Efficacy and safety of atypical antipsychotics for psychosis in Parkinson's disease: A systematic review and Bayesian network meta-analysis

Introduction: We performed a systematic review and Bayesian network meta-analysis to clarify the relative efficacy and safety of pimavanserin compared to atypical antipsychotics for psychosis in Parkinson's disease (PD). Methods: PubMed, Embase, Cochrane Central Register of Controlled Trials, and Japana Centra Revuo Medicina Web were searched for relevant articles until October 31, 2019. Eligible randomized controlled trials were synthesized for efficacy (Brief Psychiatry Rating Scale [BPRS] and Clinical Global Impression Scale [CGI-S]) and safety (Unified Parkinson's Disease Rating Scale part III [UPDRS-III] and dropouts due to adverse events). The mean differences (BPRS, CGI-S, and UPDRS-III) or odds ratios (dropouts due to adverse events) between each active drug and placebo were estimated and summarized as means and 95% credible intervals, respectively. Results: We identified 17 relevant trials. Clozapine showed significant efficacy (BPRS, -5.6 [-8.4 to -2.7] and CGI-S, -1.2 [-1.7 to 0.7]), with low impact on motor functions (UPDRS-III, -1.1 [-3.8 to 1.5]), but an increase in dropouts due to adverse events (2.9 [0.9 to 9.6]) as compared to placebo. Pimavanserin also showed significant efficacy (CGI-S, -0.5 [-0.9 to -0.2]) and similar impact on motor functions (UPDRS-III, 0.2 [-1.4 to 1.9]), but a tendency of increase in dropouts due to adverse events (2.2 [0.5 to 12.4]) as compared to placebo. Conclusions: Clozapine showed an efficacy with low impact on motor functions that was consistent with previous reports. Although the efficacy of pimavanserin may be inferior to that of clozapine, it had a favorable profile for the treatment of psychosis in PD.