A Case for Ending Rifampicin Use in Endocarditis

Background: Rifampicin has been employed in various endocarditis antibiotic regimens, particularly in prosthetic, staphylococcus aureus (SA) and culture-negative cases. However, this practice is not without controversy. Purpose: To perform a meta-analysis aimed at ascertaining the efficacy and safety of adjunctive rifampicin use in endocarditis. Methods: We systematically searched MEDLINE, Embase, Web of Science, Cochrane Library and Google Scholar, from inception to June 1st, 2021, for studies comparing rifampicin-containing antibiotic regimens with those comprising no rifamycin derivatives, in the setting of endocarditis. The primary endpoint was all-cause death, while relapse, bail-out valvular surgery, microbiological failure (lasting positivity of blood cultures) and hepatotoxicity (fivefold or higher increase in transaminase levels) were secondary outcomes. SA infections were investigated separately, with respect to the primary endpoint. Mantel-Haenszel odds ratios (ORs) were pooled using traditional meta-analytic techniques, under a random-effects model. Results: 2 randomized, 1 non-randomized prospective and 6 retrospective studies, featuring 886 patients (of whom 322 allocated to rifampicin), were included. While 8 studies addressed staphylococcal endocarditis (with 3 of them dealing exclusively with SA infections), prosthetic and culture-negative cases were only focused on 1 study each. There were 96 deaths (from 7 studies), 32 relapses (from 6 studies), 35 bail-out valvular surgeries (from 4 studies), 14 microbiological failures (from 2 studies) and 32 hepatotoxicity cases (from 5 studies). Rifampicin use was associated with a tendency towards higher all-cause mortality (OR 1.63, 95% CI 0.85-3.12, P 0.14, i 2 12%), which held true for SA infections (OR 1.63, 95% CI 0.67-3.92, P 0.28, i 2 20%), and bail-out valve surgery (OR 2.03, 95% CI 0.65-6.34, P 0.22, i 2 5%). Relapse and microbiological failures rates were similar between groups (OR 1.25, 95% CI 0.59-2.65, P 0.56, i 2 0% and OR 1.3, 95% CI 0.24-6.90, P 0.76, i 2 41%, respectively). Hepatotoxicity odds were greater in the rifampicin arm (OR 3.24, 95% CI 1.35-7.77, P 0.008, i 2 0%). Conclusion: Currently available evidence does not support rifampicin use in endocarditis.