Relapse and its modifiers after antipsychotic discontinuation in first-episode non-affective psychosis: A systematic review and meta-analysis

The long-term relapse risk after antipsychotic discontinuation, relative to maintenance therapy, remains unclear in adults with first-episode non-affective psychosis (FENAP) stabilized on antipsychotics. This pairwise meta-analysis employing a random-effects model included randomized controlled trials (RCTs) that compared antipsychotic discontinuation with maintenance treatment in adults with stabilized FENAP. Relapse rates were compared at matched time points (1, 2, 3, 6, 9, 12 [primary outcome], 15, 18, 21, and 24 months) between the discontinuation and maintenance groups to more accurately investigate the temporal relapse trend. Risk ratios (RRs) and absolute risk reductions (ARRs) with 95% confidence intervals (CIs) were calculated. This review identified 12 RCTs that included 1133 adults (60.1% male; mean age: 27.3 years). No statistically significant difference in relapse rates was observed between the maintenance and discontinuation groups at 1 month. However, most participants in the discontinuation group were still receiving antipsychotics at 1 month due to gradual tapering. Significant differences were observed at all subsequent time points. At 12 months, the RR of relapse in the maintenance group versus the discontinuation group was 0.45 (95% CI: 0.35–0.57; p < 0.001; I² =11.1%). Relapse rates at 12 months were 21.0% and 53.3% in the maintenance and discontinuation groups, respectively. From 2 to 24 months, RRs remained stable (0.45–0.54). The ARR was 6.0% at 2 months, gradually increasing to 20.0% by 6 months and 32.0% by 12 months, and remaining stable through 24 months. In conclusion, continuing antipsychotic treatment in clinically stable FENAP significantly reduces the risk of relapse for up to 24 months.