric Unit of Oncology 28: 1911–18. Kasr El-Aini Center of Radiation Oncology and Nuclear Burkitt, D. 1958. “A Sarcoma Involving the Jaws of African Medicine (NEMROCK).” Journal of the Eyptian National Children.” British Journal of Surgery 46: 218–23. Cancer Institute 17: 308–14. Chandra, J., R. Naithani, V. Singh, Y. K. Saxena, M. Sharma, Abuidris, D. O., M. E. Elimam, F. M. Nugud, E. M. Eligali, and others. 2008. “Developing Anticancer Chemotherapy M. E. Ahmed, and others. 2008. “Wilms Tumour in Sudan.” Services in a Developing Country: Hodgkin Lymphoma Pediatric Blood & Cancer 50: 1135–37. Experience.” Pediatric Blood & Cancer 51: 485–88. Al Sheyyab, M., A. Bateiha, S. E. Kayed, and B. Hajjawi. 2003. “The Chantada, G., F. Doz, C. B. Antoneli, R. Grundy, F. F. Clare Incidence of Childhood Cancer in Jordan: A Population- Stannard, and others. 2006. “A Proposal for an International Based Study.” Annals of Saudi Medicine 23: 260–63. Retinoblastoma Staging System.” Pediatric Blood & Cancer Antillon, F., F. Baez, J. C. Barrantes Zamorra, L. C. Fu, B. Moreno, 47 (6): 801–05. and others. 2005. “AMOR: A Proposed Cooperative Effort Chantada, G., S. Luna-Fineman, R. S. Sitorus, M. Kruger, to Improve Outcomes of Childhood Cancer in Central T. Israels, and others. 2013. “SIOP-PODC Recommendations America.” Pediatric Blood & Cancer 45 (2): 107–10. for Graduated-Intensity Treatment of Retinoblastoma in Antillon-Klussmann, F., P. Valverde, C. Garrido, M. Castellanos, Developing Countries.” Pediatric Blood & Cancer 60 (5): P. De Alarcon, and others. 2010. (abstract). “Treatment 719–27. of Acute Lymphoblastic Leukemia in Limited Income Chantada, G. L., I. Qaddoumi, S. Canturk, V. Khetan, Z. Ma, Country: The Experience of the Unidad Nacional de and others. 2011. “Strategies to Manage Retinoblastoma in Oncologis Pediatrica (UNOP) of Guatemala.” Pediatric Developing Countries.” Pediatric Blood & Cancer 56 (3): Blood & Cancer 55: 861. 341–48. Arora, R. S., T. Eden, and B. Pizer. 2007. “The Problem of Crawford, J. R., T. J. Macdonald, and R. J. Packer. 2007. Treatment Abandonment in Children from Developing “Medulloblastoma in Childhood: New Biological Advances.” Countries with Cancer.” Pediatric Blood & Cancer 49: The Lancet Neurology 6 (12): 1073–85. 941–46. Creutzig, U., M. Zimmermann, D. Reinhardt, M. Dworzak, Ayoub, L., L. Fu, A. Pena, J. M. Sierra, P. C. Dominguez, and J. Stary, and others. 2004. “Early Deaths and Treatment- others. 2007. “Implementation of a Data Management Related Mortality in Children Undergoing Therapy for Program in a Pediatric Cancer Unit in a Low Income Acute Myeloid Leukemia: Analysis of the Multicenter Country.” Pediatric Blood & Cancer 49 (1): 23–27. Clinical Trials AML-BFM 93 and AML-BFM 98.” Journal of doi:10.1002/pbc.20966. Clinical Oncology 22 (21): 4384–93. Baez, F., B. F. Fossati, E. Ocampo, V. Conter, A. Flores, and Davidson, A., P. Hartley, F. Desai, J. Daubenton, H. Rode, others. 2002. “Treatment of Childhood Wilms’ Tumor and others. 2006. “Wilms Tumour Experience in a South without Radiotherapy in Nicaragua.” Annals of Oncology African Centre.” Pediatric Blood & Cancer 46: 465–71. 13: 944–48. Day, S. W., J. Garcia, F. Antillon, J. A. Wilimas, L. M. Mckeon, Baez, F., E. Ocampo, V. Conter, A. Flores, T. Gutierrez, and and others. 2012. “A Sustainable Model for Pediatric others. 1997. “Treatment of Childhood Hodgkin’s Disease Oncology Nursing Education in Low-Income Countries.” with COPP or COPP-ABV (Hybrid) without Radiotherapy Pediatric Blood & Cancer 58: 163–66. in Nicaragua.” Annals of Oncology 8: 247–50. Day, S. W., L. Segovia, P. Viveros, A. Banfi, G. K. Rivera, and Bhakta, N., A. L. C. Martiniuk, S. Gupta, and S. C. Howard. others. 2011. “Development of the Latin American Centre 2012. “The Cost-Effectiveness of Treating Paediatric Cancer for Pediatric Oncology Nursing Education.” Pediatric Blood in Low-Income and Middle-Income Countries: A Case- & Cancer 56: 5–6. Study Approach Using Acute Lymphocytic Leukaemia in de Camargo, B., J. de Oliveira Ferreira, R. de Souza Reis, Brazil and Burkitt Lymphoma in Malawi.” Archives of Disease S. Ferman, M. de Oliveira Santos, and others. 2011. in Childhood. doi:10.1136/archdischild-2011-301419. “Socioeconomic Status and the Incidence of Non-Central Treating Childhood Cancer in Low- and Middle-Income Countries 141 Nervous System Childhood Embryonic Tumours in Brazil.” Gupta, S., S. Yeh, A. Martiniuk, C. G. Lam, H. Y. Chen, BMC Cancer 11: 160. and others. 2013. “The Magnitude and Predictors of De Campo, J. F. 1986. “Ultrasound of Wilms’ Tumor.” Pediatric Abandonment of Therapy in Pediatric Acute Leukemia in Radiology 16: 21–24. Middle-Income Countries: A Systematic Review and Meta- Dimaras, H., K. Kimani, E. A. O. Dimba, P. Gronsdahl, Analysis.” European Journal of Cancer 49: 2555–64. A. White, and others. 2012. “Retinoblastoma.” The Lancet Hadley, G. P. 2010. “Can Surgeons Fill the Void in the 379 (9824): 1436–46. Management of Children with Solid Tumours in Not- Dunkel, I. J., H. S. L. Chan, R. Jubran, G. L. Chantada, Developing Countries?” Pediatric Blood & Cancer 55: 16–17. S. Goldman, and others. 2010. “High-Dose Chemotherapy Hadley, L. G., B. S. Rouma, and Y. Saad-Eldin. 2012. “Challenge with Autologous Hematopoietic Stem Cell Rescue for of Pediatric Oncology in Africa.” Seminars in Pediatric Stage 4B Retinoblastoma.” Pediatric Blood & Cancer 55 (1): Surgery 21: 136–41. 149–52. Harif, M., S. Barsaoui, S. Benchekroun, L. Boccon-Gibod, Ellison, L. F., L. Pogany, and L. S. Mery. 2007. “Childhood and R. Bouhas, and others. 2005. “Treatment of Childhood Adolescent Cancer Survival: A Period Analysis of Data from Cancer in Africa: Preliminary Results of the French-African the Canadian Cancer Registry.” European Journal of Cancer Paediatric Oncology Group.” Archives de Pediatrie 12: 851–53. 43: 1967–75. Harif, M., S. Barsaoui, S. Benchekroun, R. Bouhas, P. Doumbe, Ethier, M-C., E. Blanco, T. Lehrnbecher, and L. Sung. 2011. and others. 2008. “Treatment of B-Cell Lymphoma with “Lack of Clarity in the Definition of Treatment-Related LMB Modified Protocols in Africa: Report of the French- Mortality: Pediatric Acute Leukemia and Adult Acute African Pediatric Oncology Group (GFAOP).” Pediatric Promyelocytic Leukemia as Examples.” Blood 118: 5080–83. Blood & Cancer 50: 1138–42. Fajardo-Gutierrez, A., S. Juarez-Ocana, G. Gonzalez- Hartman, D. S., and R. C. Sanders. 1982. “Wilms’ Tumor Miranda, V. Palma-Padilla, R. Carreon-Cruz, and others. versus Neuroblastoma: Usefulness of Ultrasound in 2007. “Incidence of Cancer in Children Residing in Ten Differentiation.” Journal of Ultrasound in Medicine 1: Jurisdictions of the Mexican Republic: Importance of the 117–22. Cancer Registry (A Population-Based Study).” BMC Cancer Hesseling, P., T. Israels, M. Harif, G. Chantada, and E. Molyneux. 19 (7): 68. 2012. “Practical Recommendations for the Management Ferlay, J., H. R. Shin, F. Bray, D. Forman, C. Mathers, and of Children with Endemic Burkitt Lymphoma (BL) in a others. 2010. “Estimates of Worldwide Burden of Cancer in Resource Limited Setting.” Pediatric Blood & Cancer 60 (3): 2008: GLOBOCAN 2008.” International Journal of Cancer 357–62. doi:10.1002/Pbc.24407. 127: 2893–917. Hesseling, P., E. Molyneux, S. Kamiza, T. Israels, and Ganesan, P., L. Kumar, V. Raina, A. Sharma, S. Bakhshi, R. Broadhead. 2009. “Endemic Burkitt Lymphoma: and others. 2011. “Hodgkin’s Lymphoma—Long Term A 28-Day Treatment Schedule with Cyclophosphamide and Outcome: An Experience from a Tertiary Care Cancer Intrathecal Methotrexate.” Annals of Tropical Paediatrics Center in North India.” Annals of Hematology 90: 1153–60. 29 (1): 29–34. Gavidia, R., S. L. Fuentes, R. Vasquez, M. Bonilla, M-C. Ethier, Hesseling, P., E. Njume, F. Kouya, T. Katayi, P. Wharin, and others. 2012. “Low Socioeconomic Status Is Associated and others. 2012. “The Cameroon 2008 Burkitt with Prolonged Times to Assessment and Treatment, Sepsis Lymphoma Protocol: Improved Event-Free Survival with and Infectious Death in Pediatric Fever in El Salvador.” Plos Treatment Adapted to Disease Stage and the Response to One 7 (8): E43639. Induction Therapy.” Pediatric Hematology and Oncology Gobin, Y., I. J. Dunkel, B. P. Marr, S. E. Brodie, and 29 (2): 119–29. D. H. Abramson. 2011. “Intra-Arterial Chemotherapy for Hines-Thomas, M. R., S. C. Howard, M. M. Hudson, the Management of Retinoblastoma: Four-Year Experience.” M. J. Krasin, S. C. Kaste, and others. 2010. “Utility of Archives of Ophthalmology 129 (6): 732–37. Bone Marrow Biopsy at Diagnosis in Pediatric Hodgkin’s Graf, N., M. F. Tournade, and J. de Kraker. 2000. “The Role of Lymphoma.” Haematologica 95: 1691–96. Preoperative Chemotherapy in the Management of Wilms’ Hodgson, D. C., M. M. Hudson, and L. S. Constine. 2007. Tumor: The SIOP Studies.” International Society of Pediatric “Pediatric Hodgkin Lymphoma: Maximizing Efficacy and Oncology. Urologic Clinics of North America 27: 443–54. Minimizing Toxicity.” Seminars in Radiation Oncology 17: Green, D. M. 2004. “The Treatment of Stages I–IV Favorable 230–42. Histology Wilms’ Tumor.” Journal of Clinical Oncology Howard, S. C., D. Campana, E. Coustan-Smith, F. Antillon, 22: 1366–72. M. Bonilla, and others. 2005. “Development of a Regional Gupta, S., F. A. Antillon, M. Bonilla, L. Fu, S. C. Howard, and Flow Cytometry Center for Diagnosis of Childhood others. 2011. “Treatment-Related Mortality in Children Leukemia in Central America.” Leukemia 19: 323–25. with Acute Lymphoblastic Leukemia in Central America.” Howard, S. C., D. P. Jones, and C. H. Pui. 2011. “The Tumor Cancer 117: 4788–95. Lysis Syndrome.” New England Journal of Medicine 364: Gupta, S., M. Bonilla, S. L. Fuentes, M. Caniza, S. C. Howard, 1844–54. and others. 2009. “Incidence and Predictors of Treatment- Howard, S. C., M. L. Metzger, J. A. Wilimas, Y. Quintana, C. H. Pui, Related Mortality in Paediatric Acute Leukaemia in El and others. 2008. “Childhood Cancer Epidemiology in Salvador.” British Journal of Cancer 100 (7): 1026–31. Low-Income Countries.” Cancer 112: 461–72. 142 Cancer Howard, S. C., M. Pedrosa, M. Lins, A. Pedrosa, C. H. Pui, Preoperative Chemotherapy in Wilms’ Tumor: Results of and others. 2004. “Establishment of a Pediatric Oncology an International Society of Paediatric Oncology (SIOP) Program and Outcomes of Childhood Acute Lymphoblastic Clinical Trial.” Journal of Clinical Oncology 1: 604–09. Leukemia in a Resource-Poor Area.” Journal of the American Lewis, N., J. L. Young, P. Hesseling, P. Mccormick, and Medical Association 291 (20): 2471–75. N. Wright. 2012. “Epidemiology of Burkitt’s Lymphoma Hunger, S. P., X. Lu, M. Devidas, B. M. Camitta, P. S. Gaynon, in Northwest Province, Cameroon, 2003–2010.” Paediatrics N. J. Winick, and others. 2012. “Improved Survival for and International Child Health 32 (2): 82–85. Children and Adolescents with Acute Lymphoblastic Liu, Y., J. Chen, J. Tang, S. Ni, H. Xue, and others. 2009. “Cost Leukemia between 1990 and 2005: A Report from the of Childhood Acute Lymphoblastic Leukemia Care in Children’s Oncology Group.” Journal of Clinical Oncology Shanghai, China.” Pediatric Blood & Cancer 53 (4): 557–62. 30 (14): 1663–69. doi:10.1002/pbc.22127. Hunger, S. P., D. Reyes, O. Negrin, M. Montero, L. De La Rosa, Liu, L., H. L. Johnson, S. Cousens, J. Perin, S. Scott, and and others. 2011. “Decreased Early Mortality and others. 2012. “Global, Regional, and National Causes Increased Survival with Less Intensive Therapy for of Child Mortality: An Updated Systematic Analysis Acute Lymphoblastic Leukemia (ALL) in the Dominican for 2010 with Time Trends Since 2000.” The Lancet 379 Republic.” Abstract. Pediatric Blood & Cancer 57: 761. (9832): 2151–61. Hunger, S. P., L. Sung, and S. C. Howard. 2009. “Treatment Lowe, L. H., B. H. Isuani, R. M. Heller, S. M. Stein, J. E. Johnson, Strategies and Regimens of Graduated Intensity for and others. 2000. “Pediatric Renal Masses: Wilms Tumor Childhood Acute Lymphoblastic Leukemia in Low-Income and Beyond.” Radiographics 20: 1585–603. Countries: A Proposal.” Pediatric Blood & Cancer 52 (5): Lozano, R., H. Wang, K. J. Foreman, J. K. Rajaratnam, 559–65. M. Naghavi, and others. 2011. “Progress towards Israels, T. 2012. “Wilms Tumor in Africa: Challenges to Cure.” Millennium Development Goals 4 and 5 on Maternal and Pediatric Blood & Cancer 58: 3–4. Child Mortality: An Updated Systematic Analysis.” The Israels, T., E. Borgstein, M. Jamali, J. de Kraker, H. N. Caron, Lancet 378 (9797): 1139–65. and E. M. Molyneux. 2009. “Acute Malnutrition Is Common Luo, X. Q., Z. Y. Ke, X. Q. Guan, Y. C. Zhang, L. B. Huang, in Malawian Patients with a Wilms Tumor: A Role for and others. 2008. “The Comparison of Outcome and Cost Peanut Butter.” Pediatric Blood & Cancer 53: 1221–26. of Three Protocols for Childhood Non-High Risk Acute Israels, T., E. Borgstein, D. Pidini, G. Chagaluka, J. de Kraker, Lymphoblastic Leukemia in China.” Pediatric Blood & and others. 2012. “Managment of Children with a Wilms Cancer 51 (2): 204–09. Tumor in Malawi, Sub-Saharan Africa.” Journal of Pediatric Luo, X. Q., Z. Y. Ke, L. B. Huang, X. Q. Guan, Y. C. Zhang, Hematology/Oncology 34: 606–10. and X. L. Zhang. 2009a. “High-Risk Childhood Acute Israels, T., C. Moreira, T. Scanlan, E. Molyneux, Lymphoblastic Leukemia in China: Factors Influencing the S. Kampondeni, and others. 2013. “Clinical Guidelines Treatment and Outcome.” Pediatric Blood & Cancer 52 (2): for the Management of Children with Wilms Tumour in a 191–15. Low Income Setting.” Paediatrics and International Child ———. 2009b. “Improved Outcome for Chinese Children Health 60 (1): 5–11. with Acute Promyelocytic Leukemia: A Comparison of Two Kulkarni, K. P., and R. K. Marwaha. 2010. “Pattern and Protocols.” Pediatric Blood and Cancer 53 (3): 325–28. Implications of Therapy Abandonment in Childhood Madani, A., S. Zafad, M. Harif, M. Yaakoubi, S. Zamiati, and Acute Lymphoblastic Leukemia.” Asian Pacific Journal of others. 2006. “Treatment of Wilms Tumor According to Cancer Prevention 11: 1435–36. SIOP 9 Protocol in Casablanca, Morocco.” Pediatric Blood Kutluk, T., A. Varan, N. Buyukpamukcu, L. Atahan, M. Caglar, & Cancer 46 (4): 472–75. and others. 2006. “Improved Survival of Children with Magrath, I., E. Steliarova-Foucher, E. Sidnei, R. C. Ribeiro, Wilms Tumor.” Journal of Pediatric Hematology/Oncology M. Harif, and others. 2013. “Pediatric Cancer in Low- 28: 423–26. Income and Middle-Income Countries.” The Lancet Lange, B. J., F. O. Smith, J. Feusner, D. R. Barnard, P. Dinndorf, Oncology 14: 3104–16. and others. 2008. “Outcomes in CCG-2961: A Children’s Maris, J. M. 2010. “Recent Advances in Neuroblastoma.” New Oncology Group Phase 3 Trial for Untreated Pediatric England Journal of Medicine 362: 2202–11. Acute Myeloid Leukemia: A Report from the Children’s Marjerrison, S., C. V. Fernandez, V. E. Price, E. Njume, and Oncology Group.” Blood 111 (3): 1044–53. P. Hesseling. 2012. “The Use of Ultrasound in Endemic Leal-Leal, C., H. Dilliz-Nava, M. Flores-Rojo, and J. Robles-Castro. Burkitt Lymphoma in Cameroon.” Pediatric Blood & Cancer 2011. “First Contact Physicians and Retinoblastoma in 58 (3): 352–55. Mexico.” Pediatric Blood & Cancer 57: 1109–12. Masera, G., F. Baez, A. Biondi, F. Cavalli, V. Conter, and others. Leander, C., L. C. Fu, A. Pena, S. C. Howard, C. Rodriguez- 1998. “North-South Twinning in Paediatric Haemato- Galindo, and others. 2007. “Impact of an Education Oncology: The La Mascota Programme, Nicaragua.” The Program on Late Diagnosis of Retinoblastoma in Lancet 352 (9144): 1923–36. doi:S0140673698070779 [Pii]. Honduras.” Pediatric Blood & Cancer 49 (6): 817–19. Masera, G., T. Eden, M. Schrappe, J. Nachman, H. Gadner, Lemerle, J., P. A. Voute, M.-F. Tournade, C. Rodary, and others. 2004. “Statement by Members of the Ponte Di J. F. Delemarre, and others. 1983. “Effectiveness of Legno Group on the Right of Children to Have Full Access Treating Childhood Cancer in Low- and Middle-Income Countries 143 to Essential Treatment for Acute Lymphoblastic Leukemia.” Ngoma, T., M. Adde, M. Durosinmi, J. Githang’a, Y. Aken’Ova, Pediatric Blood & Cancer 43 (2): 13–14. doi:10.1002 and others. 2012. “Treatment of Burkitt Lymphoma in /Pbc.20135. Equatorial Africa Using a Simple Three-Drug Combination Matthay, K. K., J. G. Villablanca, R. C. Seeger, D. O. Stram, Followed by a Salvage Regimen for Patients with Persistent R. E. Harris, and others. 1999. “Treatment of High- or Recurrent Disease.” British Journal of Haematology Risk Neuroblastoma with Intensive Chemotherapy, 158 (6): 749–62. Radiotherapy, Autologous Bone Marrow Transplantation, Nguyen, K., M. Devidas, S. C. Cheng, M. La, E. A. Raetz, and 13-Cis-Retinoic Acid.” New England Journal of Medicine and others. 2008. “Factors Influencing Survival after 341 (16): 1165–73. Relapse from Acute Lymphoblastic Leukemia: A Children’s Mauz-Korholz, C., D. Hasenclever, W. Dorffel, K. Ruschke, Oncology Group Study.” Leukemia 22 (12): 2142–50. T. Pelz, and others. 2010. “Procarbazine-Free OEPA- doi:Leu2008251 [Pii]10.1038/Leu.2008.251. COPDAC Chemotherapy in Boys and Standard OPPA- Olweny, C. L., E. Katongole-Mbidde, C. Kiire, S. K. Lwanga, COPP in Girls Have Comparable Effectiveness in Pediatric I. Magrath, and others. 1978. “Childhood Hodgkin’s Hodgkin’s Lymphoma: The GPOH-HD-2002 Study.” Disease in Uganda: A Ten Year Experience.” Cancer 42: Journal of Clinical Oncology 28 (23): 3680–86. 787–92. Metzger, M., A. Billett, and M. P. Link. 2012. “The Impact of Ortega, J. J., L. Madero, G. Martin, A. Verdeguer, P. Garcia, Drug Shortages on Children with Cancer: The Example and others. 2005. “Treatment with All-Trans Retinoic of Mechlorethamine.” New England Journal of Medicine Acid and Anthracycline Monotherapy for Children with 367 (26): 2461–63. Acute Promyelocytic Leukemia: A Multicentre Study by Metzger, M. L., and J. S. Dome. 2005. “Current Therapy for the PETHEMA Group.” Journal of Clinical Oncology 23: Wilms’ Tumor.” The Oncologist 10 (10): 815–26. 7632–40. Metzger, M. L., S. C. Howard, L. C. Fu, A. Pena, R. Stefan, and Ortiz, R., S. Abad, A. Pena, F. Spreafico, J. Wilimas, and others. 2003. “Outcome of Childhood Acute Lymphoblastic others. 2012. “Treatment of Wilms Tumor (WT) Leukaemia in Resource-Poor Countries.” The Lancet within the Central American Association of Pediatric 362 (9385): 706–08. Hematology/Oncology (AHOPCA): Report from Moreira, C., M. N. Nachef, S. Ziamati, Y. Ladjaj, S. Barsaoui, Guatemala, Honduras, El Salvador, and Nicaragua: and others. 2012. “Treatment of Nephroblastoma in Causes of Treatment Failure.” Pediatric Blood & Cancer Africa: Results of the First French African Pediatric 59 (6): 971. Oncology Group (GFAOP) Study.” Pediatric Blood & Parkin, D. M., E. Kramarova, G. J. Draper, E. Masuyer, Cancer 58 (1): 37–42. J. Michaelis, and others. 1998. International Incidence of Moricke, A., M. Zimmermann, A. Reiter, G. Henze, Childhood Cancer. Vol. II. Lyon, France: International A. Schrauder, and others. 2010. “Long-Term Results of Agency for Research on Cancer. Five Consecutive Trials in Childhood Acute Lymphoblastic Patte, C., A. Auperin, M. Gerrard, J. Michon, R. Pinkerton, and Leukemia Performed by the ALL-BFM Study Group from others. 2007. “Results of the Randomized International 1981 to 2000.” Leukemia 24: 265–84. doi:Leu2009257 FAB/LMB96 Trial for Intermediate Risk B-Cell Non- [Pii]10.1038/Leu.2009.257. Hodgkin Lymphoma in Children and Adolescents: It Is Mostert, S., R. S. Arora, M. Arreola, P. Bagai, P. Friedrich, and Possible to Reduce Treatment for the Early Responding others. 2011. “Abandonment of Treatment for Childhood Patients.” Blood 109 (7): 2773–80. Cancer: Position Statement of a SIOP PODC Working Patton, G. C., C. Coffey, C. Cappa, D. Currie, L. Riley, Group.” The Lancet Oncology 12 (8): 719–20. and others. 2012. “Health of the World’s Adolescents: Mostert, S., M. N. Sitaresmi, C. M. Gundy, V. Janes, Sutaryo, and A Synthesis of Internationally Comparable Data.” The A. J. Veerman. 2010. “Comparing Childhood Leukaemia Lancet 379 (9826): 1665–75. Treatment before and after the Introduction of a Parental Pediatric Oncology Group of Ontario. 2012. Provincial Education Programme in Indonesia.” Archives of Disease in Pediatric Oncology Satellite Program. http://www.pogo.ca Childhood 95 (1): 20–25. /care/satelliteprograms. Mostert, S., M. N. Sitaresmi, C. M. Gundy, Sutaryo, and Pedrosa, F., M. Bonilla, A. Liu, K. Smith, D. Davis, and others. A. J. Veerman. 2006. “Influence of Socioeconomic Status on 2000. “Effect of Malnutrition at the Time of Diagnosis on Childhood Acute Lymphoblastic Leukemia Treatment in the Survival of Children Treated for Cancer in El Salvador Indonesia.” Pediatrics 118 (6): E1600–06. and Northern Brazil.” Journal of Pediatric Hematology/ Murphy, S. B. 1978. “Childhood Non-Hodgkin’s Lymphoma.” Oncology 22 (6): 502–05. New England Journal of Medicine 299: 1446–48. Perez-Cuevas, R., S. V. Doubova, M. Zapata-Tarres, S. Flores- Nachman, J. B., R. Sposto, P. Herzog, G. S. Gilchrist, S. L. Wolden, Hernandez, L. Frazier, and others. 2013. “Scaling Up Cancer and others. 2002. “Randomized Comparison of Low-Dose Care for Children without Medical Insurance in Developing Involved Field Radiotherapy and No Radiotherapy for Countries: The Case of Mexico.” Pediatric Blood & Cancer Children with Hodgkin’s Disease Who Acheive a Complete 60: 196–203. doi:10.1002/Pbc.24265. Response to Chemotherapy.” Journal of Clinical Oncology Perilongo, G., R. Maibach, E. Shafford, L. Brugieres, P. Brock, and 20 (18): 3765–71. others. 2009. “Cisplatin versus Cisplatin Plus Doxorubicin 144 Cancer for Standard-Risk Hepatoblastoma.” New England Journal Rodriguez-Galindo, C., M. W. Wilson, G. Chantada, L. Fu, of Medicine 361: 1662–70. I. Qaddoumi, and others. 2008. “Retinoblastoma: One Prucker, C., A. Attarbaschi, C. Peters, M. Dworzak, U. Potschger, World, One Vision.” Pediatrics 122 (3): E763–70. and others. 2009. “Induction Death and Treatment-Related Sackmann-Muriel, F., P. Zubizarreta, G. Gallo, M. Socpinaro, Mortality in First Remission of Children with Acute D. Alderete, and others. 1997. “Hodgkin Disease in Lymphoblastic Leukemia: A Population-Based Analysis Children: Results of a Prospective Randomized Trial in of the Austrian Berlin-Frankfurt-Munster Study Group.” a Single Institution in Argentina.” Medical and Pediatric Leukemia 23: 1264–69. Oncology 29: 544–52. Pui, C. H., D. Campana, D. Pei, W. P. Bowman, J. T. Sandlund, Sala, A., F. Antillon, P. Pencharz, R. Barr, and others and others. 2009. “Treating Childhood Acute Lymphoblastic for the AHOPCA Consortium. 2005. “Nutritional Leukemia without Cranial Irradiation.” New England Status in Children with Cancer: A Report from the Journal of Medicine 360: 2730–41. AHOPCA Workshop Held in Guatemala City, August Pui, C. H., and S. C. Howard. 2008. “Current Management and 31–September 5, 2004.” Pediatric Blood & Cancer 45: Challenges of Malignant Disease in the CNS in Paediatric 230–36. Leukaemia.” The Lancet Oncology 9 (3): 257–68. Sandlund, J. T., T. Fonseca, T. Leimig, L. Verissimo, Pui, C. H., D. Pei, A. S. Pappo, S. C. Howard, C. Cheng, C. W. Berard, and others. 1997. “Predominance and and others. 2012. “Treatment Outcomes in Black and White Characteristics of Burkitt Lymphoma among Children Children with Cancer: Results from the SEER Database with Non-Hodgkin Lymphoma in Northeastern Brazil.” and St. Jude Children’s Research Hospital, 1992 through Leukemia 11 (5): 743–46. 2007.” Journal of Clinical Oncology 30: 2005–12. Schilling, F. H., C. Spix, F. Berthold, R. Erttmann, N. Fehse, Pui, C. H., L. L. Robison, and A. T. Look. 2008. “Acute and others. 2002. “Neuroblastoma Screening at One Year of Lymphoblastic Leukemia.” The Lancet 371 (9617): Age.” New England Journal of Medicine 346: 1047–53. 1030–43. Siddiqui, N., B. Ayub, F. Badar, and A. Zaidi. 2006. “Hodgkin’s Razack, R., P. Michelow, G. Leiman, A. Harnekar, J. Poole, Lymphoma in Pakistan: A Clinico-Epidemiological Study and others. 2011. “An Interinstitutional Review of the of 658 Cases at a Cancer Center in Lahore.” Asian Pacific Value of FNAB in Pediatric Oncology in Resource-Limited Journal of Cancer Prevention 7: 651–55. Countries.” Diagnostic Cytopathology 40 (9): 770–76. Siegel, R., D. Naishadham, and A. Jemal. 2013. “Cancer Ribeiro, R., and C. H. Pui. 2005. “Saving the Children: Statistics, 2013.” CA: A Cancer Journal for Clinicians 63 (1): Improving Childhood Cancer Treatment in Developing 11–30. Countries.” New England Journal of Medicine 352: Silverman, L. B., K. E. Stevenson, J. E. O’Brien, B. L. Asselin, 2158–60. R. D. Barr, and others. 2010. “Long-Term Results of Dana- Ribeiro, R. C., E. Steliarova-Foucher, I. Magrath, J. Lemerle, Farber Cancer Institute ALL Consortium Protocols for T. Eden, and others. 2008. “Baseline Status of Paediatric Children with Newly Diagnosed Acute Lymphoblastic Oncology Care in Ten Low-Income or Mid-Income Leukemia (1985–2000).” Leukemia 24 (2): 320–34. Countries Receiving My Child Matters Support: doi:Leu2009253 [Pii]10.1038/Leu.2009.253. A Descriptive Study.” The Lancet Oncology 9: 721–29. Sitaresmi, M. N., S. Mostert, R. M. Schook, and A. J. Veerman. Ries, L. A., M. A. Smith, J. G. Gurney, M. Linet, T. Tamra, and 2010. “Treatment Refusal and Abandonment in Childhood others, editors. 1999. Cancer Incidence and Survival among Acute Lymphoblastic Leukemia in Indonesia: An Analysis Children and Adolescents: United States SEER Program of Causes and Consequences.” Psychooncology 19 (4): 1975–1995. NIH Pub. No. 99–4649. Bethesda, MD: National 361–67. doi:10.1002/Pon.1578. Cancer Institute SEER Program. Smith, M. A., N. L. Seibel, S. F. Altekruse, L. A. G. Ries, Rivera-Luna, R., C. Correa-Gonzalez, E. Altamirano-Alvarez, D. L. Melbert, and others. 2010. “Outcomes for F. Sanchez-Zubieta, R. Cardenas-Cardos, and others. 2012. Children and Adolescents with Cancer: Challenges for the “Incidence of Childhood Cancer among Mexican Children Twenty-First Century.” Journal of Clinical Oncology 28: Registered under a Public Medical Insurance Program.” 2625–34. International Journal of Cancer 132: 1646–50. doi:10.1002/ Sripada, P. V., S. G. Tenali, M. Vasudevan, S. Viswanadhan, Ijc.27771. D. Sriraman, and others. 1995. “Hybrid (COPP/ABV) Rodriguez-Galindo, C. 2011. “The Basics of Retinoblastoma: Therapy in Childhood Hodgkin’s Disease: A Study of 53 Back to School.” Pediatric Blood & Cancer 57 (7): 1093–94. Cases during 1989–1993 at the Cancer Institute, Madras.” doi:10.1002/Pbc.23305. Pediatric Hematology Oncology 12: 333–41. Rodriguez-Galindo, C., G. Chantada, B. G. Haik, and St. Jude Children’s Research Hospital. 2012. St. Jude Children’s M. W. Wilson. 2007a. “Treatment of Retinoblastoma: Research Hospital Cure4Kids: Pediatric Oncology Education. Current Status and Future Perspectives.” Current Treatment St. Jude Children’s Research Hospital. http://www Options in Neurology 9 (4): 294–307. .cure4kids.org. ———. 2007b. “Retinoblastoma: Current Treatment and Stefan, D. C., and D. Stones. 2009. “How Much Does It Cost Future Perspectives.” Current Treatment Options in to Treat Children with Hodgkin Lymhpoma in Africa?” Neurology 9 (4): 294–307. Leukemia & Lymphoma 50: 196–99. Treating Childhood Cancer in Low- and Middle-Income Countries 145 Stiller, C. A., and D. M. Parkin. 1996. “Geographic and Ethnic Wang, Y. R., R. M. Jin, J. W. Xu, and Z. Q. Zhang. 2011. Variations in the Incidence of Childhood Cancer.” British “A Report about Treatment Refusal and Abandonment in Medical Bulletin 52: 682–703. Children with Acute Lymphoblastic Leukemia in China, Tang, Y., X. Xu, H. Song, S. Yang, S. Shi, and others. 2008. 1997–2007.” Leukemia Research 35 (12): 1628–31. “Long-Term Outcome of Childhood Acute Lymphoblastic WHO (World Health Organization). 2003. Making Choices in Leukemia Treated in China.” Pediatric Blood & Cancer 51: Health: WHO Guide to Cost-Effectiveness Analysis, edited by 380–86. T. Tan-Torres Edejer, R. Baltussen, T. Adam, R. Hutubessy, Testi, A. M., A. Biondi, F. L. Coco, M. L. Moleti, F. Giona, A. Acharya, and others. Geneva: WHO. and others. 2005. “GIMEMA-AIEOP AIDA Protocol for ———. 2013. WHO Model List of Essential Medicines for the Treatment of Newly Diagnosed Acute Promyelocytic Children. WHO, Geneva. http://www.who.int/medicines Leukemia (APL) in Children.” Blood 106: 447–53. /publications/essentialmedicines/en/index.html. Traore, F., C. Coze, J.-J. Atteby, N. Andre, C. Moreira, and others. Wilde, J. C., W. Lameris, E. H. Van Hasselt, E. Molyneux, 2011. “Cyclophosphomide Monotherapy in Children with H. A. Heij, and others. 2010. “Challenges and Outcome Burkitt Lymphoma: A Study from the French-African of Wilms’ Tumour Management in a Resource- Pediatric Oncology Group (GFAOP).” Pediatric Blood & Constrained Setting.” African Journal of Paediatric Surgery Cancer 56 (1): 70–76. 7: 159–62. Van Litsenburg, R. R., C. A. Uyl-De Groot, H. Raat, G. J. Kaspers, Woessmann, W., K. Seidemann, G. Mann, M. Zimmermann, and R. J. Gemke. 2011. “Cost-Effectiveness of Treatment B. Burkhardt, and others. 2005. “The Impact of the of Childhood Acute Lymphoblastic Leukemia with Methotrexate Administration Schedule and Dose Chemotherapy Only: The Influence of New Medication in the Treatment of Children and Adolescents with and Diagnostic Technology.” Pediatric Blood & Cancer B-Cell Neoplasms: A Report of the BFM Group Study 57 (6): 1005–10. doi:10.1002/Pbc.23197. NHL-BFM95.” Blood 105 (3): 948–58. Veerman, A. J., Sutaryo, and Sumadiono. 2005. “Twinning: Womer, R. B., D. C. West, M. D. Krailo, P. S. Dickman, A Rewarding Scenario for Development of Oncology B. Pawel, and others. 2012. “Randomized Controlled Trial Services in Transitional Countries.” Pediatric Blood & of Interval-Compressed Chemotherapy for the Treatment Cancer 45 (2): 103–06. doi:10.1002/Pbc.20390.0. of Localized Ewing Sarcoma: A Report from the Children’s Viana, M. B., R. A. Fernandes, B. M. De Oliveira, M. Murao, Oncology Group.” Journal of Clinical Oncology 30 (33): C. De Andrade Paes, and others. 2001. “Nutritional and 4148–54. Socio-Economic Status in the Prognosis of Childhood Acute Woods, W. G., K. Kobrinsky, J. D. Buckley, J. W. Lee, J. Sanders, Lymphoblastic Leukemia.” Haematologica 86 (2): 113–20. and others. 1996. “Time-Sequential Induction Therapy Vujanic, G. M., A. Kelsey, C. Mitchell, R. S. Shannon, and Improves Postremission Outcome in Acute Myeloid P. Gornali. 2003. “The Role of Biopsy in the Diagnosis Leukemia: A Report from the Children’s Cancer Group.” of Renal Tumors of Childhood: Results of the UKCCSG Blood 87 (12): 4979–89. Wilms Tumor Study 3.” Medical and Pediatric Oncology 40: Zsiros, J., R. Maibach, E. Shafford, L. Brugieres, P. Brock, 18–22. and others. 2010. “Successful Treatment of Childhood Vujanic, G. M., B. Sandstedt, A. Kelsey, and N. J. Sebire. High-Risk Hepatoblastoma with Dose-Intensive 2009. “Central Pathology Review in Multicenter Trials and Multiagent Chemotherapy and Surgery: Final Results Studies: Lessons from the Nephroblastoma Trials.” Cancer of the SIOPEL-3HR Study.” Journal of Clinical Oncology 115: 1977–83. 28 (15): 2584–90. 146 Cancer Chapter 8 Liver Cancer Hellen Gelband, Chien-Jen Chen, Wendong Chen, Silvia Franceschi, Sir Andrew Hall, W. Thomas London, Katherine A. McGlynn, and Christopher P. Wild INTRODUCTION cancer mortality (746,000 deaths in 2012) (Ferlay and others 2013). Incidence and mortality rates vary greatly, Deaths from liver cancer are common, especially in East mirroring the uneven distribution of major risk factors. In Asia and Pacific, South Asia, and parts of Sub-Saharan most high-rate liver cancer areas, the dominant risk factors Africa, largely as a result of infection decades ago. As the toll are chronic infection with HBV and consumption of foods from other cancers is likely to climb in the coming decades, contaminated with the mycotoxin aflatoxin B1. In contrast, however, liver cancer incidence and mortality rates should in most low-rate areas, the major risk factors are infec- fall, as generations vaccinated against the hepatitis B virus tion with the hepatitis C virus (HCV), excessive alcohol (HBV)—the cause of most liver cancers globally—reach consumption, obesity, and diabetes. HBV and HCV have middle and old age. Much still needs to be done and it is been classified by the International Agency for Research feasible and affordable to hasten the decline. Much can on Cancer (IARC) as carcinogenic to humans (Group 1). also be done to address other causes of liver cancer— The most common histologic type of primary liver including some on the rise, in particular, obesity-related cancer, hepatocellular carcinoma (HCC), arises from non-alcoholic fatty liver disease (NAFD)—in the coming the epithelial liver cells known as hepatocytes. Globally, years and decades. The latter half of the twentieth century approximately 80–85 percent of primary liver cancers witnessed the identification of the main causes of liver can- are HCCs; the rates of primary liver cancer and rates of cer and deployment of the first cancer prevention vaccine HCC are roughly equivalent. Intrahepatic cholangio- for humans. All of the risk factors that lead to cirrhosis carcinoma, which arises from cholangiocytes—epithelial cause at least as many noncancer deaths as cancer deaths. cells that line the bile duct—is the second most com- Controlling these risk factors would not only reduce the mon type of primary liver cancer, but it accounts for incidence of liver cancer; it would also reduce the incidence only 10–12 percent of primary liver cancer worldwide. of cirrhosis and its other complications, notably, end-stage Infection with liver flukes (flatworms) is a major cause liver disease and portal hypertension. of cholangiocarcinoma in high-incidence regions. GLOBAL BURDEN OF LIVER CANCER Incidence and Mortality Rates1 Primary liver cancer—cancer originating in the liver—is The highest national liver cancer incidence rates in the sixth most commonly occurring cancer in the world the world are found in East Asia and Pacific and Sub- (782,000 cases in 2012) and the second largest cause of Saharan Africa (maps 8.1 and 8.2). Approximately Corresponding author: Hellen Gelband, gelband@cddep.org 147 Map 8.1 Age-Standardized Liver Cancer Incidence Rates for Women, 2012 IBRD 41383 | FEBRUARY 2015 Liver Cancer in Women, 0–69 years ASR (World) Indcidence Greater than 4.9 2.6 –4.9 1.9–2.6 1.2–1.9 Less than 1.2 No data Source: Ferlay and others 2013. Note: ASR = age-standardized rate. Map 8.2 Age-Standardized Liver Cancer Incidence Rates for Men, 2012 Liver Cancer in Men, 0–69 years ASR (World) Indcidence Greater than 9.3 5.5–9.3 4–5.5 2.7–4 Fewer than 2.7 No data Source: Ferlay and others 2013. Note: ASR = age-standardized rate. 148 Cancer 85 percent of the total liver cancer burden is concen- 5 per 100,000 and 10 per 100,000—in some countries trated in these areas; China alone—because of high of central Europe (for example, Greece and Italy) and rates and a very large population—bears over half of the central Asia (for example, Kazakhstan, the Kyrgyz global burden (Ferlay and others 2013). Although more Republic, Pakistan, and Turkmenistan). Not all low- liver cancers occur in China than any other country, the and middle-income countries (LMICs) have high HBV country with the highest liver cancer incidence rate in infection rates. A prime example is India, which has the world is Mongolia (78.1 per 100,000 population), historically had low HBV infection rates and, as a result, with an incidence rate more than three times as high low liver cancer incidence rates. as that of China (22.3 per 100,000) (Ferlay and others At all incidence levels, almost all countries report rates 2013). These exceedingly high rates are reportedly the in men that are twofold to threefold higher than rates result of high rates of infection with HBV, HCV, or both, in women. The greatest gender disparity in incidence, as well as co-infection of HBV carriers with hepatitis D however, is not reported by countries with the highest virus (HDV) (Oyunsuren and others 2006). liver cancer rates, such as Mongolia (where the ratio in The prevalence of HBV infection in Mongolia is likely men to women is 1.6) (men, 97.8; women, 61.1), but by to decline as the results of a childhood HBV vaccination countries with intermediate rates, such as France (men, program that began in 1991 decrease the prevalence of 11.3; women, 2.5) and Spain (men, 9.9; women, 2.4). new infections. However, the continued high rate of An exception to the general predominance in men occurs HCV infection may result in liver cancer incidence rates in Central America and Mexico, where the rates of both increasing for some years (Dondog and others 2011). genders are low and not very different. Although the Alcohol consumption rates are reportedly higher in reasons for higher rates in men in most regions are not Mongolia than in many Asian populations, a factor that completely understood, the differences may be partly may contribute to the liver cancer burden (Alcorn 2011). explained by the sex-specific prevalence of risk factors. In addition to China, other East Asian and Pacific Men are more likely to be chronically infected with countries and economies with incidence rates greater HBV and HCV, consume alcohol, and smoke cigarettes. than 20 per 100,000 include Taiwan, China; the Republic Whether androgenic hormones or increased genetic sus- of Korea; the Lao People’s Democratic Republic; Thailand; ceptibility also predispose men to the development of liver and Vietnam. Thailand has high rates of both HCC and cancer is unclear (Hsieh and others 2007). intrahepatic cholangiocarcinoma in the northeastern part of the country, where liver fluke infection is com- mon. Countries in Sub-Saharan Africa with incidence Age-Specific Incidence rates greater than 20 per 100,000 include The Gambia Liver cancer incidence rates increase with age in all and Guinea; rates in the Democratic Republic of Congo populations, with the highest rates in those ages (10.6 per 100,000), Ghana (11.1 per 100,000), and Guinea- 75 years and older. The age-specific curves look some- Bissau (13.4 per 100,000) are almost as high (Ferlay and what different in various regions, but in no area do the others 2013). Although the reported incidence of liver rates decline among older persons (men and women cancer in Sub-Saharan Africa is lower than the incidence combined). In a low-rate area, such as northern in East Asia and Pacific, at least some of the difference Europe, rates are generally very low before age 40 and arises from underdiagnosis and the historic paucity of then rise exponentially with age. In high-rate areas well-functioning, reliable cancer registries. Another rea- of East Asia and Pacific, rates become elevated in son for the lower incidence in Sub-Saharan Africa is that childhood and continue to rise with age. In contrast, in babies are less likely to become infected with HBV in the the high-rate area of Sub-Saharan Africa, rates increase perinatal period compared with babies in Asia (Marinier until age 55 years and then plateau until age 70 years. and others 1985). The later age at infection results in lower The reasons for the slightly different patterns in Asia HBV replication rates in Sub-Saharan Africa than in East and Sub-Saharan Africa may be related to competing Asia and Pacific (Evans and others 1998). Nevertheless, causes of mortality and differences in mean ages at in both East Asia and Pacific and Sub-Saharan Africa, the HBV infection and/or differences in HBV viral repli- major risk factor for liver cancer is chronic HBV infection. cation patterns (Evans and others 1998). Notable exceptions are the Arab Republic of Egypt and Japan, in which HCV is the dominant risk factor. In contrast to these high-rate areas, rates are low Incidence by World Bank Economic Group in North and South America and northern Europe. Globally, the single greatest determinant of liver cancer Incidence rates in these areas are generally less than prevalence in any country is the prevalence of chronic 5 per 100,000. Rates are intermediate—typically between HBV infection. As chronic HBV infection has historically Liver Cancer 149 Figure 8.1 Age-Standardized Liver Cancer Incidence and Mortality Rates Rates have been declining in some eastern Asian in Men and Women, by World Bank Income Classification, 2012 countries for several reasons. In Japan, the large cohort of individuals infected with HCV in the 1930s and 1940s is a. Men b. Women dying off, and the rate of HCV-related HCC is declining Low income accordingly (Tanaka and others 2002). In China, where HBV is the dominant risk factor, HBV vaccination of Lower middle income (except India) newborns was introduced in the mid-1980s but ramped Upper middle income up only after 2000. Vaccine recipients are still too young (except China) to have their rates greatly affect the rates in the overall High income population. It is more likely that the rates in China have declined because exposure to aflatoxin B1 in the diet China has decreased as a result of shifting from a corn-based to a rice-based diet (Sun and others 2013). The declining India rates of HCC in younger age groups reported in Jiangsu 30 20 10 0 10 20 30 Province, a high aflatoxin B1 area, support this hypothe- Age-standardized rate (W) per 100,000, 0–69 years sis (Chen and others 2006; Wang and others 2010). Incidence Mortality Prognosis and DALYs Source: Ferlay and others 2013. The prognosis for liver cancer, even in HICs, is unfa- been much more common in LMICs, liver cancer has vorable. In the United States, the one-year survival rate been more prevalent in these countries (figure 8.1). In is less than 50 percent; the five-year survival rate is only general, the incidence rates decrease as a country’s per 16 percent (NCI 2013). As with all cancers, survival is best capita gross domestic product (GDP) increases. Among when detected early—for localized liver cancer, five-year the low-income countries (LICs), 13 of the 34 (38 percent) survival is 29 percent, but it falls to 3 percent for cancers have rates of 10 per 100,000 or greater. In the lower- detected late (NCI 2013). Survival is even less favorable middle-income group, excluding India, 18 of 49 countries in LMICs. Mortality rates in all locations are roughly (37 percent) have rates of 10 per 100,000 or higher. In the equivalent to incidence rates, and the number of years upper-middle-income group, excluding China, only 2 of lived with disability is very small. Accordingly, disability- 55 countries (4 percent) have rates of this magnitude; in adjusted life years (DALYs) caused by ill health, disability, the high-income countries (HICs), only 2 of 75 countries or early death from liver cancer are almost identical to (3 percent) have rates of 10 per 100,000 or higher. the years of life lost (YLLs) because of liver cancer. The A rapidly expanding economy does not have an imme- age-standardized rate of DALYs is highest in the high- diate effect on a country’s HBV carrier rate and does not incidence areas of eastern Asia (744 per 100,000 for men; greatly alter its liver cancer incidence in the short term. For 277 per 100,000 for women) and western Africa (451 per example, although China has undergone rapid economic 100,000 for men; 213 per 100,000 for women); it is lowest development in the late twentieth century, it still has a in the low-incidence areas of Europe (110 per 100,000 high liver cancer rate. With economic growth, however, for men; 45 per 100,000 for women) and North America comes the ability to reduce future liver cancer rates by (109 per 100,000 for men; 39 per 100,000 for women). funding universal HBV vaccination of newborns and afla- toxin abatement programs, ensuring that the blood supply Major Risk Factors is free of HCV, and possibly treating HCV infection. Hepatitis B Virus An estimated two billion people alive today have been Trends infected with HBV; about 360 million of these are From 1983–87 through 1998–2002, liver cancer incidence chronically infected (carriers) (Dienstag 2008). Routes increased in many areas of the world. Increases were nota- of transmission vary by life stage, but HBV is never ble in northern Europe, India, Israel, North and South spread by air, food, or water. Neonates may be infected America, Oceania, and most countries in southern by their mothers, if the mothers are infectious carriers. Europe. In contrast, incidence rates declined in most During childhood, transmission can occur among chil- eastern Asian countries. Increases in incidence in HICs dren living in close proximity, although the precise route and India have been linked to HCV infection, increasing is unclear. The virus is found in the blood and in most rates of obesity and diabetes, and improved treatment of body secretions. In adult life, the virus can be spread cirrhosis (which reduces the risk of death from cirrhosis, sexually via semen and vaginal fluid or by contami- leaving individuals at persistent risk of HCC). nated needles, frequently the result of intravenous drug 150 Cancer misuse. Blood transfusion was an important source of estimates that about 180 million people, some 3 percent infection before the introduction of donor selection and of the world’s population, are infected with HCV, of serological screening of donated blood. which 130 million are chronic carriers. At least three The natural history of chronic infection is the million to four million people are newly infected each development of chronic hepatitis, then cirrhosis, and year; this compares with the approximately 360 million finally, liver cancer. However, especially in Sub-Saharan people chronically infected with HBV. HCV prevalence Africa, cancer can develop without underlying cirrhosis. is highest (10 percent or more) in Pakistan (Ahmad Aflatoxin exposure multiplies the risk of liver cancer in 2004), Egypt (Attia 1998), Mongolia (Dondog and chronic carriers by a factor of twofold to tenfold. The others 2011), and some parts of China (Gao and others interval from the initial HBV infection to the develop- 2011). Prevalence is also high in some parts of Italy ment of cancer is in the range of 5–75 years, with most (Fusco and others 2008) and Japan (Tanaka and others cases manifesting after several decades. In 1994, IARC 2002). Globally, 10 million intravenous drug users are declared HBV a human carcinogen (Group 1). HCV-positive, with prevalence in this group exceeding The age of infection is critically important in deter- 60 percent in most countries (Nelson and others 2011). mining whether the infection becomes chronic. Newborns HCV transmission is primarily from contaminated infected by their mothers have an 80–90 percent probabil- blood or blood products; however, HCV can be acquired ity of becoming carriers. Those infected in the first five from sexual and household contacts. Acute infection years of life beyond the perinatal period have a 20–50 with HCV is usually asymptomatic; in approximately percent probability; those infected in adult life have a 75 percent of people, HCV persists as a chronic infec- probability of less than 10 percent. This is the inverse of tion, in which HCV RNA and serum HCV antibodies the risk of developing acute hepatitis, which manifests in (anti-HCV) can be detected in the blood (Alter and one-third of adult infections with the characteristic fever, Seeff 2000). Approximately 20 percent of individuals jaundice, and lethargy, whereas it is very unusual follow- with chronic HCV will develop HCC by age 75 (Huang ing infection in children under age five years. and others 2011). Acute hepatitis carries a risk of death of 1 percent or Evidence suggests that HCV prevalence is substan- less. Chronic infection increases the risk of primary liver tially underestimated for several reasons: cancer by tenfold to fiftyfold. • Country-specific data are absent, notably in LICs. In some populations in the pre-vaccination era, the • Surveillance systems that focus on acute HCV infec- prevalence of chronic HBV infection was 10–15 percent tion are insensitive because the infection is rarely of the adult population. This was the case in China and symptomatic. surrounding East Asian countries, much of Sub-Saharan • Prevalence surveys tend to include mainly young adults Africa, and the Amazon forest. These high carriage rates (for example, blood donors or pregnant women), resulted from very high rates of infection in the early while transmission of HCV through contaminated years of life, either perinatally from mothers or in early blood and needles is more probable in adulthood and childhood from other children (WHO 2004). In China, old age than in childhood (Dondog and others 2011). some 40 percent of carriers were infected perinatally by their mothers. In contrast, in Sub-Saharan Africa, The seroprevalence of HCV and HBV in cases of HCC only 10 percent of carriers were infected perinatally and can be used to estimate HCV and HBV prevalence where 90 percent were infected through child-to-child trans- accurate population-based data are lacking. In a large mission during the first few years of life, but overall, meta-analysis of seroprevalence (Raza and others 2007), rates were very high. These differences resulted from a higher proportion of cases had HBV than HCV in most the fact that about half of Chinese women who were countries in East Asia and Pacific, Latin America and HBV-positive remained infectious into adult life, whereas the Caribbean, South Asia, and Sub-Saharan Africa. In only 10 percent of HBV-positive African women did so contrast, in most European countries, Japan, the United (Marinier and others 1985). States, and, among low-resource countries, Egypt and Rates of liver cancer in Chinese carriers are higher Pakistan, HCV was more frequent than HBV (figure 8.2). than in Sub-Saharan African carriers. The reasons for The proportion of HCV-associated HCC cases is grow- this are unclear but may be related to the differences in ing steadily in other countries and economies, for exam- age at infection. ple, Taiwan, China (Lu and others 2006). HCV epidemics have accompanied the increase in Hepatitis C Virus the availability of injections and blood transfusions HCV, a blood-borne RNA virus identified in 1988, in several countries (Prati 2006). Massive and unsafe was declared a definite human carcinogen by IARC injection campaigns have occurred in many countries, in 1994. The World Health Organization (WHO) for example, the anti-schistosomal treatments in Japan Liver Cancer 151 Figure 8.2 HCV and HBV Presence in HCC Cases: Countries and Economies Where HCV Predominates or Is Increasing a. United States b. Italy c. Japan (n = 4,005) (n = 1,675) (n = 7,823) 100 100 100 90 90 90 80 80 80 70 70 70 60 60 60 Percent Percent Percent 50 50 50 40 40 40 30 30 30 20 20 20 10 10 10 0 0 0 e th th g g th e g V e V tiv sA sA tiv sA V Bo Bo tiv HC Bo HC HC ga ga HB HB HB ga ti- ti- Ne ti- Ne Ne an an an d. Taiwan, China e. Pakistan f. Egypt , Arab Rep. (n = 9,205) (n = 779) (n = 1,811) 100 100 100 90 90 90 80 80 80 70 70 70 60 60 60 Percent Percent Percent 50 50 50 40 40 40 30 30 30 20 20 20 10 10 10 0 0 0 th th ive ive e Ag Ag th g V V CV tiv sA Bo Bo HC Bo HC s s at at i-H ga HB HB HB g g ti- ti- Ne Ne Ne t an an an Source: Adapted from Raza and others 2007. Note: Includes studies that reported seroprevalence of both hepatitis B surface antigen (HBsAg) and anti-HCV, alone and in combination, for at least 20 hepatocellular carcinoma cases; updated through 2010. HBV = hepatitis B virus; HCV = hepatitis C virus; HCC = hepatocellular carcinoma. 152 Cancer starting in the 1920s and in Egypt in the 1960s, and the exposure among individuals not chronically infected use of intravenous stimulants starting in Japan in the with HBV (Omer and others 2004; Wu and others 1940s (Tanaka and others 2002). Unsafe blood transfu- 2009). In contrast to aflatoxins and HBV, there has been sions and medical procedures have tended to increase in little focus on the potential for interaction with HCV. times of military conflicts. By the end of the 1980s, most Knowledge about aflatoxin and risk of liver cirrhosis chronic transfusion recipients, and virtually all patients is also limited (Kuniholm and others 2008). However, receiving clotting factor concentrates, had been infected exposure to aflatoxins in childhood has been linked to by HCV (Prati 2006). New cases of HCV infection greatly hepatomegaly (enlarged liver) (Gong and others 2012). decreased in high-resource countries in the 1990s after Improved understanding of the biochemistry of the introduction of HCV testing for blood donors, inac- aflatoxin-cell interactions strongly supports the asso- tivation procedures for blood derivatives, and disposable ciation between aflatoxin exposure and liver cancer. needles and syringes. Following bioactivation, AFB1 binds predominantly to However, HCV-related HCC continued to increase as guanine in DNA, leading to a mutation from guanine to those people who were infected decades ago aged. Trends thymine. In HCCs collected from areas where aflatoxin in HCV prevalence were generally dominated by cohort exposure is high, up to half have been shown to harbor effects: an estimated 73 percent of HCV infections this type of mutational change in one specific location in the United States, for example, involve individuals within the tumor suppressor gene TP53. This same born between 1945 and 1965 (Smith and others 2012). mutation is extremely rare in liver tumors from regions In some low-resource countries, the spread of HCV has where aflatoxin exposure is low (Wild and Gong 2010). continued and may undermine the future benefits of Based on animal, epidemiological, and mechanistic HBV immunization (Dondog and others 2011). data, IARC has classified mixtures of naturally occur- ring aflatoxins as carcinogenic to humans (Group 1) Aflatoxins (IARC 2002, 2012a). An association between early life Aflatoxins are secondary metabolites of the fungal species exposure to aflatoxins and impaired child growth has Aspergillus flavus and A. parasiticus. These toxins contam- been reported and may represent a significant additional inate many staple cereals and oilseeds, with particularly disease burden globally (Gong and others 2002, 2004). high levels found on maize (corn) and groundnuts (pea- Aflatoxin exposure is ubiquitous in many of the nuts) (IARC 2002). Contamination occurs during crop poorest populations worldwide and is a cause of human cultivation and increases postharvest under poor storage liver cancer. Aflatoxins appear to be more potent among conditions, in which high humidity and temperature pro- HBV chronic carriers than among noncarriers, but there mote fungal growth and toxin production. An estimated may also be an increased risk in the absence of HBV. 4.5 billion people worldwide are exposed to aflatoxins Given more than 350 million chronic HBV carriers world- (Williams and others 2004); the highest exposures are in wide, many in areas with endemic aflatoxin consumption, LICs in East and Southeast Asia and Sub-Saharan Africa, the need for reduction of aflatoxin exposure is highly rel- where regulatory control is weak. evant for liver cancer prevention. Adverse effects of early The naturally occurring aflatoxins are aflatoxins B1, B2, life exposure to aflatoxins add to the public health con- G1, and G2 (AFB1, AFB2, AFG1, and AFG2, respectively). cerns related to these potent, naturally occurring toxins. AFB1 is the most abundant, toxic, and carcinogenic. Aflatoxins M1 and M2 (AFM1 and AFM2, respectively), Alcohol the hydroxylation products of AFB1 and AFB2, respec- In 1988, IARC classified alcohol as a Group 1 human tively, are found in milk and milk products (IARC 2002). carcinogen of the liver, causing HCC (IARC 2012a). Aflatoxins induce toxicity and tumors in the liver in a The National Institute of Alcohol Abuse and Alcoholism wide range of animal species. Epidemiological studies in (NIAAA) has also documented that prolonged heavy different populations have established aflatoxins as a risk drinking is associated with primary liver cancer (NIAAA factor for HCC (IARC 2002, 2012a). Prospective cohorts 1993), mainly through cirrhosis (Corrao and Arico have provided the most convincing evidence, with the 2000). The dose-response relationship for the amount increased risk highest among individuals who were also of alcohol consumed and the risk of HCC has been chronically infected with HBV (Qian and others 2013; explored in a meta-analysis (Corrao and others 2004), Wang and others 1996; Wu and others 2009). The effect which produced a relative risk of liver cirrhosis of 27 and of aflatoxins in the absence of chronic HBV infection of HCC of 1.8 for the heaviest drinkers (100 grams daily) is more difficult to assess because of their common compared with non-drinkers. The incidence of liver can- co-occurrence in many populations. Nevertheless, some cer increases by 0.7 per 1,000 for every additional drink, studies have reported an increased risk with aflatoxin defined as 14.0 grams (0.6 ounces) of pure alcohol by Liver Cancer 153 NIAAA, regularly consumed per day (Allen and others prevalence of infection varies greatly. In Thailand in 2009). Alcohol-associated liver cirrhosis is the most 1980–81, approximately 35 percent of the population important risk factor for HCC in populations with low in the northeast was infected; prevalence in the south prevalence of HBV and HCV, as in the United States and was close to zero. The national average was 14 percent. northern Europe (Boffetta and Hashibe 2006). Alcohol- By 2001, presumably as a result of treatment and control, related HCC without preexisting cirrhosis is rare. prevalence in the northeast had fallen to 15 percent and Alcohol is synergistic with tobacco consumption and the national average to 10 percent. chronic viral hepatitis (either HBV or HCV) in causing The variation in O. viverrini infection in Thailand is HCC (Haenel 1989). Alcohol as a solvent increases echoed in the variation in mortality rates from cholan- the exposure of hepatocytes to carcinogens such as giocarcinoma, which varies more than 12-fold from a low 4-aminobiphenyl and polycyclic aromatic hydrocar- in Prachuap province in the south to a high in Nakhon bons in tobacco smoke (IARC 2012a). However, alcohol Phanom province in the northeast (Sripa and others 2011). consumption remains an independent risk factor for In 2009, IARC estimated that 35 million people were HCC after adjustment for multiple host and viral infected with C. sinensis in China, Korea, Lao PDR, factors (Chen, Yu, and Liaw 1997). The risk of develop- Thailand, the Russian Federation, and Vietnam (IARC ing HBV/HCV-related HCC is about twice as high for 2009). Infections are occasionally reported in Malaysia habitual alcohol drinkers as for non-drinkers (Chen and and Singapore. In China, prevalence varies, with the Chen 2002). highest rates in Guangdong (16.4 percent) and Guangxi The carcinogenicity of alcoholic beverages does not (9.8 percent) in southern China and in Korean ethnic seem to vary with the type of beverage; the effect appears communities in northeastern Heilongjiang province to be caused by ethanol itself. After ingestion, ethanol is (4.7 percent) (Shin and others 2010). converted by the enzymes alcohol dehydrogenase and The spread of these liver flukes is restricted by the cytochrome P450 2E1 into acetaldehyde, which is oxidized distribution of the two definitive hosts other than by aldehyde dehydrogenase (ALDH) to acetate (IARC humans—particular species of snail (mainly of the genus 2012b). Acetaldehyde is a plausible candidate for the car- Bothynia) and cyprinid fish—and the cultural practice of cinogenic effects of alcoholic drinks through two mech- eating raw fish (either fresh or fermented fish can harbor anisms. It forms adducts (bonds) with DNA, leading to fluke metacercariae that infect humans after ingestion). mutations, and it can increase cell proliferation (Boffetta The transmission cycle also requires eggs from fish- and Hashibe 2006). Ethanol-caused hepatocellular injury eating hosts, which emerge in feces, to contaminate the can result in enhanced fibrogenesis and, finally, cirrhosis. freshwater bodies inhabited by the snails and fish—the A deficiency of ALDH—and subsequent inefficient oxi- consequences of the poor sanitation that is synonymous dation of acetaldehyde to acetate—substantially increases with poverty. Other fish-eating mammals, notably dogs the risk of alcohol-related liver cancer. and cats, play some role in maintaining the cycle, but their part is thought to be relatively unimportant. Liver Flukes Opisthorchis viverrini and Clonorchis sinensis—liver Obesity flukes—are the strongest risk factors for cholangio- Obesity is discussed last because it currently is not an carcinoma, accounting for about 15 percent of liver important risk factor in LMICs. However, like other cancers globally. O. viverrini was classified by IARC as obesity-related health problems, cirrhosis from NAFD a human carcinogen in 1994, and C. sinensis was so is on the rise in HICs (Baffy, Brunt, and Caldwell classified in 2009. These small, but not microscopic 2012; Michelotti, Machado, and Diehl 2013; Vanni and (1–2.5 centimeters by 0.3–0.5 centimeter), worms are Bugianesi 2014). All expectations are that in the coming prevalent only in parts of eastern Asia, but infection in decades it will have a similar impact in LMICs. HCC is some places is extraordinarily common—historically the cancer most strongly affected by NAFD. reaching virtually entire populations in some areas— and possibly one-sixth of those with chronic infection develop cholangiocarcinoma, starting in middle age. INTERVENTIONS The worms, protected in the bile duct, live surprisingly long lives—up to two decades or more for C. sinensis but Hepatitis B Virus possibly less than 10 years for O. viverrini if untreated The major intervention is primary prevention of chronic (Sithithaworn and Haswell-Elkins 2003). infection by vaccination (vaccination has no impact on O. viverrini infects at least 10 million people in established chronic infection). The vaccine consists of Lao PDR and Thailand and an unknown number in the surface antigen of the virus, an antigen that circulates Cambodia and Vietnam (Sripa and others 2011). The in the blood of carriers. The first vaccine was made by 154 Cancer separating this antigen from the blood of carriers and intravenous drug users has become the predominant sterilizing it. Although this plasma-derived vaccine is source of HCV acquisition in high-resource countries still used in some parts of the world, in most places it has (Nelson and others 2011). In contrast, most infections in been replaced by recombinant hepatitis B vaccine, which low-resource countries are iatrogenic, acquired through uses a manufactured surface antigen. contaminated blood and contaminated injections of The plasma-derived vaccine first became available in medications. Many low-resource countries, mostly in 1982 and initial trials in high-risk populations showed it East Asia and Pacific, South Asia, and Sub-Saharan Africa, to be safe and effective. It has subsequently been intro- do not systematically screen blood donations for HCV, duced in 168 countries as part of the routine childhood although screening for HBV and human immunodefi- vaccination program. However, this has been a gradual ciency virus (HIV) is common (Prati 2006). Blood safety process over the intervening 30 years. Three doses of the in low-resource countries is additionally threatened by vaccine are required to ensure protection against acute lack of voluntary, nonpaid blood donors; inadequate hepatitis and chronic infection. These can be given at supplies of medical instruments and laboratory reagents; intervals of one month. Most countries use a schedule and lack of infrastructure. Many transfusions could be of doses at two, three, and four months of age, either as avoided with the use of appropriate measures to optimize a separate injection or as part of a combination vaccine the patient’s own blood volume before surgery and to with diphtheria, pertussis, and tetanus. Early studies minimize blood loss during surgery (Goodnough 2013). showed that administration of the vaccine in the first More than 16 billion injections are administered annu- 24 hours of life could prevent transmission from a highly ally in low-resource countries (Prati 2006). Especially high infectious mother. The recommended WHO schedule is rates of injections have been reported in Mongolia and three or four doses of vaccine, with the initial dose within some countries of the former Soviet Union; they have 24 hours of birth (WHO 2009). also been reported in Pakistan (Ahmad 2004) and some The HBV vaccine is supported by The Vaccine Aliance Sub-Saharan African countries (Attia 1998). The most (Gavi) in countries eligible because of low income levels. frequently injected medications, nearly all of which could By the end of 2011, Gavi estimated that the vaccines it be taken orally, include antibiotics, vitamins, iron, and has supported would save 5.5 million lives, 3.7 million of analgesics, as well as treatments for nonspecific symptoms them attributable to the HBV vaccine. such as headache, fatigue, or fever. Injections represent Dramatic reductions in the prevalence of HBV carriage an important source of revenue for providers and are have been demonstrated in early adopters of universal vac- encouraged by the popular belief that injections are more cination. In China, for example, vaccination has reduced effective than oral administration. Insufficient hygiene, the prevalence of carriage from greater than 10 percent inappropriate use of multiple-dose medication vials, and to less than 1 percent (Cui and others 2010). In Taiwan, sharing bottles of intravenous solution also contribute to China, mortality rates from infant fulminant hepatitis, the spread of HCV. Finally, HCV transmission can occur chronic liver disease, and HCC have declined more than through traditional medicine (for example, acupuncture 90 percent among those vaccinated since a nationwide and scarring) and outside health care settings (for exam- HBV immunization began (Chiang and others 2013). ple, tattooing). Preventing infection is the most affordable option to reduce HCV-related diseases in low-resource Screening for and Treating HBV countries (box 8.1). In addition to vaccination, testing blood donors for chronic infection and excluding positive blood has Screening and Treating Chronic HCV eliminated this relatively small but important source Much progress has been made in the management of of infection. A global program to reduce needle reuse chronic hepatitis C, in terms of eradicating the virus and to ensure the safe disposal of used needles has also and reducing hepatic and extrahepatic complications reduced the risk of infection. and HCV-related deaths. Increasingly efficacious regi- More recently, antiviral treatment of those chronically mens have been developed to eradicate the virus. The infected has been shown to reduce the risk of liver cancer. first was interferon monotherapy, which was introduced Studies are underway in countries with a high prevalence in 1990; a combination of ribavirin and pegylated inter- of HBV in adults to screen for carriage, evaluate those feron was introduced in 2002; and in 2011, a protease found infected, and treat those with significant disease. inhibitor (telaprevir or boceprevir) was added to ribavi- rin and pegylated interferon (Morgan and others 2013). A new generation of direct-acting antiviral drugs, Hepatitis C Virus the first two of which entered the market in late 2013— Although no vaccines against HCV are available, most sofosbuvir and simeprevir—is promising. Compared HCV transmission could be avoided. Needle sharing by with the best previous regimens, they are significantly Liver Cancer 155 One of the factors allowing for low-priced antiret- Box 8.1 rovirals for HIV has been the economies of scale that come with a hugely expanded market. Hill and others Interventions to Prevent Blood-Borne (2014) estimate the minimum production costs of the two available HCV drugs, the two in late-stage Infections, Including HCV development, and ribavirin, using the history of HIV • Increase awareness of the importance of drugs as a guide. Using these costs, they estimate the HCV and other blood-borne infections and costs of complete combination 12-week regimens the means of preventing them, among the at from US$78 to US$166 for the least expensive, to population, health care workers, and tradi- US$232 to US$454 for the most expensive. Recognizing tional healers. that companies require compensation for their invest- • Use oral treatments instead of injections and ment and profit, these are not suggested as realistic use alternatives to blood transfusions when- sales prices; however, with tiered pricing (charging ever possible. higher prices in richer countries) and high volumes, • Use safe injection practices in the manage- affordable prices can be envisioned and should be ment of sharp waste. pursued by the global community for this now highly • Provide services to intravenous drug users, curable infection. including access to sterile needles and syringes. Widespread treatment implies screening, as well, • Make blood supplies safe by recruiting vol- which currently is recommended for relatively few untary donors and screening all donated individuals. Should treatment be made affordable, the blood for markers of HCV, HBV, and HIV. recommended populations should be expanded. Note: HBV = hepatitis B virus; HCV = hepatitis C virus; HIV = human immu- nodeficiency virus. Aflatoxins Human exposure to aflatoxins can be reduced in several ways. Aspergillus spp. infects crops during cultivation, but more effective in clearing the virus (in trials, up to toxins accumulate postharvest under poor storage condi- 100 percent of patients cleared, including multiple viral tions. Accordingly, preharvest and postharvest interven- genotypes), are taken for half the treatment period of tions can contribute to controlling aflatoxins (Groopman, previous regimens (12 versus 24 weeks), have few side Kensler, and Wild 2008; Wild and Hall 2000). Aflatoxin effects, and are taken orally (with another oral antiviral control measures are applied to commercial crops to drug, ribavirin) rather than having an injectable com- meet the stringent regulatory demands to limit exposure ponent. At least two other similar drugs are in late-stage in HICs. Industrial-scale sorting, optimal storage, and development. Hepatitis C medications were put on the extensive aflatoxin testing programs combine to limit WHO Essential Medicines List in 2015. They are all, exposure in wealthier nations. In contrast, little is done however, priced well above what is affordable in LMICs; to reduce the exposure of the populations in LICs, where they would have serious budgetary implications, even in aflatoxin levels are high (Pitt and others 2012). HICs. In the United States in 2014, sofosbuvir is priced The most effective measure to reduce exposure is to at US$84,000 for a course of treatment, and simeprivir avoid consumption of contaminated foods or to reduce at US$66,000. Because the previous regimens are also dependence on them. In China, for example, economic expensive (US$25,000–US$45,000), relatively few HCV- development led to a shift from consumption of maize to infected people anywhere have been treated. In 2010, an rice, which is far less susceptible to aflatoxin contamina- estimated 7 percent of those with chronic infection in tion. This dietary shift has been linked to reduced expo- France were treated, 3 percent in the United Kingdom, sure to aflatoxins and falling HCC rates (Chen and others and 0.3 percent in Russia (Razavi and others 2013). 2013). However, many of the poorest populations remain The only way the new anti-HCV drugs can benefit trapped by poverty and the lack of dietary alternatives. most infected individuals is if they can be made afford- Preharvest mycotoxin control includes a wide range able, particularly in LMICs. Patents on these drugs will of good agricultural practices to reduce crop stress— not begin to expire until 2025. In the meantime, several such as irrigation; early sowing; low plant density; and hundred thousand deaths per year could be averted by the use of fungicides, pesticides, and insecticides—as finding an acceptable means of setting lower prices, well as the identification of fungus-resistant strains, at least for LMICs, similar to the success story of low- genetic engineering of crop resistance, and biocontrol priced antiretrovirals for HIV. (Pitt and others 2012). Biocontrol is one of the most 156 Cancer vigorously investigated approaches (Mehl and others the world. From 2003 to 2005, average annual alcohol 2012), although it is too early to know whether it will pro- consumption per adult (over age 15) was highest in duce usable interventions (Atehnkeng and others 2008). Europe and Central Asia (12.2 liters of pure alcohol); The approach relies on competitive exclusion: strains followed by North and South America (8.7 liters), the of Aspergillus that do not produce aflatoxin (atoxigenic western Pacific (6.3 liters), Sub-Saharan Africa (6.2 liters), strains) are introduced into the soil, and they compete and southeast Asia (2.2 liters); it was lowest in the eastern with native aflatoxin-producing spores for colonization Mediterranean (0.7 liters) (WHO 2008). of crops. The cost can be high; ensuring access and sus- Excessive alcohol consumption can take the form of tainability at the subsistence or small-farm level is critical. heavy drinking, binge drinking, or any drinking by preg- Increased aflatoxin during storage is a major problem nant women or underage youth (CPSTF 2013). Heavy with dietary staples, such as maize and groundnuts in drinking is defined as more than two drinks per day on LMICs. Improved storage presents the simplest and most average for men or more than one drink per day on aver- affordable opportunity for limiting exposure in such age for women. Binge drinking is defined as five or more settings. In studies in West Africa, basic improvements drinks during a single occasion for men, or four or more in sorting, drying, and storing the groundnut crop in drinks during a single occasion for women. West Africa resulted in marked reductions in aflatoxin Several interventions have been developed to reduce contamination, feasibly and cost-effectively (Turner and excessive alcohol consumption. The U.S. Community others 2013; Wu and Khlangwiset 2010a, 2010b), sug- Preventive Services Task Force has reviewed the effec- gesting that simple, inexpensive approaches can offer tiveness of current intervention strategies, although not significant benefits. Nevertheless, challenges remain in commenting on the use of interventions outside the scaling up and implementing such approaches. United States. The recommended effective interventions Once present in a food commodity, the toxins are that are most likely to be applicable in LMICs include relatively resistant to destruction during preparation the following: and cooking. An exception is the use of alkaline meth- ods (nixtimalization), for example, using lime for maize • Increase alcohol taxes. tortilla preparation, as is done by some populations in • Maintain limits on days of sale. Latin America and the Caribbean (Pitt and others 2012). • Maintain limits on hours of sale. An alternative to primary prevention is to modify • Enforce laws against privatizing retail alcohol sales. the effects of toxins once ingested, either by preventing • Regulate alcohol outlet density. absorption or by modifying metabolism (Groopman, • Enhance enforcement of laws prohibiting sales to Kensler, and Wild 2008). Reduced absorption has been minors. achieved in animal feeds by the incorporation of clays into The effectiveness of other interventions, such feeds to bind aflatoxin in the gastrointestinal tract; this as overservice law enforcement initiatives—efforts to process remains at the pilot phase in humans (Wang and increase enforcement of laws against serving intoxicated others 2008) and may be limited to emergency situations individuals—and responsible beverage service training, is of high contamination and food insufficiency. In terms not supported by sufficient evidence to recommend them. of altered metabolism, several compounds have been Unfortunately, very little information is available about explored—including chlorophyllin, oltipraz, and broccoli the effects of these and alternative interventions in LMICs. sprout extract (Groopman, Kensler, and Wild 2008)—but these have not been translated to widespread application. Liver Flukes Infection with liver flukes can be prevented, and infec- Alcohol tions can be treated with a single dose of praziquantel, Alcohol is found in beer, wine, and liquor, as well as in an inexpensive and relatively safe drug. Infections are some medicines, mouthwashes, household products, detected by the presence of eggs excreted in stool, a stan- and essential oils (for example, scented liquids taken dard laboratory procedure. Once cholangiocarcinoma from plants). The alcohol content of various alcoholic develops, like HCC, treatment of the cancer is almost beverages varies: 3–7 percent for beers and hard ciders; always futile. 9–15 percent for wines and sake; 16–20 percent for wines The simplest means of preventing infection appears fortified with liquors, such as port; and 35–40 percent to be stopping the consumption of raw fish. This for liquor or distilled spirits, such as gin, rum, vodka, approach, however, has not been simple, because raw fish and whiskey (IARC 2012b). Alcohol consumption var- dishes are well-established components of the cuisine of ies among adults over age 15 years in different parts of certain cultures. Liver Cancer 157 Control programs in Thailand started in the 1950s. and more countries have begun vaccinating; now, 168 of In the late 1980s, the Thai government initiated a new 190 WHO member countries have implemented univer- three-pronged control strategy that includes screening sal hepatitis B vaccination in newborns and infants. and treatment with single-dose praziquantel (which has Chronic HBV can be treated with medications, a cure rate of 95 percent or higher), education programs although current therapy seldom results in a cure to encourage cooking fish before eating, and practicing (Aman and others 2012). The antiviral agents currently hygienic defecation (Jongsuksuntigul and Imsomboon used include lamivudine, adefovir, tenofovir, telbivu- 2003). The large decreases in infection prevalence in dine, entecavir, and the two immune system modulators, Thailand are attributed to these combined efforts. interferon alpha-2a and pegylated interferon alpha-2a. Finally, work is proceeding on a vaccine against Lifetime treatment inhibits virus replication, mini- O. viverrini and is considered a feasible development mizing liver damage and reducing the risk of HCC (Sripa and others 2011). (Matsumoto and others 2005). In an Indian study, the incremental cost-effectiveness ratio (ICER) per Treatment of HCC QALY gained associated with interferon use compared with usual care was six times the per capita GDP of India Liver resection and liver transplantation are the gold (Aggarwal, Ghoshal, and Naik 2002); this cost is not con- standards for treating HCC, but even with the best treat- sidered cost-effective according to the WHO definition ment, survival is poor. In the United States, five-year of cost-effectiveness, as an ICER less than three times the relative survival from liver cancer was 16 percent during per capita GDP of the country. Even if drug therapy were the 2000s, with a plurality of patients diagnosed with cost-effective in HICs, it may be unaffordable in LMICs. early-stage disease. Survival from late-stage disease was For example, the ICER per QALY gained for lamivudine less than 5 percent (NCI 2013). in HICs ranged from 0.024 per capita GDP in Australia Some new loco-regional ablative methods, notably, (Crowley and others 2000) to 0.819 per capita GDP in percutaneous radiofrequency ablation, may represent the United States (Yuan and others 2008). In LMICs, the safer and cheaper alternatives for small lesions (less than ICER per QALY gained for lamivudine could be as high 5.0 centimeters) (Zhang and Chen 2010) and may even- as three times the per capita GDP (Lui and others 2010). tually be commonly used in LMICs. Therefore, the current strategy of treating patients from a young age is very expensive for LMICs, although this could change when the price of these drugs declines. COSTS AND COST-EFFECTIVENESS OF Future studies should explore the cost-effectiveness of INTERVENTIONS treating older patients before the onset of cirrhosis, Fortunately—and unlike most other major cancers— possibly using on-off treatment that may be significantly preventive interventions for all of the major risk factors are less costly but have similar long-term clinical benefits. feasible. However, as is the case for other cancers and many other conditions, costs and cost-effectiveness are not well HCV Control studied in LMICs. We present the evidence as documented in the literature, focusing first on prevention and then on There is no effective vaccine for HCV, but viral transmis- treatment of liver cancer. We could find no economic anal- sion can be prevented by reducing exposure to infected ysis of liver fluke control. (Discussion of excessive alcohol blood products and unsafe medical procedures, the main consumption prevention is covered in DCP3 volume 4, routes of infection in LMICs (Kermode 2004). Screening Mental, Neurological, and Substance Use Disorders.) blood donors and improving the safety of medical pro- cedures can significantly reduce HCV infection (Chanzy and others 1999; Wang and others 2013). The main HBV Control laboratory test for detecting HCV in blood donations Vaccination against HBV is highly cost-effective in is the third-generation HCV antibody enzyme-linked HBV-endemic countries. The economic evidence from immunosorbent assay, which has high sensitivity and China suggests that universal HBV vaccination is cost- specificity (Colin and others 2001). saving compared with not vaccinating (Hutton, So, and Unsafe injections, which could be eliminated by Brandeau 2010). Even in India, a low HBV-endemic single-use needles and syringes, are another impor- country, universal HBV vaccination is also highly cost- tant route for HCV transmission in LMICs. A recent effective at US$13.22 per quality-adjusted life year cost-benefit analysis reported that introducing auto- (QALY) gained (Aggarwal, Ghoshal, and Naik 2003). disable syringes (which prevent reuse by locking mech- Since the introduction of vaccines in the 1980s, more anisms or other mechanical means) for all medical 158 Cancer injections in India would be highly cost-effective, costing treating HCC are almost entirely lacking. A thorough US$46–US$48 per DALY (Reid 2012). search of the medical literature identified just three Standard antiviral therapy for HCV before the studies of health expenditures for HCC treatment, all new generation of drugs—pegylated interferon plus in HICs. The average treatment cost in Russia for an ribavirin—cures more than half of all patients (Hartwell HCC patient in 2008 was US$10,400, almost equal to and Shepherd 2009). However, the treatment takes the per capita GDP in that year. Inpatient care accounted months and is expensive, making it cost-effective only for 90 percent of the total. The average indirect cost in patients in HICs who have progressive liver disease associated with HCC was US$707, in which productiv- placing them at high risk of HCC (Shepherd and others ity loss accounted for 26 percent and social welfare for 2004). The new drugs discussed earlier in this chapter are disability accounted for 74 percent (Omelyanovsky and currently even more expensive, but more effective, are all others 2010). oral, and are taken for a shorter duration. If costs can be The second study is from Taiwan, China, and reports brought down, they could become affordable in LMICs. on national insurance claims data for 2,873 patients diagnosed between 1996 and 2002 (Lang and others Aflatoxin Control 2008). More than 30 percent of the patients had died within one year of diagnosis. The highest expenditures The economic burden of HCC in LMICs can be reduced accrued during the six months before death, equaling by controlling the risk of food contamination by afla- US$7,183, or half of Taiwan, China’s per capita GDP in toxin. Preharvest and postharvest interventions are 2002. The projected 10-year costs were the equivalent of possible. the per capita GDP. Good farming practices to reduce crop stress (for A recent Canadian study investigated the health example, irrigation; early sowing; low plant density; use care costs of 2,341 HCC patients in Ontario, Canada, of fungicides, pesticides, and insecticides) reduce afla- from 2002 to 2008. Average five-year net costs were toxin production in addition to providing other more US$77,509, or US$15,502 annually, about 30 percent of direct benefits, but the practices may be costly. For exam- Canada’s 2008 per capita GDP. ple, advanced irrigation systems cost US$740–US$940 Without studies from LMICs, it is impossible to per acre in the United States (Burt 2000). know what the economic burden of treating HCC would Introducing atoxigenic strains of Aspergillus is poten- be, but we can make some inferences. First, we know tially cost-effective in LMICs (Khlangwiset and Wu that patients in LMICs are much less likely to receive 2010; Wu and Khlangwiset 2010a, 2010b). In Nigeria, treatment than patients in HICs. Most HCCs, even in local atoxigenic strains of A. flavus were inoculated on HICs, are diagnosed at late stages. Second, for those who maize and substantial reductions in aflatoxin levels are treated, the dollar costs might be somewhat lower were achieved (Atehnkeng and others 2008); data from than in the studies cited, assuming lower input prices. additional field trials are awaited. Sun drying and hand However, it is likely to be an even larger fraction of per sorting, together with other postharvest measures, were capita GDP—possibly many multiples—making the effective in reducing aflatoxins in the groundnut crop burden even heavier for what is, in most cases, treatment in Guinea and were considered cost-effective (Wu and that does not save the patient’s life. Khlangwiset 2010b). Postharvest interventions include sorting, adequate drying, and good storage conditions for grains. In Summary of Costs large-scale commercial farming, mechanical blanch- The direct and indirect costs of HCC in LMICs are ing and sorting is highly cost-effective in the United higher compared with costs in HICs because the disease States, completely eliminating aflatoxin at a cost of is more frequent, it occurs at younger ages, and diagnosis US$150–US$170 per ton (Dorner and Lamb 2006). is often delayed. However, their reliance on electricity and technical The most cost-effective approaches to preventing support puts them out of reach in many LMICs. The HCC are the following: cost-effectiveness of artificial drying to reduce aflatoxin is highly correlated with the costs of fuel and electricity • Reducing the prevalence of HBV through vaccination and moisture content in harvested crops. of infants, including a birth dose, and uninfected adolescents. Costs of Treating HCC • Preventing HCV transmission by blood donor The clinical burden of HCC in LMICs is relatively well screening and safe injection practices, including auto- understood, but analyses of health expenditures for disable syringes. Liver Cancer 159 • Controlling aflatoxin through postharvest and attractive, especially if preventive measures have possibly some preharvest (genetically resistant maize reduced transmission. and/or atoxigenic strains of Aspergillus) practices. • HCC and cholangiocarcinoma treatments, which are • Educating people to eat fully cooked food in areas usually unsuccessful for long-term survival in any with high prevalence of liver fluke infection to pre- country, particularly because the diseases are not usu- vent cholangiocarcinoma. ally diagnosed until late stages. Liver cancer treatment • Treating liver fluke infection with single-dose cannot be recommended as an important control praziquantel. strategy for LMICs. The economic attractiveness of these interventions Approaches that are not currently cost-effective are may change, however, as new technologies are introduced the following: and the prices of existing effective technologies fall. • Antiviral therapies for HBV and HCV, which are very expensive and require extended periods of treatment (HBV requires lifetime treatment). They are used CONCLUSIONS by only a small fraction of infected individuals in Liver cancer is among the least successfully treated of HICs and unlikely to be cost-effective in LMICs. all cancers but among the most readily prevented. Each However, as treatment efficacy improves and prices major risk factor is amenable to modification by feasible moderate, these interventions could become more and cost-effective (and some effective but expensive) interventions. Some of these should become more affordable in LMICs in the medium term; for example, Table 8.1 Liver Cancer Risk Factors and Feasible Interventions drugs that are currently expensive will eventually be Cost- much less expensive. All of these interventions will pre- Risk factor Intervention effectiveness vent morbidity and mortality from other liver diseases and, in the case of alcohol, from a number of other HBV HBV vaccine +++ conditions. The risk factors and interventions are sum- HBV treatment −/+ marized in table 8.1. Blood supply protection +++ HCV Blood supply protection +++ NOTES Safe medical injection +++ Programs to provide clean needles for injected- World Bank income classifications as of July 2014 are as ++ drug users follows, based on estimates of gross national income per capita for 2013: HCV treatment −/+ • Low-income countries: US$1,045 or less Aflatoxin Preharvest good agricultural practices to reduce ++ • Middle-income countries: crop stress • Lower-middle-income: US$1,046–US$4,125 Fungus-resistant strains ++ • Upper-middle-income: US$4,126–US$12,745 Biocontrol Not yet known • High-income countries: US$12,746 or more Postharvest sorting, storing, drying +++ 1. Maps and figures in this chapter are based on incidence and mortality estimates for ages 0 to 69, consistent Improved grain storage +++ with reporting in all Disease Control Priorities volumes. Postharvest treatment of grain + Global cancer statistics are estimates for the year 2012 Postingestion effect modification – and have been provided by the International Agency for Research on Cancer from its GLOBOCAN 2012 data- Alcohol Taxation +++ base. Observable population-based data were derived Legal/regulatory interventions +++ from Cancer Incidence in Five Continents, 10th edition Behavioral interventions ++ (Forman and others 2013) and for trends over time from CI5 Plus (http://ci5.iarc.fr/CI5plus/Default.aspx). The Liver flukes Public education to avoid raw fish ++ discussion of burden (including risk factors), however, Screen for infection and treat with praziquantel ++ includes all ages unless otherwise noted. Interventions also Note: Scale from not cost-effective (−) to highly cost-effective (+++) in low- and middle-income apply to all age groups, except where age ranges or cutoffs countries. HBV = hepatitis B virus; HCV = hepatitis C virus. are specified. 160 Cancer REFERENCES Chen, C. J., and D. S. Chen. 2002. “Interaction of Hepatitis B Virus, Chemical Carcinogen, and Genetic Susceptibility: Aggarwal, R., U. C. Ghoshal, and S. R. Naik. 2002. “Treatment Multistage Hepatocarcinogenesis with Multifactorial of Chronic Hepatitis B with Interferon-Alpha: Cost- Etiology.” Hepatology (Baltimore, MD) 36 (5): 1046–49. Effectiveness in Developing Countries.” The National doi:10.1053/jhep.2002.37084. Medical Journal of India 15 (6): 320–27. http://www.ncbi Chen, C. J., M. W. Yu, and Y. F. Liaw. 1997. “Epidemiological .nlm.nih.gov/pubmed/12540064. Characteristics and Risk Factors of Hepatocellular Carcinoma.” ———. 2003. “Assessment of Cost-Effectiveness of Universal Journal of Gastroenterology and Hepatology 12 (9–10): Hepatitis B Immunization in a Low-Income Country with S294–308. http://www.ncbi.nlm.nih.gov/pubmed/9407350. Intermediate Endemicity Using a Markov Model.” Journal Chen, J.-G., J. Zhu, D. M. Parkin, Y. H. Zhang, J. H. Lu, and of Hepatology 38 (2): 215–22. http://www.ncbi.nlm.nih.gov others. 2006. “Trends in the Incidence of Cancer in Qidong, /pubmed/12547411. China, 1978–2002.” International Journal of Cancer 119 (6): Ahmad, K. 2004. “Pakistan: A Cirrhotic State?” The Lancet 364 1447–54. doi:10.1002/ijc.21952. (9448): 1843–44. doi:10.1016/S0140-6736(04)17458-8. Chen, J.-G., P. A. Egner, D. Ng, L. P. Jacobson, and others. 2013. Alcorn, T. 2011. “Mongolia’s Struggle with Liver Cancer.” The “Reduced Aflatoxin Exposure Presages Decline in Liver Lancet 377 (9772): 1139–40. http://www.ncbi.nlm.nih.gov Cancer Mortality in an Endemic Region of China.” Cancer /pubmed/21465699. Prevention Research 6: 1038. doi:10.1158/1940-6207. Allen, N. E., V. Beral, D. Casabonne, S. W. Kan, G. K. Reeves, Chiang, C.-J., Y.-A. Yang, S.-L. You, M.-S. Lai, and C.-J. Chen. and others. 2009. “Moderate Alcohol Intake and Cancer 2013. “Thirty-Year Outcomes of the National Hepatitis B Incidence in Women.” Journal of the National Cancer Immunization Program in Taiwan.” JAMA 310 (9): 974–76. Institute 101 (5): 296–305. http://www.ncbi.nlm.nih Colin, C., D. Lanoir, S. Touzet, L. Meyaud-Kraemer, F. Bailly, .gov/pubmed/19244173. and C. Trepo. 2001. “Sensitivity and Specificity of Third- Alter, H. J., and L. B. Seeff. 2000. “Recovery, Persistence, and Generation Hepatitis C Virus Antibody Detection Assays: Sequelae in Hepatitis C Virus Infection: A Perspective on An Analysis of the Literature.” Journal of Viral Hepatitis 8 (2): Long-Term Outcome.” Seminars in Liver Disease 20 (1): 87–95. http://www.ncbi.nlm.nih.gov/pubmed/11264728. 17–35. http://www.ncbi.nlm.nih.gov/pubmed/10895429. Corrao, G., and S. Arico. 2000. “Alcohol and Liver Cirrhosis: Aman, W., S. Mousa, G. Shiha, and S. A. Mousa. 2012. Old Questions and Weakly Explored Opportunities.” “Current Status and Future Directions in the Management Addiction 95 (8): 1267–70. http://doi.wiley.com/10.1046 of Chronic Hepatitis C.” Virology Journal 9: 57. http://www /j.1360-0443.2000.958126715.x. .pubmedcentral.nih.gov/articlerender.fcgi?artid=3325870 Corrao, G., V. Bagnardi, A. Zambon, and C. La Vecchia. 2004. &tool=pmcentrez&rendertype=abstract. “A Meta-Analysis of Alcohol Consumption and the Risk Atehnkeng, J., P. S. Ojiambo, T. Ikotun, R. A. Sikora, P. J. Cotty, and of 15 Diseases.” Preventive Medicine 38 (5): 613–19. http:// others. 2008. “Evaluation of Atoxigenic Isolates of Aspergillus www.ncbi.nlm.nih.gov/pubmed/15066364. Flavus as Potential Biocontrol Agents for Aflatoxin in Maize.” CPSTF (Community Preventive Services Task Force). 2013. Food Additives & Contaminants. Part A, Chemistry, Analysis, “Preventing Excessive Alcohol Consumption.” The Control, Exposure & Risk Assessment 25 (10): 1264–71. http:// Community Guide [Internet]. U.S. Government. http:// www.ncbi.nlm.nih.gov/pubmed/18608502. www.thecommunityguide.org/alcohol/index.html. Attia, M. A. 1998. “Prevalence of Hepatitis B and C in Egypt Crowley, S. J., D. Tognarini, P. V. Desmond, and M. Lees. 2000. and Africa.” Antiviral Therapy 3 (Suppl. 3): 1–9. http://www “Cost-Effectiveness Analysis of Lamivudine for the Treatment .ncbi.nlm.nih.gov/pubmed/10726051. of Chronic Hepatitis B.” PharmacoEconomics 17 (5): 409–27. Baffy, G., E. M. Brunt, and S. H. Caldwell. 2012. “Hepatocellular http://www.ncbi.nlm.nih.gov/pubmed/10977384. Carcinoma in Non-Alcoholic Fatty Liver Disease: An Cui, F., L. Li, S. C. Hadler, F. Wang, H. Zheng, and others. 2010. Emerging Menace.” Journal of Hepatology 56 (6): 1384–91. “Factors Associated with Effectiveness of the First Dose of doi:10.1016/j.jhep.2011.10.027. Hepatitis B Vaccine in China: 1992–2005.” Vaccine 28 (37): Boffetta, P., and M. Hashibe. 2006. “Alcohol and Cancer.” The 5973–78. http://www.ncbi.nlm.nih.gov/pubmed/20637773. Lancet Oncology 7 (2): 149–56. http://www.ncbi.nlm.nih Dienstag, J. L. 2008. “Hepatitis B Virus Infection.” New England .gov/pubmed/16455479. Journal of Medicine 359 (14): 1486–500. doi:10.1056 Burt, C. M. 2000. “Selection of Irrigation Methods for /NEJMra0801644. http://www.ncbi.nlm.nih.gov/pubmed Agriculture.” Environmental and Water Resources /18832247. Institute (U.S.). On-Farm Irrigation Committee. Reston, Dondog, B., M. Lise, O. Dondov, B. Baldandorj, and VA: American Society of Civil Engineers. http://mobius S. Franceschi. 2011. “Hepatitis B and C Virus Infections .missouri.edu:2082/record=b14299016~S0. in Hepatocellular Carcinoma and Cirrhosis in Mongolia.” Chanzy, B., D. L. Duc-Bin, B. Rousset, P. Morand, C. Morel- European Journal of Cancer Prevention 20 (1): 33–39. http:// Baccard, and others. 1999. “Effectiveness of a Manual www.torna.do/s/Dondov-O/. Disinfection Procedure in Eliminating Hepatitis C Dorner, J. W., and M. C. Lamb. 2006. “Development and Virus from Experimentally Contaminated Endoscopes.” Commercial Use of Afla-Guard(®), an Aflatoxin Biocontrol Gastrointestinal Endoscopy 50 (2): 147–51. http://www.ncbi Agent.” Mycotoxin Research 22 (1): 33–38. http://www.ncbi .nlm.nih.gov/pubmed/10425404. .nlm.nih.gov/pubmed/23605499. Liver Cancer 161 Evans, A. A., A. P. O’Connell, J. C. Pugh, W. S. Mason, Hartwell, D., and J. Shepherd. 2009. “Pegylated and Non- F. M. Shen, and others. 1998. “Geographic Variation in Pegylated Interferon-Alfa and Ribavirin for the Treatment Viral Load among Hepatitis B Carriers with Differing of Mild Chronic Hepatitis C: A Systematic Review and Risks of Hepatocellular Carcinoma.” Cancer Epidemiology, Meta-Analysis.” International Journal of Technology Biomarkers & Prevention 7 (7): 559–65. http://www.ncbi Assessment in Health Care 25 (1): 56–62. doi:10.1017 .nlm.nih.gov/pubmed/9681522. /S0266462309090084. Ferlay, J., I. Soerjomataram, M. Ervik, R. Dikshit, S. Eser, and Hill, A., S. Khoo, J. Fortunak, B. Simmons, and N. Ford. others. 2013. “GLOBOCAN v1.0, Cancer Incidence and 2014. “Minimum Costs for Producing Hepatitis C Direct- Mortality Worldwide: IARC CancerBase No. 11” [Internet]. Acting Antivirals for Use in Large-Scale Treatment Access International Agency for Research on Cancer, Lyon, France. Programs in Developing Countries.” Clinical Infectious http://globocan.iarc.fr. Diseases 58 (7): 928–36. doi:10.1093/cid/ciu012. Forman D., F. Bray, D. H. Brewster, C. Gombe Mbalawa, Hsieh, Y.-C., W.-S. Lee, P.-L. Shao, L.-Y. Chang, and L.-M. Huang. B. Kohler, M. Piñeros, E. Steliarova-Foucher, R. Swaminathan, 2007. “The Transforming Streptococcus Pneumoniae in the and J. Ferlay, eds. 2013. Cancer Incidence in Five Continents, 21st Century.” Chang Gung Medical Journal 31 (2): 117–24. Vol. X (electronic version). IARC, Lyon. http://ci5.iarc.fr. http://www.ncbi.nlm.nih.gov/pubmed/18567411. Fusco, M., E. Girardi, P. Piselli, R. Palombino, J. Polesel, and Huang, Y-T., C-L. Jen, H-I. Yang, M-H. Lee, J. Su, and others. others. 2008. “Epidemiology of Viral Hepatitis Infections 2011. “Lifetime Risk and Sex Difference of Hepatocellular in an Area of Southern Italy with High Incidence Rates of Carcinoma among Patients with Chronic Hepatitis B Liver Cancer.” European Journal of Cancer 44 (6): 847–53. and C.” Journal of Clinical Oncology 29 (27): 3643–50. http://www.ncbi.nlm.nih.gov/pubmed/18313290. doi:10.1200/JCO.2011.36.2335. Gao, X., Q. Cui, X. Shi, J. Su, Z. Peng, and others. 2011. Hutton, D. W., S. K. So, and M. L. Brandeau. 2010. “Cost- “Prevalence and Trend of Hepatitis C Virus Infection Effectiveness of Nationwide Hepatitis B Catch-up among Blood Donors in Chinese Mainland: A Systematic Vaccination among Children and Adolescents in China.” Review and Meta-Analysis.” BMC Infectious Diseases 11: Hepatology (Baltimore, MD) 51 (2): 405–14. http://www 88. http://www.pubmedcentral.nih.gov/articlerender.fcgi? .pubmedcentral.nih.gov/articlerender.fcgi?artid=3245734 artid=3079653&tool=pmcentrez&rendertype=abstract. &tool=pmcentrez&rendertype=abstract. Gong, Y. Y., K. Cardwell, A. Hounsa, S. Egal, P. C. Turner, IARC (International Agency for Research on Cancer). 2002. and others. 2002. “Dietary Aflatoxin Exposure and “Some Traditional Herbal Medicines, Some Mycotoxins, Impaired Growth in Young Children from Benin and Naphthalene and Styrene.” IARC Monographs on the Togo: Cross Sectional Study.” BMJ (Clinical Research Evaluation of Carcinogenic Risks to Humans, Volume 82. Ed.) 325 (7354): 20–21. http://www.pubmedcentral.nih. IARC, Lyon, France. gov/articlerender.fcgi?artid=116667&tool=pmcentrez&- ———. 2009. “Biological Agents.” IARC Monographs on the rendertype=abstract. Evaluation of Carcinogenic Risks to Humans, Volume 100B. Gong, Y., A. Hounsa, S. Egal, P. C. Turner, A. E. Sutcliffe, IARC, Lyon, France. and others. 2004. “Postweaning Exposure to Aflatoxin ———. 2012a. “A Review of Human Carcinogens. E. Personal Results in Impaired Child Growth: A Longitudinal Study Habits and Indoor Combustions.” Monographs on the in Benin, West Africa.” Environmental Health Perspectives Evaluation of Carcinogenic Risks to Humans, Volume 100: 112 (13): 1334–38. http://www.pubmedcentral.nih.gov 585. IARC, Lyon, France. /articlerender.fcgi?artid=1247526&tool=pmcentrez&rend- ———. 2012b. “A Review of Human Carcinogens: Chemical ertype=abstract. Agents and Related Occupations.” Monographs on the Gong, Y. Y., S. Wilson, J. K. Mwatha, M. N. Routledge, Evaluation of Carcinogenic Risks to Humans, 100F. http:// J. M. Castelino, and others. 2012. “Aflatoxin Exposure May monographs.iarc.fr/ENG/Monographs/vol100F/index.php. Contribute to Chronic Hepatomegaly in Kenyan School Jongsuksuntigul, P., and T. Imsomboon. 2003. “Opisthorchiasis Children.” Environmental Health Perspectives 120 (6): Control in Thailand.” Acta Tropica 88 (3): 229–32. 893–96. http://www.pubmedcentral.nih.gov/articlerender doi:10.1016/j.actatropica.2003.01.002. .fcgi?artid=3385435&tool=pmcentrez&rendertype=abstract. Kermode, M. 2004. “Unsafe Injections in Low-Income Country Goodnough, L. T. 2013. “Blood Management: Transfusion Health Settings: Need for Injection Safety Promotion Medicine Comes of Age.” The Lancet 381 (9880): 1791–92. to Prevent the Spread of Blood-Borne Viruses.” Health http://www.ncbi.nlm.nih.gov/pubmed/23706789. Promotion International 19 (1): 95–103. http://www.ncbi Groopman, J. D., T. W. Kensler, and C. P. Wild. 2008. .nlm.nih.gov/pubmed/14976177. “Protective Interventions to Prevent Aflatoxin-Induced Khlangwiset, P., and F. Wu. 2010. “Costs and Efficacy of Public Carcinogenesis in Developing Countries.” Annual Review Health Interventions to Reduce Aflatoxin-Induced Human of Public Health 29: 187–203. http://www.ncbi.nlm.nih.gov Disease.” Food Additives & Contaminants: Part A 27 (7): /pubmed/17914931. 998–1014. doi:10.1080/19440041003677475 Haenel, H. 1989. “Alcohol Drinking.” IARC Monographs on the Kuniholm, M. H., O. A. Lesi, M. Mendy, A. O. Akano, Evaluation of Carcinogenic Risks to Humans. Volume 44. O. Sam, and others. 2008. “Aflatoxin Exposure and Viral IARC, Lyon, France. Hepatitis in the Etiology of Liver Cirrhosis in The Gambia, 162 Cancer West Africa.” Environmental Health Perspectives 116 (11): NIAAA (National Institute on Alcohol Abuse and Alcoholism). 1553–57. http://www.pubmedcentral.nih.gov/articlerender 1993. “Alcohol and Cancer.” Alcohol Alert No. 21-1993. .fcgi? artid=2592277&tool=pmcentrez&rendertype http://pubs.niaaa.nih.gov/publications/aa21.htm. =abstract. Omelyanovsky, V. V., M. V. Avksentieva, I. Krysanov, and Lang, H.-C., J.-C. Wu, S.-H. Yen, C.-F. Lan, and S.-L. Wu. 2008. O. Ivakhnenko. 2010. “PCN53 Analysis of Socioeconomic “The Lifetime Cost of Hepatocellular Carcinoma: A Claims Burden of Hepatocellular Carcinoma in Russia.” Data Analysis from a Medical Centre in Taiwan.” Applied Value in Health 13 (7): A260–61. doi:10.1016/S1098- Health Economics and Health Policy 6 (1): 55–65. http:// 3015(11)71948-9. http://www.valueinhealthjournal.com www.ncbi. nlm.nih.gov/pubmed/18774870. /article/S1098-3015(11)71948-9. Lu, S.-N., W.-W. Su, S.-S. Yang, T.-T. Chang, K.-S. Cheng, and Omer, R. E., A. Kuijsten, A. M. Y. Kadaru, F. J. Kok, M. O. Idris, others. 2006. “Secular Trends and Geographic Variations and others. 2004. “Population-Attributable Risk of Dietary of Hepatitis B Virus and Hepatitis C Virus-Associated Aflatoxins and Hepatitis B Virus Infection with Respect to Hepatocellular Carcinoma in Taiwan.” International Journal Hepatocellular Carcinoma.” Nutrition and Cancer 48 (1): of Cancer 119 (8): 1946–52. http://www.ncbi.nlm.nih.gov 15–21. http://www.ncbi.nlm.nih.gov/pubmed/15203373. /pubmed/16708389. Oyunsuren, T., F. Kurbanov, Y. Tanaka, A. Elkady, R. Sanduijav, Lui, Y. Y., K. K. Tsoi, V. W. Wong, J.-H. Kao, J.-L. Hou, and oth- and others. 2006. “High Frequency of Hepatocellular ers. 2010. “Cost-Effectiveness Analysis of Roadmap Models Carcinoma in Mongolia; Association with Mono-, or in Chronic Hepatitis B Using Tenofovir as the Rescue Co-Infection with Hepatitis C, B, and Delta Viruses.” Journal Therapy.” Antiviral Therapy 15 (2): 145–55. http://www of Medical Virology 78 (12): 1688–95. doi:10.1002/jmv.20755. .ncbi.nlm.nih.gov/pubmed/20386069. Pitt, J. I., C. P. Wild, R. A. Baan, W. C. A. Gelderblom, Marinier, E., V. Barrois, B. Larouze, W. T. London, A. Cofer, and J. D. Miller, R. T. Riley, and F. Wu. 2012. “Improving others. 1985. “Lack of Perinatal Transmission of Hepatitis Public Health through Mycotoxin Control.” IARC B Virus Infection in Senegal, West Africa.” Journal of Scientific Publication, No. 158. IARC, Lyon, France. http:// Pediatrics 106 (5): 843–49. http://aje.oxfordjournals.org apps.who.int/bookorders/anglais/detart1 .jsp?codlan=1 /content/147/5/478.full.pdf. &codcol=73&codcch=158#. Matsumoto, A., E. Tanaka, A. Rokuhara, K. Kiyosawa, Prati, D. 2006. “Transmission of Hepatitis C Virus by Blood H. Kumada, and others. 2005. “Efficacy of Lamivudine for Transfusions and Other Medical Procedures: A Global Preventing Hepatocellular Carcinoma in Chronic Hepatitis Review.” Journal of Hepatology 45 (4): 607–16. http://www B: A Multicenter Retrospective Study of 2795 Patients.” .ncbi.nlm.nih.gov/pubmed/16901579. Hepatology Research 32 (3): 173–84. http://www.ncbi.nlm Qian, G. S., R. K. Ross, M. C. Yu, J. M. Yuan, Y. T. Gao, and .nih.gov/pubmed/16024289. others. 2013. “A Follow-up Study of Urinary Markers of Mehl, H. L., R. Jaime, K. A. Callicott, C. Probst, N. P. Garber, Aflatoxin Exposure and Liver Cancer Risk in Shanghai, and others. 2012. “Aspergillus Flavus Diversity on Crops People’s Republic of China.” Cancer Epidemiology, and in the Environment Can Be Exploited to Reduce Biomarkers & Prevention 3 (1): 3–10. http://www.ncbi.nlm Aflatoxin Exposure and Improve Health.” Annals of the New .nih.gov/pubmed/8118382. York Academy of Sciences 1273: 7–17. http://www.ncbi.nlm Raza, S. A., G. M. Clifford, and S. Franceschi. 2007. “Worldwide .nih.gov/pubmed/23230832. Variation in the Relative Importance of Hepatitis B Michelotti, G. A., M. V. Machado, and A. M. Diehl. 2013. and Hepatitis C Viruses in Hepatocellular Carcinoma: “NAFLD, NASH and Liver Cancer.” Nature Reviews A Systematic Review.” British Journal of Cancer 96: 1127–34. Gastroenterology and Hepatology 10: 656–65. doi:10.1038 doi:10.1038/sj.bjc.6603649. /nrgastro.2013.183. Razavi, H., C. Estes, K. Pasini, E. Gower, and S. Hindman. 2013. Morgan, R. L., B. Baack, B. D. Smith, A. Yartel, M. Pitasi, “51 HCV Treatment Rate in Select European Countries in and Y. Falck-Ytter. 2013. “Eradication of Hepatitis C 2004–2010.” Journal of Hepatology 58 (null): S22–23. http:// Virus Infection and the Development of Hepatocellular dx.doi.org/10.1016/S0168-8278(13)60053-7. Carcinoma: A Meta-Analysis of Observational Studies.” Reid, S. 2012. “Estimating the Burden of Disease from Unsafe Annals of Internal Medicine 158 (5 Pt 1): 329–37. http:// Injections in India: A Cost-Benefit Assessment of the www.ncbi.nlm.nih.gov /pubmed/23460056. Auto-Disable Syringe in a Country with Low Blood- NCI (National Cancer Institute). 2013. “SEER Stat Fact Sheets: Borne Virus Prevalence.” Indian Journal of Community Liver and Intrahepatic Bile Duct.” U.S. National Institutes Medicine 37 (2): 89–94. http://www.pubmedcentral.nih of Health, Bethesda, MD. http://seer.cancer.gov/statfacts .gov/articlerender.fcgi?artid=3361807&tool=pmcentrez /html/livibd.html#survival. &rendertype=abstract. Nelson, P. K., B. M. Mathers, B. Cowie, H. Hagan, D. D. Jarlais, Shepherd, J., H. Brodin, C. Cave, N. Waugh, A. Price, and and others. 2011. “Global Epidemiology of Hepatitis B others. 2004. “Pegylated Interferon Alpha-2a and -2b in and Hepatitis C in People Who Inject Drugs: Results Combination with Ribavirin in the Treatment of Chronic of Systematic Reviews.” The Lancet 378 (9791): 571–83. Hepatitis C: A Systematic Review and Economic Evaluation.” http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid Health Technology Assessment (Winchester, England) 8 (39): =3285467&tool=pmcentrez&rendertype=abstract. iii–iv, 1–125. Liver Cancer 163 Shin, H.-R., J.-K. Oh, E. Masuyer, M.-P. Curado, V. Bouvard, Carcinoma in Taiwan.” International Journal of Cancer 67 (5): and others. 2010. “Epidemiology of Cholangiocarcinoma: 620–25. http://www.ncbi.nlm.nih.gov/pubmed/8782648. An Update Focusing on Risk Factors.” Cancer Science 101 Wang, L., Y. Wang, S. Jin, Z. Wu, D. P. Chin, and others. (3): 579–85. doi:10.1111/j.1349-7006.2009.01458.x. 2008. “Emergence and Control of Infectious Diseases in Sithithaworn, P., and M. Haswell-Elkins. 2003. “Epidemiology China.” The Lancet 372 (9649): 1598–605. doi:10.1016 of Opisthorchis Viverrini.” Acta Tropica 88 (3): 187–94. /S0140-6736(08)61365-3. http://www.ncbi.nlm.nih.gov/pubmed/14611873. WHO (World Health Organization). 2004. “UNICEF/WHO Smith, B. D., R. L. Morgan, G. A. Beckett, Y. Falck-Ytter, Workshop on the Expanded Programme on Immunizations D. Holtzman, and others. 2012. “Recommendations for the in the Pacific.” WHO, Regional Office for the Western Identification of Chronic Hepatitis C Virus Infection among Pacific. http://iris.wpro.who.int/handle/10665.1/1663. Persons Born during 1945–1965.” MMWR. Recommendations ———. 2008. “Global Information System on Alcohol and and Reports: Morbidity and Mortality Weekly Report. Centers Health.” WHO, Geneva. http://apps.who.int/globalatlas for Disease Control 61 (RR-4): 1–32. http://www.ncbi.nlm /default.asp. .nih.gov/pubmed/22895429. ———. 2009. “Weekly Epidemiological Record: Hepatitis B Sripa, B., J. M. Bethony, P. Sithithaworn, S. Kaewkes, E. Mairiang, Vaccines.” WHO Position Paper 40 (84): 405–20. and others. 2011. “Opisthorchiasis and Opisthorchis- Wild, C. P., and Y. Y. Gong. 2010. “Mycotoxins and Human Associated Cholangiocarcinoma in Thailand and Laos.” Disease: A Largely Ignored Global Health Issue.” Acta Tropica 120 (Suppl. 1): S158–68. doi:10.1016/j Carcinogenesis 31 (1): 71–82. http://www.pubmedcentral.nih .actatropica.2010.07.006. .gov/articlerender.fcgi?artid=2802673&tool= pmcentrez Sun, Z., T. Chen, S. S. Thorgeirsson, Q. Zhan, J. Chen, &rendertype=abstract. and others. 2013. “Dramatic Reduction of Liver Cancer Wild, C. P., and A. J. Hall. 2000. “Primary Prevention of Incidence in Young Adults: 28 Year Follow-Up of Hepatocellular Carcinoma in Developing Countries.” Etiological Interventions in an Endemic Area of China.” Mutation Research/Reviews in Mutation 462 (2–3): 381–93. Carcinogenesis 34 (8): 1800–05. http://www.ncbi.nlm.nih doi:10.1016/S1383-5742(00)00027-2. .gov/pubmed/23322152. Williams, J. H., T. D. Phillips, P. E. Jolly, J. K. Stiles, C. M. Jolly, Tanaka,Y., K. Hanada, M. Mizokami, A. E. T.Yeo, J. Wai-Kuo Shih, and D. Aggarwal. 2004. “Human Aflatoxicosis in Developing and others. 2002. “A Comparison of the Molecular Clock Countries: A Review of Toxicology, Exposure, Potential of Hepatitis C Virus in the United States and Japan Predicts Health Consequences, and Interventions.” American Journal That Hepatocellular Carcinoma Incidence in the United of Clinical Nutrition 80 (5): 1106–22. http://www.ncbi.nlm States Will Increase over the Next Two Decades.” Proceedings .nih.gov/pubmed/15531656. of the National Academy of Sciences 99 (24): 15584–89. Wu, F., and P. Khlangwiset. 2010a. “Evaluating the Technical http://www.pubmedcentral.nih.gov/articlerender.fcgi? Feasibility of Aflatoxin Risk Reduction Strategies in Africa.” artid=137760&tool=pmcentrez&rendertype=abstract. Food Additives & Contaminants. Part A, Chemistry, Analysis, Turner, P. C., A. Sylla, Y. Y. Gong, M. S. Diallo, A. E. Sutcliffe, Control, Exposure & Risk Assessment 27 (5): 658–76. and others. 2013. “Reduction in Exposure to Carcinogenic doi:10.1080/19440041003639582. Aflatoxins by Postharvest Intervention Measures in West ———. 2010b. “Health Economic Impacts and Cost- Africa: A Community-Based Intervention Study.” The Effectiveness of Aflatoxin-Reduction Strategies in Africa: Lancet 365 (9475): 1950–56. http://www.ncbi.nlm.nih.gov Case Studies in Biocontrol and Post-Harvest Interventions.” /pubmed/15936422. Food Additives & Contaminants. Part A, Chemistry, Analysis, Vanni, E., and E. Bugianesi. 2014. “Obesity and Liver Cancer.” Control, Exposure & Risk Assessment 27 (4): 496–509. http:// Clinics in Liver Disease 18 (1): 191–203. doi:10.1016/j www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2858 .cld.2013.09.001. 428&tool=pmcentrez&rendertype=abstract. Wang, J.-F., W.-Y. Lin, F. Jiang, W. Meng, and F.-M. Shen. 2010. Wu, H.-C., Q. Wang, H.-I. Yang, H. Ahsan, W.-Y. Tsai, and “Analysis of Time Trend of Hepatocellular Carcinoma others. 2009. “Aflatoxin B1 Exposure, Hepatitis B Virus Mortality in Haimen City of Jiangsu Province from 1993 Infection, and Hepatocellular Carcinoma in Taiwan.” Cancer to 2006.” Zhonghua Liu Xing Bing Xue Za Zhi = Zhonghua Epidemiology, Biomarkers & Prevention 18 (3): 846–53. Liuxingbingxue Zazhi 31 (7): 727–32. http://www.ncbi.nlm doi:10.1158/1055-9965.EPI-08-0697. .nih.gov/pubmed/21162831. Yuan, Y., U. H. Iloeje, J. Hay, and S. Saab. 2008. “Evaluation Wang, J.-T., T.-H. Wang, J.-T. Lin, C.-Z. Lee, J.-C. Sheu, and of the Cost-Effectiveness of Entecavir versus Lamivudine others. 2013. “Effect of Hepatitis C Antibody Screening in in Hepatitis BeAg-Positive Chronic Hepatitis B Patients.” Blood Donors on Post-Transfusion Hepatitis in Taiwan.” Journal of Managed Care Pharmacy 14 (1): 21–33. Journal of Gastroenterology and Hepatology 10 (4): 454–58. Zhang, Y.-J., and M.-S. Chen. 2010. “Role of Radiofrequency http://www.ncbi.nlm.nih.gov/pubmed/8527713. Ablation in the Treatment of Small Hepatocellular Wang, L.-Y., M. Hatch, C.-J. Chen, B. Levin, S.-L. You, and Carcinoma.” World Journal of Heptalogy 2 (4): 146–50. others. 1996. “Aflatoxin Exposure and Risk of Hepatocellular doi:10.4254/wjh.v2.i4.146. 164 Cancer Chapter 9 Cancer Pain Relief James Cleary, Hellen Gelband, and Judith Wagner INTRODUCTION effective (see Foley and others 2006 for a comprehensive listing). All but analgesic medications and psychological Despite substantial effort and expenditure, at least one- approaches require access to well-developed health care third of patients diagnosed with cancer in high-income facilities; these are usually available in large urban areas countries (HICs) die of their disease within a few years of of middle-income countries (MICs), although not nec- diagnosis (National Cancer Institute 2009). In low- and essarily in sufficient numbers, but they may not exist at middle-income countries (LMICs), two-thirds succumb, all in low-income countries (LICs). For example, many because the cancer types prevalent in LMICs tend to countries have no radiotherapy centers, and many have have poor prognoses, most cancers are advanced when only one center (International Atomic Energy Agency diagnosed, and even for curable cancers few people have Directory of Radiotherapy Centres, http://www-naweb access to effective cancer treatment. For rich and poor .iaea.org/nahu/dirac/default.asp). Palliative surgery and everywhere, cancer can cause pain and severe distress, palliative radiotherapy are discussed further in chapters especially during the last few months of life. Cancer- 13 and 14, respectively. The focus of this chapter is pain related pain is not the exclusive domain of those who control medication, which can relieve most cancer pain die of cancer. Even many who are cured of their disease and can be delivered at home, even in remote areas. live with the long-term effects of the disease and its treat- Since 1990, the World Health Organization (WHO) ment; many of them live with pain, as do people with a and other bodies have offered definitions of palliative range of chronic conditions other than cancer. care. These definitions differ in specifics but share a For the majority of cancer patients in LMICs, the common vision of care that emphasizes effective pain most effective and feasible intervention for pain con- relief and a team approach to care throughout the trol is medication. For mild pain, over-the-counter, course of the illness (Cleary and Carbone 1997; Foley inexpensive analgesic medicines can provide adequate and Gelband 2001; Morrison and Meier 2011; WHO relief. When these nonopioids no longer relieve pain, 1990). The primary goal of palliative care is improving then weak opioids, such as codeine, may work. Cancer the quality of life of patients and those around them; it patients most often experience worsening pain as their is not the prolongation of life or the hastening of death. cancer progresses; 70–80 percent progress to severe pain, Access to pain relief has been declared a human right which only strong opioid medicines can relieve. (Brennan, Carr, and Cousins 2007; Gwyther, Brennan, Other approaches are effective for specific pain and Harding 2009; International Pain Summit of the indications; the most widely applicable are palliative International Association for the Study of Pain 2011; radiotherapy and surgery. Chemotherapy, neuro- Lohman, Schleifer, and Amon 2010). logic, psychological, and other approaches also can be Corresponding author: James Cleary, MD, FRACP, FAChPM, University of Wisconsin, jfcleary@wisc.edu 165 From a global perspective, the growth of palliative Pain Society has advocated an approach based on the care has been largely limited to HICs, which also rank mechanism of pain together with its severity, but this high on the Human Development Index (HDI). The group also emphasized that all patients with cancer availability of palliative care—using the availability of should have access to opioids, as needed (Miaskowski opioid medicines as a surrogate—is correlated with a and others 2005). country’s HDI. At the low end, the availability is almost How prevalent is the need for pain relief among nil, and repeated surveys have shown that this availabil- cancer patients in LMICs? Cancer deaths are rising ity changed only marginally between 2006 and 2011 throughout the world as progress is made against (Gilson and others 2013). infectious diseases and as the world’s population ages. In the previous edition of Disease Control Priorities In 2012, 5.3 million people died of cancer in LMICs, in Developing Countries, Foley and others (2006) doc- compared with 2.9 million in HICs. In Sub-Saharan umented the global problem of low access to adequate Africa, 715,000 new cases and 542,000 cancer deaths pain relief in LMICs. Since then, a few countries have occurred in 2008; these numbers are projected to improved access, but these accomplishments are spo- nearly double by 2030 due to population growth and radic; in many countries, the change is negligible. Now, aging (Ferlay and others 2010). The consensus among there is both cause for optimism and the view of a long researchers is that 60–90 percent of patients with road ahead. Efforts to support leaders in reforming advanced cancer experience moderate to severe pain policy and clinical practice in LMICs have grown and (Cleeland and others 1988; Cleeland and others 1996; provide a basis for improvements (Cherny and others Daut and Cleeland 1982; Foley 1979, 1999; Stjernsward 2013; Cleary, Radbruch, and others 2013). and Clark 2003). The intensity, degree of pain relief, This chapter describes the current state of and effect of pain vary according to the type of cancer, pain relief in LMICs, consistent with WHO’s use treatment, and personal characteristics, but prevalence of opioid consumption as a surrogate for access to and severity of pain usually increase with the progres- palliative care in the Global Monitoring Framework sion of the disease. for Noncommunicable Diseases (WHO 2013a). We Foley and others (2006) estimated that about describe the gaps in pain control access across coun- 80 percent of people dying of cancer would experience tries, analyze the barriers to improving its delivery, and moderate or severe pain that requires opioid medication describe the costs and benefits that might accrue from for relief for an average of 90 days before death. This removing the barriers. estimate amounts to 425 million days of cancer pain Evidence summarized in this chapter focuses on that could be relieved by opioids in LMICs each year the modest costs and substantial benefits of providing currently. pain control, and it supports increased efforts in the short term. Pain control medication and other aspects of palliative care can lead, rather than follow, increased efforts in cancer treatment, relying on interventions STATUS OF PAIN CONTROL IN LMICs that are part of a more advanced cancer control and In an ideal world, palliative care and pain control would treatment infrastructure. be one component of a cancer care system, but in nearly all LICs and for the rural poor in many MICs, palliative care—if it exists at all—is more likely to be indepen- CANCER PATIENTS’ NEED FOR PAIN dent of cancer services, and patients receive little or no primary cancer treatment. Unfortunately, despite the CONTROL MEDICATION inclusion of morphine and codeine on WHO’s essential Patients with cancer usually experience pain at some medicines list (WHO 2013b), the programs to deliver points during their illness, increasingly toward the end. them are likely to be undeveloped, and patients go Mild and moderate pain often can be controlled with without relief. commonly available analgesics, such as acetaminophen In 2006, 66 percent of the world’s population lived and ibuprofen, but progressively stronger medications in countries that had virtually no consumption of are needed to control cancer pain as the illness pro- opioids, 10 percent in countries with very low con- gresses and pain becomes more severe. Opioids—such sumption, 3 percent in countries with low consumption, as codeine and morphine—are invariably needed toward and 4 percent in countries with moderate consump- the end of life. This progression is embodied in WHO’s tion (Seya and others 2011). Only 7.5 percent of the Three-Step Analgesic Ladder, developed in 1986 to guide world’s population lived in countries with consumption clinicians (WHO 1986). More recently, the American levels defined as adequate. The level of adequacy of 166 Cancer access for a country was highly correlated with its HDI Figure 9.1 Comparison of Opioid Consumption in Morphine Equivalents (R2 = 0.7583) (Seya and others 2011). without Methadone To what extent is the need for cancer pain relief met (mg per capita) under these circumstances? Even with increases in cer- 45 tain areas, the starting levels are so low that the most (mg/capita, mor equiv without methadone) recent levels are only a fraction of the per capita use in 40 HICs (figure 9.1). 35 This global assessment is supported by new data Opioid consumption 30 from sites in 26 countries representing all World Bank income levels. The International Association 25 for Hospice and Palliative Care (IAHPC) conducted 20 its first round of the Opioid Price Watch (De Lima and others 2014), which reports on availability of 15 opioids, as well as the prices that consumers pay for 10 those medicines. The survey was conducted in the following manner: surveyors visited the pharmacy 5 closest to a public health facility treating patients 0 with life-threatening conditions. If that pharmacy 80 82 84 86 88 90 92 94 96 98 00 02 04 06 08 10 19 19 19 19 19 19 19 19 19 19 20 20 20 20 20 20 had no opioids in stock, or if the chief pharmacist did not wish to participate, the surveyors visited the World AMRO-North America WPRO next closest pharmacy; this process continued until at AFRO EMRO SEARO least one opioid was found or the surveyor concluded Source: Cleary, Radbruch, and others 2013. that none would be available anywhere. At least one Note: AFRO = Africa; AMRO-North America = Latin America and Caribbean (America not including opioid was available in the first pharmacy sampled in North America); EMRO = Middle East (Eastern Mediterranean); SEARO = Southeast Asia; WPRO = Western Pacific. all seven HICs. In three of the eight low-income sites in six countries (including three sites in Tanzania), no opioids were found, even after visiting an average of 4.5 pharmacies; where they were found, many fewer kinds were available than in HICs. In three of the sam- 4a. Hospice-palliative care services at a stage of prelimi- ple countries—Moldova, Nepal, and Sudan—opioids nary integration into mainstream service provision for outpatients were available only from hospital 4b. Hospice-palliative care services at a stage of pharmacies with permission to dispense them, limiting advanced integration into mainstream service access geographically. No information was available on provision. consumption in Moldova or Sudan, and very low con- sumption was reported per capita in Nepal, suggesting Palliative care was in stage 4a or 4b primarily in HICs, that most cancer patients have no access. with only a handful of LMICs (mainly MICs) in the highest category. In 2011, most countries still had no services, were in a capacity-building mode, or had only isolated services. AVAILABILITY OF BROADLY DEFINED PALLIATIVE CARE SERVICES The level of palliative care services available to can- BARRIERS TO PAIN CONTROL IN LMICs cer patients was assessed in 2007 by the International The near-total lack of access to opioid drugs for pain Observatory of End of Life Care (Clark and others relief in most LICs and many MICs is best understood 2007) and again in 2011 (Lynch, Connor, and Clark by examining the barriers to their supply (Cherny and 2013). Four categories were defined in 2007, and two others 2013; Foley and others 2006; Sloan and Gelband subcategories were added in 2011 (designated 3b and 2007). In HICs, modern palliative care, including 4b, following): access to opioid medicines and other methods of pain control, has often developed as an adjunct to cancer 1. No known hospice-palliative care activity care programs. In LICs and many MICs, the same 2. Capacity-building activity pattern has been difficult to follow because care pro- 3a. Isolated palliative care provision grams simply do not exist in most places. However, 3b. Generalized palliative care provision cancer control programs are not the only things that Cancer Pain Relief 167 are absent, sparse, or overstretched. Medical and phar- availability of opioids for the relief of pain. Often, macy resources are lacking in numbers and quality; the logistics of pain treatment with opioids is so medicines are costly and are most often paid for out burdensome or complex for physicians, nurses, and of pocket. Patients in some places are reluctant to use pharmacists that they assume it is an impossible task addictive drugs, even in the last stages of illness, and and do not pursue it. This problem has been recog- clinicians who have received little training in appropri- nized since the 1980s by INCB, WHO, the Council ate prescribing for cancer patients may reinforce this of Europe, and Human Rights Watch (Cherny and reluctance. others 2010). The greatest and most widespread barrier, how- In 2000, WHO, in collaboration with INCB, devel- ever—by far—comes from national regulations to con- oped guidelines for national authorities to scrutinize trol the nonmedical use of narcotics. These regulations their regulatory systems for barriers that could impede make it difficult or impossible for clinicians to prescribe access (WHO 2000). The movement to diagnose barriers and patients to obtain opioids for relief of cancer pain. to access has led to several exemplary national reforms. These regulations have been put in place in response It appears that the very act of diagnosing regulatory to the Single Convention on Narcotic Drugs of 1961 impediments to opioids for cancer pain is a strong first (INCB 1961) and amended by the 1972 Protocol. The step toward reform. Single Convention is an international treaty to ensure The international organizations that have collab- the availability of opioids for medical and scientific orated in the Global Opioid Policy Initiative (GOPI)1 needs while preventing the illicit production of, traf- have published a detailed country-specific analysis in ficking in, and nonmedical use of narcotic drugs. The five regions of the world of the availability and regula- Convention established the International Narcotics tory restrictions on seven opioid analgesic formulations Control Board (INCB) in 1968 as an independent, (Cleary, De Lima, and others 2013; Cleary, Powell, and quasi-judicial organization to implement the Single others 2013; Cleary, Radbruch, and others 2013): Convention. The Single Convention requires that all governments, • Codeine even nonsignatories, estimate the amounts of opioids • Immediate-release morphine (liquid or tablet) needed for medical and scientific purposes and report • Controlled-release oral morphine annually on imports, exports, and consumption. It sets • Injectable morphine out the following principles on which countries can base • Oxycodone their own policies and regulations: • Transdermal fentanyl • Immediate-release methadone • Individuals must be authorized to dispense opioids by virtue of their professional license or be specially IAHPC deems these agents essential for treatment of licensed to do so. pain in palliative care (De Lima and Doyle 2007). • Opioids may be transferred only between authorized Respondents commonly reported that these opioids parties. were included in national formularies. The reality, how- • Opioids may be dispensed only with a medical ever, was substantial variation in true availability and prescription. many regulatory barriers to access. • Security and records are required. Although the picture we present is lack of progress on a broad scale, we can point to positive developments Despite the Convention’s recognition of the need that have improved the availability or delivery of opioid for opioid use in pain control—and the demonstra- medicines in particular situations (Gilson and others tion by organizations such as Hospice Africa Uganda 2013). Cost has been reduced by the domestic man- that the needs of patients can be met even in remote ufacture of morphine tablets or solutions in Ethiopia areas while respecting these limits—the concern over (2010), Jordan (2004), and Vietnam (2009). In Kenya illegal narcotics has tipped the scale against the legit- (2010), a tax on morphine powder was eliminated. In imate medical needs of patients. Many countries have Ethiopia, India, Kenya, and Vietnam, providers and established regulations that go well beyond those policy makers have been educated about the use of and required by the Convention. Overzealous drug con- need for opioids. trollers and poorly considered laws and regulations National cancer policies or palliative care policies to restrict the diversion of medicinal opioids into have been created in a number of countries, laying the illicit markets profoundly interfere with the medical groundwork for greater opioid availability. This has 168 Cancer been the case in Ethiopia (2004), Kenya (2011), Nepal care costs, compared with usual cancer care, because it (2009), and Rwanda (2010). Jamaica took steps in lowers hospitalization rates for patients who are termi- 2010 and 2011 to clarify and facilitate opioid distribu- nally ill (Simoens and others 2010). No specific model tion through improved regulation. In Uganda (2004), of palliative care appears to be superior to any other the National Drug Policy and Authority statute was (García-Pérez and others 2009). amended to allow specially trained palliative care nurses An important difference exists in the availability and clinical officers to prescribe morphine, greatly of pain relief in HICs and LMICs, however. Even expanding the pool of providers and thus the popula- when palliative care programs do not exist in HICs, tion with access. medications for pain relief are usually available to In Georgia (2008), Mongolia (2004), and Vietnam patients with cancer through conventional medical (2008), the number of days allowed for an opioid care. Patients in LMICs are often unable to obtain prescription was increased; in some cases, the eligible such medications in any health care setting. Yet, it is patient population was expanded, for example, from telling that even when pain medications are readily only patients with the most advanced cancers to those available through conventional cancer care, organized with certain types of acute and chronic pain. programs for palliative care can enhance the effective- ness of pain control and lower the costs to the medical care system. Health care providers who focus on and ECONOMICS OF PALLIATIVE CARE IN LMICs are trained in cancer pain relief can reduce health care costs, at least in HICs, and improve the quality We consider economics from two perspectives: that of of life for patients and their caregivers (Amery and the health care system and that of the patients. In coun- others 2009). tries with comprehensive, publicly funded health care, Most patients in LICs and many in MICs do not the governments pay for all or most of the services and routinely receive care for cancer in hospitals; many commodities needed for palliative care. The patients pay receive no cancer-specific care at all. Where this is the either nothing or a predetermined copay. case, improving access to pain relief at the end of life In settings in which the public sector does not yet may increase health care costs, because adding care to provide or subsidize the cost of services, patients may a baseline of little or no care clearly involves new costs. be responsible for all costs. The costs may be mainly For the minority of cancer patients who receive end- market-driven, but they also may include taxes, tar- of-life care in hospitals, improving the availability of iffs, and other government add-ons. In the case of pain medicine through dismantling regulatory barriers, palliative care, the cost of pain control medications educating professionals, and integrating palliative care is most important. Patients everywhere inevitably programs may reduce health care costs, as such measures bear other costs, including transportation and the have done in HICs. opportunity costs of family caregivers who miss work. A useful first step in assessing the cost of palliative How high a barrier these costs impose varies with the care to a health system is to assess the cost of the most economic situation of the families and the support basic oral opioid medicine—oral morphine. Foley and available. colleagues estimated the cost of oral morphine, and medicines to treat side effects, sufficient for pain relief in the last three months of life in three countries— Health System Perspective Chile, Romania, and Uganda—at between US$0.48 and Evaluating the costs and effectiveness of palliative care US$0.98 per day (US$ 2004) (Foley and others 2006). in LMICs presents challenges of the most basic kind. In the United States, the average community No single model of palliative care delivery can be imple- pharmacy’s acquisition cost of immediate-release oral mented across all countries; each country presents with morphine tablets sufficient for three days at 60 mcg a unique constellation of health care resources and per day was US$1.20 in 2014. More sophisticated challenges. dosing forms may cost more to manufacture; new, Studies comparing the costs of organized palliative patented formulations include in their price a pre- care programs for patients with cancer with care deliv- mium for intellectual property. A fentanyl patch, for ered in the absence of such programs are limited to example, compares at US$4.09 for similar pain relief.2 HICs, where conventional cancer care is already well Costs vary widely by country, depending on such organized. Researchers reviewing such studies have con- factors as whether medicines can be produced within cluded that organized palliative care tends to save health the country or must be imported, whether they face Cancer Pain Relief 169 import duties, and the nature of the existing medicine (17th edition, WHO 2011) and additional medicines on distribution system. An unfortunate circumstance is the IAHPC list of essential palliative care medicines: that the factors that make opioids more expensive tend to be prevalent in LMICs. • Fentanyl: transdermal patch In some cases, making more effective pain relief med- • Hydromorphone: injectable, oral liquid, oral solid ications available to patients with cancer by removing immediate release, oral solid sustained release regulatory barriers would involve no costs other than • Methadone: oral liquid, oral solid the cost of the medicine itself. Where oral morphine is • Morphine: injectable, oral liquid, oral solid already available and affordable, for example, the addi- • Oxycodone: oral solid immediate release, oral solid tional cost of providing other medicines deemed essential sustained release for palliative care by IAHPC3 would be limited largely to the drug costs alone, since the care delivery and narcot- The price of opioids, when available at all, was highest, ics control mechanisms would already be in place. The in absolute terms, in the poorest countries, except in training for professional staff in the appropriate use of certain LMICs where they are free to patients; however, palliative medicines would be an ongoing cost, but that these are countries with extremely low consumption, cost would be spread across the population of patients meaning effectively no availability. In the Opioid Price with cancer who are served. Watch sample, these no-cost but no-availability coun- In areas with no effective access to any opioids for tries were Nepal (LIC), Sudan (LMIC), and Romania these patients, getting the most appropriate medicine (upper-middle income). to patients is likely to involve additional clinic or home The least expensive drug globally, according to the visits with trained personnel, as well as new controls to international buyer reference price, was the immediate- secure the medicines from theft or abuse. release morphine oral solid; this was not the least expen- Since 1993, Hospice Africa Uganda’s (HAU) model sive formulation (standardized to a 30-day average dose) program has offered palliative care to patients with in many countries. Fentanyl patches were less expensive cancer, AIDS, and, increasingly, other (mainly non- in China (Chengdu), Germany, Guatemala, India, the communicable) conditions. Most patients are seen at Islamic Republic of Iran, Norway, Poland, Spain, the home, but HAU also provides hospice care at inpa- Philippines, and the United Kingdom, suggesting that tient facilities (HAU 2013). The service is nurse-led, these governments subsidize at least some of these med- with physician backup. It includes basic needs support, icines and do so differentially. including food, clothing, blankets, and transport costs; pain and symptom control, including morphine; and other support services. HAU estimated that the cost of adding a children’s palliative care program at a hospi- CONCLUSIONS tal for children and at HAU’s clinic was approximately There is no palliative care without pain control. US$75 (US$ 2007) per child per year. About one-third of The benefits of opioid medicines for pain relief to that cost was for medicines and pharmacy consumables patients with cancer and their families, regardless of (Amery and others 2009). their whereabouts, are real and universally observable. These benefits have been known for decades, yet the vast majority of those dying in pain cannot get these medi- Patient Perspective cines, even though the basic forms are inexpensive, oral, For patients, the most serious problem is that opioids and relatively easy to administer at home. The move- are not available at any price, even if prescribed. Where ment to declare pain relief a human right reflects the they are available, price is the next consideration, but recognition that its benefits extend to the core of human reliable information on the price of pain medicines to dignity. The question becomes not whether pain relief consumers has been lacking in LMICs. This deficit was with opioids is worth its cost, but what steps will most the impetus for IAHPC to begin the Opioid Price Watch, quickly and efficiently facilitate access to these medica- an availability and price survey in a sample of 26 low-, tions for those who need them, even in the absence of lower-middle-, and upper-middle-income countries and organized cancer control efforts. HICs (De Lima and others 2014). The endpoints were It has taken the inspired efforts of pioneers to build the availability and price of opioids to the consumer on palliative care capacity in LMICs, mainly in the pri- a single day in each site. vate faith and philanthropic sectors, but involving, as The survey included five opioids (13 formulations), it must, government. According to the GOPI report, including those on WHO’s List of Essential Medicines many international and regional organizations are 170 Cancer focusing on palliative care, with efforts to find, train, Consortium, Multinational Association of Supportive and support the leaders in each country or area. The Care in Cancer, Myanmar Oncology Society, Open Society investment in leadership required to reform national Foundations, Worldwide Palliative Care Alliance. and regional policies is a real cost that must be sup- 2. Sources: Prices obtained from the April 9, 2014, Survey of Drug Acquisition Costs Paid by Retail Community ported. The next steps identified by GOPI (Cleary, Pharmacies, http://www.medicaid.gov/Medicaid-CHIP Radbruch, and others 2013), which they refer to as -Program-Information/By-Topics/Benefits/Prescription the “cornerstone trinity,” are medication availability, -Drugs/Survey-of-Retail-Prices.html. Equivalent daily education, and policy reform. The costs of policy mea- doses are based on data provided by the Ventura County sures are front-loaded; once policies supporting the Health Care Agency, http://www.vchca.org/docs/hospitals widespread availability of pain medications to patients /fentanyl-patch-protocol-(1).pdf?sfvrsn=0. with cancer are in place, they will need only routine 3. http://hospicecare.com/resources/palliative-care-essentials monitoring. /iahpc-essential-medicines-for-palliative-care. These conclusions are robust even in the absence of large numbers of studies and supporting data about the extent of palliative care available in LMICs. The evidence supporting the effectiveness of pain REFERENCES control medicines is strong and applicable worldwide, Amery, J., C. J. Rose, J. Holmes, J. Nguyen, and C. Byarugaba. but validation in LMICs would provide support to 2009. “The Beginnings of Children’s Palliative Care in expanded efforts. Studies detailing the lack of ser- Africa: Evaluation of a Children’s Palliative Care Service vices, the personal consequences of dying in pain, in Africa.” Journal of Palliative Medicine 12 (11): 1015–21. and the increasing burden of deaths from cancer and Brennan, F., D. B. Carr, and M. Cousins. 2007. “Pain Management: other noncommunicable diseases will help to per- A Fundamental Human Right.” Anesthesia and Analgesia 105 suade policy makers and funders of the importance (1): 205–21. doi:10.1213/01.ane.0000268145.52345.55. of palliative care. Cherny, N. I., J. Baselga, F. de Conno, and L. Radbruch. 2010. “Formulary Availability and Regulatory Barriers to It is appropriate to note that progress has been made Accessibility of Opioids for Cancer Pain in Europe: A in the past few years, but the main message remains Report from the ESMO/EAPC Opioid Policy Initiative.” that very few of the increasing numbers of people Annals of Oncology 21 (3): 615–26. doi:10.1093/annonc dying from cancer in LMICs can expect to die without /mdp581. debilitating pain. Yet the interventions needed could Cherny, N. I., J. Cleary, W. Scholten, L. Radbruch, and J. Torode. be made available affordably everywhere, in a relatively 2013. “The Global Opioid Policy Initiative (GOPI) Project to short time, if given high enough priority and modest Evaluate the Availability and Accessibility of Opioids for the resources. Management of Cancer Pain in Africa, Asia, Latin America and the Caribbean, and the Middle East: Introduction and Methodology.” Annals of Oncology 24 (Suppl. 11): xi7–13. NOTES doi:10.1093/annonc/mdt498. Clark, D., M. Wright, J. Hunt, and T. Lynch. 2007. “Hospice World Bank income classifications as of July 2014 are as and Palliative Care Development in Africa: A Multi-Method follows, based on estimates of gross national income per capita Review of Services and Experiences.” Journal of Pain and for 2013: Symptom Management 33 (6): 698–710. doi:http://dx.doi .org/10.1016/j.jpainsymman.2006.09.033. • Low-income countries: US$1,045 or less Cleary, J. F., and P. P. Carbone. 1997. “Palliative Medicine in the • Middle-income countries: Elderly.” Cancer 80 (7): 1335–47. http://www.ncbi.nlm.nih • Lower-middle-income: US$1,046–US$4,125 .gov/pubmed/9317188. • Upper-middle-income: US$4,126–US$12,745 Cleary, J., L. De Lima, J. Eisenchlas, L. Radbruch, J. Torode, • High-income countries: US$12,746 or more and others. 2013. “Formulary Availability and Regulatory Barriers to Accessibility of Opioids for Cancer Pain in Latin 1. African Organisation for Research and Training in Cancer, America and the Caribbean: A Report from the Global African Palliative Care Association, Asia Pacific Hospice Opioid Policy Initiative (GOPI).” Annals of Oncology 24 Palliative Care Network, Chinese Society of Clinical (Suppl. 11): xi41–50. doi:10.1093/annonc/mdt502. Oncology, Foundation Akbaraly, Madagascar, Help the Cleary, J., R. A. Powell, G. Munene, F. N. Mwangi-Powell, Hospices, Indian Association of Palliative Care, International E. Luyirika, and others. 2013. “Formulary Availability and Association for Hospice & Palliative Care, Japanese Society of Regulatory Barriers to Accessibility of Opioids for Cancer Medical Oncology, Latin American and Caribbean Society of Pain in Africa: A Report from the Global Opioid Policy Medical Oncology, Latin American Association for Palliative Initiative (GOPI).” Annals of Oncology 24 (Suppl. 11): Care, Malaysian Oncology Society, Middle East Cancer xi14–23. doi:10.1093/annonc/mdt499. Cancer Pain Relief 171 Cleary, J., L. Radbruch, J. Torode, and N. I. Cherny. 2013. Availability for Cancer Pain Relief and Palliative Care: “Next Steps in Access and Availability of Opioids for the More Than a Function of Human Development.” Palliative Treatment of Cancer Pain: Reaching the Tipping Point?” Medicine 27 (2): 105–14. doi:10.1177/0269216312461973. Annals of Oncology 24 (Suppl. 11): xi60–64. doi:10.1093 Gwyther, L., F. Brennan, and R. Harding. 2009. “Advancing /annonc/mdt504. Palliative Care as a Human Right.” Journal of Pain and Cleeland, C. S., J. L. Ladinsky, R. C. Serlin, and N. C. Thuy. Symptom Management 38 (5): 767–74. doi:10.1016/j 1988. “Multidimensional Measurement of Cancer Pain: .jpainsymman.2009.03.003. Comparisons of US and Vietnamese Patients.” Journal of Pain HAU (Hospice Africa Uganda). 2013. Hospice Africa Uganda and Symptom Management 3 (1): 23–27. http://linkinghub Annual Report 2012–2013. Kampala: HAU. .elsevier.com/retrieve/pii/0885392488901340?showall=true. INCB (International Narcotics Control Board). 1961. Single Cleeland, C. S., Y. Nakamura, T. R. Mendoza, K. R. Edwards, Convention on Narcotic Drugs, 1961. http://www.incb.org J. Douglas, and others. 1996. “Dimensions of the Impact of /incb/en/narcotic-drugs/1961_Convention.html. Cancer Pain in a Four Country Sample: New Information International Pain Summit of the International Association from Multidimensional Scaling.” Pain 67 (2–3): 267–73. for the Study of Pain. 2011. “Declaration of Montréal: doi:http://dx.doi.org/10.1016/0304-3959(96)03131-4. Declaration that Access to Pain Management Is a Daut, R. L., and C. S. Cleeland. 1982. “The Prevalence and Fundamental Human Right.” Journal of Pain and Palliative Severity of Pain in Cancer.” Cancer 50 (9): 1913–18. Care Pharmacotherapy 25 (1): 29–31. doi:10.3109/1536028 De Lima, L., and D. Doyle. 2007. “The International 8.2010.547560. Association for Hospice and Palliative Care List of Lohman, D., R. Schleifer, and J. J. Amon. 2010. “Access to Essential Medicines for Palliative Care.” Journal of Pain and Pain Treatment as a Human Right.” BMC Medicine 8: 8. Palliative Care Pharmacotherapy 21 (3): 29–36. doi:10.1080 doi:10.1186/1741-7015-8-8. /J354v21n03_05. Lynch, T., S. Connor, and D. Clark. 2013. “Mapping Levels of De Lima, L., T. Pastrana, L. Radbruch, and R. Wenk. 2014. Palliative Care Development: A Global Update.” Journal of “Cross-Sectional Pilot Study to Monitor the Availability, Pain and Symptom Management 45 (6): 1094–106. http:// Dispensed Prices, and Affordability of Opioids around dx.doi.org/10.1016/j.jpainsymman.2012.05.011. the Globe.” Journal of Pain and Symptom Management. Miaskowski, C., J. Cleary, R. Burney, P. Coyne, R. Finley, and doi:10.1016/j.jpainsymman.2013.12.237. others. 2005. Guidelines for the Management of Cancer Ferlay, J., H.-R. Shin, F. Bray, D. Forman, C. Mathers, and Pain in Adults and Children. Glenview, IL: American Pain others. 2010. “Estimates of Worldwide Burden of Cancer in Society. 2008: GLOBOCAN 2008.” International Journal of Cancer Morrison, R. S., and D. E. Meier. 2011. “The National Palliative 127 (12): 2893–917. doi:10.1002/ijc.25516. Care Research Center and the Center to Advance Palliative Foley, K. 1979. “Pain Syndromes in Patients with Cancer.” In Care: A Partnership to Improve Care for Persons with Advances in Pain Research and Therapy, edited by K. Foley, Serious Illness and Their Families.” Journal of Pediatric J. Bonica, and V. Ventafridda, 59–75. New York: Raven Hematology/Oncology 33 (Suppl. 2): S126–31. doi:10.1097 Press. /MPH.0b013e318230dfa. ———. 1999. “Pain Assessment and Cancer Pain Syndromes.” National Cancer Institute. 2009. SEER Cancer Statistics Review In Oxford Textbook of Palliative Medicine, 2nd ed., edited by 1975–2010. National Cancer Institute, Bethesda, MD. D. Doyle, G. Hank, and N. MacDonald, 310–31. New York: http://seer.cancer.gov/csr/1975_2010/browse_csr.php Oxford University Press. ?sectionSEL=2&pageSEL=sect_02_table.09.html. Foley, K., and H. Gelband. 2001. Improving Palliative Care for Seya, M.-J., S. F. A. M. Gelders, O. U. Achara, B. Milani, and Cancer. Washington, DC: National Academies Press. W. K. Scholten. 2011. “A First Comparison between the Foley, K. M., J. L. Wagner, D. E. Joranson, and H. Gelband. Consumption of and the Need for Opioid Analgesics at 2006. “Pain Control for People with Cancer and AIDS.” Country, Regional, and Global Levels.” Journal of Pain and In Disease Control Priorities in Developing Countries, 2nd Palliative Care Pharmacotherapy 25 (1): 6–18. doi:10.3109 ed., edited by D. Jamison, J. G. Breman, A. R. Measham, /15360288.2010.536307. G. Alleyne, M. Claeson,D. B. Evans, P. Jha, A. Mills, and Simoens, S., B. Kutten, E. Keirse, P. Berghe, P. Vanden, and oth- P. Musgrove, 981–93. Washington, DC: Oxford University ers. 2010. “The Costs of Treating Terminal Patients.” Journal Press and World Bank. of Pain and Symptom Management 40 (3): 436–48. http:// García-Pérez, L., R. Linertová, R. Martín-Olivera, P. Serrano- linkinghub.elsevier.com/retrieve/pii/S0885392410003635 Aguilar, and M. A. Benítez-Rosario. 2009. “A Systematic ?showall=true. Review of Specialised Palliative Care for Terminal Patients: Sloan, F. A., and H. Gelband. 2007. Cancer Control Which Model Is Better?” Palliative Medicine 23 (1): 17–22. Opportunities in Low- and Middle-Income Countries: Cancer doi:10.1177/0269216308099957. Control. Washington, DC: National Academies Press. Gilson, A. M., M. A. Maurer, V. T. LeBaron, K. M. Ryan, and http://books .google.com/books?hl=en&lr=& J. F. Cleary. 2013. “Multivariate Analysis of Countries’ ;id=rJ6LAD7l8nMC&oi=fnd&pg=PA1&dq Government and Health-Care System Influences on Opioid =Cancer+Control+Opportunities+in+Low-+and+Middle 172 Cancer +income+countries&ots=IypmpWQGQf&sig=1 ———. 2000. Achieving Balance in National Opioids Control DGx__xZ_SLYTB5fxNsRS2TI1PM. Policy: Guidelines for Assessment. Geneva: WHO. http:// Stjernsward, J., and D. Clark. 2003. “Palliative Medicine: apps.who.int/iris/bitstream/10665/66496/1/WHO_EDM A Global Perspective.” In Oxford Textbook of Palliative _QSM_2000.4.pdf?ua=1. Medicine, 3rd ed., edited by D. Doyle, G. Hanks, and ———. 2011. “WHO Model List of Essential Medicines.” 17th N. I. Cherny, 1199–222. New York: Oxford University ed., WHO, Geneva. Press. ———. 2013a. Draft Comprehensive Global Monitoring WHO (World Health Organization). 1986. Cancer Pain Relief. Framework and Targets for the Prevention and Control of Geneva: WHO. Noncommunicable Diseases. Geneva: WHO. ———. 1990. Cancer Pain Relief and Palliative Care. Report of ———. 2013b. “WHO Model List of Essential Medicines.” a WHO expert committee. Geneva: WHO. 18th ed., WHO, Geneva. Cancer Pain Relief 173 Global Hazards of Tobacco and the Benefits of Chapter 10 Smoking Cessation and Tobacco Taxes Prabhat Jha, Mary MacLennan, Frank. J. Chaloupka, Ayda Yurekli, Chintanie Ramasundarahettige, Krishna Palipudi, Witold Zaton ´ksi, Samira Asma, and Prakash C. Gupta INTRODUCTION of tobacco control and conclude by reviewing the cur- rent state of global tobacco control implementation. Tobacco use kills approximately five million people annually worldwide, accounting for over 20 percent of all deaths of adult men and 5 percent of deaths of EPIDEMIOLOGY OF SMOKING-ATTRIBUTABLE adult women. As death rates from causes not attrib- DISEASES uted to tobacco are falling, the proportion of all adult deaths due to smoking will rise. In the 20th century, Approximately 1.3 billion people smoke worldwide; over 100 million tobacco deaths occurred; nearly 70 percent 80 percent of smokers reside in LMICs. Smoked tobacco were in high-income countries (HICs) and the for- accounts for about 97 percent of all tobacco sales glob- mer socialist economies of Europe. In contrast, in the ally (Euromonitor International 2013), mostly in the 21st century, tobacco is expected to kill about one billion form of cigarettes, or in the case of South Asia, in the people, mostly in low- and middle-income countries form of bidis, which typically contain about one-fourth (LMICs). as much tobacco as cigarettes. Inhaling tobacco smoke Widespread use of a few powerful interventions causes a greater diversity and incidence of disease than affecting tobacco price, information, and regulations chewing tobacco. Active smoking is also more hazardous could prevent tens of millions of premature deaths over than exposure to secondhand smoke, although second- the next few decades. hand smoke contributes significantly to some diseases This chapter starts with the epidemiology of (IARC 2004). smoking-related diseases, focusing on contemporary estimates of the hazards of smoking and the benefits of cessation, and then describes current and future smok- Substantial Delay from Smoking Uptake to Excess ing patterns, including the rapid emergence of electronic Mortality cigarettes. We next turn to interventions to rapidly After smoking becomes common in a population of raise cessation rates in LMICs, in particular, higher young adults, it may take more than half a century to excise taxes on tobacco products. We discuss the cost- assess reliably the full risks of mortality (Jha and Peto effectiveness, cost-benefit, and poverty considerations 2014). Five recent studies in Japan, the United Kingdom, Corresponding author: Prabhat Jha, Center for Global Health Research, St. Michael’s Hospital and Dalla Lana School of Public Health, University of Toronto, Prabhat.Jha@utoronto.ca. 175 and the United States have examined large populations rose substantially from 12-fold in the 1960s to 24-fold of men and women who began to smoke regularly when in the 1980s; it stabilized at 25-fold in the 2000s. By they were young and never quit. These five studies find contrast, the risks of lung cancer for women who smoked a twofold to threefold increased risk of death among versus never-smokers rose later (Thun and others 2013). smokers, leading to a reduction in lifespan of at least one In the 1960s, it was threefold; in the 1980s, 13-fold; and decade (Doll and others 2004; Jha, Ramasundarahettige, in the 2000s, 26-fold (with similar ratios in women in and others 2013; Pirie and others 2013; Sakata and the United Kingdom in the 2000s). This is because the others 2012; Thun and others 2013). The same studies typical 60-year-old female smoker in the United States included individuals who quit smoking. Those who in the 2000s had smoked since early adult years, whereas stopped before age 40 avoided about 90 percent of the those who were smokers in the 1960s had not. excess risk of death of those who continued. Smokers who do not start in early adult life have much smaller hazards in middle and old age. Table 10.1 summarizes Cancer and Other Diseases Caused by Smoking the main findings for individuals, and figure 10.1 shows More than two-thirds of the 47 million deaths among the risks by gender for Japan, the United Kingdom, adults over age 25 years worldwide in 2012 were caused and the United States, as well as comparable risks among by cancer, vascular and respiratory diseases, and tuber- men who smoke cigarettes in India. culosis (WHO 2013a). Because smoking causes many of Men born in the United Kingdom in the first quarter these diseases, overall mortality from smoking—rather of the 20th century were the first major population to than cause-specific mortality—is increasingly used as smoke regularly from early adult life; by 1970, the nation a measure of total smoking risk (U.S. Department of had the highest tobacco-attributable death rates in the Health and Human Services 2014). About 50 percent of world (Peto and others 1992). Sir Richard Doll’s study the current five million smoking-related deaths world- of doctors born during the first half of the 20th century wide occur in LMICs, and about 80 percent of these and followed for the second half showed a 10-year loss smoking deaths occur in men, but this is chiefly because of life expectancy among those who continued to smoke of the lag in women’s uptake of smoking. (figure 10.1). In 2012, 14 million new cases of cancer were diagnosed For women, smoking became common later in the and about eight million cancer deaths occurred world- United Kingdom and the United States, beginning wide. Over 30 percent of the cancer deaths in middle-aged with women born in the 1940s and 1950s. The full men and about 10 percent of those in women are due to risks in women have been measured only early in the smoking (Ferlay and others 2013). Smoking accounted for 21st century. By the 1980s, the lung cancer risk ratio the vast majority of the 1.6 million deaths from lung can- in men who smoked versus men who never smoked cer in 2012 (1.1 million in men and 0.5 million in women). Table 10.1 Three Main Implications for Individuals Who Become Cigarette Smokers in Adolescence or Early Adult Life 1. The risk is big, if they continue smoking. • Continued smoking eventually kills at least half of men and women who smoke. Among persistent cigarette smokers, whether men or women, the overall relative risk of death throughout middle age and well into old age is at least twofold higher than otherwise similar never-smokers. Among smokers of a given age, more than half of those who die in the near future would not have done so at never-smoker death rates. • On average, smokers lose at least one decade of life. This average combines a zero loss for those not killed by tobacco with the loss of much more than one decade for those who are killed by it. 2. At least half of those killed are middle aged (ages 30–69 years), losing many years of life. • Some of those killed in middle age might have died anyway, but others might have lived on for another 10, 20, 30, or more years. • On average, those killed in middle age lose about 20 years of never-smoker life expectancy. 3. Stopping smoking works to reduce health risks. • Those who stop before age 40 avoid more than 90 percent of the excess risk among those who continue to smoke. Those who stop before age 30 avoid nearly all of the smokers’ excess risk. • Those who have smoked cigarettes since early adult life but stop at 30, 40, 50, or 60 years of age gain, respectively, about 10, 9, 6, and 4 years of life expectancy, compared with those who continue smoking. Sources: Doll and others 2004; Jha, Ramasundarahettige, and others 2013; Pirie and others 2013; Sakata and others 2012; Thun and others 2013. 176 Cancer Figure 10.1 Loss of a Decade of Life among Current Cigarette Smokers versus Never-Smokers Who Are Middle Aged: Men and Women in the United Kingdom and the United States, Men in Japan, and Men in India a. UK Men b. UK Women Percentage of survival from age 35 Percentage of survival from age 35 100 100 Never-smokers Never-smokers 80 80 Current smokers 10 60 60 years Current smokers 10 40 years 40 20 20 0 0 30 40 50 60 70 80 30 40 50 60 70 80 Age (years) Age (years) c. US Men d. US Women Percentage of survival from age 25 Percentage of survival from age 25 100 100 Never-smokers Never-smokers 80 80 Current smokers Current smokers 60 60 11 years 12 years 40 40 20 20 0 0 30 40 50 60 70 80 30 40 50 60 70 80 Age (years) Age (years) e. Japanese Men f. Indian Men Percentage of survival from age 35 Percentage of survival from age 35 100 Never-smokers 100 Never-smokers 80 80 60 Current smokers 9 60 Current smokers 9 years years 40 40 20 20 0 0 30 40 50 60 70 80 30 40 50 60 70 80 Age (years) Age (years) Sources: Data adapted from various studies. Men in the United Kingdom: Doll and others 2004; women in the United Kingdom: Pirie and others 2013; men and women in the United States: Jha Ramasundarahettige, and others 2013; men in Japan: Sakata and others 2012; men in India: Jha and others 2008. The International Agency for Research on Cancer (IARC) periods, and the lung cancer incidence and mortality rates cancer registry data reveal that the age-standardized are accordingly lower. mortality rate from lung cancer is highest in men in However, where reliably measured, the proportion Europe, Northern America, Australia/New Zealand, and of cancer deaths due to smoking is substantial in Japan and lower in LMICs, reflecting the long duration of LMICs. In China, smoking caused about 40 percent of smoking in the former populations and earlier age of ini- cancer deaths in men and 20 percent in women ages tiation. Men in LMICs have not yet smoked for prolonged 35 or older in 2010 (Chen and others, forthcoming). Global Hazards of Tobacco and the Benefits of Smoking Cessation and Tobacco Taxes 177 In India, smoking caused about 32 percent and a substantial reduction. Smoking-attributable deaths 6 percent of cancer deaths at ages 30–69 years in men at ages 35–69 years in men declined by three-quarters and women, respectively, in 2010. Smoking appears to from 1970 to 2010 (from 70,000 to 16,000 deaths). Men have a synergistic effect on the chronic viral infections who died in 1970 had begun smoking between 1920 and that cause liver and cervical cancers. 1940, near the peak of the smoking rate, when cessation Smoking causes about three times as many noncancer was uncommon. The decline was caused by fewer men deaths as it does cancer deaths. Cardiovascular disease is beginning to smoke from 1950 to 1970; a substantial the leading cause of smoking-attributable deaths world- proportion of this cohort ultimately quit smoking. wide (Jha 2009). Smoking is a significant risk factor for The cancer and all-cause death rates from smoking fatal and nonfatal heart attack and stroke. In HICs, about among women in the United Kingdom and the United one-half of the male and one-third of the female deaths States peaked much later, around 1995, and they have at ages 35–69 years from chronic lung disease are due to since reached a plateau. Similar declines have been seen smoking. In China, chronic lung disease accounted for in the proportion of all deaths due to tobacco. Notably, a quarter of all tobacco deaths among men and women cancer death rates due to smoking have fallen substan- in 2010 (Chen and others, forthcoming). Among those tially in men and have not risen in women. Death rates ages 30–69 years in India in 2010, over 30 percent of from cancers not caused by smoking have declined, due deaths among men and 10 percent of deaths among in part to screening and treatment of common cancers women from chronic lung disease were due to smoking. (Gelband and others 2015). Increased risks of tuberculosis death and nonfatal tuberculosis among smokers have been observed in coun- tries where tuberculosis remains common, most notably, Comparably High Mortality Risks from Smoking in India. In India, nearly 40 percent of tuberculosis Emerging in LMICs deaths among middle-aged males—about 120,000 deaths Cigarette smoking rose substantially among men in annually—are attributable to smoking. Tuberculosis is many HICs in the first few decades of the 20th century the leading cause of smoking deaths in rural areas; and subsequently increased in women in most HICs and acute heart attack is the leading cause in urban areas. men in LMICs. Currently, there are about 1.3 billion Subclinical infection with the tubercle bacillus is wide- smokers worldwide (Giovino and others 2012), although spread, and smoking appears to facilitate the progression slightly lower estimates have been proposed (Ng and from silent to active clinical disease. Thus, smoking might others 2014). A simple formula is that every metric contribute to the spread of tuberculosis (Gajalakshmi tonne of tobacco produces about one million cigarettes, and others 2003; Jha and others 2008). which leads to one death (Jha and Peto 2014). The magnitude of the tobacco epidemic in a given country depends on the average daily consumption of Effects of Cessation on Total Mortality in Individuals cigarettes or bidis, which is generally lower in LMICs and Populations than in HICs; in India, about 80 percent of current Large numbers of adult males and fewer adult females smoking is of the lower-risk bidis, but these are being have quit smoking in HICs, providing the opportunity replaced by higher-risk cigarettes (Jha and others 2011). to study the effect of quitting at various ages on subse- Relative to nonsmokers in India, male bidi smokers lose quent mortality. The results of various studies indicate a roughly six years, female bidi smokers lose about eight 90 percent reduction in the excess risk of death among years, and male cigarette smokers lose about 10 years those who quit smoking by age 40; for those quitting by of life. The 10-year loss of life among Indian male cig- age 30, the benefit approaches the rates of never-smokers arette smokers is about as extreme as that in the recent (figure 10.2, panel a) (Jha, Ramasundarahettige, and studies noted, despite the fact that Indian men smoke others 2013). The benefits of cessation are remarkably fewer cigarettes per day and start later in life than do similar in studies in Japan, the United Kingdom, and the men in HICs. The age at starting smoking is generally United States (table 10.1). later in LMICs than HICs. Urban Chinese men, how- The large individual benefits of smoking cessation ever, have begun to start as young as men in the United translate into major reductions in smoking-attributable States (Chen and others, forthcoming). If similar shifts deaths from all causes and from cancer in the overall to smoking at younger ages occur in India and other population. In the United Kingdom, there are now populations, the eventual hazards of smoking might twice as many ex-smokers as current smokers over age well be greater. Epidemiological studies have confirmed 50. Reliable indirect methods (Peto and others 2012) the elevated age-specific, smoking-attributable risks in to calculate tobacco-attributable deaths demonstrate Bangladesh (Alam and others 2013); China, and Hong 178 Cancer Figure 10.2 Effect of Smoking Cessation on Survival at Various Ages, Men and Women, United States, 2006–12 a. Quitting at ages 25–34 b. Quitting at ages 35–44 100 100 Never-smokers Never-smokers Percentage survival from age 30 Percentage survival from age 40 80 80 10 years gained 9 years 60 Current smokers 60 Current smokers gained Quitters 40 40 Quitters 20 20 0 0 30 40 50 60 70 80 90 30 40 50 60 70 80 90 Age (years) Age (years) c. Quitting at ages 45–54 d. Quitting at ages 55–64 100 100 Never-smokers Never-smokers Percentage survival from age 50 Percentage survival from age 60 80 80 6 years 60 60 4 years gained Current smokers gained Current smokers 40 Quitters 40 Quitters 20 20 0 0 30 40 50 60 70 80 90 30 40 50 60 70 80 90 Age (years) Age (years) Source: Adapted from Jha, Ramasundarahettige, and others 2013. Kong SAR, China (Gu and others 2009; Chen and and 2050 (Peto and Lopez 2001). Further estimations others, forthcoming; Lam and others 1997); India (Jha are more uncertain, but based on current initiation and others 2008); and South Africa (Sitas and others and cessation rates and projected population growth, 2013). Accordingly, the future mortality risks from from 2050 to 2100, there would be, conservatively, an 2010 to 2050 in many LMICs might be greater than additional 550 million tobacco deaths, an average of those from 1940 to 1980 in the United Kingdom or the 10 million deaths per year. Of the estimated one billion United States. smoking-attributable deaths in this century, most will Overall global death rates have fallen sharply over be in LMICs. In contrast, there were “only” 100 million the past four decades, particularly from 2000 to 2010 tobacco deaths in the 20th century, mostly among (Norheim and others 2014). Greater declines have been those born before or around World War II in HICs seen in childhood and infectious diseases than for dis- and Eastern Europe. Already, 80–90 million smoking eases made more common by tobacco. Hence, while it deaths will have occurred from 2000 to 2015. Peto’s is certain that tobacco will account for an increasing estimates of 50–60 million smoking deaths from 2000 proportion of a falling overall total of premature deaths to 2010 and about 400 million tobacco-attributable before age 70 years, the future projections of the absolute deaths from 2010 to 2050 remain plausible. Indeed, total from tobacco use are less certain. the chief uncertainty is not if tobacco deaths will reach Sir Richard Peto estimated that global tobacco about 10 million a year, but when, with the most likely deaths will total about 450 million between 2000 scenario around 2030 to 2035 (Peto and Lopez 2001). Global Hazards of Tobacco and the Benefits of Smoking Cessation and Tobacco Taxes 179 SMOKING AND CESSATION PATTERNS high because they include people who quit either WORLDWIDE because they are too ill to continue smoking or because of early symptoms of tobacco-attributable The epidemiological data on the consequences of smok- illness, such as respiratory disease. A good measure ing help predict, with reasonable certainty, that deaths of the success of tobacco control is a rising propor- from smoking will be high in LMICs unless current tion of adults quitting in middle age (ages 45–64) smokers quit and potential new smokers do not start. when they can expect personal gains in health. In the This section reviews the statistics on current smoking European Union and the United States, where ces- and cessation across countries. sation has become common, about as many adults in this age group are former smokers as are current Current Smoking smokers. About 60 percent of all ex-smokers reside in HICs. By contrast, most LMICs, except Brazil, have The overall global volume of legal cigarette sales rose far fewer former than current smokers at these ages. from 5.1 trillion in 1990 to 5.9 trillion in 2012, largely Cessation among women continues to lag men in due to population growth. Although per-person con- nearly every country. sumption has fallen worldwide, it has risen in many LMICs (Euromonitor International 2013). About 80 percent of the 1.3 billion smokers age 15 years or INTERVENTIONS TO RAISE CESSATION older in the world live in LMICs; over half reside in RATES RAPIDLY eight regions or countries—Bangladesh, Brazil, China, the European Union, India, Indonesia, the Russian Cessation by today’s smokers is the only practicable Federation, and the United States (table 10.2). The three- way to avoid a substantial proportion of tobacco deaths fold hazards of lifelong smoking can be applied reliably worldwide before 2050. Halving the worldwide per to the cohort of roughly 620 million younger current capita adult consumption of tobacco by 2025 (akin to and would-be smokers in selected countries. This appli- the declines in adult smoking in the United Kingdom cation suggests that at least half, over 300 million, of this over the past three decades) would prevent approxi- cohort will be killed by smoking unless they quit. mately 160 million to 180 million tobacco deaths over The Global Adult Tobacco Survey, of people ages the next few decades. In contrast, halving the percent- 15 years and older in selected countries, indicates that age of children who become prolonged smokers (from the proportion of males who smoke ranges from over about 30 percent to 15 percent over two decades) would 67 percent in Indonesia to about 7 percent in Nigeria. In prevent 20 million deaths over the next few decades; its most LMICs, male smoking is far more common than main effect would be to lower mortality rates by 2050 female smoking. Among adult women, for example, and beyond (Jha and Chaloupka 1999; Peto and Lopez the proportion of smokers ranges from 24 percent in 2001). Table 10.3 summarizes the effectiveness of the Poland to less than 1 percent in Nigeria (Giovino and major interventions. others 2012). Tobacco Taxation Smoking Cessation Patterns Aggressive taxation is the key strategy for LMICs to To determine smoking prevalence in a population, reduce smoking at a rate faster than that achieved by individuals are divided into three categories: current HICs. Tobacco taxes and consumption are strongly smokers, ex-smokers, and never-smokers. Ex-smoking inversely related worldwide. Well over 100 studies prevalence is a good measure of cessation at a pop- demonstrate a strong negative relationship between ulation level. An increase in cessation, along with an cigarette pricing and consumption (Chaloupka, Yurekli, increasing proportion of never-smokers, reduced adult and Fong 2012; Jha and Chaloupka 1999). We review five smoking prevalence in the United Kingdom between key aspects of smart taxation: 1950 and 2005, from 70 percent to 25 percent in men and from 40 percent to 20 percent in women. In the • Price elasticity of demand for tobacco United States, among men ages 60–64, there are about • Affordability, the relationship of price to income four times as many ex-smokers as current smokers (Jha, growth Ramasundarahettige, and others 2013). Similar rates of • Importance of smart tax structure, including excise cessation have been reported in most HICs. taxes The prevalence of male ex-smoking in LMICs is • Implementation experience of large tax increases much lower. Even reported figures may be misleadingly • Signaling of prices to consumers. 180 Cancer Table 10.2 Prevalence and Number of Current and Future Smokers below age 35 and Expected Deaths, Selected Countries Current smoking prevalence, ages Approximate number 15+ (percent) of deaths in current Current Current and future and future smokers smokers, ages smokers, ages ages <35, unless Country Male Female Total 15+ (millions) 0–34 (millions) smokers quit a China (2010) 53 2 28 317 193 97 a India (2009) 24 3 14 122 95 48 EU-28 (2012)b 32 22 27 115 54 27 Indonesia (2011)a 67 3 35 61 58 29 c United States (2011) 22 17 20 50 26 13 a Russian Federation (2008) 60 22 39 47 32 16 Brazil (2008)a 22 13 17 26 19 10 a Bangladesh (2009) 45 2 23 25 25 13 a Philippines (2008) 48 9 28 18 22 11 Turkey (2008)a 48 15 31 17 18 9 a Vietnam (2010) 47 1 24 17 14 7 a Mexico (2009) 25 8 16 14 14 7 Thailand (2009)a 46 3 24 13 8 4 a Ukraine (2010) 50 11 29 11 8 4 a Egypt, Arab Rep. (2009) 38 1 19 11 12 6 Argentina (2012)a 29 24 22 7 7 4 Canada (2011)d 20 16 17 5 2 1 a Malaysia (2011) 44 15 23 5 5 3 Nigeria (2012)a 7 1 4 4 6 3 Subtotal (HICs) 170 ~80 ~40 Subtotal (LMICs) 715 ~540 ~270 Total 885 ~620 ~310 Note: For future smokers in low- and middle-income countries (LMICs), we apply the smoking prevalence at ages 25–34 from the Global Adult Tobacco Survey to the United Nations 2012 population under age 25 years, plus current smokers at ages 25–34. For future smokers in high-income countries (HICs), we apply the smoking prevalence at ages 18–24 or 20–24 to the under-25 population, with an assumed 1 percent annual decline due to decreased uptake in these countries. a. WHO-GATS, various years, various Country Reports. b. EU-28: Zaton´ski and others 2012, ages 18+. c. United States: Centers for Disease Control and Prevention 2009–11, ages 18+. d. Canada: Health Canada 2012, ages 15+. Price Elasticity of Demand for Tobacco of cigarettes constitutes a larger relative proportion of Raising tobacco taxes to achieve a 50 percent increase income. However, in China and Russia, price elasticity in tobacco prices decreases consumption by about estimates are closer to zero, in part due to a rapid rise in 20 percent in HICs and by at least as much in most affordability. LMICs (that is, price elasticity of −0.4). Price elasticity Half or more of the effect of prices on demand results estimates vary more in LMICs than in HICs, ranging from increased quitting; the remaining effect results from −0.15 to −0.9, but most studies find results to be from reducing the amount of tobacco smoked. Higher concentrated in the −0.20 to −0.60 range. In theory, taxes increase the number of attempts to quit smoking people in LMICs should be more sensitive to price and and the success of those attempts; in the United States, nonprice tobacco control interventions because the cost a 10 percent increase in price results in 11–13 percent Global Hazards of Tobacco and the Benefits of Smoking Cessation and Tobacco Taxes 181 Table 10.3 Interventions and Their Effectiveness in Reducing Tobacco Consumption Intervention Effectiveness Large, regular increases in excise taxes that reduce affordability 50 percent higher prices reduce consumption by approximately 20 percent (10 percent quit, 10 percent reduce the amount smoked) Mass media counter-advertising; warning labels; plain packaging; and Not quantified but does increase cessation rates epidemiological studies, such as deaths from smoking on death records Complete bans on tobacco advertising, promotion, and sponsorship Reduces consumption by approximately 15 percent in LMICs Complete ban on smoking in public places, including workplaces Reduces consumption by 3–14 percent Cessation support for smokers, brief medical advice, pharmacotherapy At six months unaided cessation is about 2–5 percent; brief medical advice doubles the quit rates (to 4–8 percent); medications triple the quit rates (to 8–12 percent) Antismuggling technologies: local language labels, improved tax 10 percent higher spending on antismuggling efforts reduces administration, and increased customs and international efforts to target smuggling by 5 percent and consumption by 2 percent smuggling shorter smoking duration, or a 3 percent higher proba- growth, particularly in fast-growing countries such as bility of cessation (Tauras 1999). Recent analyses of the China and Vietnam. Global Adult Tobacco Survey data and the Global Youth Tobacco Survey find that cessation is generally price Importance of Smart Excise Tax Structure inelastic, but it is still significantly positively related to Most LMICs have low levels of taxation; they also have price, while initiation is generally price elastic (Kostova inefficient tax structures that contribute to increased and others 2011; Shang and others 2014). consumption. For example, China has a relatively low Higher cigarette prices are particularly effective in price elasticity of demand for cigarettes; smokers are reducing consumption among less educated and lower- likely to switch to a wide range of cheaper brands rather income individuals and in preventing young smokers than quit smoking. This low measured price elasticity from moving beyond experimentation into regular, in part reflects the underlying tax structure and income addicted smoking. A comprehensive study using mul- growth. Further, there are several types of taxes: excises, tiple waves of the Global Youth Tobacco Survey in 17 which are based on quantity or weight (for example, the LMICs concluded that a 50 percent increase in cigarette tax per pack of 20 cigarettes); ad valorem taxes, which are prices would result in a reduction of youth prevalence based on the value of tobacco products (for example, a of almost 40 percent. Furthermore, the estimated price specific percentage of manufacturers’ prices for tobacco elasticity of youth smoking was about −2.0 (Kostova and products); and other taxes (for example, import duties). others 2011; Nikaj and Chaloupka 2014). Specific excise taxes are more important insofar as they differentiate tobacco product prices from other Affordability prices more than broader taxes do. A high reliance on Affordability is a concept that captures the interaction ad valorem and similar taxes by most LMICs creates between consumers’ income levels and tobacco prices. large price gaps and increases incentives to switch Typically, affordability is defined as per capita gross to cheaper products. China, like Indonesia, uses a domestic product (GDP) relative to the wholesale complex tiered system with small specific, different price index for bidis or cigarettes. As income rises ad valorem rates that rise with the cost of ciga- relative to price, affordability increases. From 1990 to rettes (Barber and others 2008; Hu and others 2008). 2006, cigarettes became less affordable in 59 percent Indeed, the tobacco industry usually offers such advice of 32 HICs, but only in 38 percent of 45 LMICs. to ministries of finance to promote complex, tiered Particularly since 2003, cigarettes in some LMICs have taxation systems so as to decrease the impact of tax quickly become more affordable. In India, bidis were increases on sales (Jha 2015) nearly three times more affordable in 2011 than in By contrast, empirical research from HICs finds that 1990, while cigarettes were about twice as affordable high uniform, specific excise taxes are more likely to (Blecher and van Walbeek 2009; Jha and others 2011). discourage switching among different types of tobacco Thus, tax policy needs to take into account income products, are easier to administer, and produce a steadier 182 Cancer stream of excise revenue (Chaloupka and others 2010). Figure 10.3 Inverse Relationship of Consumption and Price in the United The exact impact of high excises depends on market States, France, and South Africa, Various Years conditions, industry efforts to counter the tax hike, a. United States and the effectiveness of tax administration. Increases in 30,000 $4.25 the specific excise tax decrease the relative differences Total sales (million packs) Real (inflation-adjusted) $3.75 price, 2007 US dollars 28,000 between higher- and lower-priced cigarettes, effectively $3.25 26,000 increasing the public health impact. The main weakness 24,000 $2.75 is that such excise taxes need to be adjusted periodically for inflation, which is often higher in LMICs than in 22,000 $2.25 HICs. Thus, a complementary strategy is to raise the 20,000 $1.75 excise every year in excess of inflation and income 18,000 $1.25 growth to reduce affordability so that the number of 1970 1973 1976 1979 1982 1985 1988 1991 1994 1997 2000 2003 2006 Year ex-smokers increases every year. Sales Price Implementation of Large Tax Increases Real (inflation-adjusted) price, 1990 = 100 b. France Powerful policy interventions to tax and regulate con- 6 300 Cigarettes per adult per day sumption and to inform consumers have reduced con- 5 sumption in most HICs. The United Kingdom and the 4 200 United States each took about 35 years, and Canada 3 about 25 years, to halve per adult cigarette consump- tion, from about 10 to about five per adult per day 2 100 (Forey and others 2013). However, France took only 1 15 years to halve consumption. France’s uptake of 0 0 smoking was chiefly after World War II, and its prev- 1950 1960 1970 1980 1990 2000 2010 alence rose until the mid-1980s. From 1990 to 2005, Year cigarette consumption fell from about six cigarettes c. South Africa Real (inflation-adjusted) price, 1990 = 100 2.5 700 per adult per day to three (figure 10.3). This sharp Cigarettes per adult per day decline was mostly due to a sharp increase in excise 600 2.0 tobacco taxation starting in 1990. These excise tax 500 increases raised the inflation-adjusted price threefold. 1.5 400 Among men, the corresponding lung cancer rates at 1.0 300 ages 35–44, which are a good measure of recent smok- 200 ing in the population, fell sharply from 1997 onward. 0.5 100 During this period, revenues in real terms rose from 0 0 about 6 billion to 12 billion euros. 1960 1970 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 In HICs, about 64 percent of the retail price of the Year most popular brands consists of the excise tax; in most Consumption Price LMICs, the proportion of the excise tax is well below Sources: Based on data from the Tobacco Institute (United States) and personal communications 50 percent (figure 10.4). from C. Hill (France) and C. van Walbeek (South Africa). Excise taxes tend to be higher and account for a greater share of retail prices in HICs, but LMICs are increasing the proportion of excise tax to retail prices, retail price fell to about 20 percent around 1990, but it as seen in HICs. In Turkey, the excise share as a propor- subsequently rose to nearly 40 percent. Consumption tion of the retail price rose to 67 percent; in Thailand, it fell from about four to two cigarettes per adult per day. rose to 58 percent; the rise contributed to a decline in Poland’s tax increases have doubled the real price of adult smoking prevalence rates in both countries (WHO cigarettes and decreased consumption. Mauritius has 2013b; WHO-GATS, various years). raised excise taxes by about 30 percent, which has The Mexican government raised the tobacco excise reduced consumption. More recently, the doubling of in 2012 (WHO 2013b). This rise contributed to an excise taxes in the Philippines is expected to raise the estimated decrease in cigarette sales of 30 percent, which average price by 70 percent and reduce consumption by may avoid about 100,000 deaths over the next 30 years about 40 percent (Jha and Peto 2014). (Hernandez-Villa and others, personal communication). The World Health Organization (WHO) recommends In South Africa, the excise tax as a percentage of the that excise taxes account for 70 percent of the retail price Global Hazards of Tobacco and the Benefits of Smoking Cessation and Tobacco Taxes 183 Figure 10.4 Retail Price of the Most Popular Brand and the Proportion Signaling the Effects of Price Increases Due to Excise Taxes by Country Income Levels, 2012 Tobacco consumption changes in response to (US$) announced taxation changes (Becker, Grossman, and Murphy 1994; Chaloupka 1991; Gruber and Köszegi 6 2001). For example, even the debate on a tax increase in Switzerland, which was not adopted by popu- lar vote, led to decreased consumption (Kenkel and 5 Chen 2000). There are likely two mechanisms for this decline. The first is rational price expectations. Price and taxation per pack (PPP dollar) Excise tax 64% Consumers respond to future price expectations for 4 cigarettes, despite their addictive properties. Second is the signal that governments and peers give about the undesirability of smoking, leading people to recon- sider cessation. In the United States, the higher price 3 responsiveness by youth has been linked in part to peer effects, in which one quitter who might be more responsive to price increases influences other teenagers 2 Excise tax to do the same (Tauras and Chaloupka 2004). 39% Thus, governments can announce future tobacco tax Excise tax 37% targets to decrease current and future consumption, 1 in much the same way that central banks announce inflation targets. For example, Australia and New Zealand have opted to raise tax rates to ensure that 0 affordability decreases. France started a similar policy High income Middle income Low income in 1991, increasing cigarette prices by 5 percent or more Country income level in excess of inflation every year. As a result, French Excise tax per pack Other taxes Base price ex-smoking prevalence at ages 45–64 is now well above the European Union average (Jha 2013; Zaton ´ ski and Source: Adapted from WHO 2013b. Man ´czuk 2010). Note: Prices are expressed in dollars adjusted for purchasing power parity (PPP). In low- and middle-income countries, tripling specific excise taxes on tobacco would approximately double street prices, because non-excise taxes and retailer mark-up would also rise. In the European Union, more complex variations of specific excise taxes on tobacco are used. Data are for 48 high-income countries, Health Information and Counter-Advertising 95 middle-income countries, and 30 low-income countries. With more than 40,000 studies on smoking and health published over the past five decades, it is easy to assume of cigarettes and that countries raise excise levels more that the health consequences of tobacco are well known than inflation levels and income growth to reach that worldwide (IARC 2004; U.S. Department of Health and point (WHO 2010). Based on WHO data for the most Human Services 2014). Although this is often the case popular brand, achieving this level of excise would result in HICs, even in these countries many smokers mini- in the weighted price per pack rising from US$0.8 per mize the personal relevance of these risks. Awareness of pack to US$1.4 in low-income countries, from U$1.5 per the hazards of smoking and the benefits of cessation is pack to US$2.6 in middle-income countries, and from much lower in most LMICs. In India, few smokers know US$6.3 per pack to US$10.6 in HICs. Such excise tax that 70 percent of smoking deaths occur during pro- increases, while large, have already been achieved in some ductive middle age or that the average number of years countries, including Canada, France, the Philippines, of life lost from cigarette smoking is nearly 10 years. Poland, and South Africa, and in some states in the The Global Adult Tobacco Survey revealed that only United States. 23 percent of Chinese adults know that smoking causes An increase in cigarette taxes of 25 and 50 percent strokes, heart attacks, and lung cancer. The survey also globally would raise cigarette tax revenues by revealed the widespread belief in most countries that 14 percent and 25 percent, respectively, as the fall in smoking does not cause strokes: over 50 percent in India, demand is less than proportional to the price increase 39 percent in Mexico, 33 percent in Russia, 30 percent in in most countries (WHO 2013b). The overall increase Vietnam, and 27 percent in Brazil. in government revenue would be about US$100 billion Smoking patterns in Western countries have changed (Jha and Peto 2014). in response to control policies and increased information. 184 Cancer Data on tobacco hazards help build public support for Bans on Advertising and Promotion control measures, such as higher prices and bans on Cigarettes are among the most heavily advertised and advertising and promotion. A systematic review found promoted products in the world. Cigarette compa- that nine of 11 mass media campaigns evaluated had nies spend over US$9 billion annually on advertising reduced smoking prevalence or increased cessation rates and promotion in the United States alone, spending (Bala and others 2013). Decreases in smoking prevalence has risen in recent years (Federal Trade Commission were largest in HICs, where coverage of issues related to 2015). The spending in the United States is relevant tobacco in the news media is consistent. For example, the globally because it funds research and industry mar- 1962 report by the British Royal College of Physicians, keting strategies on advertisement and promotion of and the 1964 U.S. Surgeon General’s Report, in combi- cigarettes globally. nation with the publicity that followed each publication, In HICs, partial bans on tobacco advertising have reduced consumption by 4–9 percent initially, and by had little effect on consumption, as the industry 15–30 percent in the longer term in both countries, and shifts to other media, price-reducing promotions, indeed in other countries, such as Switzerland (Kenkel or sponsorship of events, such as rock concerts. and Chen 2000). However, comprehensive bans reduce consumption Prominent, rotating pictorial warning labels on by 6–7 percent, taking into account differences in tobacco products are effective at portraying risks to price and nonprice control interventions. For exam- smokers and can reach even illiterate individuals, which ple, complete tobacco advertising bans in Norway and is important in countries such as India, where half the Finland reduced smoking demand by 9–16 percent smoking deaths occur among the uneducated. Thailand and 7 percent, respectively. Such bans may be twice changed the warnings from 30 percent text-only to as effective in LMICs. In a study of 50 LMICs, 50 percent pictorial in 1996; subsequent surveys noted comprehensive bans reduced consumption by 14.4– an increase from 34 percent to 54 percent of people who 15.5 percent, and by about 5.5 percent in the broader claimed that the warnings made them think about health sample of 76 countries. Limited bans still had some risks “a lot,” with 31–44 percent “a lot” more likely to impact, in part because the marginal impact of bans quit (International Tobacco Control Project 2009). The is greater where no or little tobacco control has impact has been similar in other countries. occurred (Blecher 2008). In 2011, Australia became the first country to legis- late plain packaging for tobacco products. Plain pack- aging aims to standardize the look of cigarette packages Restrictions on Smoking in Public Places by mandating the removal of all brand imagery, Restrictions on smoking in public places are primarily including logos and trademarks. Manufacturers are intended to reduce nonsmokers’ exposure to passive required to print the brand name in a required size, tobacco smoke. However, comprehensive restrictions font, and place and to include prominent picto- in HICs also raise attempts to quit, so that overall rial health warnings. This goes beyond the pictorial consumption falls by 3–4 percent. Tobacco consump- warning labels used in Canada and introduced in tion significantly decreased after laws restricting smok- the United States. The result is that smokers perceive ing were implemented in Germany, Ireland, and the their plain-packaged cigarettes to be lower quality and Netherlands, but not in France (Mons 2011). to experience lower satisfaction. Smokers were also According to a systematic review of studies (most more likely to consider quitting (Wakefield and others from HICs), workplace and community smoke-free 2013). Experimental research demonstrates that plain policies reduce tobacco use prevalence by a median packaging enhances the effectiveness of health warn- of 14 percent. These policies are most effective when ings, helps dispel false beliefs, and reduces the appeal strong social norms help make smoking restric- of smoking (Hammond 2010). tions self-enforcing (IARC 2009). Reduced hospital Reliable reporting of smoking deaths is possible with admissions for cardiac events and an improvement the simple addition of a smoking status question to in some health indicators consistently occur after South Africa’s death certificates, for example, “Was the smoking is banned in public places. This appears to dead person a smoker five years ago?,” which is asked be more related to increased cessation—in partic- of living respondents (Sitas and others 2013). A simi- ular, among working-age adults, in whom smoking lar strategy in India obtains the smoking status of the causes a large proportion of the acute heart attacks deceased and the respondent during household surveys (Jha and others 2009)—and less related to reduced of the causes of death (Dikshit and others 2012; Jha and environmental smoke exposure triggering acute others 2008). heart attacks. Global Hazards of Tobacco and the Benefits of Smoking Cessation and Tobacco Taxes 185 Smoking Cessation Treatments mind that most LMICs have far less regulatory capacity Most of the ex-smokers in the world have quit unaided. currently, interim policies should try to encourage ces- However, unaided quit rates are only about 2–3 percent sation and avoid the pathways to use of manufactured at six months. Physician support or telephone- or cigarettes. Key policies include the following: (a) restrict Internet-based counseling and cessation support can advertising and promotion of e-cigarettes comparable increase these low rates. A systematic review noted to comprehensive bans on cigarette advertising, even that Internet and mobile telephone programs roughly though this is a more blunt instrument than ideally doubled short- and long-term self-reported quitting suited to helping adult smokers quit; (b) ensure that (Whittaker and others 2012). Another systematic review bans on smoking in public places also cover e-cigarettes; from pooled data in 17 trials demonstrated a significant and (c) ensure, as much as possible, separation of the increase in the rate of quitting in those who received ownership of these companies by cigarette companies brief physician advice compared with no advice, increas- (Jha and Peto 2014). Taxation strategies for e-cigarettes ing unaided quit rates by another 1–3 percent (Stead and are feasible (Huang, Tauras and Chaloupka 2014), but others 2013). Pharmacological treatments, including would need to raise the price of cigarettes even more, so nicotine replacement therapies, bupropion, and vareni- as to encourage cessation. cline, further improve the likelihood of quitting, with success rates two to three times higher than when phar- maceutical treatments are not employed (Hartmann- Supply-Side Interventions Boyce and others 2014). In addition, over-the-counter In contrast to the effective interventions designed access to such medications increases access and decreases to reduce demand, scant evidence exists to indicate cost. Cytisine, a cessation drug used commonly in the that restricting supply can effectively reduce consump- former socialist economies, was more effective than a tion. Limitation of youth access to tobacco products, placebo for smoking cessation in Poland. As cytosine cross-border trade restrictions, and crop substitution is much less expensive than standard drugs, it might be and diversification are largely ineffective in reducing practicable in LMICs (West and others 2011). consumption, given that supply will always respond to demand (Jacobs and others 2000). Recent debate has included phasing in outright bans on sales to birth Electronic Cigarettes cohorts, such as anyone born after 2000. These strate- In recent years, there has been rapid development of gies have not yet been tested (Reuter 2013). Importantly, e-cigarettes or other noncombustible products. In the most of the deaths from smoking before 2050 will occur United States, e-cigarettes are now more commonly used in current smokers, so to the extent these efforts draw than traditional cigarettes among high-school students political attention away from taxation and regulation (National Institute of Drug Abuse 2015). In most LMICs, aimed at encouraging cessation and reducing initiation, these products are mostly unregulated and not subject to they could be counterproductive. traditional tobacco control policies. It remains unknown However, controlling cigarette smuggling is effec- whether these products result mostly in adult cessation, or tive. An estimated 6–11 percent of the 5.9 trillion whether they also are significant gateway or bridge prod- cigarettes sold globally in 2006—about 600 billion ucts that might increase the uptake of cigarette smoking cigarettes—enter the market without being taxed. by youth or diminish cessation by adults. No studies have This amounts to approximately US$50 billion in lost yet reliably documented if young e-cigarette smokers revenue, excess consumption, and increased deaths actually quit after a few years. Clinical trials confirm that (Joossens and others 2009). The main determinant of e-cigarettes, used as nicotine replacement, can raise adult smuggling is not price differences from different tax cessation rates, comparable to other nicotine replacement regimes, but corruption, organized criminal networks, products (Bullen and others 2013). Multinational tobacco and weak tax administration. companies are expanding into the nontobacco nicotine A common misconception by governments and pol- delivery business. However, access to e-cigarettes remains icy makers is that illicit trade will increase as cigarette more limited in LMICs than HICs. taxes rise. Clearly, higher taxes increase the financial Much more epidemiological and economic research is incentives for smugglers, but these claims ignore factors needed in this rapidly changing field. The United States of equal or greater importance in making smuggling and other high-income countries are discussing attractive, such as the following: (a) weak governance nuanced regulation that would allow some promotion and lack of high-level commitment by governments; of e-cigarettes to adults, but not for children. Bearing in (b) ineffective customs and excise administration; 186 Cancer (c) corruption and complicity of cigarette manufactur- Cardiovascular, Respiratory, Renal, and Endocrine ers; (d) presence of informal sectors and distribution Disorders) on labor and capital, which determine eco- channels; and (e) population perceptions and socioeco- nomic output, from 2011 to 2030 (table 10.4). Using nomic status. conservative estimates attributing about 33 percent of A study by Yurekli and Sayginsoy (2010) showed no cardiovascular disease, about 50 percent of cancers, and clear correlation between the illicit trade market and the 60 percent of chronic respiratory diseases to tobacco tax rate for the most popular cigarettes from a sample use, the total economic loss from tobacco is expected of 76 countries. By contrast, a stronger correlation was to be about US$12.7 trillion dollars over the next few seen between illicit trade markets and weak governance; decades. This loss translates to about 1.3 percent of countries with strong governance experience lower illicit GDP spent on tobacco-associated diseases every year, trade levels than those with weak governance. or approximately US$0.9 trillion in 2010 terms. In HICs, companies have been convicted in criminal courts for encouraging smuggling. In addition to har- Cost-Effectiveness Analysis monizing prices among countries, effective measures to counter smuggling include prominent tax stamps Tobacco control is highly cost-effective. Significant price and warning labels in local languages, better meth- increases and comprehensive tobacco control measures ods for tracking cigarettes through the distribution are cost-effective in all WHO regions, according to 2002 chain, aggressive enforcement of antismuggling laws, analysis of data from 2000 (table 10.5). and stronger penalties (IARC 2011). For 23 LMICs, an increased real price of cigarettes Spain provides a good example of effective measures to reduce smoking prevalence by 10 percent, in com- to reduce smuggling. Spain raised its investment in intel- bination with mid-range estimates of nonprice inter- ligence tenfold over five years, increased customs activity ventions, would reduce the smoking prevalence rate in border areas, and developed international collabo- by 20 percent. Over three decades, about three million rations to target smuggling (Joossens and Raw 2008). deaths from cardiovascular disease would be averted: As a result, the market share of smuggled cigarettes fell two million from respiratory disease and one million from 16 percent to 2 percent, and tax revenues more from cancer (assuming price elasticity ranges from than doubled, netting US$68 for every dollar spent on −0.40 to −1.20). The cost of implementation would be smuggling control. A 10 percent increase in price, paired US$0.04 to US$0.32 per person, which would be largely, with a 10 percent increase in spending on smuggling if not completely, offset by the increased revenue from controls, would decrease smuggling by 5 percent, reduce the tax (Asaria and others 2007). consumption by 2 percent, and increase tax revenues A US$0.50 increase in the tax on cigarettes and small by nearly 8 percent (Yurekli and Sayginsoy 2010). The cigars, keeping pace with inflation and the growth of Canadian government raised cigarette taxes in 2014, and people’s income, would reduce the federal budget deficit also funded US$100 million for better police enforce- in the United States by about US$42 billion through 2021 ment against smuggling (Jha 2014). Even in the presence (Baumgardner and others 2012). Tax revenues would be of smuggling, tax increases will reduce consumption and higher and spending on Medicare slightly lower, although increase revenue. For example, South Africa saw a rise in spending on Social Security would rise slightly as more reported smuggling from 0 percent to 6 percent in the people would live longer. Even in the long run, there years when it raised excise taxes, but revenues continued would be a net positive budgetary impact, given that the to rise (Van Walbeek 2006). higher revenues from the tax would exceed any increase COST-EFFECTIVENESS OF TOBACCO Table 10.4 Economic Costs of Tobacco, 2010–30 CONTROL US$ trillions Costs Chronic Total due Region/disease Vascular Cancer lung Total to tobacco Tobacco is a major contributor to the large and Low and middle- 9 5 2 16 7 increasing global burden of noncommunicable dis- income countries eases (NCDs). A recent paper uses the value of lost output approach to estimate the impact of NCDs High-income 7 3 3 13 6 countries on worldwide GDP (Jha, Nugent, and others 2013). It employs the WHO’s EPIC model to simulate the World 16 8 5 29 13 macroeconomic effects of NCDs (see DCP3 volume 5, Source: Adapted from Jha, Nugent, and others 2013. Global Hazards of Tobacco and the Benefits of Smoking Cessation and Tobacco Taxes 187 Table 10.5 Range of Cost-Effectiveness Values for Price Increases, Nicotine Replacement Therapies, and Nonprice Interventions, by World Bank Regions, 2000 (2002 US$/DALY saved) NRTs with effectiveness Nonprice interventions with 33% price increase of 1–5% effectiveness of 2–10% World Bank region Low and high estimate Low and high estimate Low and high estimate East Asia and Pacific 2–30 65–864 40–498 Europe and Central Asia 3–42 45–633 55–685 Latin America and the Caribbean 6–85 53–812 109–1,361 Middle East and North Africa 6–89 47–750 115–1,432 South Asia 2–27 54–716 34–431 Sub-Saharan Africa 2–26 42–570 33–417 World 13–195 75–1,250 233–2,916 Source: Adapted from Jha, Chaloupka, and others 2006. Note: Country economies are categorized according to 2002 World Bank regions. DALY = disability-adjusted life year; NRT = nicotine replacement therapy. in longevity-related spending. Another study demon- on tobacco than do richer groups, but the higher price strated that tobacco tax increases are cost-effective from responsiveness by the poor in China showed that after a the health care perspective, even factoring in the medical 50 percent price increase, those in the lowest two quin- costs from years of life gained (van Baal and others 2007). tiles of income would gain 5 and 1 percent in net income, whereas only the higher income quintiles would lose income after the tax increase on tobacco (Verguet and Poverty Considerations others 2015). Main and others (2008) conclude that tax and price policies reduce inequalities, but they find Smoking in most countries is more common in lower- that cessation policies might increase inequities, given income and lower-education groups (Palipudi and others the greater likelihood of quitting among higher social 2012), and smoking causes greater disease burdens in the groups. In LMICs with low levels of awareness of smok- poor than in the rich. In several HICs and Poland, smok- ing risks and higher illiteracy levels, pictorial warning ing deaths account for at least one-half of the differences labels might help to reduce inequalities. in the middle-age risk of death between men who are richer and more educated versus men who are poorer and less educated (Jha, Peto, and others 2006). IMPLEMENTATION OF TOBACCO CONTROL A recent report from the Asian Development Bank INTERVENTIONS highlights the equity implications of tobacco taxation for five high-burden countries: China, India, the Philippines, Case Study: Tobacco Control in Uruguay Thailand, and Vietnam (ADB 2012). This study found In Uruguay, smoking is a major cause of avoidable mor- that a 50 percent increase in price (resulting from excise tality; in 2004, smoking contributed to 14 percent of the tax increases of 75 percent to 122 percent) would decrease country’s total deaths (Sandoya and Bianco 2011). the number of current and future smokers by almost Uruguay ratified the Framework Convention 67 million, save 27 million people from tobacco-related on Tobacco Control (FCTC) in 2004; by 2005, the deaths, and generate in excess of US$24 billion in addi- country began to implement increasingly comprehen- tional revenue annually (an increase of 143–178 percent sive tobacco control measures. Starting with banning above existing cigarette tax revenue). Importantly, each tobacco advertising except at the point of sale and country’s poorest socioeconomic groups would under- tobacco sponsorships, Uruguay outlawed smoking in take a relatively small proportion of the extra tax burden enclosed public spaces and workplaces. It also required but would gain a substantial proportion of the health primary health care providers to give free diagnosis and benefits from smoking reduction. treatment of tobacco dependence, and stipulated that The ratio of health benefits obtained by the poor to pictograms with health warnings must cover 80 percent the additional taxes paid by the poor ranges from 1.4 to of the front and back of cigarette packages. Terms like 9.5. Poorer income groups spend more of their income light, mild, and low in tar were banned. Following a 188 Cancer sequence of tax increases, the real consumer price of a technical support. The highest priority is countering pack of cigarettes rose by 88 percent from January 2003 the active influence of the tobacco industry, which to December 2010. seeks to secure complex tax regimes favoring certain Abascal and others (2012) evaluated the impact of segments of the tobacco market, as well as to lobby for these tobacco control measures by comparing Uruguay initiatives to confuse governments on tobacco taxes with neighboring Argentina. Per-person consumption (Jha and Alleyne 2015). of cigarettes fell by 4.3 percent annually in Uruguay but The 2013 World Health Assembly called on increased by 0.6 percent in Argentina; the prevalence of governments to decrease the prevalence of smoking tobacco use in adolescents decreased by 8 percent annu- by one-third by 2025 (WHO 2011); doing so would ally in Uruguay and decreased 2.5 percent in Argentina; avoid more than 200 million deaths from tobacco and the prevalence of tobacco use by adults decreased over the remainder of the century (Jha 2009; Jha and by 3.3 percent annually in Uruguay and decreased by Peto 2014). However, few governments are investing 1.7 percent in Argentina. resources in tobacco control measures. HICs spent The impact of specific interventions is difficult to the largest amount (US$1.4 per capita in 2010), which estimate, but studies suggest that comprehensive tobacco was less than 1 percent of the revenues from tobacco control policies should be adopted. Reductions in taxes. Middle-income countries spent a great deal less tobacco use of the size seen in Uruguay (approximately (a little more than US$0.1 per capita in 2010); low- 23 percent over six years) would have a significant income countries spent about US$0.1 per capita in 2010 impact on the future global burden of tobacco-associated (WHO 2013b). diseases. Bloomberg Philanthropies and the Bill & Melinda Gates Foundation have pledged, collectively, nearly US$700 million to fund global tobacco control pro- Globalization and Tobacco Control grams. Effective use of these funds could avoid a sub- A new major challenge to tobacco control is the global- stantial number of deaths in the coming decades as a ization of the tobacco industry. Globalization increas- result of increased adult cessation, and even more deaths ingly challenges strong domestic tobacco control policies could be avoided in the second half of the 21st century under various trade and investment agreements (in from lower increases in youth smoking rates. addition to challenges in national courts). For example, the major multinational tobacco companies have sued CONCLUSIONS: AVOIDABLE TOBACCO the government of Uruguay for its aggressive tobacco control policies. Australia’s plain packaging legislation DEATHS BEFORE 2050 is being challenged by Philip Morris Asia under the Earlier estimates (Jha 2009; Jha, Chaloupka, and oth- bilateral investment treaty between Hong Kong SAR, ers, 2006) have examined the potential impact of a China, where the corporation is based, and Australia, 70 percent price increase and a 10 percent reduction as well as by the Dominican Republic, Honduras, and in tobacco consumption from nonprice interventions, Ukraine through the World Trade Organization (WTO) such as bans on public smoking and information mea- (Oliver 2015). Similarly, Philip Morris International is sures, among the global cohort of 1.1 billion smokers in challenging Uruguay’s graphic warning labels and limit 2000. Price increases have the greatest impact on future on brand variations under a bilateral investment treaty tobacco mortality; a 70 percent higher price would between Switzerland and Uruguay, while Indonesia won prevent more than 110 million deaths, or one-fourth of its WTO case against the United States’ ban on clove- all expected premature deaths from tobacco worldwide. flavored cigarettes. Of the avoided deaths, about 25 million would be from cancer and 50 million would be from vascular disease. Nonprice interventions would prevent 35 million deaths. International Initiatives Worldwide, a one-third reduction in smoking could The main vehicle to accelerate tobacco control is be achieved by doubling the inflation-adjusted price of WHO’s FCTC, the first global treaty on public health, cigarettes; in many LMICs, this price increase could be which has been signed by 180 countries. The FCTC has realized by tripling the real excise tax on tobacco. Other specific provisions for the introduction of the strategies nonprice interventions could help to reduce consump- with proven effectiveness discussed in this chapter. The tion and make the substantial increases in real excise main limitation of the FCTC is that it is largely a state- taxes politically acceptable. The main challenge remains ment of intent; the specific actions needed to imple- to try to bring forward the time when large numbers of ment the provisions in each country require ongoing current smokers quit. Global Hazards of Tobacco and the Benefits of Smoking Cessation and Tobacco Taxes 189 NOTES Blecher, E. 2008. “The Impact of Tobacco Advertising Bans on Consumption in Developing Countries.” Journal of Health World Bank Income Classifications as of July 2014 are as fol- Economics 27 (4): 930–42. lows, based on estimates of gross national income (GNI) per Blecher, E. H., and C. P. van Walbeek. 2009. “Cigarette capita for 2013: Affordability Trends: An Update and Some Methodological Comments.” Tobacco Control 18 (3): 167–75. doi:10.1136 • Low-income countries (LICs) = US$1,045 or less /tc.2008.026682. • Middle-income countries (MICs) are subdivided: Bullen, C., C. Howe, M. Laugesen, H. McRobbie, V. Parag, • lower-middle-income = US$1,046 to US$4,125 and others. 2013. “Electronic Cigarettes for Smoking • upper-middle-income (UMICs) = US$4,126 to Cessation: A Randomised Controlled Trial.” The Lancet 382 US$12,745 (9905): 1629–37. doi:10.1016/s0140-6736(13)61842-5. • High-income countries (HICs) = US$12,746 or more. Centers for Disease Control and Prevention. 2009–11. National Health Interview Survey (NHIS) 2009–11. U.S. Centers for Acknowledgments: This chapter is dedicated to the late Sir Disease Control and Prevention. http://www.cdc.gov/nchs Richard Doll, who would have turned 100 years of age on /nhis.htm. October 28, 2012. We thank Yul Dorothea, Cindy Gauvreau, Chaloupka, F. J. 1991. “Rational Addictive Behaviour and Hellen Gelband, Emmanuel Guindon, Paul Isenman, Tricia Cigarette Smoking.” Journal of Political Economy 99 (4): Moser, and Ken Warner for comments; Joy Pader and Leslie 722–42. Newcombe for their editorial assistance; and Catherine Hill Chaloupka, F. J., R. Peck, J. A. Tauras, X. Xu, and A. Yurekli. for the French data and Corne van Walbeek for the South 2010. Cigarette Excise Taxation: The Impact of Tax Structure African data. Sources of support: Disease Control Priorities 3, on Prices, Revenues and Cigarette Smoking. NBER Working the U.S. National Institutes of Health, the Canadian Institutes Paper 16287, National Bureau of Economic Research, of Health Research, and the Canada Research Chairs Program. Cambridge, MA. http://www.nber.org/papers/w16287. Chaloupka, F. J., A. Yurekli, and G. T. Fong. 2012. “Tobacco Taxes as a Tobacco Control Strategy.” Tobacco Control 21 (2): 172–80. doi:10.1136/tobaccocontrol-2011-050417. REFERENCES Chen, Z., R. Peto, M. Smith, and others, for the China Kadoorie Abascal, W., E. Esteves, B. Goja, F. G. Mora, A. Lorenzo, and Biobank Collaborative Group. Forthcoming. “Chinese others. 2012. “Tobacco Control Campaign in Uruguay: Mortality from Tobacco in the 21st century: Nationwide A Population-Based Trend Analysis.” The Lancet 380: Prospective Study of 0.5 Million Adults.” The Lancet. 1575–82. Dikshit, R., P. C. Gupta, C. Ramasundarahettige, V. Gajalakshmi, ADB (Asian Development Bank). 2012. Tobacco Taxes: A Win– L. Aleksandrowicz, and others. 2012. “Cancer Mortality in Win Measure for Fiscal Space and Health. Manila: ADB. India: A Nationally Representative Survey.” The Lancet 379 Alam, D. S., P. Jha, C. Ramasundarahettige, P. K. Streatfield, (9828): 1807–16. doi:10.1016/s0140-6736(12)60358-4. L. W. Niessen, and others. 2013. “Smoking-Attributable Doll, R., R. Peto, J. Boreham, and I. Sutherland. 2004. “Mortality Mortality in Bangladesh: Proportional Mortality Study.” in Relation to Smoking: 50 Years’ Observations on Male Bulletin of the World Health Organization 91 (10): 757–64. British Doctors.” British Medical Journal 328 (7455): 1519. Asaria, P., D. Chishom, C. Mathers, M. Ezzati, and R. Beaglehole. doi:10.1136/bmj.38142.554479.AE. 2007. “Chronic Disease Prevention: Health Effects and Euromonitor International. 2013. Cigarettes: Global . Financial Costs of Strategies to Reduce Salt Intake and Euromonitor International 2012. http://www.euromonitor Control Tobacco Use.” The Lancet 370: 2044–53. .com/tobacco. Australian Bureau of Statistics. 2013. Australian Health Survey Federal Trade Commission. 2015. Federal Trade Commission (AHS) 2011–13: Tobacco Smoking. Australian Bureau of Cigarette Report for 2012. https://www.ftc.gov/system/files Statistics. http://www.abs.gov.au/australianhealthsurvey. /documents/reports/federal-trade-commission-cigarette Bala, M. M., L. Strzeszynski, R. Topor-Madry, and K. Cahill. -report-2012/150327-2012cigaretterpt.pdf. 2013. “Mass Media Interventions for Smoking Cessation Ferlay, J., I. Soerjomataram, M. Ervik, R. Dikshit, S. Eser, in Adults.” Cochrane Database of Systematic Reviews 6: and others. 2013. “GLOBOCAN 2012 v1.0.” In Cancer Cd004704. doi:10.1002/14651858.CD004704.pub3. Incidence and Mortality Worldwide: IARC CancerBase Barber, S., S. M. Adioetomo, A. Ahsan, and D. Setyonaluri. No. 11 [Internet]. Lyon, France: International Agency for 2008. Tobacco Economics in Indonesia. Paris: International Research on Cancer. Union Against Tuberculosis and Lung Disease. Forey, B. A., J. Hamling, J. Hamling, A. Thorton, and P. N. Lee. Baumgardner, J., L. T. Bilheimer, M. B. Booth, W. J. Carrington, 2013. International Smoking Statistics (ISS): Web Edition. N. J. Duchovny, and others. 2012. “Cigarette Taxes and the http://www.pnlee.co.uk/ISS.htm. Federal Budget—Report from the CBO.” New England Gajalakshmi, V., R. Peto, T. S. Kanaka, and P. Jha. 2003. “Smoking Journal of Medicine 367: 2068–70. and Mortality from Tuberculosis and Other Diseases in Becker, G. S., M. Grossman, and K. M. Murphy. 1994. “An India: Retrospective Study of 43,000 Adult Male Deaths and Empirical Analysis of Cigarette Addiction.” American 35,000 Controls.” The Lancet 362 (9383): 507–15. doi:S0140 Economic Review 84 (3): 396–418. -6736(03)14109-8 [pii] 10.1016/S0140-6736(03)14109-8. 190 Cancer Gelband, H., P. Jha, R. Sankaranarayanan, C. L. Gauvreau, and Jha, P. 2009. “Avoidable Global Cancer Deaths and Total Deaths S. Horton. 2015. “The Costs, Affordability and Feasibility from Smoking.” Nature Reviews Cancer 9: 655–64. of an Essential Package of Cancer Control Interventions in ———. 2013. “The 21st Century Benefits of Smoking Low- and Middle-Income Countries.” In Disease Control Cessation in Europe.” European Journal of Epidemiology Priorities (third edition): Volume 3, Cancer, edited by 28 (8): 617–19. doi:10.1007/s10654-013-9835-6. H. Gelband, P. Jha, R. Sankaranarayanan, and S. Horton. Jha, P., and G. Alleyne. 2015. “Effective Global Tobacco Control Washington, DC: World Bank. in the Next Decade.” Canadian Medical Association Journal. Giovino, G. A, S. A. Mirza, J. M. Samet, P. C. Gupta, M. J. Jarvis, doi:10.1503/cmaj.150261. and others. 2012. “Tobacco Use in 3 Billion Individuals from Jha, P., and F. J. Chaloupka. 1999. Curbing the Epidemic: 16 Countries: An Analysis of Nationally Representative Governments and the Economics of Tobacco Control. Cross-Sectional Household Surveys.” The Lancet 380: Washington, DC: World Bank. 668–79. Jha, P., F. J. Chaloupka, J. Moore, V. Gajalakshmi, P. C. Gupta, Gruber, J., and B. Köszegi. 2001. “Is Addiction ‘Rational’? and others. 2006. “Tobacco Addiction.” In Disease Control Theory and Evidence.” The Quarterly Journal of Economics Priorities in Developing Countries, 2nd ed., edited by 116 (4): 1261–303. doi:10.1162/003355301753265570. D. T. Jamison, J. G. Breman, A. R. Measham, G. Alleyne, Gu, D., T. N. Kelly, X. Wu, J. Chen, J. M. Samet, and others. M. Claeson, D. Evans, P. Jha, A. Mills, and P. Musgrove, 2009. “Mortality Attributable to Smoking in China.” New 869–86. Washington, DC: World Bank and Oxford England Journal of Medicine 360 (2): 150–59. doi:360/2/150 University Press. [pii] 10.1056/NEJMsa0802902. Jha, P., E. Guindon, R. A. Joseph, A. Nandi, R. M. John, and Hammond, D. 2010. “‘Plain Packaging’ Regulations for Tobacco others. 2011. “A Rational Taxation System of Bidis and Products: The Impact, the Standardizing, the Color, and Cigarettes to Reduce Smoking Deaths in India.” Economic Design of Cigarette Packs.” Salud Publica de Mexico 52 and Political Weekly 46 (4.2): 44–51. (Suppl.): 226–32. Jha, P., B. Jacob, V. Gajalakshmi, P. C. Gupta, N. Dhingra, and Hartmann-Boyce, J., L. F. Stead, K. Cahill, and T. Lancaster. others. 2008. “A Nationally Representative Case-Control 2014. “Efficacy of Interventions to Combat Tobacco Study of Smoking and Death in India.” New England Addiction: Cochrane Update of 2013 Reviews.” Addiction Journal of Medicine 358 (11): 1137–47. doi:NEJMsa0707719 109 (9): 1414–25. doi:10.1111/add.12633. [pii] 10.1056/NEJMsa0707719. Health Canada. 2012. Canadian Tobacco Use Monitoring Survey Jha, P., P. Mony, J. A. Moore, and W. Zaton ´ski. 2009. “Avoidance (CTUMS). http://www.hc-sc.gc.ca/hc-ps/tobac-tabac of World-Wide Vascular Deaths and Total Deaths from /research-recherche/stat/index-eng.php. Smoking.” In Evidence-Based Cardiology, 3rd ed., edited by Hu, T-W., Z. Mao, J. Shi, and W. Chen. 2008. Tobacco Taxation S. Yusuf, J. Cairns, A. J. Camm, E. L. Fallen, and B. J. Gersh. and Its Potential Impact in China. Paris: International Union Oxford, UK: Wiley-Blackwell. Against Tuberculosis and Lung Disease. Jha, P., R. Nugent, S. Verguet, D. E. Bloom, and R. J. Hum. Huang J., J. Tauras, and F. J. Chaloupka. 2014. “The Impact 2013. “Chronic Disease Control and Prevention.” In of Price and Tobacco Control Policies on the Demand Global Problems, Smart Solutions—Costs and Benefits, for Electronic Nicotine Delivery Systems.” 2014. edited by B. Lomborg. Cambridge, UK: Cambridge Tobacco Control 23 (suppl 3): iii41–47. doi:10.1136/ University Press. tobaccocontrol-2013-051515. Jha, P., and R. Peto. 2014. “Global Effects of Smoking, of IARC (International Agency for Research on Cancer). 2004. Quitting, and of Taxing Tobacco.” The New England Journal IARC Monographs on the Evaluation of the Carcinogenic of Medicine 370 (1): 60–68. doi:10.1056/NEJMra1308383. Risks of Chemicals to Humans. Volume 83, Tobacco Smoke ———. 2014. “Tax Hike on Tobacco Could Save 25,000 Lives.” and Involuntary Smoking. Lyon, France: IARC, WHO Press. The Star [Toronto], February 25. http://www.thestar.com ———. 2009. Tobacco Control: Evaluating the Effectiveness /opinion/commentary/2014/02/25/tax_hike_on_tobacco of Smoke-Free Policies. IARC Handbooks of Cancer _could_save_25000_lives.html. Prevention. Lyon, France: IARC. ———. 2015. “Deaths and Taxes: Stronger Global Tobacco ———. 2011. Effectiveness of Tax and Price Policies for Tobacco Control by 2025.” The Lancet 385 (9972): 18–20. Control. Edited by International Agency for Research Jha, P., R. Peto, W. Zaton ´ski, J. Boreham, M. J. Jarvis, and on Cancer. Volume 14, IARC Handbooks of Cancer A. D. Lopez. 2006. “Social Inequalities in Male Mortality Prevention. Lyon, France: IARC, WHO Press. and in Male Mortality from Smoking: Indirect Estimation International Tobacco Control Policy Evaluation Project. 2009. from National Death Rates in England and Wales, Poland, FCTC Article 11 Tobacco Warning Labels: Evidence and and North America.” The Lancet 368 (9533): 367–70. Recommendations from the ITC Project. Waterloo, ON: Jha, P., C. Ramasundarahettige, V. Landsman, B. Rostron, International Tobacco Control Policy Evaluation Project. M. Thun, and others. 2013. “21st-Century Hazards of Jacobs, R., H. F. Gale, T. C. Capehart, P. Zhang, and P. Jha. 2000. Smoking and Benefits of Cessation in the United States.” “The Supply-Side Effects of Tobacco Control Policies.” The New England Journal of Medicine 368 (4): 341–50. In Tobacco Control in Developing Countries, edited by doi:10.1056/NEJMsa1211128. P. Jha and F. J. Chaloupka, 311–341. Oxford, UK: Oxford Joossens, L., D. Merriman, H. Ross, and M. Raw. 2009. How University Press. Eliminating the Global Illicit Cigarette Trade Would Increase Global Hazards of Tobacco and the Benefits of Smoking Cessation and Tobacco Taxes 191 Tax Revenue and Save Lives. Paris: International Union Peto, R., A. D. Lopez, J. Boreham, and M. Thun. 2012. Against Tuberculosis and Lung Disease. “Mortality from Smoking in Developed Countries, 1950– Joossens, L., and M. Raw. 2008. “Progress in Combating 2005 (or later): United Kingdom.” http://www.ctsu.ox.ac Cigarette Smuggling: Controlling the Supply Chain.” .uk/~tobacco/C4308.pdf. Tobacco Control 17 (6): 399–404. doi:tc.2008.026567 [pii] Peto, R., A. D. Lopez, J. Boreham, M. Thun, and C. Heath, Jr. 10.1136/tc.2008.026567. 1992. “Mortality from Tobacco in Developed Countries: Kenkel, D., and L. Chen. 2000. “Consumer Information and Indirect Estimation from National Vital Statistics.” The Lancet Tobacco Use.” In Tobacco Control in Developing Countries, 339 (8804): 1268–78. doi:0140-6736(92)91600-D [pii]. edited by P. Jha and F. J. Chaloupka, 177–214. Oxford, UK: Pirie, K., R. Peto, G. K. Reeves, J. Green, and V. Beral. 2013. Oxford University Press. “The 21st Century Hazards of Smoking and Benefits of Kostova, D., H. Ross, E. Blecher, and S. Markowitz. 2011. Stopping: A Prospective Study of One Million Women “Is Youth Smoking Responsive to Cigarette Prices? Evidence in the UK.” The Lancet 381 (9861): 133–41. doi:10.1016 from Low- and Middle-Income Countries.” Tobacco Control /s0140-6736(12)61720-6. 20: 419–24. Reuter, P. 2013. “Can Tobacco Control Endgame Analysis Lam, T. H., Y. He, L. S. Li, S. F. He, and B. Q. Liang. 1997. Learn Anything from the US Experience with Illegal “Mortality Attributable to Cigarette Smoking in China.” Drugs?” Tobacco Control 22 (Suppl. 1): i49–51. doi:10.1136 Journal of the American Medical Association 278: 1505–08. /tobaccocontrol-2012-050809. Main, C., S. Thomas, D. Ogilvie, L. Stirk, M. Petticrew, and Sakata, R., P. McGale, E. J. Grant, K. Ozasa, R. Peto, and others. 2008. “Population Tobacco Control Interventions others. 2012. “Impact of Smoking on Mortality and Life and Their Effects on Social Inequalities in Smoking: Placing Expectancy in Japanese Smokers: A Prospective Cohort an Equity Lens on Existing Systematic Reviews.” BMC Study.” British Medical Journal 345: e7093. doi:10.1136 Public Health 8: 178. doi:10.1186/1471-2458-8-178. /bmj.e7093. Mons, U., G. E. Nagelhout, S. Allwright, R. Guignard, B. van Sandoya, E., and E. Bianco. 2011. “Mortality from Smoking den Putte, M. Willemsen, and others. 2011. “Adoption and Second-Hand Smoke in Uruguay.” Revista Uruguaya de of Home Smoking Bans after the Implementation of Cardiologia 26: 201–06. National Smoke-Free Laws: Results from the ITC Europe Shang, C., F. J. Chaloupka, N. Zahra, and G. T. Fong. 2014. Surveys.” Paper presented at the meeting of the European “The Distribution of Cigarette Prices Under Different Conference on Tobacco or Health, Amsterdam, the Tax Structures: Findings from the International Tobacco Netherlands, March. Control Policy Evaluation (ITC) Project.” Tobacco Control 23 National Institute of Drug Abuse. 2015. Monitoring the (Suppl. 1): i23–29. doi:10.1136/tobaccocontrol-2013-050966. Future 2014 Survey Results. http://www.drugabuse.gov Sitas, F., S. Egger, D. Bradshaw, P. Groenewald, R. Laubscher, /related-topics/trends-statistics/infographics/monitoring and others. 2013. “Differences among the Coloured, White, -future-2014-survey-results. Black, and Other South African Populations in Smoking- Ng, M., M. K. Freeman, T. D. Fleming, M. Robinson, Attributed Mortality at Ages 35–74 Years: A Case-Control L. Dwyer-Lindgren, and others. 2014. “Smoking Prevalence Study of 481,640 Deaths.” The Lancet 382 (9893): 685–93. and Cigarette Consumption in 187 Countries, 1980–2012.” doi:10.1016/s0140-6736(13)61610-4. JAMA 311 (2): 183–92. doi:10.1001/jama.2013.284692. Stead, L. F., D. Buitrago, N. Preciado, G. Sanchez, Nikaj, S., and F. J. Chaloupka. 2014. “The Effect of Prices on J. Hartmann-Boyce, and others. 2013. “Physician Advice Cigarette Use among Youths in the Global Youth Tobacco for Smoking Cessation.” Cochrane Database of Systematic Survey.” Nicotine & Tobacco Research 16 (Suppl. 1): S16–23. Reviews 5: Cd000165. doi:10.1002/14651858.CD000165 doi:10.1093/ntr/ntt019. .pub4. Norheim, O. F., P. Jha, K. Admasu, T. Godal, R. J. Hum, and Tauras, J. A. 1999. “The Transition to Smoking Cessation: others. 2014. “Avoiding 40% of the Premature Deaths in Evidence from Multiple Failure Duration Analysis.” NBER Each Country, 2010–30: Review of National Mortality Working Paper 7412. National Bureau of Economic Trends to Help Quantify the UN Sustainable Development Research, Cambridge MA. Goal for Health.” The Lancet 385 (9964): 239–52. Tauras, J. A., and F. J. Chaloupka. 2004. “Impact of Tobacco Oliver J. 2015. Last Week Tonight with John Oliver: Tobacco. Control Spending and Tobacco Control Policies on https://www.youtube.com/watch?v=6UsHHOCH4q8. Adolescents’ Attitudes and Beliefs about Cigarette Smoking.” Palipudi, K. M., P. C. Gupta, D. N. Sinha, L. J. Andes, Evidence-Based Preventive Medicine 1 (2): 111–20. S. Asma, and others. 2012. “Social Determinants of Thun, M. J., B. D. Carter, D. Feskanich, N. D. Freedman, Health and Tobacco Use in Thirteen Low and Middle R. Prentice, and others. 2013. “50-Year Trends in Income Countries: Evidence from Global Adult Tobacco Smoking-Related Mortality in the United States.” The New Survey.” PLoS One 7 (3): e33466. doi:10.1371/journal England Journal of Medicine 368 (4): 351–64. doi:10.1056 .pone.0033466. /NEJMsa1211127. Peto, R., and A. D. Lopez. 2001. “The Future Worldwide Health U.S. Department of Health and Human Services. 2014. The Effects of Current Smoking Patterns.” In Critical Issues in Health Consequences of Smoking: 50 Years of Progress. Global Health, edited by E. C. Koop, C. E. Pearson, and A Report of the Surgeon General. Atlanta, GA: Centers for M. R. Schwarz. New York: Jossey-Bass. Disease Control and Prevention. 192 Cancer van Baal, P. H., W. B. Brouwer, R. T. Hoogenveen, and WHO (World Health Organization). 2010. WHO Technical T. L. Feenstra. 2007. “Increasing Tobacco Taxes: A Cheap Manual on Tobacco Tax Administration. Geneva: WHO. Tool to Increase Public Health.” Health Policy 82 (2): 142–52. ———. 2011. Scaling Up Action Against Non-Communicable Van Walbeek, C. P. 2006. “Industry Responses to the Tobacco Diseases: How Much Will It Cost? Geneva: WHO. Excise Tax Increases in South Africa.” South African Journal ———. 2013a. Global Health Observatory 2013. http://www of Economics 74: 110–22. .who.int/gho/mortality_burden_disease/en/. Verguet, S., C. L. Gauvreau, S. Mishra, M. MacLennan, ———. 2013b. WHO Report on the Global Tobacco Epidemic. S. M. Murphy, and others. 2015. “The Consequences of Geneva: WHO. Raising Tobacco Tax on Household Health and Finances WHO-GATS (World Health Organization–Global Adult among Richer and Poorer Smokers in China: An Extended Tobacco Survey). Various years. Global Adult Tobacco Cost-Effectiveness Analysis.” The Lancet Global Health Survey. WHO. http://www.who.int/tobacco/surveillance 3 (4): e206-e216. doi:10.1016/S2214-109X(15)70095-1. /gats/en/. Wakefield, M. A., L. Hayes, S. Durkin, and R. Borland. 2013. Yurekli, A., and O. Sayginsoy. 2010. “Worldwide Organized “Introduction Effects of the Australian Plain Packaging Cigarette Smuggling: An Empirical Analysis.” Journal of Policy on Adult Smokers: A Cross-Sectional Study.” BMJ Applied Economics 42 (5): 545–61. Open 3 (7): e003175. doi:10.1136/Bmjopen-2013-003175. Zaton ´czuk. 2010. “Tobacco Smoking and ´ski, W. A., and M. Man West, R., W. Zaton´ski, M. Cedzynska, D. Lewandowska, J. Pazik, Tobacco-Related Harm in the European Union with Special and others. 2011. “Placebo-Controlled Trial of Cytisine for Attention to the New EU Member States.” In Tobacco: Smoking Cessation.” New England Journal of Medicine 365 Science, Policy, and Public Health, edited by P. Boyle, (13): 1193–200. N. Gray, J. Henningfield, J. Seffrin, and W. A. Zaton ´ski. Whittaker, R., H. Mcrobbie, C. Bullen, R. Borland, A. Rodgers, Oxford, UK: Oxford University Press. and others. 2012. “Mobile Phone-Based Interventions Zaton´ski, W., K. Przewozniak, U. Sulkowska, R. West, and for Smoking Cessation.” Cochrane Database of Systematic A. Wojtyla. 2012. “Tobacco Smoking in Countries of the Reviews 11: Cd006611. doi:10.1002/14651858.CD006611 European Union.” Annals of Agricultural and Environmental .Pub3. Medicine 19 (2): 181–92. Global Hazards of Tobacco and the Benefits of Smoking Cessation and Tobacco Taxes 193 Chapter 11 Cancer Services and the Comprehensive Cancer Center Mary Gospodarowicz, Joann Trypuc, Anil D‘Cruz, Jamal Khader, Sherif Omar, and Felicia Knaul INTRODUCTION integrating theme for the framework. The chapter con- cludes by detailing the benefits that a comprehensive Most countries and numerous global and local organi- cancer center provides to a country’s cancer control and zations are addressing the challenges of cancer (Blanchet health care efforts. and others 2013; Knaul, Alleyne, Atun, and others 2012), including the development of comprehensive national cancer control programs designed to reduce the number Cancer System Functions of cancer cases and deaths and to improve the quality of Cancer system planning includes the development of life of cancer patients through evidence-based strategies population-based cancer plans, at the national or lower for prevention, early detection, diagnosis, treatment, levels. Cancer plans address all aspects of cancer con- and palliation. A national cancer control program trol, including cancer registries, practice and operating addresses the functions and delivery of many compo- standards, research, health care education and practice nents of cancer control (http://www.who.int/cancer standards, certification and accreditation of service pro- /nccp/en/). The delivery of most services is anchored viders, and system performance. in comprehensive cancer centers (Gralow and others 2012; Hensher, Price, and Adomakoh 2006; Sloan and Gelband 2007). Cancer System Components This chapter describes an optimal framework for The World Health Organization (WHO) (2006b) rec- a comprehensive cancer center, which can be a free- ommends that all nations have a cancer control plan standing dedicated institution, a program within an that includes these components: prevention, screening, academic health science center or a community hospital, diagnosis, treatment, survivorship, and palliative and or a group of hospitals providing an integrated program. end-of-life care (figure 11.1). The first section presents an overview of the frame- Many cancer control components are provided work for a comprehensive cancer center, which includes in comprehensive cancer centers, regardless of a three levels that are embedded within a comprehensive country’s resource level. WHO and others have rec- cancer system. Detailed information on each level is ommended that every country aim to have at least presented, followed by a discussion of quality as an one publicly supported cancer center that advances Corresponding author: Mary Gospodarowicz, MD FRCPC FRCR (Hon), Princess Margaret Cancer Centre, University of Toronto Mary.Gospodarowicz@rmp.uhn.on.ca 195 the broad objectives of control; provides exemplary By strengthening health system capacity, cancer centers patient care, appropriate to local circumstances and go beyond treating cancer as a vertical, disease-specific resources; and concentrates the specialized human program, to enable a diagonal approach that cuts across and technical resources of the country (Gralow and horizontal initiatives that target system-wide constraints others 2012; Knaul, Gralow, and others 2012; Sloan to address the overall goals of the health system (Knaul, and Gelband 2007). Alleyne, Piot, and others 2012). The capacity to develop comprehensive cancer systems varies with available resources, national governance, management effective- Comprehensive Cancer Centers in LMICs ness, public accountability, engagement of civil society, Many low- and middle-income countries (LMICs) are and other factors (Knaul, Alleyne, Atun, and others 2012; developing comprehensive cancer centers with public WHO 2012). or private resources. Patients can be managed directly Although this goal will take time to attain in at the centers; for many aspects of treatment, they can many countries, it is being successfully achieved in be managed in less specialized hospitals and local health multiple settings (Knaul, Gralow, and others 2012). clinics, with the center providing oversight and care For example, the King Hussein Cancer Center in plans. Comprehensive cancer centers educate health care Jordan, an upper-middle-income country, progressed professionals and the public, and they conduct research from offering limited access to poorly organized, on the causes, prevention, diagnosis, and treatment of low-quality cancer services to providing internation- cancer (National Cancer Institute 2012). ally accredited cancer care, engaging in cancer-related Comprehensive cancer centers can act as focal points education and research, leading national control for cancer control nationally (Sloan and Gelband 2007) planning efforts, and contributing to regional and and influence cancer and health system development. global efforts (box 11.1). Figure 11.1 Cancer Control Components Palliative /supportive Prevention Screening Diagnosis Treatment Survivorship /end-of-life care Source: Adapted from Cancer Care Ontario 2013. Box 11.1 King Hussein Cancer Foundation and Center, Jordan The King Hussein Cancer Foundation in Amman, The center evolved by: Jordan, is an independent, nongovernmental, non-profit organization that oversees the operations • Reversing the brain drain by convincing of the cancer center. The hospital first opened in accomplished clinical and executive leaders 1997 as the Al-Amal Center or Center of Hope and working in high-income countries to return was renamed the King Hussein Cancer Center in to the region to create the foundation for and September 2002. The center treats all types of cancer expansion of excellent cancer care in adults and children from the Middle East and • Designing and building a well-functioning and North Africa. appropriately equipped physical facility box continues next page 196 Cancer Box 11.1 (continued) • Raising the standard of care in surgery, systemic/ by JCI’s Clinical Care Program in cancer. The chemotherapy, radiation therapy, nursing center helped organize the Ministry of Health’s oncology, bone marrow transplantation, and national early detection and awareness program. psycho-oncology It is also leading an effort to establish a national • Adopting policies and procedures to ensure cancer control planning program. Internationally, effective, efficient, safe operations the European Arab Society of Oncology has recog- • Establishing cancer education, training, and nized the center as a Cancer Center of Excellence public awareness programs, including oncology for the training of cancer health workers from the fellowships and residency programs region. The center has signed agreements with Petra • Developing a research program University to establish the first diploma program • Collaborating with other centers to improve in tobacco dependence treatment in the region, cancer care, training, and research; these include as well as with the German Jordan University to St. Jude Children’s Research Hospital, H. Lee establish a diploma program in nursing oncology. Moffitt Cancer Center, and MD Anderson The center is a WHO collaborating center. The Cancer Center in the United States; the Hospital center and foundation are active in the Union for Sick Children and Princess Margaret Cancer for International Cancer Control (UICC) and are Centre in Canada; National Cancer Institute in helping other countries in the region to collaborate the Arab Republic of Egypt; American University with UICC. of Beirut, Lebanon; Augusta Victoria Hospital in The center continues to develop to meet increas- Israel; Stefan-Morsch Foundation in Germany; ing patient demand from Jordan and the region. and Leeds Cancer Centre in the United Kingdom. Construction is underway to double capacity by mid-2016. Capacity building is ongoing with In 2006, the King Hussein Cancer Center was recruitment of additional staff, including in cancer accredited as a hospital by Joint Commission subspecialties, as is strengthening of cancer research International (JCI); in 2007, the center was certified and education activities. FRAMEWORK FOR A COMPREHENSIVE disease control, symptom control—recommending CANCER CENTER appropriate interventions, and setting out the optimal timeframes for instituting and completing treatment. The comprehensive cancer center has three layers: clin- The patient care plans vary from simple to complex, and ical management, clinical services, and core services may require a range of services. (figure 11.2). Errors in clinical decisions can lead to increased The framework provides a reference point for plan- morbidity and disability, increased costs, and even ning for a comprehensive center, even if this is achieved premature death. For example, a recent study reported incrementally as funding and capacity are built up. that almost one-quarter of children with acute myeloid leukemia in El Salvador, Guatemala, and Honduras died from largely avertable treatment-related mortality Clinical Management (Gupta and others 2012). Patient Care Plans Clinical management sets standards for clinical decision Clinical Practice Guidelines making and formulating patient care plans. Patient care Clinical practice guidelines are developed to assist prac- plans are based on the histopathologic and/or molecular titioners and patients in deciding on appropriate care diagnosis identifying the type of cancer; the anatomic for their circumstances (Hensher, Price, and Adomakoh disease extent or stage; and the individual patient’s char- 2006). Comprehensive cancer centers play a leadership acteristics, such as age, comorbidities, and performance role in developing and promoting treatment guidelines status. Determining the best clinical management for locally and nationally. Center clinicians and research- cancer patients involves defining the goals of care—cure, ers work with professional organizations to develop Cancer Services and the Comprehensive Cancer Center 197 Figure 11.2 Framework for a Comprehensive Cancer Center r ce ol can co nsive ntr e reh mp Co s ce rvi se re s ce Co rvi se al nic Cli nt ge l na ica Comprehensive me ma Clin cancer center guidelines for a wide range of scenarios. Some exam- United States–based National Comprehensive Cancer ples include the U.S. Preventive Services Task Force for Network, which collaborates to produce international screening guidelines and Cancer Care Ontario’s Program adaptations and translations of its guidelines that may in Evidence-Based Care, which produces evidence-based include modifications based on local circumstances.4 guidance documents.1 Guidelines are not limited to Other examples include efforts in India to establish therapeutic interventions and include indications for a wide range of guidelines adapted to local resource medical imaging and other diagnostic interventions2 availability,5 and consensus group recommendations and for symptom management. Nursing and other for imaging techniques for head and neck cancers in allied health professions develop guidelines to organize Singapore and South Asia, as developed by Wee and and direct care.3 Guidelines must be adapted for use colleagues depending on resource availability (Wee and in resource-constrained settings (Anderson and others others 2009). Mexico has developed a series of Normas 2008; Kerr and Midgley 2010). Officiales Mexicanas (Official Mexican Standards) that Many clinical practice guidelines have been adopted guide cancer services and finance. and adapted for use in LMICs (Gralow and others 2012; Konduri and others 2012). Although the focus of most of these initiatives is broader than the compre- Clinical Services hensive cancer centers, their impact has influenced the Clinical service departments include facilities, equip- care of patients within centers. A well-known interna- ment, skilled personnel, and policies and procedures tional example is the Breast Health Global Initiative, to deliver diagnostic, treatment, or supportive care. which has developed evidence-based, economically The cancer center works to integrate these services feasible, and culturally appropriate guidelines for effectively. For example, there is no point in offering breast health and cancer control in LMICs (chapter screening if positive results cannot be followed up with 3, this volume [Anderson and others 2015]; El Saghir definitive diagnostic tests and, if needed, treatment. and others 2011; Varughese and Richman 2010). The Access to the full range of clinical services is critical matrix guideline spans the spectrum of breast health for timely and appropriate cancer diagnosis and treat- care, from early detection to treatment and palliation, ment. A timely and accurate diagnosis is critical, because and considers the available resources at each stage early detection makes the difference between a curable (Sloan and Gelband 2007). Another example is the cancer and an untreatable one. 198 Cancer Many clinical service departments require special control initiatives. These processes are especially useful accreditation and are subject to external review and in limited-resource settings. control, such as in radiation protection and safety for imaging and radiotherapy, external accreditation Pathology and Laboratory Medicine for laboratory services, and cell therapy. Accreditation Pathology and laboratory medicine, including blood standards may be regional or national, or they may be banking, are essential for diagnosing cancer by exam- international (Econex 2010).6 ining patient biologic specimens. Laboratory medicine services include pathology, hematology, biochemistry, Office- and Clinic-Based Ambulatory Care microbiology, and, increasingly, cytogenetic and molec- The initial patient encounter with a cancer system often ular testing, services that are not specific to cancer. happens in an office or clinic, where cancer-related Pathology and laboratory medicine services require ambulatory procedures, such as clinical visits, physical facilities equipped to handle biological specimens with examinations, Pap smears, blood samples, or endoscop- appropriate precautions, and specialized equipment to ies, can take place. Guidelines help to determine when process and analyze tissues, blood, serum, and body and where these procedures should occur and how fluids. Basic pathology can include the capability for they should be provided properly and safely by trained specimen fixation, embedding into paraffin, tissue slicing, staff. Ambulatory facilities may need special equipment and staining; modern facilities must include immunohis- to address the needs of various patient populations, tochemistry, flow cytometry, and molecular and cytoge- for example, special examining tables for gynecologic netic testing (Gralow and others 2012). Given that the malignancies, and chairs and special endoscopic equip- diagnosis of cancer, especially rare cancers, is complex, ment for assessing head and neck cancers. Depending subspecialty expertise or access to such expertise via on the activity and jurisdiction, special facilities to sup- international networks is required. Collaborative or twin- port these procedures may be required and should be ning initiatives have been developed to support pathology accredited. services in LMICs. Examples include a Ghana-Norway partnership as part of the Breast Health Global Initiative Medical Imaging (Diagnostic Radiology) (Masood and others 2008) and Partners in Health, Imaging is a critical technology for diagnosis, to assess which includes clinics in a number of LMICs (Haiti and the effects of cancer treatment and complications, mon- Rwanda) with close ties to the Brigham and Women’s itor for the recurrence of cancer, and screen the general Hospital and Dana-Farber Cancer Institute (Carlson and population for cancerous conditions. It also is used to others 2010). Laboratories require specialized accredita- guide interventional procedures such as biopsies under tion to ensure that processes are in place to optimize the ultrasound, computed tomography (CT), or magnetic quality of specimen procurement and reporting.7 resonance imaging (MRI); securing of vascular access; and therapeutic interventions, such as embolization and Surgery high frequency ultrasound tumor ablation. Imaging for Surgery is a fundamental element of cancer treatment cancer ranges from conventional X-rays to ultrasound, and its pertinence to LMICs is the subject of chapter 13 CT, MRI, and molecular imaging (nuclear medicine) of this volume (Dare and others 2015). Well-established with position emission tomography, frequently com- interventions have proven effective in reducing surgical bined with CT. risk and provide promising strategies to improve out- Imaging services require equipment and specialized comes (Weiser and Gawande 2015). The introduction staff, such as radiologists and radiology technologists, of standardized practices, such as the Surgical Safety to operate and maintain the equipment. International Checklist endorsed by WHO, has improved the out- organizations, such as the International Atomic Energy comes of surgical procedures in countries at all economic Agency, the International Society of Radiology, and levels (Farmer and others 2010; Gawande 2009; Haynes WHO, have developed and published standards and and others 2009; Lingard and others 2008; Lingard and guidelines for the safe installation, operation, and use others 2011; World Alliance for Patient Safety 2008). The of imaging equipment. This information is used to use of comprehensive standard policies and procedures create national and regional standards (for example, facilitates safe and efficient operations. Radiation Safety Institute of Canada 2014; Zaidi 2010). Picture archiving and communication systems and web- Radiation Therapy based systems allow for offsite evaluation and reporting Radiation therapy, or radiotherapy, involves the use of images and are useful in management of care in of ionizing radiation for therapeutic purposes. It is remote communities, remote mentoring, and quality the subject of chapter 14 in this volume. Strategies to Cancer Services and the Comprehensive Cancer Center 199 improve access to radiotherapy in LMICs have been friends, and caregivers of survivors (Centers for Disease suggested, including offering basic treatment techniques Control and Prevention 2013). and optimizing fractionation to increase the throughput Increasingly, as treatments become more successful on radiotherapy machines, encouraging competitive and life expectancy increases, patients face new issues. pricing, and supporting long-distance mentorship for With improvements in access to and quality of care, this programs in remote areas (Gralow and others 2012). will increasingly be the case in LMICs, where survivor- For many reasons, cobalt machines have frequently ship services are currently unavailable. been considered more appropriate for LMICs (IAEA Psychosocial support can be provided to patients 2008, 2010). These views are changing as access to more and their families by a broad range of people, depend- sophisticated technologies is improving. Although linear ing on the level of need. Complex mental health issues accelerators require a more reliable power supply, cobalt and social matters can benefit from the engagement units present a higher radiation safety risk and require of other health professionals, including primary care frequent source replacement, which presents a hazard providers, community health workers, spiritual guides, and additional expense. volunteers, friends, families, and other lay individuals. Comprehensive cancer centers should have a survivor- Systemic Cancer Therapy or Chemotherapy ship program with a range of professionals, including In systemic cancer therapy or chemotherapy, drugs psychiatrists, psychologists, social workers, nurses, ther- are distributed in the body through the bloodstream. apists, nutritionists, and educators, as well as patients in These drugs include chemotherapy, administered treatment and long-term survivors. intravenously or orally; hormones; and immune and molecular-targeted therapies. Systemic therapy is used alone or in combination with surgery and radiother- Core Services apy to reduce recurrence, improve survival (Gralow Core services are delivered by departments of admin- and others 2012; Valentini, Barba, and Gambacorta istration and management, human resources, informa- 2010), and help preserve organs. Chemotherapy alone tion technology and management, physical facilities, is used in hematologic cancers and in most metastatic pharmacy, infection prevention and control, qual- cancers. Chemotherapy facilities can be used for other ity assurance, and finance. The level of core services intravenous treatments, transfusions, minor proce- depends on the size of the center and whether it is a dures such as bone marrow biopsies, thoracentesis, designated standalone facility, part of a larger hospital, paracentesis, and lumbar puncture for cancer and or a consortium of providers. In the latter two instances, noncancerous conditions. Systemic therapy can usually the core services may not be specific to cancer and be delivered in specialized ambulatory facilities in hos- may be used for the management of other diseases and pital outpatient units and community-based medical injuries. Generally, the core services must meet accred- offices or clinics.8 itation and licensing standards and guidelines and Chemotherapy drugs may be expensive, although a are usually included in the hospital accreditation. The host of agents are now off patent and can be effective and lack of investment in core services leads to poor access used extensively in LMICs (Konduri and others 2012). to and performance of clinical services (Grimes and Several of the cancers endemic to the lowest income set- others 2011), including poor quality, inefficient use of tings are amenable to treatment with relatively low-cost resources, and negative impacts on health (Mavalankar chemotherapy, but treatment cost is still a major barrier and others 2005). for many in LMICs. External challenges to core service infrastructure can paralyze the best clinical service. For example, Palliative Care and Supportive Care long-term increases in the price of petroleum needed Palliative care aims to prevent or relieve suffering, pro- for medical supplies; transportation of goods, per- vide early identification and assessment of symptoms, sonnel, and patients; and fuel for lighting, heating, and address other physical, psychosocial, and spiritual cooling, and medical equipment may have significant issues (WHO 2006a). It is the subject of chapter 9 of this adverse impacts on health sectors in LMICs (Dalglish, volume (Cleary, Gelband, and Wagner 2015). Poulsen, and Winch 2013). In addition, the absence of robust supply chain management may result in Survivorship delays in pathology reporting because of the lack of Survivorship is defined as the care of persons diagnosed reagents, and insufficient maintenance may result in with cancer, from the time of diagnosis throughout their equipment breakdowns that limit access to imaging or lives, as well as the impact of cancer on family members, radiotherapy. 200 Cancer Administration and Management In cancer centers, IT includes health records; oper- Cancer care is complex and requires skilled and account- ational systems, such as human resources, pharmacy, able leadership and management at all levels. Generally, supplies, and equipment; financing; and other systems. hospitals with better management have better clinical IT also includes telemedicine and mobile informa- outcomes, and good management practices help to tion and communication technologies, such as cell preserve or enhance the quality of care (Carter, Dorgan, phones (mHealth), which improve access to services. and Layton 2011). Useful frameworks exist to help guide Telemedicine initiatives have the potential to decrease the development and ongoing excellence of admin- disparities in cancer care between resource-poor and istration and management. For example, the United resource-rich institutions by developing resources— States–based Baldrige Performance Excellence Program human capital and telecommunication infrastructure— focuses on performance excellence in leadership, stra- that link institutions with different levels of funding and tegic planning, customer focus, workforce focus, oper- expertise (Hazin and Qaddoumi 2010). ations focus, results and measurement, and analysis Although IT requires funding for capital, training, and knowledge management (Baldrige Performance ongoing maintenance, and technical backup, cancer Excellence Program, National Institute of Standards and centers need reliable electronic systems to manage the Technology, and U.S. Department of Commerce 2011); high volumes of information; inform safe, efficient, and the program has a self-assessment tool. effective care; and improve access. The systems can be especially important in LMICs for linking comprehen- Human Resources sive centers to more remote areas and less specialized Cancer centers require appropriately trained and licensed centers, as well as for linking to international exper- clinicians and administrative and support staff. Centers tise and networks (Knaul, Bhadelia, and others 2012; need to recruit and retain staff and provide professional Shekelle, Morton, and Keeler 2006). and career development opportunities to maintain com- petence and develop new skills. Core human resource Pharmacy services include identifying the roles and responsibilities Pharmacy services focus on safe and effective medica- of the range of positions within the center, setting com- tion use and include managing practice; adhering to pol- pensation and benefit levels, developing performance icies on medication use; optimizing medication therapy; evaluations, setting up management and supervisory procuring drug products and managing inventory; pre- structures, and providing conflict resolution services. paring, packaging, and labeling medications; delivering Making the best use of human resources means medications; monitoring medication use; evaluating the maximizing their impact. Human resources can be effectiveness of the medication-use system; and con- increased in LMICs and remote areas by using non- ducting research (American Society of Health-System specialists or general medical professionals working Pharmacists 2013). under specific conditions. This practice promotes task- Cancer pharmacy services reflect specialized knowl- shifting and optimizes the use of sparse, highly skilled edge about the medications used for cancer, manage- personnel. For example, the use of community health ment of cancer complications, treatment side effects, workers, expert patients, and clinical officers (Knaul, and drug toxicities. The complexity of caring for patients Bhadelia, and others 2012) and, in some countries, with cancer; the costs of chemotherapy; the potential traditional healers who play an important role in influ- for severe drug toxicity and medication errors; and the encing people’s health care decisions (Price and others requirements for safe preparation, administration, and 2012) will enhance the capacity for health care delivery. disposal of cytotoxic drugs highlight the important role Teleservices, such as telepathology, teleradiology, and of pharmacies in cancer centers, regardless of a country’s virtual consultation can offer support and guidance in resource level (Wiernikowski 2013). The International cancer to nonspecialists by tapping large international Society of Oncology Pharmacy Practitioners has devel- networks of highly trained professionals. oped Standards of Oncology Pharmacy Practice that take into account realities from resource-rich and resource- Information Technology and Management poor settings. Information technology (IT) refers to systems and their applications, for example, computer hardware and soft- Infection Control ware and telecommunications that collect, store, use, Infection control is a core service that focuses on pre- and share information. Information management refers venting and controlling infections in cancer patients, to organizing, linking, analyzing, and presenting data to including advice on the care of patients with infec- guide decisions. tions, especially those acquired in the cancer center. Cancer Services and the Comprehensive Cancer Center 201 Main infection prevention and control tactics include and standards for areas such as fire safety and radi- complying with hand hygiene, disinfecting and steril- ation protection, occupational health and safety, and izing surfaces and equipment, investigating and mon- infection prevention and control. Organizations and itoring suspected infections, managing difficult cases providers usually determine how the other ancillary and outbreaks, wearing personal protective equipment, services will be provided, depending on local circum- and vaccinating and educating health care providers. stances and resources. For centers, this includes introducing prevention bundles,9 improving compliance with hand hygiene, making prudent use of antimicrobials, translating CANCER CENTERS AND QUALITY OF CARE: research results into practice, and upgrading the capa- bilities of the microbiology laboratory (Raka 2010). AN INTEGRATING THEME Having well-developed and resourced centers and sys- Quality Assurance tems does not guarantee higher quality (Chalkidou and Cancer care has many potential risks. Complex clinical others 2014; WHO 2006c). Indeed, high-quality care can management using multiple treatment paths and multi- be achieved in centers with minimal resources. ple health care providers highlights the importance of a Many organizations have highlighted issues and centerwide commitment to a quality and safety agenda impacts of quality in health care (IOM 2000, 2001) and and ongoing performance improvement. cancer care (IOM 2013). Poor quality of care can lead Centers need to select appropriate indicators to to increased injury, morbidity, disability, and death for monitor and assess the quality and effectiveness of their patients. It also has financial, physical, and psychological structures (for example, setting and facilities), pro- impacts on patients and families; financial impacts on cesses (range of care), and outcomes (patients’ recovery, health care institutions and systems, especially if addi- restoration of function, and survival) (Donabedian tional health services are needed; and economic impacts 1966). Information systems should capture baseline on societies (IOM 2000). Definitions and frameworks, performance measures for each indicator and track along with quality measures, may also be influenced by changes over time. Cancer centers should regularly a variable focus on structures, processes, and outcomes monitor performance, identify problem areas, and focus of quality. improvement efforts in these areas. A review of conceptual quality frameworks in six Organisation for Economic Co-operation and Finance Development member countries (OECD, Kelley, and All cancer centers need competent financial systems to Hurst 2006) identified the most commonly used dimen- monitor revenues and expenses. Sources of funding vary sions of quality: widely and can include national and subnational gov- ernment funding; private user payments, either through • Effectiveness: The degree of achieving desirable out- health insurance or out of pocket; revenue-generating comes, given the correct provision of evidence-based practices, for example, retail and parking; and philan- health care services to all who could benefit but not thropic support from external donors. Available finances to those who would not benefit dictate the services that can be provided. Centers need • Safety: The degree to which health care processes systems that allow effective and efficient operations and avoid, prevent, and ameliorate adverse outcomes or ensure appropriate quality services to optimize the use injuries that stem from the processes of health care of funds. itself; closely related to effectiveness, albeit distinct in its emphasis on preventing unintentional adverse Additional Key Supports events for patients Additional key supports required in the cancer cen- • Responsiveness: The way a system treats people to ter include the following: equipment and technology meet their legitimate non-health expectations; also support services, supplies and materials management, known as patient-centeredness supply chain processes, patient transport, fire safety • Accessibility: The ease with which health services are and radiation protection, occupational health and reached; can be physical, financial, or psychological safety, and security. In areas of violence or conflict, and requires health services to be a priority and security services may be especially important for available patients and their families, as well as for guarantee- • Equity: Closely related to accessibility but assesses ing the safety of health inputs and avoiding robbery. health system financing and outcomes and health National and regional bodies generally set policies status 202 Cancer • Efficiency: The system’s optimal use of available Catalyzing the Development of Effective National resources to yield maximum benefits or results; Cancer Control Systems speaks to a system’s ability to function at lower cost A center’s critical mass of clinical management expertise without diminishing attainable and desirable results. and clinical and core services results in effective and efficient quality cancer control. Cancer centers can lead Other dimensions of quality identified included the development of regional systems of cancer care, with acceptability (related to patient-centeredness), appro- care ranging from very complex to basic interventions priateness (related to effectiveness), competency or and community-based care. Centers can contribute to capability (related to effectiveness), continuity (related national cancer control efforts by being a credible voice to patient-centeredness), and timeliness (related to for public education on prevention and the signs, symp- patient-centeredness). toms, and treatment of cancer. This contribution is espe- External accreditation, regulatory, licensing, and pro- cially important in LMICs, since many people present fessional and evidence-based clinical practice organiza- with advanced or metastatic disease. The establishment tions and bodies require cancer centers to meet quality of regional cancer centers in every state of India illustrates standards for organizations and how they should operate. the important contribution of these centers to supporting Countries or regions may have general accreditation an effective national cancer control system (box 11.2). standards as well as service-specific credentialing bodies10 (Econex 2010). These external organizations provide cancer centers with arm’s-length quality reference points Training Health Care Professionals to guide their operations. Accreditation is also an exter- Comprehensive centers play a significant role educat- nal motivator for quality reform and is consistently seen ing a country’s health care professionals. In addition as an effective driver for quality in LMICs (Barnett and to providing specialty training for individual profes- Hort 2013). sions, centers provide training on interprofessional Box 11.2 Regional Cancer Centers in India India’s active National Cancer Control Program was The main functions of RCCs are cancer detection launched by the government in 1975 and revised and diagnosis, treatment, aftercare and rehabilita- in 1984. The main focus is primary prevention tion, education and training, cancer registration, and early detection of cancer, which includes the and research. RCC core requirements include divi- following: sions of surgical oncology, radiation oncology, and medical oncology, with support from the depart- ments of anesthesiology, pathology, cytopathology, • Tobacco control measures to prevent tobacco- hematology, biochemistry, and radiologic diagnosis, related cancers with appropriate equipment and staff. • Screening for cancers of the uterine cervix, mouth, and breast Oncology units in medical colleges form an impor- • Extending and strengthening therapeutic services tant link between RCCs and the more peripheral nationally through regional cancer centers (RCCs) health infrastructure, that is, district hospitals, Tehsil and medical colleges, including dental colleges. (regional) hospitals, and primary health centers. This three-tier model will help to make cancer care The objectives of the program are to be met by accessible across all socioeconomic groups and geo- creating one RCC in every state and developing graphical areas. oncology units in existing medical colleges across At the peripheral level, a district cancer control the country. program was launched in 1990/91 with elements box continues next page Cancer Services and the Comprehensive Cancer Center 203 Box 11.2 (continued) of health education, early detection, training A wide disparity exists in the level of cancer care of medical and paramedic personnel, palliative across various centers in India. Efforts are under- treatment and pain relief, coordination, and way to create a national cancer grid linking major monitoring. oncology centers across the country to facilitate the following: Although the national cancer control program has been beneficial, given the geographic expanse and • Development of a cooperative cancer manage- the vast population, cancer care facilities remain ment network for the transfer of standard unavailable to the majority of the population from treatment guidelines and expertise lower socioeconomic strata and those living in • Facilitation of uniform standards for education, remote areas. For example, global standards require training, and human resource development in two radiotherapy treating units per 100,000 popu- cancer care lation; currently, India has 0.4 radiotherapy units • Creation of a cooperative oncology research net- per 100,000. work to conduct studies of national importance. team-based care. Other hospitals, community clinics, effective quality clinical management practices that are and primary care can provide training placement transferrable to all health care services. In addition, many opportunities (Debas and others 2006). Trained pro- of the clinical and core services in cancer centers—such fessionals can take on various roles and responsibili- as diagnostic imaging, pathology, surgery, and palliative ties throughout the country. LMICs that wish to train care—can support other clinical programs. Similarly, their own doctors need at least one teaching hospital the referral systems that cancer centers establish with a (Hensher, Price, and Adomakoh 2006), which, in most continuum of providers can meet other health needs. instances, would include the comprehensive cancer Innovative financing of cancer services through com- center. Given that every developing country will not be prehensive cancer centers can drive efforts to develop able to train a full complement of health professionals financial protection in health as part of universal health on its own (Frenk and others 2010) or train staff in coverage. In Mexico, for example, pediatric and women´s highly specialized skills, comprehensive cancer centers, cancers were among the first to be included in Seguro especially in developed countries, can be part of edu- Popular, the national public insurance program focused cation consortia that extend beyond national borders. on the poor. The visibility and effectiveness of these For example, when the treatment of pediatric malig- efforts helped to develop confidence among citizens, leg- nancy was expanded in Chile to include bone marrow islators, and policy makers alike regarding the feasibility transplantation, clinical staff needed specialty train- and importance of establishing financial protection in ing to support the development of this new program health (Atun and Knaul 2012; Knaul, González-Pier, and (Palma and others 2006). In collaboration with St. others 2012). Jude Children’s Research Hospital, Memphis, Tennessee, pediatric oncologists, nurses, and other specialists— immunologists, hematologists, intensivists, pathol- Contributing to Global Health ogists, and medical technologists—received training Comprehensive cancer centers can make important con- from international institutions, including St. Jude, Vall tributions to global health and health systems. Centers d’Hebron Hospital in Barcelona, and the Hospital de can contribute to broad global efforts to improve health Clínicas in Curitiba, Brazil. The experiences and survival (Frenk and Moon 2013). International health organi- outcomes of the program have been positive. zations that cross national boundaries can benefit from the participation of centers in such areas as research and development and sharing of information for ongoing Supporting the Development of Effective Health Care learning (Blanchet and others 2013; Jamison, Frenk, and Systems Knaul 1998). Comprehensive cancer centers guide and support the Successfully developing comprehensive cancer development of effective health systems. Centers model centers in LMICs requires locally developed and 204 Cancer driven approaches that consider national and subna- development of cancer centers and help to achieve a tional resources and circumstances. Gupta and others country’s cancer goals (Gralow and others 2012; Sloan (chapter 7, this volume) identify the basic personnel and Gelband 2007). Furthermore, research suggests and infrastructure requirements for the ideal dedicated that twinning improves cancer survival in LMICs childhood cancer treatment center in an LMIC setting. (Hazin and Qaddoumi 2010). Box 11.3 provides The authors note that satellite centers can be especially examples of beneficial twinning relationships. important for decreasing the abandonment of treat- Other LMICs have raised funds locally to finance the ment for children and recognize that much treatment development of cancer center services. Box 11.4 presents occurs despite the lack of ideal centers. Many LMICs the experience of establishing the Fakous Cancer Center have leveraged the experience, expertise, and resources in the Arab Republic of Egypt, which integrates cancer of high-income countries to develop cancer services. treatment with primary health care to help prevent and For example, twinning relationships can facilitate the treat cancer in a low-resource setting. Box 11.3 Twinning Relationships St. Jude Children’s Research Hospital’s survival rate for children with acute lymphoblastic International Outreach Program and 20 Partners leukemia increased from 10 percent to 70 percent.a The St. Jude Children’s Research Hospital’s International Outreach Program improves the Fred Hutchinson Cancer Research Center and survival rates of children with catastrophic ill- Uganda Cancer Institute nesses worldwide by transferring knowledge, tech- The Uganda Cancer Institute (UCI), the only cancer nology, and organizational skills to countries and treatment and training facility in the country regions, so they can become self-sufficient and of 32 million people, partnered with the Fred successfully treat children close to home. The Hutchinson Cancer Research Center in Seattle, program involves local communities, supports the Washington, to establish the UCI/Hutchinson development of regional expertise and diagnostic Center Cancer Alliance in 2004. The alliance capabilities, partners with medical institutions focuses on developing effective prevention and and fundraising organizations, and facilitates the treatment strategies for infection-associated involvement of other agencies and organizations cancers through the following activities: to support key programs and the education of local personnel. • Conducting advanced research in infection- Located in Memphis, Tennessee, the program related cancers to improve understanding of the has pediatric oncology twinning programs with pathogenesis of these diseases and to develop 20 partner sites in 14 countries, including Brazil, and test more effective and safer treatment and Chile, China, Costa Rica, Ecuador, El Salvador, prevention regimens Guatemala, Honduras, Jordan, Lebanon, Mexico, • Improving clinical capacity by providing medical Morocco, the Philippines, and the República support and revised clinical protocols for those Bolivariana de Venezuela. The results have been with infectious cancers significant; survival rates for childhood cancers • Training cancer specialists, scientists, and support increased and the rate of abandonment of treat- staff in Uganda to increase local human capacity ment decreased. For example, the abandonment for clinical care and research at UCI and provid- rate in El Salvador dropped from 13 percent to ing a training environment for United States– 3 percent from 2010 to 2012, and the five-year based personnel in Uganda.b box continues next page Cancer Services and the Comprehensive Cancer Center 205 Box 11.3 (continued) Victoria Hospice and B. P. Koirala Memorial Canada. Funds have been raised to help support Cancer Hospital, Bharatpur, Nepal the hospital’s patient care services and increase The International Network for Cancer Treatment health professional education. Medical supplies and Research established the Palliative Access have been purchased, local staff have been hired and Program to assist developing countries in initiating trained, travel funds have been provided for staff and sustaining palliative care programs. In 2007, training opportunities, and educational material the B. P. Koirala Memorial Cancer Hospital, Nepal’s has been provided and adapted. In addition, the national cancer hospital, expressed a desire to partners exchange mutually beneficial knowledge twin with a hospice to help expand its patient and expertise in palliative care.c care services, develop education and research, and a. For additional information, see: http://www.stjude.org/stjude/v/index.jsp introduce home and community-based palliative ?vgnextoid=2f166f9523e70110VgnVCM1000001e0215acRCRD&vgnextchannel care services. The hospital—which has a 12-bed =e41e6fa0a9118010VgnVCM1000000e2015acRCRD. b. For additional information, see: http://www.fredhutch.org/en/labs/vaccine-and inpatient palliative care unit and provides inpatient -infectious-disease/international-programs/global-oncology/uganda/uci-fred and outpatient consultations—twinned with -hutch.html. Victoria Hospice in Victoria, British Columbia, c. For additional information, see https://sites.google.com/site/nepalhospicetwin. Box 11.4 The Fakous Cancer Center In Fakous district in the northeast of the Arab treatment to less developed parts of the country. Republic of Egypt, breast cancer is the most com- The center has 80 beds; three operating rooms; mon cancer. Until the center was opened in 1992, the an eight-bed intensive care unit; basic diagnos- closest cancer treatment for the largely poor popu- tic facilities with conventional X-ray; ultrasound lation of 660,000 was the National Cancer Institute for ultrasound-guided biopsy, mammography, and in Cairo. It was a three-hour trip and a world apart, endoscopy; and a histopathology unit equipped and most cancer patients went untreated. to provide cytology, tissue analysis, and hormone One of many challenges confronted in building receptor assays, as well as treatment modali- the Fakous Cancer Center was financing. Using ties. Social support of cancer patients’ families is “crowd sourcing,” one million Egyptian pounds also provided. (US$330,000) was raised in the first two months, The center’s outpatient facilities provide free clinical and donations continued to come in. A second consultations for poor patients, who constitute the challenge was finding doctors to work in the center. majority in this region—nearly 230,000 outpatients In place of permanent staff, specialists from the in the past 22 years. The inpatient wards have seen National Cancer Institute and various universities 29,000 patients admitted. come to the center to perform surgery and provide Care at the center is reflected in survival statistics: other specialized treatment. The Fakous Cancer for women treated for breast cancer in 2008, the Center has become a center of excellence in train- five-year survival is 89 percent for stage I, 77 percent ing as well as treatment. The third challenge, the for stage II, 71 percent for stage III, and 19 percent retention of good nursing staff, was accomplished for stage IV. A recent study of the experience of the through the establishment of a nursing school. center also documents stage shift at diagnosis from The Fakous model integrates third-level services the time the center was opened through 2007–08 with primary health care, taking prevention and (Omar and others 2013). 206 Cancer CONCLUSIONS for Testing and Calibration Laboratories, http://www .nabl-india.org/. The optimal framework for establishing a comprehen- 8. See, for example, http://www.asco.org/institute-quality sive cancer center provides the nucleus around which an /asco-ons-standards-safe-chemotherapy-administration. entire cancer control program can be developed. Many 9. Bundles focus on aseptic procedures that potentially LMICs are developing comprehensive cancer centers carry a high risk of hospital-related infection, for exam- supported with public and private resources, and these ple, catheter-associated bloodstream infection, catheter- countries are using locally driven approaches appropri- associated urinary tract infection, ventilator-associated pneumonia, and surgical site infections. ate to their local circumstances. Most important, they 10. A number of accreditation bodies have international are having significant impacts on advancing cancer accreditation programs to inform centers in coun- control and improving the health of their populations. tries where national accreditation does not exist. See, for example, Joint Commission International, http:// www .jointcommissioninternational.org/achieve NOTES -accreditation/; Accreditation Canada International, http://www.internationalaccreditation.ca/en/home.aspx; World Bank Income Classifications as of July 2014 are as fol- and National Accreditation Board for Hospitals and lows, based on estimates of gross national income (GNI) per Healthcare Providers, http://www.nabh.co/. capita for 2013: • Low-income countries (LICs) = US$1,045 or less • Middle-income countries (MICs) are subdivided: REFERENCES a) lower-middle-income = US$1,046–US$4,125 American Society of Health-System Pharmacists. 2013. b) upper-middle-income (UMICs) = US$4,126–US$12,745 “ASHP Guidelines: Minimum Standard for Pharmacies in • High-income countries (HICs) = US$12,746 or more. Hospitals.” American Journal of Health-System Pharmacy 70: 1619–30. 1. For the Cochrane Collaboration, see http://www Anderson, B. O., J. Lipscomb, R. H. Murillo, and D. B. Thomas. .cochrane.org/cochrane-reviews. For the U.S. Preventive 2015. “Breast Cancer.” In Disease Control Priorities (third Services Task Force, see http://www.uspreventiveservices edition): Volume 3, Cancer, edited by H. Gelband, P. Jha, taskforce.org/recommendations.htm. For Cancer Care R. Sankaranarayanan, and S. Horton. Washington, DC: Ontario, see https://www.cancercare.on.ca/cms/One World Bank. .aspx?portalId=1377&pageId=10144. Anderson, B. O., C. H. Yip, R. A. Smith, R. Shyyan, S. F. Sener, 2. For example, see the Royal College of Radiologists, and others. 2008. “Guideline Implementation for Breast http://www.rcr.ac.uk/index.aspx, and the American Healthcare in Low-Income and Middle-Income Countries: College of Radiology, http://www.acr.org/Quality-Safety Overview of the Breast Health Global Initiative Global /Standards-Guidelines. Summit 2007.” Cancer 113 (8 Suppl.): 2221–43. doi:10.1002 3. For example, see the Oncology Nursing Society, http://www /cncr.23844. .ons.org/ClinicalResources; European Oncology Nursing Atun, R., and F. M. Knaul. 2012. “Innovative Financing: Society, http://www.cancernurse.eu/education/guidelines Local and Global Opportunities.” In Closing the Cancer .html; and Association of Oncology Social Work, http://www Divide: An Equity Imperative, edited by F. M. Knaul, .aosw.org/aosw/Main/professionals/standards-of-practice J. Gralow, R. Atun, and A. Bhadelia, for the Global Task /AOSWMain/Professional-Development/standards-of-practice. Force on Expanded Access to Cancer Care and Control in aspx?hkey=51fda308-28bd-48b0-8a75-a17d01251b5e. Developing Countries, chapter 8. Boston, MA: Harvard 4. See http://www.nccn.org/international/international Global Equity Initiative. _adaptations.asp. Baldrige Performance Excellence Program, National Institute 5. See https://tmc.gov.in/clinicalguidelines/clinical.htm. of Standards and Technology, and U.S. Department of 6. For example, see the Joint Commission International, Commerce. 2011. 2011–2012 Health Care Criteria for http://www.jointcommissioninternational.org/achieve Performance Excellence. Gaithersburg, MD: Baldrige. http:// -accreditation/; National Accreditation Board for Hospitals www.nist.gov/baldrige/publications/archive/2011_2012 and Healthcare Providers International, http://www.nabh _healthcare_criteria.cfm. .co/Index.aspx; and Accreditation Canada International, Barnett, C., and K. Hort. 2013. “Approaches to Regulating http://www.internationalaccreditation.ca/en/home.aspx. the Quality of Hospital Services in Low- and Middle- 7. For example, see the International Federation of Clinical Income Countries with Mixed Health Systems: Chemistry and Laboratory Medicine, http://www.ifcc A Review of Their Effectiveness, Context of Operation .org/executive-board-and-council/regional-federations and Feasibility.” Working Paper Series 32, The Nossal /efcc-european-federation-of-clinical-chemistry/; National Institute for Global Health, University of Melbourne, Accrediting Agency for Clinical Laboratory Sciences, Australia. http://ni.unimelb.edu.au/__data/assets/pdf http://www.naacls.org/; and National Accreditation Board _file/0010/834418/WP_32.pdf. Cancer Services and the Comprehensive Cancer Center 207 Blanchet, N., M. Thomas, R. Atun, D. T. Jamison, F. Knaul, in Countries of Low and Middle Income: A Call to and others. 2013. “Global Collection Action in Health: Action.” The Lancet 376 (9747): 1186–93. doi:10.1016 The WDR+20 Landscape of Core and Supportive /S0140-6736(10)61152-X. Functions.” Working Paper, The Lancet Commission on Frenk, J., L. Chen, Z. A. Bhutta, J. Cohen, N. Crisp, and Investing in Health. others. 2010. “Health Professionals for a New Century: Cancer Care Ontario. 2013. “The Cancer Journey.” https:// Transforming Education to Strengthen Health Systems in www.cancercare.on.ca/ocs/csoverview/. an Interdependent World.” The Lancet 376: 1923–58. Carlson, J. W., E. Lyon, D. Walton, W.-C. Foo, A. C. Sievers, and Frenk, J., and S. Moon. 2013. “Governance Challenges in others. 2010. “Partners in Pathology: A Collaborative Model Global Health.” The New England Journal of Medicine 368 to Bring Pathology to Resource Poor Settings.” American (10): 936–42. Journal of Surgical Pathology 34 (1): 118–23. Gawande, A. 2009. The Checklist Manifesto: How to Get Things Carter, K., S. Dorgan, and D. Layton. 2011. “Why Hospital Right. London: Metropolitan Books. Management Matters.” Health International, McKinsey’s Gralow, J. R., E. Krakauer, B. O. Anderson, A. Ilbawi, P. Porter, Health Systems and Services Practice, McKinsey and and others. 2012. “Core Elements for Provision of Cancer Company, London. Care and Control in Low and Middle Income Countries.” Centers for Disease Control and Prevention. 2013. “Basic In Closing the Cancer Divide: An Equity Imperative, edited Information about Cancer Survivorship.” http://www.cdc by F. M. Knaul, J. Gralow, R. Atun, and A. Bhadelia, for the .gov/cancer/survivorship/basic_info/index.htm. Global Task Force on Expanded Access to Cancer Care and Chalkidou, K., P. Marquez, P. K. Dhillon, Y. Teerawattananon, Control in Developing Countries, 125–65. Boston, MA: T. Anothaisintawee, and others. 2014. “Evidence-Informed Harvard Global Equity Initiative. Frameworks for Cost-Effective Cancer Care and Prevention Grimes, C. E., K. G. Bowman, C. M. Dodgion, and C. B. D. Lavy. in Low, Middle, and High-Income Countries.” The Lancet 2011. “Systematic Review of Barriers to Surgical Care Oncology 15 (3): e119–31. http://dx.doi.org/10.1016/S1470 in Low-Income and Middle-Income Countries.” World -2045(13)70547-3. Journal of Surgery 35: 941–50. Cleary, J., H. Gelband, and J. Wagner. 2015. “Cancer Pain Relief.” Gupta, S., M. Bonilla, P. Valverde, L. Fu, S. C. Howard, and In Disease Control Priorities (third edition): Volume 3, others. 2012. “Treatment-Related Mortality in Children Cancer, edited by H. Gelband, P. Jha, R. Sankaranarayanan, with Acute Myeloid Leukaemia in Central America: and S. Horton. Washington, DC: World Bank. Incidence, Timing and Predictors.” European Journal of Dalglish, S. L., M. N. Poulsen, and P. J. Winch. 2013. Cancer 48: 1363–69. “Localization of Health Systems in Low- and Middle- Haynes, A. B., T. G. Weiser, W. R. Berry, S. R. Lipsitz, Income Countries in Response to Long-Term Increases in A. H. Breizat, and others. 2009. “A Surgical Safety Checklist Energy Prices.” Globalization and Health 9: 56. http://www to Reduce Morbidity and Mortality in a Global Population.” .globalizationandhealth.com/content/9/1/56. New England Journal of Medicine 360: 491–99. Dare, A. J., B. O. Anderson, R. Sullivan, C. S. Pramesh, Hazin, R., and I. Qaddoumi. 2010. “Teleoncology: Current and C.-H. Yip, and others. 2015. “Surgical Services for Future Applications for Improving Cancer Care Globally.” Cancer Care.” In Disease Control Priorities (third edi- The Lancet Oncology 11 (2): 204–10. tion): Volume 3, Cancer, edited by H. Gelband, P. Jha, Hensher, M., M. Price, and S. Adomakoh. 2006. “Referral R. Sankaranarayanan, and S. Horton. Washington, DC: Hospitals.” In Disease Control Priorities in Developing World Bank. Countries, 2nd ed., edited by D. T. Jamison, J. G. Breman, Debas, H. T., R. Gosselin, C. McCord, and A. Thind. 2006. A. R. Measham, G. Alleyne, M. Claeson, D. B. Evans, P. Jha, “Surgery.” In Disease Control Priorities in Developing A. Mills, and P. Musgrove, 1229–43. Washington, DC: Countries, 2nd ed., edited by D. T. Jamison, J. G. Breman, World Bank and Oxford University Press. A. R. Measham, G. Alleyne, M. Claeson, D. B. Evans, P. Jha, IAEA (International Atomic Energy Agency). 2008. Setting A. Mills, and P. Musgrove, 1245–59. Washington, DC: Up a Radiotherapy Programme: Clinical, Medical Physics, World Bank and Oxford University Press. Radiation Protection and Safety Aspects. Vienna: IAEA. Donabedian, A. 1966. “Evaluating the Quality of Medical ———. 2010. AGaRT (The Advisory Group on increasing Care.” Milbank Memorial Fund Quarterly 44 (3 Suppl.): access to Radiotherapy Technology in low and middle 166–203. income countries). Brochure. http://cancer.iaea.org Econex. 2010. Accreditation of Healthcare Providers. Health /documents/AGaRTBrochure.pdf. Reform Note 2. http://econex.co.za/publication/health-reform IOM (Institute of Medicine). 2000. To Err Is Human: Building -note-2/. a Safer Health System, edited by L. T. Kohn, J. M. Corrigan, El Saghir, N. S., C. A. Adebamowo, B. O. Anderson, and M. S. Donaldson. Washington, DC: National Academies R. W. Carlson, P. A. Bird, and others. 2011. “Breast Press. Cancer Management in Low Resource Countries (LRCs): ———. 2001. Crossing the Quality Chasm: A New Health Consensus Statement from the Breast Health Global System for the 21st Century. Washington, DC: National Initiative.” The Breast 20 (Suppl. 2): S3–11. Academies Press. Farmer, P., J. Frenk, F. M. Knaul, L. N. Shulman, G. Alleyne, ———. 2013. Delivering High-Quality Cancer Care: and others. 2010. “Expansion of Cancer Care and Control Charting a New Course for a System in Crisis. Washington, 208 Cancer DC: National Academies Press. http://www.nap.edu/catalog Anesthesiologists to Reduce Failures in Communication.” .php?record_id=18359. Archives of Surgery 143 (1): 12–17. Jamison, D. T., J. Frenk, and F. Knaul. 1998. “International Masood, S., L. Vass, J. A. Ibarra, B. M. Ljung, H. Stalsberg, and Collective Action in Health: Objectives, Functions, and others, on behalf of the Breast Health Global Initiative Rationale.” The Lancet 351: 514–17. Pathology Focus Group. 2008. “Breast Pathology Guideline Kerr, D. J., and R. Midgley. 2010. “Can We Treat Cancer for a Implementation in Low- and Middle-Income Countries.” Dollar a Day? Guidelines for Low-Income Countries.” New Cancer 113 (Suppl. 8): 2297–304. England Journal of Medicine 363 (9): 801–03. Mavalankar, D. V., K. V. Ramani, A. Patel, and P. Sankar. Knaul, F. M., G. Alleyne, R. Atun, F. Bustreo, J. R. Gralow, and 2005. Building the Infrastructure to Reach and Care for others. 2012. “Strengthening Stewardship and Leadership to the Poor: Trends, Obstacles and Strategies to Overcome Expand Access to Cancer Care and Control.” In Closing the Them. Ahmedabad, India: Center for Management of Cancer Divide: An Equity Imperative, edited by F. M. Knaul, Health Services, Indian Institute of Management. http:// J. Gralow, R. Atun, and A. Bhadelia, for the Global Task www. iimahd .ernet.in /publications/data/2005-03-01 Force on Expanded Access to Cancer Care and Control mavalankar.pdf. in Developing Countries, 311–42. Boston, MA: Harvard National Cancer Institute (National Institutes of Health, U.S. Global Equity Initiative. Department of Health and Human Services). 2012. The Knaul, F. M., G. Alleyne, P. Piot, R. Atun, J. R. Gralow, and National Cancer Program: Managing the Nation’s Research others. 2012. “Health System Strengthening and Cancer: Portfolio. An Annual Plan and Budget Proposal for Fiscal A Diagonal Response to the Challenge of Chronicity.” Year 2013. Bethesda, MD: National Institutes of Health In Closing the Cancer Divide: An Equity Imperative, edited (NIH Publication No. 13-7957). by F. M. Knaul, J. Gralow, R. Atun, and A. Bhadelia, for the OECD (Organisation for Economic Co-operation and Global Task Force on Expanded Access to Cancer Care and Development), E. Kelley, and J. Hurst. 2006. “OECD Health Control in Developing Countries, 95–124. Boston, MA: Care Quality Indicators Project Conceptual Framework Harvard Global Equity Initiative. Paper.” OECD, Paris. Knaul, F. M., A. Bhadelia, R. Bashshur, A. J. Berger, Omar, A., M. Abdelgawad, M. Aboserea, and O. Omar. 2013. A. Binagwaho, and others. 2012. “Innovative Delivery of “Breast Cancer Experience at Fakous Cancer Center (FCC).” Cancer Care and Control in Low-Resource Scenarios.” In Pan Arab Journal Oncology 6 (3): 12–17. Closing the Cancer Divide: An Equity Imperative, edited by Palma, J., C. Mosso, C. Paris, M. Campbell, X. Tong, and others. F. M. Knaul, J. Gralow, R. Atun, and A. Bhadelia, for the 2006. “Establishment of a Pediatric HSCT Program in a Global Task Force on Expanded Access to Cancer Care and Public Hospital in Chile.” Pediatric Blood Cancer 46: 803–10. Control in Developing Countries, 165–95. Boston, MA: Price, A. J., P. Ndom, E. Atenguena, J. P. M. Nouemssi, and Harvard Global Equity Initiative. R. W. Ryder. 2012. “Cancer Care Challenges in Developing Knaul, F. M., E. González-Pier, O. Gómez-Dantés, D. García- Countries.” Cancer 118: 3627–35. Junco, H. Arreola-Ornelas, and others. 2012. “The Quest for Radiation Safety Institute of Canada. 2014. Canadian and Universal Health Coverage: Achieving Social Protection for Provincial Regulatory Documents. http://www.radiation All in Mexico.” The Lancet 380 (9849): 1259–79. safety.ca/resources/regulatory-documents. Knaul, F. M., J. R. Gralow, R. Atun, A. Bhadelia, J. Frenk, Raka, L. 2010. “Prevention and Control of Hospital-Related and others. 2012. “Closing the Cancer Divide: Overview Infections in Low and Middle Income Countries.” The and Summary.” In Closing the Cancer Divide: An Equity Open Infectious Diseases Journal 4: 125–31. Imperative, edited by F. M. Knaul, J. Gralow, R. Atun, and Shekelle P., S. C. Morton, and E. B. Keeler. 2006. “Costs A. Bhadelia, for the Global Task Force on Expanded Access and Benefits of Health Information Technology. Rockville to Cancer Care and Control in Developing Countries, 3–28. (MD): Agency for Healthcare Research and Quality.” Boston, MA: Harvard Global Equity Initiative. Evidence Reports/Technology Assessments, 132. http:// Konduri, N., J. Quick, J. R. Gralow, M. Samiei, P. Castle, and www.ncbi.nlm.nih.gov/books/NBK37988/ and Benefits of others. 2012. “Access to Affordable Medicines, Vaccines, Health Information Technology. and Health Technologies.” In Closing the Cancer Divide: Sloan, F. A., and H. Gelband. 2007. Cancer Control Opportunities An Equity Imperative, edited by F. M. Knaul, J. Gralow, in Low- and Middle-Income Countries. Washington, DC: R. Atun, and A. Bhadelia, for the Global Task Force National Academies Press. on Expanded Access to Cancer Care and Control in Valentini, V., M. C. Barba, and M. A. Gambacorta. 2010. “The Developing Countries, 197–256. Boston, MA: Harvard Role of Multimodality Treatment in M0 Rectal Cancer: Global Equity Initiative. Evidence and Research.” European Review for Medical and Lingard, L., G. Regehr, C. Cartmill, B. Orser, S. Espin, and Pharmacological Sciences 14 (4): 334–41. others. 2011. “Evaluation of a Preoperative Team Briefing: Varughese, J., and S. Richman. 2010. “Cancer Care Inequity for A New Communication Routine Results in Improved Women in Resource-Poor Countries.” Reviews in Obstetrics Clinical Practice.” BMJ Quality & Safety 20: 475–82. and Gynecology 3 (3): 122–32. Lingard, L., G. Regehr, B. Orser, R. Reznick, G. R. Baker, and Wee, J. T., B. O. Anderson, J. Corry, A. D’Cruz, K. C. Soo, others. 2008. “Evaluation of a Preoperative Checklist and others. 2009. “Management of the Neck after and Team Briefing among Surgeons, Nurses, and Chemoradiotherapy for Head and Neck Cancers in Asia: Cancer Services and the Comprehensive Cancer Center 209 Consensus Statement from the Asian Oncology Summit http://www .who .int/management/quality/assurance/ 2009.” The Lancet Oncology 10 (11): 1086–92. QualityCare_B.Def.pdf. Weiser, T. G., and A. Gawande. “Excess Surgical Mortality: ———. 2012. Assessing National Capacity for the Prevention Strategies for Improving Quality of Care.” 2015. In Disease and Control of Noncommunicable Diseases: Report of the Control Priorities (third edition): Volume 1, Essential 2010 Global Survey. Geneva: WHO. Surgery. Washington, DC: World Bank. Wiernikowski, J. T. 2013. “Equitable Access to Chemotherapy, WHO (World Health Organization). 2006a. Palliative Supportive Care Drugs and Pharmaceutical Care: A Care: Cancer Control—Knowledge into Action. WHO Challenge for Low- and Middle-Income Countries.” Cancer Guide for Effective Programs: Module 5. Geneva: WHO. Control 146. http://globalhealthdynamics.co.uk/cc2013 http://www.who.int/cancer/publications/cancer_control /wp-content/uploads/2013/04/146-isopp.pdf. _palliative/en/index.html. World Alliance for Patient Safety. 2008. Implementation Manual: ———. 2006b. Planning: Cancer Control—Knowledge into WHO Surgical Safety Checklist, 1st ed. Geneva: World Action. WHO Guide for Effective Programmes: Module 1. Health Organization. http://www.who.int/patientsafety Geneva: WHO. http://www.who.int/cancer/publications /safesurgery/ss_checklist/en/index.html. /cancer_control_planning/en/index.html. Zaidi, Z. 2010. “Accreditation Standards for Medical Imaging ———. 2006c. Quality of Care: A Process for Making Services.” Indian Journal of Radiology and Imaging 20 (2): Strategic Choices in Health Systems. Geneva: WHO. 89–91. 210 Cancer Chapter 12 Screening for Cancer: Considerations for Low- and Middle-Income Countries Terrence Sullivan, Richard Sullivan, and Ophira M. Ginsburg INTRODUCTION prevent one death from that cancer, balanced with how many people who undergo screening have a positive The goal of cancer screening is to detect cancer or or abnormal test result when they do not have cancer precancerous lesions in asymptomatic individuals at (false-positive test), and how many have a normal result a point when cancer is more likely to be prevented when they in fact do have cancer (false-negative test). or cured than if the patient waited for symptoms to The number of individuals with positive results who will develop (Morrison 1992). A screening intervention can actually proceed to follow up and receive treatment is a be successful only if the disease is more likely to be cured critical issue to consider for a given population. Other when detected early, and for which effective treatment for critical considerations include the cost effectiveness of a early-stage disease is available, affordable, and acceptable screening intervention when moving from initial trials to the individual, the community, and the jurisdiction of to scale and the health system requirements needed to interest. This chapter briefly describes the principles and ensure the success of a given program. (See chapters 11, pitfalls of cancer screening, based largely on the expe- 16, and 17 for more on health systems.) rience in high-income countries (HICs); summarizes In this chapter, we selected three cancer sites for the evidence for screening “best buys” relevant to low- which there is the most evidence for screening effective- and middle-income countries (LMICs); and highlights ness in LMICs—breast, cervical, and colorectal—and opportunities to avoid some of the costly and vexatious three promising candidate conditions. problems associated with screening in HICs and LMICs. The chapter focuses principally on existing projects and recent literature on cancer screening in LMICs. DEFINITIONS OF AND CRITERIA FOR Policy considerations regarding whether and in what CANCER SCREENING manner to implement a cancer screening program should be based on systematic evaluation of several Opportunistic versus Organized Screening factors, including at a minimum: the burden of the Opportunistic screening or case finding occurs when cancer in the population of interest (those at risk), an asymptomatic individual actively seeks a screening the cost effectiveness of the proposed screening inter- procedure or a health professional offers a screening test vention, and how well a given screening test performs to an asymptomatic individual. in the target population. How well the test works can Organized screening occurs when there is an orga- be judged by how many individuals must be screened to nized, population-based program with a structured Corresponding author: Terrence Sullivan, PhD, terrence.sullivan@utoronto.ca 211 public health approach. Organized screening has six Cervical cancer screening should begin only after a elements (IARC 2005): woman has become sexually active. When considering the choice of screening method, HPV screening is not advised • An explicit policy that specifies eligible age categories, until a woman is 30 years of age, as younger women are methods, and screening intervals more likely to naturally “clear” the virus through the • A defined target population immune system. Overtreatment, particularly of young • A dedicated and responsible management team women, may lead to fertility problems in the future responsible for implementation (chapter 4 in this volume [Denny and others 2015]). • Associated teams for decision and care The optimal frequency or interval for cancer • Specified methods for quality assurance screening depends on the capacity of the health • Screening methods to identify cancer occurrence in system, as well as the cancer’s natural history, which the target population. includes the rate of growth. Fast-growing cancers are less amenable to screening, while slower-growing, In population-based screening, the elements of the indolent cancers with a more predictable natural screening pathway are planned for an entire population history (for example, colonic polyps or cervical pre- and are delivered, monitored, and evaluated for effec- cancerous lesions) are more obvious candidates for tiveness and quality to ensure that the benefits are max- a screening intervention (Esserman, Thompson, and imized in a cost-effective way. Although the approach to Reid 2013). Breast cancer has many different subtypes implementation may be phased or staged geographically (for example, estrogen and/or progesterone receptor or by age intervals, the intention for population screen- positive and negative, her2neu positive and nega- ing is to capture all at-risk individuals in the appropriate tive) with a broad range of growth rates, patterns of age interval. Organized screening is expensive and can spread (metastases), and prognoses (Carey and others succeed only if adequate resources exist to achieve the 2006; Van de Vijver and others 2002). This complex full trajectory of screening, with program quality assur- natural history of breast cancer and the expense of ance, including effective reach to all in the target popu- subtyping breast cancer are among the reasons for lation group (appropriate age, gender, and risk category) the ongoing debate regarding the utility of screening and follow-up for disease assessment, diagnosis, and mammography in HICs. treatment if disease is discovered. It is important to consider potential sources of bias High-risk screening targets known subpopulations when evaluating the effectiveness of organized cancer of men or women who may be at considerably higher screening programs. Three such forms of bias are lead- risk for specific cancer because of their genetic or risk time bias, length bias, and overdiagnosis. exposure backgrounds. In HICs, such high-risk screen- ing has included known single-gene mutations asso- • Lead-time bias ciated with breast or ovarian cancer—such as BRCA1 Survival time for cancer patients is usually measured and BRCA2 mutations, or family history of breast from the day the cancer is diagnosed until the day or ovarian cancer—as well as similarly rare forms they die. Patients are often diagnosed after they of hereditary colon cancers. In LMICs, a pragmatic have symptoms. If a screening test leads to a diag- example of screening of high-risk groups in South and nosis before symptoms develop, the survival time is Southeast Asia for oral cancer could apply to heavy increased because the date of diagnosis is earlier. This smokers and drinkers who chew betel, areca nut, paan, increase in survival time makes it seem as though and gutka. screened patients are living longer when that may The target age range of a screening program depends not be the case. This is called lead-time bias. Screened on several factors, including the following: patients may die at the same time they would have without the screening test. • Burden of the cancer in a given population Lead-time bias has been a particular challenge • Age-specific trends of the cancer, which may vary for screening with prostate specific antigen in HICs. widely between countries As part of the American Board of Internal Medicine’s • Screening modality, the type of test used, for exam- Choosing Wisely campaign, the American Society ple, visual inspection with acetic acid (VIA) versus of Clinical Oncology added prostate screening to its human papilloma virus (HPV) testing or combina- updated “Top Five List” of oncology practices that tions of these for cervical cancer should be stopped because they are not supported by • Considerations regarding the capacity of local health the evidence or are considered wasteful (Schnipper systems. and others 2013). 212 Cancer • Length bias Another source of potential bias is apparent when Box 12.1 screening detects mostly indolent, slowly progressive tumors while missing the more aggressive ones. As an Principles of Early Disease Detection example, some types of breast cancer are indolent and can be asymptomatic for years; others are much more Condition aggressive and have a far shorter asymptomatic period. • The condition should be an important health The latter are more likely to cause symptoms between problem. screening intervals and may cause a patient to seek • There should be a recognizable latent or early symp- medical attention prior to ever participating in cancer tomatic stage. screening. Consequently, a screening test will detect • The natural history of the condition, including devel- more slow-growing than fast-growing cancers, giving opment from latent to declared disease, should be ade- a false impression that screening lengthens survival, quately understood. when in fact it is merely detecting a subset of a more treatable disease (Family Practice Notebook 2011). Test • There should be a suitable test or examination. • Overdiagnosis • The test should be acceptable to the population. Interest in cancer screening in LMICs is growing at a time when concerns about overdiagnosis and Treatment overtreatment, with resulting costs to the health care • There should be an accepted treatment for patients with system, as well as the psychosocial, physical, and eco- recognized disease. nomic risks incurred by individuals are increasingly a matter of concern in HICs. Overdiagnosis is the Screening Program diagnosis of disease that will never cause symptoms • There should be an established policy on whom to treat or death during a patient’s lifetime. It can be viewed as patients. as a side effect of testing for early forms of disease • Facilities for diagnosis and treatment should be that may turn people into patients unnecessarily and available. The cost of case-finding, including diagnosis may lead to treatments that do no good and perhaps and treatment, should be economically balanced in do harm. This is especially relevant for breast and relation to possible expenditure on medical care as a prostate cancer (Esserman, Thompson, and Reid whole. 2013; Welch, Schwartz, and Woloshin 2011; Yaffe and • Case-finding should be a continuing process and not a Pritchard 2014). Overdiagnosis in breast screening is “once and for all” project. discussed further in the section on breast screening in this chapter. Another important debate in HICs is about how Source: Adapted from Wilson and Junger 1968. much screening causes harm from a false-positive screening test, which often leads to significant wait- times for additional imaging tests and/or a tissue biopsy for what ultimately proves to be a benign find- ing. False-positive screens must be balanced against the SYSTEM REQUIREMENTS FOR ORGANIZED benefits conferred by finding screen-detected cancers CANCER SCREENING that genuinely extend survival and reduce mortality. Infrastructure, Education, and Advocacy The reality in many LMICs is quite different from Criteria for Cancer Screening that in HICs with longstanding cancer health pro- Screening for cancer can be effective if the criteria are motion efforts and organized screening programs. met. The Wilson-Junger (1968) criteria (box 12.1) set Organized approaches to screening risk straining the out a series of considerations that, notwithstanding burden on already thin health care and public health updates in an era of molecular and genetic diagnostics, resources. Delayed presentation for cancer is the norm remain worthy criteria to help make an assessment. in many LMICs and within low-resource or geo- Modern variants of the criteria extend to the consider- graphically remote regions in upper-middle-income ation of genetic susceptibility, in addition to preclinical countries. This delay exists for a variety of struc- disease or precursors (Goel 2001). tural, equity, and sociocultural reasons (Knaul, Frenk, Screening for Cancer: Considerations for Low- and Middle-Income Countries 213 and Shulman 2011; Story and others 2012). Structural and accurate pathology result. For those with a cancer obstacles include the following: diagnosis, appropriate referral for evidence-based and resource-appropriate treatment planning is essential, • Strained health infrastructure, for example, the lack begging again the capacity to make it so in LMICs. of available human and technical resources for proper Those with a negative screening test should be offered diagnosis and disease management “invitations” for their next round of screening, according • Long distances and poor road conditions that render to local guidelines (for example, a woman age 60 years proper care inaccessible who is of average population risk can be invited by mail • Sociocultural barriers, including extreme poverty, or telephone to schedule her next screening mammo- myths, and stigma about cancer gram two years from the last negative screen). • Gender inequity, which is especially relevant to breast and cervical cancer (Errico and Rowden 2006; Ginsburg 2013; Price and others 2012; Vorobiof, Cost-Effectiveness Considerations Sitas, and Vorobiof 2001) Cost considerations should include excess direct and indirect health care expenditures for cancers detected at Such obstacles underscore the need to incorporate an advanced stage, including out-of-pocket expenses and a range of decisions in LMICs to inform the optimal caregivers’ time away from work. Any analysis should approach to screening. Options vary from an oppor- also consider the case for such investment, describing tunistic case-finding approach, to a population-based macroeconomic cost models and potential savings from screening model, to a high-risk screening approach. treatment and prevention of cancer sites for which pre- Regardless of the approach taken, a new cancer screening vention or early detection can have the largest impact on program will contribute to increasing the number of morbidity and mortality (Knaul, Frenk, and Shulman prevalent cases. This additional burden of disease can 2011). Estimated losses are presented with more- or less- be substantial and should be viewed as a potential strain conservative estimates of avoidable deaths. on local capacity at all levels—public health, primary According to these models for 2010, global care, diagnostic, and treatment facilities. In regions with investments in cancer care and control might have severely constrained health infrastructure, the effects saved from US$10 million to US$230 million in of the screening program must be carefully consid- disability-adjusted life years (DALYs), or US$531 million ered prior to planning and implementing an organized to almost US$1 trillion in value of statistical life. screening program. Decisions regarding the choice of Further, Knaul, Frenk, and Schulman (2011) high- cancer sites, screening strategies, and target populations light greater cost savings from adopting a prevention/ should be informed not only by cost considerations, but early detection-and-treatment approach versus a also by an understanding of the local burden of disease, treatment-only approach for breast and cervical cancer. sociocultural contexts, health systems, infrastructure, Cancer screening policy may be framed in terms of human resource capacity, community acceptability, and investments, although the timeline to downstream local political will. benefits (such as DALYs saved or citizens remaining Irrespective of the approach to screening, to scale in the workforce longer) will certainly outspan the up organized screening projects, initial plans require political cycle and will depend on how robust and effec- rigorous evaluation as well as knowledge translation and tive the screening program becomes. Cost-effectiveness exchange to all relevant stakeholders, including commu- analysis should also consider the opportunity costs nity agencies and patient advocacy groups. Whether in of not screening, specifically for cancers where early low-, middle-, or high-income settings, key factors for detection and appropriate treatment may significantly community acceptance and success include early and improve survival rates, such as breast, cervical, and high levels of engagement with community and medi- colorectal cancer. cal leaders, education, advocacy, and the establishment of adequate infrastructure and information systems to promote screening and capture initial diagnosis, treat- Ethical Considerations ment, and active follow-up information. Follow-up for In addition to economic considerations, ethical obli- those with a positive (for example, abnormal) screening gations require jurisdictions to ensure that benefits test should include a well-developed care pathway to outweigh harms and that the diagnostic and treatment ensure timely referrals for further evaluation, which may resources are sufficient to justify from the outset the include another imaging modality (for example, breast initiation of a screening program. Recently, some inves- ultrasound), a biopsy, or surgery, as well as a timely tigators have suggested that the informed population’s 214 Cancer preference should also be a factor in such deliberations patterns—fewer children, later first childbirth, and (Harris and others 2011). The informed population may shorter breastfeeding periods. These are factors that raise have a fair say in the design and buy-in for new screening the risk of breast cancer (Corbex, Burton, and Sancho- programs, but countries with established cancer screen- Garnier 2012; Porter 2008). ing policies may find it problematic to separate informed The debate regarding overdiagnosis is of particular preferences from the popular view that earlier detection relevance to breast cancer screening; it is estimated that is invariably better. from 10 to 30 percent of breast cancer detected through This viewpoint is attributable in part to what Gilbert population-based screening mammography may never Welch refers to as the popularity paradox, whereby have resulted in clinically significant disease but triggers the very modest benefits of some forms of screening full-scale treatment. are interpreted by the individuals who have detected A moderate view is that despite some limitations in early-stage disease as having had their disease cured or all screening studies, breast screening mammography survival improved as a function of screening (Welch, has benefits that outweigh harms, and a frank discussion in Raffle and Gray 2007). Few cancer care professionals should take place between health care providers and and few screening policy makers will counter this view their patients, so that each woman can make an informed publicly because there is no simple way other than the decision. In 2012, the National Cancer Institute (U.S.) fullness of time to fully determine whether the disease is convened a task force to address overdiagnosis in cancer “cured.” Nor is there much compassion to be earned for screening. The task force concluded that while screening calling into question patients who optimistically, but in is intended to detect early-stage cancer to improve the many cases mistakenly, believe they have had their dis- likelihood of cure, finding more indolent cancers with ease cured. Counterintuitively, the greater the extent of “better biology” also contributes to better outcomes. overdiagnosis and overtreatment, the greater the number The task force suggested that policies be developed to of screened individuals who believe they owe their lives help mitigate the problems of overdetection and over- to the screening program. While the popularity paradox treatment, “while maintaining those gains by which early has been identified in HICs, this experience may provide detection is a major contributor to decreasing mortality cautionary advice to LMICs that are contemplating and locally advanced disease,” (Esserman, Thompson, establishing screening programs. By contrast, high-risk and Reid 2013, p. 798) and recommend that health care areas in LMICs consist of specific countries, regions, and providers and patients openly discuss the issues, which subpopulations that bear the disproportionate burden of the media should better understand and communicate premature mortality in a range of lethal cancers, includ- to the public. ing liver, stomach, esophagus, and oral cancer. Despite these controversies, breast cancer mortality has been declining in many HICs where mammo- gram screening programs have been in place for over CANCER SCREENING CANDIDATES IN LMICs 20 years (OECD 2011). Many agree that this reflects a combination of newer effective therapies, improved Overall and site-specific cancer mortality rates can breast awareness, and advocacy campaigns, but the be gender-specific. For women in LMICs, breast and relative contribution of each of these factors is diffi- cervical cancer are the leading causes of cancer death, cult to isolate (Kalager, Adami, and Bretthauer 2014). followed by lung, stomach, and liver cancer. For men in Of relevance to LMICs, Kalager, Adami, and Bretthauer’s these geographical areas, lung, liver, stomach, esophagus, commentary on the Canadian National Breast Screening and colon cancer represent the highest mortality burden. Study 25-year follow-up noted that the study lacked The following sections explore the value of screening a “mammogram only” arm, which limited the ability among several of these candidate conditions. to determine the effects of clinical breast examination (CBE) alone versus mammography alone. The authors allude to the potential risk of generalizing to other coun- Breast Cancer tries and suggest that early detection may be of greater Breast cancer, the most common cancer in women benefit in communities where most breast cancers worldwide, is the leading cause of cancer deaths in present clinically with more advanced disease. In regions women in most jurisdictions with reliable data. More where no such advocacy and awareness campaigns than 50 percent of breast cancer deaths occur in LMICs.1 exist, it remains unclear how much early detection for These rates will continue to grow with development breast cancer (or other cancers for which screening is (Bray and others 2012), which has gone hand in hand promoted in HICs) can be achieved by a combination with the Westernization of diets and reproductive of advocacy and awareness campaigns to reduce stigma Screening for Cancer: Considerations for Low- and Middle-Income Countries 215 and overcome cancer myths, and by implementing local programs to teach the value of early detection as lower-cost but potentially effective screening interven- well as risk factors and breast health (Anderson 2013). tions such as CBE. Evaluation goals are included for each resource level The use of mammography for mass screening for for public education and awareness, as well as detection breast cancer requires expensive machinery, with its methods. Recognizing that great differences in health own measurable risk, adequate distribution of radiolo- systems and infrastructure often exist within countries, gists and radiographers, and complex quality controls. most notably from urban centers to rural areas, strat- Moreover, as overall incidence rates remain lower in ification is based on on-the-ground capacities, rather LMICs relative to HICs and the average age of women than a single country-level determination, such as gross with breast cancer is lower than in HICs, the overall domestic product per capita. benefit-to-harm ratio will be correspondingly lower There is an important role for improved breast cancer whether mammography or simpler techniques, such as awareness among the general population in LMICs as well CBE with a skilled trainee, are used. as primary care practitioners; this can be an entry point to A recent systematic review of economic analyses any early detection program. In the absence of evidence of breast cancer control in LMICs concludes that the of the benefits from a systematic assessment of CBE- evidence base for guidance on screening modality based organized screening, we will await the final results (for example, CBE versus mammography), the frequency from the Mumbai trial (Mittra and others 2010) and the of screening, and the target population is limited and of Trivandrum trial of CBE in India (Sankaranarayanan poor quality (Zelle and Baltussen 2013). Anderson and and Bofetta 2010) for any definitive story on CBE as an others in chapter 3 explore in detail the most promising organized screening tool. Notwithstanding the absence of the early detection studies reviewed by Zelle and of definitive experimental evidence for implementing Baltussen and recommend that early detection pro- organized CBE-based screening as a preferred approach grams in LMICs be carefully designed to facilitate early to screen for breast cancer, there is value in trying to phase evaluation. strengthen primary care skill in CBE to improve early Self-screening or breast self-examination in LMICs case-finding and diagnostic activity among symptomatic appears to present greater harms than benefits based on women, since the large majority of breast cancers are one large Asian trial (Thomas and others 2002). A lower diagnosed in women with breast lumps. risk of mortality or advanced breast cancer was found in a meta-analysis of breast self-examination only in stud- ies of women with breast cancer who reported practicing Cervical Cancer breast self-examination before diagnosis (Hackshaw and Cervical screening may have the greatest potential for Paul 2003); no difference was found in the death rate in screening-detected reductions in cancer mortality in studies on women who detected their cancer during an less developed regions, where about 85 percent of all examination. Despite conflicting evidence for CBE in new cases and 87 percent of deaths from cervical cancer some low-income and lower-middle-income country occur (Ferlay and others 2013). The incidence of cervical settings (Nguyen and others 2013; Pisani and others cancer is highly correlated with country income group, 2006), Anderson and others in chapter 3 of this volume the prevalence of high-risk subtypes of the causal agent note that a case remains to be made for CBE as a means HPV (particularly HPV 16 and 18, which account for of stage shifting, especially in populations where the approximately 70 percent of the case burden), and average tumor size at presentation is considerably larger whether the country or region of interest has had than that in most of the breast screening studies to date. a longstanding population-based screening program Reasonable evidence suggests that formal training in (see chapter 4 in this volume). In terms of DALYs, which CBE for primary care professionals can improve the depend also on the average age at which individuals are sensitivity of the procedure and reduce the number of affected, cervical cancer ranks highest by, and is cor- false positives (Vetto and others 2002). related with, a lower human development index, a com- The Breast Health Global Initiative has developed an posite measure that includes life expectancy, education, evidence-based, resource-stratified approach to early and income (Soerjomataram and others 2012). detection and screening, as well as diagnosis, treatment, Despite the efficacy of cytology-based mass screening and most recently, supportive care and quality of life programs, Papanicolaou, or Pap, testing is costly, com- (Anderson 2013). Recommendations for resource allo- plex, and requires robust health systems. Chapter 4 in cation include not only the screening modalities such this volume notes the poor penetration of widespread as CBE, mammography, and diagnostic ultrasound, Pap testing owing to such costs. The unequal burden of but culturally-sensitive and linguistically-appropriate mortality as a consequence reflects unequal access in less 216 Cancer developed countries. Newer and less expensive strategies of adult women, even only three times per lifetime, can to prevent cervical cancer have been evaluated and the avert a significant proportion of cervical cancer cases introduction of new HPV vaccines offers real prevention in a cost-effective manner. In addition to many other prospects for the first time. critical inputs to health decisions, such as political will VIA in combination with cryotherapy (screen- and cultural acceptability, evidence on the cost effective- and-treat) was trialed in a demonstration project in ness and affordability of HPV vaccination and screening Ghana and was well accepted by the communities involved from rigorous model-based analyses can help to inform (Blumenthal and others 2007). This effort underlines decision makers and stakeholders in their deliberations the value of simple and effective technologies for low- of how best to prevent cervical cancer worldwide. resource settings despite inadequate coverage and sig- nificant numbers lost to follow-up. A one-time screening at 35 years of age with VIA or HPV testing reduced the Colorectal Cancer lifetime risk of cervical cancer by approximately 25 to Lambert, Sauvaget, and Sankaranarayanan (2009) 36 percent and cost less than US$500 per year of life saved advance a strong argument that the burden of col- (Goldie and others 2005). orectal cancer, while high and growing in HIC regions Two exciting trials reporting on test-and-treat models (about 12 percent of deaths from cancer), remains low in India (Sankaranarayanan and others 2009) and South on the list of common cancers and primary causes of Africa (Denny and others 2010) have highlighted the cancer-related mortality in less developed regions (about superiority of a screen-and-treat approach that uses 6 percent of deaths from cancer). Lambert, Sauvaget, relatively more expensive HPV testing over VIA, whether and Sankaranarayanan argue that the expense of mount- followed by colposcopy in the Indian trial or cryotherapy ing a mass screening effort in most LMICs is not cur- in the South African trial. The Indian trial showed that rently justified, given the significant costs of colonoscopy a single round of HPV testing can reduce the incidence and follow-up services. The authors do allow that the of advanced cancers and deaths from cervical cancer. growth of more Western lifestyles in large urban centers The South African study showed benefits in the VIA in upper-middle-income countries may represent areas group, but HPV DNA testing most effectively reduced where colon screening may be more justifiable. the incidence of advanced invasive cancer that devel- By contrast, as noted in chapter 16 of this volume, oped more than 12 months after cryotherapy. HPV at least one report suggests that screening colonoscopy DNA testing, with or without VIA, shows the greatest may be cost effective in Sub-Saharan Africa (Ginsberg promise; however, given the current state of pathology and others 2012), at least in the urban areas of infrastructure and cost considerations for less developed upper-middle-income countries, where the incidence regions and, in particular, for rural populations in LICs of colorectal cancer is increasing because of population and lower-middle-income countries, the introduction aging and the adoption of Western lifestyles. of mass screening with VIA may be the most prudent The International Colorectal Cancer Screening real-world approach. Network (2013), which works to document and stan- In addition, several combination modes of preventive dardize the best jurisdictional approaches to colorectal HPV vaccination in preadolescent girls, combined with screening, identifies the need for screening program various screening measures in adult women, appear experience on every continent, although membership is promising as a comprehensive method to reduce the currently limited to more developed regions. Research burden of cervical cancer and reduce HPV infection. In in progress may offer a range of promising and less Sub-Saharan Africa in particular, a strong case exists for invasive methods to detect early-stage colon cancer, screening with VIA and rapid HPV tests to ramp up pre- which may offer better options to reduce colon cancer vention and detection services to screen, treat, or refer. mortality in LMICs. This approach would allow for the opportunity to deal A phased introduction of colorectal cancer screen- with any other gynecological issues. Populations with ing by immune sensitive fecal occult blood testing in coterminous HPV and HIV infections are at highest risk Thailand, beginning with a pilot evaluation in Lampang and have the highest need for cervical cancer prevention province, shows promise for reducing colon cancer focus (Sahasrabuddhe and others 2012). mortality. The program is based on a five-year interval With respect to the cost effectiveness of cervical for immune fecal testing, which is supportable by the screening programs, recent analyses demonstrate that health system infrastructure and appropriate, given the there are promising opportunities to prevent cervical relatively lower colon cancer rates compared with other cancer in different world settings. As stated in chapter 4, countries with screening programs (Khuhaprema and HPV vaccination for preadolescent girls and/or screening others 2014). Screening for Cancer: Considerations for Low- and Middle-Income Countries 217 Promising New Candidates The most promising evidence in LMICs suggests that Simple visual screening methods in high-risk areas the highest priorities are breast and cervical cancer for oral cancer in South Asia and Southeast Asia rep- screening, possibly followed by colorectal cancer or resent an excellent example of pragmatic screening stomach cancer screening, if programmatic infra- (Sankaranarayanan and others 2005). These cancers structure can be established in a stepwise fashion. are highly linked to tobacco and alcohol consumption, Carefully planned programs for breast screening, as well as to chewing betel and areca nut and paan and according to local context and resource capacity as gutka (see chapter 5). Increasing evidence suggests that highlighted in the Breast Health Global Initiative HPV is a risk factor in oral, head, and neck cancers. documents, and VIA with cryosurgery or colposcopy Most cost-effectiveness studies come from HICs, but (with or without HPV testing, where available) can one very promising study from India suggests that oral appropriately be recommended as first cancer screen- cancer screening by visual inspection has an incre- ing priorities in LMICs. mental cost-effectiveness ratio of US$835 per life year saved (Subramanian and others 2009). Further, the Role of Innovation authors note that the most prudent approach for lim- ited resource settings should include only higher risk Many opportunities already exist to exploit the poten- populations, such as heavy users of tobacco and alcohol. tial impact of programs for early detection and There is now some trial evidence that visual screening screening. Considering a given screening strategy for can reduce oral cancer mortality in users of tobacco and which locally-sourced evidence demonstrates at least alcohol (Sankaranarayanan and others 2013). proof of concept in terms of efficacy and cost effec- By contrast, in HICs, a recent assessment from the tiveness, transition-to-scale projects can take advan- U.S. Preventive Services Task Force has concluded that tage of a variety of innovative approaches to optimize the case for mass screening for oral cancers in the rela- participation, follow-up for an abnormal screening, tively lower risk United States is insufficient to justify the as well as monitoring for treatment-related toxicities harms of mass screening of asymptomatic adults (Moyer and survivorship care. These approaches include and U.S. Preventive Services Task Force 2014). telemedicine; telepathology; institutional twinning; Gastric cancer is a close tie with liver cancer as the task-shifting; and “m-Health” (WHO 2011a, 2011b), second leading cancer-related cause of death and is a models of care enhanced by the use of mobile phones, particular challenge in the East Asia and Pacific region. which are widely available and affordable in most Promising programs are being mounted in Japan and LMICs (Ginsburg 2013). Large technical platforms the Republic of Korea and in trials in China to screen can give way to cloud applications, which allow for for the bacterium Helicobacter pylori, the cause of a large easy and secure storage and compilation of infor- fraction of gastric cancer, and to eradicate infections mation for screening programs. However, this still detected. However, H. pylori eradication, which reduces requires a basic information and communications gastric cancer risk, is hampered by emerging regional technology infrastructure, computer availability, and antimicrobial resistance to antibiotics used to treat it up-to-date software, which are missing in many and the lack of a means to target a high-risk population. countries. Gastric cancer remains a screening and prevention can- Similarly, not all screening activity needs to involve didate in need of more refined trials (Park and others only primary care physicians or specialty care providers, 2013). if reliable evidence is used to build from project-to-scale programs. In this fashion, trained community care workers, nurses, and other care providers can assist in SPECIAL CONSIDERATIONS FOR CANCER building capacity, promoting screening activities, and SCREENING IN LMICs being effective screening agents in LMICs. Screening Priorities in LMICs Promising cancer screening candidates in each of the Diagonal Approach to Strengthen Health Systems less developed regions warrant attention. The joint From a programmatic perspective, the breast and report of the World Economic Forum/Harvard School cervical cancer studies also indicate some merit in an of Public Health and the World Health Organization integrated approach to screening under the umbrella of highlights a set of affordable, feasible, and cost-effective maternal or reproductive health policy, as suggested in intervention strategies to reduce the economic impact the trial in Mumbai (Mittra and others 2010) and the of noncommunicable diseases in LMICs (WEF 2011). approach taken in Morocco.2 The Global Task Force on 218 Cancer Expanded Access to Cancer Care and Control (2011) and quality assurance processes. The heterogeneity of has championed this diagonal approach, “the proactive, cancer screening policies across LMICs reflects differ- supply-driven provision of a set of highly cost-effective ences in health care structures as well as the political interventions on a large scale that bridges health clinics and cultural factors that shape policy. The governance and homes” (Sepúlveda 2006). While the age intervals mechanisms for the development of screening pol- best chosen for a first screening intervention may not icy may also vary. Some countries use a legislative be an exact match, for women undergoing simultane- approach to mandate screening, thus opening policy ous screening efforts, there is at least the prospect of development to political influence. In other coun- having both screening procedures performed during tries, policy development is delegated to technocrats the same visit, a predictor of better participation than who rely on advice from expert committees or ad multiple visits, an observation now being mimicked hoc groups. Screening policy in LMICs often involves in HICs. “Pink Ribbon Red Ribbon” (UNAIDS 2011) several layers of organization, including transnational is an example of a program that added breast and actors (for example, the United Nations Population cervical cancer screening to an existing program for Fund), national health ministries, and experts in var- another health condition, namely, HIV. HIV-positive ious disciplines, as well as prominent domestic and women have a greater chance of developing invasive global advocacy groups. The respective roles of author- cervical cancer and higher mortality rates than their ities is another source of variation, as is the degree of HIV-negative counterparts. This type of program can reliance on or participation in the development of pol- address the needs of a group at particularly high risk icy at the transnational level (for example, guidelines) with a low marginal cost. and how this is shaped by the institutional processes of Such a diagonal approach is not limited to wom- decision making (Flintcroft 2011). en’s health services. Integrating cancer screening into Cancer screening policies sometimes run counter to existing health programs can also help to build pri- what would seem to reflect the best evidence (Nutley, mary care capacity. Harnessing the synergies between Walter, and Davies 2007). Knowing the diversity of traditionally vertical programs can build platforms factors (political, social, and economic) that contribute onto which additional preventive and wellness care to the development of the health policies reinforces the (such as vaccinations, smoking cessation, and nutri- understanding that the way that research influences tional counseling) may be added to reduce the inci- policy is not linear and not necessarily determined dence and mortality from other cancers (such as by the quality of the research alone (Humpreys and lung, stomach, or oral), as well as other high-burden Piot 2012; Nightingale and Scott 2007). In fact, many chronic noncommunicable diseases. Modeling such accepted models of public health policy making for programs can also help to convince policy makers that cancer screening (research utilization) have likened cancer screening and cancer control in general will not the process to a complex dance (Edwards 2001), necessarily siphon off scarce resources from compet- a garbage can of ideas waiting to be needed (Cohen, ing health priorities. March, and Olsen 1972), and parallel streams awaiting a social, political, or economic reason to stimulate a convergence and therefore the formation of policy Policy Considerations for Cancer Screening in LMICs (Kingdon 2003). The policy process can become more Cancer screening policies and the programs they create transparent and outcomes more predictable (although become part of established health care systems in gov- still respecting national differences) with standardized ernments and societies, each with its own norms and decision making systems that encompass the principles standards. Although the evidence base may be global of health technology assessment for all policy decisions for any particular cancer, policies and programs vary that involve weighing evidence on effectiveness, costs, considerably across the globe, not just between coun- and societal factors. tries according to level of wealth, but among countries Ultimately, the success of a cancer screening policy of similar economic status. Some differences mirror the and its associated program depends not only on the evi- huge variation in the incidence of different cancers, but dence base, but also on the willingness of the public to many are caused by differential weighting of evidence take part in the screening process. This, in turn depends and other factors. to a great extent on how the benefits and risks of the Developing a screening policy for cancer involves procedure are communicated (McCormack and others many decisions, including choice of diagnostic 2011) and how the program fits within the health care technologies and follow-up interventions, the age system and with other health messages, including cancer groups targeted, referral and enrollment strategies, prevention. Screening for Cancer: Considerations for Low- and Middle-Income Countries 219 ACKNOWLEDGMENT edition): Volume 3, Cancer, edited by H. Gelband, P. Jha, R. Sankaranarayanan, and S. Horton. Washington, DC: The authors express special thanks to Signy Roland World Bank. for her support in the compilation and extraction of Denny, L., L. Kuhn, C. C. Hu, W. Y. Tsai, and T. C. Wright Jr. the studies discussed in this chapter and to Sandra 2010. “Human Papillomavirus-Based Cervical Cancer Hofstetter for her support in compiling the manuscript. Prevention: Long-Term Results of a Randomized Screening Trial.” Journal of the National Cancer Institute 102 (20): 1–11. Edwards, M. 2001. Social Policy, Public Policy: From Problems NOTES to Practice. St. Leonards, NSW: Allen & Unwin. Errico, K. M., and D. Rowden. 2006. “Experiences of Breast World Bank income classifications as of July 2014 are as Cancer Survivor-Advocates and Advocates in Countries follows, based on estimates of gross national income per capita with Limited Resources: A Shared Journey in Breast Cancer for 2013: Advocacy.” The Breast Journal 12 (Suppl. 1): 111–16. Esserman, L., I. Thompson Jr., and B. Reid. 2013. “Overdiagnosis • Low-income countries: US$1,045 or less and Overtreatment in Cancer: An Opportunity for • Middle-income countries: Improvement.” Journal of the American Medical Association • Lower-middle-income: US$1,046–US$4,125 310 (8): 797–98. doi:10.1001/jama.2013.108415. • Upper-middle-income: US$4,126–US$12,745 Family Practice Notebook. 2011. Length Bias. http://www • High-income countries: US$12,746 or more .fpnotebook.com/Prevent/Epi/LngthBs.htm. Ferlay J, Soerjomataram I, Ervik M, Dikshit R, Eser S, Mathers C, 1. http://www.who.int/cancer/detection/breastcancer/en Rebelo M, Parkin DM, Forman D, Bray, F. GLOBOCAN 2012 /index1.html. v1.0, Cancer Incidence and Mortality Worldwide: IARC 2. “Importance of the Screening,” Fondation Lalla Salma CancerBase No. 11 [Internet]. Lyon, France: International Prévention et Traitement des Cancers, accessed March 8, Agency for Research on Cancer; 2013. Available from: http:// 2014, http://www.contrelecancer.ma/en/importance_du globocan.iarc.fr, accessed on 23/11/2014. _depistage. Flintcroft, T. 2011. “Getting Evidence into Policy: The Need for Deliberative Strategies?” Social Science & Medicine 72: 1039–46. Ginsberg, G. M., J. A. Lauer, S. Zelle, S. Baeten, and R. Baltussen. REFERENCES 2012. “Cost Effectiveness of Strategies to Combat Breast, Anderson, B. 2013. “Context-Relevant Guidelines in Cancer Cervical, and Colorectal Cancer in Sub-Saharan Africa and Care: Breast Cancer Early Detection, Diagnosis, and South East Asia: Mathematical Modeling Study.” British Treatment in Low and Middle Income Countries.” In Medical Journal 344: e614. Cancer Control 2013, edited by Ian McGrath, 36–40. http:// Ginsburg, O. M. 2013. “Breast and Cervical Cancer Control globalhealthdynamics.co.uk/cc2013/wp-content /uploads in Low and Middle-Income Countries: Human Rights /2013/04/36-40-Benjamin-O-Anderson_2013.pdf. Meet Sound Health Policy.” Journal of Cancer Policy 1 (3): Blumenthal, P. D., L. Gaffikin, S. Deganus, R. Lewis, M. Emerson, e35–e41. and others. 2007. “Cervical Cancer Prevention: Safety, Global Task Force on Expanded Access to Cancer Care and Acceptability, and Feasibility of a Single-Visit Approach in Control in Developing Countries. 2011. Closing the Divide: Accra, Ghana.” American Journal of Obstetrics and Gynecology A Blueprint to Expand Access in Low- and Middle-Income 196 (4): 407.e1–9. Countries. Boston, MA: Harvard Global Equity Initiative. Bray, F., A. Jemal, N. Grey, J. Ferlay, and F. Forman. 2012. “Global http://ghsm.hms.harvard.edu/files/assets/Programs Cancer Transitions According to the Human Development /NewbornHealth/files/ccd_report_bw.pdf. Index (2008–2030): A Population-Based Study.” The Lancet Goel, V. 2001. “Appraising Organized Screen Programmes for 13: 790–801. Testing for Genetic Susceptibility to Cancer.” British Medical Carey, L. A., C. M. Perou, C. A. Livasy, L. G. Dressler, D. Cowan, Journal 322 (7295): 1174–78. and others. 2006. “Race, Breast Cancer Subtypes, and Goldie, S., G. Gaffikin, J. Goldhaber-Fiebert, A. Gordillo-Tobar, Survival in the Carolina Breast Cancer Study.” Journal of the C. Levin, and others. 2005. “Cost-Effectiveness of Cervical- American Medical Association 295 (21): 2492–502. Cancer Screening in Five Developing Countries.” New Cohen, M. D., J. G. March, and J. P. Olsen. 1972. “A Garbage England Journal of Medicine 353: 2158–68. Can Model of Organizational Choice.” Administrative Hackshaw, A. K., and E. A. Paul. 2003. “Breast Self-Examination Science Quarterly 17: 1–25. and Death from Breast Cancer: A Meta-Analysis.” British Corbex, M., R. Burton, and H. Sancho-Garnier. 2012. “Breast Journal of Cancer 88: 1047–53. Cancer Early Detection Methods for Low and Middle Harris, R., G. F. Sawaya, V. A. Moyer, and N. Calonge. 2011. Income Countries.” The Breast 21 (4): 428–34. “Reconsidering the Criteria for Evaluating Proposed Denny, L., R. Herrero, C. Levin, and D. B. Thomas. 2015. Screening Programs: Reflections from 4 Current and “Cervical Cancer.” In Disease Control Priorities (third Former Members of the U.S. Preventive Services Task 220 Cancer Force.” Epidemiologic Reviews 33 (1): 20–35. doi:10.1093 Nutley, S. M., I. Walter, and H. T. O. Davies. 2007. Using /epirev/mxr005. Evidence: How Research Can Inform Public Services. Bristol, Humpreys, K., and P. Piot. 2012. “Scientific Evidence Alone UK: The Policy Press. Is Not Sufficient Basis for Health Policy.” British Medical OECD (Organisation for Economic Co-operation and Journal 344: e1316. Development). 2011. Screening, Survival and Mortality IARC (International Agency for Research on Cancer). 2005. IARC for Breast Cancer. http://www.oecd-ilibrary.org/docserver Handbooks of Cancer Prevention: Cervix Cancer Screening. /download/8111101ec048.pdf?expires=1394159448&id=id Lyon, France: IARC. http://www.iarc.fr/en/publications &accname=guest&checksum=2FC8D19B124E6033BD7AF- /pdfs-online/prev/handbook10/HANDBOOK10.pdf. 114C1A8E2CB. International Colorectal Cancer Screening Network. Park, J. Y., D. Forman, E. R. Greenberg, and R. Herrero. 2013. 2013. “ICRCSN Publications.” http://icrcsn.ceu.ox.ac.uk “Helicobacter Pylori Eradication in the Prevention of /publications. Gastric Cancer: Are More Trials Needed?” Current Oncology Kalager, M., H. Adami, and M. Bretthauer. 2014. “Too Much Reports 15 (6): 517–25. Mammography.” British Medical Journal 348: g1403. Pisani, P., D. M. Parkin, C. Ngelangel, D. Esteban, L. Gibson, doi:http://dx.doi.org/10.1136/bmj.g1403. and others. 2006. “Outcome of Screening by Clinical Khuhaprema, T., S. Sangrajrang, S. Lalitwongsa, Examination of the Breast in a Trial in the Philippines.” V. Chokvanitphong, T. Raunroadroong, and others. 2014. International Journal of Cancer 118 (1): 149–54. “Organised Colorectal Cancer Screening in Lampang Porter, P. 2008. “‘Westernizing’ Women’s Risks? Breast Cancer Province, Thailand: Preliminary Results from a Pilot in Lower-Income Countries.” New England Journal of Implementation Programme.” British Medical Journal 4: Medicine 358: 213–16. doi:10.1056/NEJMp0708307. e003671. doi:10.1136/bmjopen-2013-003671. Price, A. J., P. Ndom, E. Atenguena, J. P. Mambou Nouemssi, Kingdon, J. 2003. Agendas, Alternatives and Public Policies. 2nd and R. W. Ryder. 2012. “Cancer Care Challenges in ed. New York: Longman. Developing Countries.” Cancer 118 (14): 3627–35. Knaul, F., J. Frenk, and L. Shulman. 2011. Closing the doi:10.1002/cncr.26681. Cancer Divide: A Blueprint to Expand Access in Low and Raffle, A., and M. Gray. 2007. Screening: Evidence and Practice. Middle Income Countries. Boston, MA: Harvard Global Oxford, UK: Oxford University Press. Equity Initiative. http://gtfccc.harvard.edu/icb/icb.do Sahasrabuddhe, V. V., G. P. Parham, M. H. Mwanahamuntu, ?keyword=k69586&pageid=icb.page334798. and S. H. Vermund. 2012. “Cervical Cancer Prevention in Lambert, R., C. Sauvaget, and R. Sankaranarayanan. 2009. Low- and Middle-Income Countries: Feasible, Affordable, “Mass Screening for Colorectal Cancer Is Not Justified Essential.” Cancer Prevention and Research 5 (1): 11–17. in Most Developing Countries.” International Journal of Sankaranarayanan, R., and P. Bofetta. 2010. “Research on Cancer 125 (2): 253–56. Cancer Prevention, Detection and Management in Low McCormack, L. A., K. Treiman, D. Rupert, P. Williams-Piehota, and Middle Income Countries.” Annals of Oncology E. Nadler, and others. 2011. “Measuring Patient-Centered 21: 1935–43. Communication in Cancer Care: A Literature Review and Sankaranarayanan, R., B. M. Nene, S. S. Shastri, K. Jayant, the Development of a Systematic Approach.” Social Science R. Muwonge, and others. 2009. “HPV Screening for Cervical & Medicine 72 (7): 1085–95. Cancer in Rural India.” New England Journal of Medicine Mittra, I., G. A. Mishra, S. Singh, S. Aranke, P. Notani, and 360: 1385–94. others. 2010. “A Cluster Randomized, Controlled Trial of Sankaranarayanan, R., K. Ramadas, S. Thara, R. Muwonge, Breast and Cervix Cancer Screening in Mumbai, India: G. Thomas, and others. 2013. “Long Term Effect of Visual Methodology and Interim Results after Three Rounds of Screening on Oral Cancer Incidence and Mortality in a Screening.” International Journal of Cancer 126 (4): 976–84. Randomized Trial in Kerala, India.” Oral Oncology 49 (4): doi:10.1002/ijc.24840. 314–21. Morrison, A. S. 1992. Screening in Chronic Disease. New York: Sankaranarayanan, R., K. Ramadas, G. Thomas, R. Muwonge, Oxford University Press. S. Thara, and others. 2005. “Effect of Screening on Oral Moyer, V. A., on behalf of the U.S. Preventive Services Task Cancer Mortality in Kerala, India: A Cluster-Randomised Force. 2014. “Screening for Oral Cancer: U.S. Preventive Controlled Trial.” The Lancet 365 (9475): 1927–33. Services Task Force Recommendation Statement.” Schnipper, L. E., G. H. Lyman, D. W. Blayney, J. R. Hoverman, Annals of Internal Medicine 160 (1): 55–60. doi:10.7326/ D. Raghavan, and others. 2013. “American Society of M13-2568. Clinical Oncology 2013 Top Five List in Oncology.” Journal Nguyen, L. H., W. Laohasiriwong, J. F. Stewart, P. Wright, Y. T. B. of Clinical Oncology 31 (34): 4362–70. Nguyen, and P. C. Coyte. 2013. “Cost-Effectiveness Analysis Sepúlveda, J. 2006. “Foreword.” In Disease Control Priorities of a Screening Program for Breast Cancer in Vietnam.” in Developing Countries, 2nd ed., edited by D. T. Jamison, Value in Health Regional Issues 2 (1): 21–28. J. G. Breman, A. R. Measham, G. Alleyne, M. Claeson, Nightingale, P., and A. Scott. 2007. “Peer Review and the D. B. Evans, P. Jha, A. Mills, and P. Musgrove, xii–xv. Relevance Gap: Ten Suggestions for Policymakers.” Science Washington, DC: World Bank and Oxford University and Public Policy 34: 543–553. Press. Screening for Cancer: Considerations for Low- and Middle-Income Countries 221 Soerjomataram, I., J. Lortet-Tieulent, D. M. Parkin, Training Results in Objective Improvement in CBE Skills.” J. Ferlay, C. Mathers, and others. 2012. “Global Burden Journal of Cancer Education 17 (3): 124–27. of Cancer in 2008: A Systematic Analysis of Disability- Vorobiof, D. A., F. Sitas, and G. Vorobiof. 2001. “Breast Cancer Adjusted Life Years in 12 World Regions.” The Lancet Incidence in South Africa.” Journal of Clinical Oncology 19 380 (9856): 125. (Suppl. 18): 125–27. Story, H. L., R. R. Love, R. R. Salim, A. J. Roberto, WEF (World Economic Forum). 2011. From Burden to Best J. L. Krieger, and others. 2012. “Improving Outcomes Buys: Reducing the Economic Impact of Non-Communicable from Breast Cancer in a Low-Income Country: Lessons Disease in Low and Middle-Income Countries. Geneva, from Bangladesh.” International Journal of Breast Cancer. Switzerland: WEF. http://www.who.int/nmh/publications doi:10.1155/2012/423562. /best_buys_summary.pdf. Subramanian, S., R. Sankaranarayanan, B. Bapat, T. Somanathan, Welch, G., L. Schwartz, and S. Woloshin. 2011. Overdiagnosed: G. Thomas, and others. 2009. “Cost-Effectiveness of Oral Making People Sick in the Pursuit of Health. Boston, MA: Cancer Screening: Results from a Cluster Randomized Beacon Press. Controlled Trial in India.” Bulletin of the World Health WHO (World Health Organization). 2011a. mHealth: New Organization 87 (3): 200–06. Horizons for Health through Mobile Technologies: Based on Thomas, D. B., D. L. Gao, R. M. Ray, W. W. Wang, C. J. Allison, the Findings of the Second Global Survey on eHealth (Global and others. 2002. “Randomized Trial of Breast Self- Observatory for eHealth Series, Volume 3). http://whqlibdoc Examination in Shanghai: Final Results.” Journal of the .who.int/publications/2011/9789241564250_eng.pdf. National Cancer Institute 94 (19): 1445–57. ———. 2011b. E-health in Low- and Middle-Income UNAIDS (Joint United Nations Programme on Countries: Findings from the Center for Health Market HIV/AIDS). 2011. The George W. Bush Institute, Innovations. http://www.who.int/bulletin/volumes/90/5 the U.S. Department of State, Susan G. Komen for /11-099820/en/. the Cure®, and UNAIDS Announce New Women’s Wilson, J. M. G., and G. Junger. 1968. “Principles and Practice Health Initiative. http://www.unaids.org /en /resources of Screening for Disease.” Public Health Papers 34, WHO, /presscentre / pressreleaseandstatementarchive /2011 Geneva. /september/20110913pinkribbonredribbon/. Yaffe, M. J., and K. I. Pritchard. 2014. “Overdiagnosing Van de Vijver, M. J., Y. D. He, L. J. van’t Veer, H. Dai, Overdiagnosis.” The Oncologist 19: 103–06. doi:10.1634 A. A. M. Hart, and others. 2002. “A Gene-Expression /theoncologist.2014-0036. Signature as a Predictor of Survival in Breast Cancer.” New Zelle, S., and R. Baltussen. 2013. “Economic Analyses of England Journal of Medicine 347: 1999–2009. Breast Cancer Control in Low- and Middle-Income Vetto, J. T., J. K. Petty, N. Dunn, N. C. Prouser, and D. F. Austin. Countries: A Systematic Review.” Systematic Reviews 2: 20. 2002. “Structured Clinical Breast Examination (CBE) doi:10.1186/2046-4053-2-20. 222 Cancer Chapter 13 Surgical Services for Cancer Care Anna J. Dare, Benjamin O. Anderson, Richard Sullivan, C. S. Pramesh, Cheng-Har Yip, Andre Ilbawi, Isaac F. Adewole, Rajendra A. Badwe, and Cindy L. Gauvreau INTRODUCTION of the population are unable to access even the most basic surgical care (Funk and others 2010). Surgery is a fundamental modality for curative and The projected increase in the cancer burden in LMICs palliative treatment of most cancers in countries across over the next 20 years (see chapter 2 in this volume) all income settings. In high-income countries (HICs), necessitates that all countries give consideration to the where the most common solid organ malignant cancers, establishment of surgical services with adequate capacity such as breast and colon cancers, are more likely to be to meet current and future needs. In general, significant successfully diagnosed at early stages, surgical resec- capital investment in surgical infrastructure, equipment, tion provides definitive locoregional control of the and personnel is needed in LICs, especially those in primary tumor. This approach has significant curative Sub-Saharan Africa (LeBrun and others 2014). In MICs, potential when combined with appropriately selected improved coordination, regulation, financial risk pro- adjuvant systemic treatment and radiotherapy. In low- tection, and strategic planning for cancer and surgical and middle-income countries (LMICs), where locally services are requisites to improve service delivery and advanced or metastatic cancer is a common initial dis- outcomes (Goss and others 2014). Surgical capacity ease presentation, surgical resection or debulking may be building takes time, particularly with respect to develop- one of the few available modalities to achieve reasonable ing the surgical workforce. Efforts to strengthen surgical palliative disease control. services in LMICs should be strategically proactive to Surgery has not received sufficient attention in the facilitate the provision of safe, effective, and accessible cancer control discussion in LMICs (Goss and oth- surgical cancer care for current and future patients. ers 2014; Purushotham, Lewison, and Sullivan 2012). This chapter discusses the public sector delivery of With many competing health priorities and significant surgical cancer services in resource-constrained environ- financial constraints, surgical services in these settings ments. We describe the current status of surgical services are given low priority within national health plans and for cancer care in LMICs, analyze the barriers to care, and are allocated few resources from domestic accounts or outline the surgical delivery platforms available to coun- international development assistance programs (Bae, tries at different resource and income levels. Key consider- Groen, and Kushner 2011; Farmer and Kim 2008). As a ations for policy makers relating to quality, safety, access, result, in most low-income countries (LICs), and many coverage, and economic and planning considerations in middle-income countries (MICs), access to safe, optimal the scale-up of surgical cancer services are highlighted. surgical services for cancer is poor, and large proportions Corresponding author: Anna J. Dare, MBChB, PhD, St. Michael’s Hospital and University of Toronto, Canada, DareA@smh.ca 223 BURDEN OF SURGICALLY TREATABLE for malignant bowel obstruction and fungating breast CANCERS IN LMICs cancers. To realize the therapeutic benefit of surgical care in achieving cancer cure, stage-shifting is required As many LMICs transition to higher levels of social and to address the disease burden before it becomes locally economic development, with attendant greater popula- advanced or metastatic—an objective particularly valid tion growth and improved longevity, the cancer burden in LMICs, where more than 70 percent of patients pre- amenable to surgical treatment is projected to increase sent with advanced cancer (Adebamowo and Ajayi 2000; dramatically (figure 13.1). Almost all of the common Anyanwu 2000, 2008). cancers require surgical services for histological diagno- sis if radiology-guided biopsy is not available, for resec- tion as the mainstay of curative treatment, and selectively STATUS OF SURGICAL CANCER SERVICES for palliation. IN LMICs Surgery is more effective, less complex, and less costly when performed for early-stage or locally advanced The availability of, access to, and quality of surgical cancer. Curative treatment can often be delivered within cancer care varies widely, leading to equally wide vari- a single clinical encounter and is achievable even in ations in outcomes among and within countries. Most low-resource settings. Although surgery has less of a role LICs face profound shortages of surgeons, anesthesi- in advanced stage cancer, in select cases it can provide ologists, and pathologists; inadequate equipment and improved quality and prolongation of life, for example, supplies; absent or severely dilapidated general and Figure 13.1 Estimated Number of New Cancer Cases by Country-Income Group in LMICs for Four Common Surgically Treatable Cancers, 2010 and 2030 a. Breast cancer b. Cervical cancer 800,000 400,000 700,000 350,000 600,000 300,000 Incident cases Incident cases 500,000 250,000 400,000 200,000 300,000 150,000 200,000 100,000 100,000 50,000 0 0 Low- Lower-middle- Upper-middle- High- Low- Lower-middle- Upper-middle- High- income income income income income income income income c. Colorectal cancer d. Oral cancer 1,000,000 200,000 900,000 180,000 800,000 160,000 Incident cases 700,000 Incident cases 140,000 600,000 120,000 500,000 100,000 400,000 80,000 300,000 60,000 200,000 40,000 100,000 20,000 0 0 Low- Lower-middle- Upper-middle- High- Low- Lower-middle- Upper-middle- High- income income income income income income income income 2010 2030 Source: Ferlay, Soerjomataram, and others 2013. Note: All cases of breast, cervical, colorectal, and oral cancer require the input of a surgeon or gynecologist for diagnosis and clinical management. LMICs = low- and middle-income countries. 224 Cancer surgical infrastructure; and a lack of financing and stra- situational analysis of health facilities in 24 LICs and tegic health services planning. The result is that a large 27 MICs, the average patient had to travel 100 kilometers proportion of the population is without access to even to reach a facility that could perform a basic biopsy the most basic surgical services (Funk and others 2010; diagnostic procedure (Ilbawi, Cherian, Mikkelson, Weiser and others 2008). In most MICs, surgical services Sankaranarayanan, and Sullivan, unpublished data). for cancer are more widely available, especially in major Even where surgical cancer services are available and cities, but variations in quality, inequitable coverage and accessible, prohibitive user costs and perceptions that utilization, and poor central regulation and coordina- cancer cannot be successfully treated may prevent people tion (Pramesh and others 2014; Yip and others 2011) from obtaining timely treatment (Ilbawi, Einterz, and hamper the effective provision of care. Nkusu 2013). Sociocultural belief systems and practices can affect cancer awareness and the uptake of surgical cancer Access, Distribution, and Utilization services in LMICs (Goss and others 2014). Barriers to Access timely uptake of services include fear of surgery and hos- Cancer surgery typically requires more complex infra- pital services in general, “cancer fatalism,” cultural beliefs structure, training, support services, and referral net- and social stigma related to being cut or having a body works than many basic surgical procedures. Gross part removed, poor community experiences relating to inequities between HICs and LMICs exist in access to outcomes, and costs (Daher 2012; Goss and others 2014; surgical services (Funk and others 2010; Weiser and Yip and Anderson 2007). others 2008) (table 13.1). An estimated two billion peo- ple lack access to any form of surgical care, including surgery for cancer (Funk and others 2010). Regional Quality, Safety, and Outcomes and national estimates of surgical services for cancer Wide variations exist globally in the quality and safety are not available, but access and coverage are likely to be of surgical care (see Debas and others 2015, chapter 16). significantly worse than for surgical services in general. Quality issues are particularly concerning in the context Globally, approximately seven million to eight million of surgical treatment for cancer, where achieving adequate patients require a major cancer operation each year, and resection is fundamental to the success of the procedure. at least three million additional patients require biopsies each year (Ferlay, Steliarova-Foucher, and others 2013). Infrastructure and Training In LMICs, quality and safety issues are often closely Distribution linked to basic resource deficits relating to infrastructure, The distribution of surgical services within coun- tries is also uneven in LMICs, compounding issues of access and coverage (Goss and others 2014). The Table 13.1 Disparities in Surgical Capacity between High-Income surgical workforce and surgical facilities tend to clus- and Low-Income Countries ter in urban areas (Ozgediz and others 2008). Surgical Measure High-income countries Low-income countries services for cancer, if present, typically are located at third-level facilities, often in the capital city, with poor Number of surgeons 34–97 per 100,000a 0.13–1.57 per 100,000 (Hoyler and others 2014) or nonexistent referral networks. In settings in which a large proportion of the population lives in rural areas, Number of 34–97 per 100,000a 0–4.9 per 100,000 accessing appropriate surgical cancer care can be an anesthesiologists (Hoyler insurmountable challenge. and others 2014) Number of operating > 14 per 100,000 < 2 per 100,000 Utilization rooms (Funk and others Transportation costs and the time required to access 2010) diagnostic and treatment facilities may act as deterrents Volume of operationsb 172.3 million procedures 8.1 million procedures to receiving timely care. Even traveling relatively short (Weiser and others 2008) per year (73.6 percent per year (3.5 percent distances can be a significant barrier in countries with of global total, for of global total, for poor transport infrastructure or challenging terrain. In 30.2 percent of the 34.8 percent of the global population) global population) a study of South African women presenting with breast cancer, the risk of presenting with an advanced stage can- a. High-income country data refer to “surgical providers” and include surgeons and anesthesiologists within the same estimate. cer was 1.25 times higher for every 30 kilometers traveled b. Data refer to high-health-expenditure countries and low-health-expenditure countries, which are to the diagnostic facility (Dickens and others 2014). In a correlated with income status. Surgical Services for Cancer Care 225 equipment, supplies, and sterility, as well as a lack of Partners in the Provision of Surgical Services appropriately trained providers. Poor outcomes rein- The safe provision of anesthesia is another often over- force community perceptions that cancer cannot be looked requirement of effective surgical cancer care. successfully treated with surgery. Although lumpectomy Profound differences exist in anesthetic mortality rates and modified radical mastectomy for breast cancer are between low and high Human Development Index not technically complex procedures, inadequate surgical countries (Bainbridge and others 2012); anesthetic mor- resection of tumors can significantly undermine the tality rates are reported to be as high as one in 500 in effectiveness of these procedures. Incomplete or inad- several LMICs in Sub-Saharan Africa (Glenshaw and equate breast cancer resection following lumpectomy Madzimbamuto 2005; Hansen, Gausi, and Merikebu or mastectomy has been reported at rates as high as 2000; Maman and others 2005; Walker and Wilson 2008; 15–45 percent in India and Nigeria (Agarwal and others see Debas and others 2015, chapter 15.) 2009; Thorat and others 2008; Ukwenya and others Cancer surgery is also highly dependent on two other 2008); almost 50 percent of patients who underwent major areas of clinical care: pathology and imaging. incomplete surgery in nonspecialist centers in India had Quality pathology services are central to making an surgically excisable disease left behind (Thorat and others accurate diagnosis and planning appropriate surgical 2008). Postgraduate training, which covers modern surgi- care. Imaging is required for accurately staging early, cal oncology practices and continuing medical education, curable cancers; for planning more complex operative is lacking in many LMICs. This deficit impacts not only resections; and, in some cases, for establishing the pres- proper surgical oncology technique but also appropriate ence of metastatic disease. The role of these services in decision-making, including whether surgery is indicated. the delivery of quality comprehensive cancer is discussed further in chapter 11 in this volume. Standardization of Guidelines The standardization of surgical care with guidelines, standards, and checklists can ensure a minimum level DEVELOPING SURGICAL CANCER SERVICES of quality and safety and reduce avoidable surgical AND DELIVERY PLATFORMS IN LMICs morbidity and mortality (Haynes and others 2009), even in resource-constrained settings (see Debas and Resource-Stratified Approaches others 2015, chapter 16). Current use of guidelines and Resource-stratified approaches to screening, diagnosis, standards for surgical care in LMICs varies among coun- and treatment interventions for specific cancers, which tries and facilities. A recent study of health facilities in can help countries assess the level at which they can 24 LICs and 27 MICs reported that only 22 percent of provide effective cancer services, have been presented in facilities (n = 294/1,269) had established clinical manage- previous chapters. However, policy makers developing ment guidelines for surgical care and pain relief (Ilbawi, cancer control strategies need to consider not only what Cherian, Mikkelson, Sankaranarayanan, and Sullivan, services are required, but also the platforms through unpublished data). Most clinical guidelines have been which these services can be most effectively delivered to developed in and for HICs and are not necessarily those who need them. applicable in resource-poor settings. However, for the In this section, we outline potential delivery platforms past decade, Tata Memorial Centre, a national compre- for surgical cancer services in resource-poor settings, hensive cancer center in Mumbai, India, has published its using a level-of-care approach (box 13.1). We con- own clinical guidelines and algorithms for all aspects of sider how surgical cancer services—diagnostic, curative, cancer care, including surgical and perioperative care. The palliative, and adjuvant services—may be effectively guidelines are developed and updated through annual delivered across different surgical platforms (commu- evidence-based management meetings, using interna- nity health center, first-level hospital, or third-level tional evidence and taking into account local resources hospital), using breast, cervical, oral, and bowel cancers and challenges. Tata’s guidelines (freely available for as examples. Where relevant, we consider the most reference on their institutional website, https://tmc.gov appropriate surgical platform for service delivery for .in/clinicalguidelines/clinical.htm) are now used in other countries at different income levels and according to LMICs, including Bangladesh, Kenya, and Nigeria. There the resource-stratified interventions presented in earlier is also an initiative to categorize some of these guidelines chapters.1 Finally, we discuss how quality and efficiency as “minimum,” “optimal,” and “optional,” with health demands can be balanced with access and coverage chal- care delivery platforms treating patients based on the lenges in LMICs through the appropriate deployment of platforms’ individual infrastructural and trained human surgical cancer platforms, and the referral networks and resource capabilities. service partnerships between them. Many countries are 226 Cancer Box 13.1 Situational Analysis of Surgical Cancer Services: Key Questions Key Questions for Policy Makers and Planners • Are radiotherapy services available? Where? • What is the burden of surgically treatable cancers • Where are they available in relation to surgical in the country? and adjuvant treatment? • Current • Are palliative medicines, such as opioids, reliably • Projected available? Where? • What stage of presentation is typical for each can- • What referral networks exist? What are the cer (percent early, locally advanced, disseminated)? barriers to referring patients between facilities? • What surgical platforms are currently available • What are the barriers to receiving timely and within the country? Where are they located? appropriate surgical and cancer treatment? • Do any of these platforms currently provide • Financial, geographic, sociocultural surgical cancer care: • Human resources, infrastructure, equipment • As part of a general surgical service? and supplies • As part of a dedicated cancer service? • Is it feasible to provide screening and early case • How well-resourced are these platforms? detection, given the country’s resource level and • Human resources, infrastructure, equipment, priorities? supporting services • Are there plans to do this in the medium-to- • What adjuvant therapies are available and afford- long term? able for the country’s resource level? • How will this affect surgical need? • Where are these adjuvant therapies currently delivered, if anywhere? • Who delivers them? only beginning to consider these issues; very little analy- to support the effective delivery of clusters of health sis or published country experience in LMICs is available services. to serve as an evidence base. Diagnostic, Curative, and Palliative Services Diagnosis Surgical services play a key role in cancer Guidelines for Surgical Platforms diagnosis. Biopsy, which is required for the definitive Delivery platforms refer to the structural and organiza- diagnosis of cancer, involves taking a sample of suspi- tional modes or channels of delivery for public health cious tissue by using either a needle or an open surgical and clinical services. Platforms for delivery of surgical technique and then examining the removed cells under care can be defined across four levels: a microscope. The tissue sampling aspects of the biopsy procedure can be provided in most LMICs within a • Community health center first-level hospital platform, as well as higher platforms, • First-level hospital/district hospital if the surgical providers are trained in the technique used • Second-level hospital/regional hospital and adequate means for sample fixation exist. Providing • Third-level hospital/tertiary hospital biopsy services at a first-level platform reduces delays between initial presentation and definitive diagnosis and The basic resources required for each level are sum- improves access and coverage. Because lymphadenopa- marized in table 13.2. In practice, significant variations thy has many non-neoplastic causes in LMICs (Kingham and overlap occur among levels of care. and others 2013), referral to a higher specialist service Delivery platforms for surgical cancer services coex- for the purposes of tissue sampling only is premature, ist with other platforms delivering general cancer and increases losses to follow-up, delays diagnosis, and risks surgical services, inpatient services, and primary care. overwhelming limited specialist services with nonspe- They are often co-located and operate synergistically cific referrals. Surgical Services for Cancer Care 227 Table 13.2 Platforms for Delivering Surgical Cancer Care Community health center District/first-level hospital Regional/second-level hospital Tertiary/third-level hospital • Community health center • District– or provincial-level • Referral hospital of 200–800 • Referral hospital of 300–1,500 or small rural hospital hospital, with 50–300 beds beds beds • May have a small number • Adequately equipped major • Well-equipped major and minor • Well-equipped major and of inpatient and maternity and minor operating theaters operating theaters minor operating theaters beds • Trained nonphysician or • Supported by imaging, • Advanced imaging, laboratory • Capable of performing medical officer anesthetists laboratory, and blood bank services minor surgical procedures • District medical officers services, as well as basic • Intensive care facilities under local anesthesia in surgery, senior clinical intensive care facilities • Highly specialized staff and • Paramedical staff, nurses, (nonphysician) officers in • Adequately equipped major and technical equipment midwives surgery, nurses, midwives minor operating theaters • Clinical services highly • Visiting doctors • +/− resident general • General surgeons, differentiated by function surgeon and/or obstetrician-gynecologists obstetrician-gynecologist • Often have teaching activities • Anesthesiologists • Visiting specialists • +/− specialist surgeons Source: Adapted from WHO 2003 and Debas and others 2015, chapter 12. Tissue sampling itself is not technically complex. In LICs with basic surgical resources at first-level Accurate and timely reporting of the biopsy sample by facilities, a minimum package of surgical services for a trained pathologist is the main challenge in obtaining cancer can be delivered within this platform. This pack- a diagnosis. In LICs, specimens may be taken at the age includes biopsy, surgical treatment for precancerous first-level hospital level but processed and reported at a cervical lesions and early-stage invasive cervical cancer, higher center, often within a third-level or national plat- breast cancer surgery, and resection of small oral tumors form, because of the lack of trained histopathology tech- (table 13.3). Provided there is a surgical provider famil- nicians and pathologists. This approach can maximize iar with cancer resection requirements, these procedures available resources and also promote standardized, qual- require little additional infrastructure, equipment, or ity reporting. However, it requires coordination between supplies, compared with other major general surgical the tissue sampling center, where the biopsy sample is operations routinely performed at first-level facilities. taken, and the pathology center, where the biopsy is read Treatment of some precancerous lesions can be safely and reported, to ensure timely feedback of the diagnosis. undertaken within a community or first-level platform, In MICs, histopathology services may be more even where full general surgical services are not available. widely available, and basic pathology services may Rwanda has recently published its experience scaling be provided within a first-level hospital platform. up cervical screening and treatment services across the Centralized approaches to reading and reporting cancer country using first-level or community-level facilities to biopsies are still important, however. These approaches screen, diagnose, and treat, often within a single clinical can promote the efficient use of resources; increase encounter (Binagwaho and others 2013). the range of diagnostic tests able to be performed; and In MICs, surgical services for early-stage breast, cervi- ensure standardized, quality reporting. For this reason, cal, colon, and oral cancer can often be delivered within the use of second- or third-level platforms for biopsy a first-level platform because of the greater availability of reporting is encouraged. basic surgical resources, including surgeons. This delivery can improve access and may reduce the direct nonmedical Treatment with Curative Intent In LICs, surgical ser- costs associated with seeking surgical cancer care in many vices for cancer are usually provided through second- or MICs. Appropriate training and continuing education of third-level platforms. Severe shortages of surgical infra- surgical providers at the first level is crucial, however, to structure, equipment, surgeons, anesthesiologists, and reduce the risk of inadequate or incomplete resection. supporting services preclude providing these services Advanced breast, cervical, oral, and colorectal cancers within a first-level platform. Cancer surgery is typically require advanced surgical platforms in LMICs, typi- performed by generalist surgeons, as specialist cancer cally a dedicated regional or national center providing surgeons are not available. cancer care. Advanced cancers are technically more 228 Cancer Table 13.3 Delivery Platforms for Priority Surgical Cancer Interventions in LICs and MICs Community health District (first-level) Regional (second-level) Intervention center hospital hospital Tertiary (third-level) hospital Breast cancer: LICs Diagnosis Refer to higher center Biopsy (send pathology to Biopsy ± onsite pathology, Biopsy + onsite pathology, higher center) imaging (XR, liver US), lab (CBC, imaging (XR, liver US), lab (CBC, LFT) LFT) Curative surgical treatment ″ Referral to a higher center MRM ± oophorectomy MRM ± oophorectomy Palliative surgical treatment ″ Referral to a higher center Total mastectomy Total mastectomy Adjuvant therapy ″ Hormone therapy Hormone therapy, chemotherapy Hormone therapy, chemo, RTa Breast cancer: MICs Diagnosis Refer to higher center FNA/US-guided FNAB, FNA/US-guided FNAB + onsite FNA/US-guided FNAB + onsite imaging (XR, liver US), lab pathology, imaging (XR, liver US), pathology, imaging (XR, liver US), (CBC, LFT) lab (CBC, LFT) BCS & SLNB (dye lab (CBC, LFT) BCS & SLNB (dye or radioa,b) or radioa,b) Curative surgical treatment ″ MRM ± oophorectomy MRM ± oophorectomy MRM ± oophorectomy Palliative surgical treatment ″ Total mastectomy Total mastectomy Total mastectomy Adjuvant therapy ″ Hormone therapy, 1st-line Hormone therapy, chemo, RT a Hormone therapy, chemo, RTa chemob Cervical cancer: LICs Diagnosis HPV test, VIA HPV test, VIA HPV test, VIA ± colposcopy, HPV test, VIA ± colposcopy, biopsy biopsy Curative surgical treatment Precancerous Cryotherapy Cryotherapy, LEEP Cryotherapy, LEEP, cold knife Cryotherapy, LEEP, cold knife Invasive cancer Refer to higher center Refer to higher center Simple and radical hysterectomy Simple and radical hysterectomy Palliative surgical treatment ″ ″ Adjuvant therapy ″ ″ Chemo Chemo, RTa Cervical cancer: MICs Diagnosis HPV test, VIA or HPV test, VIA or cytology, HPV test, cytology, colposcopy, HPV test, cytology, colposcopy, cytology colposcopy, biopsy biopsy biopsy Curative surgical treatment Precancerous Cryotherapy Cryotherapy, LEEP, cold knife Cryotherapy, LEEP, cold knife Cryotherapy, LEEP, cold knife Invasive cancer Refer to higher center Simple hysterectomy; Simple and radical hysterectomy Radical trachelectomy, advanced cancer, refer to hysterectomy, pelvic exenteration higher center Palliative surgical treatment ″ Adjuvant therapy ″ Chemob Chemo, RT Chemo, RT Oral cancer: LICs Diagnosis Refer to higher center Biopsy Biopsy + histopathology Biopsy + histopathology Curative surgical treatment ″ Resection of early-stage Resection of early and advanced Resection of early and advanced cancers Palliative surgical treatment ″ Refer to higher center For debulking/pain relief For debulking/pain relief Adjuvant therapy ″ ″ a RT ± chemo RTa ± chemo table continues next page Surgical Services for Cancer Care 229 Table 13.3 Delivery Platforms for Priority Surgical Cancer Interventions in LICs and MICs (continued) Community health District (first-level) Regional (second-level) Intervention center hospital hospital Tertiary (third-level) hospital Oral cancer: MICs Diagnosis Refer to higher center Biopsy Biopsy + histopathology Biopsy + histopathology Curative surgical treatment ″ Resection of early-stage Resection of early and advanced ± Resection of early and advanced cancers oncoplastics ± oncoplastics Palliative treatment ″ Refer to higher center For debulking/pain relief For debulking/pain relief Adjuvant therapy ″ ″ a RT ± chemo RTa ± chemo Colorectal cancer: LICs Diagnosis Refer to higher center gFOBT/FIT + referral for Sigmoidoscopy/colonoscopy Sigmoidoscopy/colonoscopy colonoscopy Curative surgical treatment ″ Colectomyb Colectomy, APRb, LARb Colectomy, APR, LAR Palliative surgical treatment ″ Colostomy for bowel Colostomy for bowel obstruction Colostomy for bowel obstruction obstruction Adjuvant therapy ″ Refer to higher center Chemo ± RTa Chemo ± RTa Colorectal cancer: MICs Diagnosis FOBT + referral for Colonoscopy + biopsyb Colonoscopy + biopsy Colonoscopy + biopsy colonoscopy Curative surgical treatment Refer to higher center Colectomy Colectomy, APR, LAR Colectomy, APR, LAR Palliative surgical treatment ″ Colostomy for bowel Colostomy for bowel obstruction Colostomy for bowel obstruction obstruction Adjuvant therapy ″ Chemoa,b Chemo ± RT Chemo ± RT Note: APR = abdominoperineal resection; BCS = breast-conserving surgery; CBC = complete blood count; FIT = fecal immunochemical test; FNA = fine needle aspiration; FNAB = fine-needle aspiration biopsy; FOBT = fecal occult blood test; gFOBT = guaiac fecal occult blood test; HPV = human papillomavirus; LAR = lower anterior resection; LEEP = loop electrocautery excision procedure; LFT = liver function test; LICs = low-income countries; MICs = middle-income countries; MRM = modified radical mastectomy; RT = radiotherapy; SLNB = sentinel lymph node biopsy; US = ultrasound; VIA = visual inspection of the cervix after acetic acid application; XR = x-ray; ± = with or without; ″ = repeats above. a. If available within a country’s resource level. b. Provision at this level will be dependent on the availability of appropriate equipment, supplies, monitoring, and adequately trained providers. complex to achieve adequate resection margins and Palliative surgical procedures commonly required wound closure. Platforms capable of delivering com- in LMICs include mastectomy for bulky, fungating, plex cancer and surgical care are often not available in or bleeding tumors and formation of a colostomy for LICs, especially outside the capital city. In addition to obstructing colorectal tumors. In LICs, palliative colos- these priority cancers, other complex cancers (for exam- tomy formation or mastectomy can be performed within ple, musculoskeletal, thoracic, or hepatobiliary cancers) a second-level platform, or potentially at a first-level require surgical treatment within third-level platforms, facility when resources permit. In MICs, most first-level usually by specialist surgeons. platforms are equipped to provide this level of surgical care. Palliative surgical treatment must be undertaken Treatment with Palliative Intent Palliative surgery cautiously. It should be made clear that the procedure can significantly enhance the quality of life and allow is being done to improve the quality of life, rather than patients to return home for end-of-life care. Palliative to extend it. Advanced disease has higher operative and care for all patients with advanced-stage cancer hinges postoperative risks; the risks and benefits of the proce- on access to appropriate analgesics, including opioids dure must be weighed carefully by providers and patients. (see chapter 9). Surgery also has an important role in palliation, particularly in regions in which advanced Adjuvant Treatment Considerations presentations with very large, debilitating tumors are Cancer treatment with surgery alone is only effective common. Palliative treatment should be provided within in early-stage disease. In resource-constrained settings, delivery platforms as close to patients’ homes as possible. most patients tend to present with advanced disease, and 230 Cancer adjuvant therapy is usually required in addition to sur- Table 13.4 Benefits and Risks of Centralized versus Decentralized gical resection. Strong coordination of surgical services Surgical Cancer Platforms and adjuvant services is needed to maximize outcomes, Centralized surgical platforms for Decentralized surgical platforms and additional considerations present with respect to cancer for cancer the most appropriate surgical platform for patients who require both surgical care and adjuvant therapy. Benefits Benefits In LMICs, platforms for basic surgical cancer care • Standardization of care, higher • Improved coverage and access for are likely to be more widely available than for adjuvant operative volumes, and specialist greater equity treatment, particularly radiotherapy. When planning surgical care for quality assurance • Reduced direct nonmedical and cancer services, policy makers need to consider not only • Economies of scale indirect costs to patients and where surgical services are provided, but also how these families, because of reduced travel • “One-stop shop” for cancer services are distributed in relation to where adjuvant therapy— time and productivity loss including hormonal therapy, chemotherapy, radiother- • Multidisciplinary practice for better outcomes • Reduced referral delays between apy, and biologics—can be provided. The availability presentation and definitive care of these services may dictate whether surgical treat- • Research and training activities that drive practice forward • Surgical platform more cost- ment is appropriate and the type of intervention to be effective at the first or second level performed. (Debas and others 2006) In many LMICs, the surgical providers are often Risks Risks responsible for prescribing and/or administering adju- vant therapy. This is very common when adjuvant • Reduced access and increased • Inefficient clinical services and endocrine therapy is required in the setting of breast inequity for rural versus urban duplication populations • Poor coordination and access to cancer, for example, tamoxifen. Adjuvant chemotherapy is also often given by general surgeons, physicians, and • May encourage super-specialization higher-level centers and other even patients’ families in LMICs. Ideally, chemotherapy and workforce maldistribution cancer disciplines, causing delayed should be delivered in a comprehensive cancer center or missed adjuvant care by specialist staff within a second- or third-level plat- • Poorer quality care form to ensure appropriate, high-quality care. However, these stipulations place chemotherapy out of reach for many LICs. Where significant barriers exist to accessing delivery; balancing quality and efficiency with access chemotherapy and prevent uptake, preoperative or post- and coverage demands is a key challenge in delivering operative first-line chemotherapy can be administered surgical cancer care in LMICs. Centralized, specialist by trained surgeons, general physicians, or nurses at surgical platforms for cancer services generally promote first- or second-level hospitals, using clinical guidelines quality and efficiency, whereas strengthening delivery and management algorithms to guide treatment selec- platforms peripherally tends to enhance access and cov- tion, if appropriate blood tests are available to monitor erage (table 13.4). complications. Such polyskilling (where a provider is trained to deliver more than one type of cancer care) can be used to overcome human resource shortages and Referral Networks, Service Coordination, and minimize referral delays. Partnerships The delivery of radiotherapy is limited by its avail- The delivery of surgical cancer care requires functional ability; in all LICs and most MICs, delivery requires clinical platforms, as well as strong referral networks and referral to a regional or national platform. The avail- coordination between other cancer services and providers. ability and accessibility of radiotherapy at a higher Strategies to improve the coordination and links center do not necessitate the delivery of surgical care among all platforms providing cancer services can pro- at the same center, although there may be advantages mote high-quality, standardized, and efficient surgical in doing so. cancer care. For example, India has developed a National Cancer Grid (Pramesh, Badwe, and Sinha 2014), funded by the Government of India, which links facilities pro- Centralized versus Decentralized Delivery Models viding cancer care, with the goal of standardizing the Delivery platforms for surgical cancer services must quality of care, developing uniform guidelines, reduc- necessarily be organized into an overall delivery model ing the variations in care, and facilitating exchanges of within a country. It is useful to consider the benefits expertise and experience between larger and smaller and risks of different models of surgical cancer service centers. Such links also strengthen referral capabilities Surgical Services for Cancer Care 231 and provider coordination. This is particularly impor- the available resources, current and projected, for the tant when diagnostic, surgical, and adjuvant services are scale-up of cancer care and surgical services spread across different facilities. Comprehensive cancer centers with multidisciplinary A situational analysis of current surgical and cancer cancer teams have been shown to be the most effective capacity within a country should precede policy, plan- strategy for ensuring high-quality, efficient, and appro- ning, and scale-up efforts (box 13.1). priate cancer care in HICs (chapter 11 in this volume; Yip and others 2011). The severe shortage of specialist health workers makes it almost impossible to achieve Developing the Surgical Workforce comprehensive, multidisciplinary centers currently in Human resources are a crucial component of surgical LICs and difficult to achieve in a manner that ensures cancer services, and the development of an effec- high coverage and equity in many MICs. However, even tive workforce requires proactive strategic planning in the absence of a highly specialized cancer workforce, at the national level. LICs and many MICS require some LMICs are beginning to develop regional or urgent investment in strengthening the surgical, anes- national cancer centers, drawing on expertise within thetic, and supporting cancer workforce—including general second- or third-level hospitals. Often, one or pathologists, radiotherapists, and nurses trained in two surgeons within a country become well known for perioperative and wound care. The surgical and anes- providing cancer care and serve as references for the rest thetic workforce takes time to develop—a minimum of of the country, with high numbers of patients referred 10 years from entry into medical school to qualification to them. These reference surgeons and the large urban as an accredited surgeon or anesthetist—and workforce hospitals in which they typically work can serve as a planning must take into account projected as well as major focus to drive forward cancer care within coun- current needs. Many LICs lack postgraduate surgical tries, provided they are well supported. Although not all training programs and must pay to send their doctors surgical cancer services need to be provided at this level, outside the country (and sometimes outside the region the presence of such centers may strengthen the surgical or continent) for further training after medical school. care provided at other locations through the exchange This requirement is costly and increases the likelihood of knowledge and experience and the strengthening of that the home countries will not be able to retain the referral networks. doctors upon training completion. Creating the capac- International partnerships between LMICs or ity for accredited postgraduate surgical training in LICs between LMICs and HICs also support the development has been shown to be effective and sustainable, allowing and delivery of cancer care, including surgical cancer countries to achieve national health goals (Anderson care in low-resource environments. The most effective and others 2014). international partnerships are those that seek to develop Task-shifting of general surgical procedures—for local cancer care capacity and that are closely aligned example, laparotomy, cesarean section, and fracture with local needs. The practice of short-term surgical repair—to nonphysician providers is increasingly used trips that focus on operative resection only, use entirely to overcome critical surgical workforce shortages in foreign surgical teams to deliver care, and do not partic- many LMICs. In Malawi, 93 percent of the surgical ipate in teaching or local capacity-building efforts is not workforce is composed of nonphysicians (Henry and generally an effective model for cancer care. others 2014). However, this process poses risks for developing surgical cancer services. It is generally agreed that task-shifting to nonphysicians for cancer surgery is STRENGTHENING SURGICAL SYSTEMS AND not possible owing to case complexity and quality con- BUILDING CAPACITY cerns. The failure to address the shortage of surgeons in LMICs and the overreliance on nonphysician surgical Conducting Baseline Assessment of Capacity providers to deliver surgical services will significantly At the country level, policy makers will consider several hamper the ability of countries to respond to the sub- key elements, especially when considering the most stantial projected increase in cancer requiring surgical appropriate delivery platforms: treatment in the future. Attempts to address the surgical workforce crisis need to focus on increasing the number • Burden of cancer of surgeons through recruitment and retention to ensure • Stage at diagnosis long-term success in meeting surgical needs. Training of • Availability and distribution of surgical and cancer- surgical nursing staff is also critical to ensure optimal specific resources in relation to the population and postoperative care and surgical outcomes. 232 Cancer In settings with an adequate surgical workforce, as in HICs and assist in detecting cases at stages amenable some MICs, expanding the skills of the existing work- to curative surgical treatment. However, the greatest force to provide quality surgical cancer services through overall gains are likely to come from the planned ongoing training will improve outcomes and maximize development of more basic surgical infrastructure, health gains. with good population coverage, rather than the ad hoc purchasing or donation of state-of-the art tech- nology or facilities that can be accessed by only a small Improving Infrastructure and Procurement Processes percentage of the population. Maintenance and repair The significant deficits in basic infrastructure, equip- of surgical infrastructure and equipment are major ment, supplies, and procurement processes in many challenges; an estimated 40 percent of the equipment LMICs need to be addressed early in any scale-up plans. in LMICs is out of service, compared with less than These deficits include an absence of reliable power, 1 percent in HICs (Howitt and others 2012). The water, and oxygen, as well as insufficient or dilapidated inappropriate deployment of medical technologies operating theaters and surgical and sterility equip- from HICs to LMICs is a significant contributor to ment and supplies. Attention to the development of this problem. sustainable supply chains and procurement practices is important. Improving and developing the surgical infrastructure within countries often requires capital Promoting Quality and Ensuring Safety outlays; in LICs, these costs may need to be met through Prerequisite to the scale-up of surgical cancer services is development assistance. consideration of how to promote and ensure quality and Further research is needed as to the most appro- safety. These are fundamental components for achieving priate and cost-effective infrastructure and equipment good outcomes and building community trust in cancer for surgical cancer care specific to the resource level. and surgical care. All countries can embrace the goal of In some cases, the use of technology in LMICs, for high-quality and safe surgical care, regardless of devel- example, human papillomavirus DNA testing, can opment status. Specific strategies for LMICs are listed lead to leapfrogging of cancer delivery models over in box 13.2. Box 13.2 Strategies to Improve the Quality of Surgical Cancer Services in LMICs All LMICs • CME for all surgical cancer providers • Clinical management guidelines and surgical • CME and regular courses for updates on sur- standards developed specifically for low-resource gical technique, patient selection, postopera- settings tive care, and systemic therapy • Collection of outcome data • Case fatality rates LIC-specific strategies • Risk-adjusted postoperative mortality rates • Focus on developing strong general surgical • Morbidity and mortality meetings and clinical services and referral mechanisms audits • Operation within the limits of the human • Encouraged reflection on practice and identi- and infrastructural resources to reduce poor fication of areas for improvement outcomes • Multidisciplinary approach to diagnosis and • Establishment of formal links among centers pro- treatment management viding surgical and cancer care within a country, • Local especially between different referral levels • International, for example, via telemedicine • Development of international twinning arrange- links ments box continues next page Surgical Services for Cancer Care 233 Box 13.2 (continued) • Support for training, diagnosis, and case manage- • Establishment of cancer grids or partnerships ment decisions in centers providing cancer care • Encourage collaboration and standardization • South-South of surgical care • North-South • Development of regional and national cancer • Local and international NGOs registries to track outcomes • Requirements for mandatory reporting of MIC-specific strategies case volumes, procedures, and outcomes in • Development of regional and national compre- all sectors providing surgical cancer services hensive cancer centers (government, private for-profit, and private • Provision of locally appropriate management not-for-profit) guidelines for own country Note: CME = continuing medical education; LICs = low-income countries; • Provision of training support and outreach LMICs = low- and middle-income countries; MICs = middle-income countries; clinical services for peripheral facilities NGOs = nongovernmental organizations. Scaling Up Surgical Services for Cancer and delivery within a functioning health system may be The requirements for the scale-up of surgical services weak. As countries move beyond the most basic package to meet cancer needs are country specific, dependent of cancer care delivered within a single clinical encoun- on current and projected patterns of disease, available ter, they will require complex and highly coordinated health resources and health systems capacity, amounts delivery systems, with surgical care embedded within. of domestic spending on health, and distribution of Improving governance and regulation around surgical the population. Some general recommendations can service provision will assist MICs to improve quality, be made, however, to guide policy makers based on the reduce waste and inefficiency, and promote equity. resource patterns, income level, and development status. Large imbalances between private and public sector LICs should initially focus on building general sur- provision of surgical cancer services are seen in some gical capacity and inpatient care within their health MICs, such as India. Unregulated, these imbalances systems, including investing in human resources and can drain resources (for example, higher salaries in hospital infrastructure and developing effective supply the private sector drain surgeons away from the public chains and referral networks. Without these fundamen- sector), hinder quality and transparency (for example, tals in place, it is not appropriate to embark on cancer through inappropriate, nonstandardized, or unwanted surgery–specific treatment planning. Adequate general surgical treatment), increase medical impoverishment surgical capacity will allow countries to deliver the (for example, treating patients until finances have run surgical components of the minimum cancer interven- out and then transferring them to the public sector), tion package, such as diagnosis and treatment of breast and create a two-tiered system of cancer care (Flores and cancer and treatment of precancerous cervical lesions, others 2008; Pramesh and others 2014). at the basic resource level. Importantly, it will also serve Complementing steps to improve surgical capacity is as a base for the effective scale-up of a range of cancer- the need to simultaneously focus on removing patient specific services. barriers to the uptake of surgical cancer services to In MICs with basic or limited surgical resources in improve cancer outcomes and promote equity. Delayed place, the focus should be on developing coordinated presentation increases the morbidity, mortality, and and context-specific cancer systems and services that micro- and macroeconomic costs associated with cancer. improve the quality and standards and ensure equita- As countries move to introduce financial risk protec- ble access to surgical cancer care through sound public tion and progressive universal health coverage for their policy and health governance. Many MICs have national populations, there is a need to ensure coverage for a basic health programs, services, and structures geared to the package of inpatient care, including surgical care, early in delivery of vertical programs, rather than horizontal the expansion pathway (Jamison and others 2013). health system–based approaches (Anderson and oth- Cancer care requires strong, coordinated health sys- ers 2014). Surgical care may be present, but coordination tems and services, rather than an isolated focus on 234 Cancer surgical services. Early detection and comprehensive specific to the local cost structure and, especially given treatment improve cancer outcomes. Improving the the heterogeneity of health care financing in MICs, are rate of surgical cure in LMICs requires coordinated not generalizable to other countries. The results often efforts across the health system to achieve stage-shifting, reflect the increasing ability and desire of individuals combined with efforts to improve surgical capacity to and governments to pay for a perceived (if unsupported) deliver effective treatment. For example, clinical breast qualitative improvement in outcomes, balanced against examination provided at a community-level platform increased costs of more expensive (often imported) by trained allied health workers has led to stage-shifting equipment, more highly trained personnel, and more of breast cancer in India, making it more amenable to supporting services. However, these studies also often surgical cure (Sankaranarayanan and others 2011). provide insights into and implications for the structur- ing of health care financing and equitable access. In reviews of the literature, basic surgical services ECONOMIC CONSIDERATIONS OF SURGICAL in a variety of low-resource settings were reported to CANCER CARE IN LMICs be cost-effective or very cost-effective, according to the World Health Organization threshold definitions (Chao There have been few economic evaluations of cancer and others 2014; Grimes and others 2014). Local costing care in LMICs; among these, surgical interventions studies and partial economic evaluations (for example, and surgical services have received almost no atten- where costs or effectiveness components are assessed tion. Tables 16.3 to 16.8 in chapter 16 summarize the but not directly linked) or evaluations from narrower available cost-effectiveness evidence for the detection perspectives (for example, from provider or patient per- and treatment of the priority cancers considered in this spectives rather than societal perspectives) can provide volume. Notably, surgical interventions that are feasible insights for inputs into fuller cost-effectiveness studies or at the basic, limited, and enhanced resource levels have for intervention adaptation in implementation. Detailed barely been assessed, even in upper-middle-income costing of surgery procedures, excluding preoperative and countries. Chemotherapy, in comparison, is a far more postoperative care, in a selection of hospitals of varying studied treatment modality, given concerns about its resources and settings in India showed that the salaries high cost and poor accessibility, regardless of resource and benefits of operating theater staff formed 42 percent level. For example, in a systematic review of the Tufts of the cost of a hysterectomy in a first-level hospital, com- Medical Center Cost-Effectiveness Analysis Registry of pared with 48 percent in a third-level hospital (Chatterjee cancer-related studies set mainly in HICs, 53.3 percent and Laxminarayan 2013). Overhead costs were higher at of the studies were concerned with pharmaceutical the first-level hospital, however, constituting 30 percent interventions, compared with 13.3 percent with surgical versus 20 percent of hysterectomy costs at the third- interventions (Greenberg and others 2010). level hospital. This finding suggests that if the outcomes Yet, surgery is the most significant life-saving inter- are similar, it may be equally or more cost-effective to vention in cancer treatment. Coupled with their wider perform simple hysterectomies at a first-level hospital, roles in the cancer care spectrum, as a diagnostic modal- thereby improving access for a wider population in India. ity and in palliative care, surgery and surgical services In costing breast cancer care in central Vietnam, a have the potential to be good-value choices for health lower-middle-income country, Lan and others (2013) care investment in LMICs. The expansion of surgical found the surgical treatment, while a large cost compo- interventions for solid tumors routinely found at early nent, was significantly less expensive than chemotherapy. stages is recommended in this volume based on feasi- Over a five-year course of care for breast cancer that bility, at even basic and limited resource levels, and sug- included diagnosis, initial treatment, and follow-up gested cost-effectiveness evidence from higher resource care, surgery accounted for 8.4 percent of the total cost levels (chapter 16 in this volume). (Lan and others 2013). Economic Studies Financing In the enhanced resource settings of MICs, limited In many LMICs, out-of-pocket payment for surgical cost-effectiveness analyses of comparisons between cancer services may be the main form of financing simple and enhanced surgical techniques or between (Ilbawi, Einterz, and Nkusu 2013). Lan and others (2013) surgery and other treatment modalities are emerging found that the absence of health insurance or financial (He and others 2011; Lu and others 2012; Tan and others risk protection from the costs of cancer care in Vietnam 2013). The results are set in single hospitals and are very was the main barrier to the uptake of breast cancer Surgical Services for Cancer Care 235 treatment services. The impoverishment impact of sur- the differential training of personnel. In LMICs in par- gical conditions on a household is immense, especially in ticular, there is need for costing of the surgical cancer the context of cancer. In a study in rural Bangladesh, the systems, processes, and platforms that would allow the impoverishment rates from cancer hospitalization and identification of minimized patient travel time and surgical procedures were four- to sevenfold higher than related productivity costs. the impoverishment average of 3.4 percent for all health services (Hamid, Ahsan, and Begum 2014). High user Effectiveness fees and out-of-pocket payments also increase the like- The short- and long-term effectiveness of surgical lihood that patients will not return at all for definitive services in cancer cure and control, measured at the surgical care. In a study of patients presenting at a first- national and sub-national level, could be estimated to level hospital in rural Cameroon, preoperative payment better inform cost-effectiveness analyses. The efficacy greater than US$310 and a recommended procedure for of cancer surgery may be severely compromised by poor cancer significantly increased the likelihood of patients access to supporting cancer services, including chemo- not returning for surgical care as advised following an therapy and radiotherapy, or by poor quality surgical initial assessment (Ilbawi, Einterz, and Nkusu 2013). care. However, there has been little evaluation of the potential impact of this on cancer outcomes in LMICs. Gaps in the Economic Evidence The dearth of economic evaluations for surgical cancer CONCLUSIONS services means that many knowledge gaps exist in mak- Surgical services are a central component of cancer cure ing investment decisions. This section identifies some of and control in all resource settings, playing a key role the fundamental areas that can be addressed to start an in the diagnosis, treatment, and palliation of most solid economic evidence base of cancer surgery interventions tumors. Basic surgical cancer care can be affordable and and surgical services. effective, even in countries with substantial resource constraints. This fact has not been well recognized in Burden previous dialogues on cancer control in LMICs. Country-level estimates of the health burden of resec- Major resourcing, geographic, financial, and socio- table cancer, refined by site, incidence, and stage, are cultural barriers to access to surgical cancer services exist unknown but required for the underlying foundation in many LMICs. Given the high case-fatality rates from of an economic evidence base. Estimates of avertable common malignancies such as breast cancer in LMICs, burden help direct the considerable resources needed for as well as the large projected increase in cancer inci- economic evaluations to appropriate areas of research, dence in these regions over the next 20 years, countries identify proper comparators, and give a measure against would benefit from strategic and proactive approaches which to weigh costs. In the United States, 61.4 percent to the planning and delivery of surgical cancer services. of patients admitted to hospital with a cancer diagnosis Unfortunately, very little is known about the most effec- required a surgical procedure (Rose and others 2014). tive or cost-effective delivery platforms for surgical can- Similar country-level estimates for operative cancer need cer care in LMICs to guide policy makers, or about how in LMICs are not available. Given the increasing and applicable or transferable models and lessons from HICs changing burden of cancer (see chapter 2 in this volume) are to low-resource settings. Current models of care relative to communicable diseases and among sites and delivery in LMICs have been largely developed through nations, the extent of the potential value of surgical experience, pragmatism, and consensus, rather than treatments needs to be quantified. through rigorous academic or economic evaluation. Key considerations in the scale-up of surgical cancer Costing care in LMICs that are supported by evidence include There is a general lack of costing studies on which to the urgent need to develop the surgical workforce, build cost-effectiveness studies of surgical cancer ser- improve basic general and surgical infrastructure, and vices. As a first step, those that exist for general surgical strengthen supporting services. Coordinated integration services, such as the hospital-based studies of Chatterjee of surgical services with other cancer services and the and Laxminarayan (2013), could be validated for sur- development of cancer networks and partnerships are gical cancer services. This process requires the charac- also required to promote quality and standards. terization and differentiation of cancer surgery costs Most important, efforts to improve surgical capacity versus general surgery costs, including the appropriate in LMICs need to be coupled with strategies to pro- apportioning of overhead costs to cancer surgery and mote cancer stage-shifting. Resource-appropriate efforts 236 Cancer across the health system to facilitate the early detection World Health Organization 91 (9): 697–703. doi:10.2471/ of surgically treatable cancers and reduce barriers to Blt.12.116087. timely service uptake are required to realize fully the Chao, T. E., K. Sharma, M. Mandigo, L. Hagander, S. C. Resch, curative benefits—as well as the associated social and and others. 2014. “Cost-Effectiveness of Surgery and Its Policy Implications for Global Health: A Systematic Review economic gains—that surgery can offer. and Analysis.” The Lancet Global Health 2 (6): E334–45. Chatterjee, S., and R. Laxminarayan. 2013. “Costs of Surgical Procedures in Indian Hospitals.” British Medical Journal NOTES Open 3 (6). Daher, M. 2012. “Cultural Beliefs and Values in Cancer World Bank income classifications as of July 2014 are as Patients.” Annals of Oncology 23 (Suppl. 3): 66–69. follows, based on estimates of gross national income per Debas, H. T., P. Donkor, A. Gawande, D. T. Jamison, M. Kruk, capita for 2013: and C. Mock. 2015. Essential Surgery. Volume 1 of Disease Control Priorities in Developing Countries, 3rd • Low-income countries: US$1,045 or less ed., edited by D. T. Jamison, H. Gelband, S. Horton, • Middle-income countries: P. Jha, R. Laxminarayan, and R. Nugent. Washington, DC: a) Lower-middle-income: US$1,046–US$4,125 World Bank. b) Upper-middle-income: US$4,126–US$12,745 Debas, H., R. Gosselin, C. McCord, and A. Thind. 2006. • High-income countries: US$12,746 or more “Surgery.” In Disease Control Priorities in Developing Countries, 2nd ed., edited by D. T. Jamison, J. Breman, 1. Typically, a country’s income level and development sta- A. R. Measham, G. Alleyne, M. Claeson, D. B. Evans, P. Jha, tus track with its health resource level. However, this is not A. Mills, and P. Musgrove, 1245–59. Washington, DC: always the case; there may be significant variation within World Bank and Oxford University Press. countries in resource availability, according to geography Dickens, C., M. Joffe, J. Jacobson, F. Venter, J. Schüz, and and income quintile. others. 2014. “Stage at Breast Cancer Diagnosis and Distance from Diagnostic Hospital in a Periurban Setting.” International Journal of Cancer 135 (9): 2173–82. doi:10.1002/Ijc.28861. REFERENCES Farmer, P. E., and J. Y. Kim. 2008. “Surgery and Global Health: Adebamowo, C. A., and O. O. Ajayi. 2000. “Breast Cancer in A View from beyond the OR.” World Journal of Surgery Nigeria.” West African Journal of Medicine 19 (3): 179–91. 32 (4): 533–36. Agarwal, G., P. Ramakant, E. R. Forgach, J. C. Rendón, Ferlay J., I. Soerjomataram, M. Ervik, R. Dikshit, S. Eser, J. M. Chaparro, and others. 2009. “Breast Cancer Care in and others. 2013. GLOBOCAN 2012 v1.0, Cancer Developing Countries.” World Journal of Surgery 33 (10): Incidence and Mortality Worldwide: IARC CancerBase 2069–76. No. 11. [Internet]. International Agency for Research on Anderson, F. W., S. A. Obed, E. L. Boothman, and H. Opare-Ado. Cancer, Lyon, France. http://globocan.iarc.fr, accessed on 2014. “The Public Health Impact of Training Physicians 01/September/2014. to Become Obstetricians and Gynecologists in Ghana.” Ferlay, J., E. Steliarova-Foucher, J. Lortet-Tieulent, S. Rosso, American Journal of Public Health 104 (Suppl. 1): S159–65. J. W. W. Coebergh, and others. 2013. “Cancer Incidence and doi:10.2105/Ajph.2013.301581. Mortality Patterns in Europe: Estimates for 40 Countries in Anyanwu, S. N. 2000. “Breast Cancer in Eastern Nigeria: 2012.” European Journal of Cancer 49 (6): 1374–403. A Ten Year Review.” West African Journal of Medicine 19 (2): Flores, G., J. Krishnakumar, O. O’Donnell, and E. Van Doorslaer. 120–25. 2008. “Coping with Health-Care Costs: Implications for the ———. 2008. “Temporal Trends in Breast Cancer Presentation Measurement of Catastrophic Expenditures and Poverty.” in the Third World.” Journal of Experimental and Clinical Health Economics 17: 1393–412. Cancer Research 27: 17. doi:10.1186/1756-9966-27-17. Funk, L. M., T. G. Weiser, W. R. Berry, S. R. Lipsitz, A. F. Merry, Bae, J. Y., R. S. Groen, and A. L. Kushner. 2011. “Surgery as and others. 2010. “Global Operating Theatre Distribution a Public Health Intervention: Common Misconceptions and Pulse Oximetry Supply: An Estimation from Reported versus the Truth.” Bulletin of the World Health Organization Data.” The Lancet 376 (9746): 1055–61. 89 (6): 395. Glenshaw, M., and F. D. Madzimbamuto. 2005. “Anaesthesia Bainbridge, D., J. Martin, M. Arango, and D. Cheng. 2012. Associated Mortality in a District Hospital in Zimbabwe.” “Perioperative and Anaesthetic-Related Mortality in Central African Journal of Medicine 51: 39–44. Developed and Developing Countries: A Systematic Review Goss, P. E., K. Strasser-Weippl, B. L. Lee-Bychkovsky, L. Fan, and Meta-Analysis.” The Lancet 380: 1075–81. J. Li, and others. 2014. “Challenges to Effective Cancer Binagwaho, A., F. Ngabo, C. M. Wagner, C. Mugeni, M. Gatera, Control in China, India and Russia.” The Lancet Oncology and others. 2013. “Integration of Comprehensive Women’s 15: 489–538. Health Programmes into Health Systems: Cervical Cancer Greenberg, D., C. Earle, C. H. Fang, A. Eldar-Lissai, and Prevention, Care and Control in Rwanda.” Bulletin of the P. J. Neumann. 2010. “When Is Cancer Care Cost-Effective? Surgical Services for Cancer Care 237 A Systematic Overview of Cost-Utility Analyses in Cancer: Experience from an Institute in China.” Journal of Oncology.” Journal of the National Cancer Institute 102 (2): Obstetrics and Gynaecology Research 38 (7): 1011–17. 82–88. Maman, A. F., A. Ouro-Bang’na, K. Tomta, S. Ahouangbévi, Grimes, C. E., J. A. Henry, J. Maraka, N. C. Mkandawire, and M. Chobli. 2005. “Deaths Associated with Anaesthesia and M. Cotton. 2014. “Cost-Effectiveness of Surgery in in Togo, West Africa.” Tropical Doctor 35: 220–22. Low- and Middle-Income Countries: A Systematic Review.” Ozgediz, D., M. Galukande, J. Mabweijano, S. Kijjambu, World Journal of Surgery 38 (1): 252–63. C. Mijumbi, and others. 2008. “The Neglect of the Global Hamid, S. A., S. M. Ahsan, and A. Begum. 2014. “Disease- Surgical Workforce: Experience and Evidence from Specific Impoverishment Impact of Out-of-Pocket Payments Uganda.” World Journal of Surgery 32 (6): 1208–15. for Health Care: Evidence from Rural Bangladesh.” Applied Pramesh, C. S., R. A. Badwe, B. B. Borthakur, M. Chandra, E. H. Raj, Health Economics and Health Policy 12 (4): 421–33. and others. 2014. “Delivery of Affordable and Equitable Hansen, D., S. C. Gausi, and M. Merikebu. 2000. “Anaesthesia Cancer Care in India.” The Lancet Oncology 15: E223–33. in Malawi: Complications and Deaths.” Tropical Doctor Pramesh, C. S., R. A. Badwe, and R. K. Sinha. 2014. “The 30: 146–49. National Cancer Grid of India.” Indian Journal of Medical Haynes, A. B., T. G. Weiser, W. R. Berry, S. R. Lipsitz, and Paediatric Oncology 35: 226–27. A.-H. S. Breizat, and others. 2009. “A Surgical Safety Purushotham, A. D., G. Lewison, and R. Sullivan. 2012. “The Checklist to Reduce Morbidity and Mortality in a Global State of Research and Development in Global Cancer Population.” New England Journal of Medicine 360 (5): Surgery.” Annals of Surgery 255 (3): 427–32. 491–99. Rose, J., D. C. Chang, T. G. Weiser, N. J. Kassebaum, and He, J., W. Shao, C. Cao, T. Yan, D. Wang, and others. 2011. S. W. Bickler. 2014. “The Role of Surgery in Global Health: “Long-Term Outcome and Cost-Effectiveness of Complete Analysis of United States Inpatient Procedure Frequency versus Assisted Video-Assisted Thoracic Surgery for Non- by Condition Using the Global Burden of Disease 2010 Small Cell Lung Cancer.” Journal of Surgical Oncology Framework.” Plos One 9 (2): E89693. doi:10.1371/Journal 104 (2): 162–68. .Pone.0089693. Henry, J. A., E. Frenkel, E. Borgstein, N. C. Mkandawire, and Sankaranarayanan, R., K. Ramadas, S. Thara, R. Muwonge, C. Goddia. 2014. “Surgical and Anaesthetic Capacity of J. Prabhakar, and others. 2011. “Clinical Breast Examination: Hospitals in Malawi: Key Insights.” Health Policy and Preliminary Results from a Cluster Randomized Controlled Planning doi:10.1093?heapol/czu102. Trial in India.” Journal of the National Cancer Institute Howitt, P., A. Darzi, G.-Z. Yang, H. Ashrafian, R. Atun, and 103 (19): 1476–80. others. 2012. “Technologies for Global Health.” The Lancet Tan, C., L. Peng, X. Zeng, J. Li, X. Wan, and others. 2013. 380 (9840): 507–35. “Economic Evaluation of First-Line Adjuvant Chemother- Hoyler, M., S. R. Finlayson, C. D. McClain, J. G. Meara, and apies for Resectable Gastric Cancer Patients in China.” Plos L. Hagander. 2014. “Shortage of Doctors, Shortage of Data: One 8 (12): e83396. A Review of the Global Surgery, Obstetrics, and Anesthesia Thorat, M. A., A. Rangole, M. S. Nadkarni, V. Parmar, and Workforce Literature.” World Journal of Surgery 38 (2): R. A. Badwe. 2008. “Revision Surgery for Breast Cancer.” 269–80. Cancer 113 (Suppl. 8): 2347–52. Ilbawi, A. M., E. M. Einterz, and D. Nkusu. 2013. “Obstacles Ukwenya, A. Y., L. M. Yusufu, P. T. Nmadu, E. S. Garba, and to Surgical Services in a Rural Cameroonian District A. Ahmed. 2008. “Delayed Treatment of Symptomatic Hospital.” World Journal of Surgery 37 (6): 1208–15. Breast Cancer: The Experience from Kaduna, Nigeria.” Jamison, D. T., L. H. Summers, G. Alleyne, K. J. Arrow, South African Journal of Surgery 46: 106–10. S. Berkley, and others. 2013. “Global Health 2035: A World Walker, I. A., and I. H. Wilson. 2008. “Anaesthesia in Developing Converging within a Generation.” The Lancet 382 (9908): Countries: A Risk for Patients.” The Lancet 371 (9617): 1898–955. 968–69. Kingham, P., O. I. Alatise, V. Vanderpuye, C. Casper, Weiser, T. G., S. E. Regenbogen, K. D. Thompson, A. B. Haynes, F. A. Abantanga, and others. 2013. “Treatment of Cancer S. R. Lipsitz, and others. 2008. “An Estimation of the in Sub-Saharan Africa.” The Lancet Oncology 14: E158–67. Global Volume of Surgery: A Modelling Strategy Based on Lan, N. H., W. Laohasiriwong, J. F. Stewart, N. Dinh Tung, and Available Data.” The Lancet 372 (9633): 139–44. P. C. Coyte. 2013. “Cost of Treatment for Breast Cancer in WHO (World Health Organization). 2003. Surgical Care at the Central Vietnam.” Global Health Action 6: 18872. District Hospital. Geneva: WHO. Lebrun, D. G., S. Chackungal, T. E. Chao, L. M. Knowlton, Yip, C. H., and B. O. Anderson. 2007. “The Breast Health Global A. F. Linden, and others. 2014. “Prioritizing Essential Surgery Initiative: Clinical Practice Guidelines for Management and Safe Anesthesia for the Post-2015 Development Agenda: of Breast Cancer in Low- and Middle-Income Countries.” Operative Capacities of 78 District Hospitals in 7 Low- and Expert Reviews in Anticancer Therapy 7 (8): 1095–104. Middle-Income Countries.” Surgery 155 (3): 365–73. Yip, C. H., E. Cazap, B. O. Anderson, K. L. Bright, M. Caleffi, Lu, Z., X. Yi, W. Feng, J. Ding, H. Xu, and others. 2012. and others. 2011. “Breast Cancer Management in Middle- “Cost-Benefit Analysis of Laparoscopic Surgery versus Resource Countries (MRCs): Consensus Statement from Laparotomy for Patients with Endometrioid Endometrial the Breast Health Global Initiative.” The Breast 20: S12–19. 238 Cancer Chapter 14 Radiation Therapy for Cancer David A. Jaffray and Mary K. Gospodarowicz INTRODUCTION during, or subsequent to the course of RT (Barton and others 2014). More than 14 million new cases of cancer are diagnosed Because radiation affects normal tissues and tumors, globally each year; radiation therapy (RT) has the achieving an acceptable therapeutic ratio—defined as potential to improve the rates of cure of 3.5 million the probability of tumor control versus the probability people and provide palliative relief for an additional of unacceptable toxicity—requires that the radiation 3.5 million people. These conservative estimates are dose be delivered within very tightly controlled toler- based on the fact that approximately 50 percent of all ances with less than 5 percent deviation. This controlled cancer patients can benefit from RT in the management production and precise application of radiation requires of their disease (Barton, Frommer, and Shafiq 2006; specialized equipment that is maintained and operated Barton and others 2014; Tyldesley and others 2011); by a team of trained personnel. The team includes, of these, approximately half present early enough to at a minimum, radiation oncologists to prescribe the pursue curative intent. appropriate dose, medical physicists to ensure accurate Soon after Roentgen’s discovery of X-rays in 1895, dose delivery, and radiation technologists to operate the ionizing radiation was applied to the treatment of can- equipment and guide patients through the radiation cer, with remarkable results. Carefully controlled doses process. Radiation oncologists work within multidisci- of ionizing radiation induce damage to the DNA in cells, plinary teams with medical and surgical oncologists to with preferential effects on cancer cells compared with coordinate a multidisciplinary approach to the manage- normal tissues, providing treatment benefits in most ment of cancer. A comprehensive cancer center provides types of cancer and saving lives. the full scope of RT services, ranging from externally RT is now recognized as an essential element of an applied beams of X-rays to the placement of radia- effective cancer care program throughout the world, tion-emitting sources within tumors (see chapter 11 in regardless of countries’ economic status. RT is used this volume [Gospodarowicz and others 2015]). to cure cancers that are localized; it also can provide RT is one of the more cost-effective cancer treat- local control—complete response with no recurrence ment modalities, despite the need for substantial capital in the treated area—or symptom relief in cancers that investment in the facilities and equipment. Concerns are locally advanced or disseminated (Gunderson and about the initial investment, however, have resulted in Tepper 2012). It is frequently used in combination with severely limited access in most low- and middle-income surgery, either preoperatively or postoperatively, as well countries (LMICs). Increasing the supply of RT services as in combination with systemic chemotherapy before, is critical to expanding effective cancer treatment in Corresponding author: David A. Jaffray, University of Toronto, Princess Margaret Cancer Centre, and TECHNA Institute, David.Jaffray@rmp.uhn.on.ca. 239 these settings and improving equity in access (Abdel- to escalate the radiation dose, with resulting improved Wahab and others 2013; Fisher and others 2014; Goss outcomes and reduced toxicity. Now 3D CRT is the and others 2013; Jaffray and Gospodarowicz 2014; standard approach in most countries. However, in some Rodin and others 2014; Rosenblatt and others 2013). low-income countries, the introduction of basic 2D radiotherapy would still save many lives and reduce suf- fering in thousands of patients with advanced cancers. USES OF RADIATION THERAPY The use of high-dose RT has been limited by RT is an essential element of curative treatment of can- the dose delivered to adjacent normal tissues, espe- cers of the breast, prostate, cervix, head and neck, lung, cially those areas with limited radiation tolerance, and brain, as well as sarcomas. The first four cancers are called critical normal structures. Continued progress common in LMICs (Barton and others 2014; Delaney, in computerization of RT planning and delivery allows Jacob, and Barton 2005b; Engstrom and others 2010; shaping the radiation field to deposit higher doses to Gregoire and others 2010; Petrelli and others 2014; tumors and further sparing the surrounding normal Pfister and others 2013; Ramos, Benavente, and Giralt tissues. These newer techniques—intensity modulated 2010; Souchon and others 2009; Tyldesley and others radiation therapy (IMRT) and stereotactic RT—allow 2011). RT is also used extensively in the management a therapeutic dose of RT to be delivered in a few high- of prostate cancer (Delaney, Jacob, and Barton 2005a; dose treatments and result in a higher probability of Tyldesley and others 2011). tumor eradication; they have been successfully applied Patients with hematologic malignancies are primar- in the management of brain metastasis and lung, ily treated with chemotherapy, but they also access RT bone, and paraspinal tumors. IMRT is being gradually resources (Barton and others 2014). Total body irradia- introduced in many centers and is the preferred treat- tion is used in the treatment of leukemia in the context ment for cancers of the prostate, as well as, head and of bone marrow transplantation. Localized RT is applied neck, where it has been shown to improve outcomes in many lymphomas to optimize local disease control significantly. and cure; palliative RT is extremely useful in multiple myeloma and lymphomas. RT is increasingly used to Concurrent Chemotherapy and Radiation Therapy control selected metastases. In short, RT both saves lives and alleviates suffering associated with cancer. The use of concurrent chemotherapy and RT has sig- nificantly improved tumor eradication and survival in several cancers. It may improve local control, result in Radiation Therapy Alone organ preservation, and eradicate distant microscopic RT as the sole therapy is used in the treatment of local- metastases. This combination therapy has proven ben- ized tumors, such as early-stage cancer of the larynx or eficial in treating cancers of the lung, cervix, head and prostate; non-melanoma skin cancer; head and neck neck, vulva, and anal canal (Benson and others 2012; cancers; and radiosensitive tumor types, such as semi- Chen and others 2013; Glynne-Jones and Renehan 2012; noma and lymphomas (Hoppe and others 2012; Motzer Gregoire and others 2010; Koh and others 2013; Petrelli and others 2009). In more advanced disease stages, RT and others 2014). is used before, during, or after surgery and is frequently combined with chemotherapy, either as concurrent or adjuvant treatment. Radiation Therapy as Adjuvant Treatment Prior to the development of sophisticated comput- RT is commonly used as adjuvant treatment following erized treatment planning systems, RT was planned surgery, especially in the case of incomplete resection. using clinical information and conventional X-rays Postoperative radiation is commonly used in cancers (2D RT) for field placement verification. This approach of the head and neck, rectum, breast, and lung, as well resulted in the use of large radiotherapy fields that as soft tissue sarcomas (Gunderson and Tepper 2012). assured coverage of the tumor, but also resulted in RT is also used after chemotherapy as the mainstay of limiting toxicity. With the introduction of computerized treatment when chemotherapy alone was not expected tomography (CT) scanners and computerized treatment to result in cure, such as for locally advanced breast planning, fields were shaped (3D conformal radiation cancer or bladder cancer, or as adjuvant treatment to therapy, 3D CRT) to correspond to the tumors; the use potentially curative chemotherapy, such as for Hodgkin of smaller fields resulted in less toxicity and the ability and non-Hodgkin lymphomas. 240 Cancer Radiation Therapy in Metastatic Disease several days, or high-dose rate brachytherapy, where the RT is beneficial in providing palliation to patients with single dose of radiation is delivered within minutes. metastatic disease. It is highly effective in controlling Radioisotope therapy may be applied in the radio- bleeding and pain, as well as the symptoms result- therapy department or in the nuclear medicine depart- ing from compression of the nerves, spinal cord, or ment. The most common application of radioisotope airways. The use of RT for pain relief is particularly therapy is in the treatment of thyroid cancer using valuable; a single moderate dose (8–10 Gy) achieves radioactive iodine or in the palliation of pain from bone significant pain relief in 60–80 percent of patients. metastasis using a radioactive isotope of strontium. This benefit is of particular importance in LMICs, Less common indications employ a conjugated radio- where many patients present with advanced and isotope such as lutetium (177Lu) DOTA-TATE to target metastatic disease. somatostatin-expressing neuroendocrine tumors. DELIVERING RADIATION THERAPY Facilities RT is delivered in three ways: RT is delivered in a specially designed facility that contains specialized equipment for imaging, treatment • External beam radiation therapy: applied externally planning, and radiation delivery. Modern RT depart- through directed beams of radiation to treat the can- ments are designed to optimize patient flow through the cer deep within the body. process and contain the following elements: • Brachytherapy: applied through the insertion of radiation-emitting sources directly within the tumor • Waiting areas or adjacent body cavity. • Examination rooms • Radioisotope therapy: applied through the systemic • Imaging suites with simulators/CT-simulators injection of a radioisotope that has been designed to • Computer planning workrooms target disease. • Shielded treatment rooms for linear accelerators or 60 Co treatment units Externally applied radiation beams can be produced • Shielded high-dose rate brachytherapy suites. by several approaches: radioactive sources, such as cobalt-60, that emit gamma rays; high-energy X-rays Additional support space is required for a physics or photons produced by linear accelerators; or particle testing laboratory, equipment storage, and dedicated beams—electrons, protons, or heavier ions—accelerated environmentally controlled computer server rooms. by other types of accelerators. These machines are External beam RT is delivered using machines that equipped with accessories that are able to shape dynam- produce high-energy X-ray or electron beams. The ically the radiation beam according to beam direction, two main types of photon beams are 60Co machines as well as onboard imaging devices that can verify the or X-ray-generating linear accelerators. Cobalt units accuracy of treatment delivery. Linear accelerators are contain radioactive cobalt sources in the head of the currently the backbone of external beam RT; multiple unit that emit photons with a mean energy of 1.25 MeV. companies manufacture the technologies, offering a The source is constantly emitting and requires con- range of high-energy X-rays (4–25 MV) to enable treat- stant radiation protection; it decays gradually and ment of deep-seated tumors. requires replacement every three to five years. Linear Brachytherapy involves either temporarily or per- accelerators use electric power to generate an electron manently placing radiation-emitting sources directly beam that is accelerated to produce a high-energy within tissues or body cavities. Permanent sources photon beam. Linear accelerators require a stable decay rapidly, depositing the dose and remaining in power supply for reliable operation. Both units have the body; temporary placement uses higher-activity collimators and filters to shape the radiation beam, sources that are electromechanically guided to tumors including multileaf collimators that allow motorized within preplaced interstitial or intracavitary catheters. shaping and/or modulation of the beam shape and The source and applicators are removed after delivery intensity during treatment delivery, thereby produc- of the prescribed dose of radiation. These removable ing more conformal irradiation of the target tissues radiation sources can provide either low-dose rate bra- while minimizing normal tissue exposure. In the past chytherapy, where the source remains in the tissues for 10 years, X-ray and CT imaging capabilities have been Radiation Therapy for Cancer 241 added to these machines to allow therapists to guide information is documented in the RT chart or elec- the placement of the radiation with increased preci- tronic medical record. Images of the part of the body sion and accuracy. to be treated are obtained and stored. • Treatment plan. Once the set-up and imaging are complete, the radiation oncologist outlines the tissues Personnel that must be irradiated on images and a radiation RT requires a specially trained team of professionals technologist/dosimetrist or a medical physicist devel- that includes radiation oncologists to prescribe the dose; ops the treatment plan, using specialized planning medical physicists, trained to commission and maintain software that models the placement of radiation the equipment and develop treatment plans; radiation beams and the dose contributed by each beam to technologists or therapists to operate the treatment ensure that the prescribed dose is delivered to the units; and nurses experienced in managing patients disease, while the dose to other tissues is minimized, undergoing therapy. Biomedical engineers and com- especially critical and particularly sensitive organs. puter or information technology experts complement The individualized treatment plan is independently the team. verified, and the total dose is delivered through a Once a decision to treat a patient has been made, series of treatments (fractions) in a prearranged the team develops a treatment plan and proceeds with schedule of sessions, usually daily over several weeks, delivery. The plan is based on accepted clinical guidelines as specified in the prescription. that describe the indications for RT; the target tissues to • Treatment delivery. Once the treatment plan is devel- be irradiated; the dose and fractionation prescriptions; oped by a medical physicist and dosimetrist and support for patients during treatment; and management reviewed and approved by a radiation oncologist, the of patients after treatment, including acute and late treatment can begin. In each session, the patient is complications of treatment. positioned exactly as during the simulation. After ver- The safe and effective management of the RT system ifying the prescription, treatment plan, and patient’s requires a high level of communication and coordi- position, the radiation dose is delivered. Treatments nation of the processes and systems employed in the are frequently given five days per week; in curative prescription, design, and delivery of radiation. Local, settings, they may continue for four to six weeks. national, and international bodies provide regulations Daily treatments are commonly delivered during a and guidelines for radiation safety, dose calibration, and session lasting 10–20 minutes. quality assurance of devices, clinical practice, and moni- toring of compliance. In specific circumstances, RT is applied in a shorter schedule consisting of one to three high-dose fractions. These hypofractionated treatments can be applied with Process generous margins for symptom relief for palliation rather The process refers to all the steps from the decision to than local disease control. Alternatively, they can be treat a patient with radiation to the completion of the applied for curative intent, using high-precision (also course of radiation treatment. called stereotactic) methods, wherein the targeted volume is very small and surrounding normal tissues are avoided. • Prescription. The first step is completion of the radi- During each session, specific verification steps are ation prescription, which indicates the exact part of taken before the dose is applied. During the course of the body to be treated, as well as the dose/fraction- RT, the patient is monitored daily by technologists and at ation schedule, including the total radiation dose to least weekly by a physician; patients with acute side effects be delivered in how many fractions, at what intervals, receive supportive care, as needed. The radiation records and in what overall time period. are kept for decades and made available for review in case • Planning. The second step is initiation of the plan- further RT or other interventions, such as surgery, are ning process. Patients are positioned on an X-ray planned for the previously irradiated part of the body. imaging machine that simulates the geometry of the treatment machine, or in more modern settings, on a specially adapted CT scanner (CT simulator). Safe and Effective Operation A desired position is determined (supine, prone, arms The staff processes and equipment need to be well man- up or by the side of the body); if needed, the patient aged to ensure safe and effective care that adheres to is immobilized with a specifically designed device to best practices and evidence-based medicine. Specially secure the reproducibility of the position. The set-up trained and certified personnel are essential for safe 242 Cancer and effective treatments, as well as safe operation of systems that work to maximize the therapeutic ratio for the facility. The medical specialization requires a res- each patient. This evolution has resulted in significant idency in radiation oncology to learn evidence-based increases in the complexity of the treatment, which is practice, radiation biology, and the principles of characterized by hundreds of megabytes of treatment radiation physics. Typically, an experienced radiation data and detailed quality control activities to ensure that oncologist oversees the operations of the RT depart- the prescription is applied not only accurately, but also ment. The technological and treatment design activi- appropriately for each patient. In the interest of reduc- ties are supported by specially trained physicists, called ing costs and standardizing interventions, the field is medical physicists, with a degree in physics and addi- developing automated methods that allow high-quality tional training to acquire the specific skills required treatment plans to be designed in a few minutes. These to practice RT. Trained technologists interact with approaches promise to “bury the complexity” of the patients and operate the treatment machines to deliver current RT process, while still providing a high degree of the radiation doses. Dedicated education programs safety and personalization (Jaffray 2012). have been developed to train these staff members in a The adoption of expert systems and machine learning range of topics, including patient care, technology, and methods allows the treatment team to design and deliver radiation physics. highly personalized RT (Purdie and others 2014). This The operational team of the department has several degree of automation provides a valuable form of peer key responsibilities: review that is inexpensive and can learn from experts around the world by drawing on the clinical expertise • Ensuring that the radiation systems are safe for that has gone into large databases of existing treat- patients, the public, and staff members ment plans. The emergence of cloud-based treatment • Ensuring that the radiation equipment is appropri- planning and peer review is likely to fuse with modern ately calibrated, tested, and maintained telemedicine approaches to create more efficient delivery • Ensuring that the each patient receives appropriate and learning platforms. An additional advantage of these care through peer review of the treatment plan and cost-saving methods is that they require a standardiza- independent checks of the calculations tion in the nomenclature used to describe the treatment • Monitoring and responding to errors or variations in intent and treatment record—a benefit that is highly the delivery of care. synergistic with the adoption of medical and bioinfor- matics efforts that promise to advance clinical practice Depending on the local, national, and international (Lambin and others 2013). context, these activities may need to comply with regulations. EQUITABLE ACCESS TO RADIATION THERAPY Integration into Cancer Centers The World Health Organization recommends that all RT departments collaborate closely with departments countries develop and implement a population-based of pathology and laboratory medicine, diagnostic imag- cancer control plan. These plans are based on the infor- ing, surgery or surgical oncology, medical oncology, mation provided by cancer registries and include plans and palliative care to ensure that treatment plans are for prevention; screening and early detection; timely created based on correct diagnosis, full assessment of access to high-quality treatment, including surgery, disease extent (stage), and the medical condition of the radiotherapy, and chemotherapy; and palliative and patient. Modern clinical practice ensures the physical supportive care. and operational infrastructure is in place to allow multi- Planning RT resource provision requires detailed disciplinary cancer care. This infrastructure may include knowledge of the patterns of cancer, including differ- multidisciplinary clinics and conferences where the ent disease entities and distribution by stage. National management of the patient is discussed with all appro- cancer plans should define the number of departments priate experts—for example, oncologists, pathologists, and treatment machines that are appropriate for the and radiologists—and the amalgamation of medical current and projected cancer burden. The distribution records to facilitate communication and coordination of cancer facilities needs to consider not only the burden, of care. but also the geographic distribution of the population to RT has evolved from the direct application of a single facilitate access. beam of ionizing radiation to a cancerous lesion to image- Requisite elements of effective RT include medi- guided, computer-optimized, robotically controlled cal and professional education, training programs for Radiation Therapy for Cancer 243 support staff, and ongoing refreshment of equipment building. These innovations need to come from the and infrastructure. Specific elements that need attention technological, educational, operational, and clinical include the following: practice domains to avoid unnecessary suffering and loss of human life. • Medical education system. The training of radiation oncologists, medical physicists, and radiation ther- apists is a critical element. Without this foundation, Efforts to Address the Equity Gap shortages of professionals will lead to long waiting Ample evidence indicates severe gaps in access to RT lists, treatment delays, and compromised outcomes. In in large areas of the world. The International Atomic addition, the lack of local training programs prevents Energy Agency (IAEA) maintains a directory of all RT the establishment of a stable supply of staff to operate facilities (http://www-naweb.iaea.org/nahu/dirac/). the facilities. This lack is not only a challenge during Significant inequity exists in access to RT across the initiation of a program; it will persist as cancer services world. Map 14.1 shows one descriptor. By comparison, are ramped up to reach the level of appropriate use. access rates in high-income countries would corre- • Regulatory structure. The presence of external spond to approximately 100,000 people served by one accreditation and regulation frameworks helps to radiation treatment machine. standardize the operation of RT departments and IAEA has brought attention to the lack of ade- secure high-quality practice. Establishing these quate RT resources for several decades. Comprehensive frameworks can be particularly challenging in reviews of the resources in Europe, Latin America and resource-constrained economies, where infrastruc- the Caribbean, and Sub-Saharan Africa describe the ture is limited and political stability is an issue. limitations in centers, equipment, and staff. One pub- • Societal infrastructure. Limitations in access to a reli- lication on cancer in Sub-Saharan Africa stated that 29 able supply of electric power, climate control, service of 52 countries have no RT facilities; those that have infrastructure, and complex procurement settings facilities face severe shortages. Less than 10 percent of affected by such factors as political instability and the population in the region has access (Zubizarreta and transportation are problematic. others 2015). The barriers to the implementation of RT are Innovative approaches need to be pursued to numerous. They include perceptions that it is expen- address the numerous challenges that impede capacity sive, complex, and unlikely to succeed because of the Map 14.1 Number of People Served by One Radiotherapy Unit IBRD 41687 | JULY 2015 Number of People Served by One Radiotherapy Unit Data from International Atomic Energy Agency – DIRAC database, 07/2013 Below 500,000 500,000–1 million 1–2 million 2–5 million Over 5 million No unit No data Source: Based on data from the Directory of Radiotherapy Centres (DIRAC) database of radiation therapy equipment, International Atomic Energy Agency, Vienna, http://www-naweb.iaea.org/nahu/dirac/. 244 Cancer shortage of qualified personnel and funding. With many an investment. Real effort needs to be put into calcu- competing demands for cancer control activities, there lating the true cost and the resultant benefits of RT so is a risk that the appropriate investment in RT may not that decision makers can make informed choices. Such be made, leaving countries and patients to wrestle with approaches have been applied in advancing the global dysfunctional cancer services. HIV/AIDS effort and are being pursued by the Union IAEA has provided technical assistance, training and for International Cancer Control Global Task Force on education, and financing for equipment. Unfortunately, Radiotherapy for Cancer Control (http://www.gtfrcc these efforts have not resolved the severe limitations .org). Such approaches are the key to articulating the in access. The IAEA Programme of Action for Cancer importance and value of financial investments in can- Therapy, established in 2004 (http://cancer.iaea.org/), cer control. Moreover, these approaches immediately organized a large number of missions to assess the lead to the development of novel financing schemes to readiness of a country to develop new RT facilities. These overcome the reluctance to commit the funds for the missions assess all aspects of cancer control, since the capital investment required to improve access globally potential benefit offered by RT can be realized only in (chapter 17 in this volume [Knaul and others 2015]). the presence of adequate diagnostic facilities, surgery, chemotherapy, and supportive and palliative care. IAEA can advise governments on the optimal ways to proceed, CONCLUSIONS but the implementation depends on the political will Cancer is projected to become the number one cause of and resources devoted to cancer control. death across the globe in the next 20 years. The evidence Effective cancer planning has improved access in demonstrates that more than 40 percent of patients a number of areas, including Brazil; Ireland; Ontario, with cancer would benefit from RT; the lack of access Canada; and Poland (Chalubinska-Fendler and others will compromise the care of millions of people suffering 2014). Overall, however, such progress is lacking in from cancer if not addressed through immediate action. LMICs, and international partnerships and assistance The global community has been working hard to ensure are needed to accelerate progress to close the access quality through standardization in RT practices and gap. The U.S.-based AMPATH Program is building a provide guidance in the establishment of new treat- new cancer center in Eldoret, Kenya, and has included ment capacity (IAEA 2008). It is now critical that RT plans to implement RT as soon as possible (http:// be acknowledged as an essential element of an effective www. ampathkenya.org/our-programs/primary-care cancer control plan—and that the critical equipment, -chronic-diseases/oncology/). operations, and educational investments be provided In Latin America and the Caribbean, a unique to ensure that RT is in place to respond to the growing network of national cancer institutes has embarked on cancer burden. an initiative to improve the quality of RT in the region (http://www2.rinc-unasur.org/wps/wcm/connect/rinc /site/home). The Network of National Cancer Institutions NOTE of Latin America (RINC) initiative draws together World Bank income classifications as of July 2014 are as 18 countries to organize a regional community of best follows, based on estimates of gross national income per capita practices; exchange information and knowledge; iden- for 2013: tify needs, opportunities, and common interests; foster coordination among member countries; and promote • Low-income countries (LICs): US$1,045 or less the commitment of every country’s corresponding levels • Middle-income countries are subdivided: of government, with emphasis on the availability of the a) Lower-middle-income: US$1,046–US$4,125 financial, human, and legislative resources necessary for b) Upper-middle-income (UMICs): the development of cancer control. US$4,126–US$12,745 Ongoing efforts in India and Turkey offer prom- • High-income countries (HICs): US$12,746 or more. ise, but to date fall short of addressing the limited access to effective RT for their populations (Banerjee, Mahantshetty, and Shrivastava 2014; Goksel and REFERENCES others 2011). Abdel-Wahab, M., J. M. Bourque, Y. Pynda, J. Izewska, D. Van Although ample data describe the benefits of RT der Merwe, and others. 2013. “Status of Radiotherapy for cancer control, the cost of equipment and develop- Resources in Africa: An International Atomic Energy ment of skills seem an overwhelming challenge. This Agency Analysis.” The Lancet Oncology 14 (4): e168–75. does not need to be the case; any return begins with doi:10.1016/s1470-2045(12)70532-6. Radiation Therapy for Cancer 245 Banerjee, S., U. Mahantshetty, and S. Shrivastava. 2014. Goss, P. E., B. L. Lee, T. Badovinac-Crnjevic, K. Strasser- “Brachytherapy in India: A Long Road Ahead.” Journal of Weippl, Y. Chavarri-Guerra, and others. 2013. “Planning Contemporary Brachytherapy 6 (3): 331–35. Cancer Control in Latin America and the Caribbean.” Barton, M. B., M. Frommer, and J. Shafiq. 2006. “Role of The Lancet Oncology 14 (5): 391–436. doi:10.1016 Radiotherapy in Cancer Control in Low-Income and /s1470-2045(13)70048-2. Middle-Income Countries.” The Lancet Oncology 7 (7): Gregoire, V., J. L. Lefebvre, L. Licitra, E. Felip, and Ehns-Esmo- 584–95. doi:10.1016/s1470-2045(06)70759-8. Estro Guidelines Working Group. 2010. “Squamous Cell Barton, M. B., S. Jacob, J. Shafiq, K. Wong, S. R. Thompson, and Carcinoma of the Head and Neck: EHNS-ESMO-ESTRO others. 2014. “Estimating the Demand for Radiotherapy Clinical Practice Guidelines for Diagnosis, Treatment and from the Evidence: A Review of Changes from 2003 to 2012.” Follow-Up.” Annals of Oncology 21 (Suppl. 5): v184–86. Radiotherapy & Oncology 112 (1): 140–44. doi:10.1016/j doi:10.1093/annonc/mdq185. .radonc.2014.03.024. Epub May 12. Gunderson, L. L., and J. E. Tepper. 2012. Clinical Radiation Benson, A. B., 3rd, J. P. Arnoletti, T. Bekaii-Saab, E. Chan, Oncology. 3rd ed. Philadelphia, PA: Saunders; London: Y. J. Chen, and others. 2012. “Anal Carcinoma, Version Elsevier. 2.2012: Featured Updates to the NCCN Guidelines.” Hoppe, R. T., R. H. Advani, W. Z. Ai, R. F. Ambinder, P. Aoun, Journal of the National Comprehensive Cancer Network and others. 2012. “Hodgkin Lymphoma, Version 2.2012: 10 (4): 449–54. Featured Updates to the NCCN Guidelines.” Journal of the Chalubinska-Fendler, J., W. Fendler, J. Luniewska-Bury, National Comprehensive Cancer Network 10 (5): 589–97. W. Mlynarski, M. Spych, and others. 2014. “Tackling the IAEA (International Atomic Energy Agency). 2008. Setting Turmoil of Transformation: Radiation Oncology in Poland.” Up a Radiotherapy Programme: Clinical, Medical Physics, International Journal of Radiation Oncology, Biology, Physics Radiation Protection and Safety Aspects. Vienna: IAEA. 90 (3): 480–86. doi:10.1016/j.ijrobp.2014.05.031. Jaffray, D. A. 2012. “Image-Guided Radiotherapy: From Chen, R. C., W. U. Shipley, J. A. Efstathiou, and A. L. Zietman. Current Concept to Future Perspectives.” Nature Reviews 2013. “Trimodality Bladder Preservation Therapy for Clinical Oncology 9 (12): 688–99. doi:10.1038/nrclinonc Muscle-Invasive Bladder Cancer.” Journal of the National .2012.194. Comprehensive Cancer Network 11 (8): 952–60. Jaffray, D. A., and M. Gospodarowicz. 2014. “Bringing Delaney, G., S. Jacob, and M. Barton. 2005a. “Estimating the Global Access to Radiation Therapy: Time for a Change Optimal External-Beam Radiotherapy Utilization Rate in Approach.” International Journal of Radiation for Genitourinary Malignancies.” Cancer 103 (3): 462–73. Oncology, Biology, Physics 89 (3): 446–47. doi:10.1016/j doi:10.1002/cncr.20789. .ijrobp.2014.05.019. ———. 2005b. “Estimation of an Optimal External Beam Knaul, F., S. Horton, P. Yerramilli, H. Gelband, and R. Atun. Radiotherapy Utilization Rate for Head and Neck 2015. “Financing Cancer Care in Low-Resource Settings.” In Carcinoma.” Cancer 103 (11): 2216–27. doi:10.1002 Disease Control Priorities (third edition): Volume 3, Cancer, /cncr.21084. edited by H. Gelband, P. Jha, R. Sankaranarayanan, and Engstrom, P. F., J. P. Arnoletti, A. B. Benson, 3rd, J. D. Berlin, S. Horton. Washington, DC: World Bank. J. M. Berry, and others. 2010. “NCCN Clinical Practice Koh, W. J., B. E. Greer, N. R. Abu-Rustum, S. M. Apte, Guidelines in Oncology: Anal Carcinoma.” Journal of the S. M. Campos, and others. 2013. “Cervical Cancer.” Journal National Comprehensive Cancer Network 8 (1): 106–20. of the National Comprehensive Cancer Network 11 (3): Fisher, B. J., L. C. Daugherty, J. P. Einck, G. Suneja, M. M. Shah, 320–43. and others. 2014. “Radiation Oncology in Africa: Improving Lambin, P., R. G. Van Stiphout, M. H. Starmans, E. Rios- Access to Cancer Care on the African Continent.” Velazquez, G. Nalbantov, and others. 2013. “Predicting International Journal of Radiation Oncology, Biology, Physics Outcomes in Radiation Oncology: Multifactorial Decision 89 (3): 458–61. doi:10.1016/j.ijrobp.2013.12.032. Support Systems.” Nature Reviews Clinical Oncology 10 (1): Glynne-Jones, R., and A. Renehan. 2012. “Current Treatment 27–40. doi:10.1038/nrclinonc.2012.196. of Anal Squamous Cell Carcinoma.” Hematology/Oncology Motzer, R. J., N. Agarwal, C. Beard, G. B. Bolger, B. Boston, Clinics of North America 26 (6): 1315–50. doi:10.1016/j and others. 2009. “NCCN Clinical Practice Guidelines .hoc.2012.08.011. in Oncology: Testicular Cancer.” Journal of the National Goksel, F., O. Koc, N. Ozgul, M. Gultekin, M. Abacioglu, and Comprehensive Cancer Network 7 (6): 672–93. others. 2011. “Radiation Oncology Facilities in Turkey: Petrelli, F., A. De Stefani, F. Raspagliesi, D. Lorusso, Current Status and Future Perspectives.” Asian Pacific and S. Barni. 2014. “Radiotherapy with Concurrent Journal of Cancer Prevention 12: 2157–62. Cisplatin-Based Doublet or Weekly Cisplatin for Cervical Gospodarowicz, M., J. Trypuc, A. D’Cruz, J. Khader, S. Omar, Cancer: A Systematic Review and Meta-Analysis.” and F. Knaul. 2015. “Cancer Services and the Comprehensive Gynecologic Oncology 134 (1): 166–71. doi:10.1016/j. Cancer Center.” In Disease Control Priorities (third edi- ygyno.2014.04.049. tion): Volume 3, Cancer, edited by H. Gelband, P. Jha, Pfister, D. G., K. K. Ang, D. M. Brizel, B. A. Burtness, R. Sankaranarayanan, and S. Horton. Washington, DC: P. M. Busse, and others. 2013. “Head and Neck Cancers, World Bank. Version 2.2013. Featured Updates to the NCCN 246 Cancer Guidelines.” Journal of the National Comprehensive Cancer Centres (DIRAC) Database.” The Lancet Oncology 14 (2): Network 11 (8): 917–23. e79–86. doi:10.1016/s1470-2045(12)70556-9. Purdie, T. G., R. E. Dinniwell, A. Fyles, and M. B. Sharpe. Souchon, R., F. Wenz, F. Sedlmayer, W. Budach, J. Dunst, 2014. “Automation and Intensity Modulated Radiation and others. 2009. “DEGRO Practice Guidelines for Therapy for Individualized High-Quality Tangent Breast Palliative Radiotherapy of Metastatic Breast Cancer: Bone Treatment Plans.” International Journal of Radiation Metastases and Metastatic Spinal Cord Compression Oncology, Biology, Physics 90 (3): 688–95. doi:10.1016/j (MSCC).” Strahlentherapie und Onkologie 185 (7): 417–24. .ijrobp.2014.06.056. doi:10.1007/s00066-009-2044-2. Ramos, M., S. Benavente, and J. Giralt. 2010. “Management of Tyldesley, S., G. Delaney, F. Foroudi, L. Barbera, M. Kerba, Squamous Cell Carcinoma of the Head and Neck: Updated and others. 2011. “Estimating the Need for Radiotherapy European Treatment Recommendations.” Expert Review of for Patients with Prostate, Breast, and Lung Cancers: Anticancer Therapy 10 (3): 339–44. doi:10.1586/era.10.6. Verification of Model Estimates of Need with Radiotherapy Rodin, D., D. Jaffray, R. Atun, F. M. Knaul, M. Gospodarowicz, Utilization Data from British Columbia.” International and others. 2014. “The Need to Expand Global Access Journal of Radiation Oncology, Biology, Physics 79 (5): to Radiotherapy.” The Lancet Oncology 15 (4): 378–80. 1507–15. doi:10.1016/j.ijrobp.2009.12.070. doi:10.1016/s1470-2045(14)70121-4. Zubizarreta, E. H., E. Fidarova, B. Healy, and E. Rosenblatt. Rosenblatt, E., J. Izewska, Y. Anacak, Y. Pynda, P. Scalliet, 2015. “Need for Radiotherapy in Low and Middle Income and others. 2013. “Radiotherapy Capacity in European Countries: The Silent Crisis Continues.” Clinical Oncology Countries: An Analysis of the Directory of Radiotherapy 27 (2): 107–14. doi:10.1016/j.clon.2014.10.006. Radiation Therapy for Cancer 247 Need for National Commitments to Chapter 15 Cancer Research to Guide Public Health Investment and Practice Edward L. Trimble, Preetha Rajaraman, Ann Chao, Thomas Gross, Carol Levin, You-Lin Qiao, Timothy Rebbeck, Lisa Stevens, and Fang-hui Zhao INTRODUCTION World Bank criteria. The group issued its Call to Action, which articulates the rationale for supporting research This chapter is addressed primarily to potential funders to inform public policy, and sent the Call to Action of health research at the national, provincial, and state to the World Health Organization (WHO 2008). The levels. Health research in general, and cancer research Call to Action recommends that national governments in particular, is not a luxury reserved for high-income allocate at least 2 percent of the budgets of ministries of countries (HICs); it is a necessity for all countries across health to research and that international development the income spectrum. The extent and depth of that agencies invest at least 5 percent of development assis- research may vary by a country’s financial situation, tance funds earmarked for the health sector in research. and the topics may vary by a country’s specific burden Box 15.1 highlights many of the principles applicable to of cancers and associated risk factors. Nevertheless, cancer research. a comprehensive health research plan is foundational to Subsequent white papers and policy statements have the ability to allocate resources efficiently and effectively, made clear the importance of research on noncommu- develop human capacity and infrastructure, and identify nicable diseases and cancer to guide public policy and the appropriate technologies and medicines for health public investment (See Annotated Select Bibliography). and health services delivery. In short, robust research is essential to building evidence-based cancer prevention and control programs. STAKEHOLDERS AND THEIR KNOWLEDGE In recognition of the foundational nature of NEEDS IN CANCER RESEARCH research, ministers and representatives of ministries of health, science and technology, agriculture, edu- There are many stakeholders in cancer prevention and cation, foreign affairs, and international cooperation control, including researchers, ministers of health, phy- from 53 countries convened in 2008 in Mali at the sicians, other providers, and patients. Bamako Global Ministerial Forum on Research in The most important stakeholders are the individuals Health. Of the 53 participating countries, 38 were low- with cancer, their families, patient advocates, and those and middle-income countries (LMICs), according to at risk of cancer. Ministers of health and their colleagues Corresponding author: Edward L. Trimble, MD, MPH, trimblet@nih.org 249 RESEARCH PRIORITIES IN CANCER Box 15.1 Health Surveillance and Cancer Surveillance Making cancer surveillance an integral part of public Highlights of the 2008 Bamako Call to health surveillance, which collects information on other Action for Research on Health risk factors and diseases, will greatly facilitate efforts to improve cancer outcomes. More than 20 percent • Adopting comprehensive approaches. Ensure of cancers in LMICs, for example, are associated that research and innovations are interdisci- with chronic viral, bacterial, or helminthic infection plinary and intersectoral; engage the public (De Martel and others 2012). Comprehensive cancer sector, private sector, and civil society associ- surveillance, accordingly, requires surveillance of rele- ations in collaborations. vant preventive practices, including vaccinations, that • Setting priorities. Develop the global can influence the incidence and prevalence of infections research agenda in light of national and linked to cancer. Similarly, cancer and other noncom- regional priorities, and encourage national municable diseases, such as cardiovascular disease, dia- governments to make the development of betes, and chronic obstructive pulmonary disease, share policies for health research and innovation a number of common risk factors, such as tobacco use, a priority. obesity, poor diet, physical inactivity, alcohol consump- • Building capacity. Improve capacity at all tion, and environmental pollution. Surveillance for these relevant levels to foster research and tech- common risk factors is critical to effective cancer control. nology transfer, improve the education and The Institute for Health Metrics and Evaluation training of researchers, integrate research (2011) has developed a framework for integrating sur- into health systems, and establish systems to veillance systems across health information sources evaluate the impact of research. to help decision makers allocate resources and evalu- • Improving equity. Make greater equity a key ate interventions. The World Health Organization also element in the process. collects data on a variety of health indicators across Source: WHO 2008. diseases through its Global Health Observatory (WHO 2013). Table 15.1 sets out the range of health surveil- lance systems with relevance for cancer prevention and control. Few countries can afford to collect such health surveillance data on 100 percent of their populations. are also important stakeholders, as are physicians and Nonetheless, through judicious use of cross-sectional other members of the health care team who provide surveys in representative populations, cohort studies, care on a daily basis. Additional ministries at the and disease and death registries, health policy makers national and state levels often play roles (albeit some of can draft, implement, evaluate, and modify cancer them minor), including the ministries responsible for control plans. finance, education, science and technology, agriculture, A national system that assigns unique individual energy, customs, and foreign affairs. Universities and identifying numbers or biometrics can be used routinely other academic units, as well as hospitals and clinics at all health system encounters and vital registrations in the public and private sectors, have major roles in of births and deaths to strengthen national health sur- research. veillance systems. Such a national identification system Nongovernmental organizations are also crucial can facilitate the linkage of medical records, including partners in fostering cancer research. Specialists from records from clinic visits and hospitalizations, immu- many disciplines of research and health are involved, nization records, pathology reports, operative notes, including physicians, nurses, pharmacists, psychologists, health insurance reimbursement information, and death social workers, epidemiologists, biostatisticians, basic registration. and translational research scientists, information tech- nologists, and data managers. Cancer Registries and Pathologic Diagnosis The needs of national stakeholders should guide can- Cancer Registries The most basic public health oncol- cer research in each country. One way to recognize those ogy question is what is the burden of cancer in a city, needs is by determining what ministries of health need a state, a country, or a region? The answer begins with to know (box 15.2) and what patients and their doctors pathology laboratories, where biological specimens for need to know (box 15.3). individuals—including blood tests, diagnostic biopsies, 250 Cancer Box 15.2 What Do Ministries of Health Need to Know? The answers to the following questions are relevant What resources are now in place for cancer control? to ministries of health from low-, middle-, and • What can be done in the context of the existing high-income countries. health care system to prevent cancer and other common diseases? What is the burden of cancer in the country? • What can be done to screen for and treat cancer? • How many cancers are diagnosed each year? • How can we help cancer survivors return to being • How many people die from cancer? productive members of society? • Which are the most common cancers? • What palliative care can we provide? • Which geographic regions and populations bear the greatest cancer burden? What could we do by redeploying existing • How does the burden of cancer compare with resources? What else should we be doing and what that of other diseases? will it cost? • What are the risk factors for cancer in the • What are the “best buys” for the country in the country? context of currently available health resources? and surgical specimens—are evaluated. To these data are added cancers diagnosed on the basis of imaging studies, Box 15.3 as well as cancers diagnosed on the basis of physical signs and patient symptoms. The data for individuals What Do Patients and Doctors Need to diagnosed with cancer can then be added up to give a picture of the overall cancer burden. Cancer registry data Know? can provide descriptive and trend information about the The answers to these questions are relevant for burden of cancer in a population and enable the for- patients, families, and health care providers in mation of hypotheses about etiology that can be tested low-, middle-, and high-income countries. in analytic studies. Cancer registry data also inform the need for cancer diagnosis and treatment facilities and • What type of cancer does the patient have? allow the evaluation of cancer control interventions in • What is the extent or stage of the cancer? a population. • What are the options for treating the cancer Obtaining such data at the country level is difficult and its symptoms? and expensive. The traditional approach has been to • How can the patient gain access to appropri- start small, with one hospital, then expand to a city ate and affordable cancer therapy, treatment or county, then to a state or province, and finally to of symptoms, survivorship counseling, and other representative or high-risk populations of interest. supportive care? Population-based registries may cover a representative • If the treatment works as hoped and the portion of the geographic region of interest from which patient is cured of cancer, what steps are the larger (for example, countrywide) cancer burden can needed to help the patient reintegrate into be estimated. Furthermore, population-based registries family and work life? can be used to evaluate community health interventions • If the cancer is too advanced for curative in the region. therapy or if treatment does not cure the In the United States, for example, the National patient, will the patient benefit from pallia- Cancer Institute’s Surveillance, Epidemiology, and End tive care, including pain control? Results (SEER) Program began to capture information • How can the patient best gain access to palli- on cancer incidence from 14 percent of the coun- ative care? try’s population in 1973. The scope was subsequently expanded to track additional areas with low-income Need for National Commitments to Cancer Research to Guide Public Health Investment and Practice 251 Table 15.1 Health Surveillance Systems for Noncommunicable Diseases and Cancer Control Planning Infections • Incidence and prevalence of infections linked to cancer, such as hepatitis B virus (HBV) and C, human papillomavirus (HPV), human immunodeficiency virus (HIV), human T-lymphotropic virus-1, Epstein-Barr virus, human herpesvirus, Helicobacter pylori, and liver flukes • Uptake of prophylactic HBV and HPV vaccines Common risk factors for noncommunicable diseases • Tobacco use (smoked and oral), including exposure to secondhand smoke • Lack of proper diets, such as those with more fruits and vegetables and whole grains; exposure to known carcinogens, such as nitrates and high-temperature beverages • Alcohol intake • Obesity • Low activity level Availability of, access to, and uptake of cancer screening • Preinvasive cervical cancer • Breast cancer • Colon cancer Cancer registries • Type and stage of cancer annotated with demographic data • Primary treatment and cancer outcome Death registries • Deaths occurring due to cancer or concurrent disease after diagnosis of cancer and minority populations; as of 2013, it included pathologists must be available to review the processed approximately 28 percent of the population. The SEER material, whether onsite, at central laboratories, or from Program collects high-quality, individual-level data on remote sites via telepathology. patient demographics, primary tumor site, morphol- When a person is diagnosed with cancer, it is ogy, stage at diagnosis, first course of treatment, and important to determine the aggressiveness of the spe- follow-up for vital status (Howlader and others 2013). cific cancer and whether the cancer has spread from the A complementary program established through the U.S. original site of origin to other parts of the body. This Centers for Disease Control and Prevention in 1992 has additional information is used to assign a stage to the expanded cancer registries to cover 96 percent of the U.S. cancer, which generally ranges from stage I (the earliest population. stage, which in many cases can be cured with standard therapy) to stage IV (the most advanced stage, which is Role of Laboratories and Anatomic Pathology Accuracy most difficult to treat effectively). The ability to assign in histopathologic diagnosis of tumor specimens from a stage to newly diagnosed cases requires linking the surgery or biopsy is required to make the correct diagno- pathology report to clinical data. Without data on the sis for any one person’s cancer and help that person and extent of the disease or stage, it is not possible to provide health care providers make the appropriate treatment appropriate treatment or determine the success of inter- decision. In addition, accuracy in histopathologic diag- ventions intended to diagnose cancer at earlier stages nosis is also required to ensure the accuracy of cancer when the cancer is more successfully treated. incidence data in cancer registries. Biological specimens must be processed promptly, shortly after removal from Biobanking Biobanks and biological resource centers the human body. Pathology laboratories require trained constitute key components of cancer research. To histotechnicians and cytotechnicians, as well as func- understand the biological basis of cancer; to develop tioning instrumentation and a reliable supply chain for biomarkers for cancer risk, early detection, and prog- the equipment needed to process specimens, such as nosis; and to determine the most appropriate can- formalin, glass slides, and diagnostic reagents. Trained cer treatment based on precise diagnosis of tumor 252 Cancer characteristics, it is necessary to have access to clinically cancer mortality rate, if cause-of-death data are available annotated biologic specimens of cancer and normal (Boyle and Levin 2008; Jensen and others 1991). tissue (Vaught, Henderson, and Compton 2012). Until A less costly and less precise alternative approach to recently, some analyses required special preparation of obtaining the medically certified cause of death is that specimens, such as fresh frozen tissue. More recent devel- of the verbal autopsy, in which trained health workers opments in molecular pathology permit many studies to interview the members of a household in which a death be done on formalin-fixed, paraffin-embedded tissue. has occurred about the symptoms of the deceased person To make progress in cancer research as quickly (Institute for Health Metrics and Evaluation 2011). as possible, it is important to be able to facilitate the Dikshit and others (2012) have reported the successful use collection and analysis of such specimens. Some speci- of the verbal autopsy in the Million Death Study in India mens may need to be shipped to global or regional core to estimate mortality from cancer and other diseases. laboratories for analysis using standardized protocols. In other cases, the primary analysis may be done in the country of origin, with a small number of samples Cancer Epidemiology exchanged among countries for standardization and The application of sound epidemiologic methods is quality control of the laboratory techniques. National indispensable in cancer research. Experimental and or state regulations that prohibit any shipment of observational studies have yielded much of the current specimens outside the country or region of origin may knowledge about causation, prevention, and interven- preclude efficient analysis of those specimens and delay tion; epidemiologic studies conducted using cancer progress in research needed for cancer control. registry data have made significant contributions to the Given the increasing need to pool data and bio- understanding of rates and trends. Cancer registries pro- specimens from consortia of studies around the vide descriptive data that reveal important patterns and world to achieve adequate sample size and statisti- trends in the burden of cancer in defined populations. cal power, countries with rigid rules for data- and Registry data help to generate hypotheses that guide biospecimen-sharing will be at a disadvantage in the epidemiologic investigations that can identify poten- ability to participate in cutting-edge cancer research tial causative factors, rule out false associations, define (NRC 2011; Thun, Hoover, and Hunter 2012). Although the nature of the dose-response relationship, identify it is clear that regulations must be in place for the appro- co-factors and, in some cases, identify explanations for priate use of all samples when shared outside national late-stage diagnosis. Increasingly, epidemiology studies boundaries, facilitating processes for the timely sharing incorporate molecular biology in their design to help of biological specimens will enhance research for all. better define outcome (cancer subtypes at a given site can vary greatly) and exposure with relevant biomarkers Linking Death Registration Systems to Cancer and to identify genetic and other molecular risk factors. Registration Systems Functioning national, regional, or sample-area death Case-Control Studies registration systems are critical to a country’s ability to A great deal of knowledge on cancer epidemiology is monitor its burden of all diseases. In areas covered by generated by case-control studies, which identify cancer cancer registries, accurate death registration information cases using cancer registries or hospitals and other may serve as an important source of cancer-case finding. points of care and sample controls from the source In general, cancer registrars routinely search hospital population of cases. These epidemiologic investigations medical records for the initial diagnosis and pathology may require rapid case finding by study personnel, report of medically certified cancer deaths. in-depth interviews of cases, and controls by trained Linking information on individuals diagnosed with interviewers to assess exposure information, environ- cancer to death registries also greatly facilitates the mental sampling, and collection and analysis of bio- computation of rates of cancer survival (for example, specimens at core laboratories. Case-control studies are case-fatality rates) by tumor site and stage of disease, efficient and generally less costly than cohort studies, which otherwise would require expensive and time- particularly in the study of rare outcomes (Rothman, consuming active follow-up of individuals diagnosed Greenland, and Lash 2008). Challenges include the with cancer. As noted, the use of unique individual potential for biases, such as differential recall, and the patient identifiers can help to link the diagnosis of challenge of measuring exposures or assessing biomark- cancer with patient follow-up and, ultimately, the death ers of environmental exposure before the time of cancer of that individual. In the absence of cancer registries, diagnosis, thereby making it difficult to assess tempo- the cancer burden of a country can be estimated by the rality (Wild 2009). Need for National Commitments to Cancer Research to Guide Public Health Investment and Practice 253 Cohort Studies prospectively collect information on early life exposures Cohort studies can overcome some of the limitations and childhood cancers. This effort resulted from the of case-control and cross-sectional study designs by recognition that single studies lacked the statistical enabling the measurement of exposures at the time of power to study childhood cancers that are rare. It also cohort enrollment, often years or decades before cancer provided evidence that pre-conception and in utero development and diagnosis (Breslow and Day 1987; exposures may be important determinants of subse- Rothman and Greenland 2008). Exposure assessment quent risk of childhood and adult cancers (Brown and can be determined before the onset of disease and others 2007). thereby limit the potential for recall bias and other types Some adult cohorts that have made seminal contri- of bias inherent to case-control studies. butions to cancer epidemiology include the following: Cohort studies also contribute to health surveillance by providing the opportunity to obtain repeated mea- • The prospective cohort study of British doctors (Doll sures of multiple exposures and potential confounding and Hill 1954) factors and to measure changes in these factors over • The American Cancer Society Cancer Prevention time. Cohort studies enable estimation of the inci- Studies (Calle and others 2002; Hammond 1966; dence of outcomes of interest, including infections, Thun and others 1997) premalignant lesions, cancers, and comorbid conditions. • The United Kingdom Million Women Study (Million Follow-up of cohorts requires long-term commitment, Women Study Collaborative Group 1999) for supporting the study infrastructure and team, as • The Japan Life Span Study (Sakata and others 2012) well as for building and maintaining trust between the • The Nurses’ Health Study (Colditz, Manson, and research team and the participants in the cohort. Recent Hankinson 1997) (box 15.4) developments in information technology, including the increased uptake of mobile telephones and Internet More recently formed cohorts in LMICs include the access, have facilitated the development and mainte- following: nance of study cohort enrollment and follow-up. A cohort may be used to study multiple health • The China Kadoorie Biobank, which includes 500,000 endpoints and multiple exposures; new endpoints may adults from urban and rural areas in China (Chen be added over time, and data and biospecimens from and others 2011) (box 15.5) multiple cohorts can be pooled to obtain greater • A separate cohort of 220,000 men in China (Chen statistical power. Thun and others (2013), for exam- and others 2012) ple, pooled data from seven cohorts to analyze the • A cohort of 150,000 women and men in Mexico City long-term impact of cigarette smoking in the United (Kuri-Morales and others 2009) States. Another example is the 2004 formation of the • The Chennai Prospective Study of 500,000 adults in International Childhood Cancer Cohort Consortium Tamil Nadu, India (Gajalakshmi and others 2007; to assemble birth and child cohorts around the world to Gajalakshmi, Whitlock, and Peto 2012) Box 15.4 Cohort Study: The Nurses’ Health Studies The Nurses’ Health Study (2014a) comprises two associated biological specimens, the investigators cohorts of registered female nurses that enrolled have also been able to study risk factors for many more than 115,000 nurses in the United States. The other chronic diseases, including diabetes mellitus, first cohort began in 1976 and the second in 1989. stroke, osteoporosis, mental health, and connective These long-term epidemiologic studies were orig- tissue disease. The Nurses’ Health Study coordinat- inally designed to assess risk factors for two major ing center is recruiting a third cohort of 100,000 chronic diseases in women, namely, cancer and nurses. Participant registration and follow-up will cardiovascular disease. Due to the large sample size, be conducted entirely via Internet communication the extensive data available on each participant, and (Nurses’ Health Study 2014b). 254 Cancer Box 15.5 Cohort Study: The China Kadoorie Biobank The China Kadoorie Biobank includes 500,000 adults of the odds of prevalent cardiovascular disease recruited between 2004 and 2008 from 10 regions in (Bragg and others 2014); that only one in three China, urban and rural (Chen and others 2011). All individuals with prior cardiovascular disease participants are being followed for hospital admis- was routinely treated with any proven secondary sions, as well as cause-specific morbidity and mortal- preventive drugs (Chen and others 2014); that ity. Already, studies have been published on respiratory drinking alcohol was positively correlated with disease, depression, anxiety, diabetes, cardiovascu- regular smoking, increased blood pressure, and lar disease, alcohol consumption, physical activity, increased heart rate (Millwood and others 2013); and obesity, and interactions among these factors that major depression and generalized anxiety within the cohort (Bragg and others 2014; Chen disorder are associated with type 2 diabetes mel- and others 2014; Du and others 2013; Lewington and litus (Mezuk and others 2013); and that exhaled others 2012; Mezuk and others 2013; Millwood and carbon monoxide can be used as a biomarker others 2013; Zhang and others 2013). for assessing current smoking and exposure The studies have found, for example, that self- to indoor household air pollution (Zhang and reported diabetes was associated with a doubling others 2013). Health Communications that some cancers can be prevented, that some cancers Health communications contribute in a number of are amenable to screening, and that many cancers critical areas needed in cancer research (National can be treated successfully if diagnosed early. Such Cancer Institute 2004). First is the need to com- education and communication may help to overcome municate the importance of health research to the stigma of cancer, particularly in settings where the media, policy makers, and the general public. cancers are typically diagnosed at late stages with poor Without community recognition of the need for prognoses. health research, without appropriate levels of funding, Different communication strategies may be needed and without a regulatory framework that facilitates for different populations, based on language, levels the health research needed to guide public policy and of literacy and health literacy, access to health care, public investment in health, health research cannot socioeconomic status, cultural sensitivities, and other take place. factors. The development and validation of effective Second is the need to understand how best to com- cancer-related health communication strategies is key to municate to individuals, families, and communities that developing and implementing research that can facilitate research findings support public health recommendations cancer prevention and control. and guidance. One clear example is how best to commu- nicate the health risks associated with tobacco use, alcohol abuse, and physical inactivity. Ideally, such communica- Implementation Science tion should lead to changes in behavior that reduce the Effective means of implementing cancer prevention; risk of cancer and other chronic diseases. Other examples population-based screening; and timely and accurate are communication regarding recommended regimens cancer diagnosis, treatment, and symptom manage- for approved vaccines that prevent chronic infections ment are needed to improve cancer control. The U.S. associated with cancer, such as those for hepatitis B virus National Cancer Institute has developed a website (HBV) and human papillomavirus (HPV), and recom- (http://cancercontrolplanet.cancer.gov) with links to mendations for routine cancer screening. effective cancer control interventions, including Third is the need to identify ways to help commu- a database of research-tested intervention programs nities understand cancer. This includes understanding developed in partnership with the Substance Abuse Need for National Commitments to Cancer Research to Guide Public Health Investment and Practice 255 for implementation, as the Bamako Call to Action Box 15.6 advocates • Developing and implementing mechanisms for trans- What Questions Can Implementation Science parent and objective evaluation and prioritization of clinical research studies Answer? • Establishing systems for ethical, regulatory, and Implementation science can help obtain answers to the scientific reviews so that research can be conducted following questions for cancer prevention and control: in a timely manner and clinical studies can be com- pleted expeditiously • Revising customs inspections and policies to remove • Which tobacco control programs are most effective for restrictions on the importation of drugs, devices, and specific populations? reagents for health research • What are the most cost-effective ways to ensure that as • Developing mechanisms and resources for the effi- many children as possible are vaccinated for HBV and cient and inexpensive acquisition of drugs, devices, HPV? and reagents for health research • What is the most effective way to screen for and treat • Providing financial support—possibly from national preinvasive HPV-related cervical neoplasia? governments, state governments, nongovernmental • What is the most effective way to screen for colon organizations, or a combination—for the infrastruc- cancer? ture for public clinical trials, including protocol • What is the most effective way to provide palliative care, development, regulatory management, routine med- including pain control? ical expenses of patients, data management, quality • What are the most effective ways to ensure quality assurance, biobanking, biostatistics, and informatics control across the cancer spectrum, from screening to • Integrating clinical research into national health treatment to survivorship care? systems • Integrating education about clinical research into education and training of health care providers and Mental Health Services Administration (National Cancer Institute 2013). Local Level Different approaches may be needed for different It is important to assist the institutional leadership at the regions in the same country and between countries, local level—whether hospital, clinic, or university—to based on existing health resources, cultural norms, and appreciate the importance of health research. Doctors, other factors. In some cases, task shifting of responsibil- nurses, pharmacists, and specialists from other rele- ities from doctors to nurses, other health professionals, vant disciplines need protected time to conduct clin- and lay community health workers or from nurses to ical research. Academic tracks could be established to community health workers may be required. The princi- foster research and reward individuals for conducting ples of implementation science can guide public health clinical research. In some cases, money could be ear- interventions for cancer control and facilitate their rou- marked to pay for the additional costs associated with tine evaluation and modification, as needed, to achieve research. These costs may include additional imaging the goals (Madon and others 2007). Although more studies or specimen processing that may not be required research is needed in all the areas mentioned, much is for routine clinical care. The local study sites also need already known to be able to implement cancer control the appropriate financial and technical resources for clin- strategies to reduce the burden of cancer (box 15.6). ical research management, biobanking, and informatics. Research Training Cancer Research Collaboration For research to have a significant impact on health, National Level governments in LMICs and HICs need to invest in To optimize country-level outcomes, national policy training future scientists, clinicians, public health pro- makers will need to consider the essential elements of fessionals, and physician-scientists. Such investments in and necessary conditions for health research that require training professionals in the range of relevant disciplines a comprehensive approach that includes the following: and helping them to maintain and strengthen their research skills require effective coordination that may • Achieving a consensus that health research involve government ministries responsible for education, deserves the appropriate funding and strategies health, science and technology, and human resources, as 256 Cancer well as academic institutions, hospitals, clinics, nongov- Industry Collaboration ernmental organizations, and professional societies. Many aspects of cancer control, including prevention Principles of health research can be integrated into with vaccines, screening, diagnosis, treatment, and the core curriculums of schools of medicine, nursing, symptom management, require reliable drugs, devices, public health, pharmacy, and allied health sciences, as and reagents. Many areas of cancer need better, more well as university programs for basic sciences and social effective, more accessible, and less costly drugs, devices, sciences. This integration will ensure that all individu- and reagents. Partnering with industry will facilitate als involved in research with relevance for health learn the development and validation of novel products, as about the conduct of health research and appreciate the well as help to ensure a reliable supply chain to bring need for such research. products shown to be beneficial to routine clinical prac- Biostatisticians are a critical component of the tice. Once a product is developed and tested, it will be research team, as are basic and translational research important to work with industry partners to make the scientists, social scientists, health economists, and health product available and affordable on a population basis. communicators. There is a great need, particularly Such public-private partnering is an integral component among young scientists in low-resource settings, for of developing and translating innovations in cancer opportunities to participate in high-quality cancer research to clinical care and public health. research and to have access to nurturing mentors, whether local or remote. Other allied areas in which training and mentorship CERVICAL CANCER: EXEMPLAR OF are critically needed are the ethical conduct of cancer INTEGRATED RESEARCH research, research subject protection, scientific writing for preparation of research proposals and manuscripts, Cervical cancer provides a sterling example of how and responsible study and financial management. cancer research in virology, immunology, epidemiology, clinical care, behavioral sciences, and implementation International Collaboration science has led to effective cancer prevention and con- International collaboration in cancer research spans trol. Much of this work occurred because of interna- capacity building and joint research projects. There tional collaborations that allowed appropriate sharing of are many examples of North-South and South-South research material, data, and expertise. projects to build capacity in health and cancer research. Institutional “twinning” has been particularly success- Epidemiology and Biology ful in this regard. Joint research projects can facilitate training opportunities for investigators in LMICs. Two The link between chronic HPV infection and cervical recent reports from the Organisation for Economic cancer was established by zur Hausen and colleagues in Co-operation and Development’s Global Science Forum the early 1990s, building on earlier experimental work that highlight the need for international collaboration in strongly suggested the possibility that the two were linked clinical research to address many important health (Reid 1983; zur Hausen and de Villiers 1994). Muñoz questions (OECD 2011, 2013). and others (2002) confirmed that HPV was responsible The facilitation of such collaboration requires an for more than 99 percent of cervical cancers globally. effective national commitment to health research as Epidemiologic studies have also permitted the identifi- well as a commitment to facilitate international col- cation of additional co-risk factors for cervical cancer, laboration (Trimble and others 2009). International including exposure to tobacco smoke, both firsthand and research collaboration may require allowing specimens secondhand, chronic immunosuppression, multiparity, to be shipped to a regional or global core laboratory, as long-term use of oral contraceptives, and high-risk male well as pooling relevant information in an international partners (Schiffman and Hildesheim 2006). database. Timely scientific and ethical review is criti- cal for national studies, particularly for international collaboration (Abrams and others 2013). International Development of Vaccines to Prevent HPV Infection partnerships in cancer research require the timely The identification of HPV as a necessary causative agent recognition of scientific opportunities, available led to the development of vaccines to prevent HPV resources and study conditions, strengths of research infection and cervical cancer. Two vaccines to prevent partners, integrity, persistence, and commitment of all HPV infection demonstrated efficacy and safety in phase partners in jointly overcoming barriers to accomplish III and IV studies (Schiller, Castellsagué, and Garland research objectives. 2012). Both agents have been widely approved by drug Need for National Commitments to Cancer Research to Guide Public Health Investment and Practice 257 regulatory authorities, including the U.S. Food and Drug HPV types, as well as evidence of HPV integration. Administration and the European Medicines Agency. DNA-based tests detect the presence or absence of the The rollout of these vaccines to prevent HPV infection HPV virus genome. DNA testing for high-risk HPV types required a progression of clinical research studies, first has a high sensitivity for the detection of high-grade to confirm primary efficacy, then to validate in different cervical intraepithelial neoplasia and cervical cancer populations, followed by public health investigations (Arbyn and others 2012). The great advantages are that to determine how best to deliver and encourage uptake HPV detection assays are automated and objective and of the new vaccines to prevent HPV infection and have a greater reproducibility than cytology; as such, they cervical cancer (Program for Appropriate Technology in are a promising screening test in LMICs, which may lack Health 2012). skilled personnel. Sankaranarayanan and others (2009) These investigations incorporated studies of mes- found that a single round of testing for HPV was associ- saging, including how best to communicate to parents ated with a significant reduction in the risk of advanced the health benefits that the new vaccine offers to their cervical cancer and death from cervical cancer among daughters; studies comparing on-the-ground vaccine rural women in India. In this study, which accrued delivery programs, such as school-based versus clinic- 131,746 women ages 30–59 years in 52 villages, the other based HPV vaccination programs; and comparisons and two screening arms—cytologic testing and visual inspec- studies of different dosing schedules for HPV vaccines tion of the cervix with acetic acid—did not demonstrate (Galagan and others 2013; Lamontagne, Barge, and significant reductions in the risk of advanced cervical others 2011; Lamontagne, Thiem, and others 2013). cancer and deaths from cervical cancer. Australia, which has a national reporting system for HPV-associated warts in addition to a national registry for vaccinations, has been able to document significant Unanswered Questions in HPV-Associated Neoplasia decreases in HPV infection and genital warts among and Cervical Cancer Control teenagers and young adults following the widespread Based on the body of research to date, as well as introduction of the vaccine (Read and others 2011; the extensive programs for control of HPV-associated Tabizi and others 2012). Several second-generation vac- neoplasia, many critical research questions remain cines to prevent HPV infection are under development, unanswered (Schiller and Lowy 2014) (box 15.7). These with the goal of addressing the issue of type-restricted questions span a variety of scientific areas, including the protection and decreasing the cost of production. following: • Behavioral sciences HPV Diagnostics for Cervical Cancer Screening • Health communications Based on the understanding of the link between chronic • Health services research oncogenic HPV infection and cervical neoplasia, new • Immunology diagnostic tests have been developed to target the virus, • Implementation science including evidence of active infection with high-risk • Prevention Box 15.7 Ongoing Research Questions in HPV and Cervical Cancer Control The tremendous progress in HPV and cervical prophylactic HPV vaccine regimen be reduced? cancer control can be taken even further as these still Are one or two doses as effective as three? Will outstanding research questions are addressed. The additional booster doses be needed and when? following cervical cancer research agenda provides • Prevention, implementation science, and health opportunities for scientists in all interested countries. communications. What combination of feasibility and affordability would convince policy makers in LMICs to introduce and fund population-based • Prevention, health services research, and immunol- prophylactic HPV vaccination? ogy. How can the cost of the currently available box continues next page 258 Cancer Box 15.7 (continued) • Health communications, prevention, behavioral caring for HIV-positive individuals, well-woman sciences, and implementation science. What are programs, and programs screening for other the critical components in educational programs noncommunicable diseases, such as diabetes, for parents considering whether to permit their hypertension, and breast cancer? daughters and sons to undergo prophylactic • Health communications, screening, behavioral HPV vaccination? sciences, and implementation science. What mea- • Prevention, virology, and immunology. What is sures can be taken to ensure that no woman needed to develop second-generation prophylac- found to have an abnormal screening result is lost tic HPV vaccines that provide protection against to follow-up? infection from more HPV subtypes than included • Treatment. What can be done to improve current in the first-generation HPV vaccines? ablative therapy for preinvasive cervical and anal • Screening and virology. What is needed to develop cancer? effective screening strategies for HPV-associated • Immunology and treatment. Can therapeutic neoplasia of the oropharynx and anus? HPV immunization strategies be developed to • Screening, health services research, and imple- prevent the development of neoplasia in indi- mentation science. What types of infrastruc- viduals already infected with HPV, as well as to ture, human resource capacity, and logistical complement or replace ablative therapy for HPV- support are needed to scale up existing and associated neoplasia? new cervical neoplasia screening and treat- • Treatment and implementation science. How can ment services at multiple levels of the health current therapy, including surgery, radiation, system to meet the needs of urban and rural chemoradiation, and neoadjuvant chemotherapy, populations? be improved for women with invasive cervical • Screening and virology. What is the feasibility cancer? of developing inexpensive, highly sensitive, • Treatment, behavioral sciences, and implemen- and highly specific HPV-based tests to use as a tation science. How can the quality of life best primary screen for cervical neoplasia and chronic be maintained and enhanced in cervical cancer HPV infection in low-resource settings? survivors, including bowel, bladder, and sexual • Screening, health services research, and imple- function, as well as physical intimacy? mentation science. How can population-based • Symptom management, behavioral sciences, and screening for cervical neoplasia or chronic HPV implementation science. How can palliative care infection be more effectively integrated into be delivered most effectively to women diagnosed maternal-child health programs, programs with late-stage or recurrent cervical cancer? • Screening and cancer-associated deaths, are critical to effective • Symptom management decision making for prevention and control, as well as • Treatment priorities in cancer research. • Virology. • Next is implementation science focused on how to deliver interventions that have been shown to be effective. Perhaps the most effective method of CONCLUSIONS: “BEST BUYS” FOR CANCER cancer prevention is tobacco control. Countries at all levels of income could sponsor research focused on RESEARCH how best to reduce or eliminate use of tobacco. This What are the “best buys” for cancer research in LMICs? research should include public policy, public educa- Where should ministers of health, ministers of science tion, and smoking cessation initiatives. and technology, and other funders of research begin? • For LMICs burdened with liver cancer or cervical can- cer, implementation science focused on expanding rou- • Robust health surveillance systems, including sur- tine administration of HBV and HPV vaccinations is veillance of cancer risk factors, cancer registries, appropriate. For countries at all levels of development, Need for National Commitments to Cancer Research to Guide Public Health Investment and Practice 259 implementation science in effective methods to deliver Boyle, P., and B. Levin, eds. 2008. World Cancer Report. Lyon: palliative care is critical. Educational and training pro- IARC Press. grams for health care professionals and community Bragg, F., L. Li, M. Smith, Y. Guo, Y. Chen, and others. 2014. health workers would benefit from the inclusion of “Associations of Blood Glucose and Prevalent Diabetes with Risk of Cardiovascular Disease in 500,000 Adult Chinese: the principles of health research. How best to educate The China Kadoorie Biobank.” Diabetic Medicine 31 (5): and retain health care workers at all levels is also an 540–51. appropriate area for research. Breslow, N. E., and N. E. Day. 1987. Statistical Methods in • Countries with the ability to conduct programs for Cancer Research, Volume II: The Design and Analysis of cancer screening, early diagnosis, and treatment Cohort Studies. Scientific Publications 82. Lyon: IARC Press. could expand implementation science research to Brown, R. C., T. Dwyer, C. Kasten, D. Krotoski, Z. Li, and others. cover these areas, as well as tobacco control and 2007. “International Childhood Cancer Cohort Consortium preventive vaccines. Timely topics for research (I4C).” International Journal of Epidemiology 36 (4): 724–30. also include cost-effective strategies to screen for Calle, E. E., C. Rodriguez, E. J. Jacobs, L. Almon, A. Chao, cervical, colon, oral, esophageal, stomach, and skin and others. 2002. “The American Cancer Society Cancer cancers; to evaluate breast masses; and to provide Prevention Study II Nutrition Cohort: Rationale, Study Design, and Baseline Characteristics.” Cancer 94 (9): potentially curative therapy for preinvasive and 2490–501. invasive cancer. Chen, Y., L. Li, Q. Zhang, R. Clarke, J. Chen, and others. 2014. • The next area for research, as national resources “Use of Drug Treatment for Secondary Prevention of permit, encompasses cancer epidemiology and biology. Cardiovascular Diseases in Urban and Rural Communities Improved understanding of the risk and protective of China: China Kadoorie Biobank Study of 0.5 Million factors for specific cancers, as well as their molecular People.” International Journal of Cardiology 172 (1): 88–95. biology, is essential to design effective interventions for Chen, Z., J. Chen, R. Collins, Y. Guo, R. Peto, and others. prevention, screening, early diagnosis, and treatment. 2011. “China Kadoorie Biobank of 0.5 Million People: For example, the epidemiology and biology of prostate Survey Methods, Baseline Characteristics and Long-Term cancer is not sufficiently well understood to prevent it, Follow-Up.” International Journal of Epidemiology 40 (6): screen for it, or know how best to treat it. Similarly, the 1652–66. Chen, Z., G. Yang, A. Offer, M. Zhou, M. Smith, and others. ability to make progress on cancer control for many 2012. “Body Mass and Mortality in China: A 15-Year cancers linked to chronic infection and inflammation Prospective Study of 220,000 Men.” International Journal of requires better elucidation of the salient biology, natu- Epidemiology 41 (2): 472–81. ral history, co-factors, and protective factors. Colditz, G. A., J. E. Manson, and S. E. Hankinson. 1997. “The Nurses’ Health Study: 20-Year Contribution to the Understanding of Health among Women.” Journal of NOTE Women’s Health 6 (1): 49–62. World Bank income classifications as of July 2014 are as De Martel, C., J. Ferlay, S. Franceschi, J. Vignat, F. Bray, and follows, based on estimates of gross national income per capita others. 2012. “Global Burden of Cancers Attributable to for 2013: Infections in 2008: A Review and Synthetic Analysis.” The Lancet Oncology 13 (6): 607–15. Dikshit, R., P. C. Gupta, C. Ramsundarahettige, V. Gajalakshmi, • Low-income countries: US$1,045 or less L. Aleksandrowicz, and others. 2012. “Cancer Mortality in • Middle-income countries: India: A Nationally Representative Survey.” The Lancet 379 a) Lower-middle-income: US$1,046–US$4,125 (9828): 1807–16. b) Upper-middle-income: US$4,126–US$12,745 Doll, R., and A. B. Hill. 1954. “The Mortality of Doctors in • High-income countries: US$12,746 or more Relation to Their Smoking Habits: A Preliminary Report.” British Medical Journal 1 (4877): 1451–55. Du, H., D. Bennett, L. Li, G. Whitlock, Y. Guo, and others. REFERENCES 2013. “Physical Activity and Sedentary Leisure Time and Abrams, J. S., M. M. Mooney, J. A. Zwiebel, E. L. Korn, Their Associations with BMI, Waist Circumference, and S. H. Friedman, and others. 2013. “Implementation of Percentage Body Fat in 0.5 Million Adults: The China Timeline Reforms Speeds Initiation of National Cancer Kadoorie Biobank Study.” American Journal of Clinical Institute–Sponsored Trials.” Journal of the National Cancer Nutrition 97 (3): 487–96. Institute 105 (13): 954–59. Gajalakshmi, V., R. Peto, V. C. Kanimozhi, G. Whitlock, Arbyn, M., G. Ronco, A. Anttila, C. J. Meijer, M. Poljak, and and D. Veeramani. 2007. “Cohort Profile: The Chennai others. 2012. “Evidence Regarding Human Papillomavirus Prospective Study of Mortality among 500,000 Adults Testing in Secondary Prevention of Cervical Cancer.” in Tamil Nadu, South India.” International Journal of Vaccine 30 (Suppl. 5): F88–99. Epidemiology 36 (6): 1190–95. 260 Cancer Gajalakshmi, V., G. Whitlock, and R. Peto. 2012. “Social Diverse Regions of China: Prevalence, Patterns, and Socio- Inequalities, Tobacco Chewing, and Cancer Mortality in Demographic and Health-Related Correlates.” International South India: A Case-Control Analysis of 2,580 Cancer Journal of Epidemiology 42 (3): 816–27. Deaths among Non-Smoking Non-Drinkers.” Cancer Muñoz, N., S. Franceschi, C. Bosetti, V. Moreno, R. Hererro, and Causes and Control 23 (Suppl. 1): 91–98. others. 2002. “Role of Parity and Human Papillomavirus in Galagan, S. R., P. Paul, L. Menezes, and D. S. LaMontagne. 2013. Cervical Cancer: The IARC Multicentric Case-Control “Influences on Parental Acceptance of HPV Vaccination in Study.” The Lancet 359 (9312): 1093–101. Demonstration Projects in Uganda and Vietnam.” Vaccine National Cancer Institute. 2004. Making Health Communications 31 (30): 3072–78. Programs Work. Bethesda, MD: National Institutes of Hammond, E. C. 1966. “Smoking in Relation to the Death Health, U.S. Department of Health and Human Services. Rates of One Million Men and Women.” National Cancer ———. 2013. Cancer Control P.L.A.N.E.T. http:// Institute Monographs 19: 127–204. cancercontrolplanet.cancer.gov Howlader, N., A. M. Noone, M. Krapcho, J. Garshell, N. Neyman, NRC (National Research Council). 2011. Toward Precision and others, eds. 2013. “SEER Cancer Statistics Review, 1975– Medicine: Building a Knowledge Network for Biomedical 2010.” National Cancer Institute, Bethesda, MD, based on Research and a New Taxonomy of Disease. Committee on November 2012 SEER data submission, posted to the SEER a Framework for Developing a New Taxonomy of Disease. website, April 2013, http://seer.cancer.gov/csr/1975_2010. Washington, DC: National Academies Press. Institute for Health Metrics and Evaluation. 2011. Verbal Nurses’ Health Study. 2014a. The Nurses’ Health Study. http:// Autopsy Series. http://www.healthmetricsandevaluation www.channing.harvard.edu/nhs. .org/publications/verbal-autopsy-series. ———. 2014b. Nurses’ Health Study 3. http://www.nhs3.org. Jensen, O. M., D. M. Parkin, R. MacLennan, C. S. Muir, and OECD (Organisation for Economic Co-operation and R. G. Skeet, eds. 1991. Cancer Registration: Principles and Development). 2011. “Facilitating International Methods. Scientific Publications 95. Lyon: IARC Press. Cooperation in Non-Commercial Clinical Trials.” Global Kuri-Morales, P., J. Emberson, J. Alegre-Diaz, R. Tapia-Conyer, Science Forum, OECD. http://www.oecd.org/science R. Collins, and others. 2009. “The Prevalence of Chronic /sci-tech/49344626.pdf. Diseases and Major Disease Risk Factors at Different Ages ———. 2013. “Recommendation on the Governance of among 150,000 Women Living in Mexico City: Cross- Clinical Trials.” http://www.oecd.org/sti/sci-tech/oecd Sectional Analyses of a Prospective Study.” BioMed Central -recommendation-governance-of-clinical-trials.pdf. Public Health 9: 9. Program for Appropriate Technology in Health. 2012. Lamontagne, D. S., S. Barge, N. T. Le, E. Mugisha, M. E. Penny, “Evaluating HPV Vaccination Pilots: Practical Experience and others. 2011. “Human Papillomavirus Vaccine Delivery from PATH.” Program for Appropriate Technology in Health, Strategies That Achieved High Coverage in Low- and Seattle. http://www.rho.org/HPV-evaluating-pilots.htm. Middle-Income Countries.” Bulletin of the World Health Read, T. R., J. S. Hocking, M. Y. Chen, B. Donovan, and Organization 89 (11): 821–30B. C. S. Bradshaw. 2011. “The Near Disappearance of Genital Lamontagne, D. S., V. D. Thiem, V. M. Huong, Y. Tang, Warts in Young Women 4 Years after Commencing a K. M. Neuzil, and others. 2013. “Immunogenicity of National Human Papillomavirus (HPV) Vaccination Quadrivalent HPV Vaccine among Girls 11 to 13 Years Programme.” Sexually Transmitted Infections 87 (7): of Age Vaccinated Using Alternative Dosing Schedules: 544–47. Results 29 to 32 Months after Third Dose.” Journal of Reid, R. 1983. “Genital Warts and Cervical Cancer: Is Infectious Diseases 208 (8): 1325–34. Human Papillomavirus Infection the Trigger to Cervical Lewington S., L. Li, P. Sherliker, Y. Guo, I. Millwood, and Carcinogenesis?” Gynecologic Oncology 15 (2): 239–52. others. 2012. “Seasonal Variation in Blood Pressure and Rothman, D. J., and S. Greenland. 2008. “Cohort Studies.” Its Relationship with Outdoor Temperature in 10 Diverse In Modern Epidemiology, edited by K. J. Rothman, Regions of China: The China Kadoorie Biobank.” Journal of S. Greenland, and T. Lash, 3rd edition, 100–10. Philadelphia, Hypertension 30 (7): 1383–91. PA: Lippincott Williams and Wilkins. Madon, T., K. J. Hofman, L. Kupfer, and R. I. Glass. 2007. Rothman, K. J., S. Greenland, and T. L. Lash. 2008. “Case- “Public Health: Implementation Science.” Science 318 Control Studies.” In Modern Epidemiology, edited by (5857): 1728–29. K. J. Rothman, S. Greenland, and T. Lash, 3rd edition, Mezuk, B., Y. Chen, C. Yu, Y. Guo, Z. Bian, and others. 111–27. Philadelphia, PA: Lippincott Williams and 2013. “Depression, Anxiety, and Prevalent Diabetes in the Wilkins. Chinese Population: Findings from the China Kadoorie Sakata, R., P. McGale, E. J. Grant, K. Ozasa, R. Peto, and Biobank of 0.5 Million People.” Journal of Psychosomatic others. 2012. “Impact of Smoking on Mortality and Life Research 75 (6): 511–17. Expectancy in Japanese Smokers: A Prospective Cohort Million Women Study Collaborative Group. 1999. “The Study.” British Medical Journal 345: e7093. Million Women Study: Design and Characteristics of the Sankaranarayanan R., B. M. Nene, S. S. Shastri, K. Jayant, Study Population.” Breast Cancer Research 1 (1): 73–80. R. Muwonge, and others. 2009. “HPV Screening for Cervical Millwood, I. Y., L. Li, M. Smith, Y. Guo, L. Yang, and others. Cancer in Rural India.” New England Journal of Medicine 2013. “Alcohol Consumption in 0.5 Million People from 10 360 (14): 1385–94. Need for National Commitments to Cancer Research to Guide Public Health Investment and Practice 261 Schiffman, M. H., and A. Hildesheim. 2006. “Cervical This policy paper includes the need for implemen- Cancer.” In Cancer Epidemiology and Prevention, edited tation science research in low- and middle-income by D. Schottenfeld and J. F. Fraumeni, 1044–67. New York: countries to guide effective cancer preventions and Oxford University Press. control in these settings. Schiller, J. T., X. Castellsagué, and S. M. Garland. 2012. Hunter, D. J., and K. S. Reddy. 2013. “Noncommunicable “A Review of Clinical Trials of Human Papillomavirus Diseases.” New England Journal of Medicine 369: 1336–43. Prophylactic Vaccines.” Vaccine 30 (Suppl. 5): F123–38. The authors emphasize the need for global research Schiller, J. T., and D. R. Lowy. 2014. “Virus Infection and efforts to inform the prevention, detection, and treat- Human Cancer: An Overview.” Recent Results in Cancer ment of noncommunicable diseases. Research 193: 1–10. Institute of Medicine. 2011. “The National Cancer Policy Tabizi, S. N., J. M. Brotherton, J. M. Kaldor, S. R. Skinner, Summit: Opportunities and Challenges in Cancer Research E. Cummins, and others. 2012. “Fall in Human and Care.” National Academies Press, Washington, DC. Papillomavirus Prevalence Following a National Vaccination ———. 2013a. Delivering High-Quality Cancer Care: Charting Program.” Journal of Infectious Diseases 206 (11): 1645–51. a New Course for a System in Crisis. Washington, DC: Thun, M. J., B. D. Carter, D. Feskanich, N. D. Freedman, National Academies Press. R. Prentice, and others. 2013. “50-Year Trends in Smoking- ———. 2013b. “Delivering Affordable Cancer Care in the 21st Related Mortality in the United States.” New England Century-Workshop Summary.” National Academies Press, Journal of Medicine 368 (4): 351–64. Washington, DC. Thun, M. J., R. N. Hoover, and D. J. Hunter. 2012. “Bigger, Together these three reports from the Institute of Better, Sooner: Scaling Up for Success.” Cancer Epidemiology, Medicine outline the essential elements for effective Biomarkers, and Prevention 21 (4): 571–75. cancer research, as well as priorities for implementation Thun, M. J., C. A. Lally, J. T. Flannery, E. E. Calle, W. D. Flanders, science to guide cancer control. and others. 1997. “Cigarette Smoking and Changes in the Lozano, R., M. Naghvai, K. Foreman, S. Lim, K. Shibuya, Histopathology of Lung Cancer.” Journal of the National V. Aboyans, and others. 2012. “Global and Regional Cancer Institute 89 (21): 1580–86. Mortality from 235 Causes of Death for 20 Age Groups Trimble, E. L., J. S. Abrams, R. M. Meyer, F. Calvo, E. Cazap, in 1990 and 2010: A Systematic Analysis for the Global and others. 2009. “Improving Cancer Outcomes through Burden of Disease Study 2010.” The Lancet 380: 2095–28. International Collaboration in Academic Cancer Treatment This analysis makes clear the burden of cancer, com- Trials.” Journal of Clinical Oncology 27 (30): 5109–14. pared with other causes of death. Vaught, J. B., M. K. Henderson, and C. C. Compton. 2012. Varmus, H., and H. S. Kumar. 2013. “Addressing the Growing “Biospecimens and Biorepositories: From Afterthought to International Challenge of Cancer: A Multi-National Science.” Cancer Epidemiology, Biomarkers, and Prevention Perspective.” Science Translational Medicine 175. 21 (2): 253–55. The authors summarize the recommendations, includ- WHO (World Health Organization). 2008. “The Bamako Call ing priorities for research, of representatives of institu- to Action for Research on Health.” WHO, Geneva. http:// tions and organizations that fund and perform cancer www.who.int/rpc/news/BAMAKOCALLTOACTIONFinal research in 15 countries that include more than half of Nov24.pdf. the world’s population. ———. 2013. “Global Health Observatory.” WHO, Geneva. WHO (World Health Organization). 1996. “Ad Hoc Committee http://www.who.int/gho/database/en/. on Health Research Relating to Future Intervention Wild, C. P. 2009. “Environmental Exposure Measurement in Options: Investing in Health Research and Development.” Cancer Epidemiology.” Mutagenesis 24 (2): 117–25. Document TDR/Gen/96.1, WHO, Geneva. Zhang, Q., L. Li, M. Smith, Y. Guo, G. Whitlock, and others. This document was developed to assist decision making 2013. “Exhaled Carbon Monoxide and Its Associations with by government, industry, and other investors on the Smoking, Indoor Air Pollution and Chronic Respiratory allocation of funds for health research and development. Diseases among 512,000 Chinese Adults.” International ———. 2013. Global Action Plan for the Prevention and Control Journal of Epidemiology 42 (5): 1464–75. Of Non-Communicable Diseases. 2013–2020. ISBN 978 92 4 zur Hausen, H., and E. M. de Villiers. 1994. “Human 150623 6. Papillomaviruses.” Annual Review of Microbiology 48: 427–47. This action provides guidance on public health priori- ties and thus key implementation science issues related to cancer prevention and control, developed as part ANNOTATED SELECT BIBLIOGRAPHY of follow-up to the 2011 United Nations High-Level Farmer, P., J. Frenk, F. M. Knaul, L. N. Shulman, G. Alleyne, Meeting on Non-communicable Diseases. L. Armstrong, and others. 2010. “Expansion of Cancer Wild, C. P. 2012. “The Role of Cancer Research in Care and Control in Countries of Low and Middle Income: Noncommunicable Disease Control 2012.” Journal of the A Call to Action.” The Lancet 376: 1186–93. National Cancer Institute 104 (14): 5109–14. 262 Cancer Chapter 16 Cancer in Low- and Middle-Income Countries: An Economic Overview Susan Horton and Cindy L. Gauvreau INTRODUCTION at the same level of economic development differ because other factors intervene. Urbanization affects Health care is informed first and foremost by scientific the patterns of cancer and the ability to access care. and medical understanding of how to treat and prevent Local champions, governmental political leadership, disease. Economics can, however, provide useful insights and international partnerships can all loosen the con- to inform policy in the design and implementation straints of local economic resources. Conversely, some of the systems to provide health care, as well as in the countries are underachievers in cancer care despite process of prioritizing interventions to make the best their income level, perhaps because of leadership use of scarce resources. Treating a single cancer patient failures. may require the coordination of many inputs and may The third section reviews the cost-effectiveness of cost tens or even hundreds of thousands of dollars interventions for cancer care, where care is here defined in high-income countries (HICs). Ongoing popula- to include prevention. The cost-effectiveness of inter- tion cancer screening and early detection also require ventions has been well studied in HICs, but much less considerable coordination, including treatment for cases so in low- and middle-income countries (LMICs). This detected, and costs. Finally, although knowledge of section summarizes the literature on the economics of cancer prevention is inadequate, prevention can be a cancer care in LMICs; the section also draws on the costly endeavor—as demonstrated by the large sums literature from HICs, particularly for cancer treatment, spent on behavior change promotion (such as smok- in areas where reliable studies for LMICs are particularly ing cessation) or on vaccines to prevent cancer, such scarce. It may be possible to make inferences for one as against human papilloma virus to prevent cervical country using results from another country; the validity cancer and hepatitis B virus to prevent liver cancer—and of these inferences rises with the extent of the similar- economics can be informative. ities in the two countries. Where possible, we separate The second section of this chapter reviews how out the findings for high-income economies in Asia, the availability of resources for cancer care varies by since they are likely to be more relevant for LMICs in economic status, using the World Bank’s categories of this region than the results from North America or low-income countries (LICs), middle-income coun- Western Europe. tries (MICs) (comprising lower-middle-income coun- We use the resource grouping suggested by Anderson tries and upper-middle-income countries), and HICs. and others (see chapter 3) for the Breast Health Global At the same time, economy is not destiny. Countries Corresponding author: Susan Horton, PhD, sehorton@uwaterloo.ca 263 Initiative and apply this to other cancers. In this factors that affect health care generally, such as the framework, facility resource environments fall into four availability of health insurance (public, private, or categories of resource availability: mixed) and general administrative capability for the requisite health care systems. • Basic LMICs generally have inadequate resources for health • Limited care, which conditions what is available specifically for • Enhanced cancer care and, hence, mortality rates. From a policy • Maximal perspective, it is important to identify the priorities for investment to make maximum health gains with the These categories are correlated with the World Bank available budgetary resources. We use cost-effectiveness income groupings. LICs have a preponderance of Basic analysis to provide some guidelines, for areas where facilities, rural areas in MICs have more facilities with additional recurrent expenditures would benefit care Limited capabilities, urban areas in MICs have more (such as buying additional drugs) and for areas where facilities with Enhanced capabilities, and much of the large investments in fixed costs are required (such as population in HICs has access to facilities with Maximal setting up a specialized cancer hospital). capabilities. The implications for the availability of Some resource constraints can be overcome. Even resources specific to cancer care are described. This low-income Sub-Saharan African countries can acquire section requires some interpolation on the authors’ part and maintain radiation facilities, although ensuring because of the paucity of previous work in the area and access for patients from remote rural areas may be is subject to further validation by experts. difficult. It is more challenging, however, to advocate The fourth and final section contains conclusions, treatments that require sophisticated pathology and consisting of summary recommendations of pack- laboratory facilities. Such facilities are important for ages of cancer care appropriate for each of the four a wide range of medical conditions, for which cancer resource environments, as well as priority areas where forms only a small percent, and they require a much further research is required. The appropriateness of a larger effort and investment to set up and maintain, package is defined by feasibility (those resources can particularly the training of skilled personnel. It may be expected to exist or could exist with reasonable be completely infeasible in such countries to consider investments) and by likely cost-effectiveness (within certain types of organized screening if no insurance the limits of available data). Although there are inter- system is in place to finance the screening costs, much nationally validated resource-specific care guidelines less the treatment of the cases diagnosed. for breast cancer (the Breast Health Global Initiative), Table 16.1 provides examples of availability, by no such guidelines are available as yet for other cancers. income grouping, of some specific resources relevant The packages presented here have been validated in to cancer treatment; each resource is discussed in turn. consultation with the chapter authors of this volume Information about the availability of radiation ther- (chapters 3 through 8), but need to be further refined apy and cancer registries is available elsewhere (and by expert consultation. not included in the table); quantitative data on the availability of skilled personnel and laboratories are not easily obtained. AVAILABILITY OF CANCER CARE RESOURCES ACROSS COUNTRIES Surgery Patterns of cancer vary across countries of different Surgery is the cornerstone of treatment for many income levels (chapter 2 in this volume). Countries also solid tumors. The level of surgical skill and associated have different capabilities for cancer care, depending on resources required varies by type of cancer. Surgery for resource availability. Some of the resources for cancer earlier stage colon cancer or mastectomy for early-stage care are specific to individual cancers, for example, the breast cancer can be undertaken at reasonably well- availability of a specific drug or test kit. Other resources equipped first-level hospitals. More sophisticated are specific to cancer in general, for example, radiation facilities and skills are required for such procedures as therapy or the need for specialized medical personnel breast-conserving surgery and rectal surgery. Surgery trained in oncology. Still others are not specific to can- for certain precancerous conditions may be possible at cer but affect many kinds of medical care, including lower-level facilities; cryotherapy for cervical cancer, for imaging facilities, surgical facilities, pathology, and example, can be undertaken in clinics. Table 16.1 shows laboratory medicine services. Finally, there are broader that HICs have more than 12 times as many operating 264 Cancer Table 16.1 Resource Availability Affecting Cancer Care by Country Income Groups Low-income Lower-middle- Upper-middle- High-income Resource countries income countries income countries countries Surgical facilities per 1,000 population,a 2007–08 1.3 4.7 9.9 16.4 Out-of-pocket health expenditure (% of health 48.1 52.8 33.3 13.7 expenditures),b 2011 Availability of tamoxifen (% of countries where it 34 53 80 85 is generally available, according to knowledgeable respondents),c,d 2010 Availability of oral morphine (% of countries where 27 28 57 81 it is generally available, according to knowledgeable respondents),c,d 2010 Income range,e 2012 (US$) 1,036 and below 1,036–4,087 4,087–12,615 12,616 and above a. Funk and others 2010. b. World Bank 2013. c. WHO 2012. d. The question asked whether the medication was generally available in the public sector. e. World Bank 2013. theaters per capita as LICs (chapter 13 in this volume). and rural areas of MICs often mean that these areas can Countries may face difficult choices as to how much only afford the lowest cost (usually older, off-patent) surgical capacity to utilize for palliation for patients for regimens. In addition to the cost of the agents, che- whom there is no chance of cure, compared with those motherapy requires multiple visits to a health facility for whom there is the possibility of cure. each month to obtain the supporting blood chemistry. Facilities need the Limited level of resources to support chemotherapy, effectively restricting its use to MICs Radiotherapy and HICs. Radiotherapy is a key to improving survival for certain Some effective but modest-cost cancer medications cancers. In Afghanistan, Iraq, and Sub-Saharan Africa, should be available, even from Basic level facilities. 25 countries lack any radiation unit; other countries have As long as a single laboratory test can be undertaken one unit per five million people (IAEA 2013). Clearly, per patient to determine hormone receptor status, the radiation unit alone is not the only constraint; tamoxifen can be used, even in rural areas and LICs. sufficient trained staff members are also required. Not Pain control medication, including morphine, should surprisingly, greater availability of radiotherapy is cor- be available in all environments as long as access can related with country income. Low-income countries be controlled. can provide radiotherapy, but the main issue is that the capacity in many countries is completely insufficient to meet the need. Typically, facilities need, at a minimum, Laboratories the Limited level of resources to be able to provide Laboratories are an essential component of screening, radiotherapy. diagnosis, and treatment options. They are required for rapid, accurate results from cytology or biopsies, or from the analysis of blood chemistry for chemotherapy. Medications These services are typically not available in Basic, or Pharmaceuticals of various kinds are vital to improve even Limited, facilities. Although it is possible to send cancer survival rates, yet country income is associated specimens collected from rural residents to a major city, with the availability of these agents. Access to tamox- the results are often not obtained in a sufficiently timely ifen for breast cancer is limited in LICs, as is access to manner to provide optimal treatment. Hence, treat- pain control using oral morphine (table 16.1, using ments involving extensive laboratory support are often survey data from WHO 2012). The case is similar for not feasible in settings without facilities with Enhanced chemotherapy agents, although no quantitative data resources, as in urban areas of MICs (see Fleming in were readily available. The budget constraints in LICs DCP3 volume 9, Disease Control Priorities). Cancer in Low- and Middle-Income Countries: An Economic Overview 265 Cancer Registries the level of development. In the poorer countries, only Cancer registries, which form the basis for understand- those people working in the formal sector or for the ing and documenting patterns of cancer, are a basic tool government may be covered; coverage in rural areas is in health care service provision. In LMICs, the percent- minimal. Out-of-pocket spending on health constituted age of the population covered by a high-quality registry, 48.1 percent of health expenditures in LICs, 52.8 percent such as those in the International Agency for Research on in lower-middle income countries, 33.3 percent in Cancer’s series on Cancer Incidence in Five Continents upper-middle income countries, and only 13.7 percent (Curado and others 2007), is in the single digits; this level in HICs (table 16.1, using data for the most recent years rises to double digits in Europe; it is 80 percent or more available from World Bank 2013). Expansion of insur- only in Australia, New Zealand, and North America. ance coverage in Mexico since 2003 through Seguro Although it is not essential to have 100 percent popu- Popular—the scheme that covers those working outside lation coverage, country planning and policy setting are the formal sector—was accompanied by an increase in much more difficult in the absence of a cancer registry of coverage of cervical cancer screening and a reduction in reasonable quality that covers at least one region. the proportion of those abandoning breast cancer treat- ment (Knaul and others 2012). Chapter 17 elaborates further on the role of health financing in cancer care. Skilled Medical Personnel Although the importance of country resource levels The lack of adequate numbers of skilled medical per- for the inputs required for different aspects of cancer care sonnel in LICs affects the ability to screen for, as well as has been documented, some countries underperform to treat, cancer. LICs have few oncologists and oncology despite relatively high levels of income. The disloca- nurses, which limits treatment ability. Although some of tion of public health systems in the Russian Federation the tests involved in cancer screening are often decep- following the economic system change, combined with tively low-technology interventions (for example, Pap adverse risk factors that include the relatively high smears, clinical breast exams, and fecal occult blood consumption of fat, tobacco, and alcohol, is associated tests), the organizational skills and infrastructure to con- with high rates of incidence as well as overall cancer duct them successfully at scale and ensure appropriate mortality. For example, 25 percent of the cases of referral make screening a high-technology intervention. colon cancer are diagnosed at stage 4 and 33 percent Accordingly, organized screening programs become fea- of newly diagnosed patients die within a year of diag- sible in urban areas of MICs. Opportunistic screening, nosis (Avksentyeva 2010). Many of those diagnosed however, can occur in countries at all levels, provided are not eligible to receive reimbursement for drugs; of that the screening test involved is not too demanding; those who are eligible, drug supply problems inhibit the clinical breast examination, visual inspection with acetic success of treatment. A cancer registry has existed since acid of the cervix, and the rapid DNA test for human 1939 and, in theory, screening programs exist for at least papillomavirus (HPV) in low-resource environments five cancers. In practice, however, the lack of resourc- are all possibilities. Campaign-style screening has been ing and lack of political will are associated with poor successfully used in LMICs, for example, in the Arab outcomes in cancer care (Avksentyeva 2010). Republic of Egypt for breast cancer and in India for oral cancer; a rapid DNA test for cervical cancer is being piloted in low-resource provinces in China. In a cam- International Partnerships paign, the effort is made to screen a large number of the Local champions can enable countries to outperform vulnerable groups in a short period of time, sometimes others at similar income levels. Local champions— in local health facilities and sometimes using outreach, key individuals willing to exert their influence in for example, using mobile facilities. advocacy and/or leadership—can draw on substantial international resources that can make a difference, whether through partnership with a single other country Health Insurance or hospital or through membership in international Health insurance conditions access to all services, networks. Partnerships have been used extensively for including treatment and screening. Individuals, who pediatric cancer in particular (Sloan and Gelband 2007, often underinvest in preventive health measures, may chapter 7). Although international financial resources not see the value of paying for screening tests, particu- can be important in saving lives, such as the radiation larly if they cannot afford treatment if they are subse- facilities provided by the International Atomic Energy quently diagnosed. The proportion of the population Authority (IAEA) to selected Sub-Saharan African coun- covered by health insurance typically increases with tries, the true value of these resources lies in access to 266 Cancer expertise and support for developing guidelines and for HBV. In contrast, 16 articles were found that satis- systems. fied the inclusion criteria for the cost-effectiveness of Cancer care works best in the context of a national vaccination and/or screening for cervical cancer. This cancer plan, and political leadership and the will to pro- result may reflect the fact that international funding vide the funding for the plan are keys. The World Health has been available to investigate and promote the vac- Organization reports that, although increasing numbers cines, principally through Gavi, the Vaccine Alliance. of countries surveyed have developed cancer plans over An additional six articles for breast and colon cancer the past decade, many countries still have not dedicated were found for HICs in Asia. These six were included, resources to fund these plans (WHO 2012). since they may provide some guidance for MICs in The role of partnerships with or membership in this region. international networks matters at all levels, including the Another reason that may explain the thin litera- following examples: ture is that there are very few articles (whether for LMICs or HICs) on the cost-effectiveness of surgery, • For clinicians: World Endoscopy Society, http:// the cornerstone of cancer treatment. The effectiveness www.worldendo.org of basic surgery was established long before economic • For guideline-setting: Breast Health Global Initiative, cost-effectiveness methods were developed and surgery http://portal.bhgi.org/Pages/Default.aspx or Asia became “usual care.” Pacific Working Group on Colorectal Cancer Given the lack of cost-effectiveness data for LMICs, • For screening: International Cancer Screening the literature from other countries might prove helpful. Network, http://appliedresearch.gov.icsn The literature from HICs may provide guidance and the • For training: IAEA’s support of radiation person- literature from HICs in Asia may be useful for other nel training in eight LICs and MICs through its countries in the region. The cost-effectiveness literature Programme of Action for Cancer Therapy model has to be used cautiously, since the greater the differ- demonstration sites project, http://cancer.iaea.org ence in context (including disease patterns, prevalence, /pmds.asp usual care alternatives, costs, and comorbidities), the less reliable the comparison is likely to be. We utilized Partnerships between institutions, such as twinning cost-effectiveness findings from HICs from the web arrangements between cancer hospitals, serve a similar appendix of Greenberg and others (2010). Greenberg function. and colleagues undertook a systematic review of inter- ventions for several cancers in HICs. Their approach COST-EFFECTIVENESS yielded some useful studies, primarily for breast cancer, that have relevance for LMICs. Cost-Effectiveness Methods In tables 16.3 through 16.8, we indicate gener- Cost-effectiveness methods are described in standard ally whether an intervention is “very cost-effective,” texts (such as Drummond and others 2005); these meth- “cost-effective,” or “not cost-effective” in a given study. ods have been applied widely in LMICs for infectious A few countries have set their own decision criteria, disease, for example, where there are large flows of inter- for example, the National Institute for Health and Care national assistance. These methods have been much less Excellence in the United Kingdom. In a comprehensive well used for cancer interventions in LMICs, with the study of cost-effectiveness in Australia, Vos and oth- exception of vaccines for hepatitis B (HBV) and HPV ers (2010) categorize interventions that cost less than and new DNA tests for HPV. US$10,000 per quality-adjusted life year (QALY) as very For this volume, a systematic literature search was cost-effective; those interventions between US$10,000 undertaken to identify studies from LMICs for all aspects and US$50,000 per QALY are cost-effective; and those of care for six cancers; the literature on tobacco control over $50,000 per QALY are not cost-effective. A similar is addressed separately in chapter 10. The search covered limit (US$50,000) is often used in the United States English language articles contained in PubMed and as the dividing line between what is and is not cost- EconLit from 2000 to 2013. The detailed search terms, effective. For countries that have not established their inclusion criteria, and full table of results are available in own threshold, the Commission on Macroeconomics annex 16A. The articles are also graded for quality using and Health (2001) suggests that interventions costing a checklist based on Drummond and others (2005). less than one times the per capita gross national prod- Fewer than 15 articles were found for the cost- uct per DALY averted are very cost-effective and those effectiveness of interventions for breast, colon, liver, between one and three times per capita gross national oral, and pediatric cancers in LMICs—including four product are cost-effective. Cancer in Low- and Middle-Income Countries: An Economic Overview 267 What is very cost-effective in HICs might merit feasible in facilities with Enhanced resources is likely consideration in LICs and what is cost-effective in HICs to be available for urban populations in MICs, and might be considered in MICs. This approach presup- facilities with Maximal resources are broadly available poses that the underlying model is similar, namely, the only in some HICs. Almost all countries, irrespective of interventions are similar and the “no intervention” or income, have some facilities with Maximal resources to “standard care” alternative scenarios are also similar, as which a minority of the population has access or can be are other key parameters. Country-specific data would referred. be better in the future to guide policy. These resource categories are used as an organizing The results summarized in the following section draw framework for five of the six major cancers covered. on the cost-effectiveness analyses in other chapters of The exception is pediatric cancer, which has a similar this volume: Anderson and others (chapter 3), Denny ranking of feasibility but is determined differently. and others (chapter 4), Sankaranarayanan and others Patients with pediatric cancers, which are relatively rare, (chapter 5), Rabeneck and others (chapter 6), Gupta and are often referred to specialized facilities. Specialized others (chapter 7), and Gelband and others (chapter 8). facilities with the least expertise can successfully treat a These analyses, in turn, utilize systematic literature limited range of pediatric cancers; the range increases as surveys for the LMICs described in annex 16A. experience grows. Tables 16.3 through 16.8 summarize by resource environment the feasibility of various interventions for Cost-Effectiveness Results the six cancers considered: breast, cervical, colorectal, Table 16.2 summarizes the specific resources likely to be liver, oral, and pediatric. The cost-effectiveness evidence available at each of the four facility environments: Basic, is provided where it exists, along with the country con- Limited, Enhanced, and Maximal. Cancer care feasible text in which the data were obtained. For surgery, the in facilities with Basic resources is likely to be the norm cost-effectiveness data are virtually nonexistent, even for in LICs, care feasible in facilities with Limited resources HICs, except for new techniques, such as laparoscopic predominates for the rural population in MICs, care surgery. Data are most abundant for pharmaceuticals, Table 16.2 Cancer Care Tools in Four Resource Environments Basic • Resources for organized screening and treatment of precancer conditions do not exist; vertical programs, such as mobile services for screen-and-treat options in one or two visits, may be feasible. • Basic surgery is available but in limited supply; specialized surgery skills may be available only by referral to another facility. • Radiation therapy is very scarce or unavailable. • Chemotherapy is not feasible because of the lack of laboratory facilities for required blood work. Limited • Mobile screening units are an option; rapid DNA testing is possible, if cost is sufficiently low. • Availability of surgery is better but still limited. • Radiation therapy is scarce and patients may need to travel long distances for access. • Chemotherapy may be possible, using off-patent drugs and “classical” therapies; new techniques, such as metronomics, may be considered. Laboratory facilities are limited. • Limited treatment of precancer conditions occurs at lower-level health facilities and first-level hospitals. Enhanced • Organized screening can be considered, along with treatment of precancer conditions at facilities at different levels. • Radiotherapy and surgery are widely available. • Chemotherapy is possible and newer generations of drugs can be considered, although typically not those still on-patent. Laboratory facilities are available on site to support use of chemotherapy. • The most advanced hospitals can offer most of the care options available in high-income countries, but budgets are insufficient to make such care broadly available. Maximal • State-of-the-art treatment is available; however, even in high-income countries, health budgets still require hard choices, and private insurers or public systems may carefully ration access to the most costly therapies. Note: Resource typology based on Breast Health Global Initiative (Anderson and others, chapter 3). Most facilities in low-income countries have Basic levels of resources, facilities in rural areas in middle-income countries generally have Limited resources, most facilities in urban areas of middle-income countries have Enhanced resources, and facilities with Maximal resources are widespread only in some industrial countries. 268 Cancer since these data are often required in HICs as part of the for example, for organized screening. Options that are approval process for new drugs. not recommended in this environment are those that are Interventions listed as options in the Basic envi- not cost-effective even in HICs. ronment are limited to those that the evidence sug- Finally, in the Maximal environment, an even broader gests are very cost-effective in HICS or cost-effective range of options is available, some of which are cost- in LMICs. The range of options is broadened a little effective in those environments. Those that are currently in the Limited environment to include items that are not cost-effective—for example, some new chemother- “close to being very cost-effective” in HICs, or “possibly apy agents—may well eventually become cost-effective cost-effective in LMICs,” and the greater feasibility of once they no longer have patent protection. radiation and chemotherapy options broadens the range for consideration. CONCLUSIONS In the Enhanced environment, more interventions are feasible because of the greater availability of resources Feasibility and cost-effectiveness data suggest that and because a larger percent of the population is cancer care can and should be expanded in LMICs. located in urban areas and able to undertake treatments Table 16.9 summarizes the interventions, by cancer and that require regular visits, for example, for preopera- by resource level, which are supported by feasibility tive radiotherapy, or require more intensive follow-up, and cost-effectiveness data, noting that virtually no Table 16.3 Breast Cancer: Summary of Feasibility and Likely Cost-Effectiveness of Interventions, by Resource Level Resource level Intervention type Intervention details Cost-effectiveness (if available) Basic Detection Clinical history and CBE • CBE cost-effective in Ghana (Zelle and others 2012) Treatment, Modified radical mastectomy; — stages I and II ovarian ablation in premenopausal (stage I) or all (stage II) Test for ER and add tamoxifen if • Either tamoxifen or chemotherapy very cost-effective after surgery in feasible; add chemotherapy (classical United states for younger women (Hillner and Smith 1991; Malin and CMF or AC, EC, or FAC if blood others 2002) chemistry and CBC available) • Tamoxifen very cost-effective in the Republic of Korea (Yang and others 2010) Treatment, locally Same options as for stages I and II; — advanced add preoperative chemotherapy if resources available Limited Detection CBE with diagnostic ultrasound or • Single lifetime CBE very cost-effective in India mammography in target group • Every three years or every five years cost-effective (Okonkwo and others 2008) • Annual CBE very cost-effective in Vietnam (Nguyen and others 2013) Treatment, Breast-conserving surgery; add • Breast-conserving surgery versus modified radical mastectomy cost-effective stages I and II irradiation of chest wall for high-risk in United States (Norum and others 1997) stage II Chemotherapy (classical CMF or AC, • Very cost-effective after surgery in United States for younger women EC, or FAC if blood chemistry and (Hillner and Smith 1991; Malin and others 2002) CBC available) • Cost-effective in United States for older women (Desch and others 1993; Hillner, Smith, and Desch 1993; Malin and others 2002; Naeim and Keeler 2005) Treatment, locally Same options as for stages I and II; — advanced add irradiation of chest wall table continues next page Cancer in Low- and Middle-Income Countries: An Economic Overview 269 Table 16.3 Breast Cancer: Summary of Feasibility and Likely Cost-Effectiveness of Interventions, by Resource Level (continued) Resource level Intervention type Intervention details Cost-effectiveness (if available) Enhanced Detection Mammography every two years for • Mammography every year for women over age 50 years and every two women ages 55–69 years; every years in high-risk women ages 40–49 years cost-effective in Mexico 12–18 months for women ages (Salomon and others 2012) 40–54 years Treatment, Breast-conserving surgery and — stages I and II whole-breast irradiation Aromatase inhibitors or LH-RH • Using raloxifene (taxane) instead of tamoxifen cost-effective for some agonists (hormones) and taxanes women in United States (Armstrong and others 2001) (chemo) to replace tamoxifen and • Using letrozole instead of tamoxifen cost-effective in United States classical chemo, respectively (Delea and others 2007). Using anastrozole (aromatase) instead of tamoxifen cost-effective in Brazil (Fonseca, Araújo, and Saad 2009); very cost-effective in United States (Moeremans and others 2006); not cost-effective in Spain (Gil and others 2006) Treatment, locally Add trastuzumab for HER2/neu+ • Short-course trastuzumab very cost-effective in United States compared advanced disease with usual treatment for women with HER2+ disease (Malin and others 2002). Using letrozole (aromatase) instead of tamoxifen very cost-effective in United States for advanced disease (Dranitsaris, Verma, and Trudeau 2003; Karnon and others 2003) Maximal Detection Mammography every year for women • Mammography every three years in the Republic of Korea for women ages ages 40 years and older 45–65 years not cost-effective (Lee and others 2009) • Mammography every two years in Hong Kong SAR, China, for women ages 40–69 years cost-effective (Wong and others 2007) Treatment, Add trastuzumab for HER2/neu+ • Very cost-effective in Singapore (de Lima Lopes 2011) stages I and II disease • Cost-effective in Belgium (van Vlaenderen and others 2009), Canada (Hedden and others 2012), Italy and United States (Liberato and others 2007), Netherlands (Essers and others 2010) Treatment, stages Add growth factors and dose-dense • Growth factors cost-effective in Japan for high-risk early cancer I and II, locally chemotherapy (Ishiguro and others 2010) advanced • Not cost-effective in United States for early stage (Ramsey and others 2009) Metastatic Use bevacizumab (chemo), • Bevacizumab not cost-effective in United Kingdom (Rodgers and others fulvestrant (hormone), and growth 2011), United States (Montero and others 2012) factors (supportive) • Fulvestrant cost-effective in United Kingdom (Cameron and others 2008) Note: See individual studies for further details, such as age and hormone status, for which cost-effectiveness results were obtained. Screening and treatment typically become less cost-effective for women ages 60 years and older. Resource typology based on Breast Health Global Initiative (Anderson and others, chapter 3). Most facilities in low-income countries have Basic levels of resources, facilities in rural areas in middle-income countries generally have Limited resources, most facilities in urban areas of middle-income countries have Enhanced resources, and facilities with Maximal resources are widespread only in some industrial countries. The table is cumulative, for example, treatments that are feasible in facilities with Basic environments are also feasible in facilities with Limited, Enhanced, and Maximal resources. AC = doxorubicin (Adriamycin) and cyclophosphamide (Cytoxan); CBC = complete blood count; CBE = clinical breast examination; CMF = cyclophosphamide, methotrexate, and fluorouracil; EC = epirubicin and cyclophosphamide; ER = estrogen receptor; FAC = fluorouracil, doxorubicin (Adriamycin), and cyclophosphamide (Cytoxan); LH-RH = luteinizing hormone–releasing hormone; — = not available. cost-effectiveness data are available for surgery. The inter- expansion of HPV vaccine and DNA testing for cervical ventions provide suggestions to policy makers as to the cancer, provided that the costs can be reduced sufficiently. sequence in which to add publicly funded interventions There are methods to reduce the risk of liver cancer. The as country income increases, as part of a cancer plan. large expansion of HBV vaccination is a success story in Table 16.9 suggests that LMICs have cost-effective preventing cancer. options in cancer control. More can be done in all coun- The cost-effectiveness results suggest that a sus- tries in prevention, particularly tobacco control, and tained expansion of cancer treatment is appropriate in 270 Cancer Table 16.4 Cervical Cancer: Summary of Feasibility and Likely Cost-Effectiveness of Interventions, by Resource Level Resource level Intervention type Intervention details Cost-effectiveness (if available) Basic Prevention HPV vaccination of adolescent girls Cost-effective in LMICs if the following three conditions apply: 1. Cost per vaccinated girl is low enough (US$10–25): Gavi-eligible countries can likely achieve this 2. Incidence is high 3. Vaccine protection is long-lasting (chapter 4) Detection VIA starting at age 35 years, one • Very cost-effective (Praditsitthikorn and to three times per lifetime, or rapid others 2011, Thailand) DNA test starting at age 35 years, • Rapid DNA is very cost-effective in LMICs two or three times per lifetime, with screen-and-treat in single visit (Goldie assuming cost per HPV test is less and others 2005) but difficult to undertake than US$10 in practice • Cost-effective in MICs with two visits required (Kim and others 2008; Levin and others 2010) Diagnosis Colposcopy — Treatment, precancer Cryotherapy for suspicious — precancerous lesions Treatment, early-stage cancer LEEP, CKC, simple hysterectomy — Limited Treatment, more advanced Surgery and/or radiation therapy — stage cancer Enhanced Screening Cytology every two to three years, • Cytology may be cost-effective if quality starting at time of initiation of control is good (Kim and others 2008) sexual activity; DNA test 1–3 times • DNA testing cost-effective depending on per lifetime test cost, frequency of testing (Campos and others 2012) Prevent and screen HPV vaccination combined with Can be cost-effective in all countries, depending screening on cost per vaccinated girl and cost of screening strategy chosen (Denny and others, chapter 4) Treatment options Add chemotherapy (cisplatin) where — warranted Maximal Treatment options Trachelectomy (fertility-sparing — surgery), brachytherapy, intensity- modulated radiotherapy Note: Resource typology based on Breast Health Global Initiative (Anderson and others, chapter 3). Most facilities in low-income countries have Basic levels of resources, facilities in rural areas in middle-income countries generally have Limited resources, most facilities in urban areas of middle-income countries have Enhanced resources, and facilities with Maximal resources are widespread only in some industrial countries. The table is cumulative, for example, treatments that are feasible in facilities with Basic environments are also feasible in facilities with Limited, Enhanced, and Maximal resources. CKC = cold knife conization; HPV = human papillomavirus; LEEP = loop electrosurgical excision procedure; LMICs = low- and middle-income countries; MICs = middle-income countries; VIA = visual inspection with acetic acid; — = not available. Cancer in Low- and Middle-Income Countries: An Economic Overview 271 Table 16.5 Colorectal Cancer: Summary of Feasibility and Cost-Effectiveness of Interventions, by Resource Level Resource level Intervention type Intervention details Cost-effectiveness (if available) Basic Prevention Tax cigarettes Very cost-effective for various cancers Detect/diagnose Use barium enema where colonoscopy — not available Treatment, stages I, II, III colon Hemicolectomy and regional lymph — node dissection Treatment, stages I, II, III rectal Abdominal perineal resection with — lymph node dissection Treatment, stage IV colorectal Consider palliative surgery — Limited Treatment, stages I and II colon Hemicolectomy and regional lymph — node dissection Treatment, stage I rectal Abdominal perineal resection with — lymph node dissection Treatment, stage III colon Hemicolectomy and regional lymph • Adjuvant chemotherapy close to being very node dissection plus adjuvant cost-effective in United States versus no 5-fluorouracil adjuvant after surgery for colorectal cancer (Nostrum and others 1997) • Adjuvant chemotherapy close to being very cost-effective in United States versus no adjuvant after surgery for colorectal cancer, colon cancer (Smith and others 1993) Treatment, stages II and III rectal Abdominal perineal resection • Adjuvant chemotherapy close to being very with lymph node dissection plus cost-effective in United States versus no preoperative short course radiotherapy; adjuvant after surgery for colorectal cancer add 5-fluorouracil (Nostrum and others 1997) • Preoperative radiotherapy very cost-effective in United States versus no preoperative radiotherapy (van den Brink and others 2004) Treatment, stage IV colorectal Consider palliative surgery; consider — palliative 5-fluorouracil-based chemotherapy Enhanced Detection Organized screening (beginning with • Very cost-effective in United States (gFOBT) pilot) (Pignone, Russell, and Wagner 2005) • Cost-effective in United States (colonoscopy) (Pignone, Russell, and Wagner 2005) • Very cost-effective in high-income Asia (gFOBT, sigmoidoscopy, colonoscopy) (Park, Yun, and Kwon 2005; Tsoi and others 2008; Wong, Leong, and Leong 2004; Wu and others 2006) Treatment, stages I and II colon Hemicolectomy with en bloc removal of — at least 12 lymph nodes Treatment, stage I rectal Total mesorectal excision — Treatment, stage III colon Hemicolectomy plus removal of at least FOLFOX very cost-effective in United Kingdom 12 lymph nodes plus adjuvant FOLFOX versus 5-fluorouracil plus leucovorin (Aballéa and others 2007) table continues next page 272 Cancer Table 16.5 Colorectal Cancer: Summary of Feasibility and Cost-Effectiveness of Interventions, by Resource Level (continued) Resource level Intervention type Intervention details Cost-effectiveness (if available) Treatment, stages II and III rectal Total mesorectal excision; preoperative — chemo-radiotherapy; capecitabine or infusional 5-fluorouracil; adjuvant FOLFOX Treatment, stage IV colorectal Consider palliative surgery; palliative — radiation; FOLFOX/FOLFIRI; possibly aggressive surgery for cure Maximal Treatment, stages I and II colon Surgery: can use polypectomy for • LAC not cost-effective in United States selected stage I cancers; consider (Hayes and Hansen 2007) adjuvant 5-fluorouracil or capecitabine • Not cost-effective in United States for in high-risk stage II; can consider LAC colorectal cancer (de Verteuil, Hernández, and Vale 2007) Treatment, stage I rectal Total mesorectal excision — Treatment, stage III colon Hemicolectomy plus removal of at least FOLFOX very cost-effective in United Kingdom 12 lymph nodes plus adjuvant FOLFOX versus 5-fluorouracil plus leucovorin (Aballéa and others 2007) Treatment, stages II and III rectal Total mesorectal excision; preoperative — chemo-radiotherapy; capecitabine or infusional 5-fluorouracil; adjuvant FOLFOX Treatment, stage IV colorectal As for stage IV (Enhanced); plus • Cetuximab plus irinotecan versus active or best biological options (bevucizumab, support care not cost-effective in United aflibercept; if K-Ras wild-type States (Starling and others 2007) cetuximab, panitumumab; regorafenib) • Bevucizumab plus irrotecan and 5-fluorouracil plus leucovorin versus irrotecan and 5-fluorouracil plus leucovorin not cost- effective in United States (Tappenden and others 2007) Note: Resource typology based on Breast Health Global Initiative (Anderson and others 2014). Most facilities in low-income countries have Basic levels of resources, facilities in rural areas in middle-income countries generally have Limited resources, most facilities in urban areas of middle-income countries have Enhanced resources, and facilities with Maximal resources are widespread only in some industrialized countries. The table is cumulative, for example, treatments that are feasible in facilities with Basic environments are also feasible in facilities with Limited, Enhanced, and Maximal resources. FOLFIRI = folinic acid, fluorouracil, and irinotecan; FOLFOX = folinic acid, fluorouracil, and oxalipatin; gFOBT = guaiac fecal occult blood test; K-Ras = Kirsten rat sarcoma viral oncogene homolog; LAC = laparoscopically-assisted colectomy; — = not available. Table 16.6 Liver Cancer: Summary of Feasibility and Likely Cost-Effectiveness of Interventions, by Resource Level Resource Intervention level type Intervention details Cost-effectiveness (if available) Basic Prevention Hepatitis B vaccination of neonates (East and Cost-effective/very cost-effective in intermediate- and high- Southeast Asia); infants (Sub-Saharan Africa) prevalence countries (Beutels 2001; Ozawa and others 2012) Aflatoxin reduction through better post-harvest Cost-effective in Guinea (Khlangwiset and Wu, 2010) handling and storage Limited Prevention Hepatitis B vaccination of infants or adolescents Cost-effective/very cost-effective in intermediate- and high- prevalence countries (Beutels 2001; Ozawa and others 2012) Aflatoxin reduction through biocontrol (different Very cost-effective in Nigeria (Wu and Khlangwiset 2010), seed strains) but validation needed; however, analysis did not take account of cost of diffusion of technology table continues next page Cancer in Low- and Middle-Income Countries: An Economic Overview 273 Table 16.6 Liver Cancer: Summary of Feasibility and Likely Cost-Effectiveness of Interventions, by Resource Level (continued) Resource Intervention level type Intervention details Cost-effectiveness (if available) Prevention programs for hepatitis C through Using auto-disposable syringes is very cost-effective in reducing unsafe injections (for example, auto- India (Reid 2012) disposable syringes) and screening blood donors Prevention programs for liver flukes through — education regarding food habits and hygiene Enhanced Prevention Hepatitis B vaccination of infants, children, and Cost-effective in intermediate and high-prevalence countries adolescents (Beutels 2001; Ozawa and others 2012) Treatment Screening for and treatment with praziquantel — for liver flukes in high-prevalence regions Maximal Prevention Hepatitis B vaccination of adolescents Possibly cost-effective or not cost-effective in low- prevalence countries (Beutels 2001) Treatment Hepatitis B virus treatment with antivirals or Possibly cost-effective in HICs (chapter 8) immune system modulators Treatment Hepatitis C: pegylated interferon treatment plus Cost-effective only in select patients in HICs (chapter 8) ribavirin Treatment Various treatments of hepatocellular carcinoma Not cost-effective even in HICs because of poor survival or cholangiocarcinoma even with treatment (chapter 8) Note: Resource typology based on Breast Health Global Initiative (Anderson and others, chapter 3). Most facilities in low-income countries have Basic levels of resources, facilities in rural areas in middle-income countries generally have Limited resources, most facilities in urban areas of middle-income countries have Enhanced resources, and facilities with Maximal resources are widespread only in some industrial countries. The table is cumulative, for example, treatments that are feasible in facilities with Basic environments are also feasible in facilities with Limited, Enhanced, and Maximal resources. HICs = high-income countries; — = not available. Table 16.7 Oral Cancer: Summary of Feasibility and Likely Cost-Effectiveness of Interventions, by Resource Level Resource Intervention level type Intervention details Cost-effectiveness (if available) Basic Prevention Tobacco and alcohol taxes Cost-saving (cannot be separated from impact on other cancers) (chapter 10) Screening Visual screening by trained personnel (for • Cost-effective where prevalence is reasonably high; example, dentist or nurse) of high-risk groups very cost-effective for screening high-risk groups (India: (known or self-declared tobacco and alcohol Subramanian and others 2009) users) or by all at risk (for example, over • Cost-effective in three HICs with prevalence of 30 or more 35 years of age) in high-prevalence regions; per 100,000 in men, age-adjusted population (Netherlands: screening is sporadic rather than organized van der Meij, Bezemer, and van der Waal 2002; United Kingdom: Speight and others 2006; United States: Dedhia and others 2011) Diagnosis Visual inspection; biopsy; X-ray to diagnose — spread Treatment Surgery, no adjuvant treatment, for stages — I, II, III; availability of surgery for oral reconstruction very limited Pain control, stage IV table continues next page 274 Cancer Table 16.7 Oral Cancer: Summary of Feasibility and Likely Cost-Effectiveness of Interventions, by Resource Level (continued) Resource Intervention level type Intervention details Cost-effectiveness (if available) Limited Treatment Add postoperative radiotherapy if indicated — (stage II or III, depending on type and location of tumor) or radical radiotherapy instead of surgery (stage II or II, depending on type/ location of tumor) Palliative radiotherapy, stage IV Enhanced Screening Organized screening is possible Cost-effective (Dedhia and others 2011; Speight and others 2006; Subramanian and others 2009; van der Meij, Bezemer, and van der Waal 2002) Diagnosis CT scan to confirm spread — Treatment Surgery and/or radiotherapy or brachytherapy — and/or off-patent chemotherapy, stages II, III and IV, depending on type/location of tumor; reconstructive surgery possible Palliative chemotherapy, stage IV Maximal Diagnosis PET, MRI to determine spread if bone/soft — tissue potentially involved; can consider chemotherapy with patent drugs Note: Resource typology based on Breast Health Global Initiative (Anderson and others, chapter 3). Most facilities in low-income countries have Basic levels of resources, facilities in rural areas in middle-income countries generally have Limited resources, most facilities in urban areas of middle-income countries have Enhanced resources, and facilities with Maximal resources are widespread only in some industrialized countries. The table is cumulative, for example, treatments that are feasible in facilities with Basic environments are also feasible in facilities with Limited, Enhanced, and Maximal resources. CT = computerized tomography; HICs = high-income countries; MRI = magnetic resonance imaging; PET = positron emission tomography; — = not available. Table 16.8 Pediatric Cancer: Summary of Likely Cost-Effectiveness of Interventions, by Experience Level Experience level Cancer treated Cost-effectiveness (if available) Center with least Burkitt lymphoma Very cost-effective (Malawi: Bhakta and others 2013) expertise Hodgkin lymphoma (adolescents and adults) Very cost-effective (Norway: Norum and others 1996) Center with more Wilms tumor Cost-effective (Brazil: Bhakta and others 2013) expertise Acute lymphoblastic leukemia Very cost-effective (Netherlands: van Litsenburg and others 2011); cost-effective (China: Luo and others 2009) Intraocular retinoblastoma — Center with most Sarcomas, brain tumors, acute myeloid leukemia, — expertise high-risk neuroblastoma, other retinoblastomas Note: Sequencing of cancer is illustrative rather than exhaustive. Feasibility of treatment of pediatric cancer does not follow the same pattern as adult cancers. Pediatric cancers are rare and many low-income countries have used the approach of treatment in specialized centers. Prevention is not an important issue (other than via hepatitis B vaccination); because incidence is very low, population-level screening is not an option. — = not available. all LMICs. The expansion of capacity for surgery and Enhanced facilities predominate. Chemotherapy can radiation is a priority throughout. The use of tamoxifen be expanded to include newer regimens for breast and is feasible and cost-effective for breast cancer in LICs, and cervical cancers and chemotherapy regimens for colon newer hormone treatments can be cost-effective in MICs. and oral cancers. In MICs, the use of classical chemotherapy regimens for As treatment is scaled up, screening is more impor- breast and cervical cancer is cost-effective; in areas where tant to stage-shift treatment. LICs with largely Basic Cancer in Low- and Middle-Income Countries: An Economic Overview 275 Table 16.9 Summary Recommendations Based on Feasibility and Cost-Effectiveness, by Resource Level Cancer Intervention by resource level Primary prevention Tobacco-related Taxation of cigarettes, legislation, regulation (ALL) Cervical HPV vaccine (ALL: cost-effectiveness depends on price) Liver • HBV vaccination integrated with expanded program for immunization (B) • Neonatal (L, E); adolescent (E, M) • Screening blood donors (E, M); reducing unsafe injections (L, E, M) • Education to prevent liver fluke infection (L, E, M) • Aflatoxin reduction: post-harvest storage (B); biocontrol (L) Screening and detection (to stage-shift treatment) Breast • Clinical breast exam (B, L) • Mammography (E, M) Cervical • Visual inspection with acetic acid (B, L) • Rapid DNA test and treat in two visits (L) • DNA test, cytology (E, M) Colorectal Fecal immunochemical test (E); fecal immunochemical or endoscopy (M) Liver cancer Screen and treat for liver flukes in high-prevalence regions (E, M) Oral cancer Visual inspection (L, high-prevalence countries only) Treatment with curative intent Breast • Surgery (ALL); radiation (L, E, M) • Hormones: tamoxifen (B, L); aromatase inhibitors, LH-RH agonists (E); fulvestrant (M) • Chemotherapy: CMF or AC (B); EC or FAC (L); taxanes (E); trastuzumab (E, M) • Growth factors (M); bevacizumab (M) Cervical • Surgery (ALL); trachelectomy (M) • Cryotherapy (B, L); radiotherapy (L, E, M); brachytherapy, intensity modulated brachytherapy (M) • Chemotherapy (cisplatin) (E, M) Colorectal • Surgery: colon (ALL); rectal (L, E, M) • Radiation: preoperative, rectal (L); chemo-radiotherapy preoperative, rectal (E, M) • Chemotherapy: classical 5-fluorouracil (L); FOLFOX (E, M) Liver • Antivirals or immune system modulators for hepatitis B (M) • Hepatitis C (M, cost-effective only in select patients) • Treatment of liver cancer (M, although not generally cost-effective) Oral Surgery (ALL); radiotherapy (L, E, M); brachytherapy (E, M); chemotherapy (E, M) Pediatric • Burkitt lymphoma, Hodgkin lymphoma (specialized center, least expertise) • Wilms tumor (specialized center, more expertise) • Sarcomas, brain tumors, acute myeloid leukemia, high-risk neuroblastoma (specialized center, most expertise) table continues next page 276 Cancer Table 16.9 Summary Recommendations Based on Feasibility and Cost-Effectiveness, by Resource Level (continued) Cancer Intervention by resource level Advanced disease All • Pain control (ALL) • Home or hospice care (ALL) • Palliative radiotherapy (L, E, M, as resources allow) • Palliative surgery (as resources allow) • Palliative chemotherapy (L, classical; E, next generations; M, on-patent, as resources allow) • Aggressive treatment with curative intent (M) Note: Resource typology based on Breast Health Global Initiative (Anderson and others, chapter 3). Most facilities in low-income countries have Basic levels of resources, facilities in rural areas in middle-income countries generally have Limited resources, most facilities in urban areas of middle-income countries have Enhanced resources, and facilities with Maximal resources are widespread only in some industrial countries. Higher resource–level countries can consider any of the options from lower resource levels. The table is cumulative, for example, treatments that are feasible in facilities with Basic environments are also feasible in facilities with Limited, Enhanced, and Maximal resources. AC = doxorubicin (Adriamycin) and cyclophosphamide (Cytoxan); ALL = all resource levels; B = Basic resource level; CMF = cyclophosphamide, methotrexate, and fluorouracil; E = Enhanced resource level; EC = epirubicin and cyclophosphamide; FAC = fluorouracil, doxorubicin (Adriamycin), and cyclophosphamide (Cytoxan); FOLFOX = folinic acid, fluorouracil, and oxalipatin; HBV = hepatitis B virus; HPV = human papillomavirus; L = Limited resource level; LH–RH = luteinizing hormone–releasing hormone; M = Maximal resource level. Recommendations are based on existing cost-effectiveness data and expected availability of resources. Recommendations for basic surgery, radiation therapy, hormone therapy, and classical chemotherapy are based on expert opinion, where cost-effectiveness studies are not available, and are subject to development of infrastructure where it does not yet exist. facilities are not readily able to undertake organized but groups are undertaking cost-effectiveness studies screening. Opportunistic screening in LICs and orga- in Brazil and China, and some literature is not yet cat- nized screening in MICs can help to identify cancer egorized in PubMed. Table 16.9 also does not include earlier to increase the chance of a cure. In LICs, cost- studies after 2007 for HICs, except for breast cancer. effective screening choices include clinical breast exami- It is clear that the literature on cost-effectiveness nation, visual inspection of the cervix with acetic acid, in LMICs is thin. More studies need to be done using and visual inspection for oral cancer in high-prevalence best practice methodology, such that findings can be countries, with rapid DNA test-and-treat for cervical compared across countries. Multi-country studies with cancer potentially feasible as country income increases common assumptions are valuable to help identify or the cost of the test falls. Urban areas in MICs the types of situations where a particular interven- can consider mammography and fecal immunochem- tion is cost-effective. For screening, there are good ical testing for colon cancer, where prevalence patterns multi-country studies for cervical cancer using a com- dictate. MICs can screen in rural areas for liver flukes, mon model, but almost none for the other cancers. if prevalent. Although the WHO’s Choosing Interventions That Are For the common pediatric cancers, a case can be Cost-Effective multi-country work has been done for made for centralizing treatment, either in-country or cervical, breast, and colon cancer screening (for exam- in-region. Evidence suggests that many pediatric can- ple, Ginsberg and others 2012), this needs to be updated cers can be treated cost-effectively, even in LMICs, and using state-of-the-art models similar to the large ones scale-up is feasible. used in HICs. Future economics work on cancer is to To support countries as they develop cancer plans, cost out the ingredients required for the priority inter- more work on costing is needed. Experience with other ventions, such that costs of resource-appropriate care global health concerns facing LMICs (for example, HIV- can be estimated in individual countries. This approach AIDS and nutrition) suggests that credible estimates of can help countries to plan for and mobilize the resources total costs are important. These estimates can help to needed to implement a cancer plan. convince the international community that action is possible and may motivate the substantial mobilization of resources required. Estimating resource requirements ACKNOWLEDGMENTS will be a key next step for the global fight against cancer. Further research is needed to validate the recommen- The authors thank Hellen Gelband, Carol Levin, Joe dations for cervical, colorectal, liver, oral, and pediatric Lipscomb, and Rachel Nugent for helpful suggestions cancer made in table 16.9, including expert consulta- on earlier drafts, and Benjamin Anderson, Craig Earle, tions and updating of the systematic literature reviews. Surendra Shastri, and Rengaswamy Sankaranarayanan The literature search was conducted only in English, for useful inputs regarding specific cancers. Cancer in Low- and Middle-Income Countries: An Economic Overview 277 ANNEX 16A vol. IX. Scientific Publications 160. Lyon: International Agency for Research on Cancer. The annex to this chapter is as follows. It is available at http:// de Lima Lopes, G. 2011. “Societal Costs and Benefits of www.dcp-3.org/cancer. Treatment with Trastuzumab in Patients with Early • Annex 16A. Search terms, Inclusion Criteria, and Results HER2neu-Overexpressing Breast Cancer in Singapore.” BMC Cancer 11: 178–85. de Verteuil, R. M., R. A. Hernández, and L. Vale. 2007. “Economic Evaluation of Laparoscopic Surgery for Colorectal Cancer.” NOTE International Journal of Technology Assessment in Health World Bank income classifications as of July 2014 are as Care 23 (4): 464–72. follows, based on estimates of gross national income per capita Dedhia, R. C., K. J. Smith, J. T. Johnson, and M. Roberts. 2011. for 2013: “The Cost-Effectiveness of Community-Based Screening for Oral Cancer in High-Risk Males in the United States: • Low-income countries: US$1,045 or less A Markov Decision Analysis Approach.” Laryngoscope • Middle-income countries: 121 (5): 952–60. • Lower-middle-income: US$1,046–US$4,125 Delea, T.E., J. Karnon, O. Sofrygin, S. K. Thomas, N. L. Papo • Upper-middle-income: US$4,126–US$12,745 and V. Barghout. 2007. “Cost-effectiveness of Letrozole • High-income countries: US$12,746 or more versus Tamoxifen as Initial Adjuvant Therapy in Hormone- receptor-positive Postmenopausal Women with Early-stage Breast Cancer.” Clinical Breast Cancer 7 (8): 608–18. Desch, C. E., B. E. Hillner, T. J. Smith, and S. M. Retchin. 1993. REFERENCES “Should the Elderly Receive Chemotherapy for Node- Aballéa, S., A. Boler, A. Craig and H. Wasan. 2007. “An economic Negative Breast Cancer? A Cost-Effectiveness Analysis evaluation of oxaliplatin for the Adjuvant Treatment of Examining Total and Active Life-Expectancy Outcomes.” Colon Cancer in the United Kingdom.” European Journal of Journal of Clinical Oncology 11 (4): 777–82. Cancer 43(11): 1687–93. Dranitsaris, G., S. Verma and M. Trudeau. 2003. “Cost Utility Armstrong, K., T. M. Chen, D. Albert, T. C. Randall, and Analysis of First-line Hormonal Therapy in Advanced J. S. Schwartz. 2001. “Cost-Effectiveness of Raloxifene Breast Cancer: Comparison of Two Aromatase Inhibitors and Hormone Replacement Therapy in Postmenopausal to Tamoxifen.” American Journal of Clinical Oncology 26(3): Women: Impact of Breast Cancer Risk.” Obstetrics 289–96. Gynecology 98 (6): 996–1003. Drummond, M. F., M. J. Schulpher, G. W. Torrance, Avksentyeva, M. 2010. “Colorectal Cancer in Russia.” European D. J. O’Brien, and G. L. Stoddart. 2005. Methods for Journal of Health Economics 1: S91–S98. the Economic Evaluation of Health Care Programmes. 3rd Beutels, P. 2001. “Economic Evaluations of Hepatitis B edition. New York: Oxford University Press. Immunization: A Global Review of Recent Studies Essers, B. A., S. C. Seferina, V. C. Tjan-Heijnen, J. L. Severens, (1994–2000).” Health Economics 10: 751–74. A. Novák, and others. 2010. “Transferability of Model- Bhakta, N., A. L. C. Martiniuk, S. Gupta, and S. C. Howard. Based Economic Evaluations: The Case of Trastuzumab 2013. “The Cost-Effectiveness of Treating Paediatric Cancer for the Adjuvant Treatment of HER2-Positive Early Breast in Low-Income and Middle-Income Countries: A Case- Cancer in the Netherlands.” Value in Health 13 (4): 375–80. Study Approach Using Acute Lymphocytic Leukaemia Fonseca, M., G. T. B. Araújo, and E. D. Saad. 2009. “Cost- in Brazil and Burkitt Lymphoma in Malawi.” Archives of Effectiveness of Anastrozole, in Comparison with Disease in Children 98 (2): 155–60. Tamoxifen, in the Adjuvant Treatment of Early Breast Cameron, D. A., D. R. Camidge, J. Oyee, and M. Hirsch. 2008. Cancer in Brazil.” Revista Da Associação Medica Brasileira “Economic Evaluation of Fulvestrant as an Extra Step in 55: 410–15. the Treatment Sequence for ER-Positive Advanced Breast Funk, L. M., T. G. Weiser, W. R. Berry, S. R. Lipsitz, A. F. Merry, Cancer.” British Journal of Cancer 99 (12): 1984–90. and others. 2010. “Global Operating Theatre Distribution Campos, N. G., J. J. Kim, P. E. Castle, J. Ortendahl, and Pulse Oximetry Supply: An Estimation from Reported M. O’Shea, and others. 2012. “Health and Economic Impact Data.” Lancet 376 (9746): 1055–61. of HPV 16/18 Vaccination and Cervical Cancer Screening Gil, J. M., C. Rubio-Terres, C.A. Del, P. Gonzalez and F. Canorea. in Eastern Africa.” International Journal of Cancer 130 (11): 2006. “Pharmacoeconomic Analysis of Adjuvant Therapy 2672–84. with Exemestane, Anastrozole, Letrozole or Tamoxifen Commission on Macroeconomics and Health. 2001. WHO. in Postmenopausal Women with Operable and Estrogen Macroeconomics and Health: Investing in Health for Economic Receptor-Positive Breast Cancer” Clinical Transl Oncology Development. Geneva: World Health Organization. http:// 8 (5): 339–48. whqlibdoc.who.int/publications/2001/924154550x.pdf. Ginsberg, G. M., J. A Lauer, S. Zelle, S. Baetan and R. Baltussen. Curado, M. P., B. Edwards, H. R. Shim, H. Storm., J. Ferlay, 2012. “Cost Effectiveness of Strategies to Combat Breast, and others, eds. 2007. Cancer Incidence in Five Continents, Cervical, and Colorectal Cancer in sub-Saharan Africa and 278 Cancer South East Asia: Mathematical Modelling Study.” British Levin, C. E., J. Sellors, J.-F. Shi, L. Ma, Y.-L. Qiao, J. Ortendahl, Medical Journal 344: eb14. doi:10.1136/bmj.e614. M. K. H. O’Shea, and S. J. Goldie. 2010. “Cost-effectiveness Goldie, S. J., L. Gaffikin, J. D. Goldhaber-Fiebert, A. Gordillo- Analysis of Cervical Cancer Prevention based on a Rapid Tobar, C. Levin, and others. 2005. “Cost-Effectiveness of Human Papillomavirus Screening Test in a High-Risk Region Cervical-Cancer Screening in Five Developing Countries.” of China.” International Journal of Cancer 127: 1404–11. New England Journal of Medicine 353: 2158–68. Liberato, N. L., M. Marchetti, and G. Barosi. 2007. “Cost- Greenberg, D., C. Earle, C. H. Fang, A. Eldar-Lissai, and Effectiveness of Adjuvant Trastuzumab in Human P. J. Neumann. 2010. “When Is Cancer Care Cost-Effective? Epidermal Growth Factor Receptor 2-Positive Breast A Systematic Overview of Cost-Utility Analyses in Cancer.” Journal of Clinical Oncology. 25(6): 625–33. Oncology.” Journal of the National Cancer Institute 102 (2): Luo, X. Q., Z. Y. Ke, L. B. Huang, X. Q. Guan, Y. C. Zhang, 82–88. and others. 2009. “Improved Outcome for Chinese Hayes, J. L., and P. Hansen. 2007. “Is Laparoscopic Colectomy Children with Acute Promyelocytic Leukemia: A Comparison for Cancer Cost-Effective Relative to Open Colectomy?” of Two Protocols.” Pediatric Blood and Cancer 53: 325–28. Australia and New Zealand Journal of Surgery 77 (9): 782–86. Malin, J. L., E. Keeler, C. Wang, and R. Brook. 2002. “Using Hedden, L., S. O’Reilly, C. Lohrisch, S. Chia, C. Speers, and Cost-Effectiveness Analysis to Define a Breast Cancer others. 2012. “Assessing the Real-World Cost-Effectiveness Benefits Package for the Uninsured.” Breast Cancer Research of Adjuvant Trastuzumab in HER-2/neu Positive Breast and Treatment 74 (2): 143–53. Cancer.” Oncologist 17 (2): 164–71. Montero, A. J., K. Avancha, S. Glück, and G. Lopes. 2012. Hillner, B. E., and T. J. Smith. 1991. “Efficacy and Cost “A Cost-Benefit Analysis of Bevacizumab in Combination Effectiveness of Adjuvant Chemotherapy in Women with with Paclitaxel in the First-Line Treatment of Patients Node-Negative Breast Cancer. A Decision-Analysis Model.” with Metastatic Breast Cancer.” Breast Cancer Research and New England Journal of Medicine 324 (3): 160–68. Treatment 132 (2): 747–51. ———. 1993. “Assessing the Cost Effectiveness of Adjuvant Naeim, A., and E. B. Keeler. 2005. “Is Adjuvant Therapy Therapies in Early Breast Cancer Using a Decision Analysis for Older Patients with Node (-) Early Breast Cancer Model.” Breast Cancer Research and Treatment 25 (2): Cost-Effective?” Critical Reviews in Oncology/Hematology 97–105. 53 (1): 81–89. IAEA (International Atomic Energy Authority). 2013. Nguyen, L. H., W. Laohasiriwong, J. F. Stewart, P. Wright, “Radiotherapy Availability.” Programme of Action for Y. T. B. Nguyen, and P. C. Coyte. 2013. “Cost-effectiveness Cancer Therapy (PACT), Programme Office. http://cancer Analysis of a Screening Program for Breast Cancer in .iaea.org/agart.asp. Vietnam.” Value in Health Regional Issues 2: 21–28. Ishiguro, H., M. Kondo, S. L. Hoshi, M. Takada, S. Nakamura, Norum, J., V. Angelsen, E. Wist, and J. A. Olsen. 1996. and others. 2010. “Economic Evaluation of Intensive “Treatment Costs in Hodgkin’s Disease: A Cost-Utility Chemotherapy with Prophylactic Granulocyte Colony- Analysis.” European Journal of Cancer 32A (9): 1510–17. Stimulating Factor for Patients with High-Risk Early Breast Norum, J., J. A. Olsen and E. A. Wist. 1997. “Lumpectomy or Cancer in Japan.” Clinical Therapeutics 32 (2): 11–26. mastectomy? Is breast conserving surgery too expensive?” Karnon, J., R. Delea, S. R. Johnston, R. Smith, J. Brandman, Breast Cancer Research and Treatment 45 (1): 7–14. J. Sung and others 2003. “Cost-effectiveness of Extended Okonkwo, Q. L., G. Draisma, A. der Kinderen, M. L. Brown, Adjuvant Letrozole in Postmenopausal Women after and H. J. de Koning. 2008. “Breast Cancer Screening Policies Adjuvant Tamoxifen Therapy: the UK Perspective. in Developing Countries: A Cost-Effectiveness Analysis for Pharmacoeconomics 24 (3): 237–50. India.” Journal of the National Cancer Institute 100 (18): Kim, J. J., K. E. Kobus, M. Diaz, V. Van Minh, and S. J. Goldie. 1290–1300. 2008. “Exploring the Cost-Effectiveness of HPV Vaccination Ozawa, S., A. Mirelman, M. L. Stacka, D. G. Walker, and in Vietnam: Insights for Evidence-Based Cervical Cancer O. S. Levine. 2012. “Cost-Effectiveness and Economic Prevention Policy.” Vaccine 26: 4015–24. Benefits of Vaccines in Low- and Middle-Income Countries: Khlangwiset, P. and F. Wu. 2010. “Costs and Efficacy of Public A Systematic Review.” Vaccine 31: 108–96. Health Interventions to Reduce Aflatoxin-induced Human Park, S. M., Y. H. Yun, and S. Kwon. 2005. “Feasible Economic Disease.” Food Additives and Contaminants Part A Chemical Strategies to Improve Screening Compliance for Colorectal Analysis, Control of Exposure and Risk Assessment 27(7): Cancer in Korea.” World Journal of Gastroenterology 998–1014. 11: 1587–93. Knaul, F. M., E. E. González-Pier, O. Gómez-Dantés, D. García- Pignone, M., L. Russell, and J. Wagner, eds. 2005. Economic Junco, H. Arreola-Ornelas, and others. 2012. “The Quest for Models of Colorectal Cancer Screening in Average-Risk Adults. Universal Health Coverage: Achieving Social Protection for Washington, DC: National Academies Press. All in Mexico.” Lancet 350: 1259–79. Praditsitthikorn, N., Y. Teerawattananon, S. Tantivess, Lee, S. Y., S. H. Jeong, Y. N. Kim, J. Kim, D. R. Kang, and others. S. Limwattananon, A. Riewpaiboon, and others. 2011. 2009. “Cost-Effective Mammography Screening in Korea: “Economic Evaluation of Policy Options for Prevention and High Incidence of Breast Cancer in Young Women.” Cancer Control of Cervical Cancer in Thailand.” Pharmacoeconomics Science 100: 1105–1111. 29 (9): 781–806. Cancer in Low- and Middle-Income Countries: An Economic Overview 279 Ramsey, S. D., Z. Liu, R. Boer, S. D. Sullivan, J. Malin, “Cost-Utility Analysis of Preoperative Radiotherapy in and others. 2009. “Cost-Effectiveness of Primary versus Patients with Rectal Cancer Undergoing Total Mesorectal Secondary Prophylaxis with Pegfilgrastim in Women with Excision: A Study of the Dutch Colorectal Cancer Group.” Early-Stage Breast Cancer Receiving Chemotherapy.” Value Journal of Clinical Oncology 22 (2): 244–53. in Health 12 (2): 217–25. van der Meij, E. H., P. D. Bezemer, and I. van der Waal. Reid, S. 2012. “Estimating the Burden of Disease from Unsafe 2002. “Cost-Effectiveness of Screening for the Possible Injections in India: A Cost-Benefit Assessment of the Auto- Development of Cancer in Patients with Oral Lichen Disable Syringe in a Country with Low Blood-Borne Virus Planus.” Community Dentistry and Oral Epidemiology Prevalence.” Indian Journal of Community Medicine 37 (2): 30 (5): 342–52. 89–94. doi:10.4103/0970-0218.96093. van Litsenburg, R. R., C. A. Uyl-de Groot, H. Raat, G. J. Kaspers, Rodgers, M., M. Soares, D. Epstein, H. Yang, D. Fox, and and R. J. Gemke. 2011. “Cost-Effectiveness of Treatment others. 2011. “Bevacizumab in Combination with a of Childhood Acute Lymphoblastic Leukemia with Taxane for the First-Line Treatment of HER2-Negative Chemotherapy Only: The Influence of New Medication Metastatic Breast Cancer.” Health Technology Assessment and Diagnostic Technology.” Pediatric Blood and Cancer 1: 1–12. 57 (6): 1005–10. Salomon, J. A., N. Carvalho, C. Gutiérrez-Delgado, R. Orozco, van Vlaenderen, I., J. L. Canon, V. Cocquyt, G. Jerusalem, A. Mancuso, and others. 2012. “Intervention Strategies J. P. Machiels, and others. 2009. “Trastuzumab Treatment to Reduce the Burden of Non-Communicable Diseases of Early Stage Breast Cancer Is Cost-Effective from the in Mexico: Cost Effectiveness Analysis.” British Medical Perspective of the Belgian Health Care Authorities.” Acta Journal 344: e355. Clinica Belgica 64 (2): 100–12. Sloan, F. A., and H. Gelband, eds. 2007. Opportunities for Vos, T., R. Carter, J. Barendregt, C. Mihalopoulos, L. Veerman, Cancer Control in Low- and Middle-Income Countries. A. Magnus, L. Cobiac, M. Bertram, and A. Wallace. 2010. Washington, DC: National Academies Press. Assessing Cost-Effectiveness in Prevention: Final Report. Smith, R. D., J. Hall, H. Gurney, and P. R. Harnett. 1993. University of Queensland, Brisbane and Deakin University, “A Cost-Utility Approach to the Use of 5-Fluorouracil Melbourne. and Levamisole as Adjuvant Chemotherapy for Dukes’ C WHO (World Health Organization). 2012. Assessing Colonic Carcinoma.” Medical Journal of Australia 158 (5): National Capacity for the Prevention and Control of 319–22. Noncommunicable Diseases: Report of the 2010 Global Speight, P. M., S. Palmer, D. R. Moles, M. C. Downer, Survey. Geneva: WHO. D. H. Smith, and others. 2006. “The Cost-Effectiveness Wong, I. O. L., K. M. Kuntz, B. J. Cowling, C. L. K. Lam, and of Screening for Oral Cancer in Primary Care.” Health G. M. Leung. 2007. “Cost Effectiveness of Mammography Technology Assessment 10 (14): 1–144. Screening for Chinese Women.” Cancer 110 (885): 895. Starling, N., D. Tilden, J. White, and D. Cunningham. 2007. Wong, S. S., A. P. K. Leong, and T. Z. Leong. 2004. “Cost- “Cost-Effectiveness Analysis of Cetuximab/Irinotecan vs Effectiveness Analysis of Colorectal Cancer Screening Active/Best Supportive Care for the Treatment of Metastatic Strategies in Singapore: A Dynamic Decision Analytic Colorectal Cancer Patients Who Have Failed Previous Approach.” Studies in Health Technology Information 107 Chemotherapy Treatment.” British Journal of Cancer 96 (2): (Pt 1): 104–10. 206–12. World Bank. 2013. World Databank. World Bank, Subramanian, S., R. Sankaranarayanan, B. Bapat, T. Somanathan, Washington, DC. http://databank.worldbank.org/data G. Thomas, and others. 2009. “Cost-Effectiveness of Oral /home.aspx. Cancer Screening: Results from a Cluster Randomized Wu, G. H.-M., Y.-M. Wang, A. M.-F. Yen, J.-M. Wong, H.-C. Lai, Controlled Trial in India.” Bulletin of the World Health and others. 2006. “Cost-Effectiveness Analysis of Colorectal Organization 87 (3): 200–06. Cancer Screening with Stool DNA Testing in Intermediate- Tappenden, P., R. Jones, S. Paisley, and C. Carroll. 2007. Incidence Countries.” BMC Cancer 6: 136–48. “The Cost-Effectiveness of Bevacizumab in the First-Line Yang, J. J., S. K. Park, L. Y. Cho, W. Han, B. Park, and Treatment of Metastatic Colorectal Cancer in England and others. 2010. “Cost-Effectiveness Analysis of 5 Years of Wales.” European Journal of Cancer 17 (2487): 2494. Postoperative Adjuvant Tamoxifen Therapy for Korean Tsoi, K. K. F., S. S. M. Ng, M. C. M. Leung, and J. J. Y. Sung. 2008. Women with Breast Cancer: Retrospective Cohort Study “Cost-Effectiveness Analysis on Screening for Colorectal of the Korean Breast Cancer Society Database.” Clinical Neoplasm and Management of Colorectal Cancer in Therapeutics 32: 1122–38. Asia.” Alimentary Pharmacology and Therapeutics 28 (3): Zelle, S. G., K. M. Nyarko, W. K. Bosu, M. Aikins, L. M. Niëns, 353–63. and others. 2012. “Costs, Effects and Cost-Effectiveness of van den Brink, M., W. B. van den Hout, A. M. Stiggelbout, Breast Cancer Control in Ghana.” Tropical Medicine and E. Klein Kranenbarg, C. A. Marijnen, and others. 2004. International Health 17: 1031–43. 280 Cancer Chapter 17 Financing Cancer Care in Low-Resource Settings Felicia Knaul, Susan Horton, Pooja Yerramilli, Hellen Gelband, and Rifat Atun INTRODUCTION diagnosis, treatment, survivorship, rehabilitation, and palliative care and pain control (Hewitt, Greenfield, Cancer accounts for a rapidly growing health and and Stovall 2005; Knaul, Alleyne, and others 2012). economic burden in low- and middle-income coun- The failure to adequately manage one of the components tries (LMICs) (Knaul and others 2014). The long-term can jeopardize the entire response, resulting in prema- nature of chronic and noncommunicable diseases that ture deaths, unnecessary pain, and wasted resources. characterizes many cancers inflicts repeated financial Although responding to all facets of the continuum is onslaughts on families, intensifying the poverty-illness a daunting task, several countries have included cancer cycle. Inadequately treated illnesses deepen poverty, care in recent reforms designed to achieve universal leading to a cycle of loss of health, lack of treatment, health coverage (UHC); these reforms provide useful higher morbidity, lost income, and deeper impoverish- lessons for other countries. ment (Atun and Knaul 2012). This chapter analyzes one health system function— Many LMICs are working to achieve greater, and financing—in relation to cancer, focusing on treat- even universal, financial protection in health care, with ment. The analysis draws on experiences from several funding from domestic sources that combines public middle-income countries (MICs) in which domestic insurance and prepayment. Establishing universal finance is used and efforts are underway to achieve entitlement to key services through guaranteed ben- universal coverage. We draw lessons for other com- efits packages is a cornerstone of these efforts. These ponents of cancer care and control and highlight the countries face challenges as they strive to include cancer importance of developing strategies for financing that and other chronic and noncommunicable diseases in consider all aspects of the care continuum and strength- the package of covered services. The inclusion of cancer ening of health systems. interventions poses a specific set of challenges because Our analysis focuses on how domestic sources of of the chronic nature of the illness and the high costs funding are deployed to finance cancer care; we leave of treatment. for later work the issues of how these funds are sourced An effective response to cancer requires strengthen- and collected. Domestic funding in the vast majority ing all health system functions—stewardship, financing, of LMICs does, and will inevitably continue to, pay service provision and delivery, and resource generation— for the bulk of cancer care. We do not focus on global along the entire, six-component, care-control and regional financing and platforms; this is a topic for continuum—primary and secondary prevention, future research. These platforms are especially important Corresponding author: Felicia Knaul, PhD, Director, Miami Institute for the Americas, and Professor, Miller School of Medicine, University of Miami, Florida, United States, fknaul@miami.edu. 281 sources of finance for the poorest countries, for cata- derive over 30 percent of annual inpatient expenditures lyzing discovery and innovation and for aggregating from borrowing and asset sales, which is significantly demand to reduce the costs of medicines and vaccines. greater than the reliance on these funding mechanisms by unaffected households (Mahal and others 2013). In South Asia, the probability of incurring catastrophic FINANCIAL PROTECTION, HEALTH health expenditure from hospitalization is 1.6 times FINANCING, AND CANCER CARE1 higher for cancer than for a communicable disease, such as pneumonia (Engelgau and others 2010; Nikolic, The set of diseases that we call cancer leads to some of the Stanciole, and Zaydman 2011). most problematic financial issues in providing care for One of the most insidious aspects of this illness- chronic and noncommunicable diseases. Some cancers impoverishment cycle is that for many patients, out-of- can be prevented by changing behaviors or by con- pocket spending is wasted because it contributes nothing trolling cancer-associated infections. For other cancers to improved health. Especially in LMICs, cancer is often whose causes are unknown, the only effective control detected so late that even the most effective treatment comes from early detection and treatment. Even where will not effect a long-term cure. Second, a substantial the causes are known and somewhat controllable, early proportion of what is spent by patients buys low-quality detection and treatment remain the best course for can- or inappropriate care that is ineffective. Third, care may cers that are not prevented. For some cancers, especially be coupled with prohibitive transportation costs and those detected at later stages, even the most advanced investments of time that include long waits to access treatments are not effective and palliation is the appro- care. These difficulties are more likely to occur with a priate course of action (Gralow and others 2012). disease such as cancer, where care often requires repeated Cancer often requires relatively expensive, com- travel and months-long treatment. plex, multimodal medical treatment for extended peri- ods, leading to household impoverishment, treatment abandonment, and, too often, poor outcomes, espe- Cancer Care Financing cially if the disease is detected at a later stage or In all LMICs, most of the financing for cancer care patients cannot adhere to a full regime of treatment. and control is, and will continue to be, from domestic Yet many interventions for cancer are both effective and sources. This is especially true for MICs where external cost-effective according to today’s global metrics. Recent financing is 1 percent or less of total health expenditure. discoveries have made it possible to prevent several of An important exception is the poorest and most aid- the infection-associated cancers that disproportionately dependent countries. However, even countries as poor as affect poor people because of their exposure to commu- Ethiopia, Haiti, and Niger rely on domestic funding for nicable diseases and lack of access to early detection of more than 50 percent of their total health expenditures. precancerous lesions. For example, vaccinating against The World Health Organization (WHO) estimated that the human papillomavirus (HPV) can prevent the most in 2008, external sources covered 16.4 percent of total common cervical cancers, the vast majority of which health expenditure in LMICs (WHO 2011a). occur in LMICs. Domestic finance of health care comes from two primary sources: (1) out-of-pocket spending by families, either at the point of service or via private insurance Need for Financial Protection (the latter being much less common in LMICs), and Acute care costs, even for simple ailments, can push (2) public spending for health or broader social pro- already poor families deeper into poverty. The repeated tection organized as public insurance. Out-of-pocket and ongoing costs of a chronic illness are more spending by families is the least equitable and efficient devastating. India provides an example of the sub- means of financing health systems and often leads to stantial financial vulnerability of households to non- impoverishment. Out-of-pocket expenditures as a share communicable diseases, especially cancer. In India, the of total health expenditures is highest in LMICs— share of out-of-pocket health expenditure devoted to about 50 percent—and lowest in high-income countries noncommunicable diseases increased from 33 percent (HICs), averaging 14 percent (World Bank 2013). By to almost 50 percent from 1995 to 2004. The cost of contrast, public financing or insurance schemes that a single stay for cancer or heart disease in a public enable prepayment and pooling offer financial protection hospital is the equivalent of 40–50 percent of gross from excessive expenditures for health care and can create domestic product (GDP) per capita (Mahal, Karan, and more effective and equitable ways of organizing health Engelgau 2010). As a result, cancer-affected households system financing (Knaul and others 2006; WHO 2010). 282 Cancer The movement toward UHC is a transition to pooled, Two major paths of progressive universalism have been publicly financed health care that offers financial pro- identified, both of which use prepayment and pooling tection to families and constitutes the scaffolding that of funds to extend publicly financed insurance. The will support cancer coverage in LMICS. Achieving UHC first route drives the expansion of population coverage involves a process with overlapping stages, beginning and targets the poor by insuring health interventions with enrollment and legal coverage, which entitles all for diseases that place a high burden on this group, people to health services funded by publicly organized with no co-payment for anyone. The second variant insurance. The second stage is coverage that seeks to begins with a larger package for the poor. The defini- guarantee access to a comprehensive package of health tion of the package is pivotal and based on burden of services. The third stage is universal effective coverage disease. Recommendations include highly cost-effective that guarantees the maximum attainable health results interventions for infectious diseases and reproductive, from an appropriate package of high-quality services for maternal, newborn, and child health, as well as chronic the evolving health needs of a population. UHC implies conditions and noncommunicable diseases. For cancer, financial protection that promotes equity and efficiency the package includes interventions for prevention, early and reduces the risk of financial shocks to families by detection, treatment, and palliation, focusing on those reducing out-of-pocket payments (Knaul, González- cancers of highest burden and interventions of greatest Pier, and others 2012). potential effectiveness, especially for the poor (Jamison Financial protection toward UHC can expand in and others 2013). three ways and is often tied to growth in resources allo- In practice, countries have tended to apply a cated to health and overall growth of country income: combined or iterative approach, depending on the point of departure to UHC. The point of departure • Expansion of prepayment and risk pooling over time is often a political issue and largely determined by to cover entire populations, in some cases on a group- existing institutional arrangements and the availabil- by-group basis ity of resources. Mexico’s Seguro Popular design, for • Provision of a more comprehensive benefit package example, is based on universal population coverage of health interventions and covered conditions with no co-payment for community services, sliding • Expansion of risk pooling and financial risk pro- scale prepayment for personal health services that tection through the elimination of out-of-pocket exempts the poor, and universal population coverage expenses at the point of service delivery for the poor without prepayment for catastrophic illness; all of and for those interventions considered of high value these elements are anchored in an expanding benefit where use should not be deterred (Jamison and package of cost-effective services that includes an others 2013) increasing number of cancers. A related approach focused on enhancing equity in Turkey has been ana- These three dimensions of coverage are summa- lyzed in detail (Atun and others 2013). rized in WHO’s (2010) financing “cube” as height, Cancer—because it encompasses a set of chronic and breadth, and depth. For a health system to achieve uni- complex diseases—challenges the limits of UHC and versal coverage, the height (proportion of the service the pathways to progressive universalism. A defining cost covered), breadth (covered services), and depth characteristic of most cancers and many other chronic (proportion of the population covered) must be taken diseases is the need, on a population level, for a series into account (WHO 2010). The goal of UHC, according of interventions along the care-control continuum and to WHO, is to ensure that all people are able to obtain illness lifecycle. the health services that they need without suffering Analyzing the extent to which effective interventions financial hardship because they cannot afford to pay for for specific cancers are covered along the continuum them (WHO 2012, 2013). provides insights into the depth and breadth of the over- all package, as well as the balance between prevention and treatment. To ensure effective coverage, the benefits Country Approaches package needs to be guaranteed with permanent reve- The country-specific roadmap to UHC can take several nue sources and capacity-building commitments. Low routes. One approach that has been strongly advocated effective coverage—particularly of early detection— in the literature is what Gwatkin and Ergo term “progres- is common, even in countries with relatively com- sive universalism” (2011), which refers to the determina- plete benefit packages. This situation compromises tion to include people who are poor from the outset as final outcomes (Knaul, Chertorivski Woldenberg, and programs and policies to promote UHC are introduced. Arreola-Ornelas 2012). Financing Cancer Care in Low-Resource Settings 283 PATHS TO UNIVERSAL HEALTH CARE facilities, while all other groups (nonsalaried workers, FINANCING the unemployed, those outside the labor force, and agricultural workers, all of whom tend to be poorer) are Countries are following different paths to UHC. Some limited to usually lower quality, public facilities (or those countries, mostly of middle income, in Latin America provided by nongovernmental organizations) that may and the Caribbean (Colombia, the Dominican Republic, have user fees and that often ration care by availability Mexico, and Peru, for example) have extended public and expertise. In such cases, medication costs are fre- insurance to nonsalaried workers, the unemployed, those quently paid out of pocket. out of the labor force, and the poor; these countries are The path from this pattern of segmented coverage making adjustments to equalize benefits across groups. to universal coverage has varied. Canada and many Thailand has followed a similar path, beginning in 2002, countries in Europe (the United Kingdom and the and India is beginning this process. China has extended Nordic countries, for example) rely heavily on general coverage of the national medical insurance program taxation revenue to finance the system; others more widely, but with a high co-payment and no coverage for strongly emphasize the contribution of private health catastrophic expenditures. LMICs in Sub-Saharan Africa insurance, either voluntary or mandatory (Singapore (such as Ghana) face much greater resource constraints, and the United States). Some countries, for example, and UHC tends to be associated with a more restricted Germany, have brought together coverage of distinct package of services. We illustrate some of the differences groups to reach comprehensive coverage. and similarities among countries in their paths to UHC The Lancet Commission on Investing in Health eval- by considering the specific case of introducing coverage uated the extent to which the path toward increasing for cancer, focusing on treatment. coverage of different groups is universal and progressive This section draws on case studies of eight countries: (Jamison and others 2013). In Latin America and the China (Yerramilli and Jiang 2013), Colombia (Guerrero, Caribbean, where health care provision has been highly Amaris, and Yerramilli 2013), the Dominican Republic segmented between those covered by social security and (Rathe, Knaul, and Yerramilli 2013), Ghana (Yerramilli those not, several countries have moved to invest in and Ataguba 2013), India (Yerramilli 2013), Mexico publicly financed programs to extend pooled coverage, (Knaul, Chertorvski Woldenberg, and Arreola-Ornelas focusing on the poor and nonsalaried workers, and to 2012), Peru (Seinfeld and Pleic 2013), and Thailand reduce coverage differentials progressively. (Yerramilli and Firestone 2013). The four case studies Colombia adopted a universal social insurance plan from Latin America and the Caribbean are updates of in the 1990s, with gradual implementation, and reached earlier case studies from Atun and Knaul (2012). Salient universal coverage in 2011. This approach combined details for the eight countries are summarized in table the contributory plan for the formal sector (including 17.1. Common themes and lessons emerge from these the self-employed) with coverage for the poor and the experiences. Each country faces the challenge of includ- informal sector. The cost for the subsidized scheme ing chronic, catastrophic illnesses such as cancer in the is partly funded through general taxation, with some package for rich and poor. cross-subsidization from contributions from salaried workers and employers, with a convergence in per capita expenditures between the two sectors (Guerrero, Amaris, Health Insurance Coverage by Population Group and Yerramilli 2013). Health systems have historically built their financing Mexico has more recently followed a similar path. The schemes around sources of funding rather than health health reform of 2003 led to the Seguro Popular de Salud needs, often leaving the poor without access to pooled, (SPS), which, by 2012, provided health coverage to more public financing systems or opportunities for prepay- than 52 million Mexicans who had been ineligible for ment. One of the core ideas of UHC and progressive health care through the existing social security systems, universalism is the determination to cover the poor first with coverage of a progressively expanding number of and relieve this group of the burden of impoverishing interventions (Knaul, Chertorivski Woldenberg, and and catastrophic health spending. Arreola-Ornelas 2012). The expansion of coverage In countries that finance their health systems began with the poorest segments of the population. SPS through health or social insurance, salaried workers deliberately built on the platform of the anti-poverty and government employees are typically the first to be program Oportunidades and enhanced the coverage of covered, financed by payroll deductions supplemented a package of covered services for the poor by expanding by employer contributions. In many countries, this the package (Frenk 2006; Knaul, González-Pier, and group has access to superior social security health others 2012). 284 Cancer Table 17.1 Summary of Health Insurance Initiatives: China, Colombia, Dominican Republic, Ghana, India, Mexico, Peru, and Thailand Country (World Bank income Interventions Percentage of cost group) Program/legal underpinning Population coverage covered covered Current cancer provisions Impact China Launched in early 2000s 87% coverage in 2008 All included • Varies by county: Required coverage by • Increased access (upper-middle- • New Cooperative Medical (72% urban, 93% rural) catastrophic diseases 25% to 40% of catastrophic disease insurance by NCMS to health income) Scheme (NCMS) MFA: 93.37 million poor cost for cancer (per Ministry of Health 2012 services but no covered as of 2010 covered in rural decree): childhood leukemia, improvement of • Urban Employee Basic Medical counties breast cancer, cervical cancer, financial protection, Insurance Scheme (UEBMI) chronic lymphoid leukemia, potentially because of • UEBMI and • Urban Residents Basic Medical URBMI reimburse lung cancer, esophageal cancer, increasing costs of care Insurance Scheme (URBMI) 70% and 50%, colon cancer, and rectal cancer • Supplementary • Medical Financial Assistance respectively, insurance program too Program (MFA) for the poor, of inpatient new to be evaluated covers premiums, extra expenditures expenses for programs 1–3 • Various directives from the Ministry of Health Colombia • Universal social health Universal as of 2011 • Child vaccination, • Screening: cervical, breast, • Has protected (upper-middle- insurance system introduced in • Subsidized plan: 51% cervical cancer prostate, colorectal against catastrophic income) early 1990s, with two schemes: screening, • Treatment: hospitalization; expenditures (no details 1. Contributory for formal • Contributory plan: hospitalization, in case study) 39% chemotherapy; radiotherapy; sector chemotherapy, and most cancer drugs, • Access still limited: 2. Subsidized for informal radiotherapy, including tamoxifen and 78% of breast cancers sector, unemployed, and and most cancer paclitaxel, rituximab, and diagnosed in advanced poor drugs trastuzumab stages • High-cost subaccount created • Coverage of to pool risk among insurers catastrophic across entire population illness expanded gradually, now including some cancers, with others under consideration table continues next page Financing Cancer Care in Low-Resource Settings 285 Table 17.1 Summary of Health Insurance Initiatives: China, Colombia, Dominican Republic, Ghana, India, Mexico, Peru, and Thailand (continued) 286 Country (World Bank income Interventions Percentage of cost group) Program/legal underpinning Population coverage covered covered Current cancer provisions Impact Cancer Dominican Seguro Familiar de Salud: 2001 (full 2013: 29% covered by Explicit and • Up to 1 million • Screening: Pap smear and • No evaluation; no Republic implementation in 2007) contributory plan, 25% by comprehensive pesos/person/ mammography registry, so not possible (upper-middle- Compulsory, contributory, and subsidized plan; intention package of year (US$25,000 • Diagnosis, treatment, to evaluate impact on income) subsidized parts to reach universal community and in 2013), with physiotherapy, survival coverage personal health 20% to 30% rehabilitation, and palliative • Intended to provide goods and services co-payment care for adult and pediatric financial protection, but • Contributory cancers not yet clear whether scheme: up it has to US$2,250 • Almost half the additional for population without outpatient cancer coverage, so pay out of drugs pocket for all health care Ghana National Health Insurance Act • All eligible, but only Common, relatively Breast and cervical cancer, with • Has had relatively little (lower-middle- (2003) established the National 34% enrolled as of inexpensive Parliament considering adding effect on reducing income) Health Insurance Scheme (NHIS) 2010 interventions, but prostate and childhood cancers impoverishment caused • Premiums: sliding breast and cervical by medical care scale with no payment cancer included • In 2005, 45% of from poorest, but financing out of pocket, not implemented in including for services practice that should be free India • Rashtriya Swasthya Bima • RSBY and RAS-like • 100% covered, • RAS: Some treatments for No evaluation of coverage (lower-middle- Yojana (RSBY) (national) and programs targeted including head and neck cancers, or financial protection income) several similar state programs, to poorest, with per premiums gastrointestinal tract for example, in Andhra Pradesh, family per year limits • Treatment cancers, genitourinary Karnataka, and Tamil Nadu • NPCDCS: intention covered up system cancers, • Rajeev Aarogyasri Scheme to cover all by 2017; to Rs150,000 gynecological cancers, (RAS): 2007, state of Andhra more public health (US$2,430) soft tissue/bone tumors, Pradesh, for high-cost focus, optimized per family for thoracic and mediastinum treatments mainly for NCDs; spending on subsidized cancers, breast cancer, contributory health insurance prevention, diagnosis, coverage skin cancer, multiple schemes run by RAS for treatment, and myeloma, hepatoblastoma, • Limits higher Wilms tumor, childhood employed people palliation for contributory B-cell lymphomas, AML, • National Programme for schemes ALL, ocular cancers, Prevention and Control histiocytosis, and of Cancer, Diabetes, rhabdomyosarcoma Cardiovascular Diseases, and Stroke (NPCDCS): 2010 • Includes palliative care table continues next page Table 17.1 Summary of Health Insurance Initiatives: China, Colombia, Dominican Republic, Ghana, India, Mexico, Peru, and Thailand (continued) Country (World Bank income Interventions Percentage of cost group) Program/legal underpinning Population coverage covered covered Current cancer provisions Impact Mexico • Seguro Popular de Salud (SPS): 2012: June 2013: Through FPCE: all childhood • Catastrophic spending (upper-middle- 2004, for basic package for 52 million previously • Basic package cancers and cervical, breast, decrease of more income) all not eligible for preexisting uninsured covered covers 285 testicular, prostate, NHL, than 20% among SPS coverage through social interventions ovarian, and colorectal cancers enrollees; overall out- security (formal sector of-pocket spending employees) • FPCE covers 59 decrease for poorest interventions households • Fund for Protection against for 19 disease Catastrophic Expenses (FPCE) groups • 2006–09: increase of national program treatment of childhood • Health Insurance XXI Century cancers from 3% (2013), for children under five to 55%; three-year years of age survival, 55% for ALL and 75% for HL • Low abandonment rates Peru Universal Health Insurance Law Plan Esperanza (Project • Essential Health No out-of-pocket Plan Esperanza (Project Hope), • 21,000 cancers treated (upper-middle- (Law 29,344, 2009) with mandatory Hope) for cancer to cover Plan (no high-cost payment for cancer cancer plan: 2012 in first year at no income) health insurance for all through one all poor and vulnerable treatment) for all services • All costs covered for select personal cost for of three plans: contributory, semi- • Health Solidarity group of cancers: cervix, subsidized population contributory, and subsidized Intangible Fund breast, colon, stomach, • Decrease of out-of- (FISSAL): covers prostate, leukemia, and pocket costs from 34% common cancers, lymphoma in 2012 to 11% in the ESRD, rare • Includes health promotion, first trimester of 2013, diseases prevention, early detection, for those in contributory diagnosis, treatment, and scheme palliative care table continues next page Financing Cancer Care in Low-Resource Settings 287 288 Cancer Table 17.1 Summary of Health Insurance Initiatives: China, Colombia, Dominican Republic, Ghana, India, Mexico, Peru, and Thailand (continued) Country (World Bank income Interventions Percentage of cost group) Program/legal underpinning Population coverage covered covered Current cancer provisions Impact Thailand • 2002: National Health Security • UCS: 76% of UCS: inpatient 100% of services • Cervical cancer screening • No increase in hospital (upper-middle- Act, establishing the Universal population covered and outpatient covered (Pap smear and VIA): special sales of anti-cancer income) Coverage Scheme (UCS) • CSMBS: 7% of services, such as FFS payment to incentivize drugs after five years, • Prior to 2002: population covered free prescription use despite increases in a. Civil Servants Medical medicines, • Breast self-examination other NCD drugs (for • SSS and WCS ambulatory care, example, for diabetes, Benefits Scheme (CSMBS) together: 15% of promotion b. Social Security Scheme hospitalization, blood pressure, and population covered disease prevention, • UCS coverage of cholesterol) (SSS) radiotherapy and c. Worker Compensation and health promotion • Screening rates for chemotherapy, as well as Scheme (WCS) surgeries and critical care cervical and breast for emergency patients cancer quite low • Royal Thai Government issuance of compulsory licenses for four cancer drugs: letrozole, docetaxel, erlotinib, and imatinib Sources: Based on Guerrero, Amaris, and Yerramilli 2013; Knaul, Chertorivski Woldenberg, and Arreola-Ornelas 2012; Rathe, Knaul, and Yerramilli 2013; Seinfeld and Pleic 2013; Yerramilli 2013; Yerramilli and Ataguba 2013; Yerramilli and Firestone 2013; Yerramilli and Jiang 2013. Note: ALL = acute lymphoblastic leukemia; AML = acute myeloid leukemia; ESRD = end-stage renal disease; FFS = fee-for-service; HL = Hodgkin’s lymphoma; NCD = noncommunicable disease; NHL = non-Hodgkin’s lymphoma; VIA = visual inspection with acetic acid. Peru’s Health Insurance Law of 2009 provided cover- who do not have rights of residence. They are cov- age for nonsalaried workers through a semi-contributory ered by medical insurance in their place of origin plan and for the poor through a highly subsidized plan but do not have access to doctors where they work that includes vulnerable groups, such as children and (Goss and others 2014). The scheme for the formal elderly persons. Salaried workers continue to be covered sector is financed by payroll taxes; the other schemes through a preexisting plan (Seinfeld and Pleic 2013). require individual fixed contributions, supplemented Similarly, the Dominican Republic introduced a law by contributions from various levels of government. in 2001 (establishing the Seguro Familiar de Salud) and The local autonomy in program design has resulted commenced implementation in 2007, with the aim of in some variations in the services covered by county comprehensive coverage within a decade. As of 2013, (Yerramilli and Jiang 2013). 54 percent of the population had achieved coverage, In India, as elsewhere, schemes have existed to cover with slightly over 54 percent of this group in the con- salaried workers and their families. A national scheme tributory scheme and the remaining 46 percent in the for the poor was instituted in 2008, covering treatment subsidized scheme (Rathe, Knaul, and Yerramilli 2013). up to a relatively low annual expenditure limit. However, Many other countries in Latin America and the there is no national program for informal sector work- Caribbean have not yet adopted pro-poor health insur- ers. Some states, such as Andhra Pradesh, Karnataka, ance policies and programs, and coverage remains more and Tamil Nadu, have developed schemes with broader segmented. entitlement (Yerramilli 2013). In the MICs of Asia included in our review, coverage Ghana is one of the few Sub-Saharan African coun- is less complete than in many countries in Latin America tries with a national health insurance system, which was and the Caribbean. Singapore, now an HIC, has a scheme introduced in 2003. In theory, coverage is comprehensive, with greater reliance on private insurance, including a with payroll contributions from formal sector workers, separate catastrophic insurance scheme (Medishield) in contributions from informal sector workers on a sliding addition to the mandatory regular insurance (Medisave) scale, contributions from the poor, and exemptions for and the scheme for the poor (Medifund) (Haseltine the core poor (Yerramilli and Ataguba 2013). In practice, 2013). The Singapore scheme has been held up as a informal sector workers pay the minimum contribution good example. However, Shanghai briefly experimented and a small percentage of the poor is exempted from with a similar model and discontinued it (Dong 2003). contributions. With donor contributions, the scheme The problems in Shanghai included poor control of ran at a deficit in 2010 and 2011 (Yerramilli and Ataguba incentives for doctors and hospitals to provide expen- 2013). The Ghana case illustrates some of the issues sive treatments and extreme cases where households facing ambitious schemes in LMICs. exhausted the limits of their insurance and were unable to pay hospital bills and bury their deceased relatives. These experiences suggest that what can work in a small, Health Insurance Coverage by Services and high-income urban country or city is not necessarily Conditions Covered replicable in other settings. The second dimension of coverage is breadth—by Thailand passed the National Health Security Act in services and diseases included. All health insurance 2002, integrating five existing schemes and extending schemes have restrictions on which medical services coverage to workers in the informal sector (Yerramilli are eligible for coverage; how these are determined and Firestone 2013). The scheme covering the poor, crucially affects the equity and efficiency of a health the Voluntary Health Card, was expanded following the system. Cost-effectiveness, population health needs, and financial crisis in 1997. funding should define the package of covered services. In 2003, the Chinese government began cover- In turn, the package defines entitlement, especially once ing rural residents and nonworking urban residents universal enrollment is achieved, which tends to become (including students, children, and elderly and disabled less restrictive as country income increases. A shallow persons) by adding programs to existing schemes package, even if it covers a large proportion of the pop- for urban public and private sector employees. This ulation, is unlikely to offer protection from financial expansion increased national insurance coverage from catastrophe or to lower financial barriers to accessing 23 percent in 2003 to 87 percent in 2008 (72 percent care, particularly for cancer. of urban residents and 93 percent of rural residents) Cancer coverage often comes later in the develop- (Yerramilli and Jiang 2013) and to 97 percent by 2011 ment of these schemes. In LMICs, coverage has tended (Goss and others 2014). The main group remaining to start with cancers that affect children and women and uncovered consists of rural migrants to urban areas, that are curable with access and adherence to treatment. Financing Cancer Care in Low-Resource Settings 289 Poor quality of care, incomplete services, or waiting proactive policies to tax alcohol and tobacco; it uses times can force many patients to seek care in the private the proceeds to help fund the Thai Health Promotion sector and pay out of pocket, especially for medications, Foundation (ThaiHealth), which has been involved in even though these are officially covered by insurance. comprehensive campaigns to reduce smoking (Yerramilli Because treatment typically involves the repeated use of and Firestone 2013). chemotherapeutic agents, waiting can severely reduce China is at an earlier stage of expanding cancer treat- the effectiveness of treatment or block access entirely. ment packages. The central government required local Further, the package of covered services may not include schemes to provide coverage for treatment of specified components that are important for accessing or manag- catastrophic illnesses, including six types of cancer, as ing care, such as transport costs or medications to con- of February 2013 (Yerramilli and Jiang 2013). However, trol symptoms. Similarly, some essential treatments or because medications are not generally covered by insur- services (for example, radiotherapy) may be unavailable ance, and because of high co-payments, the extent of in the public sector, preventing patients from accessing financial protection remains limited. a complete package of care. This situation can severely Although publicly funded insurance in India, par- reduce the efficacy of the package of provided treatment. ticularly for the poor, is expanding, coverage in practice In Colombia, for example, cancer was not included remains limited. In several states, including Andhra when the program started in 1994. A year later, some Pradesh, Karnataka, and Tamil Nadu, coverage is limited cancer interventions were added. Screening for four to third-level care and the treatments included in the cancers was added in 2000, radiotherapy was added packages may not be the most effective or cost-effective in 2010, and mammography and breast biopsies were for the condition (Yerramilli 2013). Primary and sec- added in 2012. Until 2012, fewer services were covered ondary cancer prevention is largely piecemeal and orga- under the subsidized scheme than under the contribu- nized by hospitals and nongovernmental organizations. tory scheme, and access to treatment has often been an However, a national program that aims to expand issue because of geographic isolation (Guerrero, Amaris, access to and coverage of noncommunicable disease and Yerramilli 2013). prevention, including cancer education and screening, Mexico has a fund for protection against catastrophic is in the initial stages of implementation (Yerramilli expenses that has gradually covered more cancer inter- 2013). Several cancer drugs are available at modest cost ventions since 2003 (Knaul, Chertorivski Woldenberg, in India, for historic reasons and because of the large and Arreola-Ornelas 2012; Knaul, González-Pier, and domestic pharmaceutical industry (Goss and others others 2012). Initially, coverage was provided for acute 2014), providing some relief to cancer patients. lymphoblastic leukemia in children; this coverage has In Sub-Saharan Africa, coverage for cancer is more subsequently been extended to certain cervical, breast, limited still. The Ghana National Health Insurance and prostate cancers in adults. The package of cov- Scheme is restricted to the more common and inex- ered services is based on cost-effectiveness criteria but pensive procedures, and the only cancer coverage is includes some expensive components (breast cancer for breast and cervical cancer (Blanchet, Fink, and treatment, for example, includes trastuzumab for HER2- Osei-Akoto 2012). Ghana signed a memorandum with positive patients). aid partners in 2007 to commence screening for breast Peru, which has a separate fund to provide for cat- cancer using mammograms; however, this screening astrophic illnesses, launched Project Hope as part of a program has not yet been implemented (Bosu 2012). national cancer plan in 2012 (Seinfeld and Pleic 2013). In the Dominican Republic, coverage of cancer is at an early stage and specifies a fixed per capita sum for Level of Financial Protection for Cancer Services financial protection. The more advanced treatments are The third dimension of coverage is whether (and how provided in the private sector; although there is some much) patients and families contribute out of pocket coverage from the new insurance scheme, co-payments for services covered. Financial protection—based on remain relatively high (25–30 percent). New public prepayment, risk pooling, and public funding for the sector facilities are under development (Rathe, Knaul, poor—is a cornerstone of efforts to achieve UHC and is and Yerramilli 2013). the goal of many health system reforms. In Thailand, the government has aimed to expand Most countries recognize that public and community access to cancer treatment and, in addition to cover- health services are of the highest priority and should age, has obtained compulsory licenses for four cancer be universally available and fully and publicly funded. medications: letrozole, docetaxel, erlotinib, and ima- Following experiences with reforms where basic public tinib (Yerramilli and Firestone 2013). Thailand also has health services, such as vaccination, suffered because 290 Cancer funding was not explicitly protected (as in Colombia), China has begun to identify priority diseases for the countries have developed strategies to offer protected reduction of co-payments, focusing on inpatient services. financing for all covered interventions in this rubric As of 2013, childhood and chronic myeloid leukemia, as (Estevez 2012). In Mexico, Seguro Popular includes a well as breast, cervical, lung, esophageal, gastric, and col- separate and protected fund (Knaul, González-Pier, and orectal cancers, were included in these programs (Goss others 2012). Still, it has been challenging to build into and others 2014). UHC the mechanisms through which these funds grow in tandem with public and community health services, especially with the availability of new interventions to COVERAGE OF CANCER CARE: EARLY treat or prevent disease. A clear example is the HPV vac- RESULTS cine, which is essential to the future prevention of cervi- cal and several other cancers that are infection associated For cancer especially, it takes time for the benefits of and much more common in LMICs. improved coverage to translate into increased use of ser- Offering public financing for disease prevention vices and then to improvements in health. Unlike adding and health-promoting services is important, given the a vaccination or a medication for an infectious disease, importance of lifestyle and early detection in managing adding cancer services to meet new demand may require many cancers, including those that most burden LMICs. new facilities and infrastructure, specially trained med- Patients tend to underuse these services, especially ical personnel, and the trust of patients and providers. patients who are not fully informed or aware of the risks Without these elements, access to care will not improve, of unhealthy behavior or late detection; this underuse is even if it is formally part of an insurance or health care exacerbated if they also face significant barriers to access. program. We report on a few results of improved cover- age that have been recorded in the case study countries. Co-payments It is important to note that data on the impacts on The effectiveness of co-payments has been debated health—cancer survival or years of healthy life lived— for decades, because any degree of co-payment can are almost impossible to obtain for the financing of deter patients from seeking care. Further, implement- cancer treatment. In some cases, reforms are too recent ing exemptions that target the poor sounds simple to show this degree of impact. In most cases, a major but, in fact, it is difficult to achieve. For these reasons, limiting factor is the lack of data in the form of cancer many proponents advocate the use of taxes as the more registries and a dearth of formal evaluation efforts. effective and equitable means of generating revenue for financing health. UHC initiatives tend to promote sliding-scale prepayments rather than co-payment at Outcomes point of service. Improvements in access and financial protection have Co-payments generally fall as country per capita been documented for the more comprehensive reforms income increases, but they exist even in HICs. Many and those of longer duration, although few formal eval- LMICs, especially the poorest countries, rely heavily on uations have assessed health outcomes. co-payments. In the LICs of Sub-Saharan Africa—with Thailand has made explicit efforts to increase care some notable and recent exceptions, such as Rwanda— availability, for example, by issuing compulsory licenses public resources for cancer treatment and care are for some cancer drugs, expanding the number of med- severely limited and co-payments are the norm. ical school graduates, and offering incentives to doctors Co-payments often vary by type of service, being to practice in rural areas (WHO 2011b). Thailand also smaller (or zero) for services at facilities, but very large explicitly allocates a fixed per capita amount to preven- (even 100 percent) for medications. In China, the design tion, which is given to communities for local efforts, in varies by county, but co-payments of 60–80 percent can addition to national programs undertaken by the Thai be required (Yerramilli and Jiang 2013). Many countries Health Promotion Foundation (Yerramilli and Firestone set explicit limits on annual coverage per person or per 2013). Despite these efforts, less than 20 percent of household (for example, India’s national scheme for eligible women have been screened for cervical cancer the poor and China), such that treatment for cancer is (Leetongin 2011). likely to exhaust benefits and require large out-of-pocket In Mexico, improved coverage has translated to expenditures. improved survival rates for some pediatric cancers Thailand is unusual in that services provided by the and lower treatment abandonment rates than else- government sector do not require co-payment, includ- where in the region. For breast cancer, the introduc- ing prescription drugs (Yerramilli and Firestone 2013). tion of SPS was associated with reduced treatment Financing Cancer Care in Low-Resource Settings 291 abandonment rates at the National Cancer Institute. cancer patients, although the treatments were futile. Incidence of catastrophic spending has decreased, as In the Republic of Korea, fee-for-service payments for has out-of-pocket expenditure by the poor (Knaul, screening tests for colorectal cancer are used, but the Chertorivski Woldenberg, and Arreola-Ornelas 2012; reimbursement rates are not consistent with the pattern Knaul, González-Pier, and others 2012). of cost. Reimbursement rates for colonoscopy are set too There is a clear need for evaluation of programs low relative to fecal occult blood tests, making colonos- as they mature, measuring health outcomes as well as copy, on the one hand, more cost-effective than it would process outcomes. otherwise be, but, on the other hand, providing a disin- centive for its provision by service providers (Park, Yun, and Kwon 2005). Incentives The design of a financing system may create unintended CONCLUSIONS or intended incentives. In India, the annual cap on pay- ments in the scheme covering the poor means that private In many LMICs, health financing reform efforts have cancer care providers have an incentive to move patients much in common. Key elements include developing to the public sector once the benefits have been exhausted contributory and subsidized plans for various popula- (Yerramilli 2013). In Latin America and the Caribbean, tion groups, meeting the challenges of incorporating the relatively high payroll taxes on formal sector employ- and financing nonsalaried workers and the poor, and ment were a concern in the 1980s and blamed for hold- building on basic services associated with social welfare ing down the size of the formal employment sector and programs. increasing informal employment. Similar issues can arise Countries working toward UHC have established if nonsalaried and own-account workers become entitled universal entitlements to key services through guar- to the same health benefits as salaried workers, without anteed benefits packages. The countries are striving to being required to pay similar payroll taxes. These pitfalls include interventions for common cancers and other need to be recognized in system design to meet efficiency chronic and noncommunicable diseases for which there needs as well as to improve equity. are effective interventions. In Colombia, the access to and use of health services Our review of the experiences in several LMICs sug- by the poor have improved (Giedion and Uribe 2009). gests that these challenges can be met with well-designed A concerning development, however, is that patients and financing reform. Prevention, early detection, treatment, their families are successfully suing the government to and palliative care interventions for cancer can be effec- demand coverage of expensive but ineffective cancer treat- tively integrated into basic service packages covered ments, including those for late-stage cancers (Guerrero, by a combination of social insurance and tax-financed Amaris, and Yerramilli 2013). Open-ended constitutional schemes. Invariably, cancer comes after other basic and programmatic rights, combined with the desperation services have been covered, but this can happen relatively of patients and families, and the financial benefits for pro- quickly—within a decade of program initiation. ducers of on-patent drugs and expensive services provide In several countries, for example, Mexico and strong financial incentives. These costly interventions may Singapore, specific and distinct funds were established deplete available funding for the national cancer fund and to cover personal health services and catastrophic undermine the financing of the health system (Guerrero, expenses. In China and India, the design of public Amaris, and Yerramilli 2013). insurance provides for some coverage, yet treatment Incentives to provide care are also affected by whether expense ceilings leave the population vulnerable to payments to service providers are made on a capitation catastrophic health expenses. Some countries, such as or fee-for-service basis. Some insurance schemes pay Ghana, have used earmarked taxes or levies to derive on a fee-for-service basis for interventions that the resources for health. In all cases, UHC coverage is built government wishes to promote, for example, cervical up over time, adding covered populations and services cancer screening in Thailand (Srithamrongswat and and covering a greater proportion of costs, particularly others 2010). Private sector providers are reimbursed for catastrophic costs. other services on a capitation basis, with a global budget Cancer epitomizes why investment in a systems ceiling based on Diagnosis Related Groups to contain approach to chronic diseases in LMICs is strategic. costs (Garebedian and others 2012). In China in the The expansion of services and interventions discussed past, fee-for-service payment provided hospitals with in other chapters can be realized only if countries the incentive to offer expensive treatments to end-stage develop appropriate financing and insurance systems. 292 Cancer Linking each component of the cancer care control con- Garebedian, L. F., D. Ross-Degnan, S. Ratanawijitrasin, tinuum in an integrated financing plan is challenging P. Stephens, and A. K. Wagner. 2012. “Impact of Universal but necessary to guarantee the effectiveness of national Health Insurance Coverage in Thailand on Sales and Market cancer programs. Share of Medicines for Non-Communicable Diseases: An Interrupted Time Series Study.” British Medical Journal Open 1: e001686. Giedion, U., and M. Uribe. 2009. “Colombia’s Universal Health NOTES Insurance System.” Health Affairs 28 (3): 853–63. World Bank Income Classifications as of July 2014 are as fol- Goss, P. E., K. Strasser-Weippl, B. L. Lee-Bychkovsky, L. Fan, lows, based on estimates of gross national income (GNI) per J. Lie, and others. 2014. “Challenges to Effective Cancer capita for 2013: Control in China, India and Russia.” Lancet Oncology 15: 489–538. • Low-income countries (LICs) = US$1,045 or less Gralow, J. R., E. Krakauer, B. O. Anderson, A. Ilbawi, P. Porter, • Middle-income countries (MICs) are subdivided: and others. 2012. “Core Elements for Provision of Cancer a) lower-middle-income = US$1,046 to US$4,125 Care and Control in Low and Middle Income Countries.” In b) upper-middle-income (UMICs) = US$4,126 to US$12,745 Closing the Cancer Divide: An Equity Imperative, edited by • High-income countries (HICs) = US$12,746 or more. F. M. Knaul, J. R. Gralow, R. Atun, and A. Bhadalia, 123–65. Cambridge, MA: Harvard Global Equity Initiative and 1. This section is based on Atun and Knaul (2012). Harvard University Press. Guerrero, R., A. M. Amaris, and P. Yerramilli. 2013. “Colombia Study.” Harvard University Global Health Equity Initiative, updated from R. Guerrero, “Case 2: REFERENCES Colombia,” in R. Atun and F. M. Knaul, 2012, and Atun, R., S. Aydin, S. Chakraborty, S. Sümer, M. Aran, and “Innovative Financing: Local and Global Opportunities,” others. 2013. “Universal Health Coverage in Turkey: in Closing the Cancer Divide: An Equity Imperative, edited Enhancement of Equity.” The Lancet 382 (9886): 65–99. by F. M. Knaul, J. R. Gralow, R. Atun, and A. Bhadalia, doi: 10.1016/S0140-6736(13)61051-X. 257–88. Cambridge, MA: Harvard Global Equity Atun, R., and F. M. Knaul. 2012. “Innovative Financing: Local Initiative and Harvard University Press. http://hgei and Global Opportunities.” In Closing the Cancer Divide: .harvard.edu/fs/docs/icb.topic910623.files /DCP3%20 An Equity Imperative, edited by F. M. Knaul, J. R. Gralow, Input%20-Colombia.pdf. R. Atun, and A. Bhadalia, 257–88. Cambridge, MA: Harvard Gwatkin, D. R., and A. Ergo. 2011. “Universal Health Coverage: Global Equity Initiative and Harvard University Press. Friend or Foe of Equity?” The Lancet 377: 2160–61. Blanchet, N. J., G. Fink, and I. Osei-Akoto. 2012. “The Effect of Haseltine, W. A. 2013. Affordable Excellence: The Singapore Ghana’s National Health Insurance Scheme on Health Care Healthcare Story. Washington, DC: Brookings Institution Utilisation.” Ghana Medical Journal 46 (2): 76–84. Press. Bosu, W. K. 2012. “A Comprehensive Review of the Policy and Hewitt, M., S. Greenfield, and E. Stovall, eds. 2005. From Programmatic Response to Chronic Non-Communicable Cancer Patient to Cancer Survivor: Lost in Transition. Disease in Ghana.” Ghana Medical Journal 46 (2 Suppl): Washington, DC: Institute of Medicine and National 69–78. Research Council of the National Academics, The National Dong, W. 2003. “Healthcare-Financing Reforms in Transitional Academies Press. Society: A Shanghai Experience.” Journal of Health, Jamison, D. T., L. H. Summers, G. Alleyne, K. J. Arrow, Population and Nutrition 21 (3): 223–34. S. Berkley, and others. 2013. “Global Health 2035: A World Engelgau, M., K. Okamoto, K. Navaratne, and S. Gopalan. Converging within a Generation.” The Lancet 382 (9908): 2010. “Prevention and Control of Selected Chronic NCDs 1898–55. in Sri Lanka: Policy Options and Action Plan.” Health, Knaul, F. M., G. Alleyne, P. Piot, R. Atun, J. Gralow, and Nutrition and Population Discussion Paper, World others. 2012. “Health System Strengthening and Cancer: Bank, Washington, DC. http://siteresources.worldbank A Diagonal Response to the Challenge of Chronicity.” In .org/HEALTHNUTRITIONANDPOPULATION/Resources Closing the Cancer Divide: An Equity Imperative, edited by /281627-1095698140167/NCDsSriLanka.pdf. F. M. Knaul, J. R. Gralow, R. Atun, and A. Bhadalia, 95–122. Estevez, R. 2012. “The Quest for Equity in Latin America: A Cambridge, MA: Harvard Global Equity Initiative and Comparative Analysis of the Health Care Reforms in Brazil Harvard University Press. and Colombia.” International Journal of Equity and Health Knaul, F. M., H. Arreola-Ornelas, O. Méndez, M. Alsan, 11 (6). http://www.equityhealthj.com/content/11/1/6. doi: J. Seinfeld, and others. 2014. “The Global Economic 10.1186/1475-9276-11-6. Burden of Cancer.” In World Cancer Report 2014, edited Frenk, J. 2006. “Bridging the Divide: Global Lessons from by B. W. Stewart and C. P. Wild, chapter 6.7. Lyon, Evidence-Based Health Policy in Mexico.” The Lancet 368 France: IARC. http://www.iarc.fr/en/publications/books (9539): 954–61. /wcr/index.php. Financing Cancer Care in Low-Resource Settings 293 Knaul, F. M., H. Arreola-Ornelas, O. Mendez-Carniado, in R. Atun and R. M. Knaul, 2012. http://hgei.harvard C. Bryson-Cahn, J. Barofsky, and others. 2006. “Evidence .edu/fs/docs/icb.topic910623.files/DCP3%20Input% Is Good for Your Health System: Policy Reform to Remedy 20-Dominican%20Republic.pdf. Catastrophic and Impoverishing Health Spending in Seinfeld, J., and M. Pleic. 2013. “Peru Case Study.” Harvard Mexico.” The Lancet 368 (9549): 1828–41. Global Equity Initiative, updated from J. Seinfeld, “Case Knaul, F. M., S. Chertorivski Woldenberg, and H. Arreola- 4: Peru,” in R. Atun and F. M. Knaul, 2012. http://hgei Ornelas. 2012. “Case 4: Mexico.” In Closing the Cancer .harvard.edu /fs/docs/icb.topic910623.files/DCP3%20 Divide: An Equity Imperative, edited by F. M. Knaul, Input%20-Peru.pdf. J. R. Gralow, R. Atun, and A. Bhadalia, 269–70. Cambridge, Srithamrongswat, S., W. Aekplakorn, P. Jongudomsuk, MA: Harvard Global Equity Initiative and Harvard J. Thammatach-aree, W. Patcharanarumol, and others. 2010. University Press. “Funding Health Promotion and Prevention: The Thai Knaul, F. M., E. González-Pier, O. Go ´mez-Dantés, D. Garcia- Experience.” Background Paper 45 for the World Health Junco, H. H. Arreola-Ornelas, and others. 2012. “The Report 2010, World Health Organization. http://www.who Quest for Universal Health Coverage: Achieving Social .int/healthsystems/topics/financing/healthreport/Thailand Protection for All in Mexico.” The Lancet 380 (9849): No45FINAL.pdf. 1259–79. WHO (World Health Organization). 2010. World Health Report Knaul, F. M., J. R. Gralow, R. Atun, and A. Bhadalia, eds. 2010: Health Systems Financing: The Path to Universal 2012. Closing the Cancer Divide: An Equity Imperative. Coverage. WHO, Geneva. http://www.who.int/whr/2010 Cambridge, MA: Harvard Global Equity Initiative and /whr10_en.pdf?ua=1 Harvard University Press. ———. 2011a. “World Health Statistics, 2011.” WHO, Geneva. Leetongin, G. 2011. “Results Based Financing: Screening for http://www.who.int/whosis/whostat/EN_WHS2011_Full Cervical Cancer in Thailand.” World Bank Presentation. .pdf. http://siteresources.worldbank.org/INTHSD/Resources ———. 2011b. “Country Cooperation Strategy: Thailand.” /topics/415176-1217510876610/5259033-1333544225723 http://www.who.int/countryfocus/cooperation_strategy /Day5_3_Grits_Thailand_NHSO_EN.pdf. /listofccs/en/index.html. Mahal, A., A. Karan, and M. Engelgau. 2010. “The Economic ———. 2012. “What Is Universal Health Coverage? Online Implications of Non-Communicable Disease for India.” Q&A.” http://www.who.int/features/qa/universal_health Health, Nutrition and Population Discussion Paper, _coverage/en/index.html. World Bank, Washington, DC. http://siteresources.worldbank ———. 2013. World Health Report 2013: Research for Universal .org/HEALTHNUTRITIONANDPOPULATION/Resources Health Coverage. Geneva: WHO. /281627-1095698140167/Economic Implications ofNCD World Bank. 2013. World Development Indicators 2013. forIndia.pdf. Washington, DC: World Bank. http://data.worldbank Mahal, A., A. Karan, V. Fan, and M. Engelgau. 2013. “The .org/indicator/NY.GNP.PCAP.CD/countries/SG --XR? Economic Burden of Cancers on Indian Households.” PLOS page=1&display=default. One 8 (8): e71853. doi: 10.1371/journal.pone.0071853. Yerramilli, P. 2013.“The Financing of Cancer Care and Control in Nikolic I., A. Stanciole, and M. Zaydman. 2011. “Chronic Andhra Pradesh,India.”MSc thesis,London School of Tropical Emergency: Why NCDs Matter.” Health, Nutrition and Medicine and Hygiene and London School of Economics. Population Discussion Paper, World Bank, Washington, Yerramilli, P., and J. Ataguba. 2013. “Ghana Case Study.” DC. http://siteresources.worldbank.org /HEALTH Harvard Global Equity Initiative. http://hgei.harvard.edu NUTRITION ANDPOPULATION/Resources/281627 /fs/docs/icb.topic910623.files/DCP3%20Input%20-Ghana - 1 0 9 5 6 9 8 1 4 0 1 6 7 / C h r o n i c E m e r g e n c y Why N C D s .pdf. Matter.pdf. Yerramilli, P., and R. Firestone. 2013. “Thailand Case Study.” Park, S. M., Y. H. Yun, and S. Kwon. 2005. “Feasible Economic Harvard Global Equity Initiative. http://hgei.harvard.edu Strategies to Improve Screening Compliance for Colorectal /fs/docs/icb.topic910623.files/DCP3%20Input%20-% Cancer in Korea.” World Journal of Gastroenterology 11 (11): 20Thailandt.pdf. 1587–93. Yerramilli, P., and X. X. Jiang. 2013. “China Case Study.” Rathe, M., F. M. Knaul, and P. Yerramilli. 2013. “Dominican Harvard University Global Equity Initiative. http:// Republic Case Study.” Harvard Global Equity Initiative, hgei.harvard.edu/fs/docs/icb.topic910623.files/DCP3% updated from M. Rathe, “Case 3: The Dominican Republic,” 20Input%20-%20China.pdf. 294 Cancer Chapter 18 An Extended Cost-Effectiveness Analysis of Publicly Financed HPV Vaccination to Prevent Cervical Cancer in China Carol Levin, Monisha Sharma, Zachary Olson, Stéphane Verguet, Ju-Fang Shi, Shao-Ming Wang, You-Lin Qiao, Dean T. Jamison, and Jane J. Kim INTRODUCTION Inadequate Screening Services Disproportionate Burden of Disease Screening in China is opportunistic in the absence of a national cervical cancer screening program. From 2009 Cervical cancer is one of the 10 most common diseases to 2011, the national government initiated a program affecting women in China. Although the average national to provide free cervical cancer screening for 10 million estimates of the cervical cancer burden in China are rural women between ages 35 and 59 years; the program low, the burden may be underestimated because the covered only 7 percent of women because of the short- prevalence of the human papillomavirus (HPV) is ages of gynecologists and cytologists and an overbur- high. Cervical cancer mortality is heterogeneous across dened health care system (Colombara and Wang 2013; geographic settings (Li, Kang, and Qiao 2011); it is high- Qiao 2010; The Lancet 2009). China has an estimated est among poor women living in Gansu, Shanxi, and 500 million women in rural areas; scaling up preventive Shaanxi, the least developed provinces in central and services constitutes a significant public health challenge western China. (Li, Kang, and Qiao 2011), and national cervical cancer The low national cervical cancer estimates may be screening coverage remains low (Gakidou, Nordhagen, the result of the lack of a nationwide cancer registry. and Obermeier 2008; WHO 2012). Additional rea- Most registries are located in urban areas, where the sons for low screening coverage include weak health socioeconomic status of women is higher and the can- system infrastructure to support screening, diagnosis, cer disease burden is likely to be lower than in rural and treatment; limited access to health services; and areas (Shi and others 2011). The HPV prevalence has limited knowledge of cervical cancer among women in been found to be similar in rural and urban regions, less developed regions (Jia and others 2013; Qiao and but cervical cancer mortality is significantly higher others 2008). among women in rural areas. This disproportionate China’s health care system has been evolving disease burden is likely attributable to the unequal to respond to the pervasive inequity in access to availability and utilization of health services, such as health services. In 2009, China began to introduce screening and treatment. Corresponding author: Carol Levin, PhD, University of Washington, clevin@uw.edu. 295 universal health coverage (Yip and others 2012), reach- costs were US$25 per vaccinated girl. Chapter 4 in this ing relatively high coverage in urban and rural areas volume (Denny and others 2015) provides a fuller with two government-sponsored schemes. Despite the description of the cost-effectiveness of cervical cancer high coverage, the benefits are minimal and reimburse- prevention in China and other settings. ment is limited to inpatient expenses. It is unclear how recent health insurance schemes and opportunistic Extended Cost-Effectiveness Analysis screening programs have affected women’s cervical cancer treatment rates, health outcomes, and costs. Recent attention to attaining the goal of universal health Recent studies that focused on prevention indicate that coverage provides a strong rationale for exploring existing cancer prevention services do not reach women mechanisms to expand access to the prevention and in poorer rural and urban areas (Jia and others 2013; treatment of cervical cancer in China, without increas- Li, Kang, and Qiao 2011; Shi, Canfell, and others 2012). ing the financial burden of the women seeking care and This trend is likely to continue until insurance schemes paying for services (WHO 2013b). We conducted an cover outpatient services, including treatment of pre- extended cost-effectiveness analysis (ECEA) to evaluate cancer and early-stage cancer. public financing of HPV vaccination to prevent cervi- Although widespread screening with cytology has cal cancer. Importantly, the ECEA approach adds new dramatically reduced the cervical cancer burden in dimensions to conventional cost-effectiveness analysis high-income countries, low-resource settings have been through a more explicit treatment of equity and impact unable to achieve similar cancer reductions. Newer on financial risk protection—prevention of medical screening technologies are cheaper and easier to imple- impoverishment (Verguet and others 2015; Verguet, ment and scale up than cytology and can reduce the Laxminarayan, and Jamison 2015; Verguet and others cervical cancer burden among Chinese women, pro- 2013). Specifically, ECEA can evaluate publicly financed tecting them from the future costs and consequences programs by measuring program impact along four of the disease. For example, a study of cervical cancer main dimensions: screening that evaluated strategies using cervical cytol- ogy and HPV DNA testing found that screening women • Health benefits three times in their lives (between the ages of 25 and 45 • Household private expenditures averted (household years) reduced the risk of cancer by 50 percent at a cost cost savings) of US$150 per life-year saved (LYS). The most efficient • Financial risk protection provided to households strategy used a two-visit rapid HPV DNA test, with • Distributional consequences across the wealth strata screening and diagnostic assessment at a county hospital of country populations. and treatment provided during the second visit (Levin and others 2010; Li and others 2013; Shi, Canfell, and As a result, ECEA enables the quantitative inclusion others 2012; Wang and others 2013; Zhang, Pan, and of information on equity and on how much financial others 2013; Zhao and others 2013). risk protection is bought per dollar expenditure on health policy, in addition to how much health is bought (Verguet, Laxminarayan, and Jamison 2015; Verguet and HPV Vaccination others 2013). In addition to screening, HPV vaccination presents a As a consequence, the distribution of health and promising primary prevention strategy against cervical financial benefits resulting from health interventions— cancer. Several studies have concluded that screening and, by extension, from the policy instruments that women and vaccinating preadolescent girls against HPV finance them—can be examined to answer the question are cost-effective interventions in reducing the burden of whether the interventions are pro-poor. In practice, of cervical cancer in China (Canfell and others 2011; the ECEA approach can also be used to examine the Goldie and others 2008; Levin and others 2010; Shi and financial effects of interventions and policies on indi- others 2011). Canfell and others (2011) showed that viduals or families by income group and in aggregate. vaccination strategies were cost effective up to US$55 Health policies and interventions typically involve costs per vaccinated girl, with incremental cost-effectiveness to the public sector and to households. Even if a spe- ratios (ICERs) of US$2,746 per LYS when vaccination cific intervention is provided at no cost, users often was combined with screening once in a lifetime; the incur time costs if they are required to travel or wait at ICER was up to US$5,963 per LYS when combined with health facilities to receive information, treatment, or test five screenings in a lifetime. Goldie and others (2008) results; the value placed on these costs differs according found an ICER of US$1,360 per LYS when total vaccine to income level. 296 Cancer Publicly financed health interventions can help users immunity after clearance of HPV infections. The natural to avoid future costs. For example, HPV vaccination and history and model transitions have been well described cancer screening programs reduce the risk of cervical elsewhere (Kim and Goldie 2008; Kim, Ortendahl, and cancer, which might otherwise lead to medical impover- Goldie 2009). For a more extensive discussion of the ishment, devastating health consequences (for example, natural history of cervical cancer, see chapter 4 in this the death of a mother increases the mortality risk for volume (Denny and others 2015). children), or both (for example, the death of the primary All women are subject to mortality from the com- household income earner could impoverish the family). peting causes listed in the World Health Organization The objective of this analysis is to evaluate the conse- (WHO) life table estimates for China (WHO 2011). quences of public finance of HPV vaccination in China, Our approach was to calibrate the model for the cervical using the ECEA methodology. Public finance increases cancer burden of the country as a whole and also for the the uptake of the HPV vaccine, which can improve HPV 16/18 type distribution in cervical cancer, which health, reduce household medical expenditures related has been found in previous meta-analyses to be stable to cervical cancer treatment, and prevent subsequent for 16/18 at 70 percent, regardless of country. De Sanjose impoverishment. Finally, public finance can have differ- and others (2010) found that more than 30 types of HPV ential impacts among populations of different income are sexually transmitted and may lead to cervical cancer, levels. We estimate the level and distribution across most notably HPV 16 and 18, which together contribute income groups of the cervical cancer deaths averted, to approximately 70 percent of cervical cancers world- the households’ expenditures related to cervical cancer wide. Accordingly, HPV is categorized as follows: treatment averted and the costs needed to sustain the HPV program, and the financial risk protection that the • High-risk type 16 (HR-16) program provides, using a combination of indicators. • High-risk type 18 (HR-18) • Other high-risk types (HR-other), including 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, 68, 73, and 82 MATERIALS AND METHODS • Low-risk (LR) types, including 6, 11, 26, 32, 34, 40, 42, 44, 53, 54, 55, 57, 61, 62, 64, 67, 69, 70, 71, 72, 81, Model 83, and 84. We synthesized the available epidemiological, clinical, and economic data from China, using a previously We initially established model parameters using the described individual-based Monte Carlo simulation best available information on the natural history of HPV model of cervical cancer (Goldhaber-Fiebert and others infection and cervical carcinogenesis. The model was 2007; Goldie and others 2007; Kim and others 2007). The adapted to the context in China using likelihood-based model consists of health states representing important methods to fit the parameters to epidemiological data clinical stages of disease, including HPV infection, grade (figure 18A.1 in online annex 18A). Age-specific cervical of precancerous lesions, and stage of invasive cancer. cancer estimates were obtained from the GLOBOCAN We evaluated vaccination and screening as a combined (2008); data on HPV 16 and 18–type distribution in strategy in a single cohort, such that preadolescent girls CIN 2,3 and cervical cancer were from a meta-analysis who are vaccinated will also eventually receive screening. of primary data from Asia (Bao and others 2008) Cervical cancer can be detected through symptoms (figure 18A.2 in online annex 18A). The baseline natural or screening, and women with cancer survive accord- history parameters were allowed to vary over plausi- ing to stage-specific survival rates for local, regional, ble ranges. We identified sets of parameter values that and metastatic disease. This model does not consider achieved close fit to the empirical data and conducted screening strategies alone for the current cohort of older the analysis using the parameter set with the maximum women. Individual girls enter the model at age nine likelihood. Additional details of the model structure, years, before sexual debut and free of HPV infection, and calibration process, and calibration results are available they transition between health states throughout their online in annex 18A. lifetimes. Each month, women face a risk of acquiring HPV infection; once infected, they can clear their infec- tion or develop low- or high-grade lesions, categorized Strategies, Data, and Assumptions as cervical intraepithelial neoplasia, grade 1 (CIN 1) or To model the impact of cervical cancer prevention on grade 2,3 (CIN 2,3). Low-grade lesions can regress; CIN distributional benefits and financial risk protection, we 2,3 can progress to invasive cancer. These transitions are simulated screening with cytology and visual inspection determined by age, HPV type, and type-specific natural with acetic acid (VIA) at five-year intervals, beginning at An Extended Cost-Effectiveness Analysis of Publicly Financed HPV Vaccination to Prevent Cervical Cancer in China 297 age 35 years, for a cohort of one million women in each Table 18.1 Summary of Parameters Used for Modeling of the income quintiles. We assumed that the screening the Impact and Costs of a Publicly Financed HPV frequency progressively increases with income; women Vaccination Policy in China in the lowest two quintiles would be screened once in a Parameter Estimate lifetime, those in the next two income quintiles would a b be screened three times in a lifetime, and those in the Screening with cytology: frequency and coverage (%) highest quintile would be screened five times in a life- Quintile I: Once per lifetime 21 time. Consistent with assumptions made in previous Quintile II: Once per lifetime 34 analyses (Goldie and others 2001; Goldie and others Quintile III: Three times per lifetime 43 2005), cytology was assumed to occur in three visits: the initial screening (visit 1); colposcopy and possible biopsy Quintile IV: Three times per lifetime 47 for screen-positive women (visit 2); and treatment of Quintile V: Five times per lifetime 51 precancerous lesions or invasive cancer (visit 3), includ- Loss to follow-up (%)c ing loop electrosurgical excision procedure (LEEP), Quintile I 62 cold knife conization, simple hysterectomy, or simple radiotherapy, depending on lesion size or cancer stage. Quintile II 40 The VIA screening incorporated same-day screening and Quintile III 22 treatment for all women with positive screening results Quintile IV 13 as described in the Comprehensive Cervical Cancer Quintile V 5 Control guidelines (WHO 2006). c Vaccine characteristics Vaccination was assumed to occur before age 12 (prior to sexual debut), with full adherence to the three Vaccination coverage (%) 70 doses, affording complete and lifelong protection against HPV vaccine price per dose (US$) 13 HPV 16 and 18. HPV vaccine coverage was assumed at Incremental vaccine program delivery cost per fully 5 70 percent, based on current immunization rates of over immunized girl (US$) 95 percent for childhood vaccines and recent evidence Vaccine cost per fully immunized girl, including 46 on the feasibility of reaching preadolescent girls with wastage and handling (US$) HPV vaccination using facility, school-based, and out- Income and wagesd reach strategies (La Montagne and others 2011; WHO 2013a). We used screening coverage estimates by quin- Average GDP per capita (US$) 3,749 tile from the WHO Study of Global AGEing and Adult Average GDP per capitae (US$) Health (WHO 2012). In the absence of patient health Quintile I 783 utilization data for screen-positive women, we assumed Quintile II 1,633 that loss to follow-up from screening to subsequent visits for diagnosis and treatment was inversely related Quintile III 2,567 to income, with loss to follow-up rates ranging from 62 Quintile IV 3,888 to 5 percent from lowest to highest quintile, respectively. Quintile V 7,896 Recognizing that service utilization and loss to follow-up Mean wage ratef (US$) will be influenced by heterogeneity in health system, Quintile I 3 spatial, and socioeconomic factors across China’s prov- inces, we conducted a sensitivity analysis on screening Quintile II 6 coverage rates and the loss to follow-up assumptions by Quintile III 10 quintile. Table 18.1 summarizes the point estimates for Quintile IV 15 the model input parameters. Quintile V 30 Using output data from the simulation model, we Note: Income quintiles are from lowest (I) to highest (V). Monetary values are in 2009 estimated the level and distribution of deaths averted by U.S. dollars. GDP = gross domestic product; HPV = human papillomavirus. income quintile, comparing vaccination plus screening a. Frequency of screening was estimated at one time and five times per lifetime for all against screening at current coverage rates. We also esti- income quintiles in the sensitivity analysis. b. Estimates from Gakidou, Nordhagen, and Obermeyer (2008) and WHO (2012). mated the reduction in cervical cancer, incremental costs c. Estimates are assumed values. to the government equal to HPV vaccination costs minus d. Estimates from WHO Global Health Observatory. cervical cancer treatment costs averted, and patient cost e. Estimates from WHO (2012) and WHO Global Health Observatory. f. Estimates from WHO Global Health Observatory and Shi, Chen, and others (2012). savings, as well as the incremental government health care costs per death averted. Financial risk protection 298 Cancer was estimated using a combination of indicators, includ- programs that range from less than US$5 for countries ing the number of women who would avoid cervical eligible for Gavi, the Vaccine Alliance, to US$13 for Pan cancer treatment expenses and the average out-of-pocket American Health Organization (PAHO) countries (Gavi expenses averted as a share of average per capita income, 2013). Given the likelihood that China could negotiate measured by gross domestic product (GDP). We present lower public sector prices (Colombara and Wang 2013), all results by income quintile. we assumed a public sector cost of US$46 per fully immunized girl, which includes the vaccine price (three doses at US$13 per dose), vaccine wastage (2 percent), Cost Data Sources freight (6 percent), and program administration cost To estimate direct medical and nonmedical costs (US$5). The program administration cost captures the associated with screening, diagnosis, and treatment, we average cost of new delivery strategies to reach preado- used published cost data from two studies conducted lescent girls who fall outside existing routine immuni- in China (Levin and others 2010; Shi, Chen, and oth- zation programs. The program administration costs are ers 2012), where all costs were expressed in 2009 U.S. lower than the average incremental costs in recent stud- dollars. Since these studies provided cost estimates by ies in Latin America and the Caribbean and Sub-Saharan type of facility in urban and rural settings, we assumed Africa, but they are likely to reflect economies of scale an average health-seeking behavior by income quin- that are found in more densely populated Asian coun- tile and geographic setting, where in the lowest three tries (Levin and others 2013). Tables 18A.2 and 18A.3, in income quintiles rural and urban women were screened online annex 18A, summarize the cost data by quintile. at the levels of township (primary health center) and county hospital, respectively, and all women in these lower quintiles received diagnosis and treatment at the Sensitivity Analysis county hospital. We assumed that in the highest two We performed a sensitivity analysis of the findings and income quintiles, rural and urban women were screened evaluated the robustness of the results to changes in and treated at prefecture or provincial level hospitals, screening frequency per lifetime, screening coverage, respectively. loss to follow-up rates, and the cost per fully vaccinated We then applied a weighted average unit cost for girl. To accommodate the uncertainty around the uptake screening, diagnosis, and cancer treatment, based on of the vaccine and vaccine delivery costs in the case of urban and rural population proportions by income China, we conducted a sensitivity analysis and varied the quintile. To estimate the consequences for household cost between US$10 and US$100 per fully immunized and government costs, we assumed that 35 percent of girl to allow for either higher vaccine prices or higher cancer screening and treatment costs are still privately service delivery costs. Table 18.2 provides the estimates financed in China (WHO Global Health Observatory), or ranges used in the sensitivity analysis for these reflecting that many services, including outpatient ser- parameters and the way they varied by income quintile. vices, are not covered, despite mandatory health insur- Not shown are the estimates for cancer treatment costs, ance schemes. Direct nonmedical patient time costs for which were uniformly increased by 50 and 100 percent transportation and waiting were based on time estimates for all income quintiles. from Shi, Chen, and others (2012), using an updated national average wage rate in China ranging from US$3 to US$30 per day for the lowest to highest quintile, RESULTS respectively. The average wage rate is equal to average per capita income divided by 255 workdays per year at eight We estimate that adding preadolescent HPV vaccina- hours per day (Shi, Chen, and others 2012). We obtained tion at 70 percent coverage to current screening will GDP data from the World Bank and used the consumer yield a 44 percent cancer reduction across all income price index to deflate all costs to 2009 U.S. dollars. Per quintiles, as shown in table 18.3. Although the relative capita income for each quintile is the proportion of cancer reduction is constant across income groups, the GDP accrued to each income quintile using estimates absolute numbers of cervical cancer deaths averted and from the World Bank and PovCalNet, an online poverty the financial risk protection from HPV vaccination are analysis tool, divided by the total population per quintile highest among women in the lowest quintile; women in (World Bank 2013a, 2013b). the bottom income quintiles received relatively higher Although Merck and GlaxoSmithKline’s commer- benefits compared with those in the upper income cially available HPV vaccines are not yet approved in quintiles. HPV vaccination averts 15 percent more China, both are offered at low prices for public sector detected cancer cases and 18 percent more deaths in the An Extended Cost-Effectiveness Analysis of Publicly Financed HPV Vaccination to Prevent Cervical Cancer in China 299 Table 18.2 Sensitivity Analysis Parameter Estimates income quintiles, ranging from 60 percent among the and Ranges for HPV Vaccine and Service Delivery lowest income quintile to 30 percent among the highest Costs quintile. (US$) At a vaccine cost of US$46 per fully immunized girl and 70 percent coverage, the incremental cost is approxi- Point Estimate mately US$160 million for a single cohort of five million Parameter estimate or range girls. At the relatively low levels of cancer screening and Screening with cytology: frequencya,b and coverage (%) treatment in China, government intervention costs do Quintile I: Once per lifetime 21 21–70 not vary by wealth strata, since these medical savings are Quintile II: Once per lifetime 34 34–70 offset by the publically financed HPV vaccination costs. Given China’s low reported rates of cervical cancer Quintile III: Three times per lifetime 43 43–70 screening, the model results and relative relationships Quintile IV: Three times per lifetime 47 47–70 across income quintiles are robust to changes in assump- Quintile V: Five times per lifetime 51 51–70 tions about screening frequency, screening coverage, Loss to follow-upc (%) and loss to follow-up (table 18.4). As expected, changes in the cost per fully immunized girl do not have an Quintile I 62 15, 39 impact on deaths averted, cancer reduction, or financial Quintile II 40 15, 24 risk protection, assuming that 70 percent coverage is Quintile III 22 15, 22 maintained. At US$10 per fully vaccinated girl, the cost Quintile IV 13 15, 17 per death averted ranges from US$2,161 for the lowest income quintile to US$2,608 for the highest income Quintile V 5 11, 15 quintile. At US$100 per vaccinated girl, the cost per Vaccine cost per fully immunized girl, death averted increases to US$24,000 for the lowest including wastage and handling (US$) 46 10–100 income quintile to more than US$29,000 for the high- Note: Income quintiles are from lowest (I) to highest (V). Monetary values are in 2009 est income quintile (table 18.4). Universal coverage of U.S. dollars. HPV = human papillomavirus. a. The frequency of screening is estimated at one time and five times per lifetime for all the HPV vaccination becomes even more favorable for income quintiles in the sensitivity analysis. individuals in the lower income quintiles and provides b. Estimates from Gakidou, Nordhagen, and Obermeyer (2008) and WHO (2012). greater relative financial risk protection when treatment c. Estimates are assumed. costs are increased by an additional 50 or 100 percent (table 18.5). lowest compared with the highest quintile. Although in absolute dollars patient savings were higher in the DISCUSSION top income quintile compared with the lowest quintile (US$7,041,335 and US$1,633,160, respectively), the cost Despite worldwide progress in reducing the burden of savings from HPV vaccination constituted a larger share cervical cancer, more than 270,000 women still die from of per capita income among women in the bottom the disease each year; the majority of these deaths occur Table 18.3 Benefits and Costs of a Publicly Financed HPV Vaccination Policy in China Quintile Benefit or cost I II III IV V Deaths averted per million women 2,877 2,854 2,667 2,604 2,362 Government cost per million women (incremental) (US$) 31,417,285 31,420,191 31,440,420 31,446,679 31,359,970 Government cost per death averted (US$) 3,540 3,511 3,312 3,256 2,999 Treatment-seeking cases of cancer averted per million women 3,540 3,511 3,312 3,256 2,999 Patient cost savings per million women (US$) 1,633,160 2,240,688 2,785,626 4,417,303 7,041,335 Savings as percentage of total income 59 39 33 35 30 Cancer reduction (%) 44 44 43 43 44 Note: Income quintiles are from lowest (I) to highest (V). Monetary values are in 2009 U.S. dollars. HPV = human papillomavirus. 300 Cancer Table 18.4 Results of Sensitivity Analysis for HPV Vaccination Costs at US$10, US$46, and US$100 per Fully Vaccinated Girl Quintile Result I II III IV V HPV vaccination at US$10 per fully vaccinated girl Deaths averted 2,877 2,854 2,667 2,604 2,362 Government cost (incremental) (US$) 6,217,285 6,220,191 6,240,420 6,246,679 6,159,970 Government cost per death averted (US$) 2,161 2,179 2,340 2,399 2,608 Treatment-seeking cases of cancer averted 3,540 3,511 3,312 3,256 2,999 Patient cost savings (US$) 1,633,160 2,240,688 2,785,626 4,417,303 7,041,335 Savings as percentage of income 59 39 33 35 30 Cancer reduction (%) 44 44 43 43 44 HPV vaccination at US$46 per fully vaccinated girl Deaths averted 2,877 2,854 2,667 2,604 2,362 Government cost (incremental) (US$) 31,417,285 31,420,191 31,440,420 31,446,679 31,359,970 Government cost per death averted (US$) 10,920 11,009 11,789 12,076 13,277 Treatment-seeking cases of cancer averted 3,540 3,511 3,312 3,256 2,999 Patient cost savings (US$) 1,633,160 2,240,688 2,785,626 4,417,303 7,041,335 Savings as percentage of income 59 39 33 35 30 Cancer reduction (%) 44 44 43 43 44 HPV vaccination at US$100 per fully vaccinated girl Deaths averted 2,877 2,854 2,667 2,604 2,362 Government cost (incremental) (US$) 69,217,285 69,220,191 69,240,420 69,246,679 69,159,970 Government cost per death averted (US$) 24,059 24,254 25,962 26,592 29,280 Treatment-seeking cases of cancer averted 3,540 3,511 3,312 3,256 2,999 Patient cost savings (US$) 1,633,160 2,240,688 2,785,626 4,417,303 7,041,335 Savings as percentage of income 59 39 33 35 30 Cancer reduction (%) 44 44 43 43 44 Note: Income quintiles are from lowest (I) to highest (V). Monetary values are in 2009 U.S. dollars. HPV = human papillomavirus. in Asia, Latin America and the Caribbean, and Sub- yet available in China. Delaying the introduction of Saharan Africa. China accounts for 12 percent of new the HPV vaccine will result in a lost opportunity to cervical cancer cases each year (Ferlay and others 2013), prevent cervical cancer cases and deaths. A national with higher incidence and death rates in the country’s vaccination program from 2006 to 2012 of all girls ages poorest provinces. The factors contributing to the dis- 9–15 years could have prevented 381,000 cervical cancer proportionate distribution of cervical cancer disease cases and 212,000 related deaths in the coming decades include low coverage and poor quality of screening pro- (Colombara and Wang 2013). It is expected that China grams, differential access to services for screening and could negotiate HPV vaccine prices to cost-effective lev- treatment, poverty, and lack of awareness. The availabil- els of approximately US$9 to US$13 per dose, but many ity of HPV vaccines can complement existing cervical Chinese women—at least 33 percent—are not willing to cancer prevention efforts, accelerating the equity and pay more than US$3 (Li and others 2009). A successful health impacts by overcoming many of these barriers program is likely to depend on government financing. (Tsu and Levin 2008). We applied an ECEA approach to evaluate the impact The HPV vaccine holds great promise for reducing of a publically financed policy for HPV vaccination in the burden of cervical cancer, but the vaccine is not China on the distribution of health consequences and An Extended Cost-Effectiveness Analysis of Publicly Financed HPV Vaccination to Prevent Cervical Cancer in China 301 Table 18.5 Results of Sensitivity Analysis Assuming Treatment Costs Increase by 50 Percent and 100 Percent Compared with Baseline Quintile Result I II III IV V Baseline strategy Deaths averted 2,877 2,854 2,667 2,604 2,362 Government cost (incremental) (US$) 31,417,285 31,420,191 31,440,420 31,446,679 31,359,970 Government cost per death averted (US$) 10,920 11,009 11,789 12,076 13,277 Treatment-seeking cases of cancer averted 3,540 3,511 3,312 3,256 2,999 Patient cost savings (US$) 1,633,160 2,240,688 2,785,626 4,417,303 7,041,335 Savings as percentage of income 59 39 33 35 30 Cancer reduction (%) 44 44 43 43 44 Treatment costs increased by 50% Deaths averted 2,877 2,854 2,667 2,604 2,362 Government cost (incremental) (US$) 31,035,156 31,057,311 31,085,113 31,103,101 30,939,508 Government cost per death averted (US$) 10,787 10,882 11,655 11,944 13,099 Treatment-seeking cases of cancer averted 3,540 3,511 3,312 3,256 2,999 Patient cost savings (US$) 1,899,093 2,506,089 3,040,532 4,714,290 7,310,555 Savings as percentage of income 69 44 36 37 31 Cancer reduction (%) 44 44 43 43 44 Treatment costs increased by 100% Deaths averted 2,877 2,854 2,667 2,604 2,362 Government cost (incremental) (US$) 30,647,484 30,682,085 30,728,534 30,760,469 30,550,290 Government cost per death averted (US$) 10,653 10,751 11,522 11,813 12,934 Treatment-seeking cases of cancer averted 3540 3511 3312 3256 2999 Patient cost savings (US$) 2,165,026 2,771,490 3,295,438 5,011,276 7,579,775 Savings as percentage of income 78 48 39 40 32 Cancer reduction (%) 44 44 43 43 44 Note: Income quintiles are from lowest (I) to highest (V). Monetary values are in 2009 U.S. dollars. financial risk protection benefits across income levels. • Although HPV vaccination led to patient cost savings Our analysis showed that preadolescent HPV vaccina- that were small relative to the increase in government tion, added to current cervical cancer screening, could costs, all income groups experienced cost savings; reduce cancer by over 40 percent across all income importantly, there was a powerful equity effect, with groups, while providing relatively higher financial pro- higher financial risk protection in the poorest groups. tection to households in the bottom income quintiles. • Patient cost savings represent a large proportion of The low screening coverage rates reported for China poor women’s average per capita income, reaching affect the government and patient screening and treat- 60 percent among women in the bottom income ment costs, but with differential results. quintile and declining to 30 percent among women in the wealthiest quintile. • From the governmental perspective, a publically financed HPV vaccination program would increase We also estimated standard cost-effectiveness ratios net costs, with little offset from averted cervical-related (results available from the authors) and, similar to treatment costs, because of the low levels of screening. previous studies conducted in China, found that HPV 302 Cancer vaccination is cost effective across all income groups impact of mandatory health insurance by prefecture, when the cost is less than US$50 per vaccinated girl. province, and other geographic settings or across Since the vaccine is not yet available in China, we wealth strata. Our assumptions were based on aggre- assumed a cost of US$46 per vaccinated girl, using gate national estimates of private expenses, which US$13 per dose, based on the manufacturers’ offer may be out of date, given the recent rapid growth in price to PAHO for public vaccination programs in GDP per capita. Latin America and the Caribbean. The financial cost • Fourth, we estimated women’s time using a wage rate of vaccinating 70 percent of China’s current cohort derived from a national estimate of GDP per capita of 6.6 million 10-year-old girls is US$213 million. income, which may overestimate income in the low- This estimate, which accounts for less than 0.5 per- est quintile, where some rural communities are likely cent of projected health care spending of US$357 to live on US$2 per day or less. billion in 2011, would have a large financial impact on • Fifth, this analysis does not include women’s trans- China’s current Expanded Program for Immunization port costs in seeking screening, treatment, or vaccina- (EPI). The introduction of the HPV vaccine would tion; these costs are expected to be small components require a change in policy to finance the vaccine pub- of patient costs based on previous analyses in China licly, through current health insurance schemes or (Canfell and others 2011; Levin and others 2010; Shi, inclusion in EPI, which provides free childhood vac- Chen, and others 2012). cines for measles, diphtheria/tetanus/pertussis, Bacille • Sixth, we did not conduct an exhaustive evaluation Calmette–Guérin, polio, and hepatitis B. EPI manages of scenarios, including increasing screening to higher non-EPI vaccines, such as those for Japanese encephali- levels, since the objective of the analysis was to illus- tis, mumps, and rubella, but patients pay for these vac- trate the potential of an HPV vaccination program cines via user fees. A third type of “optional” vaccines, to address equity and financial risk protection and such as hepatitis A, Haemophilus influenza type B, and not to identify optimal cervical cancer prevention rotavirus, are procured and delivered outside EPI and approaches. paid for by patients without government subsidies (Liu and others 2006). An ECEA approach yields new and essential infor- mation on a policy’s ability to reduce inequity and catastrophic expenses. The approach complements Limitations of the Analysis information on value for money from traditional This analysis has several limitations: cost-effectiveness analyses. Future applications of this approach will benefit from improved information on • First, the analysis is an illustrative application of the public and private health financing, as well as from ECEA method using the best available published data disaggregated data on disease burden and health service from selected provinces, which do not fully capture utilization by key socioeconomic, demographic, and the heterogeneity in disease burden, health systems, geographic variables. socioeconomic development, and GDP per capita across China’s provinces. For example, we used data from different regions in China to estimate the cer- CONCLUSIONS vical cancer burden and costs for the whole country, leading to results that may not hold for a country as HPV vaccines have not yet been approved in China, and large as China. Accordingly, in this application, the concern is growing that the use of HPV vaccines in the results should be considered suggestive, rather than country is still a long way off (Lu 2013). A recent editorial evidence based, but the estimates can be refined for recognizing the burden of cervical cancer in China, as well specific subregions as improved data become available. as its unequal impact among women in lower income • Second, the ECEA method simulates the costs and groups, proposes a semi-mandatory HPV vaccination impacts of HPV vaccination by income quintile; how- program in China targeted to low-income, high-risk ever, there are limited data on the variation of HPV women living in regions with historically high prevalence incidence, mortality rates, loss to follow-up rates for of cervical cancer (Zhang, Li, and others 2013). This screening, and out-of-pocket health expenses related illustrative application of the ECEA approach to a pub- to cancer prevention or treatment by wealth or licly financed HPV vaccination policy provides decision income category. makers with the potential distributional consequences • Third, there is limited information on health service and financial risk protection of including cervical can- utilization; screening and treatment costs; and the cer in future health care reform investments to provide An Extended Cost-Effectiveness Analysis of Publicly Financed HPV Vaccination to Prevent Cervical Cancer in China 303 insight to policy debates in China. An ECEA can provide Denny, L., R. Herrero, C. Levin, and J. T. Kim. 2015. “Cervical policy makers with additional evidence beyond evidence Cancer.” In Disease Control Priorities (third edition): Volume 3, of effectiveness, costs, and cost-effectiveness for selective Cancer, edited by H. Gelband, P. Jha, R. Sankaranarayanan, resource allocation to the populations and provinces and S. Horton. Washington, DC: World Bank. de Sanjose, S., W. G. V. Quint, L. Alemany, D. T. Geraets, most in need in the context of public financing and the J. E. Klaustermeier, and others. 2010.“Human Papillomavirus strengthening of Chinese health reform. Genotype Attribution in Invasive Cervical Cancer: A Previous research has demonstrated that HPV Retrospective Cross-Sectional Worldwide Study.” The vaccination in China can be cost-effective at a cost of Lancet Oncology 11 (11): 1048–56. US$50 per vaccinated girl. A targeted program may Ferlay, J., H. R. Shin, F. Bray, D. Forman, C. Mathers, and others. even be affordable, given China’s plans for dramati- 2010. “Estimates of Worldwide Burden of Cancer in 2008: cally increasing health care spending in the near future GLOBOCAN 2008.” International Journal of Cancer 127 (Le Deu and others 2012; Zhang, Li, and others 2013). (12): 2893–917. doi:10.1002/ijc.25516. Ensuring high and uniform HPV vaccine uptake will Ferlay, J., I. Soerjomataram, M. Ervik, R. Dikshit, S. Eser, and likely also contribute to more equitable gains with others. 2013. GLOBOCAN 2012 v1.0, Cancer Incidence and respect to the reduction of morbidity and mortality Mortality Worldwide: IARC CancerBase No. 11 [Internet]. Lyon: IARC. http://globocan.iarc.fr. from cervical cancer and has the potential to protect Gakidou, E., S. Nordhagen, and Z. Obermeyer. 2008. women in poor households against catastrophic cervical “Coverage of Cervical Cancer Screening in 57 Countries: cancer medical expenses. Low Average Levels and Large Inequalities.” PLoS Medicine 5 (6): e132. Gavi (Vaccine Alliance). 2013. “Human Papillomavirus NOTE Vaccine Support.” http://www.gavialliance.org/support/nvs This chapter was previously published in a shorter form in /human-papillomavirus-vaccine-support. Vaccine; the original publication may be found at doi: 10.1016/j. Goldhaber-Fiebert, J. D., N. K. Stout, J. Ortendahl, vaccine.2015.02.052. It is reprinted here under the Creative K. M. Kuntz, S. J. Goldie, and others. 2007. “Modeling Commons CC-BY 4.0 License (https://creativecommons Human Papillomavirus and Cervical Cancer in the United .org/licenses/by/4.0/). Reprinting in this volume does not States for Analyses of Screening and Vaccination.” Population indicate Elsevier’s endorsement. doi:10.1016/j.vaccine. Health Metrics 5: 11. doi:10.1186/1478-7954-5-11. 2015.02.052 Goldie, S. J., M. Diaz, S. Y. Kim, C. E. Levin, H. Van Minh, World Bank Income Classifications as of July 2014 are as and others. 2008. “Mathematical Models of Cervical follows, based on estimates of gross national income (GNI) Cancer Prevention in the Asia Pacific Region.” Vaccine per capita for 2013: 26 (Suppl. 12): M17–29. Goldie, S. J., L. Gaffikin, J. D. Goldhaber-Fiebert, A. Gordillo- Tobar, C. Levin, and others. 2005. “Cost-Effectiveness of • Low-income countries (LICs) = US$1,045 or less Cervical-Cancer Screening in Five Developing Countries.” • Middle-income countries (MICs) are subdivided: New England Journal of Medicine 353 (20): 2158–68. a) lower-middle-income = US$1,046 to US$4,125 Goldie, S. J., J. J. Kim, K. Kobus, J. D. Goldhaber-Fiebert, b) upper-middle-income (UMICs) = US$4,126 to US$12,745 J. Salomon, and others. 2007. “Cost-Effectiveness of HPV • High-income countries (HICs) = US$12,746 or more. 16, 18 Vaccination in Brazil.” Vaccine 25 (33): 6257–70. Goldie, S. J., L. Kuhn, L. Denny, A. Pollack, and T. C. Wright. 2001. “Policy Analysis of Cervical Cancer Screening REFERENCES Strategies in Low-Resource Settings: Clinical Benefits Bao, Y. P., N. Li, J. S. Smith, Y. L. Qiao, and Asian Cervical and Cost-Effectiveness.” Journal of the American Medical Cancer Prevention Supervisory Board. 2008. “Human Association 285 (24): 3107–15. Papillomavirus Type Distribution in Women from Asia: IARC (International Agency for Research on Cancer). 2012. A Meta-Analysis.” International Journal of Gynecological http://www.iarc.fr/. Cancer 18 (1): 71–79. Jia, Y., S. Li, R. Yang, H. Zhou, Q. Xiang, and others. 2013. Canfell, K., J. F. Shi, J. B. Lew, R. Walker, F. H. Zhao, and “Knowledge about Cervical Cancer and Barriers of others. 2011. “Prevention of Cervical Cancer in Rural Screening Program among Women in Wufeng County, a China: Evaluation of HPV Vaccination and Primary HPV High-Incidence Region of Cervical Cancer in China.” PLoS Screening Strategies.” Vaccine 29 (13): 2487–94. One 8 (7): e67005. Colombara, D. V., and S.-M. Wang. 2013. “The Impact of HPV Kim, J. J., and S. J. Goldie 2008. “Health and Economic Vaccination Delays in China: Lessons from HBV Control Implications of HPV Vaccination in the United States.” Programs.” Vaccine 31 (38): 4057–59. New England Journal of Medicine 359 (8): 821–32. 304 Cancer Kim, J. J., K. M. Kuntz, N. K. Stout, S. Mahmud, L. L. Villa, Shi, J. F., K. Canfell, J. B. Lew, and Y. L. Qiao. 2012. and others. 2007. “Multiparameter Calibration of a Natural “The Burden of Cervical Cancer in China: Synthesis of History Model of Cervical Cancer.” American Journal of the Evidence.” International Journal of Cancer 130 (3): Epidemiology 166 (2): 137–50. 641–52. Kim, J. J., J. Ortendahl, and S. J. Goldie 2009. “Cost-Effectiveness Shi, J. F., K. Canfell, J. B. Lew, F. H. Zhao, R. Legood, and of Human Papillomavirus Vaccination and Cervical Cancer others. 2011. “Evaluation of Primary HPV-DNA Testing Screening in Women Older than 30 Years in the United in Relation to Visual Inspection Methods for Cervical States.” Annals of Internal Medicine 151 (8): 538–45. Cancer Screening in Rural China: An Epidemiologic and LaMontagne, D. S., S. Barge, N. T. Le, E. Mugisha, M. E. Cost-Effectiveness Modelling Study.” BMC Cancer 11: 239. Penny, and others. 2011. “Human Papillomavirus Vaccine doi:10.1186/1471–2407–11–239. Delivery Strategies That Achieved High Coverage in Shi, J. F., J. F. Chen, K. Canfell, X. X. Feng, J. F. Ma, and Low- and Middle-Income Countries.” Bulletin of the others. 2012. “Estimation of the Costs of Cervical Cancer World Health Organization 89 (11): 821–30. Screening, Diagnosis and Treatment in Rural Shanxi Le Deu, F., R. Parekh, F. Zhange, and G. Zhou. 2012.“Healthcare Province, China: A Micro-Costing Study.” BMC Health in China: Entering Unchartered Waters.” McKinsey & Co. Services Research 12: 123. http://www.mckinsey.com/insights /health_systems_and The Lancet. 2009. “Women’s Health in Rural China.” The Lancet _ services / health_care_in_china_entering_uncharted 374 (9687): 358. doi:10.1016/S0140-6736(09)61394-5. _waters. Tsu, V., and C. Levin. 2008. “Making the Case for Cervical Levin, C. E., H. V. Minh, J. Odaga, S. S. Rout, N. T. N. Diep, and Cancer Prevention: What about Equity?” Reproductive others. 2013. “Incremental Costs of Strategies to Deliver Health Matters 16 (32): 104–12. Human Papillomavirus Vaccine to Young Adolescent Girls Verguet, S., C. Gauvreau, S. Mishra, M. MacLennan, S. Murphy, in Peru, Uganda, and Viet Nam.” Bulletin of the World and others. 2015. “The Consequences of Tobacco Tax on Health Organization 91: 585–92. Household Health and Finances in Rich and Poor Smokers Levin, C. E., J. Sellors, J. F. Shi, L. Ma, Y. L. Qiao, and others. 2010. in China: An Extended Cost-Effectiveness Analysis.” “Cost-Effectiveness Analysis of Cervical Cancer Prevention The Lancet Global Health 3 (4): e206-16. doi:10.1016 Based on a Rapid Human Papillomavirus Screening Test /s2214-109x(15)70095-1. in a High-Risk Region of China.” International Journal of Verguet, S., R. Laxminarayan, and D. T. Jamison. 2015. Cancer 127 (6): 1404–11. “Universal Public Finance of Tuberculosis Treatment in Li, J., L.-N. Kang, and Y.-L. Qiao. 2011. “Review of the Cervical India: An Extended Cost-Effectiveness Analysis.” Health Cancer Disease Burden in Mainland China.” Asian Pacific Economics 24: 318–22. Journal of Cancer Prevention 12: 1149–53. Verguet, S., S. Murphy, B. Anderson, K. A. Johansson, R. Glass, Li, J., L. K. Li, J.-F. Ma, L. H. Wei, M. Niyazi, and others. and others. 2013. “Public Finance of Rotavirus Vaccination 2009. “Knowledge and Attitudes about Human in India and Ethiopia: An Extended Cost-Effectiveness Papillomavirus (HPV) and HPV Vaccines among Women Analysis.” Vaccine 31 (42): 4902–10. Living in Metropolitan and Rural Regions of China.” Wang, J. L., Y. Z. Yang, W. W. Dong, J. Sun, H. T. Tao, and Vaccine 27 (8): 1210–15. others. 2013. “Application of Human Papillomavirus Li, R., Q. Zhou, M. Li, S. M. Tong, M. He, and others. in Screening for Cervical Cancer and Precancerous 2013. “Evaluation of Visual Inspection as the Primary Lesions.” Asian Pacific Journal of Cancer Prevention 14 Screening Method in a Four-Year Cervical (Pre-) Cancer (5): 2979–82. Screening Program in Rural China.” Tropical Doctor 43 WHO (World Health Organization). 2006. Comprehensive (3): 96–99. Cervical Cancer Control: A Guide to Essential Practice. Liu, W., J. D. Clemens, J. Yang, and Z. Xu. 2006. “Immunization Geneva: WHO. against Japanese Encephalitis in China: A Policy Analysis.” ———. 2011. “WHO Life Tables.” http://www.who.int/gho Vaccine 24: 5178–82. /mortality_burden_disease/life_tables/life_tables/en. Lu, Y. 2013. “HPV Vaccines among Chinese Women Still a ———. 2012. Study on Global AGEing and Adult Health— Long Way Off.” People’s Daily Online, July 24. http://english 2007/2010, WAVE 1. Geneva: WHO. http://apps.who.int .people.com.cn/90882/8339695.html. /healthinfo/systems/surveydata/index.php/catalog/13. Qiao, Y. L. 2010. “Perspective of Cervical Cancer Prevention ———. 2013a. WHO Vaccine-Preventable Diseases: Monitoring and Control in Developing Countries and Areas.” Chinese System 2013 Global Summary. Geneva: WHO. Journal of Cancer 29 (1): 1–31. ———. 2013b. The World Health Report 2013: Research for Qiao, Y. L., J. W. Sellors, P. S. Eder, Y. B. Bao, J. M. Lim, and Universal Health Coverage. Geneva: WHO. others. 2008. “A New HPV-DNA Test for Cervical Cancer ———. “Global Health Observatory (GHO) data.” http:// Screening in Developing Regions: A Cross-Sectional Study www.who.int/gho/en/. of Clinical Accuracy in Rural China.” The Lancet Oncology World Bank. 2013a. World Development Indicators 2013. 9 (10): 929–36. Washington, DC: World Bank. An Extended Cost-Effectiveness Analysis of Publicly Financed HPV Vaccination to Prevent Cervical Cancer in China 305 ———. 2013b. “PovCalNet: An Online Analysis Tool for A Multicenter National Survey in China.” Vaccine 31 (32): Global Poverty Monitoring.” Development Research Group, 3244–49. World Bank, Washington, DC. http://iresearch.worldbank. Zhang, W. J., F. Li, Y. H. Wang, D. Simayi, A. Saimaiti, org/PovcalNet/index.htm. and others. 2013. “The Case of Semi-Mandatory HPV Yip, W. C., W. C. Hsiao, W. Chen, S. Hu, J. Ma, and others. 2012. Vaccination in China.” Nature Biotechnology 31: 590–91. “Early Appraisal of China’s Huge and Complex Health Care Zhao, F. H., J. Jeronimo, Y. L. Qiao, J. Schweizer, W. Chen, Reforms.” The Lancet 379 (9818): 833–42. and others. 2013. “An Evaluation of Novel, Lower-Cost Zhang, S. K., X. F. Pan, S. M. Wang, C. X. Yang, X. H. Gao, Molecular Screening Tests for Human Papillomavirus and others. 2013. “Perceptions and Acceptability of in Rural China.” Cancer Prevention Research 6 (9): HPV Vaccination among Parents of Young Adolescents: 938–48. 306 Cancer DCP3 Series Acknowledgments Disease Control Priorities, third edition (DCP3) com- We thank the many contractors and consultants piles the global health knowledge of institutions who provided support to specific volumes in the form and experts from around the world, a task that of economic analytical work, volume coordination, required the efforts of over 500 individuals, including chapter drafting, and meeting organization: the Center volume editors, chapter authors, peer reviewers, advi- for Disease Dynamics, Economics, & Policy; Center sory committee members, and research and staff assis- for Chronic Disease Control; Center for Global Health tants. For each of these contributions we convey our Research; Emory University; Evidence to Policy Initiative; acknowledgement and appreciation. First and fore- Public Health Foundation of India; QURE Healthcare; most, we would like to thank our 31 volume editors University of California, San Francisco; University of who provided the intellectual vision for their volumes Waterloo; University of Queensland; and the World based on years of professional work in their respective Health Organization. fields, and then dedicated long hours to reviewing each We are tremendously grateful for the wisdom and chapter, providing leadership and guidance to authors, guidance provided by our advisory committee to the and framing and writing the summary chapters. editors. Steered by Chair Anne Mills, the advisory com- We also thank our chapter authors who collectively mittee assures quality and intellectual rigor of the high- volunteered their time and expertise to writing over est order for DCP3. 160 comprehensive, evidence-based chapters. The U.S. Institute of Medicine, in collaboration with We owe immense gratitude to the institutional spon- the Interacademy Medical Panel, coordinated the peer- sor of this effort: The Bill & Melinda Gates Foundation. review process for all DCP3 chapters. Patrick Kelley, Gillian The Foundation provided sole financial support of Buckley, Megan Ginivan, and Rachel Pittluck managed the Disease Control Priorities Network. Many thanks this effort and provided critical and substantive input. to Program Officers Kathy Cahill, Philip Setel, Carol The World Bank External and Corporate Relations Medlin, and (currently) Damian Walker for their Publishing and Knowledge division provided excep- thoughtful interactions, guidance, and encouragement tional guidance and support throughout the demanding over the life of the project. We also wish to thank production and design process. We would particularly Jaime Sepúlveda for his longstanding support, including like to thank Carlos Rossel, the publisher; Mary Fisk, chairing the Advisory Committee for the second edition Nancy Lammers, Devlan O’Connor, Rumit Pancholi, and, more recently, demonstrating his vision for DCP3 and Deborah Naylor for their diligence and expertise. while he was a special advisor to the Gates Foundation. Additionally, we thank Jose de Buerba, Mario Trubiano, We are also grateful to the University of Washington’s Yulia Ivanova, and Chiamaka Osuagwu of the World Department of Global Health and successive chairs Bank for providing professional counsel on communi- King Holmes and Judy Wasserheit for providing a cations and marketing strategies. home base for the DCP3 Secretariat, which included Several U.S. and international institutions contrib- intellectual collaboration, logistical coordination, and uted to the organization and execution of meetings that administrative support. supported the preparation and dissemination of DCP3. 307 We would like to express our appreciation to the reproductive and maternal health volume consulta- following institutions: tion November 2013) • National Cancer Institute cancer consultation • University of Bergen, consultation on equity (June (November 2013) 2011) • Union for International Cancer Control cancer • University of California, San Francisco, surgery consultation (November 2013, December 2014) volume consultations (April 2012, October 2013, February 2014) Carol Levin provided outstanding governance for • Institute of Medicine, first meeting of the Advisory cost and cost-effectiveness analysis. Stéphane Verguet Committee to the Editors ACE (March 2013) added invaluable guidance in applying and improving • Harvard Global Health Institute, consultation the extended cost-effectiveness analysis method. Shane on policy measures to reduce incidence of non- Murphy, Zachary Olson, Elizabeth Brouwer, and Kristen communicable diseases (July 2013) Danforth provided exceptional research assistance and • Institute of Medicine, systems strengthening meeting analytic assistance. Brianne Adderley ably managed the (September 2013) budget and project processes. The efforts of these indi- • Center for Disease Dynamics, Economics, and Policy viduals were absolutely critical to producing this series (Quality and Uptake meeting September 2013, and we are thankful for their commitment. 308 DCP3 Series Acknowledgments Volume and Series Editors VOLUME EDITORS middle-income countries. Aided by collaborative rela- tionships with other international and national organi- Hellen Gelband is Associate Director for Policy at the zations and individual researchers, he provides technical Center for Disease Dynamics, Economics & Policy assistance to countries developing public health policies (CDDEP). Her work spans infectious disease, particu- that include cancer. His manuals on early detection of larly malaria and antibiotic resistance, and noncommu- cervical cancer have been translated into several lan- nicable disease policy, mainly in low- and middle-income guages, including Chinese, French, Hindi, Portuguese, countries. Before joining CDDEP, then Resources for the Spanish and Turkish, and others. Future, she conducted policy studies at the (former) Congressional Office of Technology Assessment, the Susan Horton is Professor at the University of Waterloo Institute of Medicine of the U.S. National Academies, and holds the Centre for International Governance and a number of international organizations. Innovation (CIGI) Chair in Global Health Economics in the Balsillie School of International Affairs there. Prabhat Jha is the founding director of the Centre for She has consulted for the World Bank, the Asian Global Health Research at St. Michael’s Hospital and Development Bank, several United Nations agencies, holds Endowed and Canada Research Chairs in Global and the International Development Research Centre, Health in the Dalla Lana School of Public Health at among others in work carried out in over 20 low- and the University of Toronto. He is lead investigator of middle-income countries. She led the work on nutrition the Million Death Study in India, which quantifies the for the Copenhagen Consensus in 2008, when micro- causes of death and key risk factors in over two million nutrients were ranked as the top development priority. homes over a 14-year period. He is also Scientific She has served as associate provost of graduate studies Director of the Statistical Alliance for Vital Events, at the University of Waterloo, vice-president academic at which aims to expand reliable measurement of causes of Wilfrid Laurier University in Waterloo, and interim dean death worldwide. His research includes the epidemiol- at the University of Toronto at Scarborough. ogy and economics of tobacco control worldwide. Rengaswamy Sankaranarayanan is head of the Early Detection and Prevention Section and the Screening SERIES EDITORS Group at the World Health Organization’s International Agency for Research on Cancer (IARC) in Lyon, France. Dean T. Jamison He was trained in radiation oncology at the University Dean Jamison is a Senior Fellow in Global Health of Pittsburgh and the University of Cambridge. He is Sciences at the University of California, San Francisco, an expert in evaluating early detection strategies and and an Emeritus Professor of Global Health at the has conducted trials of screening interventions for cer- University of Washington. He previously held aca- vical, oral, and breast cancers in a number of low- and demic appointments at Harvard University and the 309 University of California, Los Angeles; he was an econo- Hellen Gelband mist on the staff of the World Bank, where he was lead See the list of Volume Editors. author of the World Bank’s World Development Report 1993: Investing in Health. He was lead editor of DCP2. He holds a PhD in economics from Harvard University Susan Horton and is an elected member of the Institute of Medicine See the list of Volume Editors. of the U.S. National Academy of Sciences. He recently served as Co-Chair and Study Director of The Lancet’s Prabhat Jha Commission on Investing in Health. See the list of Volume Editors. Rachel Nugent Rachel Nugent is a Research Associate Professor in Ramanan Laxminarayan the Department of Global Health at the University Ramanan Laxminarayan is Vice President for Research of Washington. She was formerly Deputy Director of and Policy at the Public Health Foundation of India, and Global Health at the Center for Global Development, he directs the Center for Disease Dynamics, Economics Director of Health and Economics at the Population & Policy in Washington, D.C., and New Delhi. His Reference Bureau, Program Director of Health and research deals with the integration of epidemiological Economics Programs at the Fogarty International Center models of infectious diseases and drug resistance into of the National Institutes of Health, and senior econ- the economic analysis of public health problems. He was omist at the Food and Agriculture Organization of one of the key architects of the Affordable Medicines the United Nations. From 1991–97, she was associate Facility for malaria, a novel financing mechanism to professor and department chair in economics at Pacific improve access and delay resistance to antimalarial Lutheran University. She has advised the World Health drugs. In 2012, he created the Immunization Technical Organization, the U.S. government, and nonprofit orga- Support Unit in India, which has been credited with nizations on the economics and policy environment of improving immunization coverage in the country. He noncommunicable diseases. teaches at Princeton University. 310 Volume and Series Editors Contributors Isaac F. Adewole Wendong Chen University of Ibadan, Ibadan, Nigeria University of Toronto, Toronto, Canada Hemantha Amarasinghe James Cleary Institute of Oral Health, Maharagama, Sri Lanka University of Wisconsin, Madison, Wisconsin, United States Benjamin O. Anderson Fred Hutchinson Cancer Research Center, University of Anil D’Cruz Washington, Seattle, Washington, United States Tata Memorial Hospital, Mumbai, India Federico G. Antillon Anna J. Dare Universidad Francisco Marroquin, Guatemala City, Centre for Global Health Research, St. Michael’s Guatemala Hospital, University of Toronto, Toronto, Canada Samira Asma Lynette Denny Centers for Disease Control and Prevention, Atlanta, University of Cape Town, Groote Schuur Hospital, Georgia, United States Cape Town, South Africa Rifat Atun Craig Earle Harvard University, Cambridge, Massachusetts, United Cancer Care Ontario, Ontario Institute for Cancer States Research, University of Toronto, Toronto, Canada Rajendra A. Badwe Silvia Franceschi Tata Memorial Centre, Mumbai, India International Agency for Research on Cancer, Lyon, Freddie Bray France International Agency for Research on Cancer, Lyon, Cindy L. Gauvreau France Centre for Global Health Research, St. Michael’s Frank J. Chaloupka Hospital, University of Toronto, Toronto, Canada University of Illinois at Chicago, Chicago, Illinois, Hellen Gelband United States Center for Disease Dynamics, Economics & Policy, Ann Chao Washington, DC, United States National Cancer Institute, Bethesda, Maryland, Ophira M. Ginsburg United States Dalla Lana School of Public Health, University of Chien-Jen Chen Toronto, Toronto, Canada; BRAC University, Dhaka, National Taiwan University, Taipei, Taiwan, China Bangladesh 311 Mary K. Gospodarowicz Jane J. Kim Princess Margaret Cancer Centre, Toronto, Canada Harvard University, Cambridge, Massachusetts, United States Thomas Gross National Cancer Institute, Bethesda, Maryland, Felicia Knaul United States Miami Institute for the Americas, and Miller School of Prakash C. Gupta Medicine, University of Miami, Florida, United States Healis Sehksaria Institute for Public Health, Mumbai, Carol Levin India University of Washington, Seattle, Washington, Sumit Gupta United States Toronto Hospital for Sick Children, Toronto, Canada Joseph Lipscomb Sir Andrew Hall Emory University, Atlanta, Georgia, United States International Agency for Research on Cancer, Lyon, W. Thomas London France Fox Chase Cancer Center, Philadelphia, Pennsylvania, Mhamed Harif United States CHU Mohammed VI, Marrakesh, Morocco Mary MacLennan Rolando Herrero Centre for Global Health Research, St. Michael’s International Agency for Research on Cancer, Lyon, France Hospital, University of Toronto, Toronto, Canada Susan Horton Katherine A. McGlynn University of Waterloo, Waterloo, Ontario, Canada National Cancer Institute, Bethesda, Maryland, United States Scott C. Howard University of Tennessee Health Sciences Center, Monika L. Metzger Memphis, Tennessee, United States St. Jude Children’s Research Hospital, Memphis, Tennessee, United States Stephen P. Hunger Children’s Hospital of Philadelphia, Philadelphia, Raul H. Murillo Pennsylvania, United States National Cancer Institute of Colombia, Bogotá, Colombia Andre Ilbawi MD Anderson Cancer Center, Houston, Texas, Zachary Olson United States University of Washington, Seattle, Washington, United States Trijn Israels Academisch Medisch Centrum, Amsterdam, the Sherif Omar Netherlands Cairo University, Cairo, Arab Republic of Egypt David A. Jaffray Krishna Palipudi Princess Margaret Cancer Centre, TECHNA Institute, Centers for Disease Control and Prevention, Atlanta, Toronto, Ontario Georgia, United States Dean T. Jamison C. S. Pramesh University of California, San Francisco, and University Tata Memorial Centre, Mumbai, India of Washington, Seattle, Washington, United States You-Lin Qiao Prabhat Jha Cancer Hospital, Chinese Academy of Medical Sciences, Centre for Global Health Research, St. Michael’s Peking Union Medical College, Beijing, China Hospital, Dalla Lana School of Public Health, Linda Rabeneck University of Toronto, Toronto, Canada Cancer Care Ontario, University of Toronto, Toronto, Newell Johnson Canada Griffith University, Gold Coast, Australia Preetha Rajaraman Jamal Khader National Cancer Institute, Bethesda, Maryland, King Hussein Cancer Center, Amman, Jordan United States 312 Contributors Kunnambath Ramadas David B. Thomas Regional Cancer Centre, Trivandrum, India Fred Hutchinson Cancer Research Center, University of Washington, Seattle, Washington, United States Chinthanie Ramasundarahettige Centre for Global Health Research, St. Michael’s Edward L. Trimble Hospital, University of Toronto, Toronto, Canada National Cancer Institute, Bethesda, Maryland, United States Timothy Rebbeck Joann Trypuc University of Pennsylvania, Philadelphia, Pennsylvania, Princess Margaret Cancer Centre, Toronto, Canada United States Stéphane Verguet Carlos Rodriguez-Galindo Department of Global Health and Population, Harvard University, Cambridge, Massachusetts, Harvard T. H. Chan School of Public Health, Boston, United States Massachusetts, United States Rengaswamy Sankaranarayanan Judith Wagner International Agency for Research on Cancer, Lyon, Independent economic consultant, Bethesda, Maryland, France United States Monisha Sharma Shao-Ming Wang University of Washington, Seattle, Washington, United Cancer Institute, Chinese Academy of Medical Sciences, States Peking Union Medical College, Beijing, China Ju-Fang Shi Christopher P. Wild Cancer Institute, Chinese Academy of Medical Sciences, International Agency for Research on Cancer, Lyon, Peking Union Medical College, Beijing, China France Pooja Yerramilli Isabelle Soerjomataram Harvard Global Equity Initiative, Harvard University, International Agency for Research on Cancer, Lyon, Cambridge, Massachusetts, United States France Cheng-Har Yip Lisa Stevens University of Malaya Medical Centre, Kuala Lumpur, National Cancer Institute, Bethesda, Maryland, Malaysia United States Ayda Yurekli Sujha Subramanian Independent economist, Ithaca, New York, United States RTI International, Waltham, Massachusetts, Witold Zaton´ski United States Maria Sklodowska-Curie Cancer Centre, Warsaw, Poland Richard Sullivan Ann G. Zauber Kings College London, King’s Health Partners, London, Memorial Sloan-Kettering Cancer Center, New York, United Kingdom New York, United States Terrence Sullivan Fang-Hui Zhao University of Toronto, Canadian Partnership Against Cancer Institute/Hospital, Chinese Academy of Medical Cancer, Toronto, Canada Sciences, Beijing, China Contributors 313 Advisory Committee to the Editors Anne Mills, Chair Roger Glass Professor, London School of Hygiene & Tropical Director, Fogarty International Center, National Medicine, London, United Kingdom Institutes of Health, Bethesda, Maryland, United States Olusoji Adeyi Amanda Glassman Director, Health, Nutrition and Population Global Director, Global Health Policy, Center for Global Practice, World Bank, Washington, DC, United States Development, Washington, DC, United States Kesetebirhan Admasu Glenda Gray Minister of Health, Addis Ababa, Ethiopia Executive Director, Perinatal HIV Research Unit, Chris Hani Baragwanath Hospital, Johannesburg, South George Alleyne Africa Director Emeritus, Pan American Health Organization, Washington, DC, United States Demissie Habte Chair of Board of Trustees, International Clinical Ala Alwan Epidemiological Network, Addis Ababa, Ethiopia Director, World Health Organization, Regional Office for the Eastern Mediterranean, Cairo, Arab Republic of Richard Horton Egypt Editor, The Lancet, London, United Kingdom Rifat Atun Edward Kirumira Professor, Global Health Systems, Harvard University, Dean, Faculty of Social Sciences, Makerere University, Boston, Massachusetts, United States Kampala, Uganda Zulfiqar Bhutta Peter Lachmann Chair, Division of Women and Child Health, Aga Khan Professor, University of Cambridge, Cambridge, University Hospital, Karachi, Pakistan United Kingdom Agnes Binagwaho Lai Meng Looi Minister of Health, Kigali, Rwanda Professor, University of Malaya, Kuala Lumpur, Malaysia Mark Blecher Senior Health Advisor, South Africa Treasury Adel Mahmoud Department, Cape Town, South Africa Senior Molecular Biologist, Princeton University, Princeton, New Jersey, United States Patricia Garcia Dean, School of Public Health, Universidad Peruana Anthony Measham Cayetano Heredia, Lima, Peru World Bank (retired) 315 Carol Medlin Richard Skolnik Children’s Investment Fund Foundation, London, Lecturer, Health Policy Department, Yale School United Kingdom of Public Health, New Haven, Connecticut, United States Alvaro Moncayo Researcher, Universidad de los Andes, Bogotá, Colombia Stephen Tollman Jaime Montoya Professor, University of Witwatersrand, Johannesburg, Executive Director, Philippine Council for Health South Africa Research and Development, Taguig City, the Philippines Jürgen Unutzer Ole Norheim Professor, Department of Psychiatry, University of Professor, University of Bergen, Bergen, Norway Washington, Seattle, Washington, United States Folashade Omokhodion Damian Walker Professor, University College Hospital, Ibadan, Nigeria Senior Program Officer, Bill & Melinda Gates Foundation, Seattle, Washington, United States Toby Ord President, Giving What We Can, Oxford, United Kingdom Ngaire Woods K. Srinath Reddy Director, Global Economic Governance Program, President, Public Health Foundation of India, New Oxford University, Oxford, United Kingdom Delhi, India Nopadol Wora-Urai Sevkat Ruacan Professor, Department of Surgery, Phramongkutklao Dean, Koc University School of Medicine, Istanbul, Turkey Hospital, Bangkok, Thailand Jaime Sepúlveda Kun Zhao Executive Director, Global Health Sciences, University Researcher, China National Health Development of California, San Francisco, California, United States Research Center, Beijing, China 316 Advisory Committee to the Editors Reviewers Nicolas Andre, MD, PhD Dan Greenberg, PhD Pediatric Oncology, Children’s Hospital of La Timone, Associate Professor and Chairman, Department AP-HM Marseille, France of Health Systems Management, Guilford Glazer Faculty of Business Management & Faculty of Ronald D. Barr, MD Health Sciences, Ben-Gurion University of the Negev, Professor of Pediatrics, Pathology, and Medicine, Beersheba, Israel McMaster University, Hamilton, Ontario, Canada Raymond Hutubessy, PhD Nazmi Bilir, MD Senior Health Economist, Initiative for Vaccine Professor of Public Health, Faculty of Medicine, Research, World Health Organization, Geneva, Hacettepe University, Ankara, Turkey Switzerland Miguela A. Caniza, MD M. Tezer Kutluk, MD, PhD Associate Member, Director of Infectious Diseases, Professor of Pediatrics and Pediatric Oncologist, International Outreach, Department of Infectious Hacettepe University Cancer Institute, Ankara, Turkey Diseases, International Outreach Program, St. Jude Children’s Research Hospital, Memphis, Tennessee, Cédric Mahé, PhD United States Senior Director, Global Epidemiology, Sanofi Pasteur, Lyon, France Phaik-Leng Cheah, MBBS, MRC Path, FRC Path, MPath, MD Donald Maxwell Parkin, MD, PhD Professor, Department of Pathology, Faculty of Senior Epidemiologist, Centre for Cancer Medicine, University of Malaya, Kuala Lumpur, Prevention, Wolfson Institute of Preventive Medicine, Malaysia Queen Mary University of London, London, United Kingdom Maria Paula Curado, MD, PhD Senior Researcher, International Prevention Research Aloka Pathirana MS, FRCS (Eng), FRCS (Edin) Institute, and Epidemiologist, Accamargo Cancer Professor in Surgery, University of Sri Jayewardenepura, Center, International Research Center, Sao Paulo, Brasil Nugegoda, Sri Lanka Henry Ddungu, MD, consultant Sevket Ruacan, MD Uganda Cancer Institute, Kampala, Uganda Professor of Pathology, Koc University School of Medicine, Istanbul, Turkey Nagi S. El Saghir, MD, FACP Professor of Clinical Medicine, Director, Breast Center Vikash Sewram, PhD (Medicine, Chemistry, of Excellence, NK Basile Cancer Institute, Division and Physiology), MPH (Cancer Epidemiology), of Hematology/Oncology, Department of Internal PhD (Public Health) Medicine, American University of Beirut Medical Director, African Cancer Institute, Stellenbosch Center, Beirut, Lebanon University, Stellenbosch, South Africa 317 Frank A. Sloan, PhD Sidney J. Winawer, MD J. Alexander McMahon Professor of Health Policy Paul Sherlock Chair in Medicine, and Management and Professor of Economics, Duke Gastroenterology and Nutrition University, Durham, North Carolina, United States Service, Department of Medicine, Memorial Sloan-Kettering Cancer Center; Professor Verna Dnk Vanderpuye, MBCHB, FWACS of Medicine, Weill Medical College, Consultant Radiation and Clinical Oncologist, Cornell University, Ithaca, New York, Department of Radiation Oncology, Korle Bu Teaching United States Hospital, Accra, Ghana Cheng-Har Yip, MBBS, FRCSEd, FRCS Katsuri Warnakulasuriya, Oral Med BDS, FDSRCS, Consultant breast surgeon, Breast Centre, Sime Darby PhD, Dip, DSc Medical Centre, Selangor, Malaysia Professor of Oral Medicine and Experimental Oral Pathology, King’s College London, London, United Kingdom 318 Reviewers Index Boxes, figures, maps, notes, and tables are indicated by b, f, m, n, and t respectively. A oral cancer and, 85, 87, 89 abandonment of therapy as risk factor, 37 for breast cancer, 266, 292 Algeria, cervical cancer in, 71 for childhood cancer, 124, 125, 125t, 126–27b, 127, ALL (acute lymphoblastic leukemia). See leukemia 130, 205, 205b, 291 Amarasinghe, Hemantha, 85 due to expense, 282 American Board of Internal Medicine’s Choosing Abascal, W., 189 Wisely campaign, 212 accreditation programs, 125, 195, 199, 200, 203, American Cancer Society Cancer Prevention 207n10, 244 Studies, 254 acute lymphoblastic leukemia (ALL). See leukemia American Dental Association (ADA), 91 acute promyelocytic leukemia (APL), 126b American Pain Society, 166 ADA (American Dental Association), 91 American Society of Clinical Oncology, 111, 212 Adami, H., 215 AMPATH Program (U.S.), 245 Adewole, Isaac F., 223 Anderson, Benjamin O., 45, 216, 223, 263, 268 adjuvant treatment considerations for surgery, anesthesia, administration of, 137–38, 224, 226 230–31 Antillon, Federico G., 121 advertising of cigarettes, 185, 189 antiretroviral therapy, 73, 156 affordability of cancer services, 15. See also catastrophic antiviral therapy for HBV and HCV, 155–56, 158, health expenditures 159, 160 abandonment of therapy due to inability to APL (acute promyelocytic leukemia), 126b afford, 127 areca nut. See nut chewing anti-HBV drugs, 158 Argentina anti-HCV drugs, 156 colorectal cancer in, 11, 108 Afghanistan, radiation therapy in, 265 tobacco control, compared to Uruguay, 189 aflatoxins, 11, 150, 151, 153, 156–57, 159 aromatase inhibitor (AI), 12b, 54, 60 Africa. See also specific regions and countries asbestos exposure, 12, 38 cervical cancer prevention approaches in, 80 Asia. See also specific regions and countries types of cancer in, 29 breast cancer in, 216 AHOPCA (Central American Association of Pediatric cervical cancer in, 71, 71m Hematology Oncology), 130 childhood cancer in, 121 air pollution, 12, 38 health insurance coverage in, 289 alcohol consumption liver cancer in, 147, 150 binge drinking, 157 oral cancer in, 85, 87 liver cancer and, 153–54, 157 types of cancer in, 29 319 Asian Development Bank, 188 incidence and death rates, 28f Asma, Samira, 175 radiation therapy for, 240 Atun, Rifat, 281, 284 underdiagnosis of childhood brain tumors, 121 Australia/New Zealand Bray, Freddie, 23 cancer registries in, 266 Brazil cigarettes in breast cancer treatment in, 60 plain packaging of, 185, 189 childhood cancer in tax rates, 184 cost-effectiveness of treatment, 139 cost-effectiveness study in, 267 mortality rates, 123 lung cancer in, 177 treatment, 127b oral cancer in, 88 colorectal cancer incidence rates in, 103 reporting system for HPV infection and costs of intervention packages in, 13, 277 vaccinations in, 258 oral cancer in, 85 Austria, cervical cancer screening in, 74 percent of total health spending to cancer Axios Healthcare Development, 78 control in, 15 radiation therapy in, 245 B smoking in, 180 Badwe, Rajendra A., 223 knowledge of health information, 184 Baltussen, R. M., 59, 216 tobacco use in, 10 Bamako Global Ministerial Forum on Research in breast cancer, 45–68. See also Breast Health Health (2008), 249 Global Initiative Call to Action for Research on Health, 249, 250b, 256 abandonment of therapy, 266, 292 Bangladesh access to diagnostic and treatment facilities, 225 cervical cancer in, 71, 74 assessment of local situation, 57–59 costs of cancer treatments in, 7, 236 social and cultural barriers, 58 smoking in, 178, 180 target group identification, 58–59 Belarus, oral cancer in, 85 breast awareness education, 50t, 52, 215 benefits of cancer control, 17. See also cost-effectiveness breast self-examination, 52, 216 “best buys” chemotherapy, 11, 12b, 54–55, 55–56t, 275 in cancer services, 8 clinical breast examination (CBE), 51–52, 60, 63, in research, 259–60 215–16, 235, 266, 277 betel nut. See nut chewing, as oral cancer risk clinical evaluation, 53–54 bevacizumab, 13 clinic-based cluster randomized trial, 62–63 Bhakta, N., 138–39 combination screening-treatment interventions, BHGI. See Breast Health Global Initiative 60–61 Bhutan, HPV vaccination in, 78 comparison of modalities, 61 Bill & Melinda Gates Foundation, 189 cost-effectiveness of interventions, 13, 13f, 59–61, binge drinking, 157 268, 269–70t biobanks and biological resource centers, 252–53 dedicated cancer center (Egypt), 206b biopsy procedures, 54, 63, 227–28. See also diagnosis, treatment, and patient triage, 63 tissue sampling diagnosis guidelines, 53–54, 53t biostatisticians, 257 diagnostic imaging, 54 bladder cancer disparities in global outcomes, 45–47 radiation therapy for, 240 early-stage treatment of, 8, 11, 29, 60 schistosomiasis and, 12 cost-effectiveness of, 13, 13f blood donations/transfusions and HCV risk, costs of, 14t 153, 155, 158 early detection, 45, 47, 49, 50t, 60 Bloomberg Philanthropies, 189 in LMICs, 12b Bolivia, HPV vaccination in, 78 scaling up, 16 bone marrow transplantation, 129, 197b, 204, 240 economic analyses and future of breast health B. P. Koirala Memorial Cancer Hospital (Nepal), 206b care in LMICs, 63–64 brachytherapy, 241 estrogen receptor protein, testing for, 11, 12b brain cancer field studies, 61–63 320 Index by geographical region, 29 hormonal and reproductive factors, 38 high-risk screening, 212 HPVs and hepatitis B and C viruses and, 36–37 incidence and death rates, 6t, 8f, 27–29, 28f, 45, by income group, 27–29, 31, 31–35f 46m, 215 infections, cancers associated with, 36–37, 37f disproportionate death rates of the young, 45 key indicators of, 24–27 by income group, 31f, 47–48f, 224f incidence, 25 transitions in low- and middle-income countries, income as proxy for human development, 26 29–32 mortality, 25–26 intervention packages, 9t proportion of burden attributable to risk factors, literature review, 59–60 26–27 mammographic screening. See mammograms obesity and, 37–38 overdiagnosis concerns, 213, 215 occupational risk factors, 38 population-based randomized trial, 62 overview, 6–7, 6t, 8f, 8t, 17, 23–24 radiation therapy for, 11, 54, 55–56t, 60, 240 physical activity, lack of, 38 rate of curing, 4b projected (2030), 9t, 17, 31–32, 36f resource limitations, need for guidelines to address, risk factors, 33–34 48–49 sun exposure and skin cancer, 38 risk factors and risk reduction strategies, 47–48 surgically treatable cancers in LMICs, 224, 224f screening for, 10–11, 212, 215–16 surveillance mechanisms, need for, 39–40 cost-effectiveness of, 13f tobacco-related cancer, 32, 34–36 gender inequity as issue, 214 transitions of various types of cancers, 23, 29–32 HIV-positive women, 219 Burkitt, Denis, 134 maternal or reproductive health policy as Burkitt lymphoma, 11, 123, 129t, 134–37, 136t umbrella for, 219 priority in LMICs, 218 C surgery, 54, 223, 226, 228, 229t, 230 Cambodia survivor groups, formation of, 58, 58f cervical cancer in, 71 systemic pharmacotherapy, 54–57 HPV vaccination in, 78 temporal trends in stage of disease, 61–62 liver fluke infection in, 154 tissue sampling, 54, 63 Cameroon treatment guidelines, 54–57, 55–56t costs of cancer treatments in, 7, 236 tumor markers, 54 HPV vaccination in, 78 Breast Health Global Initiative (BHGI) Canada design of, 48–49 Canadian National Breast Screening Study, 215 early detection guidelines, 49–57, 50t Cancer Care Ontario’s Program in Evidence-Based effectiveness of, 267 Care, 198 evidence-based, resource-stratified approach, 216 childhood leukemia in, 122 Ghana-Norway partnership as part of, 199 colorectal screening program (ColonCancerCheck) guidelines developed by, 12b, 198, 218 in, 108 implementation, 57–59 liver cancer, cost of treatment in, 159 resource levels, 49b, 263–64 radiation therapy in, 245 Bretthauer, M., 215 smoking in British doctors, cohort study of, 254 excise tax increases, 184 British Royal College of Physicians, 185 inverse relationship of consumption and burden of cancer, 23–44. See also specific types of cancer price, 183, 183f alcohol consumption and, 37 smuggling and black market sales of asbestos exposure and, 38 cigarettes, 187 compared to other noncommunicable diseases, taxation to finance universal health coverage in, 284 24, 24f Victoria Hospice (Victoria, British Columbia), 206b diet and, 37–38 cancer awareness diversity of cancer by type and geographical breast awareness education, 50t, 52, 215 region, 29 cultural practices and, 225 environmental factors, 38 public awareness of, 16, 255 Index 321 Cancer Care Ontario’s Program in Evidence-Based radiation therapy and, 275 Care, 198 research support and, 250b cancer death rates. See also specific types of cancer surgery and, 232–35, 275 decreases (2000 to 2010), 6, 8t, 17 case-control studies, 253 in high-income countries (HICs), 2–3m, 6t, 8f catastrophic health expenditures, 7, 282, 284, 290 as key indicator of burden of cancer, 25–26 Centers for Disease Control and Prevention (U.S.), 252 in low- and middle-income countries (LMICs), Central American Association of Pediatric Hematology 2–3m, 6t, 8f Oncology (AHOPCA), 130 projected (2030), 9t, 17, 245 Central Asia by region, 27f alcohol consumption in, 157 cancer incidence. See also specific types of cancer cervical cancer in, 69 as key indicator of burden of cancer, 25, 27f HPV vaccination in, 79 most common cancers by gender, 30m liver cancer in, 149 projected (2030), 36f Central Europe registries compiling information on, 251 colorectal cancer in, 102 Cancer Incidence in Five Continents (IARC), 25, 57, 266 liver cancer in, 149 Cancer Intervention and Surveillance Modeling oral cancer in, 88 Network (CISNET), 111 cervical cancer, 69–84 cancer registries. See registries burden of, 69–71, 216, 295, 300–301 cancer research. See research support global, 69, 70f cancer screening, 211–22. See also specific types of cancer regional, 71, 71m, 72f care pathway for positive results, development of, 214 chemotherapy, 11, 78, 275 cost-effectiveness of, 214, 275–76 cost-effectiveness of interventions, 8, 13, 78–79, 81, criteria for, 213 217, 268, 271t diagonal approach to, 218–19 cytology classification and terminology, 72–73, effectiveness of, 10–11, 275–76 297–98 ethical considerations, 214–15 early-stage treatment of, 8, 11 false-positive screening tests, 213 scaling up, 16 follow-up strategies, 214 epidemiology and biology research, 257 high-risk screening, 212 by geographical region, 29 infrastructure, education, and advocacy, 213–14 HIV and, 73, 219 innovation, role of, 218 HPV infection, integrated research on link to, lead-time bias and, 212 257–59, 258–59b. See also HPV (human length bias and, 213 papillomavirus) infection opportunistic vs. organized screening, 108, 211–12, HPV vaccination programs for. See HPV (human 266, 277 papillomavirus) vaccination organized screening requirements, 108, 213–15 incidence and death rates, 6t, 8f, 27–29, 28f, 69, overdiagnosis concerns, 213, 215 215, 216 overview, 211 by income group, 34f, 72f, 224f policy considerations, 219 median age at death, 71 population-based screening, 212 trends in, 29 priorities in LMICs, 218 integrated research exemplar, 257–59 structural obstacles to, 214 natural history of, 72 target age range of, 212 overview, 69 Wilson-Junger criteria for, 213, 213b Pap smears and, 69, 73, 74, 216 cancer services for comprehensive cancer center. See radiation therapy for, 11, 78, 240 comprehensive cancer centers rate of curing, 4b cancer surveillance systems. See surveillance systems research support for, 259–60 cancer survival rates, 291. See also specific types of cancer screening for, 10, 69, 73–77, 74t, 216–17 in high-income countries (HICs), 1, 6t case study of upscaling VIA, 75 in low- and middle-income countries (LMICs), 6t combined HPV vaccination and screening, Canfell, K. H., 296 80–81, 217 capacity building. See also scaling up cost-effectiveness of, 80 322 Index cryotherapy (screen-and-treat approach), 217 ineffectiveness of prevention and screening, 123 gender inequity as issue, 214 International Incidence of Childhood Cancer HIV-positive women, 219 (IARC), 123 HPV testing, 75–77 in LMICs, 123, 125t, 127, 128, 139 maternal or reproductive health policy locally adapted treatment protocols, 126–27, 139 as umbrella for, 219 Mexican national health insurance coverage of, 1 priority in LMICs, 218 mortality rates, 123 self-collection by women, 76 outcome evaluation, 127–28 target age for, 212 rate of curing, 4b triage of positive HPV tests, 76–77 retinoblastoma, 137–38 visual inspection with acetic acid (VIA), 74–75, scaling up, 277 217, 266, 277, 297–98 spillover effect from pediatric to adult surgery, 228, 229t oncology, 123 treatment of, 9t, 78 treatment of, 11, 123–25 costs of, 14t dedicated centers, 123–24, 124t, 128 cetuximab, 13 infrastructure needed in LMICs, 125t Chaloupka, Frank. J., 175 principles of, 126–28 Chao, Ann, 249 specific cancers, 128–38 chemotherapy treatment facilities, organization of, 9t, 16, 125t breast cancer, 11, 12b, 54–55, 55–56t, 275 twinning programs, 124–25, 130, 140, 266 Burkitt lymphoma, 135–37 underdiagnosis of childhood brain tumors, 121 cervical cancer, 11, 78, 275 Wilms tumor, 11, 126, 133–34, 133t, 135t childhood cancer, 124, 125t, 126, 128, 129t Children’s Oncology Group in North America, 133 colorectal cancer, 110, 111, 112, 114, 275 Chile comprehensive cancer center services, 200, 231 opioids for pain relief in, 169 in conjunction with radiation therapy, 240, 243 training of health care professionals in, 204 in conjunction with surgery, 231 China health insurance coverage and, 290 air pollution in, 38 Hodgkin lymphoma, 131–32, 132t breast cancer in, 52, 60 oral cancer, 94, 275 cervical cancer in retinoblastoma, 138 burden of, 301 scaling up, 16, 275 screening program, 295–96 Wilms tumor, 133–34 childhood cancer treatment in, 125 Chen, Chien-Jen, 147 acute lymphoblastic leukemia (ALL), 128, 130–31 Chen, Wendong, 147 chronic lung disease in, 178 Chennai Prospective Study, 254 colorectal cancer in, 103, 111 childhood cancer, 121–46. See also leukemia cost-effectiveness of HPV vaccination in, 295–305. abandonment of therapy, 124, 125, 125t, 126–27b, See also HPV vaccination 127, 127b, 130, 205, 205b, 291 costs of intervention packages in, 277 achievability of cure, 123 dietary changes in, 150, 156 burden in LMICs of, 121–23 HBV infection in, 151 Burkitt lymphoma, 11, 123, 129t, 134–37, 136t HBV vaccination in, 150, 155, 158 characteristics of most common childhood HCV infection in, 151 cancers, 129t HPV vaccination in, 295–305. See also HPV chemotherapy, 124, 125t, 126, 128, 129t vaccination cost-effectiveness of interventions, 13, 13f, 138–40, liver cancer in, 149, 150 139–40f, 268, 275–76t, 277 liver fluke infection in, 154 costs of interventions, 14t national health insurance in, 284, 285t, 289, 290, 292 future directions for progress, 140 cancer surgery coverage, 7 in HICs, 121, 126 co-payments, 291 Hodgkin lymphoma, 129t, 131–33, 132t, 240 fee-for-service basis, 292 incidence rates, 121 opioid accessibility in, 170 inclusion in cancer registries, 121, 122f oral cancer in, 88, 94 Index 323 percent of total health spending to cancer intervention packages, 9t control in, 15 low-income countries recommendations, 114 smoking in, 179, 180 middle-income countries recommendations, 115 affordability, 182 overview, 101 implications of tobacco taxation, 188 physical inactivity, relationship to, 38 knowledge of health information, 184 radiation therapy for, 110 male smoking prevalence, 30, 35, 177 recommendations, 114–16, 115t price elasticity, 181 research support for, 260 related cancer deaths in, 177 risk factors, 105 starting age of men, 178 screening for, 11, 101, 105–9, 217 stomach cancer in, 218 ColonCancerCheck (Canadian program), 108 types of cancer in, 29 colonoscopy, 106–7, 217 China Kadoorie Biobank, 254, 255b computerized tomographic colonography chronic lung disease, 178 (CTC), 106 chronic trauma, as risk factor for oral cancer, 90 cost-effectiveness of, 11, 111–14 cigarettes. See tobacco taxes; tobacco use fecal DNA, 106 cirrhosis, 147, 150, 151, 153 fecal immunochemical test (FIT), 105–6, 277 CISNET (Cancer Intervention and Surveillance flexible sigmoidoscopy (FS), 106 Modeling Network), 111 guaiac fecal occult blood test (gFOBT), 105, 108 Cleary, James, 165 high-risk screening, 212 clinical breast examination. See breast cancer IARC guidelines, 107, 108–9 Clinton Health Access Initiative, 16 ICRCSN pilot program, 108 Clonorchis sinensis, 36 International Cancer Screening Network, 109 Cochrane Collaboration, 91 International Digestive Cancer Alliance, 109 codeine, 165, 166, 168 organized vs. opportunistic screening, 108 cohort studies, 254, 254–55b performance of organized screening Colombia programs, 108 breast cancer in, 62 USMSTF guidelines, 108 childhood cancer in, 16 USPSTF guidelines, 107–8 universal health care in, 284, 285t World Endoscopy Organization, 109 cancer coverage, 290, 291 staging of, 109–10 incentives to improve coverage, 292 barium enema, 109 colonoscopy, 106–7, 217 computerized tomography, 109–10 colorectal cancer (CRC), 101–19 surgery, 110, 223, 228, 230, 230t burden of, 101–5, 217 treatment of, 110–11 chemotherapy, 110, 111, 112, 114, 275 international partnership arrangements to cost-effectiveness of interventions, 7b, 13, 13f, 114, support in LMICs, 111 217, 268, 272–73t metastatic colorectal cancer, 110–11 diagnosis of, 109 Commission on Macroeconomics and Health, 7b diet, lifestyle, and medications as risk factors for, community health centers, 226–27, 228–30t 105, 114, 217 community health workers, 200, 201, 256, 290 European Guidelines for Quality Assurance in community involvement, 214 Colorectal Cancer Screening and Diagnosis comprehensive cancer centers, 195–210 (IARC), 107, 108 accreditation programs, 207n10 genetic factors, 105 additional key supports, 202 high-income countries recommendations, 116 administration and management, 201 high-risk screening, 212 affordability and domestic financing of, 15 incidence and death rates, 6t, 8f, 27–29, 28f, 101–5 cancer system components, 195–96, 196f by gender, 102m, 103–4f, 215 cancer system planning, 195 by income group, 33f, 103–5, 105f, 224f chemotherapy, 200, 231 transitions in low- and middle-income clinical management, 197–98, 198f countries, 29 clinical practice guidelines, 197–98 trends, 102–3 patient care plans, 197 324 Index clinical services, 198–200, 198f of interventions, 4b, 7b, 13–14, 14t, 277 core services, 198f, 200–202 priority of reduction for essential intervention financial systems, 202 packages, 16 framework, 197–202, 198f of surgery, 235–36 global health, contribution to, 204–5 Council of Europe, 168 human resources, 201 CRC. See colorectal cancer infection control, 201–2, 207n9 Cuba, oral cancer in, 92, 94 information technology and management, 201 cultural practices in LMICs, 196 breast cancer and, 58 medical imaging (diagnostic radiology), 199 cancer awareness and, 225 office- and clinic-based ambulatory care, 199 cancer screening and, 214 palliative care, 200 cervical cancer and, 71 pathology and laboratory medicine, 199 fatalism and, 225 pharmacy services, 201 surgery and, 225 psychosocial support, 200 Cure4Kids, 125 quality assurance, 202 customs reform to allow for importation of health quality of care with minimal resources, 202–5 research components, 253, 256 accessibility, 202 cytosine, 186 catalyzing development of national cytotoxic regimens, 12b systems, 203 Czech Republic, oral cancer in, 87 effectiveness, 202 efficiency, 203 D equity, 202 Dare, Anna J., 223 responsiveness, 202 D’Cruz, Anil, 195 safety, 202 death certificates, indicating smoking habits of radiation therapy, 199–200, 243 deceased, 185 regional cancer centers, establishment of, 203 death rates. See also cancer death rates role in support of development of effective top 10 causes of death worldwide, 24, 24f health care systems, 204 death registration systems, linking to cancer supportive care, 200 registration systems, 253. See also registries surgery, 199, 232 delayed presentation for cancer as norm in LMICs, survivorship programs, 200 213–14, 234, 282, 290 systemic cancer therapy, 200 Democratic Republic of Congo, liver cancer in, 149 training of health care professionals, 203–4 Denny, Lynette, 69, 268 Congo, Democratic Republic of, liver cancer in, 149 diagnosis and staging, 11. See also specific types of co-payments, 291 cancer cost-effectiveness, 13, 13f, 267–69. See also specific role of laboratories and anatomic pathology, 252 types of cancer surgery’s diagnostic role, 227 cancer screening, 214, 275–76 UICC Tumor, Nodes, Metastasis (TNM) staging estimates of, 7b system, 92 HPV vaccination, cost-effectiveness in China, diagnostic imaging 295–305. See also HPV (human breast cancer, 54 papillomavirus) vaccination comprehensive cancer center services, 199 methods, 267–68 follow-up to positive screening test with, 214 results based on level of resources, oral cancer and, 92 268–69, 268t surgery’s reliance on, 226 studies of, 263, 267 diet, relationship to cancer, 37–38 surgery, 235–36 colorectal cancer, 105 costs. See also affordability of cancer services; liver cancer, 150, 156 financing; specific types of cancer digital mammography, 51 of cancer drugs, 12b Dikshit, R., 253 catastrophic health expenditures, 7, 282, 284, 290 disability-adjusted life years (DALYs) DCP3 analysis of, 5b cervical cancer and, 71, 216 Index 325 cost-effectiveness and DALYs averted, 13, 214 cervical cancer screening in, 74 liver cancer and, 150 domestic funding in, 282 Disease Control Priorities opioid accessibility in, 168–69 HPV (human papillomavirus) vaccination in European Guidelines for Quality Assurance in Colorectal essential package of, 8 Cancer Screening and Diagnosis (IARC), 107, publication background of, 5b 108, 109 DNA testing for HPV. See HPV (human European Medicines Agency, 258 papillomavirus) infection Europe/European Union. See also specific regions Doll, Richard, 176 and countries Dominican Republic alcohol consumption in, 157 childhood cancer treatment in, 130 cancer registries in, 266 universal health care in, 284, 286t, 289 cervical cancer in, 69 drug costs. See affordability of cancer services HPV vaccination in, 79 drug therapy. See also chemotherapy; pain relief liver cancer in, 150 breast cancer, 55–56t, 55–57 lung cancer in, 177 comprehensive cancer centers, 201 radiation therapy in, 244 constraints on, 265 smoking in, 180 as risk factor, 105 smoking cessation treatment, 186 F universal health care, coverage of, 290 Fakous Cancer Center (Egypt), 206b Dutch Childhood Oncology Group, 130–31 false-positive screening tests, 213 farming practices. See aflatoxins E FCTC (Framework Convention on Tobacco Control), Earle, Craig, 101 10, 89, 188, 189 East Asia and Pacific fee-for-service payments, 292 aflatoxins in, 153 fentanyl patch, 168, 169, 170 HBV infection in, 151 Ferlay, J., 25 HCV infection in, 155 financing, 281–94. See also health insurance liver cancer in, 147, 149 cancer care and, 15, 282–83 stomach cancer in, 218 co-payments, 291 Eastern Europe country approaches, 283 colorectal cancer in, 102 domestic funding, 282 oral cancer in, 85, 87 need for financial protection, 282 types of cancer in, 29 out-of-pocket spending by families, 282, 290 e-cigarettes, 186 overview, 281–82 Egypt progressive universalism, 283 breast cancer in, 206b public financing, 282, 290–91 cervical cancer in, 71 research support and, 256 Fakous Cancer Center, 206b surgery and, 235–36 HCV infection in, 151, 153 Finland, ban on advertising of cigarettes in, 185 incidence of HBV vs. HCV in, 151, 152f first-level hospitals liver cancer in, 149 intervention packages, 9t National Cancer Institute, 206b surgery in, 227, 228t endrocrine therapy, 54–55, 55–56t flukes. See liver fluke infection environmental factors, 38 Foley, K. M., 166 epidemiologic studies, 253–57, 260 Fonseca, M., 60 Epstein-Barr virus, 36, 131 Food and Drug Administration (U.S.), 258 Ergo, A., 283 food and grain storage. See aflatoxins erythroplakia, 90 Framework Convention on Tobacco Control (FCTC), esophagus. See oral cancer 10, 89, 188, 189 estrogen receptor protein, testing for, 11, 12b, 45 France ethical considerations of cancer screening, 214–15 HCV treatment in, 156 Ethiopia liver cancer in, 149 326 Index oral cancer in, 85 globalization smoking in comprehensive cancer centers, contribution to, cigarette taxes, 10, 184 204–5 inverse relationship of consumption and price, initiatives for cancer control, 16–17 183, 183f tobacco use and, 189 restrictions in public places, 185 Global Monitoring Framework for Noncommunicable Franceschi, Silvia, 147 Diseases (WHO), 166 Fred Hutchinson Cancer Research Center (Seattle, Global Opioid Policy Initiative (GOPI), 168, 170–71 Washington), 205b Global Task Force on Expanded Access to Cancer Care Frenk, J., 214 and Control, 218–19 Global Task Force on Radiotherapy for Cancer G Control, 16 Gakidou, E., 74 Global Youth Tobacco Survey, 182 Gambia, liver cancer in, 149 Goldie, S. J., 80, 296 Gardasil Access Program, 78 Gospodarowicz, Mary K., 195, 239 gastric cancer. See stomach cancer Greenberg, D., 111, 267 Gauvreau, Cindy L., 1, 223, 263 Gross, Thomas, 249 Gavi (The Vaccine Alliance), 10, 16, 36, 78–80, 155, 299 Guatemala Gelband, Hellen, 1, 147, 165, 268, 281 childhood cancer treatment in, 127b, 130, 197 gender differences Unidad Nacional de Oncología Pediátrica cancer incidence and mortality by geographical (UNOP), 140 region, 29, 215 opioid accessibility in, 170 delay in screening for women, 214 Guinea liver cancer, 148m, 149 cervical cancer in, 71 lung cancer, 29 liver cancer in, 149 most common cancers by gender, 30m, 215 Guinea-Bissau, liver cancer in, 149 oral cancer, 86t, 87–88m Gupta, Prakash C., 175 smoking and associated cancer risks, 35–36, 176, Gupta, Sumit, 121, 205 177f, 180 Guyana, cervical cancer in, 71 genetic factors Gwatkin, D. R., 283 colorectal cancer, 105 high-risk screening and, 212 H liver cancer, 149 Haiti oral cancer, 89–90 domestic funding in, 282 retinoblastoma, 137 HPV vaccination in, 78 Georgia, opioid accessibility in, 169 Hall, Andrew, 147 Germany Harif, Mhamed, 121 opioid accessibility in, 170 HBV. See hepatitis B and C viruses smoking restrictions in public places in, 185 HCC (hepatocellular carcinoma). See liver cancer Ghana HCV. See hepatitis B and C viruses breast cancer screening in, 290 HDI (Human Development Index), 31, 71, 81n2, cervical cancer screening in, 217 166, 226 Ghana-Norway partnership as part of Breast Health HDV (hepatitis D virus), 149 Global Initiative, 199 head and neck cancers HPV vaccination in, 78 imaging techniques for, 198 liver cancer in, 149 radiation therapy for, 240 universal health care in, 284, 286t, 289, 290 health care professionals taxes to finance, 292 research support role of, 250, 251b Ginsburg, Ophira M., 211 screening by community care workers GlaxoSmithKline, 299 and nurses, 218 Global Adult Tobacco Survey, 180, 182, 184 shortage of skilled medical personnel, 232–33, 266 Global Fund for AIDS, TB and Malaria, 16 training of, 203–4 Global Initiative for Cancer Registry Development, 39 health communications, 255 Index 327 health insurance lung cancer in, 29 as constraint, 266, 282 palliative care in, 11, 167, 169 coverage by population group, 284, 289 percent of total health spending to cancer fee-for-service basis, 292 control in, 15 incentives, 292 prostate cancer in, 29 results of coverage, 291–92 retinoblastoma in, 138 Seguro Popular (Mexico), 1, 125, 204, 266, 283, 284, smoking in 287t, 291 advertising of cigarettes, ban on, 185 services and conditions covered, 289–90 e-cigarettes, 186 universal health coverage of cancer services, 4b, 234, health information and counter-advertising, 281, 283, 284, 289 184–85 Health Insurance Plan of Greater New York, 51 public places, restrictions on, 185 health surveillance systems. See surveillance systems tobacco control, investment in, 189 heart attack, smoking’s relationship to, 178, 185, 187 tobacco excise taxes, 182–83 hepatitis B (HBV) and C (HCV) viruses, 154–56 types of cancer prevalent in, 27 antiviral therapy for, 155–56, 158, 159, 160 Hill, A., 156 burden of cancer linked to, 36–37 HIV. See also antiretroviral therapy liver cancer, 149–53 cervical cancer and, 73 communication about, 255 screening HIV-positive women for breast and incidence of HBV vs. HCV in, 151, 152f cervical cancer, 219 prevention of HCV, 156b Hodgkin lymphoma, 129t, 131–33, 132t, 240 screening for, 155–56 Honduras treatment of HBV, 155. See also hepatitis B cervical cancer in, 71 vaccination childhood leukemia in, 16, 122, 197 cost-effectiveness of, 158 Hong Kong SAR treatment of HCV, 155–56 colorectal cancer in, 111 cost-effectiveness of, 158–59 oral cancer in, 88 hepatitis B (HBV) vaccination, 4b, 154–55 smoking risks in, 178–79 burden of cancer and, 36 hormonal and reproductive factors, 38 communication about, 255 Horton, Susan, 1, 101, 263, 281 cost-effectiveness of, 8, 13, 13f, 158, 267 Hospice Africa Uganda, 168, 170 costs of, 14t hospice-palliative care, 167 global coverage, 10, 147 hospitals. See also comprehensive cancer centers plasma-derived vaccine, 15 intervention packages, 9t scaling up, 15, 149 research role of, 250 hepatitis D virus (HDV), 149 surgery in, by hospital level, 227, 228t hepatocellular carcinoma (HCC). See liver cancer upgrading, 16 herpes virus type 8, 36 households, impact on, 7–8. See also catastrophic health Herrero, Rolando, 69 expenditures high-income countries (HICs) Howard, Scott C., 121 advanced cancer treatment in, 12 HPV (human papillomavirus) infection breast cancer in, 45, 48, 215 burden of cancer linked to, 36–37, 69, 71, 72, 282, Burkitt lymphoma in, 135 300–301 cancer death rates in, 2–3m, 6t, 8f, 24, 27–28, 165 communication about, 255 cancer survival rates in, 1 DNA testing for, 233, 258, 266, 267 cervical cancer in, 70f future research needs, 258–59, 258–59b childhood cancer in, 11, 121 integrated research establishing link to cervical colorectal cancer in, 101, 108, 116 cancer, 257–59 cost-effectiveness of cancer interventions in, 7b, 268 oral cancer and, 87, 88, 90, 218 HCV infection in, 155 screening for, 10, 75–77, 212 high-risk screening for breast or ovarian in China, 296–97 cancer in, 212 HPV (human papillomavirus) vaccination liver cancer in, 150 case studies of implementation, 78 328 Index combined HPV vaccination and screening, 80–81, 217 cancer death rates in, 7 communication about, 255 cancer surgery costs in, 7 cost-effectiveness in China, 295–305 cervical cancer in, 71, 217, 258 benefits and costs of publicly financed program, childhood cancer in, 121, 122 299–301, 300t chronic lung disease in, 178 cost assumption, 303 costs of intervention packages in, 13, 282 cost data sources, 299 guidelines adapted to local resource delay in approving HPV vaccines, 303 availability in, 198 extended cost-effectiveness analysis, 296–97 HBV infection rates in, 149 limitations of analysis, 303 HBV vaccination in, 158 model, 297, 298t HCV treatment, cost-effectiveness of, 159 policy reform to expand present Expanded HPV testing and reduction in cervical cancer in, 258 Program for Immunization, 303 liver cancer in, 149, 150 results, 299–300, 302t Million Death Study, 39 sensitivity analysis, 299, 300–301t National Cancer Control Program, 203–4b strategies, data, and assumptions, 297–98 National Cancer Grid, 16, 231 cost-effectiveness of, 13, 13f, 79–80, 217, 267 opioid accessibility in, 168, 170 in China, 295–305. See also above: cost- oral cancer in, 11, 85, 89, 92, 94, 95 effectiveness in China percent of total health spending to cancer costs of, 14t, 282 control in, 15 in DCP3 Cancer essential package, 8 radiation therapy in, 245 efficacy, 4b, 77 regional cancer centers in, 203–4b Gavi subsidy of, 36 retinoblastoma in, 137 national programs, 10, 69 smoking in, 178, 180 public health challenges to implementing, 77 cigarette taxes, 10, 188 scaling up, 15, 217 knowledge of health information, 184–85 universal health coverage of, 291 related cancer deaths in, 178 Human Development Index (HDI), 31, 71, 81n2, surgery, private vs. public sector provision of 166, 226 services in, 234 human immunodeficiency virus. See HIV tuberculosis in, 178 human rights, pain relief as, 165 universal health care in, 284, 286t, 289, 290, 292 Human Rights Watch, 168 cap on payments, 292 human T-cell lymphotropic virus type 1, 36 co-payments, 291 Hungary, oral cancer in, 87, 88 Indonesia Hunger, Stephen P., 121 acute lymphoblastic leukemia (ALL) in, 128 breast cancer detection in, 51 I childhood cancer treatment in, 127b IAHPC (International Association for Hospice and smoking in, 180 Palliative Care), 167, 170 male smoking prevalence, 30 IARC. See International Agency for Research on Cancer WTO case on U.S. ban of clove-flavored ICRCSN (International Colorectal Cancer Screening cigarettes, 189 Network), 108, 217 types of cancer in, 29 Ilbawi, Andre, 223 inequality in health, contribution of cancer control to imaging. See diagnostic imaging reducing, 17 implementation science, 255–56 infections, cancers associated with, 36–37, 37f. See also as “best buy” for cancer research, 259 HPV (human papillomavirus) infection INCB (International Narcotics Control Board), 168 information technology and management, 201, 218 incidence. See cancer incidence; specific types of cancer informed consent to mammographic screening, 51 income as proxy for human development, 26 injected drug administration and HCV infection, India 151, 155, 156b acute lymphoblastic leukemia (ALL) in, 128 Institute for Health Metrics and Evaluation, 250 breast cancer in, 45–47, 52, 226, 235 International Agency for Research on Cancer (IARC) Mumbai trial, 216, 218 on alcohol as liver carcinogen, 153 Index 329 on breast self-examination, 52 prevention. See prevention strategies Cancer Incidence in Five Continents, 25, 57, 266 resource requirements, 14, 14t on cervical cancer screening, 73 screening. See cancer screening on clinical breast examination, 51 surgery. See surgery on colorectal cancer screening, 107, 108–9 intestinal cancer. See stomach cancer estimating cancer incidence and death rate, 6 intravenous drug users, HCV infection in, 151, global initiative for cancer registry, 40, 40m 155, 156b GLOBOCAN, 7b, 25, 27, 81n1, 297 Iran, opioid accessibility in, 170 International Incidence of Childhood Cancer, 123 Iraq, radiation therapy in, 265 on liver flukes, 154 Ireland on lung cancer in men in HICs, 177 radiation therapy in, 245 on mammogram screenings, 50 smoking restrictions in public places in, 185 Monographs on the Evaluation of Carcinogenic Israel, liver cancer in, 150 Risks to Humans, 12 Israels, Trijn, 121 national cancer mortality databank of, 26 Italy radiation facilities provided to Sub-Saharan HCV infection in, 151 Africa by, 266 incidence of HBV vs. HCV in, 151, 152f Working Group on Cervical Cancer Screening, 73 International Association for Hospice and Palliative J Care (IAHPC), 167, 170 Jaffray, David A., 239 International Association of Cancer Registries, 25 Jamaica International Atomic Energy Agency (IAEA), 199, 244 cervical cancer in, 71 Programme of Action for Cancer Therapy, 245, 267 opioid accessibility in, 169 International Cancer Screening Network, 109, 267 Jamison, Dean T., 295 International Childhood Cancer Cohort Japan Consortium, 254 cancer death rates in, 24 International Classification of Diseases, 25, 26 quitting smoking’s effect on, 178 International Colorectal Cancer Screening Network smoking’s effect on, 175–76, 177f (ICRCSN), 108, 217 colorectal cancer screening in, 108 International Digestive Cancer Alliance, 109 HCV infection in, 151 International Federation of Gynecology and incidence of HBV vs. HCV in, 151, 152f Obstetrics, 78 liver cancer in, 149 International Narcotics Control Board (INCB), 168 lung cancer in, 177 International Network for Cancer Treatment and mammography in, 63 Research, 206b oral cancer in, 88 International Observatory of End of Life Care, 167 stomach cancer in, 218 international partnerships. See also twinning programs Japan Life Span Study, 254 colorectal cancer and, 111 Jha, Prabhat, 1, 175 effectiveness of, 266–67 Johnson, Newell, 85 research support, 257 Jordan International Society of Oncology Pharmacy childhood leukemia in, 122 Practitioners, 201 King Hussein Cancer Foundation, 196, International Society of Pediatric Oncology 196–97b (SIOP), 125 opioid accessibility in, 168 Abandonment of Treatment Working Group, 127 International Society of Radiology, 199 K interventions. See also specific types of cancer Kalager, M., 215 cost-effectiveness of, 4b, 7b, 13, 13f Kaposi sarcoma, 29 diagnosis and treatment, 11 Kenya essential packages of, 8–13, 9t HPV vaccination in, 78 implementation challenges for, 15–16 opioid accessibility in, 168–69 local priority conditions, 11–13 radiation therapy in, 245 palliative care. See pain relief Khader, Jamal, 195 330 Index Kim, Jane J., 69, 295 in LMICs, 129–30 King Hussein Cancer Foundation (Jordan), 196, treatment of, 129–30 196–97b acute myeloid leukemia, 197 Knaul, Felicia, 195, 214, 281, 284 acute promyelocytic leukemia (APL), 126b Korea, Republic of incidence and death rates, 6t, 28f breast cancer treatment in, 60 childhood, 121–23, 122f colorectal cancer in, 108, 111 radiation therapy for, 240 fee-for-service payments in, 292 leukoplakia, 90 liver cancer in, 149 Levin, Carol, 69, 249, 295 liver fluke infection in, 154 Libya, cervical cancer in, 71 mammograms in, 60 lifestyle oral cancer in, 88 breast cancer and, 45, 47 stomach cancer in, 218 cancer and, 23–24, 29, 32, 38 thyroid cancer in women in, 29 colorectal cancer and, 105, 114, 217 Lipscomb, Joseph, 45 L liver cancer, 147–64 laboratories. See also pathology aflatoxins and, 11, 150, 151, 153, 156–57, 159 in comprehensive cancer centers, 199 alcohol consumption and, 153–54, 157 constraints, 265 cost-effectiveness of interventions, 13, 13f, 158–60, research role of, 252 268, 273–74t Ladner, J., 78 costs of treating, 159–60 Lambert, R., 114, 217 global burden of, 147–50 Lan, N. H., 235 HCC (hepatocellular carcinoma) as most Lancet Commission on Investing in Health, 284 common type, 147 Lao People’s Democratic Republic incidence and death rates, 6t, 8f, 27–29, 28f, HPV vaccination in, 78 147–49, 148m liver cancer in, 149 age-specific incidence, 149 liver fluke infection in, 154 by gender, 148m, 149, 215 larynx. See oral cancer by income groups, 149–50, 150f Latin America and the Caribbean. See also specific trends, 150 countries liver fluke infection, 11–12, 149, 154, 157–58, 277 aflatoxins in, 157 obesity and, 147, 154 alcohol consumption in, 157 prognosis and DALYs, 150, 158 cervical cancer in, 71, 71m research support for, 259–60 health care coverage, variations in, 284, 289 risk factors, 150–54, 160t. See also hepatitis liver cancer in, 149, 150 B and C viruses Network of National Cancer Institutions of Latin treatment of, 9t, 158, 160, 160t. See also hepatitis America (RINC) initiative, 245 B vaccination oral cancer in, 85 vaccination programs for, 32, 154–56. See also radiation therapy in, 244 hepatitis B vaccination types of cancer in, 29 London, W. Thomas, 147 lead-time bias, 212 low- and middle-income countries (LMICs) Lee, S., 60 air pollution in, 38 length bias, 213 alcohol consumption in, 157. See also alcohol Lesotho, HPV vaccination in, 78 consumption leukemia breast cancer in, 45, 215–16. See also breast cancer acute lymphoblastic leukemia (ALL), 11, 127, compared to HICs, 216 128–31 cost-effective interventions, 59–61 characteristics of, 129t early-stage treatment, 12b chemotherapy, 129–30 incidence and death rates, 27, 59 costs of treatment, 130–31, 290 Burkitt lymphoma in, 134. See also Burkitt diagnosis of, 129 lymphoma in HICs, 128–29 cancer incidence and death rates in, 2–3m, 4b, Index 331 6t, 8f, 17, 27, 165, 166. See also specific air pollution and, 38 types of cancer asbestos exposure and, 38 cancer screening priorities in, 218. See also cancer burden of, 29–30 screening gender differences, 29, 177–78, 215 cancer statistics, need to compile and monitor in, 24, by geographical region, 29, 177 26, 39 incidence and death rates, 8f, 27–29, 28f, 215 cancer survival rates in, 1 by income group, 29 cancer transition in, 29–32 radiation therapy for, 240 cervical cancer in, 69, 70f. See also cervical cancer Luo, X. Q., 131 regional burden, 71, 71m Luxembourg, cervical cancer screening in, 74 screening programs, 75, 216–17 lymphoma. See also Burkitt lymphoma; Hodgkin treatment, 78 lymphoma childhood cancer in, 123. See also childhood cancer incidence and death rates, 28f colorectal cancer in, 108. See also colorectal cancer radiation therapy for, 240 international partnership arrangements to support treatment, 111 M screening programs, 217 MacLennan, Mary, 175 comprehensive cancer centers in, 196. See also Madagascar comprehensive cancer centers cervical cancer screening in, 75 cost-effectiveness of interventions in, 13, 13f, HPV vaccination in, 78 267–69, 270, 275, 276–77t, 277. See also Malawi cost-effectiveness Burkitt lymphoma in, 137 costs of cancer control in, 4b, 14t. See also costs cervical cancer in, 71, 75 delayed presentation for cancer as norm in, 213–14, childhood cancer in, 139 234, 282, 290 HPV vaccination in, 78 economic overview, 263–64. See also nonphysicians performing surgery in, 232 cost-effectiveness Malaysia financing cancer care in, 15, 281 cervical cancer prevention approaches in, 80 HPV vaccination programs in, 10, 77. See also HPV liver fluke infection in, 154 vaccination referral hospital in, 61 inadequate resources in, 264 mammograms, 10–11, 29, 45, 49–51, 51t, 57–58, 63, intervention packages in, 9–10 212, 214, 215–16, 277, 290 laboratories in, 265 mastectomy, 54, 55t liver cancer in, 150. See also liver cancer mate drinking, as risk factor for oral cancer, 90 lung cancer in, 177. See also lung cancer maternal health policy as umbrella for breast and noncommunicable diseases, death rates in, 24, 24f cervical cancer, 219 oral cancer in, 85, 89. See also oral cancer Mauritius, inverse relationship of consumption and cost-effectiveness assessments of preventing and price of smoking in, 183 screening, 94–96 McGlynn, Katherine A., 147 future research needs, 96 MD Anderson Cancer Center at the University pain control in, 166–71. See also pain relief of Texas, 111 radiation therapy in, 239, 265. See also radiation medical records, 250 therapy medications. See drug therapy retinoblastoma in, 137 men, cancer incidence and mortality of. See gender surgery in, 223, 229–30t, 233–34b, 264–65. differences See also surgery Merck, 78, 299 tobacco use in, 23, 35–36. See also tobacco use methodology, 7b affordability and, 182 Metzger, Monika L., 121 smoking cessation patterns, 180 Mexico Westernized lifestyle adopted in, 24, 29, 217 breast cancer, national health insurance coverage of, low-income countries (LICs). See low- and middle- 1, 291 income countries (LMICs) cancer services guidelines in, 198 lung cancer catastrophic health expenditures in, 290 332 Index cervical cancer in, 80, 266 cancer research, commitments to, 249–62. See also childhood cancer in, 1, 125, 204 research support liver cancer in, 149 cancer services, universal health coverage of, 4b, 234 Oportunidades program in, 284 collaboration, 256 percent of total health spending to cancer control effectiveness of, 267 in, 15 intervention packages, 9t Seguro Popular (national public insurance MIC national programs, 234 program), 1, 125, 204, 266, 283, 284, 287t, 291 Native Americans, retinoblastoma among, 137 smoking in NCDs. See noncommunicable diseases cigarette taxes in, 10 neck cancer. See head and neck cancers inverse relationship of consumption and price, 183 needle biopsies, 54, 63 knowledge of health information, 184 Nepal survival rates due to improved coverage in, 291 B. P. Koirala Memorial Cancer Hospital, 206b universal health care in, 284 cervical cancer in, 71 m-Health models, 218 HPV vaccination in, 78 Middle East and North Africa. See also specific countries opioid accessibility in, 169, 170 bladder cancer in, 12 pain medications in, 167 cervical cancer in, 71 Netherlands, smoking restrictions in public colorectal cancer in, 111 places in, 185 HPV vaccination in, 79 Network of National Cancer Institutions of Latin types of cancer in, 29 America (RINC) initiative, 245 middle-income countries (MICs). See low- and newborns, HPB vaccination for. See hepatitis B middle-income countries (LMICs) vaccination Million Death Study (India), 39 New Zealand. See Australia/New Zealand ministries’ research role, 249–50, 251b NGOs. See nongovernmental organizations mobile phones, use of, 218 NIAAA (National Institute of Alcohol Abuse and Moldova, pain medications in, 167 Alcoholism), 153–54 Mongolia Nicaragua, cervical cancer in, 71 HBV vaccination program in, 149 Niger HCV infection in, 151, 155 domestic funding in, 282 liver cancer in, 29, 149 HPV vaccination in, 78 opioid accessibility in, 169 Nigeria morphine, 166, 168, 170, 265t acute lymphoblastic leukemia (ALL) in, 128 mortality rates. See cancer death rates aflatoxins in, 159 Muñoz, N., 257 breast cancer in, 226 Murillo, Raul H., 45 cervical cancer screening in, 75 Muslim countries costs of intervention packages in, 13 alcohol consumption levels in, 37 non-alcoholic fatty liver disease (NAFD), 147, 154 breast cancer screening in, 58 noncommunicable diseases (NCDs) Myanmar, cervical cancer screening in, 74 allocation of development assistance for, 16 cancer in comparison to other causes of N death, 24, 24f NAFD (non-alcoholic fatty liver disease), 147, 154 surveillance systems for, 250–51, 252t Namibia, childhood cancer in, 121 tobacco as major contributor to, 187 National Cancer Institute (U.S.), 215, 255 nongovernmental organizations (NGOs) Surveillance, Epidemiology, and End Results (SEER) cancer prevention activities and funding of, Program (U.S.), 251–52 139, 290 National Comprehensive Cancer Network, 111, 198 research role of, 250, 256 National Institute for Health and Care Excellence (UK), surgical capacity support from, 234b 267 Nordhagen, S., 74 National Institute of Alcohol Abuse and Alcoholism Nordic countries. See also specific countries (NIAAA), 153–54 cervical cancer screening in, 73 national strategies liver cancer in, 149, 150 Index 333 oral cancer in, 88 overview, 85 taxation to finance universal health coverage posttreatment follow-up, 94 in, 284 prognosis, 94 Northern Africa. See Middle East and North Africa radiation therapy for, 92–94, 240 Northern Ireland’s greenways urbanization project, research support for, 260 effect on chronic diseases, 38 risk factors, 85–90, 218 North-South collaborations, 16, 257 screening for, 11, 90, 91–92, 218 Norway health care personnel performing visual advertising of cigarettes, ban on, 185 screening, 91 Ghana-Norway partnership as part of Breast high-risk screening, 212 Health Global Initiative, 199 self-examination, 91–92 opioid accessibility in, 170 visual screening, 90–91, 277 Nurses’ Health Study, 254, 254b surgery, 92–94, 228, 229–30t nut chewing, as oral cancer risk, 85, 87, 88–89 complications of, 94 nutrition as risk factor for oral cancer, 89 survival rates, 88 treatment of, 92–94 O viruses as risk factor, 90 Obermeyer, Z., 74 oral submucous fibrosis (OSMF), 89, 90 obesity. See also lifestyle Organisation for Economic Co-operation and relationship to cancer, 37–38 Development’s Global Science Forum, 257 liver cancer, 147, 154 ovarian cancer occult blood test. See colorectal cancer (CRC) death rates from, 6t occupational risk factors, 38 high-risk screening, 58, 212 Olson, Zachary, 295 overdiagnosis concerns, 213, 215 Omar, Sherif, 195 oxycodone, 168, 170 opioid accessibility in, 168 opioid medications, 11, 165–69, 167f. See also pain relief P Opioid Price Watch (IAHPC), 167, 170 PAF (population attributable fraction), 26–27 Opisthorchis viverrini, 36 pain relief, 165–73 Oportunidades program (Mexico), 284 availability of broadly defined palliative care oral cancer, 85–99 services, 167 alcohol consumption associated with, 37, 89 barriers in LMICs to, 167–69 chemotherapy, 94, 275 cancer patients’ need for, 166 chronic trauma as risk factor, 90 comprehensive cancer center services, 200 cost-effectiveness assessments of interventions, cost-effectiveness of, 9, 169 94–96 costs of, 14t, 169–70 cost-effectiveness of interventions, 94–96, 95t, 268, Global Opioid Policy Initiative (GOPI), 168, 274–75t 170–71 defined, 85 as human right, 165 early diagnosis and staging of, 92, 93t International Narcotics Control Board (INCB), 168 future research needs, 96 in LMICs, 166–67 genetic factors, 89–90 opioid medications, availability of, 11, 166–67, 167f high-risk screening, 212 Opioid Price Watch (IAHPC), 167, 170 HPV infection and, 87, 88, 90, 218 overzealous regulation restricting access to, 168 incidence and death rates, 6t, 28f, 85–88 patient perspective, 170 by gender, 86t, 87–88m, 215 Pakistan by income group, 87f, 224f acute lymphoblastic leukemia (ALL) in, 128 intervention packages, 9t HCV infection in, 151 management of, 92 incidence of HBV vs. HCV in, 151, 152f natural history of, 90 oral cancer in, 87, 88, 94 nutrition as risk factor, 89 Palestinian Authority, breast cancer screening in, 58 oral potentially malignant disorders (OPMDs), Palipudi, Krishna, 175 screening for, 91 palliative care. See also pain relief 334 Index radiation therapy and, 165, 240, 241 cost reduction for essential intervention surgery with palliative intent, 230 packages, 16 pancreas cancer, 28f research support, 17 Pap smears, 69, 73, 74, 216 technical assistance, 16–17 Papua New Guinea, oral cancer in, 85 progressive universalism, 283 Partners in Health, 199 prostate cancer pathology early detection as factor, 29 comprehensive cancer center services, 199 by geographical region, 29 research role of, 252 incidence and death rates, 6t, 27, 28f surgery’s reliance on, 226 by income group, 32f telepathology, 218 transitions in low- and middle-income PBCRs (population-based cancer registries), 25, countries, 29 39, 251 overdiagnosis concerns, 213 pediatric cancer. See childhood cancer physical inactivity, relationship to, 38 Pediatric Oncology in Developing Countries (PODC) radiation therapy for, 240 committee of International Society of Pediatric research support for, 260 Oncology (SIOP), 125 screening for, 10 Peru lead-time bias and, 212 cervical cancer prevention approaches in, 80 prostate-specific antigen (PSA) testing, 10–11 universal health care in, 284, 287t, 289, 290 psychosocial support, 200 Peto, Richard, 179 public awareness of cancer control interventions, pharmacotherapy. See chemotherapy 16, 255 pharmacy services. See also drug therapy public places, restrictions on smoking in, 185, 188 comprehensive cancer centers, 201 pharynx. See oral cancer Q Philip Morris, 189 Qiao, You-Lin, 249, 295 Philippines quality-adjusted life years (QALYs), 7b colorectal cancer incidence rates in, 103 opioid accessibility in, 170 R smoking in Rabeneck, Linda, 101, 268 implementation of excise tax increases, 184 Radiation Safety Institute of Canada, 199 implications of tobacco taxation, 10, 188 radiation therapy (RT), 239–47. See also specific types inverse relationship of consumption and of cancer price, 183 access to, 243–45, 244m physical inactivity. See lifestyle as adjuvant treatment, 231, 240 “Pink Ribbon Red Ribbon” program, 219 adoption of expert systems and machine learning Poland methods, 243 opioid accessibility in, 170 brachytherapy, 241 radiation therapy in, 245 comprehensive cancer center services, 199–200, 243 smoking in concurrent chemotherapy and, 240 cessation treatments, 186 constraints on, 265 implementation of excise tax increases, 184 cost-effectiveness of, 239 inverse relationship of consumption and critical normal structures, limited tolerance of, 240 price, 183 delivery methods, 241–43 popularity paradox, 215 equipment specialized for, 239, 265 population attributable fraction (PAF), 26–27 externally applied radiation beams, 241 population-based cancer registries (PBCRs), 25, 39, 251 facilities designed for, 241–42 population-based randomized trial, 62 infrastructure requirements, 244, 265 Pramesh, C. S., 223 integration into cancer centers, 243 prevention strategies, 1, 10. See also specific types intensity modulated radiation therapy (IMRT), 240 of cancer in multidisciplinary approach to cancer primary health clinic or mobile outreach, 9t management, 239 priorities for international support, 16–17 oral cancer, 94 Index 335 overview, 239–40 industry collaboration, 257 palliative treatment and, 165, 240, 241 international collaboration, 257 personnel and radiation therapists, 242, 244 laboratories and anatomic pathology, 252 process, 242 linking death registration systems to cancer radioisotope therapy, 241 registration systems, 253 safety and quality assurance, 242–43 local level collaboration, 256 scaling up, 15, 275 ministries’ role in, 249–50 as sole therapy, 240 national level collaboration, 256 stereotactic RT, 240 overview, 249 uses of, 240–41 priorities in cancer research, 250–57 Rajaraman, Preetha, 249 research training, 256–57 Ramadas, Kunnambath, 85 stakeholders’ knowledge needs and, 249–50, 251b Ramasundarahettige, Chintanie, 175 transport of specimens among countries, 253, 256 raw fish consumption, 157 resource limitations Reach for Recovery, 58 availability across countries, 264–67 Rebbeck, Timothy, 249 comprehensive cancer centers, 202–5 rectal cancer, 110. See also colorectal cancer need for guidelines to address, 48–49 referrals, 214 surgery, 226–27, 227b surgery, 231–32 retinoblastoma, 137–38 regional cancer centers, establishment of, 203, 203–4b Rodriguez-Galindo, Carlos, 121 registries Romania, opioid accessibility in, 169, 170 childhood cancer incidence rates in, 121, 122f Russia death registration systems, linking to cancer alcohol consumption levels in, 37 registration systems, 253 breast cancer detection in, 52 global initiative for cancer registry through IARC, colon cancer incidence and mortality in, 266 40, 40m HCV treatment in, 156 importance of, 266 health insurance coverage in, 266 population-based cancer registries (PBCRs), 25, liver cancer, cost of treatment in, 159 39, 251 liver fluke infection in, 154 types of data in, 250–52 oral cancer in, 88 reproductive health policy as umbrella for breast smoking in, 180 and cervical cancer, 219 knowledge of health information, 184 Republic of Korea. See Korea, Republic of price elasticity, 181 research support, 17, 249–62 Rwanda “best buys,” 259–60 cervical screening and treatment services in, 228 biobanking, 252–53 HPV vaccination program in, 78 budget allocation recommended for, 249 opioid accessibility in, 169 Call to Action for Research on Health (2008 Rwandan Women’s Interassociation Study and Bamako Forum), 249, 250b Assessment, 73 cancer registries, data in, 250–52 case-control studies, 253 S cervical cancer as integrated research exemplar, St. Jude Children’s Research Hospital (Memphis, 257–59 Tennessee), 130, 136, 204, 205b cohort studies, 254, 254–55b Sankaranarayanan, Rengaswamy, 1, 85, 114, 217, 258, 268 collaboration, 256 Sauvaget, C., 114, 217 education to include, 256 Sayginsoy, O., 187 epidemiologic studies, 253–57, 260 scaling up health care professionals’ role in, 250 cancer screening projects, 214 health communications, 255 chemotherapy, 16, 275 health surveillance and cancer surveillance, 250–51 childhood cancer, 277 hospitals’ role in, 250 hepatitis B vaccination, 15 implementation science, 255–56, 256b HPV vaccination, 15 as “best buy” for cancer research, 259 surgical systems, 234–35 336 Index Schistosoma haematobium, 36 oral cancer in, 85, 87, 89, 212, 218 schistosomiasis, 12 tobacco use in, 175 screening. See cancer screening; specific types of cancer Southeast Asia secondhand smoke, 175 aflatoxins in, 153 SEER (Surveillance, Epidemiology, and End Results) alcohol consumption in, 157 Program, 102 breast cancer in, 60 Seguro Popular (Mexico), 1, 125, 204, 266, 283, 284, colorectal cancer in, 38, 114 287t, 291 oral cancer screening in, 87, 218 self-examination high-risk screening, 212 breast self-examination, 52, 216 South-South collaborations, 16, 257 oral cancer screening, 91–92 Spain Sharma, Monisha, 295 liver cancer in, 149 Shi, Ju-Fang, 295 opioid accessibility in, 170 Shulman, L., 214 smuggling and black market sales of cigarettes Sierra Leone, HPV vaccination in, 78 in, 187 simeprevir, 155–56, 159 specialized cancer centers/units. See also comprehensive Singapore cancer centers colorectal cancer in, 108, 111 childhood cancer dedicated centers, 123–24, head and neck cancers, imaging techniques for, 198 124t, 128 health insurance coverage in, 289 intervention packages, 9t liver fluke infection in, 154 scaling up, 16 oral cancer in, 88 Standards of Oncology Pharmacy Practice, 201 Single Convention on Narcotic Drugs of 1961, 168 Stevens, Lisa, 249 SIOP (International Society of Pediatric stomach cancer Oncology), 125 diet and consumption associated with, 37 skilled medical personnel, 266 incidence and death rates, 6t, 8f, 27, 28f, 215, 218 skin cancer, 38, 240, 260 decline in, 29 Slovak Republic, oral cancer in, 87, 88 by income group, 35f smoking. See tobacco use research support for, 260 smuggling and black market sales of cigarettes, 186–87 schistosomiasis and, 12 sociocultural beliefs. See cultural practices screening for, 218 Soerjomataram, Isabelle, 23 stroke, smoking’s relationship to, 178, 184 sofosbuvir, 155–56, 159 Subramanian, Sujha, 85 South Africa Sub-Saharan Africa breast cancer and access to treatment in, 225 alcohol consumption in, 157 cervical cancer prevention approaches in, 80, 217 antiretroviral therapy in, 73 national health insurance package in, 15 Burkitt lymphoma in, 123, 134 smoking in cancer incidence and death rates in, 166 cigarette taxes, 10, 184 cancer trends in, 29 inverse relationship of consumption and price, cervical cancer in, 23, 71, 71m 183, 183f childhood cancer in, 121 reporting on death certificates, 185 costs of intervention packages in, 13–14 smuggling and black market sales of cigarettes, 187 HBV infection in, 151 South Asia HCV infection in, 155 catastrophic health expenditures in, 282 HPV vaccination in, 10, 79 colorectal cancer in, 38, 102 liver cancer in, 147, 149, 151 HBV infection in, 151 national health insurance coverage in, 290 HCV infection in, 155 radiation therapy in, 244, 265, 266 head and neck cancers in, 198 resource constraints in, 264 high-risk screening for oral cancer in, 212 retinoblastoma in, 137 imaging techniques guidelines for head and neck surgical infrastructure needed in, 223 cancers, 198 Substance Abuse and Mental Health Services liver cancer in, 147 Administration, 255–56 Index 337 Sudan treatment with palliative intent, 230 cervical cancer in, 71 utilization of services, 225 opioid accessibility in, 170 workforce development for, 232–33 oral cancer, smokeless tobacco products as Surveillance, Epidemiology, and End Results (SEER) factor in, 89 Program, 102 pain medications in, 167 surveillance systems, 250–51, 252t Sullivan, Richard, 211, 223 as “best buy” for cancer research, 259 Sullivan, Terrence, 211 need for, 24, 26, 39–40 surgery, 223–38 survivorship programs, 200 access to, 225, 225t Sweden adjuvant treatment considerations, 230–31, 240 alcohol consumption levels in, 37 anesthesia, safe provision of, 226 cervical cancer treatment in, 16 biopsy procedures, 227–28 Switzerland, smoking in breast cancer, 12b, 54, 55–56t, 223 knowledge of health risks associated with, 185 burden of surgically treatable cancers in LMICs, 224, possible tax increase on cigarettes, 184 224f, 236 treaty with Uruguay as basis for lawsuit, 189 capacity building, 232–35 centralized vs. decentralized delivery models, T 231, 231t Taiwan cervical cancer, 78 colorectal cancer in, 108, 111 colon cancer, 110, 223 HCV infection in, 151 comprehensive cancer center services, 199 hepatitis B vaccination in, 36 costs and cost-effectiveness of, 235–36, 267 incidence of HBV vs. HCV in, 151, 152f diagnostic role of, 227 liver cancer in, 149, 159 distribution of services, 225 oral cancer in, 85, 87, 92 guidelines for surgical platforms, 227–31 tamoxifen, 12b, 54, 60, 231, 265t imaging in conjunction with, 226 Tanzania infrastructure and training, 225–26, 232–33, 264–65 cervical cancer screening in, 75 need for further study to assess effectiveness of, 235, HPV vaccination in, 78 236, 267 technical assistance, need to prioritize, 16–17 nonphysician providers, task shifting to, 232 telemedicine, 218 oral cancer, 92–94 telepathology, 218 complications of, 94 Thailand overview, 223 care availability, outcomes of increasing, 291 pain relief and, 165 cervical cancer prevention approaches in, 80–81, 292 pathology in conjunction with, 226 colorectal cancer in, 103, 108, 217 platforms for delivery, 227–31, 228–30t financing of cancer interventions from taxes in, 290 private vs. public sector provision of services, 234 liver cancer in, 149 procurement processes, 233 liver fluke infection in, 149, 154, 158 quality and safety, 225–26, 233, 233–34b oral cancer in, 88, 94 rectal cancer, 110 smoking in referral networks, service coordination, and implications of tobacco taxation, 188 partnerships, 231–32 inverse relationship of consumption resources and price, 183 availability across countries, 264–65, 265t knowledge of health information, 185 stratified approaches by level of care, 226–27, universal health care in, 284, 288t, 289, 290 227b co-payments, 291 scaling up, 15, 16, 234–35 Thomas, David B., 45 shortage of surgeons in LMICs, 232–33 Three-Step Analgesic Ladder (WHO), 166 standardization of guidelines, 225 Thun, M. J., 254 status in LMICs, 224–25 thyroid cancer surgical missions, 232 by geographical region, 29 treatment with curative intent, 228–30 incidence and death rates, 28f 338 Index tissue sampling, 54, 63, 227–28 interventions to increase, 180–87. See also tobacco taxes, 180–84 tobacco taxes ad valorem taxes, 182 patterns, 180 affordability and, 182 smuggling and black market sales of cigarettes, cost-effectiveness of, 13, 13f, 187 186–87 discouraging youth smoking, 10, 182 supply-side interventions, 186–87 financing of cancer interventions from, 15, 290 tuberculosis and, 178 implementation of large tax increases, 183–84, 183f Uruguay case study of tobacco control, 188–89 poverty considerations, 188 U.S. ban of clove-flavored cigarettes, WTO suit price elasticity of demand for tobacco, 181–82 over, 189 signaling effects of price increases, 184 WTO suits involving cigarettes, 189 smart excise tax structure, 182–83, 184f traditional medicine, 155 tobacco use, 175–93. See also oral cancer; tobacco taxes training of health care professionals, 203–4 advertising of cigarettes, 185, 189 in radiation therapy, 244 cancer and other diseases resulting from, 176–78 surgeons and surgical staff, 225–26, 232–33 chronic lung disease and, 178 transport of specimens among countries, 253, 256 colorectal cancer and, 32 trastuzumab, 12b, 290 cost-effectiveness of tobacco control, 187–88, Trimble, Edward L., 249 187–88t Trypuc, Joann, 195 current global volume of, 180 tuberculosis, 176, 178 e-cigarettes, 186 Tufts Medical Center Cost-Effectiveness Analysis epidemiology of smoking-attributable diseases, Registry of cancer-related studies, 235 175–79 tumor lysis syndrome (TLS), 135 Framework Convention on Tobacco Control Tunisia, cervical cancer in, 71 (FCTC), 10, 89, 188, 189 Turkey future burden of cancer and, 32, 34–36, 179 financing approach in, 283 gender differences in, 29–30, 176, 177f, 180 radiation therapy in, 245 Global Adult Tobacco Survey, 182, 184 smoking, inverse relationship of consumption globalization and, 189 and price in, 183 Global Youth Tobacco Survey, 182 twinning programs individuals who start smoking in adolescence or childhood cancer treatment, 124–25, 130, 140 early adult life, 176, 176t, 177f, 181t development of cancer centers and, 205 international initiatives, 189 examples of, 205–6b, 267 intervention packages, 9t innovative approaches and, 218 cost-effectiveness of, 8 pathology and laboratory medicine, 199 costs of, 14t survival rates improved by, 205 effectiveness of, 10, 182t mortality rates, 35, 175, 179 U avoidable deaths before 2050, 189 Uganda cessation of use, effect on, 178 alcohol consumption levels in, 37 in LMICs, 178–79 cancer trends in, 29 smoking’s delayed effect on, 175–76 cervical cancer in, 23, 75 plain packaging mandate, 185, 189 opioid accessibility in, 169 population attributable fraction (PAF) and, 26 Uganda Cancer Institute, 205b poverty considerations for, 188 ultrasound in breast cancer diagnosis, 54, 216 public places, restrictions on smoking in, 185, 188 Unidad Nacional de Oncología Pediátrica (UNOP, as risk factor, 23, 85, 87, 88–89 Guatemala), 140 smokeless tobacco use, 88–89 Union for International Cancer Control (UICC), 197b smoking cessation Global Task Force on Radiotherapy for Cancer counseling and cessation support Control, 245 programs, 186 Tumor, Nodes, Metastasis (TNM) staging system, 92 health information and counter-advertising, United Kingdom 184–85 asbestos exposure and lung cancer in, 38 Index 339 breast cancer, new drug treatments for, 13 colorectal cancer in, 11 British doctors, cohort study of, 254 oral cancer in, 85 British Royal College of Physicians, 185 tobacco control in, 188–89 HCV treatment in, 156 U.S. Community Preventive Services Task Force, 157 National Institute for Health and Care Excellence, 267 U.S. Multi-Society Task Force on CRC/American opioid accessibility in, 170 Cancer Society (USMSTF), 107, 108 oral cancer in, 88 U.S. Preventive Services Task Force (USPSTF), 10–11 quitting smoking’s effect on mortality in, 178 on breast self-examination, 52 salt intake, related to cancer rate in, 37 on clinical breast examination, 51 smoking in on colorectal screenings, 107–8 effect on mortality in, 175–76, 177f on mammogram screenings, 50, 51t inverse relationship of consumption and price, on oral cancer, 91, 218 183, 183f on screening guidelines, 198 quitting, effect on cancer, 178, 180 U.S. Surgeon General’s Report, 185 taxation to finance universal health coverage in, 284 V United Kingdom Million Women Study, 254 vaccines. See hepatitis B (HBV) vaccination; HPV United Nations (UN) (human papillomavirus) vaccination demographic estimates of cancer mortality, 7b Verguet, Stéphane, 295 Sustainable Development Goals for 2030, 1 Victoria Hospice (Victoria, British Columbia), 206b United Nations Population Fund, 219 Vietnam United Nations World Population Prospects, 26 breast cancer in United States clinical breast examinations, 60 acute lymphoblastic leukemia (ALL) in, 128 surgical treatment costs, 235 alcohol consumption in, 157 liver cancer in, 149 cancer registries in, 266 liver fluke infection in, 154 colorectal cancer incidence rates in, 102 opioid accessibility in, 168–69 cost-effectiveness in, 267 smoking in incidence of HBV vs. HCV in, 151, 152f affordability, 182 liver cancer in, 149, 150 implications of tobacco taxation, 188 liver cirrhosis in, 154 knowledge of health information, 184 lung cancer in, 177 Vos, T., 267 asbestos exposure and, 38 opioids for pain relief in, 169 W oral cancer, smokeless tobacco products as factor Wagner, Judith, 165 for, 89 Wang, Shao-Ming, 295 smoking in, 180 Wee, J. T., 198 advertising of cigarettes, 185 Welch, Gilbert, 215 ban of clove-flavored cigarettes, WTO suit over, West Africa, aflatoxins in, 157 181–82 WHO. See World Health Organization e-cigarettes, 186 Wild, Christopher P., 147 effect on mortality in, 176, 177f Wilms tumor, 11, 126, 133–34, 133t, 135t implementation of excise tax increases, 184, 187 Wilson-Junger screening criteria, 213, 213b inverse relationship of consumption and price, women. See also gender differences 183, 183f breast cancer. See breast cancer price elasticity, 181–82 cervical cancer. See cervical cancer quitting, effect on mortality, 178, 179f ovarian cancer. See ovarian cancer U.S. Surgeon General’s Report, 185 World Development Report (1993), 5b youth quitting, 184 World Economic Forum/Harvard School of Public surgery in, 236 Health, 218 universal health coverage. See health insurance World Endoscopy Organization, 109 urbanization, 263, 277 World Endoscopy Society, 267 Uruguay World Health Assembly 340 Index premature death, goal to reduce, 1 on imaging guidelines, 199 smoking, goal to reduce, 189 on life table estimates for China, 297 World Health Organization (WHO). See also mortality databank, 26 International Agency for Research on Cancer MPOWER report on global tobacco epidemic, 36 (IARC) on national cancer control planning, 195, 243, 267 on breast cancer guidelines, 48 on noncommunicable disease best buys for LMICs, 8 cervical cancer screening, VIA demonstration on palliative care, 165, 168 project, 75 Study of Global AGEing and Adult Health, 298 Choosing Interventions That Are Cost-Effective Surgical Safety Checklist, 199 (WHO-CHOICE) framework, 59, 60, 277 Three-Step Analgesic Ladder, 166 on cigarette excise taxes, 183–84 WHO Action Plan for the Global Strategy for the collection of cancer incidence, 39 Prevention and Control of Noncommunicable on cost-effective intervention strategies for NCDs in Diseases, 39 LMICs, 218 World Trade Organization (WTO), 189 cost effectiveness recommendations, 60, 138, 235 on effectiveness of tobacco control, HPV and HBV Y vaccination, and opportunistic cervical cancer Yerramilli, Pooja, 281 screening, 17 Yip, Cheng-Har, 223 EPIC model to simulate the macroeconomic effects youth smoking, 10, 182, 184, 186 of noncommunicable diseases, 187, 187t Yurekli, Ayda, 175, 187 Essential Medicines List, 156, 166, 170 Framework Convention on Tobacco Control Z (FCTC), 10, 89, 188, 189 Zambia, cervical cancer in, 71, 75 Global Health Estimates, 7b Zaton´ksi, Witold, 175 Global Health Observatory, 250 Zauber, Ann G., 101 Global Monitoring Framework for Zelle, S. G., 59, 60, 216 Noncommunicable Diseases, 166 Zhao, Fang-hui, 249 on HBV vaccine, 155 Zimbabwe on HCV infection, 151 cancer trends in, 29 on health expenditures covered by external sources, cervical cancer, increases in, 23 282, 283 zur Hausen, H., 257 Index 341 ECO-AUDIT Environmental Benefits Statement The World Bank Group is committed to reducing its environmental footprint. In support of this commitment, the Publishing and Knowledge Division leverages electronic publishing options and print-on-demand technology, which is located in regional hubs worldwide. Together, these initiatives enable print runs to be lowered and shipping distances decreased, resulting in reduced paper consumption, chemical use, greenhouse gas emissions, and waste. The Publishing and Knowledge Division follows the recommended standards for paper use set by the Green Press Initiative. The majority of our books are printed on Forest Stewardship Council (FSC)–certified paper, with nearly all containing 50–100 percent recycled content. The recycled fiber in our book paper is either unbleached or bleached using totally chlorine free (TCF), processed chlorine free (PCF), or enhanced elemental chlorine free (EECF) processes. More information about the Bank’s environmental philosophy can be found at http://crinfo.worldbank.org/wbcrinfo/node/4.