Steroidomic Differences Among Clozapine, Olanzapine, and Risperidone in Male Patients with First-Episode Psychosis, Including a Treatment-Resistant Schizophrenia Subgroup

Clozapine (Clo) is the most effective antipsychotic for treatment-resistant schizophrenia (TRS), yet the mechanisms underlying its unique clinical profile remain insufficiently understood. Neuroactive steroids (NASs) influence neuronal excitability, neuroplasticity, inflammation, and stress responses—domains disrupted in schizophrenia—and may contribute to antipsychotic pharmacodynamics. This pilot study examined whether Clo exhibits a distinctive steroidomic pattern compared with olanzapine (Ola) and risperidone (Ris) in male patients with first-episode psychosis, including a clinically defined TRS subgroup receiving Clo. Serum collected after group-specific treatment durations (Clo: 17.7 ± 12.4; Ola: 13.9 ± 1.9; Ris: 12.9 ± 1.3 months) was analyzed by GC–MS/MS, with steroidogenic enzyme activities inferred from precursor–product molar ratios as proxies of metabolic balance. Statistical evaluation included age-adjusted ANCOVA, OPLS, multiple regression, and trend-based analysis of coordinated pathway-level shifts. Although Ola increased several steroid concentrations, its overall and unconjugated trends did not differ significantly from Clo (p = 0.515 and p = 0.409). At the multivariate level, however, the two profiles were significantly discriminated (OPLS: R2Y = 50.8%, CV-ANOVA p = 0.011), indicating that the difference lies in the coordinated pattern rather than the global magnitude of change. Clo showed a distinctive exploratory pattern: lower DHEA, higher 5α-reduced pregnane NAS, altered androgen turnover, reduced CYP17A1 lyase proxy activity, and a mild but statistically significant increase in the cortisol-to-cortisone ratio, and was clearly distinguished from Ris by broad steroid increases. This may indicate a shift toward a profile with potential GABAergic and neuroprotective features, although interpretations remain exploratory given the small, cross-sectional, male-only sample and the non-independence of steroid pathways. Neurosteroidogenesis may thus contribute to antipsychotic response variability and offer candidate biomarkers, warranting validation in larger longitudinal studies.