Outcomes of Neuroleptic Malignant Syndrome With Depot Versus Oral Antipsychotics: A Systematic Review and Pooled, Patient-Level Analysis of 662 Case Reports

Objective: This systematic review and pooled, patient-level analysis of neuroleptic malignant syndrome (NMS) case reports and series compared NMS characteristics and outcomes during long-acting injectable antipsychotic (LAI) versus oral antipsychotic (OAP) treatment.

Data Sources: Two authors independently searched MEDLINE, Embase, Cochrane, CINAHL, and PsycINFO databases for articles in English from database inception until October 9, 2018.

Study Selection: Case reports with author-defined NMS during ongoing antipsychotic treatment or within 1 injection interval of LAIs in adults aged 18-65 years.

Data Extraction: Demographic, clinical, treatment and outcome data were independently extracted following PRISMA guidelines. NMS severity was rated using the Francis-Yacoub scale. Characteristics and outcomes of NMS were compared when occurring during LAI versus OAP treatment, adjusting for significant between-group differences.

Results: Of 662 reported cases (median age = 36 years, male = 61.2%), 122 (18.4%) involved LAIs (second-generation antipsychotic [SGA] LAIs [SGA-LAIs] = 10, 1.5%), whereas 540 (81.6%) involved OAPs (SGA-OAPs = 159, 24.0%). The 2 groups did not differ in age, illness duration, comorbidities, or presence or severity of NMS symptoms (median Francis-Yacoub score: LAIs = 26 vs OAPs = 23, P = .8276). Antipsychotic formulation was not significantly associated with longer duration of hospitalization (LAIs = 5.0 weeks vs OAPs = 3.8 weeks, P = .8322), post-NMS sequelae (LAIs = 8.8% vs OAPs = 7.0%, P = .7489), or death (LAIs = 10.7% vs OAPs = 6.7%, P = .0861). When different, post hoc confounder-adjusted models were used, duration of NMS (but not hospitalization for NMS) was longer with LAIs than with OAPs (median = 2.6 vs 1.8 weeks, P = .0339), driven by FGAs rather than SGAs.

Conclusions: These data, plus the fact that only 10 published NMS cases exist with SGA-LAIs, should mitigate safety concerns regarding LAIs, but results should be interpreted cautiously since they are based on case reports.