Meta-analysis comparing efficacy of benzoyl peroxide, clindamycin, benzoyl peroxide with salicylic acid, and combination benzoyl peroxide/clindamycin in acne.

Abstract

Comparative efficacy of the multiple treatments containing benzoyl peroxide (BPO) and clindamycin (CL) is not established. We compared the efficacy of topical 5% BPO, 1% to 1.2% CL, 5% BPO with salicylic acid (SA) preparation, and combination BPO/CL in acne lesion reduction. A meta-analysis was conducted using the Cochrane collaboration guidelines in accordance with the PRISMA statement. A total of 23 studies including 7309 patients were used in the meta-analysis. At 2 to 4 weeks, 5% BPO + SA had statistically greater percent lesion reductions over other groups (weighted mean inflammatory lesion reduction: BPO = 33.4%, CL = 21.5%, BPO + SA = 55.2%, BPO/CL = 40.7%, placebo = 7.3%; weighted mean noninflammatory lesion reduction: BPO = 19.1%, CL = 10.0%, BPO + SA = 42.7%, BPO/CL = 26.2%, placebo = 6.7%). At 10- to 12-week end points, 5% BPO + SA and BPO/CL were similar, with overlapping confidence intervals (weighted mean inflammatory lesion reduction: BPO = 43.7%, CL = 45.9%, BPO + SA = 51.8%, BPO/CL = 55.6%, placebo = 26.8%; weighted mean noninflammatory lesion reduction: BPO = 30.9%, CL = 32.6%, BPO + SA = 47.8%, BPO/CL = 40.3%, placebo = 17.0%). Trial heterogeneity, publication bias, and deficits in the reporting of individual primary studies may affect results. At early time points, 5% BPO + SA had the best profile. BPO/CL was only incrementally better than BPO alone but was superior to CL alone. At later time points, 5% BPO + SA was similar to BPO/CL.