Although the VOICE trial failed to show that tenofovir vaginal gel protects against HIV-1 infection in women — likely due to poor adherence to the study drug — new analyses suggest the product did guard against infection among the women who used it.
“Use of antiretroviral (ARV) drugs in pre-exposure prophylaxis (PrEP) has been shown to be protective against the risk of sexual transmission of HIV in four clinical trials,” Jeanne M. Marrazzo, MD, MPH, professor of medicine at the University of Washington, and colleagues wrote. “Based on these results, daily use of [Truvada (emtricitabine/tenofovir disoproxil fumarate, Gilead Sciences; FTC/TDF)] ... was approved for HIV prevention by the FDA in July 2012
“Unexpectedly, two PrEP trials using the same agent in African women, FEM-PrEP and VOICE (MTN 003), reported futility in intent-to-treat analyses,” the researchers said. “These seemingly contradictory findings are very likely explained by different degrees of adherence achieved in trials and differences in levels of adherence to daily oral dosing that is required for protection between heterosexual women and men who have sex with men, emphasizing the critical relationship between adherence and PrEP efficacy in HIV prevention.”
Jeanne M. Marrazzo
VOICE, conducted from 2009 to 2012, included more than 5,000 women (mean age, 25.3 years) from 15 sites in South Africa, Uganda and Zimbabwe. The randomized, placebo-controlled trial assessed the efficacy of daily use of oral Viread (tenofovir disoproxil fumarate, Gilead Sciences; TDF), oral FTC-TDF and 1% vaginal tenofovir (TFV) gel.
Measurements of adherence through self-reports and the number of returned pills or gel applicators were generally overestimated, according to the researchers. For example, less than 40% of women had detectable levels of TFV in plasma compared with 90% who reported product use.
“In addition to providing a more accurate measure of adherence, pharmacological evidence of product use offers the possibility of estimating causal prevention effect among adherers under suitable assumptions, even if the primary intent-to-treat analyses yields null results on effectiveness,” the researchers wrote.
However, investigating the protective effect of TFV against HIV infection in the minority of women who adhered to the drug during follow-up is problematic, Marrazzo and colleagues said, because drug levels in plasma can only be measured in women belonging to the intervention arm, not the placebo arm, offering no comparable group of adherers. Moreover, any comparisons used to examine the causal relationship between TFV use and HIV infection are compromised by potential confounding.
“In essence, confounding in this setting means characteristics associated with adherence can also be directly related to the risk for HIV infection itself,” the researchers wrote.
Therefore, Marrazzo and colleagues developed a novel approach to assess whether the adjustment for confounding variables in regression models — an approach used in previous PrEP trials — effectively removed confounding. To do so, they based their analysis on two proxy measures of adherence in the VOICE trial: TFV detected in plasma at least once during follow-up, and detection of the drug at 3 months of follow up.
Their results showed that after adjusting for baseline predictors of HIV risk, there was no residual confounding when comparing adherers in the active drug arm and those in the placebo group. According to the researchers, the RR for HIV prevention among women who had detectable TFV in plasma at least once during follow-up is 0.53 (P = .038). The RR among those with detectable levels of the drug at 3 months is 0.4 (P = .045).
“While we used the VOICE study as an illustrative example, this approach should be generally applicable to placebo-controlled PrEP trials with drug detection as a proxy measure of adherence in the active product arm,” Marrazzo and colleagues wrote. – by David Jwanier and John Schoen
Disclosure: The researchers report no relevant financial disclosures.