Intranasal esketamine: real-world clinical practice in treatment-resistant depression and factors associated with treatment response

Treatment-resistant depression (TRD) remains common and difficult to manage, and standard antidepressants often do not produce a rapid or sufficient response. Esketamine, the S-enantiomer of ketamine, has been developed as a faster-acting option and has shown efficacy in randomised trials, but those trials enrolled selected patients and may not reflect the complexity of routine inpatient practice. The extracted text also notes that evidence in older adults and in people with psychiatric comorbidity has been mixed or limited, and that questions remain about who benefits most, how long induction should continue, and how well the treatment works outside trial conditions. T. and colleagues set out to evaluate the real-world effectiveness and tolerability of intranasal esketamine in a naturalistic inpatient TRD cohort, and to examine demographic and clinical factors associated with treatment response. In particular, they aimed to bridge the gap between trial data and routine care by using retrospective chart data from a relatively large hospital sample. The study also explored whether variables such as age, sex, psychiatric comorbidity, lifetime electroconvulsive therapy, and the number of treatment sessions were related to outcomes, including response, partial response, remission, and suicidal ideation.

This was a retrospective, uncontrolled chart review of inpatients with treatment-resistant depression treated with intranasal esketamine at the University Hospital Münster, Germany. The sample comprised 101 patients, with a subset of 36 patients also having symptom ratings after the eighth treatment session to assess extended induction effects. Treatment was delivered in a routine inpatient setting and was not blinded; ratings were made by treating physicians and a psychologist during standard clinical care. Esketamine was administered using the FDA-approved intranasal spray, usually started at 56 mg and increased to 84 mg in later sessions; for patients older than 65 years, treatment generally began at 28 mg. Dose choice and adjustment were made by a senior psychiatrist experienced in TRD. Patients gave informed consent and received extensive information before treatment. The mean induction period was 6.6 weeks, and the induction phase was fully inpatient; about half of the cohort later entered a maintenance regimen. Depression severity was measured before treatment and after induction using the Montgomery-Åsberg Depression Rating Scale (MADRS) and Beck Depression Inventory-II (BDI-II). Suicidality was assessed using item 10 of the MADRS. Response was defined as at least a 50% reduction in MADRS, partial response as a 30-50% reduction, and remission as MADRS 10 or below at follow-up. The researchers analysed pre- to post-treatment change with paired t-tests and effect sizes using Cohen's d. Repeated-measures ANOVA was used to test change over time and the influence of between-subject factors including age, sex, psychiatric comorbidity, and lifetime electroconvulsive therapy, with number of treatment sessions entered as a covariate. Binary logistic regression was used to examine which factors predicted response, partial response, and remission.

Across the cohort, depressive symptoms improved significantly after esketamine treatment. MADRS scores fell by 10.7 points (95% confidence interval 8.3-13.0), and BDI-II scores fell by 11.5 points (95% confidence interval 9.2-13.9), both with large effect sizes. The MADRS suicidality item also decreased significantly, from 2.5 to 1.6, with a mean change of -0.9 (p<0.001). Using the prespecified outcome definitions, 28.8% of patients achieved response, 52.5% achieved at least partial response, and 19.3% reached remission. Repeated-measures analyses confirmed a significant effect of time on MADRS scores, without significant time-by-factor interactions. Depression severity was higher in women and in those younger than 60 years, while for BDI-II there was a significant interaction with age, indicating greater self-rated improvement in older patients. In logistic regression, no covariate significantly predicted response overall. However, in the discussion of the modelling results, older age was associated with higher odds of remission (OR 4.06, p=0.041), and male sex was associated with higher odds of partial response (OR 3.71, p=0.012). The extracted text notes that the confidence intervals around these odds ratios were large, implying substantial uncertainty. In the exploratory analysis of extended induction, the 36 patients with baseline, session 8, and end-of-treatment ratings appeared to benefit from continuing beyond eight sessions, especially older adults and those with psychiatric comorbidity. Patients younger than 60 years and those without comorbidity showed little additional improvement. The mean induction duration was 6.6 weeks, and maintenance treatment had a mean interval of 4.9 weeks and a mean duration of 30.3 weeks. At the time of analysis, 72.7% had stopped esketamine treatment, while others were still in maintenance. The mean total number of sessions was 23.5. No treatment-related serious adverse events were observed. Premature discontinuation occurred in 8.9% of patients, most often early in treatment, commonly because of dissociation or sedation; two patients switched because of non-response, and two left inpatient care early. One non-fatal suicide attempt occurred during treatment.

The authors interpret these findings as showing that intranasal esketamine produced meaningful symptom improvement in a real-world inpatient TRD population, including a reduction in suicidal ideation. They emphasise that the response and remission rates were lower than those seen in manufacturer-led phase 3 trials, but they consider the results broadly consistent with earlier randomised evidence and with other observational cohorts involving clinically complex patients. They frame the study as adding practical inpatient data to the growing real-world evidence base. T. and colleagues note that their findings align with prior trial data showing rapid antidepressant effects of esketamine, although they also acknowledge that some randomised studies had mixed results in specific subgroups. They state that their response and remission rates fall within the range reported in earlier real-world studies, particularly in populations with high comorbidity and complex treatment histories. The authors highlight older age and psychiatric comorbidity as potentially relevant to treatment tailoring. They suggest that extending the induction phase beyond eight sessions may be especially useful for older patients and for those with comorbid psychiatric illness, and they argue that treatment duration and frequency should be individualised during optimisation and maintenance rather than fixed rigidly. Regarding suicidality, they caution that the observed reduction in the MADRS suicidality item should be interpreted carefully. They note that earlier trials in acute suicidal ideation showed rapid symptom improvement, but regulatory guidance does not establish esketamine as preventing suicide or independently reducing suicidal behaviour. They therefore advise continued routine risk management and hospitalisation when clinically indicated. For safety, the authors state that their lack of treatment-related serious adverse events is consistent with the established safety profile of esketamine, where dissociation and sedation are common but usually transient. They also acknowledge that adverse events were not systematically assessed, so tolerability data are limited by routine-care ascertainment. They further note one non-fatal suicide attempt, reinforcing the need for ongoing monitoring. The extracted text also indicates several limitations and uncertainties. The study was retrospective, uncontrolled, and open-label in routine practice, so causal inference is limited. The sample size was modest for subgroup and interaction analyses, and the authors note that some analyses may have been underpowered. The extract also states that the study was one of the larger inpatient cohorts and therefore useful for real-world translation, but that the setting and non-blinded assessments constrain generalisability and interpretation.