Important additional details were presented at the International AIDS Conference today regarding the CAPRISA 004 tenofovir gel microbicide trial. The data provoked an emotional response from the audience, ultimately prompting a prolonged standing ovation. Salim Abdool Karim described the HSV-2 analysis, reporting that the 95% confidence intervals around the 51% efficacy result spanned 30-78% (p=.003). Several sensitivity analyses were also presented: adjustment for every known confounder (such as condom use) still showed efficacy of 47% (95% CI: 33-83%, p=.005) and considering four women with equivocal HSV-2 serology as infected only reduced the efficacy to 47% (95% CI 33-83%, p=.006).
Pharmacologist Angela Kashuba gave a talk outlining several significant associations between tenofovir concentrations and efficacy. Average drug levels in cervicovaginal fluid were significantly lower in women who acquired HIV compared to those who remained uninfected: 1 ng/mL (range: 0-290,734) compared to 520 ng/mL (0-1,338,079). The proportion of women with detectable levels was also significantly different between the two groups: 45% vs. 96%. For the HSV-2 efficacy analysis, there was a significant association between having a tenofovir level over 10,000 ng/mL; 24% of uninfected women were above this threshold compared to 6% that became HSV-2 positive. Kashuba addressed the issue of mechanism of action by stating that while the concentration of tenofovir required to inhibit HSV-2 in vitro is very high and not achievable with oral dosing, it can be obtained in tissues with gel administration.
Overall the presentations solidified the impression that the CAPRISA 004 results are robust and real. The pharmacology data is important and likely to offer crucial clues to improving the protective efficacy of the compound. The HSV-2 results may also have profound implications given the doubling in risk of HIV acquisition that is associated with HSV-2 seropositivity. Researcher Sheena McMormack nevertheless cautioned that there are some limitations; the lower bound of the confidence interval for the HIV result, at 6%, is not as high as is ideally desirable and CAPRISA 004 was a relatively small trial in a single population. But McMormack also stressed the need for rapid confirmation of the findings, because – as Salim Abdool Karim pointed out – the intervention could have the ability to prevent 1.3 million new infections and 800,000 deaths in South Africa alone.
Comments