Intranasal esketamine represents a novel therapeutic option for Treatment-Resistant Depression (TRD). However, its medium-to-long-term risk-benefit profile in real-world clinical practice remains controversial. Limited data is available regarding its effectiveness, particularly in patients who have not responded to electroconvulsive therapy (ECT), which is the focus of this study.
MethodsWe conducted a 6-month observational study investigating intranasal esketamine treatment in the first TRD patients managed at our care centers. We compared outcomes between patients who had not responded to prior ECT and those for whom ECT had not been attempted.
ResultsThe sample comprised 60 patients exhibiting high chronicity (mean episode duration: 41.2 ± 54.7 months) and complexity, with psychiatric comorbidity observed in 37 cases (61.6%). The mean Montgomery-Åsberg Depression Rating Scale (MADRS) score decreased from a baseline of 33.9 ± 7.8 to 15.7 ± 9.6 at 6 months. No significant differences in therapeutic responses were observed in patients who had not responded to ECT during the current episode.
ConclusionsThis real-world study suggests the opportunity to promptly test intranasal esketamine in ECT refractory subtype of TDR patients.
Approximately one-third of patients with Major Depressive Disorder (MDD) do not respond to conventional treatments and are thus categorized as having Treatment-Resistant Depression (TRD).1,2 The widely accepted definition of TRD is the failure to respond to two adequate trials of antidepressant medications, with each trial involving sufficient dosage and duration.3,4 However, this definition has been criticized for its failure to consider trials of structured psychotherapies and other neuromodulatory antidepressant strategies, including electroconvulsive therapy (ECT).2,5 The absence of response to ECT has been proposed as the most severe form of TRD, given that it has been considered one of the most effective biological therapeutic alternatives for TRD1,6 .
There is a significant unmet clinical need among patients with TRD who fail to respond to ECT. This subgroup comprises approximately 30–40 % of patients with TRD treated with ECT, though prevalence rates vary according to clinical definitions and population characteristics.7 The prognosis for ECT non-responders is poor; the majority persist with severe depressive symptomatology and functional impairment, with only a minority achieving remission via alternative interventions within 6–24 months8 ECT non-responders represent a distinct 'high refractoriness' phenotype, having failed multiple pharmacological and psychotherapeutic strategies in addition to the most potent biological intervention available.7-9 This cohort is characterized by protracted episode duration, greater cumulative treatment failures, and a frequent absence of clinical markers associated with favorable ECT outcomes, such as psychotic features.7,8,10 The persistence of symptoms and the scarcity of robust evidence for subsequent modalities underscore the urgent need for novel therapeutic paradigms tailored to this population.8,9,11
The therapeutic efficacy of esketamine in patients with TRD who have failed to respond to ECT may be attributed to its unique pharmacodynamic profile.2-4,6 As an N-methyl-d-aspartate (NMDA) receptor antagonist, esketamine modulates glutamatergic neurotransmission and promotes synaptic plasticity through pathways distinct from the monoaminergic systems targeted by conventional antidepressants and the generalized seizure induction of ECT.12 This mechanistic divergence suggests that ECT refractoriness is not a definitive predictor of esketamine non-response, as the underlying neurobiological targets remain functionally independent.4-6
Recent real-world evidence suggests that intranasal esketamine induces statistically and clinically significant reductions in depressive symptomatology among patients with treatment-resistant depression (TRD) who are refractory to ECT. Medium-term data indicate sustained therapeutic benefits.6,13,14 Although response rates in ECT non-responders may be lower compared to less refractory TRD cohorts, esketamine remains an effective intervention, notably in reducing depression severity and suicidal ideation.6,13 Discontinuation rates in real-world cohorts are moderate, primarily attributed to insufficient efficacy, adverse events (particularly dissociation, dizziness, nausea, and sedation), or logistical barriers.15-17 The tolerability profile aligns with results from randomized controlled trials; while serious adverse events remain rare, the frequency of transient hypertension and dissociation necessitates rigorous clinical monitoring.18-21 Current literature is constrained by limited sample sizes within ECT non-responder subgroups and short follow-up periods.6,13,18 Furthermore, longitudinal data regarding the impact of repeated esketamine exposure on cognitive function and suicidality in this highly refractory population remain sparse.19,22–25 Consequently, while esketamine represents a viable therapeutic alternative for ECT non-responders, further longitudinal studies are required to establish its long-term efficacy and safety profile in this specific subpopulation.
In summary, while esketamine has demonstrated promising efficacy for Treatment-Resistant Depression (TRD) in rigorous clinical trials, its real-world effectiveness and safety profile remain poorly defined.23 Uncertainty persists regarding long-term outcomes and generalizability across diverse patient phenotypes.4,23 Notably, although ECT is typically reserved for the most refractory cases, ECT non-responders are frequently excluded from pivotal esketamine trials.6 Furthermore, current evidence suggests that ECT may surpass esketamine in efficacy for certain TRD populations, though findings remain inconsistent.2,26 Consequently, it remains unclear whether intranasal esketamine provides clinical benefit for patients who have failed to respond to ECT, and whether such patients can achieve sustained remission during maintenance therapy. The present study addresses this gap by evaluating the real-world, 6-month effectiveness of intranasal esketamine specifically in a cohort of ECT non-responders.
Our primary hypothesis posits that ECT non-responders will exhibit a significantly attenuated therapeutic response on the MADRS scale at the 6-month follow-up compared to patients who did not undergo ECT during the current episode. Furthermore, we hypothesize no significant differences in the safety or tolerability profiles of intranasal esketamine when comparing current-episode ECT non-responders to ECT-naïve patients.
Material and methodsA cohort of the first 60 patients treated with intranasal esketamine in Mallorca (Spain) was analyzed. Despite the lack of data regarding eligible patients who declined treatment, all those who clinically initiated the protocol provided consent for this retrospective study. The study sample comprised outpatients diagnosed with TRD who initiated treatment with intranasal esketamine at Hospital Universitario Son Llátzer (HUSLL) or Hospital Universitario Son Espases (HUSE). Patients included in the study during the recruitment period (January 2023 - March 2025) completed at least six months of treatment. Treatment initiation adhered to current Spanish reimbursement criteria: combination with an SSRI or SNRI with TRD who have not responded to at least three different antidepressant strategies, at least one of which is a combination or augmentation strategy, in the current major depressive episode. All patients provided informed consent after receiving a thorough explanation of the study's characteristics and objectives. Ethical approval was provided by the Research Ethics Committee of the Balearic Islands (IB 5653/24 Eoms).
TRD was defined as the failure to respond to at least two trials of antidepressant pharmacological treatment of adequate duration (minimum of six weeks) and dosage (maximum tolerated dose within the therapeutic range). Successive pharmacological trial failures were determined clinically. Patients with bipolar depression, psychotic depression, substance use disorders, non-severe personality disorders, or mild neurocognitive symptoms were not excluded. Patients with substance use disorders, severe personality disorders, or important neurocognitive symptoms were excluded. The current depressive episode and comorbidities were clinically diagnosed by each patient’s treating psychiatrist.
Bilateral (bitemporal) ECT was previously administered to 18 patients, with sessions conducted 2–3 times per week on alternating weekdays for a minimum of 12 sessions. Anesthesia induction was performed at the discretion of the attending anesthesiologist, following standard clinical protocols and tailored to individual patient characteristics and physical status. The electrical stimulus dose was titrated based on electroencephalographic (EEG) response, with intensity adjusted to achieve a postictal suppression index exceeding 80 %. To evaluate ECT efficacy, in addition to the postictal suppression index, the quality and duration of EEG seizure activity were considered.27,28 Non-response was defined by the attending psychiatrist following the 12-session course. Additionally, ECT failure was defined by poor treatment tolerance (as determined by the attending psychiatrist) or the withdrawal of patient consent.
The esketamine administration protocol was standardized across both sites, adhering to the regulatory requirements established by the Spanish health authorities and the manufacturer’s guidelines. The induction phase lasted between 4 and 12 weeks. During the maintenance phase, dosing frequency (once weekly or every other week) was adjusted based on the patients' clinical course and tolerability. The proportion of patients who achieved the 84 mg dose level during both the induction and maintenance phases was recorded. Although treatment was extended beyond 6 months in some instances, longitudinal data following this period were not collected. While antidepressant regimens remained stable prior to treatment initiation, the selection of specific agents, dosages, and augmentation strategies was left to the discretion of the attending psychiatrist, reflecting real-world clinical practice. Concurrent psychotherapy was delivered via diverse modalities and frequencies; where applicable, these interventions were maintained throughout the esketamine treatment phase.
All eligible patients were invited to participate during scheduled clinical visits. For patients who had completed treatment and did not have an upcoming appointment, they were contacted by telephone and offered an in-person appointment to participate. If patients were unable to attend in person but agreed to a telephone interview, their consent was made by mail, and data collection subsequently proceeded. Data was collected both prospectively and retrospectively by consulting medical records.
Montgomery-Åsberg Depression Rating Scale (MADRS) scores were collected from clinical records at baseline (T1) and after 1 month (T2), 3 months (T3), and 6 months (T4) of treatment in accordance with the established clinical protocol. MADRS is a clinician-rated scale that assesses the severity of depression across 10 items related to 10 distinct depressive symptoms.29 The scale was specifically designed to evaluate changes in the intensity/severity of depressive symptomatology, necessitating its repeated application in patients with a clinical diagnosis of depression, with score variation being the relevant data point. Therapeutic response was defined as a reduction of 50 % or more from pre-treatment scores. A MADRS total score ≤ 12 was utilized to define remission. This cutoff is higher than conventional criteria, reflecting the highly refractory nature and clinical complexity of the current sample.30,31
Trained psychiatrists overseeing the intranasal esketamine program conducted the MADRS assessments. The 6-month follow-up data includes only patients who achieved that milestone; notably, no significant intergroup differences were observed in study completion rates.
Data analysis was performed using SPSS. A significant level of p < 0.05 was applied. Results are presented as mean (standard deviation) for continuous variables and frequency ( %) for categorical variables. Comparisons between groups with and without previous ECT were conducted using the Mann–Whitney U test for continuous variables and the chi-square test for categorical variables. The choice of nonparametric tests was driven by the limited sample size, particularly in the subgroup of patients with prior ECT non-response (n = 18), and the unequal group distribution (42 vs. 18 patients), which may compromise the robustness of parametric methods. Given the real-world design of the study, analyses were conducted using available data only. Participants who completed the corresponding assessments at each time point were included in the analysis, and no imputation methods, such as last observation carried forward (LOCF), were applied. An intention-to-treat (ITT) approach was used to avoid biases.
Within-group longitudinal changes across the four assessment time points (T1-T4) were examined separately patients with prior ECT and those without prior ECT using Friedman tests for repeated measures. When the omnibus test was significant, pairwise post-hoc comparisons between time points were performed using Wilcoxon signed-rank tests with Bonferroni-adjusted p-values to control multiple comparisons. To approximate between-group differences over time, change scores (Δ) were calculated for each participant by subtracting baseline values (T1) from the subsequent time points (T2-T4). These change scores were then compared between patients with prior ECT and those without prior ECT using Mann-Whitney U tests.
ResultsThe sample comprised 60 patients exhibiting high chronicity (mean episode duration: 41.2 ± 54.7 months) and complexity, with psychiatric comorbidity observed in 37 cases (61.6 %). The mean Montgomery-Åsberg Depression Rating Scale (MADRS) score decreased from a baseline of 33.9 ± 7.8 to 15.7 ± 9.6 at 6 months. Of these, 23 patients (38.33 %) met criteria for response, and 14 (23.33 %) achieved remission. Twenty-four patients (40 %) discontinued treatment or were excluded before 6 months, primarily due to lack of efficacy. Among the 60 patients recruited, 18 had previously undergone ECT during the current depressive episode without success, either due to lack of response or poor tolerability. No significant differences in therapeutic response or discontinuation rates were observed in patients who had not responded to ECT during the current episode. Fig. 1 shows the flow diagram of the participants included.
The main sociodemographic and clinical characteristics of this subgroup are presented alongside those of the remaining patients in Table 1. Table 2 reports descriptive statistics for MADRS scores and the response and remission rates across group and time points. Fig. 2 shows the temporal trajectories of MADRS scores between patients with and without prior ECT.
Characteristics of the sample.
Note: “With prior ECT” indicates patients who received electroconvulsive therapy during the current depressive episode before esketamine initiation; “Without prior ECT” indicates patients with no such treatment during the current episode. ECT: Electroconvulsive therapy; BMI: Body mass index; SD: Standard Deviation; MDD: Major Depressive Disorder; MADRS: Montgomery-Åsberg Depression Rating Scale; AD: Antidepressants.
Descriptive statistics of MADRS scores by group and time point.
Note. MADRS, Montgomery-Åsberg Depression Rating Scale; ECT, electroconvulsive therapy.
In patients without prior ECT, a significant effect of time was observed (Friedman test: X2(3)=51.852, p ≤ 0.001), indicating changes in MADRS scores across assessments. Post-hoc Wilcoxon signed-rank tests showed significant decreases in MADRS scores between T1-T2 (p ≤ 0.001), T1-T3 (p ≤ 0.001), and T1-T4 (p ≤ 0.001), with no additional pairwise differences being significant (Table 3).Similarly, in patients with prior ECT, a significant effect of time was observed (Friedman test: X2 (3)=20.495, p ≤ 0.001), indicating changes in MADRS scores across assessments. Post-hoc Wilcoxon signed-rank tests showed significant decreases in MADRS scores between T1-T2 (p ≤ 0.001), T1-T3 (p ≤ 0.001), and T1-T4 (p = 0.003), with no additional pairwise differences being significant (Table 3).
Post-hoc pairwise comparisons of MADRS scores across time points.
Note. MADRS, Montgomery-Åsberg Depression Rating Scale; ECT, electroconvulsive therapy. Wilcoxon signed-rank tests were performed. P-values adjusted using Bonferroni correction (p-value < 0.05 / 12 = 0.004). * Statistically significant at p value < 0.004.
Change scores (Δ) were calculated to examine between-group differences in reductions of MADRS scores over time. No significant differences were observed between patients without prior ECT and those with prior ECT at any assessment point, indicating that the magnitude of MADRS score reduction over the study prior was comparable between groups (Table 4).
Between-group comparisons of change scores of MADRS scores.
Note. MADRS, Montgomery-Åsberg Depression Rating Scale; ECT, electroconvulsive therapy; T1, baseline; T2, 1 month; T3, 3 months; T4, 6 months. Mann-Whitney U tests were performed.
In a sample of 60 patients with TRD, patients who had not responded to ECT improved with intranasal esketamine to the same extent as those in whom ECT had not been administered during the current episode (see Fig. 2). No significant differences in adverse events were observed between groups (see Table 1). From a clinical-practical perspective, this finding encourages the use of intranasal esketamine in these patients, despite their highly chronic, complex, and refractory profiles. Conceptually, this supports a differential mechanism of action between the two therapies, as previously described, although commonalities have also been noted.26,32
Previous research by Ibrahim et al.33 investigated the efficacy of intravenous ketamine in 17 patients with ECT-resistant depression compared to 23 patients who had not undergone ECT. While similar acute responses were observed, measurements were confined to a 4-hour post-infusion window. More recent studies evaluating the response to intranasal esketamine in patients TRD, stratified by prior ECT resistance, have consistently demonstrated no significant differences in outcomes around the 3-to-6-month mark.3,6 Our findings corroborate and strengthen this emerging body of evidence, confirming the sustained effect for a minimum of 6 months.6,13,14
The interpretation of therapeutic outcomes in patients with ECT-refractory TRD must be contextualized within the high baseline severity and chronicity inherent to this cohort. Evidence from real-world data and clinical trials indicates that response and remission rates for intranasal esketamine are modest yet clinically significant, even in highly refractory populations.15,16,34 Specifically, 8-week remission rates approximate 25–30 %, demonstrating superiority over quetiapine augmentation in head-to-head comparisons.21 Nevertheless, the magnitude of this benefit and its clinical relevance is a subject of ongoing debate, particularly for patients with extensive prior treatment failures.19,23
Another crucial aspect to emphasize is that both in this study and others cited, intranasal esketamine was administered as an adjunctive treatment to other antidepressant therapies of proven efficacy.34 For instance, all patients were concurrently receiving antidepressant medications, and the majority were also on other augmenting psychotropic drugs. Furthermore, in our study, some patients received structured psychotherapy prior to and/or during esketamine administration (see Table 1). Therefore, the observed effect of esketamine emerged within a context where multiple therapeutic agents were utilized synergistically. This approach is strongly supported by evidence, particularly in cases of (TRD) .3 Moreover, it has been suggested that this synergy might be more potent if it is also synchronous, meaning that the administration of therapeutic agents coincides temporally.5 In fact, there is ongoing debate about whether the therapeutic outcomes of intranasal esketamine could be further improved if assisted psychotherapy were consistently implemented during its administration, similar to practices with other psychedelic compounds like psilocybin.35,36 Some studies appear to support this notion.37 Interestingly, this possibility is not feasible with ECT, which may represent an additional disadvantage. Conversely, some research indicates that patients prefer intranasal esketamine over ECT due to their tolerability, and this preference may also contribute to a better therapeutic response with the former.38,39 These arguments support the suggestion of using intranasal esketamine before ECT in the TDR therapeutic algorithm.32
Discontinuation rates for esketamine in real-world settings are substantial, frequently exceeding 40 %, which aligns with the proportion observed in our study.6,19,40 This attrition is primarily driven by tolerability issues—notably dissociation, dizziness, and nausea—alongside unmet clinical expectations, logistical barriers inherent to supervised administration, and insufficient early therapeutic response.6,40 Notably, these rates markedly surpass those observed in controlled trials, where attrition due to adverse events typically ranges from 5 % to 15 %.19,40 Service-level factors and highly selective patient inclusion criteria likely contribute to this disparity.
Selection bias remains a significant concern; recruitment and consent procedures often exclude patients with greater comorbidity or lower adherence, thereby limiting the generalizability of the findings. Furthermore, the use of combined retrospective and prospective data extraction may preferentially capture patients with higher levels of clinical engagement. The reliance on open-label designs without active comparators further restricts the ability to draw causal inferences.6,19
Statistical power for subgroup analyses is often insufficient—particularly regarding ECT non-responders—increasing the risk of Type II errors (false negatives) and precluding robust conclusions concerning differential efficacy.6 It is critical to note that the absence of statistically significant differences does not equate to proof of therapeutic equivalence.
The clinical application of these findings warrants a cautious approach. While esketamine is FDA-approved for adults with TRD following at least two failed antidepressant trials—and could be a viable alternative for ECT non-responders—patient selection must be individualized. Clinical management should prioritize rigorous monitoring for tolerability and the alignment of patient expectations regarding the duration and magnitude of benefit.19,21,23 Currently, the evidence base for ECT-refractory populations remains sparse. Prospective, placebo-controlled, longitudinal studies are required to clarify long-term outcomes and the optimal positioning of esketamine within treatment algorithms.
Beyond the modest sample size, the primary limitations of this study include its naturalistic, retrospective design, which lacks randomization and a control group. Another limitation of our study is that it includes only the first 60 patients recruited into the esketamine program. This may imply that the initial referrals were patients with greater illness severity, in whom the number of treatment lines prior to esketamine initiation exceeded those recommended by clinical practice guidelines (Table 1).41 Consequently, the observed outcomes may be more modest than those expected in patients referred to the esketamine program earlier in the course of their depressive episode. Furthermore, the relatively high attrition rate and the reliance on available-case analyses may have introduced selection bias and limited the generalizability of the findings. Nevertheless, this approach was adopted to reflect real-world clinical practice and to avoid potentially speculative data imputations within this specific population. Additionally, due to the observational nature of the study, the precise timing of treatment discontinuation was not systematically documented; consequently, a survival analysis or formal time-to-event analysis could not be performed.
ConclusionsOur findings indicate that intranasal esketamine is a feasible intervention that provides clinically meaningful benefits for patients with treatment-resistant depression (TRD) refractory to ECT. The observed response rates and tolerability profiles align with those reported in broader TRD populations.6,13,14 However, several limitations warrant consideration, notably the retrospective design, the modest sample size of ECT non-responders, the absence of a control group, and the limited follow-up period. These factors preclude definitive conclusions regarding efficacy and long-term outcomes in this highly refractory subgroup.6,18,19 Consequently, prospective controlled trials with longer longitudinal assessments are required to validate these preliminary results and to identify optimal dosing protocols, predictors of response, and treatment durability.19,22,23 Nevertheless, given the paucity of therapeutic alternatives for this population, real-world data support esketamine as a viable clinical option for ECT non-responders.
Ethical considerationsEthical approval for this study was obtained from the Research Ethics Committee of the Balearic Islands (IB 5653/24 Eoms). All procedures were conducted in accordance with the ethical standards of the relevant institutional and national research committees. Written informed consent was obtained from all participants after they received a comprehensive explanation of the study's characteristics and objectives.
FundingThis research did not receive any specific grant from funding agencies in the public, commercial, or non-profit sectors.
Data availability statementDe-identified data available on reasonable request, subject to ethical/legal constraints
Declaration of generative AI and AI-assisted technologies in the manuscript preparation processDuring the preparation of this work the authors used Gemini [Google] in order to improve language and readability. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the published article.
CRediT authorship contribution statementRocío Gómez-Juanes: Conceptualization, Methodology, Writing – original draft, Writing – review & editing. Guillermo Mompaler Lázaro: Methodology, Investigation, Writing – review & editing. Adoración Castro: Data curation, Formal analysis, Writing – review & editing. Guillem Navarra-Ventura: Data curation, Formal analysis, Writing – review & editing. Josep Antoni Mestre Quetglas: Investigation, Software, Writing – review & editing. Ana Bolon Marset: Investigation, Software, Writing – review & editing. Julia Van der Walt Monasterio: Investigation, Software, Writing – review & editing. Jaume Andreu Garcias Trullols: Investigation, Software, Writing – review & editing. Miquel Roca: Conceptualization, Supervision, Writing – original draft. Margalida Gili: Conceptualization, Project administration, Writing – original draft. Joan Salvà: Conceptualization, Methodology, Writing – original draft. Mauro Garcia-Toro: Conceptualization, Supervision, Writing – original draft, Writing – review & editing.
MR has received research funding or grants from Adamed, Janssen, Lundbeck, and Viatris. GN-V, JS, MG-T and RG-J have served as investigators in clinical trials sponsored by Janssen. The remaining authors declare no conflict of interest.
