The effect of esketamine in patients with treatment-resistant depression with and without comorbid anxiety symptoms or disorder

Comorbid anxiety is common in major depressive disorder (MDD), with estimates ranging roughly from 53% to 67% depending on the definition used. Patients who have depression with concurrent anxiety tend to show greater illness severity and chronicity, higher rates of suicidal behaviour, and poorer response and remission rates to conventional monoaminergic antidepressants. Previous small studies of intravenous ketamine have suggested antidepressant effects in patients with comorbid anxiety, but evidence remains limited and definitions of “anxious depression” vary across studies, using either syndromal (DSM-based) or dimensional (scale cut-off) approaches. Daly and colleagues performed a post hoc analysis of the TRANSFORM-2 trial to explore whether the antidepressant effect and safety profile of esketamine nasal spray given with a newly initiated oral antidepressant differ according to the presence or absence of comorbid anxiety symptoms or disorder. The specific aim was to compare change in depressive symptoms, response and remission rates, and adverse events between esketamine plus oral antidepressant and antidepressant plus placebo, stratified by comorbid anxiety status as defined for this analysis.

This report is a post hoc analysis of TRANSFORM-2, a 4-week, flexible-dose, double-blind, active-controlled, multicentre trial in outpatients with treatment-resistant depression (TRD). TRANSFORM-2 randomised participants to twice-weekly intranasal esketamine (56 or 84 mg, flexible dosing) or matching placebo, each given alongside a newly initiated open-label oral antidepressant for 4 weeks. The trial enrolled adults aged 18–64 years with recurrent MDD (DSM-5) without psychotic features and evidence of inadequate response to between one and five prior oral antidepressants. Comorbid anxiety for these post hoc analyses was defined in two ways: either a current anxiety disorder on the Mini International Neuropsychiatric Interview (MINI) at screening (including generalized anxiety disorder, panic disorder, social anxiety disorder, PTSD, or OCD) or a patient-reported GAD-7 total score ≥10 at both screening and baseline (day 1). The GAD-7 is a brief validated self-report measure of anxiety; a score ≥10 indicates at least moderate anxiety. Efficacy assessments included the Montgomery–Åsberg Depression Rating Scale (MADRS) administered at baseline and on days 2, 8, 15, 22 and 28 by off-site, independent blinded raters. Response was prespecified as ≥50% improvement in MADRS from baseline; remission was defined as MADRS ≤12. Safety evaluations recorded adverse events throughout, and dissociation was assessed with the Clinician Administered Dissociative States Scale (CADSS) at multiple dosing visits. Analyses used all randomised patients who received at least one dose of both intranasal study drug and oral antidepressant. Change in MADRS from baseline to day 28 was compared within treatment arms by paired t test and between arms by analysis of covariance (ANCOVA) with fixed effects for treatment, comorbid anxiety status, and baseline MADRS; interaction terms between treatment and comorbid anxiety were included. Response and remission at day 28 were compared by Cochran–Mantel–Haenszel tests controlling for region and class of oral antidepressant, and logistic regression models tested treatment-by-anxiety interactions and yielded odds ratios with 95% confidence intervals. Adverse events and CADSS-defined dissociation (CADSS >4) were summarised by comorbid anxiety status.

Of 227 patients randomised, 223 received study drug and were included in efficacy analyses (114 esketamine plus antidepressant; 109 antidepressant plus placebo). Completion rates for the 28-day double-blind phase were 86.0% (98/114) in the esketamine/antidepressant group and 90.8% (99/109) in the antidepressant/placebo group. Discontinuation due to adverse events was uncommon (8/114 [7.0%] esketamine group; 1/109 [0.9%] placebo group). Overall, 72.6% (162/223) met the post hoc definition for comorbid anxiety: 69.1% (154/223) had GAD-7 ≥10 at both screening and baseline, while 13.5% (30/223) met criteria for a current anxiety disorder on the MINI. Patients classified as having comorbid anxiety had longer mean duration of the current depressive episode and higher baseline MADRS scores, indicating more chronic and severe depressive symptoms; other demographic and disease characteristics were broadly similar between groups. On the primary depressive outcome, esketamine-treated patients showed substantial reductions in MADRS by day 28 irrespective of anxiety status (mean [SD] change from baseline reported in the extracted text as approximately -21.0 [12.51] for those with anxiety and -22.7 [11.98] for those without). Compared with antidepressant plus placebo, esketamine plus antidepressant produced significantly greater improvement at day 28 both in patients with comorbid anxiety (least squares [LS] mean difference -4.2, 95% CI -8.1 to -0.3) and in those without comorbid anxiety (LS mean difference -7.5, 95% CI -13.7 to -1.3). The interaction between treatment and comorbid anxiety on change in MADRS was not statistically significant (interaction p = .371), and formal tests of interaction for response and remission rates were also non-significant (p = .136 and p = .088, respectively). The antidepressant/placebo arm showed similar improvement whether patients had comorbid anxiety or not. Response and remission rates at day 28 were higher in the esketamine/antidepressant group than in the antidepressant/placebo group regardless of comorbid anxiety status. Safety findings were consistent with the broader esketamine development programme: the most common adverse events in esketamine-treated patients (>10%) included anxiety, blurred vision, dissociation, dizziness, dysgeusia, headache, nausea, paresthesia, somnolence and vertigo. Placebo-adjusted rates of anxiety, dissociation, nausea and paresthesia were numerically higher among esketamine-treated patients who had comorbid anxiety compared with those who did not, which the authors note might reflect greater somatic attentiveness among anxious patients. Most adverse events were mild or moderate and resolved on the treatment day. Two serious adverse events occurred among the 227 randomised patients (0.9%), both in patients with comorbid anxiety: a fatal road traffic accident in the esketamine/antidepressant group (investigator judged doubtfully related to study drug) and positional vertigo in the antidepressant/placebo group. Incidence of CADSS-defined dissociative symptoms was higher in the esketamine/antidepressant group than in the antidepressant/placebo group, but rates of dissociation did not differ between patients with and without comorbid anxiety.

Daly and colleagues interpret these post hoc data as supporting the efficacy of esketamine nasal spray plus an oral antidepressant in adults with TRD irrespective of co-occurring anxiety symptoms or disorders. While previous literature has shown that comorbid anxiety is generally associated with worse outcomes to standard antidepressants, the current analysis found that esketamine-treated patients with comorbid anxiety attained clinically meaningful and statistically significant improvement versus antidepressant plus placebo (LS mean difference -4.2 points on the MADRS). The investigators also note that patients with lower baseline anxiety tended to show greater absolute improvement than those with comorbid anxiety (LS mean subgroup difference 3.3 points), but the absence of a significant treatment-by-anxiety interaction indicates that the relative benefit of esketamine over active control did not differ significantly by anxiety status. The authors position these findings alongside limited prior ketamine data, which have variably indicated antidepressant effects in samples with anxious depression. They caution, however, that differences in sample size, anxiety definitions and study designs complicate direct comparisons. Several limitations acknowledged by the study team constrain interpretation: the analysis was post hoc and the definition of comorbid anxiety was not prespecified at enrollment, the two operational definitions (MINI syndromal diagnosis versus GAD-7 dimensional cut-off) capture different constructs and may not reflect true population prevalence, and the analysis did not account for differences in oral antidepressant dose or concomitant benzodiazepine use. The frequent twice-weekly site visits for both arms could have increased nonspecific therapeutic effects, potentially explaining why the antidepressant/placebo group did not show the typically poorer outcomes among anxious patients. Other limits to generalisability include exclusion of patients with moderate-to-severe substance or alcohol use disorders and those on high-dose benzodiazepines, and the relatively short (4-week) treatment duration. Given these caveats, the authors recommend cautious interpretation and call for further studies specifically designed to evaluate esketamine in patients with TRD and comorbid anxiety.

In this post hoc analysis of TRANSFORM-2, esketamine nasal spray in combination with a newly initiated oral antidepressant was more effective than a newly initiated antidepressant plus placebo in reducing depressive symptoms in adults with TRD, regardless of the presence or absence of comorbid anxiety. Because the analysis was exploratory and limited by post hoc definitions and other design constraints, the investigators conclude that additional prospective studies targeted to this patient population are warranted.