Treatment Is Prevention: The Results of HPTN 052
Among HIV-serodiscordant couples, treatment of the infected partner decreased the risk for transmission to the uninfected partner by 96%.
Combination antiretroviral therapy (ART) has long been known to drive viral replication to levels below the limit of detection, leading to significant improvements in patients' health. Given that ART also decreases the amount of HIV present in genital secretions, experts have speculated that it should also reduce sexual transmission of HIV. Several observational studies — most recently, the Partners in Prevention HSV/HIV Transmission Study (JW AIDS Clin Care Jun 7 2010) — have supported this hypothesis, but only a randomized, controlled trial could provide definitive proof. Enter the HPTN 052 study.
The results of this trial were released in May 2011, after the data and safety monitoring board (DSMB) concluded that the study had demonstrated its goal (JW AIDS Clin Care May 16 2011). Now, they are being published in the New England Journal of Medicine and simultaneously presented at this year's International AIDS Society meeting in Rome.
The HPTN 052 study involved 1763 HIV-serodiscordant couples in nine countries (5 in Africa, 2 in Asia, and 2 in the Americas). Per inclusion criteria, all the HIV-infected partners had baseline CD4 counts between 350 and 550 cells/mm3, and all were treatment naive (except for possible ART to prevent perinatal transmission). They were randomized 1:1 to receive early ART at enrollment or to delay therapy until their CD4 counts were 250 cells/mm3 on two consecutive measurements or they developed an AIDS-related illness. Nearly all the couples were heterosexual (97%), and half the HIV-infected participants were men. The median baseline CD4 count was 442 cells/mm3 in the early-therapy group and 428 cells/mm3 in the delayed-therapy group. AZT/3TC/efavirenz was the most commonly used antiretroviral regimen in the study.
At the time of the DSMB review, 90% of couples remained in the study, with a median follow-up of 1.7 years; 89% of participants in the early-therapy group had viral loads below 400 copies/mL. A total of 39 HIV transmission events occurred during follow-up: 4 in the early-therapy group and 35 in the delayed-therapy group. In 28 of these cases, genetic testing confirmed that the HIV-infected partner in the study was the source of infection: Only 1 of these cases was in the early-therapy group versus 27 in the delayed-therapy group, which translated to a 96% risk reduction with early ART.
Of interest, 17 of the 28 linked infections occurred when the infected partner had a CD4 count >350 cells/mm3, and 23 were among African couples. A high plasma viral load at baseline increased the likelihood of transmission, whereas consistent condom use decreased the likelihood.
The clinical benefits of early ART were also evident, with the delayed-therapy group experiencing significantly more clinical events (primarily, extrapulmonary tuberculosis).
Comment: This study conclusively demonstrates that treatment of HIV infection decreases the risk for sexual transmission and improves the lives of those infected, regardless of CD4-cell count. Clearly, the study has important policy implications. Is it time to recommend treatment for HIV-infected individuals in stable serodiscordant couples in order to prevent infection of their partners? I think the evidence is now quite strong (AI evidence) and we should therefore move in that direction. I also think it is probably time to stop guiding ART by CD4-cell count and just treat anyone who is willing to be treated — not only for their own health but also for the larger public health benefit.
Published in Journal Watch HIV/AIDS Clinical Care July 19, 2011
Cohen MS et al. Prevention of HIV-1 infection with early antiretroviral therapy. New Engl J Med 2011 Jul 18; [e-pub ahead of print]. (http://dx.doi.org/10.1056/NEJMoa1105243)
Hammer SM. Antiretroviral treatment as prevention. New Engl J Med 2011 Jul 18; [e-pub ahead of print]. (http://dx.doi.org/10.1056/NEJMe1107487)
To ensure that your Reader Remark is not formatted as one long paragraph, precede new paragraphs with either a blank line or an indentation.