FEBRUARY 2016 Volume 2, Issue 1 assets/JGO_2_1.pdf · 2016-12-07 · Ken G. Kornfield Emilie Gunn...

Commentary: Ethics of Clinical Trials in Low-Resource SettingsVinay Prasad et al

Development of a Breast Cancer Treatment Program in Port-au-Prince, HaitiVincent DeGennaro et al

Tuberculosis Diagnosis Delaying Treatment of CancerLeo Peter Lockie Masamba et al

FEBRUARY 2016Volume 2, Issue 1

An official publication of

ContentsCOMMENTARIES

Ethics of Clinical Trials in Low-Resource Settings: Lessons From Recent Trials in Cancer MedicineVinay Prasad, Hemanth Kumar, and Sham Mailankody......................................................................................... 1

Breast Cancer Screening in Low- and Middle-Income Countries: A Perspective From MalawiLily A. Gutnik, Beatrice Matanje-Mwagomba, Vanessa Msosa, et al ....................................................................... 4

ORIGINAL REPORTS

Breast Cancer

Development of a Breast Cancer Treatment Program in Port-au-Prince, Haiti: Experiences From the FieldVincent DeGennaro Jr, Rachel Libby, Elizabeth Patberg, et al ............................................................................... 9

Thoracic Oncology

Impact of the Bim Deletion Polymorphism on Survival Among Patients With Completely Resected Non–Small-CellLung CarcinomaJun Atsumi, Kimihiro Shimizu, Yoichi Ohtaki, et al................................................................................................. 15

Epidemiology

Tuberculosis Diagnosis Delaying Treatment of Cancer: Experience From a New Oncology Unit in Blantyre, MalawiLeo Peter Lockie Masamba, Yankho Jere, Ewan Russell Stewart Brown, et al........................................................ 26

Lung Cancer Survival Among Chinese Americans, 2000 to 2010Scarlett Lin Gomez, Juan Yang, Shih-Wen Lin, et al............................................................................................... 30

Health Policy

Increasing Access to Oral Anticancer Medicines in Middle-Income Countries: A Case Study of Private HealthInsurance Coverage in BrazilElize Massard da Fonseca, Francisco Inacio Bastos, and Gilberto Lopes ............................................................... 39

SPECIAL ARTICLE

Collision of ThreePandemics: TheCoexistence of Cervical Cancer, HIV Infection, andPrior Tuberculosis in the Sub-Saharan Country of BotswanaNicola M. Zetola, Surbhi Grover, Chawangwa Modongo, et al ................................................................................ 47

Journal of Global Oncology (ISSN 2378-9506) is published online only six times a year, bimonthly,by the American Society of Clinical Oncology, 2318 Mill Road, Suite 800, Alexandria, VA 22314.

Editorial correspondence should be addressed to David Kerr, MD, Journal of Global Oncology.Phone: 703-797-1900; Fax: 703-684-8720. E-mail: [email protected]. Internet: www.jgo.org.

Journal of Global Oncology® is a registered trademark of American Society of Clinical Oncology, Inc.

February 2016Volume 2, Issue 1

mailto:[email protected]

http://www.jco.org/podcasts

jgo Editorial RosterEDITOR-IN-CHIEF

David Kerr, MD, DSc, University of Oxford andWeill Cornell Medical College, Oxford, UnitedKingdom

ASSOCIATE EDITORS

Otis Brawley, MD, American Cancer Society,Atlanta, GA

Mary Gospodarowicz, MD, Princess MargaretCancer Centre, Toronto, Canada

Tezer Kutluk, MD, PhD, Hacettepe UniversityFaculty of Medicine and Cancer Institute,Ankara, Turkey

Gilberto Lopes, MD, MBA, Cancer Center forHospital do Coraç~ao, S~ao Paulo, Brazil

D. Cristina Stefan, MD, PhD, South AfricanMedical Research Council, Cape Town, SouthAfrica

EDITORIAL BOARD

Isaac F. Adewole, MB, BS, University of Ibadan,Ibadan, Oyo, Nigeria

Nada A. S. Al Alwan, MD, PhD, BaghdadUniversity, Baghdad, Iraq

Benjamin O. Anderson, MD, University ofWashington, Seattle, WA

Ami S. Bhatt, MD, PhD, Stanford University,Stanford, CA

Gouri Shankar Bhattacharyya, MD, PhD, DNB,MRCP, Fortis Hospital, Kolkata, India

Franco Cavalli, MD, Oncology Institute ofSouthern Switzerland, Ticino, Switzerland

Eduardo L. Cazap, MD, PhD, Latin American andCaribbean Society of Medical Oncology, BuenosAires, Argentina

James Cleary, MB, BS, FAChPM, University ofWisconsin, Madison, WI

C. Norman Coleman, MD, Chevy Chase, MD

Anil D’Cruz, Tata Memorial Centre, Mumbai,India

Paul de Souza BSc(Med), MBBS, MPH, PhD,FRACP, University of Western Sydney, SydneyNSW, Australia

Nagi S. El Saghir, MD, American University ofBeirut, Beirut, Lebanon

Ahmed Elzawawy, MD, International Campaignfor Establishment and Development of OncologyCenters and South and East MediterraneanCollege of Oncology, Port Said, Egypt

Kenneth Fleming, MB ChB, MA(Oxon), DPhil,FRCPath, FRCP, Oxford, United Kingdom

Ophira Ginsburg, MSc, MD, Public HealthUniversity of Toronto, Toronto, Canada

William J. Gradishar, MD, NorthwesternUniversity, Chicago, IL

Serigne Gueye, MD, FWACS, Cheikh Anta DiopUniversity, Dakar, Senegal

Francisco Gutierrez-Delgado, MD, PhD, FACP,Centro de Estudios y Prevencion del Cancer,Oaxaca, Mexico

Abdul Rahman Jazieh, MD, MPH, King AbdulazizMedical City, Riyadh, Saudi Arabia

Felicia Knaul, PhD, Harvard Global EquityInitiative, Boston, MA

Patrick J. Loehrer, MD, Indiana University,Indianapolis, IN

Gary H. Lyman, MD, MPH, Fred HutchinsonCancer Research Center, Seattle, WA

Danny Milner, MD, Brigham and Women’sHospital, Boston, MA

Twalib Ngoma,MD,OceanRoadCancer Institute,Dar es Salaam, Tanzania

Nir Peled, MD, PhD, FCCP, Davidoff CancerCenter, Petah Tikva, Israel

Surendra S. Shastri, MD, Tata Memorial Centre,Mumbai, India

Yuankai Shi, MD, PhD, Chinese Academy ofMedical Sciences, Beijing, China

Lawrence N. Shulman, MD, University ofPennsylvania, Philadelphia, PA

George W. Sledge Jr., MD, Stanford University,Stanford, CA

Richard Sullivan, MD, PhD, Kings HealthPartners Institute of Cancer Policy & GlobalHealth, London, United Kingdom

Edward L. Trimble, MD, MPH, Bethesda, MD

Daniel A.Vorobiof,MD,SandtonOncologyCentre,Johannesburg, South Africa

Christopher P. Wild, PhD, International Agencyfor Research on Cancer, Lyon, France

William C. Wood, MD, Emory University Schoolof Medicine, Atlanta, GA

AnnieYoung,University ofWarwick, Coventry,UKPublisher:David SampsonManaging Editors:Ken G. KornfieldEmilie Gunn

Volume 2, Issue 1, February 2016 jgo.ascopubs.org JGO – Journal of Global Oncology

http://jgo.ascopubs.org

jgo Journal of Global Oncology

JOURNAL OF GLOBAL ONCOLOGY

(ISSN 2378-9506) is published online only sixtimesa year, bimonthly, by theAmericanSociety ofClinical Oncology, 2318 Mill Road, Suite 800,Alexandria, VA 22314.

EDITORIAL CORRESPONDENCE(manuscript-related inquiries):

David Kerr, MD, Editor-in-ChiefJournal of Global OncologyPhone: 703-797-1900; Fax: 703-684-8720E-mail: [email protected]; Internet: jgo.org

AMERICAN SOCIETY OF CLINICAL ONCOLOGY(membership-related inquiries):

ASCO Member ServicesPhone: 703-299-0158Toll-free: 888-282-2552Fax: 703-299-0255E-mail: [email protected]: www.asco.orgHours: Monday-Friday, 8:30 a.m.-5:00 p.m.Eastern Time

CUSTOMER SERVICE

JGO Customer ServicePhone: 703-519-1430; Toll-free: 888-273-3508; Fax: 703-518-8155E-mail: [email protected]

ADVERTISING SALES

The Walchli Tauber Group, Inc.225 Old Emmorton Road, Suite 201Bel Air, MD 21015Phone: 443-512-8899; Fax: 443-512-8909www.wt-group.com

BUSINESS-TO-BUSINESS SALES

Rick WerdannSpringer Healthcare, LLC233 Spring StreetNew York, NY 10013Phone: 212-460-1523Mobile: 646-209-1840E-mail: [email protected]: www.SpringerHealthcare.com

FREE PUBLIC ACCESS

Journal of Global Oncology (JGO) provides freeonline access at jgo.org.

PERMISSIONS REQUESTS

Licensing, Rights, and Permissions DivisionAmerican Society of Clinical OncologyPhone: 571-483-1722Fax: 703-518-5094E-mail: [email protected]

Licensing: All articles are published under a CC BY or CC BY-NC-ND Creative Commons license, found here:https://creativecommons.org/licenses/.Copyright: Copyright © 2015 by American Society of Clinical Oncology unless otherwise indicated. All rights reserved. Nopart of this publication may be reproduced or transmitted in any form or by any means now or hereafter known, electronicor mechanical, including photocopy, recording, or any information storage and retrieval system, without permission inwriting from the Publisher. Printed in the United States of America. Journal of Global Oncology® is a registered trademarkof American Society of Clinical Oncology, Inc.

Disclaimer:The ideas and opinionsexpressed in JGO donot necessarily reflectthose of the AmericanSociety of ClinicalOncology (ASCO). Themention of any product,service, or therapy inthis publication or inany advertisement inthis publication shouldnot be construed asan endorsement of theproducts mentioned. Itis the responsibility ofthe treating physicianor other health careprovider, relying onindependent experienceand knowledge of thepatient, to determinedrug dosages and thebest treatment for thepatient. Readers areadvised to check theappropriate medicalliterature and theproduct informationcurrently provided by themanufacturer of eachdrug to be administeredto verify approveduses, the dosage,method, and durationof administration, orcontraindications.Readers are alsoencouraged to contactthe manufacturer withquestions about thefeatures or limitationsof any products. ASCOassumes no responsibilityfor any injury or damageto persons or propertyarising out of or relatedto any use of thematerial contained inthis publication or to anyerrors or omissions.



http://jgo.org


http://www.asco.org


http://www.wt-group.com


http://www.SpringerHealthcare.com

http://jgo.org


https://creativecommons.org/licenses/


jgo Current Abstracts Volume 2, Issue 1 February 2016

Development of a Breast Cancer Treatment Program inPort-au-Prince, Haiti: Experiences From the Field

abstract

Purpose The nonprofit Project Medishare launched a breast cancer treatment program in Port-au-Prince in July 2013 to address the demand for breast cancer care in Haiti. We outline the devel-opment of the program, highlight specific challenges, and discuss key considerations for othersworking in global oncology.

MethodsWe reflected on our experiences in the key areas of developing partnerships, building laboratorycapacity, conducting medical training, using treatment algorithms, and ensuring access to safe, low-costchemotherapy drugs. We also critically reviewed our costs and quality measures.

Results The program has treated a total of 139 patients with breast cancer with strong adherence totreatment regimens in 85% of patients. In 273 chemotherapy administrations, no serious exposure oradverse safety events were reported by staff. The mortality rate for 94 patients for whom we havecomplete data was 24% with a median survival time of 53 months. Our outcome data were likelyinfluenced by stage at presentation, with more than half of patients presenting more than 12 months afterfirst noticing a tumor. Future efforts will therefore focus on continuing to improve the level of care, whileworking with local partners to spread awareness, increase screening, and get more women into careearlier in the course of their disease.

Conclusion Our experiences may inform others working to implement protocol-based cancer treatmentprograms in resource-poor settings and can provide valuable lessons learned for future global oncologyefforts.

J Glob Oncol 2. © 2015 by American Society of Clinical Oncology Licensed under the Creative Commons Attribution 4.0License

continued

Vincent DeGennaro Jr

Rachel Libby

Elizabeth Patberg

et al

pp 9-14




Impact of the Bim Deletion Polymorphism on Survival Among PatientsWith Completely Resected Non–Small-Cell Lung Carcinoma

abstract

Purpose A deletion polymorphism of the Bim gene has been reported to be a prognostic factor for patientswith non–small-cell lung cancer (NSCLC) treated with epidermal growth factor receptor-tyrosine kinaseinhibitors in the Asian population. We investigated the impact of the Bim deletion polymorphism onsurvival among patients with completely resected NSCLC.

Patients and Methods The Bim polymorphism was detected by polymerase chain reaction analysis. Wemeasured overall survival (OS) and recurrence-free survival rates in 411 patients and postrecur-rence survival (PRS) in 94 patients who experienced recurrence and received additional anticancertherapy.

Results The Bim deletion polymorphism was detected in 61 patients (14.8%). OS rates were significantlylower for patients with the Bim deletion polymorphism than for those with the wild-type sequence. Onmultivariable analysis, the Bim deletion polymorphism was identified as an independent prognostic factorfor OS (hazard ratio, 1.98; 95% CI, 1.17 to 3.36; P 5 .011). Among the 94 patients who experiencedrecurrence and were treated with anticancer therapy, patients with the Bim deletion polymorphismshowed significantly poorer PRS than those with the wild-type sequence (median, 9.8 months v26.9 months, respectively; P < .001). Multivariable analysis revealed that the Bim deletion polymorphismwas an independent predictor of PRS (hazard ratio, 3.36; 95% CI, 1.75 to 6.47; P < .001). This trendremained apparent in subgroup analyses stratified by EGFR status, histology, and therapeutic modality.

Conclusion The Bim deletion polymorphism is a novel indicator of shortened PRS among patients withrecurrent NSCLC treated with anticancer therapy in the Asian population.

J Glob Oncol 2. © 2015 by American Society of Clinical Oncology Licensed under the Creative Commons Attribution 4.0 License

continued

Jun Atsumi

Kimihiro Shimizu

Yoichi Ohtaki

et al

pp 15-25




Tuberculosis Diagnosis Delaying Treatment of Cancer: ExperienceFrom a New Oncology Unit in Blantyre, Malawi

abstract

PurposeMalawi is a low-income country in sub-Saharan Africa with limited health care infrastructure andhigh prevalance of HIV and tuberculosis. This study aims to determine the characteristics of patientspresenting to Queen Elizabeth Central Hospital Oncology Unit, Blantyre, Malawi, who had been treated fortuberculosis before they were diagnosed with cancer.

Methods Clinical data on all patients presenting to the oncology unit at Queen Elizabeth Central Hospitalfrom 2010 to 2014 after a prior diagnosis of tuberculosis were prospectively recorded, and a descriptiveanalysis was undertaken.

Results Thirty-four patients who had been treated for tuberculosis before being diagnosed with cancer wereidentified between 2010 and 2014, which represents approximately 1% of new referrals to the oncology unit.Forty-one percent of patients were HIV positive. Mean duration of tuberculosis treatment before presentation tothe oncology unit was 3.6 months. The most common clinical presentation was a neck mass or generalizedlymphadenopathy. Lymphoma was the most common malignancy that was subsequently diagnosed in 23patients.

ConclusionMisdiagnosis of cancer as tuberculosis is a significant clinical problem in Malawi. This studyunderlines the importance of closely monitoring the response to tuberculosis treatment, being aware ofthe possibility of a cancer diagnosis, and seeking a biopsy early if cancer is suspected.


continued

Leo Peter LockieMasamba

Yankho Jere

Ewan Russell StewartBrown

et al

pp 26-29




Lung Cancer Survival Among Chinese Americans, 2000 to 2010

abstract

Purpose Despite being the leading cause of cancer death, no prior studies have characterized survivalpatterns among Chinese Americans diagnosed with lung cancer. This study was conducted to identifyfactors associated with survival after lung cancer in a contemporary cohort of Chinese patients with lungcancer.

Methods The study design is a prospective descriptive analysis of population-based California CancerRegistry data. Multivariable Cox proportional hazards models were used to estimate hazard ratios (HRs)for overall mortality. Participants were Chinese American residents diagnosed with first primary invasivelung cancer from 2000 to 2010 (2,216 men and 1,616 women).

Results Among Chinese men, decreased mortality was associated with care at a National Cancer Institutecancer center (HR, 0.85; 95% CI, 0.73 to 0.99) and adenocarcinoma versus small-cell carcinoma (HR,0.78; 95% CI, 0.65 to 0.92). Women had better survival compared with men (HR, 0.82; 95% CI, 0.75 to0.89), with mortality associated with never married versus currently married status (HR, 1.36; 95% CI,1.11 to 1.66), lower versus higher neighborhood socioeconomic status (HR, 1.38; 95% CI, 1.10 to 1.72comparing lowest to highest quintile), care at a cancer center (HR, 0.80; 95% CI, 0.67 to 0.96), andsquamous cell relative to small-cell carcinoma (HR, 1.60; 95% CI, 1.04 to 2.48).

Conclusion Focusing on factors associated with marital status, community socioeconomic status, andcharacteristics unique to National Cancer Institute–designated cancer centers may help to identifypotential strategies for improving the length of survival for Chinese Americans.


continued

Scarlett Lin Gomez

Juan Yang

Shih-Wen Lin

et al

pp 30-38




Increasing Access to Oral Anticancer Medicines in Middle-IncomeCountries: A Case Study of Private Health Insurance Coverage in Brazil

abstract

TheWorld Health Organization estimates that approximately 60% of the world’s new annual cancer casesoccur in Asia, Africa, and Central and South America, and that 70% of cancer deaths occur in theseregions. Although oral chemotherapy is a promising intervention for cancer treatment, given its high cost,it is usually unavailable in middle-income countries. In 2013, after strong lobbying from civil society,Brazil’s Congress passed legislation mandating that all private health insurance companies provideaccess to oral antineoplastic treatment. The decision to scale up the provision of oral chemotherapywas a watershed event in the regulation of private health insurance in Brazil. Until then, private insurers,which cover 25% of the population, were exempted from the provision of pharmaceutical drugs for homecare treatments. This article explores the political process involved in regulating the provision of oralchemotherapy medicines by private health insurers. Elements of this successful advocacy case includedinvestment in strategic communication, specialized knowledge of regulatory policy, and the ability to actvia democratic channels of political representation. In turn, the receptiveness of government branchessuch as the Congress and regulating bodies, as well as the Cancer Awareness Month campaign, opened awindow of opportunity. However, prospects for expanded access to such medicines in the public healthsystem are bleak in the short term because of the ongoing political and economic crisis.


continued

Elize Massard da Fonseca

Francisco Inacio Bastos

Gilberto Lopes

pp 39-46




Collision of Three Pandemics: The Coexistence of Cervical Cancer, HIVInfection, and Prior Tuberculosis in the Sub-Saharan Country of Botswana

executivesum

mary

Cervical cancer is the leading cause of cancer-related mortality in the developing world, where HIV andMycobacterium tuberculosis (TB) infection are also endemic. HIV infection is independently associatedwith increased morbidity and mortality among women with cervical cancer. TB is believed to increase therisk of malignancies and could cause chronic inflammation in the gynecologic tract. However, therelationship between cervical cancer and TB in settings hyperendemic for HIV is unknown. We found that18 (10%) of a cohort of 180 women with cervical cancer in Botswana had a history of TB disease. Ageand HIV infection were also associated with a history of TB disease. Our data show that prior TB disease ishighly prevalent among patients with cervical cancer infected with HIV. The coexistence of cervicalcancer, HIV infection, and prior TB infection might be higher than expected in the general population.Prospective studies are needed to better determine the impact of the collision of these three world healthepidemics.

Nicola M. Zetola

Surbhi Grover

Chawangwa Modongo

et al

pp 47-50



commentaries

Ethics of Clinical Trials in Low-ResourceSettings: Lessons From Recent Trials inCancer Medicine

Randomized clinical trials in cancer medicine areincreasingly conducted in low- andmiddle-incomecountries (LMICs). Theproperway to conduct suchstudies is complex and the subject of much de-bate.1 Two key questions outline the debate. First,what is the appropriate control arm for such stud-ies? For example, when can trials use placebo orsome control less than global best practice? Sec-ond, what obligations do trial sponsors have towardhost nations and communities?

These key questions were first debated morethan a decade ago in the context of clinical trialsin Africa, which tested shorter, cheaper coursesof zidovudine against placebo to prevent peri-natal transmission of AIDS. Critics argued thatthese placebo-controlled trials were unethicaland that control participants should have accessto the longer, full course of zidovudine usedelsewhere in the world. Proponents counteredthat only placebo-controlled trials could answerthe relevant question: whether an affordablecourse of therapy is better than the actualpractice of doing nothing, and not the ideal,but unrealistic question: whether a shortercourse is noninferior to a longer course. In re-cent months, these questions have regainedrelevance as recent trials recapitulate thesetensions.

Citing recent examples in cancer medicine, weargue that placebo-controlled trialsmay be ethicalbut only if the intervention being tested has areasonable chance of being implemented in thehost community.

WHAT IS THE APPROPRIATE CONTROL ARM FORTRIALS IN LOW-RESOURCE SETTINGS?

Advanced cervical cancer, now infrequentlyencountered in developed nations, carries asubstantial global burden.2 Unfortunately, uni-versal screening programs that use cytology-

based techniques that have been credited withtransforming outcomes have been deemed im-practical or unaffordable in low-resource set-tings.3 For this reason, several randomizedcontrolled trials have tested whether a low-cost cervical cancer screening program canimprove outcomes. In a recent large, random-ized trial conducted in India, Shastri et al3

tested whether four successive biennial visualinspections with acetic acid (VIAs) performedby public health workers could reduce cervicalcancer mortality. VIAs involve a process inwhich a trained health care worker uses a spec-ulum and applies dilute acetic acid (vinegar) tothe cervix and is able to directly visualize ab-normal preneoplastic tissue that turns white.The results of the Shastri study were positiveand showed 31% relative risk reduction incause-specific death. As a result, this interven-tion has been hailed as a realistic, affordablecervical cancer screening program with thepotential of saving 22,000 lives per year. Butbecause the study used a control arm thatreceived the local standard of care (no screen-ing), it has been criticized as unethical.4,5

However, India had already deemed a screen-ing program with cytology impractical and, inpractice, it is seldom performed. We believe thestudy is ethical andmeets the standard for suchtrials in low-resource settings. We support thisbecause in India, there is no universal cervicalcancer screening, and the burden of disease ishigh. Furthermore, the study by Shastri et alwas reviewed by the US National CancerInstitute before funding, was approved bya local institutional review board, and wasreviewed annually by the institutional reviewboard and a data safety and monitoringcommittee.

Discouraging studies like that of Shastri et al3

would limit the ability of individual nations to set

Vinay PrasadNational Cancer Institute,Bethesda,MD,andOregonHealth and SciencesUniversity, Portland, OR

Hemanth KumarGoa Medical College, Goa,India

Sham MailankodyNational Cancer Institute,Bethesda, MD, andMemorial Sloan KetteringCancer Center, New York,NY

Corresponding author:ShamMailankody, MBBS,Myeloma Service,Department of Medicine,Memorial Sloan KetteringCancer Center, 1275 YorkAvenue, NY 10065;e-mail: [email protected].

1 Volume 2, Issue 1, February 2016 jgo.ascopubs.org JGO – Journal of Global Oncology

© 2015 by American Society of Clinical Oncology Licensed under the Creative Commons Attribution 4.0 License




their own research agenda and conduct trials theydeem important. Furthermore, diffuse criticism ofscreening studies in LMICs diverts attention fromtruly unethical trials, which explore questions thatwill not benefit the country in which the trials areconducted. The study by Shastri et al provided thehighest level of evidence that VIAs save lives andcan be reasonably expected to benefit the hostpopulation in India and other LMICs with highburdens of cervical cancer and no cytology-based universal screening programs.

IS THERE A REASONABLE EXPECTATION OFBENEFIT FOR THE HOST POPULATION FROMCLINICAL TRIALS?

Some have argued that sponsors must not onlyprovide benefit to the individual patients on trialsbut must also provide health-related resources tothe host community and help with infrastructuredevelopment. Others submit that sponsors andinvestigators should ensure that participation insuch trials is voluntary, that individual participantsare provided fair benefits for their participation,and that this is the sole requirement for ethicalresearch.1

We take a middle position. Sponsors and investi-gators should not be expected to provide health-related resources and infrastructure to the hostcommunity in which they conduct clinical trials.Conducting clinical trials is expensive, and intoday’s economic environment, an expectationto provide additional, post-trial services could beprohibitive and discourage sponsors from sup-porting rational trials. We do believe, however, thatwhen trials in low-resource settings reach positiveconclusions regarding an intervention’s benefit,that intervention should be reasonably likely to beimplemented by the community in which the trialsare conducted.

This tension is best illustrated in two recent ex-amples. The first is that of afatinib (BoehringerIngelheim, Ingelheim, Germany), a next-in-classepidermal growth factor receptor (EGFR) inhibitorin patients with lung cancer harboring EGFRmu-tations. The IPASS (First Line IRESSA VersusCarboplatin/Paclitaxel in Asia) trial published in2009 found that patients with non–small-cell lungcancer (NSCLC) andactivatingmutations inEGFRhadamarkedprogression-free survival benefit (9.5v 6.3 months) and improvement in the quality oftheir life when treated with gefitinib (AstraZeneca,London, United Kingdom), an anti-EGFR–targeteddrug, compared with standard chemotherapy.6

These results led the European Medicines Agencyto approve gefitinib for this indication on June 24,

2009, and in turn led to an international bestpractice of using targeted agents as the initialtherapy in patients whose tumors harbor EGFRmutations. Yet from August 2009 to November2011, LUX-Lung 3 (BIBW 2992 [Afatinib] VersusChemotherapy as First Line Treatment in NSCLCWith EGFRMutation) and then LUX-Lung 6 (LUX-Lung6:ARandomized,Open-label,Phase III Studyof BIBW 2992 Versus Chemotherapy as First-lineTreatment for Patients With Stage IIIB or IV Ade-nocarcinoma of the Lung Harbouring an EGFRActivating Mutation) randomly assignedmore than1,200 and 910 patients, respectively, with NSCLCharboring EGFR mutations to either afatinib orcytotoxic chemotherapy.7,8 Both studies were per-formed largely in LMICs and, in effect, addressedthe question previously answered by the 2009IPASS study: Is there value in using an EGFR-targeted agent over conventional chemotherapy?From our perspective, the ethics of these trialshinged on whether afatinib was somehow moreaffordable or feasible for use in the nations wherethe trial was conducted. But unfortunately at a costof $79,000 per year of treatment, afatinib could notbe seen as affordable in nations that could notafford the $25,000 cost of gefinitib (all prices arefrom the most recent edition of the Redbook). In-stead, the results of LUX-Lung 3 and LUX-Lung 6were used to petition the US Food and Drug Ad-ministration for an approval that was granted in2013.

Another trial had a similar pattern. Traztuzumab(Genentech, San Francisco, CA) added to che-motherapy has been the standard of care in theinitial treatment of human epidermal growth factor2 (HER2)–positive metastatic breast cancer since2001. A recent study sought to determinewhetherlapatinib (GlaxoSmithKline, London, United Kingdom),a different HER2-targeted drug, also had efficacyin this indication.9 From 2006 to 2009, more than400 Chinese women were randomly assigned tolapatinib or placebo added to chemotherapy.However, because themajority of patients inChinacannot afford traztuzumab ($66,000 for 1 year oftreatment), it is not clear that lapatinib ($67,000 for1 year of treatment) represents a realistic alternative.

It is worth noting that these trials would not havesucceeded in enrolling patients in the UnitedStates, because most oncologists would not haveallowed their patients to have a 50% chance ofbeing randomly assigned to a treatment deemedinferior by previous studies.

Underwhat circumstances placebo-controlled re-search is ethical in low-resource settings is still a



contestedmatter. Fewwouldcriticizeourposition thatsuch research is clearly unethical when the questionaddressed cannot realistically benefit the nation inwhich the trial is performed.Wepropose the followingdecision aid to help determinewhich clinical trials arewarranted. First, in accordance with the seventhDeclaration of Helsinki (2013), “Medical researchinvolving human subjects may only be conducted ifthe importance of the objective outweighs the risksand burdens to the research subjects.”10 Thus, forany proposed trial, investigators should postulate thefollowing: If the results were positive, what changescould realistically be expected in the host nation, aswell as the world? and the same question if the trialwere negative. Ultimately, the decision will requireweighing thesepotentialbenefits to thesocietyagainstthe risks to participants.

Placebo-controlled trials of anticancer drugswith prices similar to those of proven alterna-tives benefit only the sponsoring companies.Instead, trials seeking to identify interventionsthat can realistically be implemented bydeveloping nations are justified. They are des-perately needed to inform pressing policydecisions facing the leaders of these nations.Condemning these studies is a perverse formof first-world paternalism. The need for clarityregarding the ethical conduct of trials in thedeveloping world is great, because recentcriticism has been misplaced, a disservice tocitizens around the world.

DOI: 10.1200/JGO.2015.000885Published online ahead of print at jgo.ascopubs.org onDecember 23, 2015

AUTHOR CONTRIBUTIONS

Manuscript writing: All authorsFinal approval of manuscript: All authors

AUTHORS’ DISCLOSURES OF POTENTIAL CONFLICTSOF INTEREST

The following represents disclosure information provided byauthors of this manuscript. All relationships are consideredcompensated. Relationships are self-held unless noted.I 5 Immediate Family Member, Inst 5 My Institution.Relationships may not relate to the subject matter of thismanuscript. For more information about ASCO’s conflictof interest policy, please refer to www.asco.org/rwc orjco.ascopubs.org/site/ifc.

Vinay PrasadNo relationship to disclose

Hemanth KumarNo relationship to disclose

Sham MailankodyNo relationship to disclose

ACKNOWLEDGMENTWe thank Tito Fojo, MD, PhD, Genitourinary MalignanciesBranch, National Cancer Institute, for his thoughtful com-ments on this article. He received no compensation.The views and opinions of the authors do not reflect those ofthe National Institutes of Health or National Cancer Institute.

REFERENCES1. Joffe S, Miller FG: Ethics of cancer clinical trials in low-resource settings. J Clin Oncol 32:3192-3196, 2014

2. Forouzanfar MH, Foreman KJ, Delossantos AM, et al: Breast and cervical cancer in 187 countries between 1980 and2010: A systematic analysis. Lancet 378:1461-1484, 2011

3. Shastri SS,Mittra I, Mishra GA, et al: Effect of VIA screening by primary health workers: Randomized controlled study inMumbai, India. J Natl Cancer Inst 106:dju009, 2014

4. Ortega B: Ethical questions linger in cervical-cancer study. USA Today, August 31, 2013. http://www.usatoday.com/story/news/nation/2013/08/31/ethical-questions-linger-in-cervical-cancer-study/2751705

5. Srinivasan S: Ethics of ‘standard care’ in randomised controlled trials of screening for cervical cancer. Indian J MedEthics 10:147-149, 2013

6. Mok TS, Wu YL, Thongprasert S, et al: Gefitinib or carboplatin-paclitaxel in pulmonary adenocarcinoma. N Engl J Med361:947-957, 2009

7. Sequist LV, Yang JC, Yamamoto N, et al: Phase III study of afatinib or cisplatin plus pemetrexed in patients with metastaticlung adenocarcinoma with EGFR mutations. J Clin Oncol 31:3327-3334, 2013

8. Wu YL, Zhou C, Hu CP, et al: Afatinib versus cisplatin plus gemcitabine for first-line treatment of Asian patients withadvanced non-small-cell lung cancer harbouring EGFRmutations (LUX-Lung 6): An open-label, randomised phase 3trial. Lancet Oncol 15:213-222, 2014

9. Guan Z, Xu B, DeSilvio ML, et al: Randomized trial of lapatinib versus placebo added to paclitaxel in the treatment ofhuman epidermal growth factor receptor 2-overexpressing metastatic breast cancer. J Clin Oncol 31:1947-1953,2013

10. World Medical Association: World Medical Association Declaration of Helsinki: Ethical principles for medical researchinvolving human subjects. JAMA 310:2191-2194, 2013


http://dx.doi.org/10.1200/JGO.2015.000885


http://www.asco.org/rwc

http://jco.ascopubs.org/site/ifc

http://www.usatoday.com/story/news/nation/2013/08/31/ethical-questions-linger-in-cervical-cancer-study/2751705

http://www.usatoday.com/story/news/nation/2013/08/31/ethical-questions-linger-in-cervical-cancer-study/2751705


commentaries

Breast Cancer Screening in Low- andMiddle-Income Countries: A PerspectiveFrom Malawi

Breast cancer burden is high in low-resourcecountries. From 1980 to 2010, new breast cancercases increased by more than 50% worldwide.1

Disease burden increased even more rapidly inlow- andmiddle-incomecountries (LMICs), wheremore than half of breast cancer cases now occur.Moreover, breast cancer disproportionately affectsyoung women in LMICs, such that 23% of newbreast cancer cases occur among women age 15to 49 years in LMICs versus 10% in high-incomecountries.1

Breast cancer mortality is also higher in LMICscompared with high-income countries, and rea-sons for this are multifactorial. One contributingfactor is a lack of breast cancer awareness andearly detection in LMICs. For example, more than90% of women with newly diagnosed breast can-cer in theUnited States have locoregional disease,whereas more than half of women with newlydiagnosed breast cancers in LMICs have stageIII or IV disease.2,3 According to an analysis fromthe 2003 World Health Survey, only 2.2% ofwomen age 40 to 69 years in LMICs had receivedany breast cancer screening.4 In addition to in-sufficient early detection, other factors contribut-ing to delayed diagnosis include poverty, culturaland religious beliefs, misconceptions about thedisease, and fear of mastectomy.5 Women’s au-tonomy in health care decision making may alsobe limited in some cultures.5

The WHO, along with many national cancercontrol programs, recommends population-based screening mammography for detectionof early-stage breast cancer in high-incomecountries, even though there continues to behonest and sometimes heated debate regardingthis recommendation.6–8 It is worthwhile to con-sider the possible benefits versus harms ofbreast cancer screening in LMICs, which havereceived far less attention. In this commentary,

we discuss breast cancer screening and earlydetection in LMICs with a particular focus onMalawi. We highlight areas of uncertainty andsuggest pragmatic strategies formoving forwardin light of current evidence gaps.

Health care systems in LMICs may face strongincentives and pressure to adopt health care in-terventions suchas screeningmammography thatarewell established in high-resource settings, withimplicit assumptions that benefits demonstratedinmore developed countries will generalize to lessdeveloped countries. Such assumptions are in-herently problematic and unrealistic in settings ofsevere resource scarcity. For example, there arecompelling reasons to believe that breast cancerscreening would perform differently in LMICs thanin high-income countries. Factors that could re-duce efficacy of breast cancer screening in LMICsinclude a younger population with lower breastcancer incidence, shorter life expectancy, moreprevalent competing causes of death, and higherprevalence of biologically aggressive subtypes forwhich patient outcomes are less likely to be af-fected by screening. Conversely, breast cancerscreening could have greater impact in LMICs if itincreases breast cancer awareness and earlydetection of symptomatic disease. For example,there may be more diffuse effects than would beexpected in resource-rich settings where stronghealth care systems and higher levels of aware-ness narrow the scope of breast cancer screeningprincipally to detection of asymptomatic disease.Indeed, for weak health care systems, it is plau-sible that effects beyond breast cancer may berealized andmay extend to cancer more generallyor to women’s health. Investments inHIV programshave similarly had far-reaching effects beyondproviding antiretroviral therapy, and antiretroviraltherapy clinics are now established vehicles foreffective delivery of many other essential health

Lily A. Gutnik

Beatrice Matanje-Mwagomba

Vanessa Msosa

Suzgo Mzumara

Blandina Khondowe

Agnes Moses

Racquel E. Kohler

Lisa A. Carey

Clara N. Lee

Satish Gopal

Corresponding author:Satish Gopal, MD,MPH, UNC–ProjectMalawi, Private Bag A-104, Lilongwe, Malawi;e-mail: [email protected].






services. InMalawi, commonly piggybackedhealthservices in HIV clinics now include cervical cancerscreening, Kaposi sarcoma treatment, nutritionalsupplementation, and reproductive health andmother-childwellness initiatives, all of which seek tomaximize impacts from initial investments for HIV.

Despite recent controversies about screeningmammography in high-income countries and ascarcity of high-quality data for this approach inLMICs, it is often assumed that wherever mam-mography is available, itmust benefit women. Thismay be the case, evenwhen screening is availableonly in the private sector without clearly definedeligibility guidelines, quality control measures, orfollow-up procedures.9 Examples of this exist inMalawi, where amajor intersection in Lilongwe (thecapital) features a billboard advertising screeningmammography in a private clinic promoted by afamous young Malawian breast-cancer survivor.However, the cost of a screening mammogram inLilongwe is approximately US$90 in a country withan annual gross domestic product per capita ofUS$253.10 Moreover, screening is often directlymarketed to and used for womenwho can pay forit, without clear eligibility criteria accounting forage, comorbidities, or projected life expectancy.In Lilongwe, mammography sponsors have distrib-uted coupons for discounted screening mammog-raphy at public breast cancer awareness events tounselected audiences of women primarily in their20s and 30s. Benefits of screeningmammographyhave not been clearly demonstrated for average-risk women in these age groups anywhere in theworld,nor is it recommended for them inconsensusguidelines.

In addition, LMICs often lack the necessary in-frastructure to ensure high-qualitymammographyand subsequent follow-up care.11 Operating amammography unit continuously requires a con-sistent supply of electricity and x-ray films, as wellas engineers, technicians, and radiologists, all ofwhich may be lacking in many LMICs. Four mam-mography units were donated to Malawi in 2012,one to each tertiary referral hospital, with the intent

to provide the first publicly available mammogra-phy services in the country, but these units haveyet to become operational.12 Mammographyscreening programs have also been estimated tocost US$16,000 to US$37,000 per life saved,which exceeds per capita health care budgets inmany LMICs by a significant margin.9,13

International guidelines recommend clinicalbreast examination (CBE) as a preferred ap-proach to screening in settings in which mam-mography screening is not available.5,14 Even inhigh-resource settings, there is some evidencethat annual CBE might be as effective as screen-ing mammography in lowering breast cancermortality.15-18 Relative advantages formammogra-phy versus CBEwith respect to implementation aredetailed in Table 1.

In LMICs, twoclinical trials in Egypt found that CBEconducted by physicians was effective and cost-effective in rural andurbanareas.19,20 InMalaysia,training rural nurses to perform CBE resulted insignificant breast cancer downstaging (77% v 37%late-stage diagnoses).21 In an ongoing cluster-randomized trial in India, CBE performed by fe-male community health care workers detectedmore early-stage (I to IIA) cancers (18.8 v 8.1per100,000women) in interventionversuscontrolvillages; no differences were observed for stageIIB and higher-stage cancers.22 A cross-sectionalstudy in Nepal comparing CBE performed byfemale community health care workers with ex-aminations by surgeons reported interobserveragreement of 64% for lump detection, with 70%sensitivity and 95% specificity.23 Moreover, mod-eling studies have suggested that CBE is cost-effective in low-resource settings.24,25

In addition to health care workers, lay volunteerscanalsobe trained toperformCBE.A study in ruralSudan screened approximately 10,000 womenage 18 years or older by using this approach.Seventeen of those screened had carcinoma in situor breast cancer, including eight with carcinomain situ and four with early-stage breast cancer. Incontrol villages, only four women self-referred for

Table 1. – Relative Advantages of Mammography Versus Clinical Breast Examination as Screening Approaches inLow- and Middle-Income Countries

Relative Advantages

Mammography Clinical Breast Examination

More proven method (at least in resource-richsettings)

Typically accessed in settings where follow-updiagnostic services are readily available andtravel distances for follow-up are short

Lower costLower technical requirementsWider implementabilityMore easily packaged withother health services



breast symptoms, three of whom had advanced-stage breast cancer.26 In Tanzania, laypersons invillages were trained to provide screening for avariety of cancers by using basic history andphysical examination. After 3 years, breast can-cer downstaging was one of the most significantresults of the program, evidenced by a 74% in-crease in stage I to II breast tumors.27

In LMICs where health care systems are signifi-cantly weakened by limited resources and humancapacity, it is worth emphasizing that anticipatedimpacts of widespread breast cancer screeningwould not be limited to detecting asymptomaticdisease. For example, in Malawi, 47% of womenwith pathologically confirmed breast cancer at thetertiary referral hospital in Lilongwe had symptomdurationsgreater than12months,28 andonly 44%of randomly selected women from rural and urbanareas in the Lilongwe district were aware of breastcancer as a disease.29 Therefore, if CBE wereeffectively scaled up throughout Malawi in a man-ner that engages communities with effective down-stream referral, anticipated benefits might be largewith respect to improved cancer awareness andearlier identification of unaddressed, prevalent,symptomatic disease. In addition, there may becollateral effects on other public health problemsapart from breast cancer, including promotion ofhealthier lifestyles among women as well as in-creased cancer awareness and destigmatization.These off-target effects of breast cancer screeningare no less important simply because they areharder to define and measure than the numberof early-stage breast cancers diagnosed.

Classical cancer screening paradigms and mes-saging must be adapted to the LMIC context. TheHIV implementation science field now championspragmatic scale-up of proven multicomponent in-terventions tomaximize population-level outcomesin LMICs. Similar approaches may be attractive forcancer screening as well. We are currently con-ducting a pilot breast cancer education and CBE

screening intervention in Lilongwe among womenattending diverse health clinics. The major objec-tives are to assess uptake and feasibility of pack-aging CBE with other health services, performancecharacteristics of CBE performed by trained laybreast health promoters, and completion rates forreferrals among women with detected abnormali-ties. These preliminary data will help inform widerscale-up of breast cancer awareness and screen-ing efforts throughout Malawi.

Even as the screening mammography debateevolves in resource-rich settings, mammography isbeing actively promoted and implemented in manyresource-limited countries in the world, includingMalawi. We believe there is agreement within theglobal health community that high breast cancerburden and mortality in LMICs require an urgentresponse, but competing health needs and localrealities require that available resources be opti-mally used to provide the best value for popula-tions overall. This may be particularly true, giventhat several breast cancer screening approachesare available that can be packaged together invarying combinations. We believe more evidenceisneeded toguide large-scalebreastcancerscreen-ing approaches in LMICs under varying socioeco-nomic and cultural conditions, and we emphasizethat although CBE has been shown to result incancer downstaging in LMIC settings, effects onbreast cancer–specific mortality remain unclear.Limited cancer diagnosis, treatment, and registra-tion throughout LMICs also limit the impact ofscreening interventions as well as metrics for theirevaluation and must be simultaneously strength-ened. We eagerly await results of ongoing studies,including our own work, to define optimal ap-proaches in Malawi, with the expectation that suc-cessful strategies here may be quite different fromthose in other LMIC settings.



Financial support: Satish GopalAdministrative support: Satish Gopal



The following represents disclosure information provided byauthors of this manuscript. All relationships are consideredcompensated. Relationships are self-held unless noted. I 5Immediate Family Member, Inst 5 My Institution.Relationships may not relate to the subject matter of this

manuscript. For more information about ASCO’s conflict ofinterest policy, please refer to www.asco.org/rwc orjco.ascopubs.org/site/ifc.

Lily A. GutnikNo relationship to disclose

Beatrice Matanje-MwagombaNo relationship to disclose

Vanessa MsosaNo relationship to disclose

Suzgo MzumaraNo relationship to disclose







Blandina KhondoweEmployment: Center for Medical Diagnostics, Lilongwe

Agnes MosesNo relationship to disclose

Racquel E. KohlerNo relationship to disclose

Lisa A. CareyResearch Funding: GlaxoSmithKline, Genentech/Roche

Clara N. LeeNo relationship to disclose

Satish GopalNo relationship to disclose

AFFILIATIONS

Lily A. Gutnik, Montefiore Medical Center/Albert Einstein College of Medicine, Bronx, NY; Lily A. Gutnik, BlandinaKhondowe, Agnes Moses, and Satish Gopal, UNC[en]Project Malawi; Beatrice Matanje-Mwagomba, Malawi Ministry ofHealth; Vanessa Msosa and Suzgo Mzumara, Kamuzu Central Hospital, Lilongwe; Suzgo Mzumara and Agnes Moses,University ofMalawi College ofMedicine, Blantyre,Malawi; Racquel E. Kohler, Gillings School of Global Public Health; LisaA. Carey, Clara N. Lee, and Satish Gopal, Lineberger Comprehensive Cancer Center; and Satish Gopal, Institute for GlobalHealth and Infectious Diseases, University of North Carolina at Chapel Hill, Chapel Hill, NC.Supported by Grants No. K01TW009488, R21CA180815, and U54CA190152 from the National Institutes of Health(S.G.), Grant No. 5R25TW009340 from the National Institutes of Health Fogarty International Center (L.G.), and by theState Department Fulbright Scholar Program.

REFERENCES1. Institute for Health Metrics and Evaluation (IHME): The Challenge Ahead: Progress and Setbacks in Breast and

Cervical Cancer. Seattle, WA, IHME, 2011

2. Coughlin SS, Ekwueme DU: Breast cancer as a global health concern. Cancer Epidemiol 33:315-318, 2009

3. Howlader N, Noone AM, KrapchoM, et al: (eds). SEER Cancer Statistics Review, 1975-2012. Bethesda, MD, NationalCancer Institute. http://seer.cancer.gov/csr/1975_2012/

4. Akinyemiju TF: Socio-economic and health access determinants of breast and cervical cancer screening in low-income countries: Analysis of the World Health Survey. PLoS One 7:e48834, 2012

5. Yip CH, Smith RA, Anderson BO, et al: Guideline implementation for breast healthcare in low- and middle-incomecountries: Early detection resource allocation. Cancer 113:2244-2256, 2008

6. Pace LE, Keating NL: A systematic assessment of benefits and risks to guide breast cancer screening decisions. JAMA311:1327-1335, 2014

7. Jørgensen KJ, Gøtzsche PC: Overdiagnosis in publicly organised mammography screening programmes: Systematicreview of incidence trends. BMJ 339:b2587, 2009

8. Welch HG, Passow HJ: Quantifying the benefits and harms of screening mammography. JAMA Intern Med 174:448-454, 2014

9. Galukande M, Kiguli-Malwadde E: Rethinking breast cancer screening strategies in resource-limited settings. AfrHealth Sci 10:89-92, 2010

10. The World Bank: GDP per capita (current US$): Malawi. http://data.worldbank.org/indicator/NY.GDP.PCAP.CD/countries/1W?display5default

11. Corbex M, Burton R, Sancho-Garnier H: Breast cancer early detection methods for low and middle income countries,a review of the evidence. Breast 21:428-434, 2012

12. Kaminjolo S: India donates cancer detecting equipment to Malawi. The Nation, February 17, 2004.

13. Harford JB: Breast-cancer early detection in low-income and middle-income countries: Do what you can versus onesize fits all. Lancet Oncol 12:306-312, 2011

14. World Health Organization (WHO): WHO position paper on mammography screening. Geneva, Switzerland, WHO,2014

15. Miller AB, Wall C, Baines CJ, et al: Twenty-five-year follow-up for breast cancer incidence and mortality of the CanadianNational Breast Screening Study: Randomised screening trial. BMJ 348:g366, 2014

16. Fletcher SW: ACP Journal Club: Annual mammography screening did not reduce long-term breast cancer mortality inwomen 40 to 59 years of age. Ann Intern Med 160:JC7, 2014

17. Jatoi I: Screening clinical breast examination. Surg Clin North Am 83:789-801, 2003

18. Anderson BO: Breast cancer: Thinking globally. Science 343:1403, 2014

19. Boulos S, Gadallah M, Neguib S, et al: Breast screening in the emerging world: High prevalence of breast cancer inCairo. Breast 14:340-346, 2005

20. Denewer A, Hussein O, Farouk O, et al: Cost-effectiveness of clinical breast assessment-based screening in ruralEgypt. World J Surg 34:2204-2210, 2010

21. Devi BC, Tang TS, Corbex M: Reducing by half the percentage of late-stage presentation for breast and cervix cancerover 4 years: A pilot study of clinical downstaging in Sarawak, Malaysia. Ann Oncol 18:1172-1176, 2007


http://seer.cancer.gov/csr/1975_2012/

http://data.worldbank.org/indicator/NY.GDP.PCAP.CD/countries/1W?display=default




22. Sankaranarayanan R, Ramadas K, Thara S, et al: Clinical breast examination: Preliminary results from a clusterrandomized controlled trial in India. J Natl Cancer Inst 103:1476-1480, 2011

23. Hyoju SK, Agrawal CS, Pokhrel PK, et al: Transfer of clinical breast examination skills to female community healthvolunteers in Nepal. Asian Pac J Cancer Prev 12:3353-3356, 2011

24. Duffy SW, Tabar L, Vitak B, et al: Tumor size and breast cancer detection: What might be the effect of a less sensitivescreening tool than mammography? Breast J 12:S91-S95, 2006

25. Okonkwo QL, Draisma G, der Kinderen A, et al: Breast cancer screening policies in developing countries: A cost-effectiveness analysis for India. J Natl Cancer Inst 100:1290-1300, 2008

26. Abuidris DO, Elsheikh A, Ali M, et al: Breast-cancer screening with trained volunteers in a rural area of Sudan: A pilotstudy. Lancet Oncol 14:363-370, 2013

27. Ngoma T, Mandeli J, Holland JF: Downstaging cancer in rural Africa. Int J Cancer 136:2875-2876, 2015

28. Kohler RE, Moses A, Krysiak R, et al: Pathologically confirmed breast cancer in Malawi: A descriptive study—Clinicalprofile of breast cancer. Malawi Med J 27:10-12, 2015

29. Kohler RE: Breast cancer in Malawi [doctoral thesis]. University of North Carolina, Chapel Hill, NC, 2015



originalreport

Development of a Breast CancerTreatment Program in Port-au-Prince,Haiti: Experiences From the Field

abstract

Purpose The nonprofit Project Medishare launched a breast cancer treatment program in Port-au-Princein July 2013 to address the demand for breast cancer care in Haiti. We outline the development of theprogram, highlight specific challenges, and discuss key considerations for others working in globaloncology.

MethodsWe reflected on our experiences in the key areas of developing partnerships, building laboratorycapacity, conducting medical training, using treatment algorithms, and ensuring access to safe, low-costchemotherapy drugs. We also critically reviewed our costs and quality measures.

Results The program has treated a total of 139 patients with breast cancer with strong adherence totreatment regimens in 85% of patients. In 273 chemotherapy administrations, no serious exposure oradverse safety events were reported by staff. Themortality rate for 94 patients for whomwe have completedata was 24%with amedian survival time of 53months. Our outcome data were likely influenced by stageat presentation, with more than half of patients presentingmore than 12months after first noticing a tumor.Future efforts will therefore focus on continuing to improve the level of care, while working with localpartners to spread awareness, increase screening, and get more women into care earlier in the course oftheir disease.

Conclusion Our experiences may inform others working to implement protocol-based cancer treatmentprograms in resource-poor settings and can provide valuable lessons learned for future global oncologyefforts.

J Glob Oncol 2. © 2015 by American Society of Clinical Oncology Licensed under the Creative Commons Attribution 4.0License

INTRODUCTION

Breast cancer is the leading cause of cancermortality amongwomen andwas responsible for521,000 deaths worldwide in 2012 alone.1 Overthe last 5 years, breast cancer incidence hasincreased by nearly 20%, with disproportion-ate growth in low-income countries.2,3 Breastcancer mortality rates are highest in low- andmiddle-income countries (LMICs), in whichrapid socioeconomic changes have led to anincrease in breast cancer diagnosis with com-paratively little promotion of early detectionand affordable treatment.4 In fact, the age-standardized mortality rates from breast cancerin developing nations are three times those indeveloped countries.5,6 These trends highlightthe disproportionate growth and impact ofbreast cancer in the developing world and em-phasize the need for a commensurate growth intreatment programs.

Like many other LMICs, Haiti struggles with thegrowing burden of breast cancer while lackingthe health care infrastructure required to identify,diagnose, and treat the disease. Treatment op-tions that do exist are often prohibitively expen-sive, and care is often fragmented because of theabsence of centralized and coordinated healthservices.

Inaneffort toexpandaccessibility to treatment, thenonprofit Project Medishare in partnership withHospital Bernard Mevs and the University of Flor-ida launched a breast cancer treatment programin July 2013. Through an innovativemix of private,public, and research-related funds, the ProjectMedishare breast cancer initiative has providedcare to 139 patients with breast cancer in nearly 2years, becoming an integral part of the oncologycommunity in Port-au-Prince. Project Medisharehas worked in Haiti since 1994 and has partneredwith Hospital BernardMevs since 2010 to provide

Vincent DeGennaro Jr

Rachel Libby

Elizabeth Patberg

Dieudina Gabriel

Samer Al-Quran

Matthew Kasher

Coy Heldermon

Karen Daily

Joseph R. Auguste

Valery C. Suprien

Judith Hurley

Affiliations appear at theend of this article.

Authors’ disclosures ofpotential conflicts ofinterest and contributionsare found at the end of thisarticle.Corresponding author:Vincent DeGennaro Jr,MD, MPH, Division ofHospital Medicine,University of FloridaCollege of Medicine,1600 Archer Rd, Room7402, Gainesville, FL32608; e-mail: [email protected].







trauma and critical care services, making thelocation an ideal place to implement a programinvolving chemotherapy. Before the implementa-tion of this program, neither ProjectMedishare norHospitalBernardMevsprovidedanykindof breastcancer care apart from mastectomies. The pro-gram has also helped provide the necessary in-frastructure to begin collecting data on theepidemiologic profile and tumor biology of breastcancer in Haiti.7

In this article, we outline the key steps inthe implementation of Project Medishare’sbreast cancer treatment program: buildingand maintaining partnerships with the localoncology community; securing laboratory ser-vices abroad while building local laboratorycapacity; using experts in oncology, nursing,and research to establish treatment algorithmsand train local staff; and purchasing andimporting safe, effective, low-cost chemother-apy and hormone therapy. We also discusscosts and other practical barriers to treatment,and we critically reflect on our own objectiveperformance as a program.

METHODS

Building Partnerships

Project Medishare’s breast cancer treatment pro-gram focused on developing strong local partner-ships, which were crucial for the program to gainacceptance in the local medical community. Be-fore launching the project, Project Medisharephysicians and organizers met with cliniciansand administrators from the Haitian Ministry ofHealth and the oncology program at theUniversityHospital of Haiti to discuss the scope of the pro-gram and request their input. Project directorsasked them to review and approve all of the pro-gram’s initial patient staging and treatment pro-tocols before they were submitted to the HaitianMinistry ofHealth for governmental approval. Proj-ect Medishare also sent patients to local healthcareproviders for verification of the clinical stagingin an effort to ensure quality control in the newprogram.

In addition, Project Medishare partnered with theSupport Group Against Cancer to offer counselingand social services to the program’s patients andhelp raise awareness of the program through localtelevision and radio campaigns. This partnershipincreased the program’s credibility in Port-au-Prince among both patients and providers andhelped reach more patients than would otherwisehave been possible.

Pathology and Laboratory Capacity

Thechallengeof limitedpathology services inHaitiis being addressed via a step-wise approach.Because the immunohistochemical studies forestrogen receptor (ER), progesterone receptor,and human epidermal growth factor receptor 2biomarkers cannot currently be performed at anylaboratory in Haiti, it was necessary to turn tointernational pathology laboratories. As an initialstopgap measure, tissues from breast biopsy andexcision specimens were placed in formalin andsent directly to partner laboratories in the UnitedStates for processing, paraffin embedding, andstaining. The specimens were hand carried atregular intervals via travelers from Port-au-Prince to Florida, which requires no special im-portation documentation.

Specimens were previously collected via openbiopsy, but core needle biopsies were imple-mented as part of the program launch. Four Hai-tian doctors were trained in obtaining samples viathree passes with core needle biopsy, typicallyusing 16- to 20-gauge needles. After finding ahigh false-negative rate in tumors that were clin-ically apparent cancer, theprogramswitched to sixquality samples with the core needle and in-creased the needle gauge to 14 or 16.

To set up a permanent and reliable system forascertaining the pathology, Project Medisharepartnered with a local pathology laboratory in Haitito process the formalin-fixed specimens intoparaffin-embedded blocks. All samples were thenevaluated by breast pathologists in the UnitedStates, and all invasive carcinomas were gradedaccording to the Nottingham Histologic Scoresystem. Electronic copies of pathology reportswere sent to the treating physicians by e-mail.All specimens with ductal carcinoma in situ orinvasive carcinoma of the breast were assayed forER, progesterone receptor, and human epidermalgrowth factor receptor 2.8

In conjunction with the Haitian Ministry of Healthand several US universities, a reference labora-tory will soon be established in Port-au-Princethat can perform immunohistochemistry for ERstatus for all specimens originating in Haiti, andtwo new pathology laboratories distributed geo-graphically throughout Haiti will open in the next12 months.

Clinical Training and Treatment Algorithms

Another critical step in the development of theprogram was training clinicians in cancer care. Atthe inception of the program, there were only two



trained subspecialty hematologists in Haiti and nomedical oncologists. Chemotherapy options forpostmastectomy patients in Port-au-Prince at thattime were limited to private clinics and UniversityHospital of Haiti, where patients had to purchasechemotherapy at high prices and often receivedsubstandard chemotherapy regimens such ascyclophosphamide/doxorubicin/fluorouracil. Byusing educational models similar to those devel-oped for addressing the HIV/AIDS epidemic insub-Saharan Africa, local health care providerswere trained to use curative chemotherapy andpalliative care.9 Training at the Project Medishareoncology program was initiated by a US board-certified internist who had previous internationalexperience helping to develop the national cancertreatment program in Rwanda.

Because gynecologists in Haiti were typically thefirst medical interface for patients with breastcancer, they were selected as the clinicians forthe Project Medishare breast cancer treatmentprogram. They received 40 hours of intensiveclassroomeducation onbreast cancer pathophys-iology, presentation, diagnosis, and treatment. Inaddition, they were trained to use chemotherapytreatment algorithms andwere required to discussall patients individually with the program directorfor the first 6 weeks of clinical application.

The development and use of treatment algorithmswas a crucial part of the oncology training. Asmentionedpreviously, theProjectMedishare teamworked in conjunction with oncologists from theUniversity of Florida and theUniversity of Miami todevelop treatment protocols specifically for Haiti.The protocols included basic staging guidelines,timing of surgical interventions, chemotherapyregimens, hormonal therapy, and radiotherapyrecommendations. The two intravenous chemo-therapy regimens are four to six cycles ofdocetaxel/cyclophosphamide or four cycles ofdoxorubicin/cyclophosphamide followed by fourcycles of paclitaxel given once every 3 weeks.These regimens are standard regimens in theUnited States and do not require growth factorsupport. Of note, dose-dense scheduling of che-motherapy is not possible in Haiti because of theunavailability of growth factor support, and weeklydosing with paclitaxel is impractical because ofsignificant issues with transportation. Trastuzu-mab is also not available because of its prohibitivecost. Tamoxifen and letrozole are available andwere used by the program in the neoadjuvant,adjuvant, and metastatic settings. The exact algo-rithm changed depending on the stage, degree oflymphatic invasion, and hormone sensitivity.

Mastectomies and other surgeries, such as bi-lateral salpingo-oophorectomies for premeno-pausal women with ER-positive, advanced-stagetumors, were performed by surgeons at HospitalBernard Mevs. Surgical and anesthesia capacityare less frequently barriers to basic breast cancercare in urban hospitals in developing countriesthanchemotherapyandothermedical treatments.Nevertheless, it is important to note that existingsurgical capacity at Hospital Bernard Mevs waskey to the rapidscalingupof theProjectMedisharebreast cancer treatment program.10

Unfortunately, radiotherapy is not available in Haiti,making it impossible to offer breast conservation forthosewithout the resources topay$2,000forprivatetreatment in the neighboring Dominican Republic.Moreover, 80% of the women treated by the pro-gram in the first 2 years were not candidates forbreast conservation therapy because of tumor size,chest wall invasion, skin ulcerations, or the pres-enceofmetastasis.7Nonetheless, radiotherapywasrecommendedaspart of the treatment algorithm forall women with stage IIB to IIIC disease after che-motherapy and mastectomy.11 Going forward, theHaitiangovernmenthasarticulateda long-termplanto build a radiotherapy center in Haiti, and ProjectMedishare continues to rely on close contacts inDominican Republic radiotherapy centers to helpdirect clinical care decisions for qualifying patients.

For breast cancer cases that fall outside the estab-lished treatment algorithms, breast oncologistsfrom the University of Florida advise the teamon the ground via e-mail and telephone. Roughly10% of patients required consultation from theseoncologists,with the remainderbeingmanagedbythe Haitian gynecologists alone. Discussion re-garding these patients and consistent communica-tion with the University of Florida breast oncologistsprovided Haitian clinicians with an ongoing educa-tion in cancer care.

Nursing protocols were developed for mixing andadministering chemotherapy. Nurses in ProjectMedishare were trained in chemotherapy mixing,administration, and extravasation protocols byHaitian pharmacologists and Haitian-Americanchemotherapy infusion nurses from Miami. TheHaitian compounding pharmacist provided over-sight in mixing the chemotherapy to ensure staffand patient safety. After the initial week-long train-ing, the nurses received continued training on aregular basis from visiting infusion nurses whoconfirmed that the quality of care remained at aconsistently high level. All staff wear personal pro-tective equipment when mixing and administering



chemotherapy, ensuring a safe environment forthe nurses, pharmacists, and patients.

After 12 months of experience, the Project Medi-share nurses designed and implemented an in-service training program for the chemotherapynurses at the public University Hospital of Haitito raise the level of care and improve patient andstaff safety at the country’s largest public hospital.With trained nurses at the two sites, the “train thetrainer” model has proved useful in the Haitiancontext, although periodic quality control by US-based nurses still occurs at present.

All providers and staff were trained to documentclinical encounters on anelectronic record systemthat was principally online but that could be usedoffline during the frequent power outages. Wire-lessUSB jumpdrivemodemsprovided Internet viamobile networks. In addition to adapting the treat-ment protocols to local resources, Project Medi-share also adaptedconsent forms, including thosefor biopsy and receiving chemotherapy treat-ments, to adhere to local standards.

Chemotherapy Supply

A critical challenge to the development of the ProjectMedishare oncology program in Haiti was the in-adequate supply of chemotherapy drugs. ProjectMedishare initially attempted to source chemother-apy drugs locally from each of three pharmacies inPort-au-Prince, but thepriceswereprohibitivelyhigh,selection was small, and quantities were limited andsubject to frequentstockoutages.Genericdrugswerepurchased from a manufacturer (Angel Biogenics,Gujarat, India) that had WHO approval. Oncologistsfrom the University of Florida and the University ofMiami, in consultationwith the team inHaiti, selecteddrug regimens that were easy to administer, did notrequiregrowthfactorsupport,andrequiredthefewestpatient visits possible. We estimated the quantities ofeach drug needed throughdemand forecasting frompartners,basedonthe localpopulationneedsandthefinancial resources available.

As oftenhappens in countrieswith poor infrastruc-ture, the imported medications were significantlydelayed in customs. Themaximumcustoms delaywas 6 weeks, which delayed patients receivingchemotherapyby asmuchas1month.During thistime, Project Medishare was again unsuccessfulin its attempt to purchase drugs locally. The nurs-ing staff implemented an intensive inventory sys-tem at Project Medishare to track usage of thedifferent drugs and ensure timely reordering withthepar level set at 3months, allowing for abuffer of1 month for customs delays. The local design and

ownership of the inventory tracking by thenursing staff contributed to the success ofthe system and helped avoid stock outsof medications in the second year of theprogram.

A significant proportion of patients with breastcancer in Haiti do not finish their prescribedchemotherapy sessions because of drug costs,andmanymoreare forced todelay treatmentwhilesavingmoney for future sessions.12 Because Proj-ect Medishare purchased generic drugs in bulk,our patients’ costs were considerably lower thanthe costs for patients at the University Hospital ofHaiti, who purchased drugs individually from pri-vate pharmacies. Project Medishare now partnerswith the University Hospital of Haiti and the Sup-port Group Against Cancer (Groupe de SupportContre le Cancer) to offer chemotherapy to thepoorer patients on a sliding scale according tomeans, thereby increasing access to care. Pa-tients are able to receive the generic drugs fromthe Project Medishare program, occasionally withfinancial assistance from Support Group AgainstCancer, and bring those drugs to the public Uni-versity Hospital to be infused. Through this part-nership, the University Hospital of Haiti hascontinued to treat patients, demonstrating thecapacity of the public health care system to effec-tively treat patients when the necessary resourcesare available.

Costs

The average cost of treatment per patient wasinitially calculated to be $1,500 per patient: $550formastectomy, $450 for chemotherapy, and$500for the combined cost of pathology, laboratory,radiology, and operating costs. In practice, the costper patient to ProjectMedisharewasapproximately$750 when dividing the total amount of moneyspent by the total number of patients treated.The cost was lower because of less care givenper patient as a result of advanced disease oruntimely death.Mastectomywas performed in only70% of patients (DeGennaro et al, manuscriptsubmitted for publication). The remaining 30% ofpatients presented after they had already had amastectomy at another institution or with stage IVdisease, which did not clinically warrant a mastec-tomy. In addition, the curative chemotherapy reg-imens included more expensive drugs such asdocetaxel and paclitaxel than the palliative regi-mens, which consisted of monotherapy with lessexpensive agents such as doxorubicin or cyclo-phosphamide. We plan to calculate the cost-effectiveness of the program as we go forward.



RESULTS

The program has enrolled 139 patients and offerscare at an average cost of $750 per patient.However, outcome data are limited and arestrongly influenced by poor predictors at time ofpresentation. For example, the initial data from thefirst year of our program show that 80%of patientspresented with stage III to IV disease, and 50%had a tumor in their breast for 12 months at thetime of presentation. The mortality rate for the 94patients for whomwe have complete data, includ-ing thosewhohadbeendiagnosedbefore the startof theprogram,was24%withmediansurvival timeof53months.Barriers to seekingclinicaldiagnosisand treatment for breast cancer in LMICs and inHaiti specifically have previously been investi-gated. Failure to recognize that a lump in thebreast might signal illness and fear of the cost oftreatment were found to be the most significantcontributors to delays in seeking care.12,13

For women receiving intravenous chemotherapy,adherence to 21-day administration intervals forthe entire four- to eight-cycle regimen is recom-mended to all patients.14,15 Traditionally, this hasbeen difficult in Haiti. In efforts to achieve the bestpossible outcomes, clinic staff were taught toemphasize the importance of timely administra-tion of chemotherapy cycles, and patients wereprovided with social services, including transpor-tation costs and nutritional support. A retrospec-tive reviewofourdatashows thatmore than77%ofchemotherapy cycles were delivered on time and85% were delivered within 1 day of the desiredtreatment date. All women who received eithercurative or palliative intravenous chemotherapyand had completed at least two cycles at the timeof the investigation were included. Results can beseen in Table 1.

The reasons for treatment delay were reviewed inhopes of finding modifiable factors. Medical rea-sons for delaying chemotherapy included fever,hemoglobin below 7 g/dL, and an absolute neu-trophil countbelow1,000.Patientswhopresentedwith these adverse eventswere given treatment forthe symptoms, and chemotherapy was resumedas soonas it was safe to do so. Supply-chaindelaysin importing chemotherapy drugs and/or procur-ing chemotherapy drugs locally led to more sig-nificant delays.

A safety review showed that in 273 chemotherapyadministrations, therewere no instances of extrav-asation. Therewere three instances of chemother-apy drugs being spilled on the floor, which wereaddressed with preapproved clean-up protocols.There were seven allergic reactions, all to doce-taxel, and those quickly went to zero after imple-menting prednisone pretreatment andadministering dexamethasone no more than 1hour before chemotherapy initiation.

CONCLUSION

Strong local partnerships, significant commitmentby the staff of Project Medishare, collaborationwith oncology institutions and laboratoriesabroad, and lessons learned from similar initia-tives all helped to make the breast cancer treat-ment program possible. Despite these efforts,there is significant room for improvement. Mostimportantly, an increase in cancer awarenessnationally is needed to create a shift in cancerstage at presentation and an increase in the curerate.7 Project Medishare is working with theHaitian Ministry of Health to expand access topathology, surgery, and chemotherapy forbreast cancer nationally by using these sametechniques. Project Medishare staff arehopeful that this cancer treatment program willdemonstrate new possibilities for hospitals inresource-poor settings and that our lessonslearned highlight important considerations and so-lutions for advancing cancer care in the developingworld.

DOI: 10.1200/JGO.2015.000364Published online ahead of print at jgo.ascopubs.org on October28, 2015


Conception and design: Vincent DeGennaro Jr, Rachel Libby,Elizabeth Patberg, Judith HurleyCollectionandassembly of data:VincentDeGennaro Jr, RachelLibby, Elizabeth Patberg, Dieudina Gabriel, Samer Al-Quran, Matthew Kasher, Joseph R. Auguste, Valery C.Suprien

Data analysis and interpretation: Vincent DeGennaro Jr,Elizabeth Patberg, Matthew Kasher, Coy Heldermon,Karen Daily, Joseph R. Auguste, Valery C. Suprien,Judith Hurley


Table 1 – Rates of Treatment Schedule Adherence

Dosing ScheduleNo. of IV Chemotherapy Doses Administered (%)

(N = 273)

Given every 21 days exactly 211 (77.3)

Given every 21 days 6 1 day 234 (85.8)

NOTE. Total No. of patients for this analysis was 58.Abbreviation: IV, intravenous.






The following represents disclosure information provided byauthors of this manuscript. All relationships are consideredcompensated. Relationships are self-held unless noted. I =ImmediateFamilyMember, Inst=MyInstitution.Relationshipsmaynot relate to thesubjectmatter of thismanuscript. FormoreinformationaboutASCO’sconflictof interestpolicy,pleasereferto www.asco.org/rwc or jco.ascopubs.org/site/ifc.

Vincent DeGennaro JrNo relationship to disclose

Rachel LibbyNo relationship to disclose

Elizabeth PatbergNo relationship to disclose

Dieudina GabrielNo relationship to disclose

Samer Al-QuranNo relationship to disclose

Matthew KasherNo relationship to disclose

Coy HeldermonNo relationship to disclose

Karen DailyNo relationship to disclose

Joseph R. AugusteNo relationship to disclose

Valery C. SuprienNo relationship to disclose

Judith HurleyNo relationship to disclose

AFFILIATIONS

Vincent DeGennaro Jr, Samer Al-Quran, Coy Heldermon, and Karen Daily, University of Florida College of Medicine, Gainesville;Rachel Libby and Judith Hurley, University of Miami Miller School of Medicine, Miami, FL; Elizabeth Patberg, Emory UniversitySchool of Medicine, Atlanta, GA; Vincent DeGennaro Jr, Dieudina Gabriel, Joseph R. Auguste, and Valery C. Suprien, ProjectMedishare, Port-au-Prince, Haiti; and Matthew Kasher, University of North Carolina School of Medicine, Chapel Hill, NC.

Supported by the Gatorade Trust through funds distributed by the Department of Medicine, University of Florida, and byProject Medishare.

REFERENCES1. World Health Organization: Media Centre: Cancer—Fact Sheet No. 297, updated February 2015. http://www.who.int/

mediacentre/factsheets/fs297/en/

2. Anderson BO, Jakesz R: Breast cancer issues in developing countries: An overview of the Breast Health GlobalInitiative. World J Surg 32:2578-2585, 2008

3. Mackay J, Eriksen M, Shafey O, et al: The Cancer Atlas. American Cancer Society, 2006

4. Ferlay J, Bray F, Pisani P, et al: GLOBOCAN 2002: Cancer Incidence, Mortality and Prevalence Worldwide. Can-cerBase No. 5, version 2.0. Lyon, France, IARC Press, 2004

5. National Cancer Institute, Surveillance, Epidemiology, and End Results (SEER) Program: SEER Cancer Statistics.http://seer.cancer.gov/mortality/

6. Bray F, McCarron P, Parkin DM: The changing global patterns of female breast cancer incidence andmortality. BreastCancer Res 6:229-239, 2004

7. Reference deleted

8. Hammond ME, Hayes DF, Dowsett M, et al: American Society of Clinical Oncology/College of American Pathologistsguideline recommendations for immunohistochemical testing of estrogen and progesterone receptors in breast cancer(unabridged version). Arch Pathol Lab Med 134:e48-e72, 2010

9. Farmer P, Frenk J, Knaul FM, et al: Expansion of cancer care and control in countries of low andmiddle income: A callto action. Lancet 376:1186-1193, 2010

10. DeGennaro V Jr, DeGennaro VA, Bitar M, et al: Building advanced surgical capacity at a hospital in Port-au-Prince,Haiti. J Craniofac Surg 26:1042-1047, 2015

11. National Cancer Comprehensive Network: Clinical Practice Guidelines in Oncology: Breast Cancer, Version 2. 2011

12. Sharma K, Costas A, Damuse R, et al: The Haiti Breast Cancer Initiative: Initial findings and analysis of barriers-to-caredelaying patient presentation. J Oncol 2013:206367, 2013

13. Ramirez AJ, Westcombe AM, Burgess CC, et al: Factors predicting delayed presentation of symptomatic breastcancer: A systematic review. Lancet 353:1127-1131, 1999

14. Electronic Medicines Compendium: Doxorubicin 2 mg/ml concentrate for solution for infusion: Summary of productcharacteristics, updated February 20, 2014

15. Electronic Medicines Compendium: Docetaxel accord 160 mg/8 ml concentrate for solution for infusion: Summary ofproduct characteristics, updated October 10, 2014




http://www.who.int/mediacentre/factsheets/fs297/en/

http://www.who.int/mediacentre/factsheets/fs297/en/

http://seer.cancer.gov/mortality/


originalreport

Impact of the Bim Deletion Polymorphismon Survival Among Patients WithCompletely Resected Non–Small-Cell LungCarcinoma

abstract

Purpose A deletion polymorphism of the Bim gene has been reported to be a prognostic factor for patientswith non–small-cell lung cancer (NSCLC) treated with epidermal growth factor receptor-tyrosine kinaseinhibitors in the Asian population.We investigated the impact of theBim deletion polymorphismon survivalamong patients with completely resected NSCLC.

Patients and Methods The Bim polymorphism was detected by polymerase chain reaction analysis. Wemeasured overall survival (OS) and recurrence-free survival rates in 411 patients and postrecur-rence survival (PRS) in 94 patients who experienced recurrence and received additional anticancertherapy.

Results The Bim deletion polymorphism was detected in 61 patients (14.8%). OS rates were significantlylower for patients with the Bim deletion polymorphism than for those with the wild-type sequence. Onmultivariable analysis, the Bim deletion polymorphism was identified as an independent prognostic factorfor OS (hazard ratio, 1.98; 95% CI, 1.17 to 3.36; P 5 .011). Among the 94 patients who experiencedrecurrence andwere treatedwith anticancer therapy, patients with theBim deletion polymorphism showedsignificantly poorer PRS than those with the wild-type sequence (median, 9.8 months v 26.9 months,respectively; P < .001). Multivariable analysis revealed that the Bim deletion polymorphism was anindependent predictor of PRS (hazard ratio, 3.36; 95% CI, 1.75 to 6.47; P < .001). This trend remainedapparent in subgroup analyses stratified by EGFR status, histology, and therapeutic modality.

Conclusion The Bim deletion polymorphism is a novel indicator of shortened PRS among patients withrecurrent NSCLC treated with anticancer therapy in the Asian population.


INTRODUCTION

Lung cancer is the leading cause of cancer deathworldwide.1 Even after radical surgery in patientswith early-stage non–small-cell lung cancer(NSCLC), 30% to 40% of patients experience re-currence within 5 years.2,3 Postoperative recurrentdisease is usually treated as metastatic NSCLC.Although molecule-targeted drug therapies suchasepidermal growth factor receptor-tyrosinekinaseinhibitors (EGFR-TKIs) have produced consider-able survival benefits in patients with both ad-vanced disease and postoperative recurrence ofEGFR-mutated NSCLC,4-8 the majority of patientseventually become refractory to these therapies.

B-cell chronic lymphatic leukemia lymphoma 2-like 11 (BCL2L11), or BIM, is a proapoptotic

member of the Bcl-2 protein family and is a keymodulator of EGFR-TKI–induced apoptosis inNSCLC cell lines.9 Ng et al10 reported a commonintronic deletion with a 2,903-base pair (bp) poly-morphism in thegeneencodingBIM.Thisdeletionpolymorphism leads to impaired expression ofBH3-containing BIM isoforms, resulting in resis-tance to EGFR-TKIs in patients with NSCLC whohave EGFR mutations. Interestingly, this deletionpolymorphism was observed only in East Asianpopulations.10 Several clinical studies of EastAsian populations have indicated that the Bimdeletion polymorphism is an independent prog-nostic factor for progression-free survival inadvanced EGFR-mutated NSCLC treated withEGFR-TKIs11,12 and cytotoxic chemotherapy.13

The Bim deletion polymorphism is expected to

Jun Atsumi

Kimihiro Shimizu

Yoichi Ohtaki

Kyoichi Kaira

Seiichi Kakegawa

Toshiteru Nagashima

Yasuaki Enokida

Seshiru Nakazawa

Kai Obayashi

Yoshiaki Takase

Osamu Kawashima

Mitsuhiro Kamiyoshihara

Masayuki Sugano

Takashi Ibe

Hitoshi Igai

Izumi Takeyoshi

Authors’ disclosures ofpotential conflicts ofinterest and contributionsare found at the end of thisarticle.Corresponding author:Kimihiro Shimizu, MD,PhD, Department ofThoracic and VisceralOrgan Surgery, GunmaUniversity GraduateSchool of Medicine, 3-39-22 Showa-machi,Maebashi, Gunma 371-8511, Japan; e-mail:[email protected].





beanovel biomarker in anticancer therapy againstinoperable NSCLC, especially adenocarcinoma.Patients with NSCLC who have recurrence aftercurative surgery have a more favorable prognosisthan those with advanced-stage disease at initialpresentation, because patients with NSCLC whohave postoperative recurrence have different char-acteristics from those with stage IV disease.14,15

However, there have been no studies regarding theprognosticpowerof theBimdeletionpolymorphismin postoperative patients with lung cancer, includ-ing those with non-adenocarcinoma histology, orthe influence of the polymorphism on postrecur-rence treatment.

We hypothesized that the Bim deletion polymor-phism affects survival among patients with post-operative recurrent NSCLC who have receivedanticancer therapy. In this study, we investigatedthe impact of the Bim deletion polymorphism onthe outcomes of patients with completely resectedNSCLC.

PATIENTS AND METHODS

Patients and data collection

A total of 565 patients with NSCLCwho underwentpulmonary resectionatGunmaUniversityHospitalbetween June 2003 and December 2013 wereidentified in our database. Among these patients,481 underwent complete resection (lobectomy orgreater with systematic lymph node dissection)without induction chemotherapy or radiotherapy.We excluded patients with residual lesions (mac-roscopically or microscopically apparent), as wellas thosewithpathologic stage IVdiseaseand thosewithout adequate documentation. Consequently,411 patients were eligible for inclusion in thisstudy. Histologic diagnoses were made on thebasis of WHO criteria,16 and disease stage wasdetermined according to the TNMClassification ofMalignant Tumors, 7th edition. This study wasapproved by the ethics committee of Gunma Uni-versity Hospital. Informed consent for a globalgenome analysis of samples was obtained fromeach patient before inclusion in the study. Insti-tutional review board approval for the study wasobtained for the analysis of Bimand other genes inthose samples.

Diagnosis of Recurrence and Survival Analysis

Patients were followed at 3-month intervals for thefirst 2 years and at 6-month intervals thereafter onan outpatient basis. Follow-up evaluation included aphysical examination, chest radiography, and bloodanalysis, including analysis of pertinent tumormarkers. Computed tomography of the chest

and abdomen or positron emission tomography-computed tomography was performed annually.When symptoms or signs of recurrence were de-tected, further evaluations were performed. Re-currence was diagnosed based on compatiblephysical examination and diagnostic imagingfindings, and the diagnosis was confirmed histo-logically when clinically feasible. The date of re-currence was defined as the date of histologicconfirmation, or in patients whose diagnosis wasbased on clinical evidence, the date of recognitionof recurrent disease by the attending physician.Local recurrence was defined as disease recur-rence at the surgical margin, ipsilateral hemi-thorax, or mediastinum. Distant metastasis wasdefined as disease recurrence in the contralaterallung or outside the hemithorax and mediastinum.

The overall survival (OS) periodwas defined as thetime between the date of surgery and the date ofdeath as a result of any cause. Patients who werelost to follow-upwere censored fromanalysis at thetime of the last negative follow-up. For the patientswho developed recurrent disease during follow-up, postrecurrence survival (PRS) was measuredfrom the date of initial recurrence to the date ofdeath as a result of any cause or the date onwhichthepatientwas last known tobealive.Recurrence-free survival (RFS) was measured from the date ofsurgery to the date of initial recurrence.

DNA Extraction and Gene Analysis

After surgical removal of the tumor, a portion ofeach sample was immediately frozen and storedat 280°C before DNA extraction. Genomic DNAwas extracted from a 3- to 5-mm cube of tumortissue by using DNA Mini Kits (QIAGEN, Hilden,Germany) and was subsequently diluted to a con-centration of 20 ng/mL. EGFR mutations in lungcancer tissue were analyzed by peptide nucleicacid–enriched sequencing, as described previ-ously.17 Presence of the Bim deletion polymor-phism was analyzed by first extracting DNA fromperipheral blood mononuclear cells by using aQIAamp DNA Blood Mini Kit (QIAGEN, Venlo,the Netherlands) followed by polymerase chainreaction assay as described previously.11

Statistical Analysis

Statistical analyseswereconductedbyusingSPSSsoftware for Windows, version 12.0 (SPSS, Chi-cago, IL) and Power and Sample Size Calcula-tion software, version 3.1.2 (http://biostat.mc.vanderbilt.edu/wiki/Main/PowerSampleSize). Allcategorical variables were analyzed by using thex2 test. Continuous variables were compared by


http://biostat.mc.vanderbilt.edu/wiki/Main/PowerSampleSize

http://biostat.mc.vanderbilt.edu/wiki/Main/PowerSampleSize


using the independent samples t test. Survival wasanalyzed by using the Kaplan-Meier method, andstatistical analysis was performed by using the log-rank test.Prognostic groupswereassessedbyusingCox proportional hazards regression analysis. Vari-ables significantly associated with OS and PRS onunivariable analysis were tested by multivariableanalysisusingaCoxproportionalhazards regressionmodel. A two-tailed P value of less than .05 wastaken to indicate statistical significance. On thebasis of previous reports,12,13,18 we assumed that13.7% of Japanese patients had the Bim deletionpolymorphism and an OS of 24.8 and 16.8months, respectively, for patients with advanced

NSCLC who received anticancer therapy in theBim wild-type and Bim deletion groups. Underthese assumptions, with a two-tailed a of .05 andpower at 0.8, 64 patients with the Bim deletionpolymorphism and 403 patients with the wild-typesequencewere required to evaluate the effect of theBim deletion polymorphism on PRS for anticancertherapy.

RESULTS

Clinicopathologic Characteristics

Patient characteristics are presented in Table 1.All patients were Japanese. Themedian age at the

Table 1 – Baseline Patient Characteristics and Bim Deletion Polymorphism Distribution

VariableAll

(n 5 411)

Bim Polymorphism Status

PWild Type

(n 5 349) No. (%)Deletion Polymorphism

(n 5 61) No. (%)

Median age, years 67.6 67.9 66.5 .294

Sex .483

Female 175 152 (86.9) 23 (13.1)

Male 236 198 (83.9) 38 (16.1)

Smoking status .086

Never smoker 157 140 (89.2) 17 (10.8)

Ever smoker 254 210 (82.7) 44 (17.3)

Histology .064*

ADC 297 259 (87.2) 38 (12.8)

SQC 93 74 (79.6) 19 (20.4)

Other 21 17 (81.0) 4 (19.0)

Tumor size, cm .124

# 3 248 215 (86.7) 33 (13.3)

. 3 163 135 (82.8) 28 (17.2)

Node metastases .007

N0 303 267 (88.1) 36 (11.9)

N1-2 108 83 (76.9) 25 (23.1)

Vascular invasion .197

Negative 256 223 (87.1) 33 (12.9)

Positive 155 127 (81.9) 28 (18.1)

Lymphatic invasion .024

Negative 238 211 (88.7) 27 (11.3)

Positive 173 139 (80.3) 34 (19.7)

Pathologic stage .001

I 275 246 (89.5) 29 (10.5)

II or III 136 104 (76.5) 32 (23.5)

EGFR gene 1.000

Wild type 276 235 (85.1) 41 (14.9)

Mutation 135 115 (85.2) 20 (14.8)

Abbreviations: ADC, adenocarcinoma; EGFR, epidermal growth factor receptor; SQC, squamous cell carcinoma.*ADC v SQC and other.



time of surgery was 67.6 years (range, 36 to 90years), and the study population consisted of 175females and236males.On thebasis of thehistologyof the lesions, the study population included 297adenocarcinomas, 93 squamous cell carcinomas,12 large-cell neuroendocrine carcinomas, sevenlarge-cell carcinomas, and two adenosquamouscellcarcinomas. With regard to diagnosis, 275 patientswere classified as pathologic stage I, and 136 pa-tientswereclassifiedas stage II or III.EGFRmutationwas detected in 135 tumors (32.8%) consisting of133 adenocarcinomas and two squamous cell car-cinomas. The Bim deletion polymorphism was de-tected in 61 patients (14.8%). The percentage ofpatients according to sex, smokinghistory, histology,and EGFR mutational status did not differ signifi-cantly between the Bim wild-type and Bim deletionpolymorphism groups, although the percentage oflymphnodemetastases,positive lymphatic invasion,andadvancedstage in thepatientswithBimdeletionpolymorphismwas significantly higher than in thosewith the wild-type Bim sequence (Table 1).

Prognostic Impact of Bim Polymorphism on OSand RFS

Factors associated with OS and RFS, as revealedthrough univariable analysis, included sex, smok-ing status, histology, vascular invasion, lymphaticinvasion, pathologic stage, EGFR gene status, andBim polymorphism. In the multivariable analysis,pathologic stage, EGFR gene mutation, and theBim deletion polymorphism were independentfactors associated with OS, and pathologic stageand lymphatic invasion were independent factorsassociated only with poor RFS (Table 2). The Bimdeletion polymorphism was independently asso-ciated with OS but not RFS in 411 patients. The5-year OS rate was significantly lower for patientswith the Bim deletion polymorphism comparedwith those with wild-type Bim (58.8% v 78.9%,

respectively; P, .001; Fig 1A). To eliminate bias,we analyzed survival by using propensity scorematching (DataSupplement).The5-yearOS in thepropensity score–matched analysis was also sig-nificantly poorer in patients withBim deletion thanin those with wild-type Bim (58.8% v 80.3%,respectively; P 5 .036; Fig 1B).

In addition, we investigated RFS among patientswho developed recurrence. As of October 2014,109 patients had experienced recurrence. Patientcharacteristics are shown in theDataSupplement.In the univariable analysis, the variables associatedwith RFS in patients with recurrence were vascularinvasion andBimdeletion polymorphism, and theseremainedas independent factors in themultivariableanalysis (Data Supplement). Furthermore, patientswith the Bim deletion polymorphism showed signif-icantly shortened RFS compared with those withwild-type Bim (median, 9.8 v 13.9 months, respec-tively; P 5 .003; Fig 2A).

Prognostic Impact of Bim Polymorphism on PRS

To determine the impact of the Bim deletion poly-morphism on outcome after recurrence, we investi-gated94(86%)of109patientswith recurrentdiseasewhoreceivedadditionalanticancer therapy, includingcytotoxic chemotherapy, EGFR-TKIs, or radiotherapywith curative intent. The characteristics of the 94patients who received anticancer therapies are sum-marized inTable3. Themedian time to follow-upwas16.4months (range,2.0 to91.8months),medianageat recurrence was 68.6 years (range, 37 to 80 years),and the patients consisted of 38 females and 56males. There were 65 patients with adenocarcinomaand 29 with non-adenocarcinomas (23 squamouscell carcinomas, four large-cell neuroendocrinecarcinomas, and two large-cell carcinomas). Sixteenpatients (17%) harbored the Bim deletion poly-morphism, and 29 patients (31%) harbored

Table 2 – Multivariable Analysis of Predictors of OS and RFS

Variable

OS RFS

HR (95% CI) P HR (95% CI) P

Female sex 1.008 (0.456 to 2.229) .985 1.056 (0.562 to 1.985) .866

Ever smoker 0.817 (0.330 to 2.020) .662 0.974 (0.487 to 1.945) .940

ADC histology 0.656 (0.375 to 1.147) .139 0.776 (0.495 to 1.218) .271

Positive vascular invasion 1.211 (0.704 to 2.084) .489 1.039 (0.675 to 1.601) .861

Positive lymphatic invasion 1.210 (0.674 to 2.171) .524 1.673 (1.032 to 2.711) .037

Pathologic stage II or III 5.213 (2.913 to 9.327) , .001 4.738 (2.994 to 7.498) , .001

EGFR mutated 0.358 (0.172 to 0.743) .006 0.762 (0.466 to 1.245) .277

Bim deletion polymorphism 1.979 (1.166 to 3.357) .011 1.231 (0.781 to 1.939) .370

Abbreviations: ADC, adenocarcinoma; EGFR, epidermal growth factor receptor; HR, hazard ratio; OS, overall survival; RFS, recurrence-free survival.



EGFR-mutated tumors. Thirty-seven patients(39%) showed local recurrence only, and 59patients (61%) showed distant recurrence. Re-currence in multiple foci was detected in 65 pa-tients (69%). Treatment for recurrence consistedof platinum-based chemotherapy in 43 patients(46%), radiotherapy in 43 patients (46%), andEGFR-TKIs in 33 patients (35%). No significantdifferences in age, sex, tumor histology, smokingstatus, pathologic stage, site of recurrence, num-ber of recurrent foci, EGFR gene status, or ther-apeutic modality were observed between patientswith the Bim deletion polymorphism and thosewith wild-type Bim (Table 3).

Univariable analysis indicated that RFS shorterthan 12 months, EGFR gene status, and Bimpolymorphism influenced PRS, all of whichremained as independent prognostic factors forPRS in the multivariable analysis (Table 4).Median PRS was 26.9 months among those withwild-type Bim and 11.4months among those withthe Bim deletion polymorphism (P, .001; Fig 2).SubsetanalysisofPRSshowedthatpatientswithwild-type Bim consistently showed prolonged survival

comparedwith thosewith the deletion polymorphismwhen stratified by EGFR gene status (mutated:median, 61.0 v 23.2 months; P , .001; Fig 3A;wild-type: median, 19.7 v 9.8 months; P 5 .001;Fig 3B) or tumor histology (adenocarcinoma: me-dian, 33.9 v 11.4 months; P5 .009; Fig 3C; non-adenocarcinoma: median, 19.7 v 9.8 months;P 5 .013; Fig 3D). When analyzed according totherapeutic modality, the median PRS was signif-icantly shorter in patients with the Bim deletionpolymorphism compared with those with thewild-type Bim or EGFR-mutated NSCLC treatedwith EGFR-TKIs (median, 38.1 v 23.2 months,respectively; P5 .007; Fig 4A), those treated withcytotoxic chemotherapy alone (median, 18.5 v6.2 months, respectively; P 5 .003; Fig 4B),and those treated with radiotherapy alone (me-dian, 26.9 v 11.4 months, respectively; P5 .046;Fig 4C). Nobiaswas observed in the distribution ofthe Bim deletion polymorphism in terms of plati-num or taxane use among the 23 patients whoreceivedcytotoxicchemotherapy.Similarly, therewasno significant difference in the distribution of theBimdeletion polymorphism according to the radiother-apy method (conventional or cyberknife) or total

A

0

Over

all S

urvi

val (

prob

abili

ty)

Time (months)

1.0

0.8

0.6

0.4

0.2

12 24 36 48 60

WT Bim (n = 350): 5-year survival rate, 78.9%Bim deletion (n = 61): 5-year survival rate, 58.8%Log-rank P < .001Hazard ratio, 2.55 (95% CI, 1.53 to 4.26)

B

0

Over

all S

urvi

val (

prob

abili

ty)

Time (months)

1.0

0.8

0.6

0.4

0.2

12 24 36 48 60

WT Bim (n = 61): 5-year survival rate, 80.3%Bim deletion (n = 61) : 5-year survival rate, 58.8%Log-rank P = .036Hazard ratio, 2.27 (95% CI, 1.03 to 4.98)

A

0

Recu

rren

ce-F

ree

Surv

ival

(p

roba

bilit

y)

Post

recu

rren

ce S

urvi

val

(pro

babi

lity)

Time (months)

1.0

0.8

0.6

0.4

0.2

12 24 36 48 60

WT Bim: 13.9 months Bim deletion: 9.8 months Log-rank P = .002 Hazard ratio, 2.21(95% CI, 1.31 to 3.73)

WT Bim: 26.9 months Bim deletion: 9.8 months Log-rank P < .001 Hazard ratio, 3.65(95% CI, 1.93 to 6.90)

B

0

Time (months)

1.0

0.8

0.6

0.4

0.2

12 24 36 48 60

Figure 1 –

Kaplan-Meier curves foroverall survival according tothe Bim polymorphism.Overall survival in (A) 411patients with non–small-cell lung cancer and (B)propensity score–matchedpatients (n5122).WT,wildtype.

Figure 2 –

Kaplan-Meier survivalcurves for patients whodeveloped recurrentdisease classifiedaccording to the Bimdeletion polymorphism. (A)Recurrence-free survival in109 patients whodeveloped recurrence. (B)Postrecurrence survival in94 patients who receivedanticancer therapy afterrecurrence. WT, wild type.



radiation dose among the 22 patients who re-ceived radiotherapy alone (Table 5).

DISCUSSION

The Bim deletion polymorphism has been investi-gated in inoperable advanced NSCLC and identifiedas a heritable factor conferring resistance toEGFR-TKIs and chemotherapy in the Asianpopulation.10-13,19 However, only one report hasexamined the impact of the Bim deletion polymor-phism on survival in patients with resectableNSCLC.18 In this study, wedemonstrated the impactof the Bim deletion polymorphism on NSCLC out-comes (survival) after complete tumor resection. TheBim deletion polymorphism was an independentunfavorable prognostic factor of OS in all patientswithNSCLCwho receivedcomplete resection,whichwas the result of shorter RFS and PRS associatedwith the Bim deletion polymorphism among thosewho developed recurrent disease. Furthermore, thePRS trend was consistent in subgroup analysesstratified by EGFR mutation status, histology, andtherapeutic modality. On the basis of the results ofthis study, we suggest that the Bim deletion poly-morphismhas a positive impact on early emergenceof metastasis and a negative impact on anticancertreatment in recurrent NSCLC.

There have been few studies regarding the biologiccharacteristics associatedwithBimdeletion polymor-phism,10 but several basic studies demonstrated thatthe BIM protein is essential for anticancer therapy-induced apoptosis. EGFR-TKI–induced apoptosis re-quires BIM protein expression in EGFR-mutatedNSCLC cell lines,9 and clinical studies have focusedon the relationship between the Bim deletion poly-morphism or Bim messenger RNA expression andEGFR-mutated NSCLC treated with EGFR-TKIs.10-13,20 Our results support the notion that theBim deletion polymorphism is an indicator of signif-icantlypooreroutcomesforEGFR-TKI therapyagainstEGFR-mutated NSCLC (Fig 4A). In terms of cytotoxicchemotherapy, BIM protein was shown to mediateapoptosis inducedbypaclitaxel inNSCLCcells and tobe a major determinant in the response of tumors topaclitaxel.21,22Wang et al23 reported that BIMplays a

Table 3 – Characteristics of Patients Who Received Anticancer Therapy After Recurrenceand Bim Deletion Polymorphism Distribution (n 5 94)

VariableAll

(n5 94)

Bim Polymorphism Status

P

Wild Type(n 5 78)No. (%)

DeletionPolymorphism

(n 5 16)No. (%)

Median age (years) 68.6 69.2 66.1 .284

Sex 1.000

Female 38 32 (84.2) 6 (15.8)

Male 56 46 (82.1) 10 (17.9)

Smoking status .405

Never smoker 33 29 (87.9) 4 (12.1)

Ever smoker 61 49 (80.3) 12 (19.7)

Histology .244*

ADC 65 56 (86.2) 9 (13.8)

SQC 23 18 (78.3) 5 (21.7)

Other 6 4 (66.7) 2 (33.3)

Pathologic stage .135

I 29 27 (93.1) 2 (6.9)

II or III 65 51 (78.5) 14 (21.5)

Recurrence-free survival, months .093

, 12 39 29 (74.4) 10 (25.6)

$ 12 55 49 (89.1) 6 (10.9)

Site of recurrence .579

Local only 37 32 (86.5) 5 (13.5)

Distant 57 46 (80.7) 11 (19.3)

Recurrent foci .573

Single 30 24 (80.0) 6 (20.0)

Multiple 64 54 (84.4) 10 (15.6)

EGFR gene 1.000

Wild type 65 54 (83.1) 11 (16.9)

Mutation 29 24 (82.8) 5 (17.2)

Treatment

Platinum-based chemotherapy 43 35 (81.4) 8 (18.6) .786

Radiotherapy 43 34 (79.1) 9 (20.9) .415

EGFR-TKI 33 30 (90.1) 3 (9.1) .160

Abbreviations: ADC, adenocarcinoma; EGFR-TKI, epidermal growth factor receptor-tyrosine kinaseinhibitor; SQC, squamous cell carcinoma.*ADC v SQC and other.

Table 4 – Univariable and Multivariable Analyses of Predictors of PRS

VariableUnivariable Analysis Multivariable Analysis

HR (95% CI) P HR (95% CI) P

RFS , 12 months 1.990 (1.172 to 3.380) .011 1.804 (1.045 to 3.117) .034

Bim deletion polymorphism 3.645 (1.925 to 6.901) , .001 3.363 (1.747 to 6.474) , .001

EGFR mutated 0.356 (0.189 to 0.668) .001 0.344 (0.183 to 0.647) .001

Abbreviations: EGFR, epidermal growth factor receptor; HR, hazard ratio; PRS, postrecurrence survival; RFS, recurrence-free survival.



critical role in cisplatin resistance, demonstratingthat BIM protein is degenerated in cisplatin-resistant but not in cisplatin-sensitive cells, andinhibition of BIM degeneration can effectively in-ducecancer cell death.Becauseexpression of theproapoptotic BH3 domain in BIM is suppressed inindividualswith theBimdeletion polymorphism,10

sensitivity tocytotoxicchemotherapymaybe low insuch patients. Consequently, as demonstratedhere and in a previous study,13 patients with theBim deletion polymorphism tend to have shortersurvival periods than thosewithwild-typeBimaftercytotoxic chemotherapy (Fig 4B).

With regard to radiotherapy-induced apoptosis,it has been reported that radiation increasesFOXO3a protein expression, leading to upregula-tion of BIM expression and apoptotic induction, areaction that is downstream of the PI3K/AKT sig-naling pathway and independent of the p53pathway.24,25 The PI3K/AKT pathway, which reg-ulates BIM expression, is expected to contributeto radiotherapy resistance, and blockade of thepathway may enhance cancer cell radiotherapysensitivity.25,26 Our results indicate that the Bimdeletion polymorphism is an indicator of poorer

radiotherapy outcomes in recurrent NSCLCafter complete resection (Fig 4C). Taken to-gether, these findings suggest that the Bim de-letion polymorphism confers resistance againsttreatment with EGFR-TKIs, chemotherapy, andradiotherapy.

The relationship betweenBIMand tumor develop-ment has been investigated in several solid tumors.Comparison of BIM levels in primary and meta-static tumors revealed progressive decreases inBIM expression in melanoma,27 renal cell carci-noma,28 and colon carcinoma cells.29 In NSCLCcells, low BIM expression was observed morefrequently in cases of advanced pathologic stage,poorer differentiation, and squamous histology,although no impact on survival was observed.30

These studies support the suggestion that BIMprotein plays an important role in suppressingtumor development. Merino et al31 recently re-ported thatBim loss does not affect proliferation orthe expression of epithelial-mesenchymal transi-tionmarkers but does increase the number of lungmetastases inbreast cancers. Theysuggested thatthe loss of Bim may be responsible for dissemina-tion of tumor cells and their colonization of distant

Figure 3 –

Kaplan-Meier curves forpostrecurrence survival(PRS) according to EGFRgene status and histology.(A) PRS in EGFR-mutatedtumors according to wild-type (WT) Bim (n 5 24) orthe Bim deletionpolymorphism (n 5 5).(B) PRS in wild-type EGFRtumors according to wild-type Bim (n 5 54) or theBimdeletion polymorphism(n 5 11). (C) PRS inadenocarcinomaaccording to wild-type Bim(n5 56) or theBimdeletionpolymorphism (n 5 9).(D) PRS in nonadenocarcinoma according towild-type Bim (n 5 22) orthe Bim deletionpolymorphism (n 5 7).



tissues.31 Subgroup analyses in our study revealedno significant differences between patients withstage I NSCLC with or without the Bim deletionpolymorphism (n 5 275) with respect to the per-centage of patients with lymphatic invasion (27.6%v 23.6%, respectively; P5 .648), to the 5-year OSrate (81.8% v 90.4%, respectively; P5 .402), or tothe 5-year RFS rate (83.1% v 70.0%, respectively;P 5 .806), consistent with a previous report.18

Furthermore, our study demonstrated that theBim polymorphism was a significant predictor ofRFS only for patients with recurrence (Data Sup-plement). These results suggest that the Bim de-letion polymorphism is associated with tumordevelopment in disseminatedormetastatic lesions,whereas it has little influence on the primary lesion.Patients with recurrence are likely to have micro-metastases at the time of surgery, and thereforeBim deletion polymorphism may somehow be in-volved in growth of these metastatic lesions. Takentogether, the Bim deletion polymorphism has littleeffect on tumor aggressiveness or survival in earlyand/or primary NSCLC but may have an impact ontumor survival in metastatic NSCLC.

To the best of our knowledge, this is the firstinvestigation of the impact of Bim deletion polymor-phism on PRS in patients with NSCLC. Previousstudies10-12,14 demonstrated that the Bim deletionpolymorphism is a prognostic factor for progression-free survival in patients with stage IIIB or IV NSCLCwho received EGFR-TKIs and chemotherapy, al-thoughall but one study13 showednoobvious impacton OS. The reasons underlying these inconsistenciesregarding the impact of the Bim deletion polymor-phism in this and previous studies are unclear. How-ever, previous studies indicated that patients withNSCLC who had recurrence after curative surgeryhad a favorable prognosis compared with those withadvanced-stage disease at initial presentation.15,32

These results suggest that although both patientgroupscanbeclassifiedasadvancedNSCLC,biologiccharacteristics,suchasEGFR-TKIand/orchemother-apy treatment outcome, may be distinct.

The Bim polymorphism may be a novel germlinebiomarker for therapy resistance in patients withadvanced NSCLC. The presence of the Bim de-letion polymorphismmay be a negative indicationfor standard therapies, with the exception of sur-gery, because such patients are at risk of devel-oping aggressive cancer refractory to EGFR-TKI,chemotherapy, and radiotherapy. Thus, patientswith unresectable or recurrent NSCLCwho harbortheBimdeletionmaybenefit from treatmentwith aBH3-mimetic drug9,10 or histone deacetylase in-hibitor33 to overcome therapy resistance.

This study had several limitations. The first andmost important onewas the small sample size. Thesurvival analysis included heterogeneous patientbackgrounds. Because the subset analysesaccording to histology or therapy modality wereperformed by using small sample sizes, this studylacked statistical power, and further investigationis requiredwitha larger sample. Second, thiswasaretrospective study. Although the indications andtherapeutic strategies for recurrent disease werereviewed by the cancer board of our department,not all patients received treatment according to the

Table 5 – Therapeutic Background of Patients Who Received Cytotoxic Chemotherapy orRadiotherapy Alone

Type of Treatment

Bim Status

PWild TypeNo. (%)

DeletionPolymorphism No. (%)

Chemotherapy 19 4

Platinum

Yes 12 (75.0) 4 (25.0) .273

No 7 (100.0) 0 (0.0)

Taxane

Yes 3 (75.0) 1 (25.0) 1.000

No 16 (84.2) 3 (15.8)

Radiation* 17 5

Conventional 15 (88.2) 4 (11.8) .637

Cyberknife 2 (66.7) 1 (33.3)

*Total dose average: patientswithwild-type sequence, 66.0Gy (range, 39-100Gy); patientswith deletionpolymorphism, 52.0 Gy (range, 30-104 Gy); P 5 .217.

A

0

Time (months)

1.0

0.8

0.6

0.4

0.2

12 24 36 48 60

Post

recu

rrenc

e Su

rviva

l (p

roba

bilit

y)EGFR-mutated NSCLC patients EGFR-mutated NSCLC patientstreated With EGFR-TKI (n = 21)treated With EGFR-TKI (n = 21)WT Bim : 38.1 monthsWT Bim : 38.1 monthsBim deletion : 23.2 monthsBim deletion : 23.2 monthsLog-rank Log-rank P = .007= .007Hazard ratio 6.70Hazard ratio 6.70(95% CI, 1.34 to 33.5)(95% CI, 1.34 to 33.5)

EGFR-mutated NSCLC patients treated With EGFR-TKI (n = 21)WT Bim : 38.1 monthsBim deletion : 23.2 monthsLog-rank P = .007Hazard ratio 6.70(95% CI, 1.34 to 33.5)

B

0

Time (months)

1.0

0.8

0.6

0.4

0.2

12 24 36 48 60

Post

recu

rrenc

e Su

rviva

l (p

roba

bilit

y)

Patients treated withchemotherapy alone (n = 23)WT Bim: 18.5 monthsBim deletion: 6.2 monthsP = .003Hazard ratio 6.34(95% CI, 1.55 to 25.97)

Patients treated with radiotherapy alone (n = 22)WT Bim: 26.9 monthsBim deletion: 11.4 monthsLog-rank P = .046Hazard ratio 3.02(95% CI, .966 to 9.415)

C

0

Time (months)

1.0

0.8

0.6

0.4

0.2

12 24 36 48 60

Post

recu

rrenc

e Su

rviva

l (p

roba

bilit

y)

Figure 4 –

Kaplan-Meier curves forpostrecurrence survival(PRS) according totherapeutic modality. (A)PRS in patients with EGFR-mutated non–small-celllung cancer (NSCLC) whoreceived epidermal growthfactor receptor-tyrosinekinase inhibitors (EGFR-TKIs) according towild-type(WT) Bim (n 5 18) or theBimdeletion polymorphism(n5 3). (B) PRS in patientstreated with cytotoxicchemotherapy aloneaccording to wild-type Bim(n5 19) or theBimdeletionpolymorphism (n 5 4). (C)PRS in patients treated withradiotherapy aloneaccording to wild-type Bim(n5 17) or theBimdeletionpolymorphism (n 5 5).



same standard. A prospectivemulticenter study isrequired to determine the clinical significance ofthe Bim deletion polymorphism with regard totherapy for advanced and recurrent NSCLC. Fi-nally, this polymorphism is observed only in Asianpopulations. Even if the significance of the Bimdeletion polymorphism is validated, the resultswould not provide any benefit to non-Asian pa-tients with NSCLC.

In conclusion, the Bim deletion polymorphismwas an indicator of poor RFS and PRS in patientswith recurrence after complete resection and is

consequently an independent unfavorable prog-nostic factor for OS in all patients with NSCLC whoreceived complete resection. The polymorphismwas associated with tumor aggressiveness andtherapy resistance in metastatic disease. If vali-dated, these results suggest that the Bim poly-morphismmaybeabiomarker of poor outcome formultimodal therapies in treating recurrent or ad-vanced NSCLC in the Asian population.



Conception and design: Kimihiro ShimizuCollection and assembly of data: Jun Atsumi, Yoichi Ohtaki,Seiichi Kakegawa, Toshiteru Nagashima, Yasuaki Enokida,KaiObayashi, Yoshiaki Takase, OsamuKawashima,MitsuhiroKamiyoshihara, Masayuki Sugano, Takashi Ibe, Hitoshi Igai,Izumi TakeyoshiDataanalysis and interpretation: JunAtsumi,KimihiroShimizu,Yoichi Ohtaki, Kyoichi Kaira, Seshiru NakazawaManuscript writing: All authorsFinal approval of manuscript: All authors


The following represents disclosure information provided byauthors of this manuscript. All relationships are consideredcompensated. Relationships are self-held unless noted. I 5Immediate Family Member, Inst 5 My Institution. Relation-ships may not relate to the subject matter of this manuscript.For more information about ASCO’s conflict of interest policy,please refer to www.asco.org/rwc or jco.ascopubs.org/site/ifc.

Jun AtsumiNo relationship to disclose

Kimihiro ShimizuNo relationship to disclose

Yoichi OhtakiNo relationship to disclose

Kyoichi KairaNo relationship to disclose

Seiichi KakegawaNo relationship to disclose

Toshiteru NagashimaNo relationship to disclose

Yasuaki EnokidaNo relationship to disclose

Seshiru NakazawaNo relationship to disclose

Kai ObayashiNo relationship to disclose

Yoshiaki TakaseNo relationship to disclose

Osamu KawashimaNo relationship to disclose

Mitsuhiro KamiyoshiharaNo relationship to disclose

Masayuki SuganoNo relationship to disclose

Takashi IbeNo relationship to disclose

Hitoshi IgaiNo relationship to disclose

Izumi TakeyoshiNo relationship to disclose

ACKNOWLEDGMENTWethankYurihaIwata,MasakiShinohara,RyosukeKonuma,andNaoKobayashi, GunmaUniversity, for their technical assistance.

AFFILIATIONS

Jun Atsumi, Kimihiro Shimizu, Yoichi Ohtaki, Kyoichi Kaira, Toshiteru Nagashima, Yasuaki Enokida, Seshiru Nakazawa, KaiObayashi, Yoshiaki Takase,Masayuki Sugano, and Izumi Takeyoshi, GunmaUniversity Graduate School ofMedicine,Maebashi,Gunma; Seiichi Kakegawa and OsamuKawashima, National Hospital Organization Nishi-GunmaHospital, Shibukawa, Gunma;and Mitsuhiro Kamiyoshihara, Takashi Ibe, and Hitoshi Igai, Maebashi Red Cross Hospital, Maebashi, Gunma, Japan

REFERENCES1. Siegel R, Naishadham D, Jemal A: Cancer statistics, 2012. CA Cancer J Clin 62:10-29, 2012

2. Sugimura H, Nichols FC, Yang P, et al: Survival after recurrent nonsmall-cell lung cancer after complete pulmonaryresection. Ann Thorac Surg 83:409-417, 2007

3. Kelsey CR, Marks LB, Hollis D, et al: Local recurrence after surgery for early stage lung cancer: An 11-year experiencewith 975 patients. Cancer 115:5218-5227, 2009







4. Maemondo M, Inoue A, Kobayashi K, et al: Gefitinib or chemotherapy for non-small-cell lung cancer with mutatedEGFR. N Engl J Med 362:2380-2388, 2010

5. Mok TS, Wu YL, Thongprasert S, et al: Gefitinib or carboplatin-paclitaxel in pulmonary adenocarcinoma. N Engl J Med361:947-957, 2009

6. Ohtaki Y, Shimizu K, Kakegawa S, et al: Postrecurrence survival of surgically resected pulmonary adenocarcinomapatients according to EGFR and KRAS mutation status. Mol Clin Oncol 2:187-196, 2014

7. Katayama T, Matsuo K, Kosaka T, et al: Effect of gefitinib on the survival of patients with recurrence of lung ade-nocarcinoma after surgery: A retrospective case-matching cohort study. Surg Oncol 19:e144-e149, 2010

8. Takenaka T, Takenoyama M, Yamaguchi M, et al: Impact of the epidermal growth factor receptor mutation status onthe post-recurrence survival of patients with surgically resected non-small-cell lung cancer. Eur J Cardiothorac Surg47:550-555, 2015

9. Cragg MS, Kuroda J, Puthalakath H, et al: Gefitinib-induced killing of NSCLC cell lines expressing mutant EGFRrequires BIM and can be enhanced by BH3 mimetics. PLoS Med 4:1681-1690, 2007

10. Ng KP, Hillmer AM, Chuah CT, et al: A common BIM deletion polymorphismmediates intrinsic resistance and inferiorresponses to tyrosine kinase inhibitors in cancer. Nat Med 18:521-528, 2012

11. Zhao M, Zhang Y, Cai W, et al: The Bim deletion polymorphism clinical profile and its relation with tyrosine kinaseinhibitor resistance in Chinese patients with non-small cell lung cancer. Cancer 120:2299-2307, 2014

12. Isobe K, Hata Y, Tochigi N, et al: Clinical significance of BIM deletion polymorphism in non-small-cell lung cancer withepidermal growth factor receptor mutation. J Thorac Oncol 9:483-487, 2014

13. Lee JH, Lin YL, Hsu WH, et al: Bcl-2-like protein 11 deletion polymorphism predicts survival in advanced non-small-cell lung cancer. J Thorac Oncol 9:1385-1392, 2014

14. Lee JK, Shin JY, Kim S, et al: Primary resistance to epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors(TKIs) in patients with non-small-cell lung cancer harboring TKI-sensitive EGFRmutations: An exploratory study. AnnOncol 24:2080-2087, 2013

15. Sekine I, Nokihara H, Yamamoto N, et al: Comparative chemotherapeutic efficacy in non-small cell lung cancerpatients with postoperative recurrence and stage IV disease. J Thorac Oncol 4:518-521, 2009

16. TravisWD, Colby TV, Corrin B, et al: Histological Typing of Lung and Pleural Tumors (ed 3). Berlin, Germany, Springer-Verlag, 1999

17. Miyamae Y, Shimizu K, Mitani Y, et al: Mutation detection of epidermal growth factor receptor and KRAS genes usingthe smart amplification process version 2 from formalin-fixed, paraffin-embedded lung cancer tissue. J Mol Diagn 12:257-264, 2010

18. Ebi H, Oze I, Nakagawa T, et al: Lack of association between the BIM deletion polymorphism and the risk of lungcancer with and without EGFR mutations. J Thorac Oncol 10:59-66, 2015

19. Ong ST, Chuah CT, Ko TK, et al: Reply: The BIM deletion polymorphism cannot account for intrinsic TKI resistance ofChinese individuals with chronic myeloid leukemia. Nat Med 20:1090-1091, 2014

20. Costa C, Molina MA, Drozdowskyj A, et al: The impact of EGFR T790M mutations and BIM mRNA expression onoutcome in patients with EGFR-mutant NSCLC treated with erlotinib or chemotherapy in the randomized phase IIIEURTAC trial. Clin Cancer Res 20:2001-2010, 2014

21. Li R, Moudgil T, Ross HJ, et al: Apoptosis of non-small-cell lung cancer cell lines after paclitaxel treatment involves theBH3-only proapoptotic protein Bim. Cell Death Differ 12:292-303, 2005

22. Tan TT, Degenhardt K, Nelson DA, et al: Key roles of BIM-driven apoptosis in epithelial tumors and rational che-motherapy. Cancer Cell 7:227-238, 2005

23. Wang J, Zhou JY, Wu GS: Bim protein degradation contributes to cisplatin resistance. J Biol Chem 286:22384-22392,2011

24. Yang JY, Xia W, Hu MC: Ionizing radiation activates expression of FOXO3a, Fas ligand, and Bim, and induces cellapoptosis. Int J Oncol 29:643-648, 2006

25. Hein AL, Ouellette MM, Yan Y: Radiation-induced signaling pathways that promote cancer cell survival (review). Int JOncol 45:1813-1819, 2014

26. Toulany M, Lee KJ, Fattah KR, et al: Akt promotes post-irradiation survival of human tumor cells through initiation,progression, and termination of DNA-PKcs-dependent DNA double-strand break repair. Mol Cancer Res 10:945-957,2012

27. Dai DL, Wang Y, Liu M, et al: Bim expression is reduced in human cutaneous melanomas. J Invest Dermatol 128:403-407, 2008

28. Zantl N, Weirich G, Zall H, et al: Frequent loss of expression of the pro-apoptotic protein Bim in renal cell carcinoma:Evidence for contribution to apoptosis resistance. Oncogene 26:7038-7048, 2007

29. Sinicrope FA, Rego RL, Okumura K, et al: Prognostic impact of bim, puma, and noxa expression in human coloncarcinomas. Clin Cancer Res 14:5810-5818, 2008



30. Sakakibara-Konishi J, Oizumi S, Kikuchi J, et al: Expression of Bim, Noxa, and Puma in non-small cell lung cancer.BMC Cancer 12:286, 2012

31. Merino D, Best SA, Asselin-Labat ML, et al: Pro-apoptotic Bim suppresses breast tumor cell metastasis and is a targetgene of SNAI2. Oncogene 34:3926-3934, 2015

32. Yoshioka H, Mitsudomi T, Morita S, et al: Final overall survival results of WJTOG 3405, a randomized phase 3 trialcomparing gefitinib (G) with cisplatin plus docetaxel (CD) as the first-line treatment for patients with non-small cell lungcancer (NSCLC) harboringmutations of the epidermal growth factor receptor (EGFR). J Clin Oncol 32, 2014 (suppl 5s;abstr 8117)

33. Nakagawa T, Takeuchi S, Yamada T, et al: EGFR-TKI resistance due to BIM polymorphism can be circumvented incombination with HDAC inhibition. Cancer Res 73:2428-2434, 2013



originalresearch

Tuberculosis Diagnosis DelayingTreatment of Cancer: Experience From aNew Oncology Unit in Blantyre, Malawi

abstract

Purpose Malawi is a low-income country in sub-Saharan Africa with limited health care infrastructure andhigh prevalance of HIV and tuberculosis. This study aims to determine the characteristics of patientspresenting to Queen Elizabeth Central Hospital Oncology Unit, Blantyre, Malawi, who had been treated fortuberculosis before they were diagnosed with cancer.

Methods Clinical data on all patients presenting to the oncology unit at Queen Elizabeth Central Hospitalfrom 2010 to 2014 after a prior diagnosis of tuberculosis were prospectively recorded, and a descriptiveanalysis was undertaken.

Results Thirty-four patients who had been treated for tuberculosis before being diagnosed with cancer wereidentified between 2010 and 2014, which represents approximately 1% of new referrals to the oncology unit.Forty-onepercentofpatientswereHIVpositive.Meandurationof tuberculosis treatmentbeforepresentation to theoncology unit was 3.6 months. The most common clinical presentation was a neck mass or generalizedlymphadenopathy.Lymphomawasthemostcommonmalignancy thatwassubsequentlydiagnosed in23patients.

Conclusion Misdiagnosis of cancer as tuberculosis is a significant clinical problem in Malawi. This studyunderlines the importance of closelymonitoring the response to tuberculosis treatment, being aware of thepossibility of a cancer diagnosis, and seeking a biopsy early if cancer is suspected.

J Glob Oncol 2. © 2016 by American Society of Clinical Oncology Licensed under the Creative Commons Attribution 4.0 License.

INTRODUCTION

Malawi is a low-income country within sub-SaharanAfrica and thus has a low number of trained medicalpersonnel.1 Outside the main government centralhospitals, most health care is delivered by nursingand clinical officer staff. Resources are scarce, andthere are high levels of HIV, with a national seropre-valence rate of 10.0% in adults age 15 to 49 years(2014 data).2 Although the country has mounted aneffectivescale-upprogramofantiretroviral therapy, therates of tuberculosis (156 per 100,000)3 and AIDS-related cancers, particularly lymphomas, are high.Cancer incidence in Malawi is estimated at 55.5 per100,000 in males and 68.8 per 100,000 in females(age-standardized rates), and themost commoncan-cer sites are Kaposi’s sarcoma, esophageal cancer,non-Hodgkin lymphoma, cervical cancer, and breastcancer.4Age-standardized incidenceof non-Hodgkinlymphomaisreportedat2.3per100,000inmalesand1.9 per 100,000 in females, considerably lower thanthe incidence of tuberculosis. Since the oncology unitatQueenElizabethCentralHospital (QECH)opened in2010, it has registeredmore than 4,000 new patientswith cancer, with diagnoses reflecting the distribution

of cancer inMalawi.4 Several of the cancers are AIDSdefining, theproportionofpatientswithcancerwhoareHIV positive is high (44% of new patients at theoncology unit were recorded as HIV positive in2013 and 2014 combined), and some are co-infectedwith tuberculosis.Although tuberculosis ratesin Malawi are reported by WHO to be lower than insome surrounding countries and are definitely drop-ping, they remain high, with tuberculosis treatmentoften based on clinical diagnosis alone.

The oncology team has been aware of somepatients presenting with malignancy who havebeen erroneously diagnosed and treated for tu-berculosis, thus delaying cancer care. The unithas been prospectively recording informationabout such instances, and we report on this.

METHODS

All patients presenting to the oncology unit atQECH from 2010 to 2014 were assessed by either aclinical officer (Y.J.), a consultant oncologist (L.P.L.M.),or both. Patients who had an erroneous tuberculosisdiagnosis that delayed their cancer diagnosis were

Leo Peter LockieMasamba

Yankho Jere

Ewan Russell StewartBrown

Dermot Robert Gorman

LeoPeter LockieMasambaand Yankho Jere, QueenElizabeth CentralHospital, Blantyre,Malawi; and Ewan RussellStewart Brown andDermotRobert Gorman, NationalHealth Service Lothian,Western General Hospital,Edinburgh, Scotland.

Authors’ disclosures ofpotential conflicts ofinterest and contributionsare found at the end of thisarticle.Corresponding author:Dermot Robert Gorman,FFPH, National HealthService Lothian, 2-4Waterloo Place, EdinburghEH1 3EG, UnitedKingdom; e-mail: [email protected]


© 2016 by American Society of Clinical Oncology Licensed under the Creative Commons Attribution 4.0 License.





identified, and clinical data that included age, HIVstatus, clinical presentation, and type of malignancywereprospectively recordedandenteredontoanExcelspreadsheet. A descriptive analysis of the data wasundertaken. Ethics approval was gained through theMalawi Health Sciences Research Board.

RESULTS

Thirty-four patients who had been treated for tu-berculosis before being diagnosed with cancerwere identified between 2010 and 2014 (sevenin 2010, nine in 2011, 11 in 2012, five in 2013, twofromJanuary throughMarch2014; Table 1). Forty-one percent of patients were HIV positive.

Mean duration of tuberculosis treatment beforeoncology presentation was 3.6 months. The meanresultantdelay incancerdiagnosiswas5.4months.This was slightly longer for men (5.9 months) thanfor women (4.5 months).

Many patients had a constellation of signs and symp-toms on presentation, including prominent neckmasses, fever, malaise, weight loss, cough, and ab-dominal pain (Table 2).Mean hemoglobinwas 9.2 g/dl.

Misdiagnoses of tuberculosis were predominantlyclinical (17 instances) but were often supported bychest x-ray (seven), other x-ray (two), ultrasoundscan(one), fine-needleaspirate(one),magneticresonanceimaging (one), and cerebrospinal fluid analysis (one).The most common site for misdiagnosis of tubercu-losis was lymph nodes (Table 3).

The eventual cancer diagnosis was confirmed byhistology or cytology in 33 of the 34 patients. Thesingle patient with Kaposi’s sarcoma had a clinicaldiagnosis. The most common diagnosis was non-Hodgkin lymphoma followed by Hodgkin lym-phoma (Table 4).

Treatment was possible for many patients, and achemotherapy regimenwas offered to 26patients.Four patients were not eligible for chemotherapy,two were given steroids, one patient had surgerycombined with chemotherapy, and treatment wasnot recorded for one patient. The regimen used forthe majority of patients was cyclophosphamide,doxorubicin, vincristine, and prednisolone.

DISCUSSION

This study confirms that delay in diagnosing cancercausedbyprevious incorrectdiagnosisand treatmentof tuberculosis is an important clinical problem inMalawi. Our figure of 34misdiagnosed patients since2010 represents approximately 1% of the oncologypatients that presented to the QECH Oncology Unit.This misdiagnosis is well understood in the literature,notably in lymphomas5 and the lungs.6 In our series,the most common malignancies that were misdiag-nosed were lymphoma followed by lung cancer. Thedelay in treatment in this series was 5.4 months, andthis study reinforces the concerns raised about in-appropriate tuberculosis care leading to delayed can-cer diagnosis in a secondMalawian central hospital.7

Therefore, cancer treatment for our patients oftenstarted at a later clinical stage in which outcomesmay have been compromised.

Malawi has a large number of patients and fewstaff, particularly in rural areas with limited inves-tigative capacity; as a consequence, the diagnosisof tuberculosis is sometimes made on clinicalgrounds alone. This contributed to misdiagnosis

Table 1 – Demographic Characteristics and HIV Status of Patients

Characteristic

Male(n = 23)

Female(n = 11)

No. % Age (years) No. % Age (years)

Age, years 32.7* 36.8

HIV positive 6 26 27.2* 8 73 33.9*

HIV negative 16 69 33.4* 2 18 55.5*

HIV status unknown 1 53 1 23

*Mean

Table 2 – Symptoms Reported on Presentation to Queen Elizabeth Central HospitalOncology Unit

SymptomsNo. of Instances in Which Sign

or Symptom Was Noted

Neck masses 8

Plus respiratory 4

Plus abdominal 2

Generalized lymphadenopathy 6

Primarily respiratory or chest 4

Primarily neurologic

Back pain 2

Peripheral neuropathy 1

Lower limb weakness 1

Headaches 1

Spinal cord compression 1

Abdominal 3

Detail missing 1

Table 3 – Site of Misdiagnosed Tuberculosis

Site of Misdiagnosed Tuberculosis No.

Abdominal 2

Disseminated 2

Tuberculosis lymph nodes 16

Pulmonary 9

Spinal 4



in 17 of 35 patients and is a common challenge forhealth care services in low-income countries, par-ticularly for cancers that share features of pre-sentation with tuberculosis.5

Themajority of symptoms and signs described arecommon between tuberculosis and malignancy,especially the lymphomas. Given the limited in-vestigative capacity and the commonpresentationof tuberculosis, the misdiagnoses are not unex-pected, and similar findings have been found andsimilar explanations for the problem have beengiven in South Africa and elsewhere in Malawi.5,7

The empirical treatment of tuberculosis is common inMalawi, but when this is undertaken for extrapulmo-naryor smear-negative tuberculosis, closemonitoringfor response is important. Patients presenting withlymphadenopathy should be considered for biopsyreferral because there is a high likelihood of cancerdiagnosis: 53% and 35% in two Malawian series.7,8

Where biopsy and histopathologic facilities areavailable, a clinical diagnosis of tuberculosis forlymphadenopathy should be discouraged.

Follow-up of patients with tuberculosis in aresource-poor setting is notoriously difficult. Here,

although the treatmentwas incorrect, this sample ofpatients often had complete or almost completecourses of tuberculosis treatment over severalmonths under some form of clinical supervision,mostly by nursing and clinical officer staff. Thiscreatesanopportunity to interveneandoffer trainingto health care staff and also provides an opportunityfor health care institutions to improve their monitor-ing of response to tuberculosis treatment. Our find-ings raise the concern that patients with potentiallytreatable cancers may miss the opportunity to haveaccess to cancer treatment because of misdiagno-sis and emphasize the importance of more cancerawareness training for all health care staff.

Ensuring the microbiologic diagnosis of tubercu-losis, promoting biopsies of patients with lymph-adenopathy, andbeingmore alert to thepossibilityof a cancer diagnosis in peoplewhowere originallydiagnosed as having tuberculosis but who do notimprove with treatment are all key to improvingcare for this group of patients. When the diagnosisis reviewed and cancer is correctly diagnosed andthen treated promptly, a successful outcome forthe patient is more likely. We recommend thathealth care workers have a low threshold for re-ferring patients for investigation for malignancy ifempirical tuberculosis treatment doesnot lead to aclinical response within 4 weeks. The MalawianMinistry of Health National Action Plan for Pre-vention and Management of Non-CommunicableDisease in Malawi 2012-20169 plans initiatives toimprove cancer knowledge in the general popu-lation and improve cancer education for healthcare providers. We hope that these and otherinitiatives will help improve outcomes for patientswith cancer in Malawi.

DOI: 10.1200/JGO.2015.000299

Published online on jgo.ascopubs.org on January 13, 2016.


Conception and design: Leo Peter Lockie Masamba, YankhoJere, Dermot Robert Gorman

Collection and assembly of data: Leo Peter Lockie Masamba,Yankho Jere, Ewan Russell Stewart Brown

Data analysis and interpretation: Leo Peter Lockie Masamba,Ewan Russell Stewart Brown, Dermot Robert Gorman


AUTHORS’ DISCLOSURES OFPOTENTIAL CONFLICTS OF INTEREST

Tuberculosis Diagnosis Delaying Treatment of Cancer:Experience From a New Oncology Unit in Blantyre, MalawiThe following represents disclosure information provided byauthors of this manuscript. All relationships are considered

compensated. Relationships are self-held unless noted. I =Immediate Family Member, Inst = My Institution. Relation-ships may not relate to the subject matter of this manuscript.For more information about ASCO’s conflict of interest policy,please refer to www.asco.org/rwc or jco.ascopubs.org/site/ifc.

Leo Peter Lockie MasambaNo relationship to disclose

Yankho JereNo relationship to disclose

Ewan Russell Stewart BrownTravel, Accommodations, Expenses: Bristol-Myers Squibb

Dermot Robert GormanNo relationship to disclose

Table 4 – Cancer Diagnoses in Patients Treated as HavingTuberculosis (No. HIV positive)

Diagnosis Female Male Total

Adenocarcinoma/carcinoma 1 (1) 2 (0) 3

Breast 1 (0) 0 1

Kaposi’s sarcoma 0 1 (1) 1

Esophagus 0 1 (0) 1

Hodgkin lymphoma 3 (0) 7 (0) 10

Non-Hodgkin lymphoma 5 (5) 6 (2) 11

Burkitt lymphoma 1 (1) 1 (1) 2

Multiple myeloma 0 2 (1) 2





http://www.jco.ascopubs.org/site/ifc


REFERENCES1. WHO Statistical Information System: World Health Statistics 2010. Geneva, Switzerland, World Health Organization,

2010

2. The Joint United Nations Programme on HIV/AIDS (UNAIDS): Countries: Malawi—HIV and AIDS estimates (2014).http://www.unaids.org/en/regionscountries/countries/malawi

3. World Health Organization: Tuberculosis (TB): Tuberculosis country profiles—Malawi 2013. http://www.who.int/tb/country/data/profiles/en/

4. Msyamboza KP, Dzamalala C, Mdokwe C, et al: Burden of cancer in Malawi; common types, incidence and trends:National population-based cancer registry. BMC Res Notes 5:149, 2012

5. Puvaneswaran B and Shoba B: Misdiagnosis of tuberculosis in patients with lymphoma. S Afr Med J 103:32-33, 2013

6. Bhatt M, Kant S, and Bhaskar R: Pulmonary tuberculosis as differential diagnosis of lung cancer. South Asian J Cancer1:36-42, 2012

7. Mabedi C, Kendig C, Liomba G, et al: Causes of cervical lymphadenopathy at Kamuzu Central Hospital. Malawi Med J26:16-19, 2014

8. Mtonga P, Masamba L, Milner D, et al: Biopsy casemix and diagnostic yield at a Malawian central hospital. Malawi MedJ 25:62-64, 2013

9. National Action Plan for Prevention and Management of Non-Communicable Disease in Malawi 2012-2016. Lilongwe,Malawi, Ministry of Health, 2013


http://www.unaids.org/en/regionscountries/countries/malawi

http://www.who.int/tb/country/data/profiles/en/

http://www.who.int/tb/country/data/profiles/en/


originalresearch

Lung Cancer Survival Among ChineseAmericans, 2000 to 2010

abstract

Purpose Despite being the leading cause of cancer death, no prior studies have characterized survivalpatterns among Chinese Americans diagnosed with lung cancer. This study was conducted to identifyfactors associated with survival after lung cancer in a contemporary cohort of Chinese patients with lungcancer.

Methods The study design is a prospective descriptive analysis of population-based California CancerRegistry data.Multivariable Cox proportional hazardsmodelswere used to estimate hazard ratios (HRs) foroverall mortality. Participants were Chinese American residents diagnosedwith first primary invasive lungcancer from 2000 to 2010 (2,216 men and 1,616 women).

Results Among Chinese men, decreased mortality was associated with care at a National Cancer Institutecancer center (HR, 0.85; 95% CI, 0.73 to 0.99) and adenocarcinoma versus small-cell carcinoma (HR,0.78; 95% CI, 0.65 to 0.92). Women had better survival compared with men (HR, 0.82; 95% CI, 0.75 to0.89),withmortality associatedwith nevermarried versuscurrentlymarried status (HR,1.36;95%CI,1.11to 1.66), lower versus higher neighborhood socioeconomic status (HR, 1.38; 95% CI, 1.10 to 1.72comparing lowest to highest quintile), care at a cancer center (HR, 0.80; 95% CI, 0.67 to 0.96), andsquamous cell relative to small-cell carcinoma (HR, 1.60; 95% CI, 1.04 to 2.48).

Conclusion Focusing on factors associated with marital status, community socioeconomic status, andcharacteristics unique to National Cancer Institute–designated cancer centers may help to identify po-tential strategies for improving the length of survival for Chinese Americans.

J Glob Oncol 2. © 2016 by American Society of Clinical Oncology License under the Creative Commons Attribution 4.0 License

INTRODUCTION

Chinese Americans are the largest Asian groupin the United States, with 4.0 million ChineseAmericans counted in the 2010 Census.1 InCalifornia, the most populous US state for Chi-nese Americans, this population increased 30%over a decade from 2000 to 2010, numberingnearly 1.5 million.2 Lung cancer is the secondand fourth most common cancer among USChinese men and women, respectively,3 andthe most common cause of cancer death forboth men and women, followed by prostate andcolorectal cancers for men and breast and co-lorectal cancer for women.4Wepreviously docu-mented differences across multiple AsianAmerican ethnic groups in survival after lungcancer among patients in the California CancerRegistry (CCR)5 and among a series of femalenever-smoker patients.6 However, no prior stud-ies, to our knowledge, have characterized thesurvival patterns specific for Chinese Ameri-cans, particularly as the biology of the diseasemay be unique among this population, with a

higher incidence of epidermal growth factor re-ceptor (EGFR) tyrosine kinase domain–activating mutations7 and, among Chinesewomen, a majority of lung cancers presentingamong never-smokers.6 Considering the highburden of disease in Chinese Americans, anexamination of survival patterns and prognosticfactors, with attention to both clinicopathologicand sociodemographic factors, may informstrategies to improve survival and potentiallyearly detection by identifying subgroups withpoor survival who may benefit from targetedscreening efforts.

To provide insight into the prognostic factors forlung cancer among the growing Chinese Americanpopulation, we used population-based CCR dataenhanced with information regarding immigrantstatus and neighborhood-level information on so-cioeconomic status (SES) and residence in ethnicenclaves to examine patterns in lung cancer sur-vival among Chinese persons in California, the USstatewith the largestChinesepopulation (one thirdof the US Chinese population).8

Scarlett Lin Gomez

Juan Yang

Shih-Wen Lin

Margaret McCusker

Alan Sandler

Manali Patel

Iona Cheng

Heather A. Wakelee

Christina A. Clarke

Scarlett Lin Gomez, JuanYang, Iona Cheng, andChristina A. Clarke, CancerPrevention Institute ofCalifornia, Fremont;Scarlett Lin Gomez andChristina A. Clarke,Stanford School ofMedicine; Scarlett LinGomez, Manali Patel, IonaCheng,HeatherA.Wakelee,and Christina A. Clarke,Stanford Cancer Institute,Stanford; and Shih-WenLin, Margaret McCusker,and Alan Sandler,Genentech, South SanFrancisco, CA.Authors’ disclosures ofpotential conflicts ofinterest and contributionsare found at the end of thisarticle.Corresponding author:Scarlett Lin Gomez, PhD,Cancer PreventionInstitute of California,2201 Walnut Ave, Ste300, Fremont, CA 94538;e-mail: [email protected].


© 2016 by American Society of Clinical Oncology License under the Creative Commons Attribution 4.0 License



METHODS

Case Selection

We obtained data for all first primary invasive lungcancers (International Classification of Diseases[ICD] for Oncology, third edition, site codes, C34.0to C34.9, excluding histologic codes 9050 to9055, 9140, and 9590 to 9992) among ChineseAmerican residents of California during the periodbetween January 1, 2000, and December 31,2010, from the CCR (comprised of four registries[San Francisco Bay Area, San Jose/Monterey, LosAngeles, and Greater California] within the Na-tional Cancer Institute [NCI] SEER program; n =4,537). We excluded patients diagnosed withoutmicroscopic confirmation (n = 390), those diag-nosed at autopsy or via death certificate (n = 59),and patients with survival time less than 30 dayswho did not undergo any first course of treatment(n = 256). The final study cohort included 2,216male and 1,616 female patients with lung cancer(total, N = 3,832). Other patient and clinical char-acteristics obtained from CCR data include ageand year of diagnosis, birthplace, sex, residentialaddress and stage at diagnosis, histologic subtype(coded using ICD for Oncology, third edition, his-tology codes as follows: small-cell carcinoma,8041 to 8045 and 8246; squamous cell carci-noma, 8051, 8052, 8070 to 8078, 8083, and8084; adenocarcinoma, 8050, 8140 to 8147,8201, 8230, 8250 to 8255, 8260, 8263, 8290,8310, 8320, 8323, 8220, 8350, 8441, 8460,8470, 8471, 8480, 8481, 8490, 8500, 8503,8507, 8550, and 8570 to 8576; large-cell carci-noma, 8011 to 8015, 8082, and 8123; and non–small-cell lung cancer [NSCLC], not otherwisespecified, 8010, 8020 to 8022, 8030 to 8035,8046, 8094, 8120, 8130, 8170, 8200, 8240 to8249, 8340, 8430, 8525, 8551, 8560, 8562,8580, 8940, 8972, and 8980), and first courseof treatment (extent of surgical resection, chemo-therapy [yes/no], and radiation [yes/no]). All dataused in this analysis came from the CCR. Smokingstatus is not collected by the cancer registry.

Information on patient race and ethnicity fromcancer registry data is primarily based on infor-mation abstracted from hospital records and usu-ally self-reported by patients,9 but for a smallproportion of patients, race and ethnicity maybebasedonassumptionsor inferencesbyhospitalpersonnel from other patient data includingmaiden name, surname, birthplace, or death re-cords. Chinese ethnicity in cancer registry dataincludes Taiwanese. Because our previous stud-ies have shown that Asian patients in the CCRwith

unknown registry birthplace are more likely to beUSborn,10,11 random imputation of nativity (US orforeign born) for patients with unknown birthplacewould thus lead to an underestimate of US-bornpatients. To more accurately impute nativity, weapplied a statistical imputation method based onthe age at issue of Social Security number (SSN),using a crosswalk file provided by the Social Se-curity Administration that indicates the year ofissuance for each SSN sequence. By comparingthe age of SSN issue with self-reported birthplacein previously interviewed cancer patients (n =1,836) and based on maximization of the areaunder the receiver operating characteristic curveand confirmation with logistic regressionmodeling,patients receiving anSSNbefore age 25 yearswereconsidered US born, and those who had receivedan SSN at or after age 25 years were consideredforeign born. This age cut point resulted in 84%sensitivity and80%specificity for assigning foreign-born status across the Asian populations.12 For ourstudy, registry-basedbirthplacedatawereavailablefor 93%of theChinese patients (72% fromhospitalrecords and 21% from death certificates). Nativitywas imputed using the method described earlierfor approximately 7% of patients without registrybirthplace information. For the remaining less than1% of patients for whom SSNs were missing orinvalid, we randomly assigned nativity based onthe overall sample’s joint distributions of race/ethnicity, sex, and age.

Patient residential address at diagnosis was geo-coded and assigned to a census block group,whichwas then linked toblockgroup–level censusmeasures. Neighborhood SES is a composite in-dex developed previously from principal compo-nent analysis, incorporating information oneducation, occupation, employment, householdincome, poverty, and rent and house values fromthe Census 2000 Summary File (for patients di-agnosed from 2000 to 2005) and American Com-munitySurvey (ACS)2007 to2011data (applied topatients diagnosed from 2006 to 2010 becauseACS replaced thedecennial census long formafter2000).13,14 Ethnic enclave is defined as a neigh-borhood that maintains more Asian ethnic moresand norms and/or is ethnically distinct from itssurrounding area. It is characterized using a com-posite index based on the following four censusindicator variables: percentage of recent immi-grants, percentage of Asian/Pacific Islander(API) language-speaking households that werelinguistically isolated, percentage of API languagespeakers with limited English proficiency, andpercentage of API population.15 For patients



diagnosed during the period from 2000 to 2005,this information was derived from the summaryfiles of Census 2000; for patients diagnosed from2006 to 2010, we used the Census 2000 databecause the component variables are lacking orunreliable in theACS.BothneighborhoodSESandethnic enclavemeasureswereclassified intoquin-tiles based on distributions across California blockgroups.

Determination of Follow-Up and Vital Status

CCR routinely collects information onpatientswithcancer through active and passive follow-up untilconfirmation of their death using linkages to datafrom the diagnosing hospital, state and nationalvital statistics databases, and other data sources.Underlying causes of death, coded by ICD, 10thedition (used in the CCR for deaths starting in1999), were obtained from death certificates,and deaths assigned codes C34.0 to C34.9 (ICD,10thedition)were identifiedasbeinga result of lungcancer. Follow-up time for overall mortality wascomputed as the number of days between the dateofdiagnosisand the first occurrenceof the followingdates: date of death, date of last known contact, orend date of follow up (December 31, 2012).

Statistical Analysis

To describe overall survival time after lung cancerdiagnosis amongChineseAmericans inCalifornia,we estimated median all-cause survival amongpatients, overall and by demographic, neighbor-hood, and tumor factors by sex. To assess theindependent influence of patient, tumor, or treat-ment characteristics on survival and identifypossible prognostic factors, we conducted Coxproportional hazards multivariable regression bysex. We tested the proportional hazards assump-tion based on correlation test of time versus scaledSchoenfeld residuals. The assumption of propor-tional hazardswas violated for chemotherapy, andthe test statistic was marginally significant forSEER summary stage (local, regional, distant, orunknown). Therefore, hazard ratios (HRs) and95% CIs were computed using stratified Coxmultivariable models, with stratification on che-motherapy and SEER summary stage, whichallowed the baseline hazards within each modelto vary by the strata variable(s). All of the indepen-dent variables of interest in Table 1 that werestatistically significant at P , .10 in unadjustedmodels were included in the multivariable model,and covariates included in the final models in-cluded year of diagnosis, age, marital status, na-tivity, neighborhood SES, neighborhood ethnic

enclave, urban or rural region, surgery type, radi-ation, cancer center, and histologic subtype. Allstatistical analyses were performed using SASversion 9.3 (SAS Institute, Cary, NC). All statisticaltests were two-sided with an a = 0.05.

RESULTS

Patient Characteristics

Among Chinese American men in California di-agnosed with lung cancer between 2000 and2010, more than half (55.8%) were age 70 yearsor older at diagnosis (Table 1). The vast majority(90.5%) were foreign born (66.2% born in China,6.1% born in Taiwan, and 4.3% born in HongKong). Most were married (80.4%), were insuredonpublic insurance (45%onMedicaid,military, orother public insurance), lived in the highest twostatewide SES quintiles (23.5% in SES quintile 4and 26.9% in quintile 5), and lived in the mostAsian ethnic neighborhoods (73.9% in quintile 5).A small proportion of patients (11.7%) were re-ported by an NCI cancer center (includes com-prehensive and noncomprehensive designation).Most male patients (60.8%) were diagnosed withdistant disease, and nearly half (46.5%) werediagnosed with adenocarcinomas, although arelatively high proportion (23.4%) had NSCLC,not otherwise specified. Twenty-two percent ofpatients received any surgery, 38.9% receivedradiation, and 49.3% received chemotherapy.

Among Chinese American female patients, slightlymore than half were diagnosed at age 70 years orolder, and nearly 90% were foreign born. Relativeto Chinese American men, a considerably lowerproportion of women (58.4%) were married atdiagnosis, whereas nearly one third (30.8%)were previously married (separated, divorced, orwidowed). The distributions of primary health in-surance, neighborhood SES, and ethnic enclavewere similar to those for men. A small proportionof female patients (12.9%) received care at anNCI-designated cancer center. Although a slightlyhigherproportionof femalepatients, relative tomalepatients, were diagnosed at local stage (14.7% v13.6%, respectively), women also had a slightlyhigher rate of distant disease compared with men(63.3% v 60.8%, respectively). Of note, a consid-erably higher proportion of tumors were adenocar-cinomas in women versus men (65.5% v 46.5%,respectively), whereas women had lower rates ofsmall-cell and squamous cell histologies.

Survival

Table 2 lists themedian survival time and adjustedHRs for overall (all-cause)mortality after diagnosis



with lung cancer. Among Chinese Americanmen,the median survival time was 13.0 months (95%CI, 12.0 to 14.2 months) overall, with minor dif-ferences between US-born patients (median,12.5 months; 95% CI, 10.5 to 17.7 months)and foreign-born patients (median, 13.0 months;95% CI, 12.0 to 14.2 months). In multivariablehazard models, independent associations withbetter overall survival were found for receivingcare at an NCI-designated cancer center andspecific histologies including adenocarcinomaand large-cell carcinoma (although the latter as-sociation was of borderline statistical signifi-cance). Married men had somewhat lowermortality thanunmarriedmen, but this associationwas not statistically significant (HR, 1.22; 95%CI,0.98 to 1.51).

AmongChineseAmericanwomen, overallmediansurvival was higher than for men (18.7 months;95% CI, 17.1 to 20.6 months) overall. In a mul-tivariable model including both men and women,the HR comparing risk of death among women tomen was 0.82 (95% CI, 0.75 to 0.89; data notshown), suggesting that the female survival benefitwas not explained by other variables. In the mul-tivariable hazard model, higher mortality was in-dependently associated with never beingmarried,living in lower SES neighborhoods, receiving carefrom facilities other than NCI-designated cancercenters, and having squamous cell histology (rel-ative to small-cell histology).

DISCUSSION

In this analysis of survival among all ChineseAmerican patients with lung cancer in Californiafrom 2000 to 2010, we found that social factorssuch as birthplace, marital status, and SES wereimportant prognostic factors for womenbut less sofor men. Among Chinese American women, me-dian survival varied substantially, by as much as12 months among those living in the highest andlowest socioeconomic groups. We also found sig-nificantly higher survival among Chinese Ameri-can women compared with Chinese Americanmen, with a 5-month difference in median sur-vival. Although we did not have data on smokingstatus among the patients with cancer, prior stud-ies in Chinese Americans have shown that themajority of female patients are never-smokers (eg,70% in one study),6 whereas the majority of malepatients are current or former smokers (eg, 86% inthe same study).6 These major discrepancies insmoking history by sex suggest either that thesurvival differences we observed may be in part

Table 1 –Characteristics of ChineseAmericanPatientsWith LungCancer by Sex, California,2000 to 2010

Characteristic

% of Patients

x2 Test PMen

(n = 2,216)Women

(n = 1,616)

Age at diagnosis, years , .01

, 60 20.2 25.9

60-69 24.0 23.3

70+ 55.8 50.8

Nativity .09

US born 9.5 11.1

Foreign born 90.5 88.9

Marital status , .01

Married 80.4 58.4

Never married 6.9 8.7

Separated/divorced/widowed 10.3 30.8

Unknown 2.4 2.0

Health insurance (primary payer source) .04

Uninsured 2.4 1.5

Private 38.8 41.6

Public/Medicaid 44.5 42.8

Medicare 11.5 11.8

Military 0.5 0.1

Unknown 2.3 2.2

Neighborhood SES, quintile , .01

1 (low SES) 13.4 10.4

2 16.9 15.2

3 19.3 18.0

4 23.5 24.3

5 (high SES) 26.9 32.1

Neighborhood ethnic enclave, quintile .03

1 (least ethnic) 1.5 0.9

2 3.4 4.1

3 6.5 7.8

4 13.0 12.9

5 (most ethnic) 73.9 71.7

Unknown 1.6 2.7

Care at NCI cancer center .26

Yes 11.7 12.9

No 88.3 87.1

Stage .04

Local 13.6 14.7

Regional 20.4 18.3

Distant 60.8 63.3

Unknown 5.1 3.7

(Continued on following page)



attributable to our inability to measure smokingas a prognostic factor and/or that lung canceramong smokers is a different, perhaps more ag-gressive disease entity than lung cancer amongnever-smokers.7 The improved survival amongwomencomparedwithmenhas been consistentlyobserved and is thought to indicate sex as anindependent prognostic factor thatmay be relatedto differences in tumor molecular or biologic pro-file, drug metabolism, and/or DNA damage

susceptibility and repair capacity.16 Social factorsare suggested from the results to be potentiallyimportant prognostic factors after lung cancer di-agnosis, although more so for Chinese womenthan for Chinese men. Improved survival amongpatients who are married compared with thosewho are unmarried has been well documentedacross multiple cancer sites,17 other health out-comes, and overall mortality.18 This association isoften suggested to be attributed to greater socialsupport among patients with cancer who are mar-ried. However, we do not observe a difference insurvival when comparing married patients withthose who were previously married (ie, separated,divorced, widowed), suggesting that additionalfactorsmay be responsible for the highermortalityamong never-married patients that is distinct frompreviously married patients. The successivelyhigher mortality with lower neighborhood SESsuggests mechanisms related to access to healthcare and other resources. Although we were ableto account for health insurance and other tumorand treatment characteristics, we were not able toaccount for quality of care, detailed treatment,comorbidities, use of palliative care, and otherfactors that may mediate the association betweenSES and survival.

We found that among both male and femaleChinese American patients, the small proportionof patients receiving care at an NCI-designatedcancer center had improved survival comparedwith those receiving care at non–cancer centerfacilities. NCI-designated cancer centers are ac-ademic centers that are characterized by theircancer research, but with relevance to cancercare, they may also provide more state-of-the-artcancer care, integrated care and tumor boards,and access to clinical trials. We cannot discountthat unmeasured sociodemographic or clinicalpatient characteristics may confound these sur-vival patterns; for example, patients with EGFRmutations may more likely be referred to an NCIcancer center for treatment.

Interestingly, becausewe included small-cell lungcancers in our analysis, we were able to comparesurvival for NSCLC histologic subtypes with sur-vival for small-cell lung cancer, and we found that,as expected, mortality for adenocarcinoma waslower than that for small-cell cancer among Chi-nese American men; however, we did not find acomparable mortality difference among Chinesewomen. In contrast, we found markedly worsesurvival for squamous cell lung cancer thansmall-cell lung cancer among women. These re-sults should be interpreted with caution, however,

Table1 –CharacteristicsofChineseAmericanPatientsWith LungCancerbySex,California,2000 to 2010 (Continued)

Characteristic

% of Patients

x2 Test PMen

(n = 2,216)Women

(n = 1,616)

Histology , .01

Small-cell carcinoma 9.6 3.7

NSCLC

Squamous 15.0 5.4

Adenocarcinoma 46.5 65.5

Large cell 4.2 3.8

NSCLC, not otherwise specified 23.4 20.5

Other lung cancer 1.4 1.1

Surgery .07

None 77.9 74.3

Wedge resection 2.7 3.2

Lobectomy 19.0 22.2

Other surgery 0.4 0.4

Radiation , .01

No 61.1 67.3

Yes 38.9 32.7

Unknown 0 0.1

Chemotherapy .92

No 48.9 49.6

Yes 49.3 48.7

Unknown 1.8 1.7

Urbanicity (census tract level) .02

Rural 0.6 0.6

Town 0.2 0.1

City 6.3 6.8

Suburban metropolitan 48.9 52.3

Metropolitan 42.5 37.6

Unknown 1.5 2.6

Deaths, No. of patients

All cause 1,832 1,235 , .01

Lung cancer 1,551 1,051 , .01

Abbreviations: NCI, National Cancer Institute; NSCLC, non–small-cell lung cancer; SES, socioeconomicstatus.



Table2–MedianSurvival andAdjustedHRs forAll-CauseMortalityAmongChineseAmericansDiagnosedWithLungCancerbySex,California, 2000 to2010

Characteristic

Men Women

Median Survival,Months (95% CI) HR (95% CI)


Total 13.0 (12.0 to 14.2) — 18.7 (17.1 to 20.6) —

Nativity

US born 12.5 (10.5 to 17.7) 1.05 (0.89 to 1.25) 23.4 (16.2 to 37.3) 0.92 (0.74 to 1.14)

Foreign born 13.0 (12.0 to 14.2) 1.0 18.4 (16.7 to 20.4) 1.0

Marital status

Married 14.0 (12.7 to 15.3) 1.0 21.8 (18.8 to 24.3) 1.0

Never married 8.7 (6.6 to 10.8) 1.22 (0.98 to 1.51) 18.7 (13.0 to 23.0) 1.36* (1.11 to 1.66)

Separated/divorced/widowed 10.1 (7.7 to 13.4) 1.10 (0.93 to 1.29) 14.3 (11.9 to 16.5) 1.08 (0.94 to 1.25)

Unknown 12.7 (8.4 to 22.3) 0.82 (0.58 to 1.15) 25.0 (14.9 to 35.2) 1.05 (0.76 to 1.44)

Health insurance

Uninsured 10.5 (6.4 to 23.8) 1.10 (0.72 to 1.69) 16.4 (4.2 to 39.5) 1.04 (0.62 to 1.74)

Private 14.7 (12.9 to 16.8) 1.0 23.9 (20.1 to 27.3) 1.0

Public/Medicaid 11.8 (10.8 to 13.2) 0.94 (0.84 to 1.06) 16.3 (13.8 to 19.2) 0.92 (0.79 to 1.06)

Medicare 12.9 (9.1 to 16.9) 0.96 (0.82 to 1.13) 16.2 (11.9 to 18.8) 0.91 (0.73 to 1.12)

Military 13.7 (4.5 to 26.2) 1.72 (0.91 to 3.23) — 0.44* (0.31 to 0.62)

Unknown 12.5 (4.4 to 18.3) 1.19 (0.89 to 1.59) 24.8 (9.6 to 34.6) 1.07 (0.73 to 1.58)

Neighborhood SES, quintile

1 (low SES) 10.4 (8.3 to 13.3) 1.16 (0.98 to 1.38) 10.9 (8.4 to 15.1) 1.38* (1.10 to 1.72)

2 11.0 (8.9 to 13.2) 1.12 (0.95 to 1.32) 17.3 (13.1 to 21.7) 1.22* (1.01 to 1.47)

3 13.3 (11.2 to 16.9) 1.11 (0.95 to 1.28) 17.3 (13.5 to 23.4) 1.18 (0.98 to 1.41)

4 13.9 (11.7 to 16.2) 1.06 (0.93 to 1.22) 19.8 (16.7 to 22.9) 0.97 (0.82 to 1.15)

5 (high SES) 15.2 (12.8 to 17.9) 1.0 23.0 (19.7 to 26.7) 1.0

Ethnic enclave, quintile

1 (least ethnic) 12.8 (4.8 to 31.9) 1.0 22.4 (5.7 to 72.6) 1.0

2 15.7 (11.0 to 25.8) 0.92 (0.59 to 1.47) 19.6 (11.8 to 35.2) 1.03 (0.56 to 1.91)

3 14.1 (11.3 to 17.1) 1.14 (0.77 to 1.71) 19.0 (15.2 to 24.3) 1.12 (0.63 to 2.00)

4 14.3 (11.3 to 19.6) 1.11 (0.75 to 1.63) 17.9 (14.7 to 23.7) 1.22 (0.69 to 2.15)

5 (most ethnic) 12.6 (11.3 to 14.0) 1.13 (0.78 to 1.64 18.2 (16.3 to 20.5) 1.01 (0.58 to 1.77)

Unknown 10.5 (5.9 to 27.5) 2.42 (0.91 to 6.43) — —

Care at NCI-designated cancer center

Yes 22.4 (18.3 to 25.9) 0.85* (0.73 to 0.99) 34.2 (22.8 to 39.8) 0.80* (0.67 to 0.96)

No 12.1 (11.2 to 13.1) 1.0 17.6 (16.0 to 19.5) 1.0

Stage

Local 64.0 (44.5 to 88.6) — — —

Regional 27.1 (23.8 to 32.1) 52.3 (42.0 to 59.8)

Distant 8.0 (7.1 to 8.7) 11.5 (10.2 to 12.9)

Unknown 16.2 (12.7 to 21.3) 16.0 (13.3 to 26.0)

Tumor histology

Small-cell carcinoma 9.7 (8.0 to 11.0) 1.0 12.0 (9.1 to 16.8) 1.0

NSCLC

Squamous 13.5 (11.4 to 15.3) 0.91 (0.74 to 1.11) 10.6 (6.0 to 14.2) 1.60* (1.04 to 2.48)

Adenocarcinoma 19.2 (16.7 to 21.2) 0.78* (0.65 to 0.92) 25.0 (22.5 to 28.9) 1.00 (0.69 to 1.45)

(Continued on following page)



considering the high proportion of tumors classi-fied here as NSCLC, not otherwise specified (23%among men and 20% among women). The pro-portion of NSCLCs diagnosed with histology nototherwise specified has steadily declined overtime19 as a result of the availability of targetedtherapies for specific lung cancer histologies; themajority of these are likely adenocarcinomas.

The primary limitation in our study involves theabsence of cancer registry data on potentiallyimportant prognostic factors for lung cancer, in-cluding specific treatments, tumor geneticmarkers (such asEGFR andALKmutation status),smoking history, and comorbid conditions thataffect treatment decisions and survival time. Sexdifferences in EGFR mutations may well explainthe survival differencesbetweenChinesemenandwomen; however, it is unlikely that EGFR muta-tions would confound the associations betweenmarital status and SES with survival. Although it ispossible that our study results are biased as aresult of misclassification of race or ethnicity, prior

research shows minimal misclassification of Chi-nese ethnicity in cancer registry data.20

In summary, despite generally poor survival forlung cancer, our study did identify several non-clinical factors associated with lung cancer sur-vival among Chinese Americans, including sex,marital status, and SES. Focusing on factors thatdiffer between female married and unmarriedpatients (eg, greater social and/or instrumentalsupport or improved economic resources) andpatients who live in low versus high SES commu-nities (eg, greater socioeconomic resources, abil-ity to access and pay for treatments), as well ascharacteristics unique to NCI-designated cancercenters (eg, presence of tumor boards, access toclinical trials), may help to identify potential strat-egies for improving the length and quality of sur-vival for Chinese Americans after diagnosis of lungcancer.

DOI: 10.1200/JGO.2015.000539Published online on jgo.ascopubs.org on January 20, 2016.


Conception and design: Scarlett Lin Gomez, Shih-Wen Lin,Margaret McCusker, Christina A. Clarke

Financial support:Scarlett LinGomez, Shih-WenLin,MargaretMcCusker, Christina A. Clarke

Administrative support: Scarlett Lin GomezProvision of study materials or patients: Scarlett Lin Gomez,Christina A. ClarkeCollection and assembly of data: Scarlett Lin Gomez, JuanYang, Christina A. ClarkeData analysis and interpretation: Scarlett Lin Gomez, JuanYang, Alan Sandler, Manali Patel, Iona Cheng, Heather A.Wakelee, Christina A. ClarkeManuscript writing: All authorsFinal approval of manuscript: All authors


LungCancer Survival AmongChineseAmericans, 2000 to 2010The following represents disclosure information provided byauthors of this manuscript. All relationships are consideredcompensated. Relationships are self-held unless noted. I =Immediate Family Member, Inst = My Institution. Relation-ships may not relate to the subject matter of this manuscript.For more information about ASCO’s conflict of interest policy,please refer to www.asco.org/rwc or jco.ascopubs.org/site/ifc.

Scarlett Lin GomezEmployment: Eurofin (I), Boehringer Ingelheim (I)Stock or Other Ownership: Amgen (I)Research Funding: GenentechTravel, Accommodations, Expenses: Genentech

Juan YangNo relationship to disclose

Table 2 – Median Survival and Adjusted HRs for All-Cause Mortality Among Chinese Americans Diagnosed With Lung Cancer by Sex, California, 2000 to2010 (Continued)

Characteristic

Men Women



Large cell 11.8 (8.4 to 19.2) 0.76 (0.55 to 1.04) 12.7 (9.6 to 19.9) 0.98 (0.60 to 1.58)

NSCLC, NOS 8.9 (7.4 to 10.6) 0.92 (0.76 to 1.10) 11.4 (8.7 to 13.1) 1.21 (0.82 to 1.78)

Other lung cancer 12.4 (8.9 to 17.6) 0.60 (0.37 to 0.97) 10.9 (5.7 to 23.9) 0.97 (0.47 to 2.01)

NOTE. HRs computed via multivariable Cox proportional hazards models stratified by SEER summary stage and chemotherapy, and adjusted for all factors shown in the table inaddition to age at diagnosis, year of diagnosis, urbanicity, surgery type, and radiation. —, Not estimated.Abbreviations: HR, hazard ratio; NCI, National Cancer Institute; NOS, not otherwise specified; NSCLC, non–small-cell lung cancer; SES, socioeconomic status.*Statistically significant at P , .05.







Shih-Wen LinEmployment: GenentechStock or Other Ownership: GenentechTravel, Accommodations, Expenses: Genentech

Margaret McCuskerEmployment: GenentechStock or Other Ownership: RocheTravel, Accommodations, Expenses: Genentech

Alan SandlerEmployment: GenentechStock or Other Ownership: Roche

Manali PatelNo relationship to disclose

Iona ChengNo relationship to disclose

Heather A. WakeleeConsulting or Advisory Role: Peregrine Pharmaceuticals, Acea,Pfizer

Research Funding: Genentech, Pfizer, Xcovery, Celgene,Bristol-Myers Squibb, Novartis, Clovis, Gilead Sciences,AstraZeneca/MedImmune, ExelixisTravel, Accommodations, Expenses: Pfizer, Genentech, ClovisOncology, Acea

Christina A. ClarkeLeadership: HealthlineStock or Other Ownership: Healthline, Tonic Health SolutionsResearch Funding: GenentechTravel, Accommodations, Expenses: Genentech

ACKNOWLEDGEMENTSupported byGenentech,which develops products that targetcancer entities described in thismanuscript. The terms of thisarrangement have been reviewed and approved by the CancerPrevention Institute of California in accordance with its policyregarding objectivity in research. The collection of cancerincidence data used in this study was supported by theCalifornia Department of Health Services as part of thestatewide cancer-reporting program mandated by CaliforniaHealth and Safety Code, Section 103885; by the NationalCancer Institute’s Surveillance, Epidemiology, and EndResults Program under contracts No. N01-PC-35136awarded to theCancer Prevention Institute of California, N02-PC-15105 awarded to the Public Health Institute,HHSN261201000140C awarded to the Cancer PreventionInstitute of California, HHSN261201000035C awardedto the University of Southern California, andHHSN261201000034C awarded to the Public Health In-stitute; and by the Centers for Disease Control and Preven-tion’s National Program of Cancer Registries, underagreements No. U55/CCR921930-02 awarded to the PublicHealth Institute and U58DP003862-01 awarded to theCalifornia Department of Public Health. The ideas andopinions expressed herein are those of the author(s), andendorsement by the State of California Department ofPublic Health, the National Cancer Institute, and theCenters for Disease Control and Prevention or theircontractors and subcontractors is not intended nor should beinferred.

REFERENCES1. US Census Bureau: 2010 Census shows Asians are fastest-growing race group. https://www.census.gov/newsroom/

releases/archives/2010_census/cb12-cn22.html

2. Asian American Center for Advancing Justice: A Community of Contrasts: Asian Americans in theUnited States: 2013.Washington, DC, Asian American Center for Advancing Justice, 2013

3. Gomez SL, Noone AM, Lichtensztajn DY, et al: Cancer incidence trends among Asian American populations in theUnited States, 1990-2008. J Natl Cancer Inst 105:1096-1110, 2013

4. Miller BA, Chu KC, Hankey BF, et al: Cancer incidence and mortality patterns among specific Asian and PacificIslander populations in the U.S. Cancer Causes Control 19:227-256, 2008

5. Chang ET, Shema SJ, Wakelee HA, et al: Uncovering disparities in survival after non-small-cell lung cancer amongAsian/Pacific Islander ethnic populations in California. Cancer Epidemiol Biomarkers Prev 18:2248-2255, 2009

6. Gomez SL, Chang ET, Shema SJ, et al: Survival following non-small cell lung cancer among Asian/Pacific Islander,Latina, and Non-Hispanic white women who have never smoked. Cancer Epidemiol Biomarkers Prev 20:545-554,2011

7. Subramanian J, and Govindan R: Lung cancer in never smokers: A review. J Clin Oncol 25:561-570, 2007

8. US Census Bureau: The Asian population: 2010. https://www.census.gov/prod/cen2010/briefs/c2010br-11.pdf

9. Gomez SL, Le GM, West DW, et al: Hospital policy and practice regarding the collection of data on race, ethnicity, andbirthplace. Am J Public Health 93:1685-1688, 2003

10. Carreon JD,Morton LM, Devesa SS, et al: Incidence of lymphoid neoplasms by subtype among six Asian ethnic groupsin the United States, 1996-2004. Cancer Causes Control 19:1171-1181, 2008

11. Gomez SL, Glaser SL, Kelsey JL, et al: Bias in completeness of birthplace data for Asian groups in a population-basedcancer registry (United States). Cancer Causes Control 15:243-253, 2004

12. Gomez SL, Quach T, Horn-Ross PL, et al: Hidden breast cancer disparities in Asian women: Disaggregating incidencerates by ethnicity and migrant status. Am J Public Health 100:S125-S131, 2010 (suppl 1)

13. Yang J, SchuppCW,Harrati A, et al: Developing an area-based socioeconomicmeasure fromAmericanCommunity Surveydata. http://www.cpic.org/files/PDF/Research_Files/Reports/CPIC_ACS_SES_Index_Documentation_3-10-2014.pdf


https://www.census.gov/newsroom/releases/archives/2010_census/cb12-cn22.html

https://www.census.gov/newsroom/releases/archives/2010_census/cb12-cn22.html

https://www.census.gov/prod/cen2010/briefs/c2010br-11.pdf

http://www.cpic.org/files/PDF/Research_Files/Reports/CPIC_ACS_SES_Index_Documentation_3-10-2014.pdf


14. Yost K, Perkins C, Cohen R, et al: Socioeconomic status and breast cancer incidence in California for different race/ethnic groups. Cancer Causes Control 12:703-711, 2001

15. Gomez SL, Glaser SL,McClure LA, et al: The California Neighborhoods Data System: A new resource for examining theimpact of neighborhood characteristics on cancer incidence and outcomes in populations. Cancer Causes Control 22:631-647, 2011

16. Wakelee HA, Wang W, Schiller JH, et al: Survival differences by sex for patients with advanced non-small cell lungcancer on Eastern Cooperative Oncology Group trial 1594. J Thorac Oncol 1:441-446, 2006

17. Aizer AA, ChenMJ, McCarthy EP, et al: Marital status and survival in patients with cancer. J Clin Oncol 31:3869-3876,2013

18. Rendall MS, Weden MM, Favreault MM, et al: The protective effect of marriage for survival: A review and update.Demography 48:481-506, 2011

19. Yu M, Feuer EJ, Cronin KA, et al: Use of multiple imputation to correct for bias in lung cancer incidence trends byhistologic subtype. Cancer Epidemiol Biomarkers Prev 23:1546-1558, 2014

20. Clegg LX, Reichman ME, Hankey BF, et al: Quality of race, Hispanic ethnicity, and immigrant status in population-based cancer registry data: Implications for health disparity studies. Cancer Causes Control 18:177-187, 2007



originalreport

Increasing Access to Oral AnticancerMedicines in Middle-Income Countries:A Case Study of Private Health InsuranceCoverage in Brazil

abstract

TheWorld Health Organization estimates that approximately 60% of the world’s new annual cancer casesoccur inAsia, Africa, andCentral andSouthAmerica, and that 70%of cancer deathsoccur in these regions.Although oral chemotherapy is a promising intervention for cancer treatment, given its high cost, it isusually unavailable in middle-income countries. In 2013, after strong lobbying from civil society, Brazil’sCongress passed legislation mandating that all private health insurance companies provide access to oralantineoplastic treatment. The decision to scale up the provision of oral chemotherapy was a watershedevent in the regulation of private health insurance inBrazil. Until then, private insurers,whichcover25%ofthe population, were exempted from the provision of pharmaceutical drugs for home care treatments. Thisarticle explores the political process involved in regulating the provision of oral chemotherapy medicinesby private health insurers. Elements of this successful advocacy case included investment in strategiccommunication, specialized knowledge of regulatory policy, and the ability to act via democratic channelsof political representation. In turn, the receptiveness of government branches such as the Congress andregulating bodies, as well as the Cancer Awareness Month campaign, opened a window of opportunity.However, prospects for expanded access to such medicines in the public health system are bleak in theshort term because of the ongoing political and economic crisis.

J Glob Oncol 2. © 2016 by American Society of Clinical Oncology License under the Creative Commons Attribution 4.0 License

INTRODUCTION

Cancer is a leading cause of morbidity and mor-talityworldwide,withapproximately14millionnewcases and 8.2 million cancer-related deaths in2012.1 More than 60% of the world’s total newannual cases occur in Asia, Africa, and Centraland South America, and these areas account for70% of the world’s cancer deaths. Since the1990s, oncology treatment has advanced fromintravenous chemotherapy to include only orallyadministered antineoplastic agents, which aremore convenient to administer and often havefeweradverseeffects thando their olderparenteralcounterparts.2 Oral drugs may also lead to lowercosts through savings related to decreased hos-pitalization times, for instance, and better qualityof life.3,4 Usually, newer technologies such as oralantineoplastics put pressure on health care bud-gets and are incorporated infrequently by healthsystems in middle-income countries because oftheir added direct cost.5

Brazil is well known for offering free and universalaccess to antiretroviral drugs, which are deliveredexclusively through its public dispensing units.6

However, far less is known and has been pub-lished about the country’s experience with other,noncommunicable diseases that require high-cost treatments, such as cancer. Unlike drugsfor AIDS, cancer medicines have been providedwithin the public health care system, purchasedout of pocket at pharmacies by individuals andfamilies, or purchased/reimbursed by private in-surance companies. This article focuses on therecent developments in oral anticancer drug cov-erage through the supplemental (private) healthinsurance system.

After a new parenterally administered anticancerdrug is approved by ANVISA, the Brazilian regu-latory agency, it gains coverage automatically inthe private health care systemafter a price is set byan interministerial commission. Historically, how-ever, oral drugs were not covered. In 2013, after

ElizeMassard da Fonseca

Francisco Inacio Bastos

Gilberto Lopes

Elize Massard da Fonesca,S~ao Paulo BusinessSchool, S~ao Paulo;Fransisco Inacio Bastos,Institute of Scientific andTechnologicalCommunication andInformation in Health, Riode Janeiro, Brazil; andGilberto Lopes,Oncoclinicas Group, S~aoPaulo, Brazil and JohnsHopkins University,Baltimore, MD.Supported by Grant No.2014/07725-3 from theSao Paulo ResearchFoundation.

Authors’ disclosures ofpotential conflicts ofinterest and contributionsare found at the end of thisarticle.Corresponding author:Elize Massard da Fonseca,PhD, Public ManagementDepartment, S~ao PauloBusiness School, RuaItapeva, 474 7th Fl., S~aoPaulo, Brazil; e-mail:[email protected]


© 2016 by American Society of Clinical Oncology License under the Creative Commons Attribution 4.0 License



strong lobbying from civil society, Congress ap-proved legislationmandating that all private healthinsurance companies provide access to oral anti-neoplastic treatment (Brazilian Federation Law12880/2013). At the same time, the NationalRegulatory Agency for Private Health Insurance(ANS) issued a resolution mandating privatehealth insurance companies to cover 37 orallyadministered oncology drugs for home cancertreatment (ANS Resolution 388/2013). BetweenJanuary 2013 and December 2014, the numberof requests for oral antineoplastic treatment in-creased by almost 90% in the private health in-surance sector.7

The decision to scale up the provision of oralchemotherapy was a watershed event in the reg-ulation of private health insurance in Brazil, whichwas until then exempted from the provision ofpharmaceuticals drugs for home care treatment.In addition, it is particularly important to highlightthat this case builds evidence for the role of civilsociety in advocating access to cancer treatmentwithin a middle-income country. Although muchhas been said regarding the relevance of civilgroups to cancer policy advocacy,8 we know littleabout the processes by which they affectedchange in regulatory policies. This article aimsto explore the political process involved in regu-lating the provision of oral chemotherapy medi-cines by private health insurance companies forcancer home care treatment in Brazil.

BACKGROUND

The Brazilian National Cancer Institute estimatesthat there will be 576,000 new cases of cancer,including nonmelanoma skin cancer, in the coun-try in 2015. Among men, prostate cancer repre-sents 22.8% of incident cases, whereas amongwomen, breast cancer represents 20.8% of allcases.9 Breast cancer has been one of the leadingcauses of death from cancer among Brazilianwomen, whereas lung and prostate cancer havebeenamong the leadingcauses of death inmen.10

Oncology treatment in public health care in Brazilis provided by 278 accredited establishments.11 Itis the responsibility of the reference service topurchase and provide chemotherapy, which isreimbursed by the state government with re-sources transferred from the Ministry of Healthand funded through general taxation.12 Between2006 and 2011, a commission was established inthe Ministry of Health to assess the inclusion ofnew technologies into the National Health Service(SUS); before that, health technology assessmentwasnot commonly performed inBrazil. During this

period, only two oral antineoplastic medicines(nilotinib and dasatinib) were incorporated intothe SUS (public) list.

Parallel to the SUS, which covers all Braziliancitizens and foreigners who are physically presentin the country (including illegal aliens), privatehealth insurance companies are also respon-sible for providing oncology treatment to theirclients. Although Brazil has one of the largestpublic health systems in the world, approxi-mately 25% of the population is also enrolledin private health insurance.13 The majority ofprivate health insurance plans are collectiveplans (approximately 80%) obtained throughwork or cooperatives, and a small and decliningpercentage are individual plans (approximately20%). The Private Health Insurance Act (Law9656/1998) regulates private insurance in Brazil.According to this act, insurance companies arerequired to cover only medicines used duringperiods of hospitalization and during ambulatoryvisits (article 12) and not during home care treat-ment (article 10). Therefore, oral antineoplasticmedicines for home care were excluded fromthe list of procedures covered by private insur-ance. These two scenarios led to an increase inthe number of lawsuits over access to cancermedicines.14

The 2013 National Health Survey, a household-based nationwide study, suggested that 49millionpeople are covered by private health insurance inBrazil. Among these, approximately 17% do notpay for their medicines because they have freeaccess to medicines through their governmentalprograms or employer’s insurance program;60.9% had to pay for medications out of pocket,16.1% paid for a percentage of the medicinesneeded, and 6% were not able to afford themedicines prescribed. As such, it is clear thatamong those people who hold private healthinsurance, a large proportion pay for 100%of theirmedicines themselves.

In 2011, three parallel events paved the way forthe inclusion of oral cancer treatments in privatehealth insurance coverage; they are divided herefor clarity. Table 1 provides a chronologic list ofthese events.

Revisionof theMinimal List of ProceduresCoveredby Private Health Insurance

Every 2 years, the ANS revises the minimal list ofprocedures with mandated coverage by privatehealth insurance companies, and this process isopen to different segments of society throughpublic consultation. In May 2011, the regulatory



Table 1 – List of Events Preceding Approval of Law 12.880/2013

Date Event Action

2011 Publishing of the ANS Regulatory Agenda for 2011-2012

Inclusion of the question of pharmaceutical assis-tance as a core issue for discussion in the ANS

Apr 15, 2011 to May 21, 2011 ANS Public Consultation No. 40 Different segments of society given the opportunityto comment on the revision of the list of proce-dures covered by private health insurance

May 2, 2011 to May 20, 2011 Instituto Oncoguia mobilizes other civil society or-ganizations and begins a mass media campaignfor the inclusion of oral chemotherapy in privatehealth insurance plans

Signatures supporting the campaign submitted tothe ANS, distribution of folders informing thepopulation about oral chemotherapy

May 11, 2011 Seminar in Congress to discuss the rights of peopleliving with cancer

The Instituto Oncoguia and the ANS invited to givespeeches

Jun 22, 2011 Presentation of Bill 352/2011 Senator Ana Amelia presents Bill 325/2011 toSenate

Aug 1, 2011 ANS Resolution 262 Updated list of procedures mandatory for inclusionin health insurance plans; oral chemotherapy isnot among them

Aug 18, 2011 Oncoguia meets with the minister of health The minister endorses the campaign on oral che-motherapy, which receives inclusion in the ANSlist of procedures

Sep 29, 2011 First meeting of the WGPA at the ANS Presentation of successful cases of pharmaceuticalcare packages offered by nonprofit, private in-surance companies

Dec 15, 2011 Congress Hearing on Bill 352/2011 Different stakeholders express their positions

Mar 27, 2012 Second meeting of the WGPA at the ANS One of the directors of the ANS, Maurıcio Ceschin,opens the meeting stating the relevance of thisissue; different stakeholders express theirpositions

Apr 28, 2012 Senate approval of Bill 352/2011 Bill 352/2011 is submitted to the Chamber ofDeputies

Apr 30, 2012 Third meeting of the WGPA at the ANS Discussions to provide input regarding a resolutionto regulate a voluntary pharmaceutical carepackage that would include oral chemotherapy inprivate health insurance plans

Sep 4, 2012 to Oct 6, 2012 ANS Public Consultation No. 49 Different segments of society invited to comment onthe regulation of the pharmaceutical carepackage

Oct 30, 2012 Resolution 310 Regulation of voluntary pharmaceutical carepackages offered by private health insurancecompanies

Nov 30, 2012 Fourth meeting of the WGPA at the ANS Presentation of ANS Resolution 310

Aug 27, 2013 Bill 3.998/2012* is approved by the CCJC of theChamber of Deputies

The CCJC acknowledges the importance of themobilization of civil society in the approval of thisproposal

Oct 21, 2013 The ANS issues Resolution 338 Update to list of procedures covered by privateinsurance plans; this time, the list includes oralchemotherapy

Oct 22, 2013 Senate approves the amendment proposed by theChamber of Deputies

Legislation submitted to President Dilma Roussefffor ratification

Nov 12, 2013 President Dilma Rousseff enacts Law 12.880/2013

Abbreviations: ANS, National Regulatory Agency for Private Health Insurance; CCJC, Commission on Constitution, Justice and Citizenship; WGPA, Working Group onPharmaceutical Assistance.*Bill 352/2011 and 3.998/2012 are the same. The bill number was modified once submitted to Congress.



agency promoted Public Consultation No. 40, in-viting comments and suggestions from the publicon the list of procedures covered by insurancecompanies.

A group called Instituto Oncoguia, which repre-sents people living with cancer, saw this as anopportunity to submit a much-sought-after claimfor these patients, that is, that insurance compa-nies should cover the cost of oral antineoplasticmedicines.Anumberofpatients reported toOnco-guia that their insurance plans did not cover oralantineoplastic drugs and that they were also de-nied access to these medicines in the publicsector, forcing them to file lawsuits to obtain treat-ment. The institute, a nonprofit organization, wasformed in 2009 by health professionals workingwith people living with cancer in the state ofS~ao Paulo. This civil society organization is spon-sored by pharmaceutical industries and otherprivate sector companies such as Google Grants,Asics, and others. In addition, the American Can-cer Society has provided training on strategicpolitical advocacy and technical assistance tocancer nongovernmental organizations (NGOs),including Oncoguia and the Brazilian Federationof Philanthropic Breast Health Institutions.15,16

Between May 2 and May 20, 2011, the InstitutoOncoguia launcheda largemassmedia campaigncalled the “Campaign for Inclusion ofOral Chemo”to put pressure on the government to include oralantineoplastic medicines on the mandated list ofANS procedures. For the duration of this cam-paign,Oncoguia coordinated a groupofmore than17 associations including medical and profes-sional societies and AIDS and hepatitis NGOs thatwere sympathetic to this demand, as well asspecific groups of patients with cancer, such asthose representing individualswith lymphomaandothers. According to their records, the social me-dia campaign received more than 900 citations,2,500 informative folders were distributed to in-form the population about their demands, and apetition was submitted to the ANS after collecting18,000 signatures. Despite these efforts, the ANSinitially denied their claim for the inclusion of oralchemotherapy drugs in the new list of procedurescovered by insurance (Resolution 262/2011). TheANS argued that this inclusion was not allowedunder the Private Health Insurance Act.16a How-ever, after this decision, the agency created aWorking Group of Pharmaceutical Assistance(WGPA) to discuss avenues for pharmaceuticalcare within private health insurance in Brazil.Instituto Oncoguia not only participated in thisworking group, but also coordinated, in tandem

with a group of other NGOs, a campaign in Con-gress to amend the Private Health Insurance Act,which we discuss in the next section.

Working Group of Pharmaceutical Assistance

The ANS noted that the theme of pharmaceuticalcare was a main concern of the “RegulatoryAgenda for 2011/2012,” which is the planninginstrument that set the priority goals for the agencyin the biennium. As a consequence, in September2011, the WGPA was created that included rep-resentation from private insurance companies,consumers, health professionals, and regulators.The motivation for the creation of the WGPAwas twofold. First, the agency was aware that alarge proportion of consumers were paying outof pocket for medicines not covered by insur-ance plans. Second, the ANS was concernedthat the regulatory framework for private healthinsurance in Brazil was encouraging a healthcare model strongly focused on hospital andambulatory care, with little focus on home carecoverage.17

During the first WGPA meeting, the ANS invitedsome private health insurance companies thatprovided a pharmaceutical care package, in ad-dition to their regular coverage, to share theirexperiences. However, these were nonprofithealth care plans (ie, self-managed by the com-pany or a third party responsible for the socialbenefit),18 and therefore, other insurance compa-nies questioned the viability of applying their ex-perience to for-profit insurance plan models.

One of the opponents to the idea of includingpharmaceutical care coverage in insurance planswas the Institute of Studies of SupplementaryHealth Care, a think tank that produces analysesfor the sector, which prepared a technical notediscussing cases of pharmaceutical coverage byprivate health insurance in several countries.19

They argued, mostly on the basis of the UnitedStates Medicare model, that there was no evi-dence to suggest that pharmaceutical coveragereduces health care expenditure, but that in con-trast, there was evidence to indicate that newmedicines increase cost and that there is an in-creased use of medicines among healthy peopleafter enrolling in such insurance coverage (ie,it would increase moral hazard). Although theirstudy did not provide a comprehensive discussionof themethodologicproceduresnecessary to drawsuch conclusions, it was widely used by privateinsurance representatives as evidence against theideaof regulating the inclusionof pharmaceuticalsin insurance plans (personal communication,



October 24, 2014). However, civil society groupssuch as the Otimismo Group, which representspatients living with viral hepatitis B and C, andOncoguia also participated in the WGPA discus-sions. These civil society groups demanded thatthe ANS request a more effective response fromprivate health insurance companies regarding theprovision pharmaceutical care.

Given the strong conflicting positions, the coordi-nator of the WGPA consulted the attorney generalfor the ANS and requested a legal opinion on thismatter. The interpretationwas that because article10 of the Private Health Insurance Regulatory Actexemptedprivate health insurances fromcoveringmedicines for home care treatment, the ANSwould be entitled to regulate only complementary,voluntary packages of pharmaceutical care.20

The WGPA was therefore able to discuss a draftresolution for pharmaceutical care packages thatcould be offered as additional, voluntary productsto willing private health insurance plans. A draftresolution was then submitted for public consul-tation (PublicConsultationNo.49/2012). The finaldocument, approved in 2012 (Resolution 310/2012), established that pharmaceutical care is avoluntary and supplementary product to healthinsurance plans, similar to products such as airambulance services. Therefore, civil societygroups were once more left with just one ave-nue for action: the Private Health Insurance Actwould have to be amended to allow access tohome pharmaceutical care in private insuranceplans. This would require a political debate inCongress.

Legislative Actions to Amend the Private HealthInsurance Act

Parallel to the discussions at the ANS, Oncoguia,together with other civil society organizations,coordinated a campaign to amend the PrivateHealth Insurance Act in Congress. The first stepwas taken in May 2011 after a meeting withSenator Ana Amelia Lemos (Progressive Party,RioGrandedoSulState),whoorganizeda seminarto discuss the rights of people living with cancer.During this event, a main discussion point was thedenied coverage of oral antineoplastics by privatehealth insurance providers.20a As a consequence,Senator Ana Amelia, sensitive to issues relatedto cancer control, tabled a bill in Congress (Bill352/2011) that would mandate that all privatehealth insurance plans include oral cancer treat-ment as a mandatory procedure. Oncoguia alsoappealed to theminister of health for support of Bill352/2011.21

In December 2011, a public hearing was held inCongress to discuss the different positions aroundthis bill.22 Several stakeholders took part, includ-ing representatives from the ANS; representativesof a reference cancer hospital; a representativefrom the Brazilian Association of Group Medicine(Abramge); a representative from the BrazilianMedical Association; and Oncoguia, representingthe interests of civil society. During this event,Abramge, with Arlindo de Almeida representingprivate health insurance companies, argued thatthere is a lack of governmental funds to supportcancer care treatment and that oral antineoplasticmedicines should be the responsibility of thefederal government. Abramgealsomentioned thatthere was a reorganization of private health in-surance companies, with more than 1,100 com-panies exiting the market since the creation of theANS in 1998. Finally, he stated that annual mar-gins for the private health insurance sector wereless than 1%. The arguments used by privateinsurance companies to emphasize the cost ofproviding oral antineoplastics, and the effects ofthe bill on the private insurance sector, were notwell received in Congress.22 On the other hand,the argument for supporting the needs of peo-ple living with cancer resonated more widely inthe Senate and Chamber of Deputies (the twohouses that form the Brazilian Congress). Ac-cess to cancer treatment is an argument thatis hard to counter when elections are aroundthe corner.

Records of Oncoguia’s and its partners’ activi-ties are publicly available online22a and suggestthat these groups constantly lobbiedmembers ofthe commissions responsible for the bill. They alsoidentified other civil society organizations thatcould help them in the advocacy process. Forinstance, in August 2013, the bill was approvedby theCommission of theConstitution, Justice andCitizenship, which received more than 250 mes-sages of support for this proposal.22b The final stepin approving the legislation was the endorsementof the Senate. The Brazilian Congress joined theinternational Breast Cancer Awareness Monthcampaign in October and, as a consequence,the president of the Senate agreed to make allbills related to cancer care a voting priority. In thiscontext, the bill was finally approved. InNovember2013, President Dilma Rousseff enacted Law12.880/2013.

Interestingly, a day before the Senate approvedthe bill, the ANS, foreseeing the decision, also de-cided to include 37 oral antineoplastic drugs in thenew revision of the list of procedures covered by



insurance plans (Resolution 338/2013). Together,the ANS resolution and Law 12.880/2013 ensurethat all consumers of private health insurance inBrazil are eligible to receive the 37 oral antineo-plastic medicines, if needed.

Using as an example the experience of Brazil incovering oral antineoplastic medicines throughprivate health insurance plans, it is possible todraw some conclusions regarding regulatorypolicy.

Civil Society Advocacy on Pharmaceutical Care

We know a great deal about the role of civil so-ciety in advocating for antiretroviral medicines inBrazil and other developing countries6,23 andabout the necessity of empowering cancer advo-cacy groups in developing countries.8 Findingsfrom this study suggest that civil society, coordi-nated by the Instituto Oncoguia, played a crucialrole in Brazil’s regulation of oral antineoplasticcoverage under private health insurance. On theother hand, there were several institutional chan-nels, including public consultations, congressio-nal hearings, and a meeting with the minister ofhealth, through which these advocacy groupscould voice their concerns. Theprocessdescribedhere suggests a democratic, pluralist process ofpolicy making.

Persuading Decision Makers

Studies of HIV/AIDS advocacy have long calledattention to the policy frame and strategic use ofideas,24,25 that is, the use of normative values andinformation to influence policy debates. In theprocess discussed here, medicines can save livesor extend the lives of people living with cancer,arguments that are persuasive and hard to de-construct. Therefore, the second explanatoryelement for the successful campaign on oral che-motherapy was the ability of civil society to frametheir demands and publicize their position indifferent political arenas.

Policy Context

The third important element in the case of oralantineoplastic medicines refers to the contextin which this proposal was formulated. Scholarsof policy argue that it is crucial to understandthe historical circumstances through which pub-lic policies are designed and implemented.26

The ANS was sensitive to the discussion of

pharmaceutical care presented by private healthinsurance companies. Although the initial out-come was unsatisfactory because it proposedvoluntary coverage only, the ANS ended up in-cluding oral antineoplastic treatment in the min-imum coverage list of procedures. In addition,given the strong support of civil society, Congresswas mobilized to endorse an amendment in thePrivate Health Insurance Act. Finally, the recog-nition and marking of Breast Cancer AwarenessMonth created a window of opportunity to finallyapprove the bill in Congress.

Few studies have detailed the consequences ofthis change in regulatory policy on the privatehealth insurance sector and on the treatment ofpeople living with cancer; however, the availabledata suggest an important increase in the percent-age of requests for oral chemotherapy in privatehealth insurance between 2013 and 2014 (ap-proximately 90%),7 and the authors are currentlyworking with data from Intercontinental MarketingServices Health to quantify the increase to accessthat has occurred since January 2014 when thecoverage became official. In the current context ofpolitical andeconomiccrisis, it is unlikely that suchinnovations could be incorporated into Brazil’spublic health system, already affectedbydramaticcuts of 70 billion reais (nearly 20 billion dollars) in2015.27 In any case, the successful reform ofprivate health insurance plans may pave theway for future changes in the public sphere.

In summary, this Brazil case study suggests thatcivil society played a crucial role in the regulatoryprocess. The main elements leading to this suc-cessful intervention were (1) the investment instrategic communication, (2) the possession ofspecialized knowledge of regulatory policy, and(3) the ability to act via democratic channels ofpolitical representation, such as public consulta-tions. In turn, the receptiveness of governmentbranches such as Congress and the ANS, and theCancer Awareness Month campaign, opened awindow of opportunity for the change desired bythese groups. The impact of civil society was less aresult of their material resources andmore a resultof their capacity to persuade decision makers,despite the strong and resourceful lobbying ofthe private health insurance sector.



Conception and design: Elize Massard da Fonseca, FranciscoInacio Bastos

Collection and assembly of data: Elize Massard da Fonseca,Francisco Inacio Bastos

Data analysis and interpretation: All authors







Increasing Access to Oral Anticancer Medicines in Middle-Income Countries: A Case Study of Private Health InsuranceCoverage in BrazilThe following represents disclosure information provided byauthors of this manuscript. All relationships are consideredcompensated. Relationships are self-held unless noted. I =Immediate Family Member, Inst = My Institution. Relation-ships may not relate to the subject matter of this manuscript.For more information about ASCO’s conflict of interest policy,please refer to www.asco.org/rwc or jco.ascopubs.org/site/ifc.

Elize Massard da FonsecaNo relationship to disclose

Francisco Inacio BastosNo relationship to disclose

Gilberto LopesHonoraria: AstraZeneca, Roche/Genentech, Merck Serono,Merck Sharp & Dohme, Fresenius Kabi, Novartis, Bristol-Myers Squibb, Janssen-Cilag

Consulting or Advisory Role:Pfizer, Bristol-Myers Squibb, Lilly/ImClone

Research Funding: Lilly/ImClone, Pfizer, AstraZeneca, MerckSharp & Dohme, Eisai

Expert Testimony: SanofiBoard Member: Instituto Oncoguia scientific board

REFERENCES1. World Health Organization: Cancer. Fact sheet No. 297. http://goo.gl/I9YRvJ

2. Bedell CH: A changing paradigm for cancer treatment: The advent of new oral chemotherapy agents. Clin J Oncol Nurs7:5-9, 2003

3. Hoff PM, Ansari R, Batist G, et al: Comparison of oral capecitabine versus intravenous fluorouracil plus leucovorin asfirst-line treatment in 605 patients with metastatic colorectal cancer: Results of a randomized phase III study. J ClinOncol 19:2282-2292, 2001

4. Faithfull S, and Deery P: Implementation of capecitabine (Xeloda) into a cancer centre: UK experience. Eur J OncolNurs 8:S54-S62, 2004

5. Bazargani YT, de Boer A, Schellens JH, et al: Selection of oncology medicines in low- and middle-income countries.Ann Oncol 25:270-276, 2014

6. Nunn A: The Politics and History of AIDS Treatment in Brazil. New York, Springer, 2008

7. Piedade A, Castro A, Alves A, et al: Impact of federal regulation on access to oral chemotherapy (OC) and hormonaltherapy (HT) in the Brazilian private health system (PHS). Presented at ISPOR 20th Annual International Meeting.Philadelphia, PA, May, 16-20, 2015

8. Goss PE, Lee BL, Badovinac-Crnjevic T, et al: Planning cancer control in Latin America and the Caribbean. LancetOncol 14:391-436, 2013

9. Instituto Nacional do Cancer: Estimativa 2014 - Incidencia de cancer no Brasil. Rio de Janeiro, Brazil, INCA, 2014

10. International Agency for Research on Cancer: Fact Sheets: Brazil. 2012. http://goo.gl/L75OXP

11. Ministerio da Saude: Implantaç~ao das Redes de Atenç~ao a Saude e outras estrategias da SAS / Ministerio da Saude.Brasılia, Brazil, Ministerio da Saude, 2014

12. Ministerio da Saude: Manual de Bases Tecnicas da Oncologia - SIA/SUS. Brasılia, Brazil, Ministerio da Saude, 2013

13. Instituto Brasileiro de Geografia e Estatıstica: Pesquisa Nacional de Saude 2013., Rio de Janeiro, Brazil, IBGE, 2013

14. Lopes G, Prado E, and Montero A: Suing the state for access to cancer medications: The Brazilian experience. J ClinOncol 2015;33:e17566

15. Camara C: Mapeamento polıtico da saude no Brasil: Um recurso para ONGs atuando em cancer demama. Sao Paulo,Brazil, American Cancer Society, 2011

16. Instituto Oncoguia: Instituto Oncoguia e American Cancer Society promoveram Simposio para Jornalistas de todo oBrasil (30/08/2011). http://goo.gl/0BXGfR

16a. Agencia Nacional de Saude Suplementar: Nota tecnica pos consulta publica n° 40: Revis~ao do rol de procedimentose eventos em Saude. http://www.ans.gov.br/images/stories/Plano_de_saude_e_Operadoras/Area_do_consumidor/nota%20ps%20consulta%20publica2.pdf

17. AgenciaNacional de Saude Suplementar: Ata da Reuni~ao do Grupo Tecnico de Assistencia Farmaceuticado dia 29 desetembro de 2011. http://goo.gl/NfamCu

18. Agencia Nacional de Saude Suplementar: Sumario Executivo de Impacto Regulatorio referente a consulta publicanumero 49.2012a 7 de junho de 2011. http://goo.gl/YNxWdk

19. Novais M, and Reis A: Seguros para Cobertura de Medicamentos – Principais Desafios e Experiencia. S~ao Paulo,Brazil, Instituto de Estudos de Saude Suplementar, 2012

20. Agencia Nacional de Saude Suplementar: Ata da Reuni~ao do Grupo Tecnico de Assistencia Farmaceutica do dia 27de março de 2012. http://goo.gl/HlrAJa




http://goo.gl/I9YRvJ

http://goo.gl/L75OXP

http://goo.gl/0BXGfR

http://www.ans.gov.br/images/stories/Plano_de_saude_e_Operadoras/Area_do_consumidor/nota%20ps%20consulta%20publica2.pdf

http://www.ans.gov.br/images/stories/Plano_de_saude_e_Operadoras/Area_do_consumidor/nota%20ps%20consulta%20publica2.pdf

http://goo.gl/NfamCu

http://goo.gl/YNxWdk

http://goo.gl/HlrAJa


20a. Portal Atividade Legislativa Comiss~oes: Comiss~ao de Assuntos Sociais audiencias publicas. http://www.senado.gov.br/atividade/comissoes/CAS/audPub.asp

21. Instituto Oncoguia: Audiencia com Alexandre Padilha, Ministro da Saude (18/08/2011). http://goo.gl/PcvBKA

22. Nacional C: Ata da 55a reuni~ao (extraordinaria) da Comiss~ao de Assuntos Sociais, da 1a Sess~ao Legislativa Ordinariada 54a Legislatura, realizada no dia 15 de dezembro de 2011. Brasılia, Brazil, Congresso Nacional, 2011

22a. Instituto Oncoguia: Campanha: Pela inclus~ao da quimio oral. http://www.oncoguia.org.br/quimioterapiaoral/

22b. Comiss~ao de constituiç~ao e justiça e de Cidadania: Ata da quadragesima oitava reuni~ao ordinaria, realizada em27 deagosto de 2013. http://www.camara.leg.br/internet/ordemdodia/integras/1121935.htm

23. Galvao J: AIDS no Brasil: Agenda de Construç~ao de Uma Epidemia. Rio de Janeiro, Brazil, Associaç~ao BrasileiraInterdisciplinar de AIDS, 2000

24. Kapstein E, and Busby J: AIDS Drugs For All: Social Movements and Market Transformations. New York, NY,Cambridge University Press, 2013. doi:10.1017/CBO9781139565967

25. Sell S, and Prakash A: Using ideas strategically: The contest between business and NGO networks in intellectualproperty rights. Int Stud Q 48:143-175, 2004

26. Pierson P: Politics in time: History, institutions and social analysis. Princeton, NJ, Princeton University Press, 2004

27. Valor Economico: Governo anuncia corte no Orçamento de quase R$ 70 bilh~oes: Valor Economico. http://www.valor.com.br/imprimir/noticia/4062878


http://www.senado.gov.br/atividade/comissoes/CAS/audPub.asp

http://www.senado.gov.br/atividade/comissoes/CAS/audPub.asp

http://goo.gl/PcvBKA

http://www.oncoguia.org.br/quimioterapiaoral/

http://www.camara.leg.br/internet/ordemdodia/integras/1121935.htm

http://www.valor.com.br/imprimir/noticia/4062878

http://www.valor.com.br/imprimir/noticia/4062878


specialarticle

Collision of Three Pandemics: TheCoexistence of Cervical Cancer, HIVInfection, and Prior Tuberculosis in theSub-Saharan Country of Botswana

executivesum

mary

Cervical cancer is the leading cause of cancer-related mortality in the developing world, where HIV andMycobacterium tuberculosis (TB) infection are also endemic. HIV infection is independently associatedwith increased morbidity and mortality among women with cervical cancer. TB is believed to increase therisk of malignancies and could cause chronic inflammation in the gynecologic tract. However, the re-lationship between cervical cancer and TB in settings hyperendemic for HIV is unknown.We found that 18(10%) of a cohort of 180womenwith cervical cancer in Botswana had a history of TB disease. Age andHIVinfection were also associated with a history of TB disease. Our data show that prior TB disease is highlyprevalent among patients with cervical cancer infected with HIV. The coexistence of cervical cancer, HIVinfection, and prior TB infection might be higher than expected in the general population. Prospectivestudies are needed to better determine the impact of the collision of these three world health epidemics.

INTRODUCTION

Cervical cancer is the third most common malig-nancy in women worldwide and the leading causeof cancer-related death for women in developingcountries. Globally, it is estimated that approxi-mately half a million women are diagnosed withcervical cancer every year, and approximately275,000 women die of the disease, 85% of whomlive in low- and middle-income countries.1 In Bot-swana, cervical cancer is the most common gyne-cologic cancer and the leading cause of cancerdeath.2,3

HIV infection accelerates the progression towardcervical cancer and likely is associated with worseclinical outcomes.4 In Botswana, 60% of patientsdiagnosed with cervical cancer are also infectedwith HIV.2,5 With cervical cancer rates rising indeveloping countries where HIV is endemic, iden-tifying and understanding the factors leading toincreased morbidity and mortality in these popu-lations should be a priority.6

The devastating impact of the colliding HIV andtuberculosis (TB) pandemics in resource-limitedsettings has been long recognized.7 HIV infectionis one of the most important risk factors for pro-gression to active TB disease. Furthermore, TB isthe number one killer of people living with HIV in

the world.5 Importantly, the impact of these pan-demics is more severe in countries with morechallenging health care systems, where TB andHIV are less controlled, and patients may havemore advanced disease states compared with thedevelopedworld. Botswana has one of the highestTB prevalence rates in the world, and it is esti-mated that 60% to 75% of people diagnosed withTB are coinfected with HIV.8

An association betweenmalignancies and TBwaspreviously established.9 TB has been implicatedin the pathogenesis of malignancies and mayinterfere with their diagnosis. TB and cancer fre-quently coexist, and the relative immunosuppres-sion caused by cancer or its treatmentmay lead toreactivation of latent TB infections, leading to in-creased morbidity and mortality.9 However, dataon the association between cervical cancer andTB are scarce.10

Despite themagnitude of the cervical cancer, HIV,and TB epidemics coexisting in many resource-limited settings, their associations and the poten-tial interactions among all three pandemics havenot been documented previously, although, aspreviously mentioned, those between HIV andcervical cancer as well as between HIV and TBhave been explored. However, there is limitedinformation on the impact of TB or the

Nicola M. Zetola

Surbhi Grover

Chawangwa Modongo

Sebathu P. Chiyapo

MemoryNsingo-Bvochora

MohanNarasimhamurthy

Lilie L. Lin

Joseph Jarvis

Sanghyuk S. Shin

Erle Robertson

Author affiliations appear atthe end of this article.

N.M.Z. and S.G.contributed equally to thiswork.

Authors’ disclosures ofpotential conflicts ofinterest and contributionsare found at the end of thisarticle.Supported by the Centerfor AIDS Research, ASCOand Conquer CancerFoundation, ASCO YoungInvestigator Award, andNational Institutes ofHealth grant Nos.R01AI097045,K01AI118559, andU54CA190158-02.The funders had no role instudy design, datacollection and analysis,decision to publish, orpreparation of themanuscript.Corresponding author:Surbhi Grover, 244GUniversity of Botswana,Gaborone, Botswana;e-mail: [email protected].






combination of TB and HIV on cervical carcino-mas. In this study, we examined the associationbetween cervical cancer and a prior history of TBby HIV serostatus in a prospective cohort of pa-tients with cervical cancer in Botswana.

METHODS

Cervical cancer diagnosis and treatment in Bot-swana is centralized at Princess Marina Hospital,the largest referral hospital in the country, locatedin the capital, Gaborone. Approximately 95%of allpatients diagnosed with any cancer in publichospitals from all over the country are referredthere for confirmation of diagnosis and treatment.In this study, we enrolled consecutive patientsreferred to Princess Marina Hospital for the treat-ment of cervical cancer between July 2013 andJanuary 2015. We collected demographic char-acteristics and cancer- and TB-specific informa-tion prospectively through patient interviews andmedical record reviews. All data were collectedbefore cancer treatment initiation. Prior TB dis-ease was defined as an episode of clinically ormicrobiologically diagnosedTB requiring initiationof TB treatment at anypoint in the past. HIV testingwas performed on all patients in accordance withBotswana’s national guidelines. Radiologic infor-mation was unavailable in most cases, and it wasnot analyzed in this study.

We first described the characteristics of patientswith cervical cancer when stratified by history ofactive TB and cervical cancer. Associations weredetermined using x2, t test, and Mann-Whitneytesting as appropriate. A logistical regressionmodel was developed to describe correlates ofprior TB disease. Variables for the model wereidentified a priori on the basis of their conceptualimportance and included age at diagnosis of cer-vical cancer, age of first sexual encounter, smok-ing history, and presence of HIV infection. Aseparate model was developed to determine theassociation of CD4 cell count categories usingHIV-negative patients as referents. Odds ratiosand 95% CI were calculated. Data processingand analyseswere performedusing Stata software(STATA, College Station, TX).This study was ap-proved by the institutional review boards at theUniversity of Pennsylvania, University of Bot-swana, and Botswana Ministry of Health. All pa-tients provided informed consent.

RESULTS

We enrolled 180 women during the study period,of whom 117 (65%) were infected with HIV at thetime of their cervical cancer diagnosis. Overall, 18

(10%) patients with cervical cancer had a historyof TB disease, and 16 (89%) of them were coin-fected with HIV. All patients with TB were treatedfor their disease. All HIV-positive patients wereeither already receiving antiretroviral treatmentor had started on antiretroviral treatment at thetime of cancer diagnosis. The main demographicand clinical characteristics of the sample stratifiedby history of TB infection are listed in Table 1. Ourbivariate analyses showed a significantly higherprevalence of HIV infection among patients withcervical cancer with a history of TB comparedwiththose without a history of TB (16 [89.8%] of 18patients v 101 [62.3%] of 162 patients; P5 .02).As listed in Table 2, this difference remainedsignificant inmultivariable analysis (adjusted oddsratio [AOR], 21.5; 95% CI, 2.16 to 214.52; P ,.01). Increasing age was also found to be signif-icantly associated with prior TB disease after ac-counting for confounders (AOR, 1.1; 95%CI, 1.01to 1.11; P 5 .02). In a separate multivariablemodel, we found that CD4 counts less than 350cells/mL (AOR, 57.3; 95% CI, 4.44 to 740.8) andgreater than 500 cells/mm3 (AOR, 39.0; 95% CI,3.15 to 483.89) were associated with prior TBdisease compared with no HIV infection.

DISCUSSION

In this study, we found that prior TB disease wascommon in our sample of patients with cervicalcancer in Botswana. HIV infection was also highlyprevalent among patients with cervical cancerwho had TB disease before their diagnosis ofcervical cancer. Our findings are not surprising,given the well-known association between HIVand TB. Nevertheless, they highlight an underap-preciated overlap of three of the most devastatingepidemics in developing countries: cervical can-cer, HIV, and TB.

Our study design does not allow any inferenceregarding the role of TB in the causal pathway ofcervical cancer, the temporal nature of events, orthepresence of interactions betweenTB,HIV, andcervical cancer thatmay alter the natural course ofany of those diseases. However, our results in-dicate that thecoexistenceofpriorTBdisease,HIVinfection, and cervical cancer is common, sug-gesting that potential interactions between two ormore of those conditions might be common aswell. Disseminated (gynecologic) TB has the po-tential to lead todelays incervical cancerdiagnosisand an increase in morbidity and mortality duringcancer treatment.11 Conversely, cervical cancerdiagnosis may delay the diagnosis and treatmentinitiation of coexisting TB.11,12 Although TB has



not been associated with the development ofcervical cancer, chronic TB infection and inflam-mation in the gynecologic tract is an importantcause of infertility in the developing world.13 Thus,the chronic inflammation due to chronic gyneco-logic TB infectionmight be a contributing factor in

the progression toward cervical cancer.11,13,14

HIV infection is associated with a higher inci-dence, as well as increased morbidity and mor-tality, of both TB and cervical cancer,independently. Whether HIV infection modifiesthe relationship between TB and cervical cancerremains to be determined.

We acknowledge the limitations of our study.Multiple sources of bias could have confoundedour findings. Our study was performed on a con-venience sample of all consecutive patients withcervical cancer referred to the only cancer treat-ment center in Botswana, and we did not have acomparison group from the general population.Therefore,weareunable todeterminewhether theoverall prevalence of prior TB infection was higherthan what would be expected from the generalpopulation with or without stratifying it by HIVstatus. Our inability to prospectively determinethe extent of active TB disease after cancer di-agnosis severely limits any inference regarding thetrue burden of TB in our population. ProspectiveTB-attributable morbidity and mortality data arehighly needed.

Despite these limitations, we believe that theseresults highlight the critical need to better under-stand the imminent collision of three of the mostimportant pandemics of our time: HIV, TB, andcancer. Although data remain scant, HIV is aleading factor increasing the incidence of bothAIDS-defining and non–AIDS-defining cancers insouthern Africa. In Botswana, 43% of all cancersand an even greater proportion of cancer deathsare attributable to HIV. Future studies are neededto determine the extent of the interactions amongthese highly prevalent conditions worldwide.



Financial support: Nicola M. Zetola, Surbhi GroverAdministrative support: Nicola M. Zetola, Surbhi GroverProvisionof studymaterialsor patients:NicolaM.Zetola,SurbhiGroverManuscript writing: All authorsFinal approval of manuscript: All authors


The following represents disclosure information provided byauthors of this manuscript. All relationships are consideredcompensated. Relationships are self-held unless noted. I 5Immediate Family Member, Inst 5 My Institution. Relation-ships may not relate to the subject matter of this manuscript.

Table 1 – Characteristics of Cervical Cancer Patients by History of Prior Tuberculosis inBotswana

Characteristic

History of PriorTuberculosis(n 5 18)

No History ofPrior Tuberculosis

(n 5 162) P

Age, median (IQR) 47.3 (41.1-57.8) 47.8 (39.2-56.2) .78

Marital status, No. (%) .24

Single 13 (86.7) 90 (67.7)

Married 2 (13.3) 41 (30.8)

Divorced or widowed 0 2 (1.5)

Age of first sexual activity,median (IQR)

18 (17-19) 18 (16-20) .39

Smoking, No. (%) 1 (5.6) 4 (2.5), .49

HIV infection, No. (%) 16 (88.9) 101 (62.3) .02

CD4 cell count, median (IQR)* 367 (261-592) 474 (343-579)

CD4 cell count, cells/mL , No. (%)* .51

, 350 6 (37.5) 22 (25.3)

351-500 3 (18.8) 26 (29.9)

. 500 7 (43.8) 39 (44.8)

Taking ART, No. (%)† 16 (100) 95 (101) .18

Viral load undetectable, No. (%) 16/16 (100) 63/66 (95.5)

Cancer type, No. (%) .53

Squamous cell carcinoma 150 (93.2) 16 (88.9)

Adenocarcinoma 11 (6.8) 2 (11.1)

Abbreviations: ART, HIV antiretroviral treatment; IQR, interquartile range.*CD4 cell count at the time of cervical cancer diagnosis.†Only available results presented. One case of papillary serous carcinoma excluded from the analyses.

Table 2 – Multivariate Analyses to Determine Factors Associated With Prior History ofTuberculosis

Factor Adjusted Odds Ratio 95% CI P

Age 1.1 1.01 to 1.11 .023

HIV 21.5 2.16 to 214.52 , .01

Smoking 4.8 0.38 to 61.91 .227

CD4 cell count

HIV negative Reference Reference Reference

, 350 57.3 4.44 to 740.8 .002

351-500 14.9 0.96 to 232.78 .054

. 500 39.0 3.15 to 483.89 .004





For more information about ASCO’s conflict of interest policy,please refer to www.asco.org/rwc or jco.ascopubs.org/site/ifc.

Nicola M. ZetolaNo relationship to disclose

Surbhi GroverNo relationship to disclose

Chawangwa ModongoNo relationship to disclose

Sebathu P. ChiyapoNo relationship to disclose

Memory Nsingo-BvochoraSpeakers’ Bureau: AstraZenecaTravel, Accommodations, Expenses: Roche, AstraZeneca

Mohan NarasimhamurthyNo relationship to disclose

Lilie L. LinSpeakers’ Bureau: Varian Medical SystemsResearch Funding: Varian Medical Systems

Joseph JarvisResearch Funding: Gilead Sciences

Sanghyuk S. ShinEmployment: Beckman Coulter

Erle RobertsonNo relationship to disclose

ACKNOWLEDGEMENTWe thank Princess Marina Hospital, Gaborone Private Hos-pital, the Botswana Ministry of Health, and the Centers forDiseaseControl andPrevention for their constant support, andall our patients,, who made this study possible.

REFERENCES

references

1. Ferlay J, Shin HR, Bray F, et al: Estimates of worldwide burden of cancer in 2008: GLOBOCAN 2008. Int J Cancer 127:2893-2917, 2010

2. Dryden-Peterson S, Medhin H, Kebabonye-Pusoentsi M, et al: Cancer incidence following expansion of HIV treatmentin Botswana. PLoS One 10:e0135602, 2015

3. Ramogola-Masire D, de Klerk R,MonareB, et al: Cervical cancer prevention inHIV-infectedwomenusing the “see andtreat” approach in Botswana. J Acquir Immune Defic Syndr 59:308-313, 2012

4. Denslow SA, Rositch AF, Firnhaber C, et al: Incidence and progression of cervical lesions in women with HIV: Asystematic global review. Int J STD AIDS 25:163-177, 2014

5. Botswana Ministry of Health: Botswana Aids Impact Survey (BAIS) IV Report, 2013

6. Chabner BA, Efstathiou J, and Dryden-Peterson S: Cancer in Botswana: The second wave of AIDS in sub-SaharanAfrica. Oncologist 18:777-778, 2013

7. Murray CJ, Ortblad KF, Guinovart C, et al: Global, regional, and national incidence and mortality for HIV, tuberculosis,andmalaria during 1990-2013: A systematic analysis for the Global Burden of Disease Study 2013. Lancet 384:1005-1070, 2014

8. Botswana Ministry of Finance and Development Planning: 2011 Population and Housing Census. Gaborone, Bot-swana, Central Statistics Office, 2011

9. Cruz AT, Airewele G, and Starke JR: Tuberculosis in pediatric oncology and bone marrow transplantation patients.Pediatr Blood Cancer 61:1484-1485, 2014

10. Divya A, Irfant A, Geetha P, et al: Cancer-like presentation of female genital tuberculosis. Indian J Tuberc 61:331-335,2014

11. KumakechW, ZambleraD, and Jolaoso A: Themultifaceted presentation of tuberculosis in gynaecology: Amasqueraderas cervical cancer as well as a cause of primary infertility in the same patient. J Obstet Gynaecol 26:178-179, 2006

12. Bekele D, and Bekuretsion Y: Tuberculosis of the cervix mimicking cervical cancer. Ethiop Med J 52:87-89, 2014

13. Zhao FH, Varanasi AP, Cunningham CA, et al: Tuberculosis and oncogenic HPV: Potential co-infections in women athigh-risk of cervical cancer in rural China. Asian Pac J Cancer Prev 12:1409-1415, 2011

14. Singh N, Sumana G, and Mittal S: Genital tuberculosis: A leading cause for infertility in women seeking assistedconception in North India. Arch Gynecol Obstet 278:325-327, 2008

Affiliations

NicolaM. Zetola, Surbhi Grover, Lilie L. Lin, and Erle Robertson, University of Pennsylvania, Philadelphia,PA; Nicola M. Zetola, Surbhi Grover, andMohan Narasimhamurthy, University of Botswana; ChawangwaModongo, Botswana-University of Pennsylvania Partnership; Sebathu P. Chiyapo, Princess MarinaHospital; Memory Nsingo-Bvochora, Gaborone Private Hospital, Gaborone, Botswana; Joseph Jarvis,London School of Hygiene and Tropical Medicine, London, United Kingdom; and Sanghyuk S. Shin,University of California–Los Angeles, Los Angeles, CA.





Date post:	12-Aug-2020
Category:	Documents
Upload:	others
View:	0 times
Download:	0 times

FEBRUARY 2016 Volume 2, Issue 1 assets/JGO_2_1.pdf · 2016-12-07 · Ken G. Kornfield Emilie Gunn...

Documents