Contraception & HIV Still searching for answers after >2 - PowerPoint PPT Presentation

Contraception & HIV Still searching for answers after >2 decades Inter –CFAR Symposium on HIV R Scott McClelland, MD, MPH Research in Women University of Washington September 20 th 2012

Overview • Global use of contraception & unmet need • Hormonal contraception & HIV acquisition • Hormonal contraception & HIV transmission • Hormonal contraception & progression of HIV disease • Current guidelines • Where do we go from here?

Global Use of Contraception • An estimated 800 million women use modern methods of contraception – 150 million use hormonal contraceptives – Others use IUDs, barriers, tubal ligation Population Reference Bureau, 2008

Reported Contraceptive Use and Unmet Need in Married Women Unmet Region 1990 2007 Need Asia 57% 67% 9% Latin 62% 72% 11% America Africa 17% 28% 22%

Unmet Contraceptive Needs Ref: Network Vol 23, number 3, 2004. Family Health International • Globally, 200 million couples not using contraception despite wanting to space or limit child bearing (WHO)

Benefits of Contraception • Avoid unplanned pregnancies – Potential for obstetrical complications – Spacing pregnancies benefits the health of women, infants, and children • Family planning empowers women, reducing gender inequality • HIV-positive women may wish to reduce risk of vertical transmission by preventing further pregnancies

J Infect Dis 1991 AOR 4.5 (95% CI 1.4-13.8) Non-OC Users OC Users

Does hormonal contraceptive use increase the risk of HIV acquisition in women?

Bradford Hill Criteria for Causation • Strength of association • Consistency • Specificity • Temporal relationship • Biological gradient • Plausibility • Experiment • Coherence • Analogy

• 1314 HIV-serodiscordant couples in which the female partner was HIV-negative • Participants in Partners in Prevention trial from seven countries in Africa • Used marginal structural modeling and Cox proportional hazards regression to assess effect of contraceptive use on HIV acquisition Lancet Infect Dis 2012

HC Use and HIV Acquisition Incidence Cox Model per 100 p-y AHR (95% CI) No HC 3.78 Reference Any HC 6.61 1.98 (1.06-3.68) Injectable 6.85 2.05 (1.04-4.04) Oral 5.94 1.80 (0.55-5.82) Heffron et al. Lancet Infect Dis 2012

• 5567 HIV-negative women • Participants in the Carraguard Phase 3 Efficacy Trial in South Africa • Used marginal structural models and Cox models to assess effect of contraceptive use on HIV acquisition – Overall and in subset 16-24 years old AIDS 2012

HC Use and HIV Acquisition Incidence Marginal Structural Models per 100 All women Women <24 y.o. p-y AHR (95% CI) AHR (95% CI) No HC 3.4 Reference Reference DMPA 4.6 1.28 (0.92-1.78) 1.68 (0.96-2.94) Net-En 3.5 0.92 (0.64-1.32) 1.36 (0.78-2.35) COC 2.8 0.84 (0.51-1.39) 1.02 (0.46-2.28) Morrison et al. AIDS 2012

• 2236 HIV-negative women • Participants screened for the PRO 2000/5 trial in South Africa • Used Cox proportional hazards regression AIDS 2012

HC Use and HIV Acquisition Cox Model AHR (95% CI) No HC Reference Injectable 2.02 (1.37-3.00) Oral 0.95 (0.62-1.46) Wand and Ramjee. AIDS 2012

Unprotected Sex Hormonal HIV Contraception Acquisition

• Sub-cutaneous progesterone implants enhanced SIV vaginal transmission in macaques 7-fold compared to placebo – Thinning of vaginal epithelium – Increased number of SIV DNA-positive cells in vaginal lamina propria – Higher plasma SIV RNA in first 3 months Nature Med 1996

SIV Vaginal Transmission Marx et al. Nature Med 1996

Blish. Am J Reprod Immunol 2011

Bradford Hill Criteria for Causation • Strength of association • Consistency • Specificity • Temporal relationship • Biological gradient • Plausibility • Coherence • Experiment • Analogy

Can we disentangle the relationships between HC and HIV Acquisition?

Experimental Design • Is it possible to conduct a randomized trial of hormonal contraception versus some control?

To trial or not to trial… Pros Cons • Potential for providing a • Difficult to identify a clear answer suitable control condition • Potential for clarifying the • Frequent switching of biological mechanisms contraceptive method • Added complexity of HIV prevention trials in setting of proven interventions • Cost: most effective use of resources?

BJOG 2012 • Decision analysis modeling risk of HIV acquisition and maternal mortality • Chad, Kenya, South Africa, Uganda • Assumes injectable progesterone associated with 2.19-fold increased HIV risk • Contraceptive methods (UN), failure rates ( Trussel Contracept 2011 ), maternal mortality & pregnancy outcome (WHO)

• If progesterone injectables removed without 70-100% of women switching to IUD or COC, there could be up to 9 additional maternal deaths per HIV infection averted Rodriguez et al. BJOG 2012

Contraception 2012 • Modeled risks of unwanted births, maternal deaths, and HIV acquisition associated with different contraceptive methods • Kenya, South Africa, Zimbabwe • Method-specific HIV risk for family planning, sex workers, and discordant couples 1-4 • Contraceptive methods (UN), failure rates ( Trussel 2007 ), maternal mortality & pregnancy outcome (WHO) 1. Heffron Lancet ID 2011, 2. Baeten AIDS 2007, 3-4. Morrison AIDS 2007 & 2010

Additional unwanted births per 100 HIV infections averted Jain. Contraception 2012

Additional maternal deaths per 100 HIV infections averted Jain. Contraception 2012

Does hormonal contraceptive use increase the risk of HIV transmission from women to men?

BMJ 1992 • 156 female index patients with 159 HIV- negative male partners – 26/114 (23%) of women who reported on contraceptive use were using COCs • 19 (12%) male partners infected • Reported no association between hormonal contraceptive use and transmission to male partner

• 2,476 HIV-serodiscordant couples in which the male partner was HIV-negative • Participants in Partners in Prevention trial from seven countries in Africa • Used marginal structural modeling and Cox regression to assess effect of contraceptive use on HIV transmission – Genetically linked HIV transmission events Lancet Infect Dis 2012

HC Use and HIV Transmission Incidence Cox Model per 100 p-y AHR (95% CI) No HC 1.51 Reference Any HC 2.61 1.97 (1.12-3.45) Injectable 2.64 1.95 (1.06-3.58) Oral 2.50 2.09 (0.75-5.84) Heffron et al. Lancet Infect Dis 2012

HC Use and Genital HIV Shedding Log10 Adjusted coefficient † copies/swab Median* (IQR) (95% CI) No HC 3.14 (2.08-3.85) Reference Any HC 3.29 (2.08-3.91) 0.14 (0.04 to 0.23) Injectable 3.38 (2.08-4.02) 0.19 (0.08 to 0.30) Oral 2.96 (2.08-3.65) -0.05 (-0.24 to 0.14) * Log10 HIV RNA copies/swab † Average difference in HIV RNA concentration Heffron et al. Lancet Infect Dis 2012

Female to Male HIV Transmission in Rakai, Uganda • 224 male HIV-/female HIV+ couples • Excluded intervals with condom use Incidence aIRR (95% CI) per 100 PY Non-HC 7.0 Reference Any HC 6.9 1.04 (0.43-2.48) OC 10.6 1.42 (0.30-6.57) Injectables 5.9 0.92 (0.34-2.50) Lutalo et al. CROI 2012

Does hormonal contraceptive use increase the rate of disease progression in HIV+ women?

Hormonal Contraception and HIV Progression Higher Risk Neutral Risk Lower Risk 1. Lavreys JID 2004 1. Kilmarx JID 2000 1. Polis AIDS 2010 2. Stringer AJOG 2007 2. Cejtin AIDS 2003 2. Heffron CROI 2012 3. Stringer AIDS 2009 3. Wang AIDS 2004 4. Richardson AIDS 2007 5. Allen J Women’s Health 2007 6. Stringer AIDS 2009 7. Morrison JAIDS 2011

• 599 postpartum women in Lusaka, Zambia • RCT IUD vs. hormonal contraception • Followed for at least 2 years • Primary endpoint: Safety and efficacy of intrauterine device vs. hormonal contraception • Secondary endpoints: – Time to CD4<200 – Time to death – Combined endpoint: time to CD4<200 or death Am J Obstet Gynecol 2007

HC vs. IUD in HIV-positive Women • Only one episode of PID in IUD arm • Higher rate of discontinuation in IUD arm • Pregnancy higher in HC vs. IUD – HR 2.2 (95% CI 1.2-2.4) • Mortality did not differ significantly – HR 1.4 (95% CI 0.7-3.0) • Faster progression to CD4<200 with hormonal contraception compared to IUD – HR 1.6 (95% CI 1.04-2.03)

HC vs. IUD in HIV-positive Women CD4 decline CD4 decline or death •Time to CD4<200 or death faster in HC vs. IUD Death • HR 1.6 (95% CI 1.1-2.3)

HC vs. IUD in HIV-positive Women • IUD appeared safe and was more effective than HC at preventing pregnancy • No difference in mortality • Time to CD4<200 longer with IUD vs. HC • Limitations – ~30% withdrew or lost to follow-up – ~30% discontinued allocated method – Generalizability? (Post-natal population)