There is a study in the recent issue of the Lancet assessing the role of anti-retroviral therapy to prevent HIV transmission in sero-discordant couples in China. This was an observational, retrospective study of almost 40,000 discordant couples with over 100,000 person years of followup. China was adhering to the general WHO guidelines of initiating retroviral therapy in people with CD4 counts of 350 or less. Their level of HIV transmission was compared to HIV discordant couples in which the HIV positive person had higher CD4 counts and were not on therapy. ( http://dx.doi.org/10.1016/S0140-6736(12)61898-4). Couples with inadequate followup visits were excluded from analysis. Treated patients were older than non-treated patients. Patients were followed up to 8 years, though the median duration of followup was only 1.2 years for non-treated couples and 2.4 years in those on therapy. Results:
–Total number of seroconversions in partners = 1631
–Rate of HIV transmission to the non-HIV partner was 2.6 per 100 person-years among those not taking antiretroviral medications (average CD4 count of 441)
–For those on treatment (average CD4 count of 168), rate of HIV transmission was 1.3 per 100 person-years\
–The corrected, calculated relative risk reduction in HIV transmission was 26%, and this was true for all demographic subgroups, was significant in the first year (though not thereafter), and was significant in couples where the HIV positive partner was infected by transfusion or heterosexual intercourse. However, the results were nonsignificant when the HIV positive partner was infected by injecting drugs.
The study provides a lot of support for the concept of treatment as prevention. There are other observational data associating decreased transmission with increased ambient use of medication, such as in San Francisco, Vancouver and S Africa. There also are data suggesting that prescribing antiretroviral drugs in HIV-positive pregnant women and breast-feeding mothers decreases mother to child transmission. One issue, however, in this Chinese study is that only patients with more advanced disease, with a CD4 count of less than 350, received therapy. There is likely a bias here, and that the treated patients were older and more likely to be sicker, and may therefore have had fewer sexual partners, be more likely to use protection, and, thereby be in a lower risk category for transmission. Although we do not know medication adherence or effect on viral loads, this study is a real-world, community-based study and is therefore perhaps more generalizable than academic-centered randomized controlled trials. For example, there was a recent report of a randomized trial (HPTN 052, N Engl J Med 2011;365:493-505),in which 1800 discordant couples from 9 countries (>50% from Africa) with HIV-infected partner having CD4 of 350-550 were randomized to immediate therapy vs delayed until either significant drop in CD4 or symptoms. In this study there were 39 HIV transmissions to partners, of which 28 were linked to the specific infected partner, and only 1 was in the early treatment group (96% risk reduction with early treatment). All told , the overall conclusion that treatment as prevention seems valid. In fact, treating at an earlier stage in sero-discordant couples, independent of CD4 count as suggested by the WHO initiative (which I sent out last week), may have even more impressive data for treatment as prevention. This all reinforces the US approach of HIV testing universally with more aggressive treatment strategies, as well as a more aggressive treatment strategy internationally.
geoff