Estimation of PrEP efficacy in HIV clinical trials

Vaccine and Infectious Disease Division

Estimation of PrEP efficacy in HIV clinical trials

Preexposure prophylaxis (PrEP) is one of many strategies for curbing infectious disease transmission. The goal is to treat at-risk populations with drug before they are exposed to a pathogen. In the case of HIV, PrEP uses antiretroviral agents. Several recent HIV clinical trials used PrEP to prevent HIV acquisition: for example, the Preexposure Prophylaxis Initiative trial (iPrEx), which measured the effect of daily antiretrovirals given to men who have sex with men; and the Centre for the AIDS Programme of Research in South Africa trial (CAPRISA 004), which evaluated the efficacy of per-coitus 1% tenofovir gel in high-risk South African women.

If a PrEP trial fails to demonstrate effectiveness, several reasons could explain this result: the PrEP confers no benefit to the patient, the subjects were not exposed to HIV, or the drug was not taken. The ideal situation in a clinical trial is 100 percent adherence (compliance) in taking the drug, but this just isn’t realistic. Therefore, trial statisticians must distinguish between a “drug effect” and “adherence effect.” In both the iPrEx and CAPRISA 004 trials, adherence was defined by drug detection in patients’ plasma.  

VIDD associate members Dr. James Dai and Dr. Elizabeth Brown, along with members Dr. Peter Gilbert and Dr. James Hughes, developed a causal estimand (a statistical model) and sensitivity analysis for PrEP efficacy using published adherence data from iPrEx and CAPRISA 004. Their estimation procedure assessed PrEP efficacy in HIV prevention among compliers: those who would have achieved a defined drug concentration had they been in the drug arm regardless of whether they received drug or placebo.

Using their model with iPrEx data, the authors found that the risk of infection for participants who had a detectable level of drug was reduced by 90%. CAPRISA 004 analysis was similar: ~90% reduction in risk for participants with a tenofovir concentration of 1000 ng/mL or above. This underscores the strength of the current study’s model for calculating PrEP efficacy, and will be of great value for future HIV prevention trials. These data illustrate the importance of monitoring participant compliance so researchers can take steps to increase drug adherence in future trials.

Dai JY, Gilbert PB, Hughes JP, Brown ER. Estimating the efficacy of preexposure prophylaxis for HIV prevention among participants with a threshold level of drug concentration. Am J Epidemiol. 2013 Feb 1;177(3):256-63.