Lumateperone as an alternative to clozapine in treatment-resistant schizophrenia: Insights from a quantitative...

Clozapine remains the only drug with superior efficacy in treatment-resistant schizophrenia (TRS) but is limited by underutilization and significant adverse effects, motivating mechanistic approaches to identify clozapine-like alternatives. We developed a 2-layer quantitative systems pharmacology model for TRS. Layer 1 integrates binding affinities at 12 receptor mechanisms with theory-driven mechanistic weights derived from contemporary clozapine pharmacology frameworks. Layer 2 extends predictions with positron emission tomography-calibrated effective brain concentrations back-calculated from published [¹¹C]raclopride observations, active metabolite modeling and dose-dependent benefit-risk integration. Eight treatments were scored on a 0-1 efficacy scale using weighted mechanism components, transformed to predicted TRS response rates via a Hill-type transformation function calibrated to clozapine. Bayesian Monte Carlo with Dirichlet priors (5000 iterations) propagated uncertainty, and model performance was externally validated against TRS trial data for 5 drugs (root mean square error 3.5%, r = 0.964). Predicted response rates formed hierarchical tiers: clozapine-lumateperone and clozapine-aripiprazole combinations ≈33%; clozapine (400 mg) 30% (95% CI, 25-34%); lumateperone (42 mg) 28% (24-31%); aripiprazole 24% (19-29%); risperidone (4 mg) 12%; olanzapine (15 mg) 10%; and haloperidol (10 mg) 2%. The positron emission tomography-calibrated dose-response extension (root mean square error 7.25%, mean absolute error 2.88% for D₂ occupancy prediction) revealed that lumateperone possesses the widest therapeutic window (14-84 mg) with an optimal benefit-risk dose of 55 mg (predicted 30.3% response, response rate-extrapyramidal symptoms ratio = 25), compared with response rate-EPS ratio = 11.5 for clozapine 400 mg. Haloperidol, olanzapine, and risperidone did not exhibit a therapeutic window that met benefit-risk criteria. Model simulations identify lumateperone as a promising alternative TRS candidate with a favorable benefit-risk profile, warranting prospective dose-finding clinical trials. SIGNIFICANCE STATEMENT: This study develops and validates a 2-layer quantitative systems pharmacology model integrating receptor-level mechanisms with positron emission tomography-calibrated dose-response to predict treatment outcomes in treatment-resistant schizophrenia. It identifies lumateperone as a clozapine-like candidate with near-comparable efficacy but a superior therapeutic window, generating clinically testable hypotheses for optimized dosing and combination strategies.