Real-world experience of esketamine use to manage treatment-resistant depression: A multicentric study on safety and efficacy (REAL-ESK study)

Treatment-resistant depression (TRD) is a common and severe condition associated with substantial healthcare costs and an elevated suicide risk. Earlier research implicates glutamatergic dysfunction in some cases of major depressive disorder, and both ketamine and its S-enantiomer esketamine—NMDA receptor antagonists that may promote neuroplasticity via mTOR/BDNF signalling—have shown antidepressant effects. Esketamine's intranasal formulation and higher NMDA affinity have enabled outpatient use, and randomised trials have supported its efficacy and safety, but these trials often exclude complex, comorbid, or frail patients and therefore leave uncertainty about performance in ordinary clinical practice. Martinotti and colleagues designed an observational, retrospective, multicentre study to assess the real-world effectiveness and safety of intranasal esketamine in adults with TRD treated across several Italian mental health services. The primary objective was to measure change in depressive symptoms at one and three months; a secondary aim was to describe adverse events and tolerability in this more heterogeneous, treatment-seeking population.

The REAL-ESK study used a retrospective, multicentre observational design and included 116 adults with TRD treated under an early-access esketamine programme across multiple Italian centres coordinated by the Università degli studi di Chieti and the University of Brescia. The analysed cohort comprised 61 females and 55 males with a mean age of 50 ± 12 years. Eligibility required age over 18, a current major depressive episode and failure to remit after at least two conventional monoaminergic antidepressant trials as judged by a psychiatrist; patients with absolute contraindications per the Italian regulatory authority (AIFA), such as untreated hypertension or prior cerebrovascular disorders, were excluded. Clinical and anamnestic data were extracted from medical records at baseline (T0), one month (T1) and three months (T2). Clinician-rated scales used were the Montgomery–Åsberg Depression Rating Scale (MADRS), the 21-item Hamilton Depression Rating Scale (HAM-D-21) and the 21-item Hamilton Anxiety Rating Scale (HAM-A). Response was predefined as a ≥50% reduction in MADRS or HAM-D-21 from baseline and remission as MADRS < 10 or HAM-D-21 < 7. Adverse events were recorded from chart notes; no structured side-effect instrument or inter-rater reliability assessment across centres was reported. Sample size was estimated with G*Power for a repeated-measures ANOVA assuming an expected response of 40% and a planned power of 95%. The extracted text reports a significance level as "0.05 %", which appears to be a typographic extraction issue and is not clearly stated; the authors used SPSS 20.0 for analyses, with two-tailed tests and a reported significance threshold of p < 0.05. Continuous changes in psychometric scores were tested with paired Student t-tests and categorical comparisons with Pearson χ2 tests. Due to treatment discontinuations and patients not yet reaching follow-up, 106 patients contributed data at T1 and 91 at T2.

At baseline the cohort had severe depressive symptoms (mean MADRS 35 ± 8.53), a long illness duration (mean 19 ± 11.05 years) and multiple prior antidepressant trials (mean 3.28 ± 1.89). Twenty-five patients (22.3%) had a history of suicide attempts. Personality disorders were present in 15% and substance use disorders in 6%; 64% had no other psychiatric comorbidity. Regarding concurrent treatments, 57 patients (49.13%) were on SSRIs, 39 (33.62%) on SNRIs and many received augmentation: 66 (56.9%) on mood stabilisers and 67 (57.7%) on antipsychotics. Treatment retention was high: 10/116 (8.62%) discontinued by T1 (seven for inefficacy, two for excessive side effects, one for severe psychomotor agitation) and a further five discontinued before T2 (4.31%); 10 patients had not yet reached T2 at analysis. Mean MADRS scores fell from 35 ± 8.53 at T0 to 22.27 ± 9.81 at T1 (t = 15.79, df = 95, p < 0.0001) and to 14.69 ± 9.88 at T2 (t = 18.07, df = 81, p < 0.0001). Suicidal ideation (MADRS item 10) declined from 2.13 ± 1.58 at baseline to 1.00 ± 0.55 at T1 and 0.94 ± 0.1 at T2 (both comparisons p < 0.0001). By the study definitions, 33 patients (28.4%) were responders at T1 and 13 (11.2%) were remitted; at T2 response rose to 68 patients (64.2%) and remission to 43 (40.6%). The increases from T1 to T2 were statistically significant for both response and remission (χ2 tests, p < 0.0001). Only 29% of T2 remitters had remitted at T1, and a substantial portion of T2 remitters had not been responders at T1. No significant baseline sociodemographic or psychometric differences were observed between three-month responders and non-responders, although responders tended toward longer current episode duration and higher baseline anxiety without reaching significance. Adverse events leading to discontinuation were recorded in three patients at T1 (2.58%), including one case of severe psychomotor agitation. The most frequently noted side effects were dissociative symptoms (39.7%), sedation (28.4%) and transient hypertension (10.3%); manic symptoms, psychomotor agitation, anxiety and headache were each uncommon (around 1.7–2.6%). Overall, 27.6% of patients reported no side effects. The investigators reported no signs of abuse, craving, withdrawal or long-term cognitive, urogenital or hepatic toxicity within the study period. Effectiveness did not differ between patients with and without psychiatric comorbidities (T1 and T2 response and remission p-values non-significant), whereas patients receiving augmentation with mood stabilisers or antipsychotics showed lower response rates (T1 p = 0.023, T2 p = 0.010).

Martinotti and colleagues interpret these findings as support for the safety, tolerability and clinical effectiveness of intranasal esketamine in a heterogeneous, treatment-seeking TRD population. They highlight that the study population was more clinically severe and more comorbid than typical RCT samples—longer illness duration, higher unemployment and frequent augmentation therapy—yet three-month response (64.2%) and remission (40.6%) rates were comparable to those reported in controlled trials. A distinctive observation was the pronounced increase in response and remission between one and three months; most three-month remitters had not remitted at one month, prompting the authors to suggest that the standard induction period should not be the sole criterion for continuing or stopping treatment in real-world practice. Possible explanations offered for the delayed responses include greater baseline illness burden requiring longer exposure to esketamine-induced synaptic plasticity (mTOR/BDNF-mediated long-term potentiation), delayed dose escalation from 56 mg to 84 mg in many patients after T1, and lower adherence in routine care compared with trial settings. Safety outcomes were judged consistent with RCT evidence: dissociation and sedation were common but transient, manic switches were rare, and no addictive behaviours were documented during the observation period. The lower effectiveness observed in patients receiving mood stabilisers or antipsychotics might reflect either greater baseline severity in those individuals or pharmacodynamic interactions that attenuate esketamine's antidepressant effect; the authors call for further study of specific agents (for example lithium, valproate, lamotrigine or atypical antipsychotics). The investigators acknowledge several limitations that temper the conclusions: the open, uncontrolled and retrospective design; absence of inter-rater reliability checks across centres; reliance on charted instead of structured adverse-event assessments; and limited generalisability despite the multicentre, clinical sample being a strength for real-world relevance. Overall, the authors conclude that their observational data complement RCT findings and support esketamine as a useful option in the therapeutic algorithm for TRD, while noting the need for further controlled and mechanistic research.

The authors conclude that in this multicentre, real-world sample esketamine nasal spray demonstrated clinical effectiveness and an acceptable safety and tolerability profile in adults with TRD. They emphasise two notable observations: a substantial proportion of patients showed later (three-month) responses that were not evident at one month, and patients with psychiatric comorbidities did not show inferior outcomes. No evidence of abuse, misuse, withdrawal phenomena or short-term organ toxicity was observed within the study period. The investigators note that these findings supplement randomized trial data but are constrained by the study's open and retrospective design.