CROI 2016 Abstract eBook

Abstract Listing

Poster Abstracts

Cancer'versus'Tuberculosis'Mortality'among'HIV8infected'individuals'in'Botswana ! Scott%Dryden,Peterson, 1,2,3 %Gita%Suneja, 4 %Heluf%Medhin, 5 %Memory%Bvochora,Nsingo, 6 %Mukendi%Kayembe, 5,7 %Neo% Tapela, 1,2,5 %%Shahin%Lockman 1,2,3% 1% Botswana%Harvard%AIDS%Institute%Partnership,%Gaborone,%Botswana;% 2% Brigham%and%Women’s%Hospital,%Boston,% MA;% 3% Harvard%T.H.%Chan%School%of%Public%Health,%Boston,%MA;% 4% University%of%Utah,%Salt%Lake%City,%UT;% 5% Botswana% Ministry%of%Health,%Gaborone,%Botswana;% 6% Department%of%Oncology,%Gaborone%Private%Hospital,%Gaborone,% Botswana;% 7% National%Health%Laboratory,%Gaborone,%Botswana%

615 Cancer Versus Tuberculosis Mortality Among HIV-Infected Individuals in Botswana Scott Dryden-Peterson 1 ; Gita Suneja 2 ; Heluf Medhin 3 ; Memory Bvochora-Nsingo 4 ; Mukendi K. Kayembe 5 ; NeoTapela 3 ; Shahin Lockman 1 1 Brigham and Women’s Hosp, Harvard Med Sch, Boston, MA, USA; 2 Univ of Utah Sch of Med, Salt Lake City, UT, USA; 3 Botswana Ministry of Hlth, Gaborone, Botswana; 4 Gaborone Private Hosp, Gaborone, Botswana; 5 Natl Hlth Lab, Gaborone, Botswana Background: With declining mortality due to infections, cancer has become the leading cause of death among HIV-infected individuals in high-income countries. However, in sub-Saharan Africa where more than two-thirds of HIV infections occur, the relative contribution of infection and malignancy to HIV-associated mortality is unknown. We sought to estimate cancer mortality since availability of antiretroviral treatment (ART) in Botswana, and compare with mortality due to tuberculosis (TB), the leading infectious cause of death in sub-Saharan Africa. Methods: Incidence by cancer type was estimated from 8479 incident cases from the Botswana National Cancer Registry during the period of ART expansion, 2003-2008. We utilized Poisson regression in an inverse probability weighted population with known HIV status and projected cancer incidence through 2013. Cancer mortality was estimated using parametric Weibull models from observed survival in a separate prospective cancer cohort in Botswana (2010-2015). Survival probabilities for each cancer type were assumed to be constant during the study period and all deaths were attributed to cancer. We utilized estimates from the WHO|Global TB Program (derived from Botswana government data) to estimate TB-HIV deaths. Results: A total of 808 patients with HIV and cancer followed for median of 12.2 months (IQR 6.1 to 24.3 months) contributed to survival estimates (1.2% loss-to-followup). Estimated 5-year survival was low: cervix 3.9%, head and neck 4.4%, breast 19.3%, non-Hodgkin lymphoma 39.7%, Kaposi sarcoma 52.1%, and combined other sites 15.1%. Mortality due Kaposi sarcoma declined over the study period (-4.2% annually, 95% CI -5.0 to -3.3%), but cervical cancer mortality increased (13.3% annually, 95% CI 11.7 to 14.9). Overall cancer mortality increased (1.2% annually, 95% CI 0.7 to 3.1%) while TB mortality declined substantially between 2003 and 2013. In 2013, projected cancer mortality (293 per 100,000, 95%CI 264-331) approximated projected TB mortality (324 per 100,000, estimate range 241-419). Conclusions: With ART coverage exceeding 90%, mortality due to cancer in HIV-infected individuals has increased in Botswana and now likely exceeds mortality due to TB. Cervical cancer mortality is rising sharply. Interventions to reduce cancer risk, establish screening programs, and improve access to treatment are urgently needed for HIV-infected individuals. BACKGROUND.'' With!declining!mortality!due!to!infec ions,!c ncer!has!become!the!leading!cause!of!death! among!HIV9infected!individuals!in!high9income!countries.!However,!in!sub9Saharan!Africa!where!more!than!two9 thirds!of!HIV!infections!occur,!the!relative!contribution!of!infection!and!malignancy!to!HIV9associated!mortality!is! unknown.!We!sought!to!estimate!cancer!mortality!since!availability!of!antiretroviral!treatment!(ART)!in! Botswana,! nd!compare! ith!mortality!due!to!tuberculosis!(TB),!the!leading!infectious!cause!of!death!in!sub9 Saharan!Africa. ' METHODS.'' Incidence!by!cancer!type!was!estimated!from!8479!incident!cases!from!the!Botsw a!National! Cancer!Registry!during!the!period!of!ART!expansion,!200392008.!!We!utilized!Poisson!regression!in!an!inverse! probability!weighted!population!with!known!HIV!status!and!projected!cancer!incidence!through!2013.!!Cancer! mortality!was!estimated!using!parametric!Weibull!models!from!observed!survival!in!a!separate!prospective! cancer!cohort!in!Botswana!(201092015).!!Survival!probabilitie !for!each!ca cer!type!wer !a sumed!to!be! constant!during!the!study!period!and!all!deaths!were!attributed!to!cancer.!!We!utilized!estimates!from!the! WHO|Global!TB!Program!(derived!from!Botswana!government!data)!to!estimate!TB9HIV!deaths. ' RESULTS.'' A!total!of!808!patients!with!HIV!and!cancer!followed!for!median!of!12.2!months!(IQR!6.1!to!24.3! months)!contributed!to!survival!estimates!(1.2%!loss9to9followup).!!Estimated!59year!survival!was!low:!cervix! 3.9%,!head!and!neck!4.4%,!breast!19.3%,!non9Hodgkin!lymphoma!39.7%,!Kaposi!sarcoma!52.1%,!and!combined! other!sites!15.1%.!!Mortality!due!Kaposi!sarcoma!declined!over!the!study!period!(94.2%!a nually,!95%!CI!95.0!to!9 3.3%),!but!cervical!cancer!mortality!increased!(13.3 !annually,!95%!CI!11.7!t !14.9).!Overall!cancer!m rt lity! increased!(1.2%!annually,!95%!CI!0.7!to!3.1%)!while!TB!mortality!declined!substantially!between!2003!and!2013.!! In!2013,!projected!cancer!mortality!(293!per!100,000,!95%CI!2649331)!approximated!projected!TB!mortality!(324! per!100,000,!estimate!range!2419419). ' CONCLUSIONS. !!With!ART!coverage!exceeding!90%,!mortality!due!to!cancer!in!HIV9infected!individuals!has! increased!in!Botswana!and!now!likely!exceeds!mortality!due!to!TB.!!Cervical!ca cer!mortality!is!rising!sharply.!! Interventio s!to!reduce!canc r!risk,!establish!scr ening!programs,!and!imp ove!access!to!t eatment!are!urg ntly! needed!for!HIV9infected!individuals.! !

Poster Abstracts

'

616 Excess Mortality Rates Among HIV-Infected Cancer Patients in the United States Anna Coghill ; Ruth Pfeiffer; Meredith Shiels; Eric Engels NCI, Rockville, MD, USA

Background: We have recently reported that HIV-infected cancer patients have a higher risk of dying from their cancer than their HIV-uninfected counterparts, providing evidence that immunosuppression plays a role in cancer progression. However, the magnitude of excess mortality resulting from this association among US HIV-infected cancer patients remains unclear. In the present study, we examine whether patients diagnosed with both HIV and cancer die at higher rates than expected based on the mortality rates for HIV and cancer individually.

249

CROI 2016

Made with FlippingBook - Online catalogs