Baseline Traits Have Not Reliably Forecast Antidepressant Response
Nearly half of patients with MDD fail to achieve adequate response to ADT,1 and prior studies of baseline predictors are largely inconsistent.2 In two phase 3 trials, adjunctive lumateperone 42 mg/day significantly improved MADRS scores versus placebo at Week 6.3,4 This pooled post hoc analysis, presented at the 39th European College of Neuropsychopharmacology Congress, evaluated potential predictors of Day 43 response.2
Two Pooled Trials, 30 Candidate Predictors
Two six-week, double-blind trials randomized adults (18–65 years) with DSM-5–defined MDD and inadequate response (ie, <50% improvement in MADRS total score) to 1–2 ADT courses in the current episode to ADT + lumateperone 42 mg/day or ADT + placebo. In the pooled analysis of 950 patients, response was defined as a ≥50% reduction in MADRS total score at Day 43. Approximately 30 variables were evaluated using three logistic regression approaches (ie, elastic net, lasso, and stepwise regression).
Treatment and Early CGI-S Change Were Associated with Response
Two variables emerged as predictors of Day 43 response in all three statistical models: adjunctive lumateperone 42 mg/day (versus placebo) and early CGI-S improvement at Day 8.
- ADT + Lumateperone: ORs ranged from 2.19 to 2.87 (P < 0.001 for all), with 95% CIs that excluded the null.
- Day 8 CGI-S improvement: ORs ranged from 2.70 to 2.78 (P < 0.001 for all), with 95% CIs that excluded the null.
- Early benefit: Patients receiving ADT + lumateperone showed significantly greater CGI-S improvement at Day 8 than those receiving ADT + placebo.
Other variables, such as psychiatric hospitalization and age at first diagnosis, were selected in stepwise regression but not consistently across models. Safety was not analyzed; the parent trials reported a consistent safety profile.3,4
Early Change Was Associated With Response, but Predictive Value Remains Untested
Clinicians need to know who will respond to adjunctive treatment, but baseline characteristics have not reliably answered that question. In this analysis, no baseline characteristic was consistently associated with Day 43 response, whereas treatment with ADT + lumateperone and early CGI-S improvement at Day 8 were. The CGI-S association was observed at the group level, but the poster did not report how accurately early improvement classified individual patients; additionally, as an on-treatment measure, it cannot guide patient selection before treatment.
Because the analysis was post hoc and the trials’ eligibility criteria may not reflect the broader clinical population, these associations need confirmation before they can inform practice. Prospective studies with broader populations and longer follow-up could test if a lack of early CGI-S improvement identifies patients unlikely to respond; such a finding could reframe the search for predictors from baseline traits to how patients respond in the first week.
References
- Mago R, Fagiolini A, Weiller E, Weiss C. Understanding the emotions of patients with inadequate response to antidepressant treatments: results of an international online survey in patients with major depressive disorder. BMC Psychiatry. 2018;18(1):33.
- Singh MK, Durgam S, Earley WR, Wilson AC, Sheehan JJ, Turkoz I. Predictors of response with adjunctive lumateperone in patients with major depressive disorder: a pooled analysis of two phase 3 trials. Poster presented at: 39th European College of Neuropsychopharmacology (ECNP) Congress; October 10-13, 2026; Munich, Germany. Poster PS02-2118.
- Durgam S, Earley W, Kozauer S, et al. Lumateperone as adjunctive therapy in patients with major depressive disorder: results from a randomized, double-blind, phase 3 trial. J Clin Psychiatry. 2025;86(4):25m15848.
- Durgam S, Earley WR, Kozauer SG, et al. Adjunctive lumateperone in patients with major depressive disorder: results from a randomized, double-blind, phase 3 trial. Am J Psychiatry. 2025;182(12):1072-1082.
