Zero HIV Infections Seen Through Condomless Sex When HIV-Positive Partner Is on Effective Treatment

Zero cases of HIV transmission were observed in mixed-status couples when the HIV-positive partner was on treatment with a viral load below 200, according to a study presented by Alison Rodger, M.D., at CROI 2014 in Boston. However, the result comes with many caveats and does not condone condomless sex, nor does it indicate that transmission is impossible with effective treatment.

The trial, known as the PARTNER study, followed 767 mixed-status couples in 75 sites across 14 European countries. The couples provided informed consent, with the negative partner knowing about the positive partner's status and risk of HIV transmission, as well as being educated about the need for consistent condom use.

To be eligible for follow-up, the couples had to report condomless sex, not using PrEP (pre-exposure prophylaxis) or PEP (post-exposure prophylaxis) and having a viral load below 200 within the last year. Data was collected every six months, in the form of a confidential risk-behavior questionnaire, along with a viral load test for the positive partner and an HIV test for the negative partner. Overall, the participants provided 894 eligible couple-years of follow up (CYFU), with an estimated total number of condomless sex acts at about 44,400.

The study included 445 heterosexual couples and 282 men who have sex with men (MSM) couples. Out of the positive partners, the median age was around 42, with at least 94% reporting treatment adherence over 90% and at least 85% reporting an undetectable viral load.

For the negative partners, the median age was also around 42, with reports of having consistent condomless sex for a median of about 2.5 years. For the men and women in heterosexual couples, diagnosis with a sexually transmitted infection (STI) other than HIV was at 5% and 6%, respectively, while reported condomless sex with partners outside of the relationship was 3% and 4%. However, for the men in MSM couples, STI diagnosis was at 16% and reported condomless sex with other partners was 34%.

Surprisingly, there were zero linked cases of HIV transmission within couples. However, here's where the numbers and statistics get a little tricky. While the estimated rate of HIV transmission within couples (per 100 CYFU) for any type of sex was zero, the confidence interval (CI) was 95% and the upper bound for this interval was 0.4, translating into a 10-year risk of about 4%.

Similarly, while the estimated rate of HIV transmission within couples (per 100 CYFU) for anal sex was zero, the confidence interval (CI) was 95% and the upper bound for this interval was 0.96, translating into a 10-year risk of about 10%.

Furthermore, looking at type of sex reported by the negative partner, all estimated rates of HIV transmission within couples (per 100 CYFU) were zero. However, the upper bounds of these confidence intervals varied:

  • For women, vaginal sex with ejaculation: 1.88.
  • For men, vaginal sex: about 1.32.
  • MSM insertive anal sex: about 1.37.
  • MSM receptive anal sex without ejaculation: about 2.5.
  • MSM receptive anal sex with ejaculation: 4, translating into a 10-year risk of about 32%.

It's also important to note that during the study, HIV transmissions that were not phylogenetically linked to the positive partner did occur through partners outside of the relationship. According to Rodger, the total number of actual transmissions will remain blinded until the end of the study, and further research will be conducted to focus on the risk of transmission specifically among MSM.

As a reminder, the upper limits of risk remain uncertain, particularly over receptive anal sex with ejaculation, so while these results are promising, they require further investigating and do not condone condomless sex.

[Editor's note: To help better understand the statistics: Joel Gallant, M.D., M.P.H., writes in his blog, "While the transmission rate was zero in this study, the presenter made a point of discussing the uncertainty around that number. [Rodger] pointed out that while their best estimate for the transmission rate is zero, they can’t exclude the possibility, with 95% certainty, that the true overall transmission rate was up to 4% over 10 years, and that the true transmission rate for receptive anal sex with ejaculation was up to 10% over 10 years. This is basically a statistical way of saying that you can’t prove a negative. As the study accumulates more couples, the statistical uncertainty will diminish, and if they continue to see no transmissions, they will be able to say with greater certainty that the estimated risk is zero."]