POST HOC ANALYSES OF ESKETAMINE NASAL SPRAY PLUS AN ORAL ANTIDEPRESSANT IN ELDERLY PATIENTS WITH TREATMENT-RESISTANT DEPRESSION

Abstract

Introduction Major depressive disorder in the elderly is correlated with lower response and remission rates, greater disability, decreased quality of life, and greater mortality from suicide; approximately 18-40% develop treatment-resistant depression (TRD). Moreover, the elderly respond less well to currently available treatments and may be more vulnerable to their adverse effects. The severity of TRD in the elderly is exemplified by a 5-fold increased use of electroconvulsive therapy, highlighting a critical need for alternative safe and effective treatments. A companion abstract discusses results from the first large phase 3 study of esketamine nasal spray in elderly patients with TRD. Post hoc analyses presented here explored factors that may have contributed to the lack of statistical significance observed in the study, despite a numerical advantage on the primary endpoint–Montgomery Asberg Depression Rating scale (MADRS) LS mean scores–in the esketamine arm vs the control arm of the trial. Methods The primary phase 3, double-blind, multicenter, active controlled study (NCT02422186), included adults ≥65 years of age (N=138) with TRD. Patients were randomized (1:1) to flexibly-dosed esketamine nasal spray (28, 56 or 84 mg twice weekly) and a new oral antidepressant (esketamine/antidepressant), or a new oral antidepressant and placebo nasal spray (antidepressant/placebo). Change from baseline was analyzed using mixed-effects model for repeated measures (MMRM). Analyses were conducted at a 2-sided significance level of 0.05. Findings from the primary analysis are described in the companion abstract. Post hoc analyses explored factors that may have affected the study outcome including (1) impact of study stage, i.e. pre interim analysis (IA) (stage 1) or post IA (stage 2): modifications were made early in the study including training of remote MADRS raters to work with elderly and site discussions related to dose that, because of timing for implementation of the changes, had the greatest impact on stage 2; (2) impact of dose assessed by stage of the IA; (3) age subgroups (65-74 and ≥75 years); (4) age of onset of depression ( Results For reference, the primary efficacy endpoint, the LS mean (95% CI) difference for change in MADRS total scores from baseline to day 28 between the esketamine/antidepressant group and the antidepressant/placebo group using MMRM was -3.6 (−7.20, 0.07; p=0.059). Post hoc analyses included (1) a marked difference in efficacy between stages of the IA: LS mean (95% CI) difference was -1.6 (-6.85, 3.70) in stage 1 vs -5.6 (-10.78, -0.32) in stage 2. The primary analysis applied equal weight to stage 1 (51 patients) and stage 2 (87 patients) effectively down-weighting the results of stage 2. Overall analysis, without adjusting for the IA, showed LS mean (95% CI) change of -4.0 (-7.71, -0.25); (2) use of maximum dose of 84mg: 52.5% at day 25 of stage 1 vs 71.8% at day 25 of stage 2; (3) age: 65-74 years LS mean (95% CI) change of -4.9 (-8.96, -0.89) vs ≥75 years -0.4 (-10.38, 9.50); (4) age at onset of depression Conclusions While MADRS improvement with esketamine/antidepressant vs treatment with antidepressant/placebo was not statistically significant in the primary analysis, a nonsignificant favourable trend was found, with a treatment effect size similar to that seen in younger adult esketamine studies. Post hoc analyses assessed factors potentially affecting the primary outcome. In the short-term study, use of lower doses earlier in the study may have decreased efficacy. In the post hoc analysis, a 95% CI of difference that did not include 0 indicated that esketamine/antidepressant was favoured over antidepressant/placebo without the corrective weighting for the IA, as well as for patients 65-74 years of age, and for those with the onset of depression at age This research was funded by This study was funded by Janssen Research & Development, LLC.