Network meta-analysis of post-exposure prophylaxis randomized clinical trials

Objectives We performed a network meta-analysis of PEP randomized clinical trials to evaluate the best regimen. Methods After MEDLINE/Pubmed search, studies were included if: (1) were randomized, (2) comparing at least 2 PEP three-drug regimens and, (3) reported completion rates or discontinuation at 28 days. Five studies with 1105 PEP initiations were included and compared ritonavir-boosted lopinavir (LPV/r) vs. atazanavir (ATV) (one study), cobicistat-boosted elvitegravir (EVG/c) (one study), raltegravir (RAL) (one study) or maraviroc (MVC) (two studies). We estimated the probability of each treatment of being the best based on the evaluation of five outcomes: PEP non-completion at day 28, PEP discontinuation due to adverse events, PEP switching due to any cause, lost to follow-up and adverse events. Results Participants were mostly men who have sex with men (n = 832, 75%) with non-occupational exposure to HIV (89.86%). Four-hundred fifty-four (41%) participants failed to complete their PEP course for any reason. The Odds Ratio (OR) for PEP non-completion at day 28 in each antiretroviral compared to LPV/r was: ATV 0.95 (95% CI 0.58-1.56; EVG/c: OR 0.65 95% CI 0.30-1.37; RAL: OR 0.68 95% CI 0.41-1.13; and MVC: OR 0.69 95% CI 0.47-1.01. In addition, the rankogram showed that EVG/c had the highest probability of being the best treatment for the lowest rates in PEP non-completion at day 28, switching, lost to follow-up or adverse events and MVC for PEP discontinuations due to adverse events. Conclusions Our study shows the advantages of integrase inhibitors when used as PEP, particularly EVG as a Single-Tablet Regimen.