Weighing 17 years of evidence: Does hormonal contraception increase HIV acquisition risk among Zambian women in discordant couples? Kristin M. Wall, PhD Department of Epidemiology Rwanda Zambia HIV Research Group Emory University, Atlanta, GA, USA

Conflict of Interest Disclosure The authors have no conflicts of interest due to financial or personal relationships that might be perceived to cause bias.

Background Hormonal contraception Prevents unintended pregnancy 1 –Prong 2 PMTCT for HIV+ women 1 Is widely used in high HIV prevalence areas 2 Use among married women in Zambia 3 : 11% OCP9% Injectable0.4% Implant

Background Extant evidence is conflicting World Health Organization technical meeting Medical Eligibility Criteria (MEC) Category 1 Emphasized dual method use for high-risk women using progestogen-only injectables 1

Study population M+F- serodiscordant couples Identified from couples’ voluntary HIV counseling and testing services in Lusaka from >16 years of age Male partner was not on ART Followed 3-monthly at the research site –Contraception methods provided –HIV testing of negative partners

Contraceptive exposures Hormonal methods: –Implant (Norplant, Jadelle) –Injectable (150 mg IM DMPA) –Combined oral contraceptive pills (OCPs) Non-hormonal control: –No method –Condoms –Copper intrauterine device (IUD) –Permanent methods

HIV infection outcomes Time to any HIV infection –Genetically linked or unlinked to the study partner Time to linked HIV infections

Analyses Multivariate Cox models Potential effect-measure modifiers: –Genital ulceration, inflammation, VL, fertility intentions, age

Analyses Multivariate Cox models Potential effect-measure modifiers: –Genital ulceration, inflammation, VL, fertility intentions, age Sensitivity analyses explored effects of: Method exposure/control categorizations Cumulative exposure Misclassification of unprotected sex Time-dependent confounding –Marginal structural models

Non- seroconverting couples (N = 1141) Seroconverting couples (N=207 linked & 45 unlinked) Total (N = 1393 M+F- couples) 82%18%

Non- seroconverting couples (N = 1141) Seroconverting couples (N=207 linked & 45 unlinked) Total (N = 1393 M+F- couples) 82%18% Baseline contraceptive method None/condoms alone 72%74% OCPs 11%12% Injectables 10% Implant 3% IUD 2%1% Permanent method 1%0%

Seroconversion rates among HIV-women in discordant relationships by method of contraception METHOD # infections CY Incidence /100 CY 95%CI p-value (2-tail) None/condoms ref OCPs

Seroconversion rates among HIV-women in discordant relationships by method of contraception METHOD # infections CY Incidence /100 CY 95%CI p-value (2-tail) None/condoms ref OCPs Injectables

Seroconversion rates among HIV-women in discordant relationships by method of contraception METHOD # infections CY Incidence /100 CY 95%CI p-value (2-tail) None/condoms ref OCPs Injectables Implant

Multivariate models of time to ANY HIV seroconversion METHOD aHR*95%CIp-value (2-tail) Non- hormonal ref Implant Injectables OCPs *Controlling for: Woman's age (per year increase) Woman's literacy in Nyanja Sperm present on a wet prep during interval Genital ulceration of woman in past 3 months Genital inflammation of woman in the past 3 months

Multivariate models of time to ANY HIV seroconversion METHOD aHR*95%CIp-value (2-tail) Non- hormonal ref Implant Injectables OCPs *Controlling for: Woman's age (per year increase) Woman's literacy in Nyanja Sperm present on a wet prep during interval Genital ulceration of woman in past 3 months Genital inflammation of woman in the past 3 months  Findings remained the same when controlling for pregnancy and/or fertility intentions

Results Similarly, the results of: Multivariate models of linked infections only –Additionally controlling for baseline pregnancy, couples’ unprotected sex in past 3 months, genital ulceration and inflammation of male partner in past 3 months, and VL All sensitivity analyses Marginal structural models did not indicate any significant increase in HIV risk for hormonal contraception users.

Measures of unprotected sex by method of contraception 3-monthly measure of: Non- hormonal % OCPs % p-value (2- tailed)* Incident pregnancy 3%5% ^ Unprotected sex29%37%^ Sperm on wet prep 6%8%^ ^p-value <0.05 OCP vs. non-HC #p-value <0.001 injectables vs. non-HC &p-value <0.001 non-HC vs. implant

Measures of unprotected sex by method of contraception 3-monthly measure of: Non- hormonal % OCPs % Injectable % p-value (2- tailed)* Incident pregnancy 3%5%1% # Unprotected sex29%37%34%# Sperm on wet prep 6%8%6% ^p-value <0.05 OCP vs. non-HC #p-value <0.001 injectables vs. non-HC &p-value <0.001 non-HC vs. implant

Measures of unprotected sex by method of contraception 3-monthly measure of: Non- hormonal % OCPs % Injectable % Implant % p-value (2- tailed)* Incident pregnancy 3%5%1%0% & Unprotected sex 29%37%34%18%& Sperm on wet prep 6%8%6%2%& ^p-value <0.05 OCP vs. non-HC #p-value <0.001 injectables vs. non-HC &p-value <0.001 non-HC vs. implant

Measures of unprotected sex by pregnancy status 3-monthly measure of: Not pregnant % Pregnant % p-value (2-tailed)* No. times unprotected sex (mean) 26^ Any unprotected sex 28%54%^ Sperm on wet prep 6%10%^ ^p-value <0.001 pregnant vs. not pregnant

Conclusions & Recommendations We found no association between hormonal contraception and HIV acquisition risk in women Reinforced condom counseling is needed during: Oral contraceptive pill use Injectable use Pregnancy These findings: Add to a controversial literature Are important when evaluating MEC Categories

Acknowledgements Rwanda Zambia HIV Research Group (RZHRG) Contributors William Kilembe Htee Khu Naw Ilene Brill Bellington Vwalika Elwyn Chomba Brent Johnson Lisa Haddad Amanda Tichacek Susan Allen Zambian Ministry of Health & District Health Management Team Study Participants & Clinic Staff

Funding National Institutes of Child Health and Development (NICHD RO1 HD40125) National Institute of Mental Health (NIMH R01 66,767) AIDS International Training and Research Program Fogarty International Center (D43 TW001042) Emory Center for AIDS Research (P30 AI050409) National Institute of Allergy and Infectious Diseases (NIAID R01 AI51231; NIAID R01 AI040951; NIAID R01 AI023980; NIAID R01 AI64060; NIAID R37 AI51231) US Centers for Disease Control and Prevention (5U2GPS000758) International AIDS Vaccine Initiative This study was made possible by the generous support of the American people through the United States Agency for International Development (USAID). The contents are the responsibility of the International AIDS Vaccine Initiative and do not necessarily reflect the views of USAID or the United States Government. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Seroconversion rates among HIV-negative women in discordant relationships by method of contraception METHOD # infections CY Incidence /100 CY 95%CI p-value (2-tail) Total None/condoms ref OCPs Injectables IUD Implant Permanent