ESKETAMINE NASAL SPRAY FOR RELAPSE PREVENTION IN PATIENTS WITH TREATMENT-RESISTANT DEPRESSION: A POST HOC ANALYSIS OF PREDICTORS OF RELAPSE IN PLACEBO-TREATED PATIENTS IN SUSTAIN-1

Background

SUSTAIN-1 (NCT02493868) was a phase 3, double-blind, multicenter, randomized withdrawal study evaluating the efficacy of continuing esketamine nasal spray plus an oral antidepressant (ESK+OAD) versus placebo nasal spray plus an OAD (PBO+OAD) in delaying relapse among adults with treatment-resistant depression (TRD). In this post hoc analysis, we evaluated the potential predictors of relapse among patients randomly assigned to the PBO+OAD arm of SUSTAIN-1, focusing on baseline demographics and disease-related characteristics that may increase vulnerability to relapse following discontinuation of ESK.

Methods

This analysis included patients who completed induction and optimization with ESK+OAD, achieved stable remission, and were subsequently randomly assigned to discontinue ESK and receive PBO+OAD, rather than continue ESK+OAD treatment, during the maintenance phase. Relapse was defined per criteria in the SUSTAIN-1 protocol (Montgomery-Åsberg Depression Rating Scale total score ≥22 for 2 consecutive assessments and/or hospitalization for worsening depression or any other clinically relevant event). We used an exploratory approach, evaluating 45 candidate predictors (baseline demographic characteristics and psychiatric history) chosen from the literature and authors’ clinical judgment. Initial screening used univariate Cox proportional hazards models; variables with nominal significance (p < 0.3) were advanced (no multiplicity adjustment). To identify independent predictors given the heterogeneity of TRD, those candidate predictors were entered into a stepwise multivariable Cox proportional hazards model, with candidates entered sequentially and removed if they became nonsignificant; the final model retained predictors with p < 0.05. Hazard ratios and 95% confidence intervals are reported.

Results

Patients in stable remission who discontinued ESK and received PBO+OAD experienced a higher risk of relapse compared with those who continued ESK+OAD treatment (ESK+OAD vs PBO+OAD, HR 0.49, 95% CI: 0.29, 0.84). Of 176 patients in stable remission, 45.3% of patients in the PBO+OAD arm experienced relapse (n = 39/86) compared with 26.7% of patients in the ESK+OAD arm (n = 24/90). In the PBO+OAD arm, risk of relapse was associated with number of prior unsuccessful antidepressant treatments (2 vs ≥3; HR 0.45; 95% CI: 0.22, 0.93; P = 0.030), Columbia-Suicide Severity Rating Scale score at screening (lifetime history of suicidal behavior vs no event; HR 4.38; 95% CI: 1.14, 16.83; P = 0.032), and baseline Clinical Global Impression-Severity (CGI-S) scale category (markedly ill vs mildly and moderately ill; HR 16.62; 95% CI 2.09, 131.89; P = 0.008; only 1 patient experienced relapse in the CGI-S mildly and moderately ill category).

Conclusions

This post hoc analysis identifies potential baseline clinical factors that could be associated with higher risk of relapse following discontinuation of ESK among patients with TRD. These results support continued ESK treatment or more vigilant monitoring strategies for patients who have ≥3 prior antidepressant treatments, have a lifetime history of suicidal behavior, or are categorized as CGI-S “markedly ill,” as they have a higher risk of relapse.