Published on 19 September 2026

Analysis · OCD and related disorders · Psychopharmacology

In treatment-resistant OCD with comorbid depression, does intranasal esketamine work faster on mood than on obsessions?

▬ Publication
BMC Psychiatry · 2026;26:469 · López-Rodríguez et al.
DOI 10.1186/s12888-026-08119-5
PMID 42036648
Scientific 63
Editorial 75

Editorial collections The Monday Morning Patient PEB Decode Tomorrow Psychiatry A Study Under the Microscope

The essentials

Eight patients with treatment-resistant obsessive-compulsive disorder and a severe comorbid major depressive episode received add-on intranasal esketamine for twelve weeks at a university reference hospital. Depression scores fell by an average of 48.8% and obsessive-compulsive scores by 30.3%, with four of eight patients responding on each dimension. The most interesting finding is not the size of these changes but their timing: mood moved within the first weeks, obsessions later, between weeks eight and twelve. Tolerability was described as satisfactory, with transient dissociation in half the patients, no serious adverse event documented and no discontinuation for an adverse effect. This should be read for exactly what it is, an eight-patient case series with no control group, in a single center, with prior treatments maintained. Nothing here allows separating what belongs to the drug from what belongs to natural evolution or to intensified monitoring.

The context

Treatment-resistant obsessive-compulsive disorder is one of the harder impasses of outpatient psychiatry. After two adequately dosed serotonin reuptake inhibitors, clomipramine, cognitive behavioral therapy with exposure and response prevention, and at least one augmentation strategy, few options remain, and the wait between successive trials is measured in months.

When a severe, equally resistant depressive episode is added to the picture, the question becomes practical: is there an option that can act on both fronts? Intranasal esketamine has an established place in treatment-resistant depression. Whether it also helps in OCD is a separate question, and it is the one this study addresses.

The study at a glance

Question (PICO)
Population
8 patients, four men and four women, mean age 47.3 years with a standard deviation of 8.8 years. Treatment-resistant OCD defined by a Yale-Brown score of 24 or higher and level 3 or 4 on the Pallanti and Quercioli resistance scale, together with a concurrent major depressive episode with a Montgomery-Åsberg score of 35 or higher. Documented failure of at least two serotonin reuptake inhibitors, clomipramine, cognitive behavioral therapy with exposure and response prevention, and at least one augmentation strategy. These failure criteria define OCD resistance; the publication sets no separate resistance criterion for the depressive episode, for which only severity is required at inclusion. Consecutive inclusions from July 2023 to August 2024 at a Spanish university reference hospital.
Intervention
Intranasal esketamine, 56 to 84 mg per session, following the antidepressant protocol approved for treatment-resistant depression: induction at two sessions per week for about four weeks, then one session per week through week twelve, with minor adjustments left to clinical judgment. The total number of sessions actually received ranged from 5 to 12 across patients. Ongoing treatments, including serotonin reuptake inhibitors and antipsychotics, were maintained throughout the observation period.
Comparator
None. This is a case series, with each patient serving as their own baseline.
Outcomes
Yale-Brown Obsessive Compulsive Scale for obsessions and compulsions, Montgomery-Åsberg Depression Rating Scale for depression, measured at baseline and at weeks 1, 4, 8 and 12. Response defined as a decrease of at least 35% on the first scale and at least 50% on the second.
Design
Prospective observational case series, with assessment timepoints and response criteria defined a priori and a favorable opinion from the hospital’s clinical research ethics committee (Acta 25/24, CEIm file 2024/2492), Oxford level of evidence 4. Two-tailed paired t-test, Wilcoxon test as a sensitivity analysis, Cohen’s dz effect size and 95% confidence intervals.

Quality control

Point checked	Judgment
Prospective design	Pre-specified protocol
FindingPlanned data collection, with assessment timepoints fixed in advance, response criteria set a priori and a favorable ethics opinion. This clearly distinguishes the study from a retrospective chart review. The publication reports no public protocol registration: only the reference of the ethics opinion is given.
Definition of the population	Strict and explicit
FindingResistance criteria are detailed on both dimensions, with numerical thresholds. The population is therefore recognizable, which matters for judging which patients this result could someday apply to.
Sample size	Eight patients
FindingAt this size, each patient accounts for one eighth of the mean, and the confidence intervals are necessarily very wide. Effect sizes calculated on eight subjects are fragile estimates, whatever their numerical value.
Absence of a comparator	Decisive
FindingWithout a control group, the drug’s effect cannot be separated from spontaneous evolution, expectation effects, the intensified clinical contact of twice-weekly sessions, or the maintained treatments.
Outcome assessment	No blinded assessment
FindingThe authors themselves state the absence of a control group, of randomization and of blinded assessment. The scales are scored by clinicians on the same team that administers the treatment, which bears directly on how the score changes should be interpreted.
Multiple analyses	Uncorrected
FindingTwo scales compared between baseline and week twelve, complemented by a nonparametric sensitivity analysis, across five measurement timepoints in total. The authors explicitly state that no correction for multiplicity was applied, given the exploratory nature of the work and a sample of eight. The p-values should therefore be read as descriptive.
Independence	Public funding
FindingFunded by the Instituto de Salud Carlos III, the Fundació La Marató de TV3 and FEDER funds, with institutional support from the CERCA program of the Generalitat de Catalunya. The authors declare no conflicts of interest. No industry funding is declared, which is worth noting for a molecule still under patent.

The results

4/8

patients responding on the obsessive-compulsive dimension, and four of eight also on the depressive dimension, after twelve weeks. The two responses do not occur at the same time.

Result	What the source reports
Depression	37.5 to 19.75, a mean decrease of 17.75 points
FindingPaired t-test with 7 degrees of freedom, statistic of 4.55 and p = 0.0026, Cohen’s dz effect size of 1.61, 95% confidence interval for the mean decrease of 8.53 to 26.97 points, Wilcoxon test at 0.018. The 48.8% decrease highlighted by the authors is the mean of individual percentage changes; relative to the group means, it is 47.3%.
Obsessions and compulsions	30.13 to 21.00, a mean decrease of 9.13 points
FindingStatistic of 4.28 with 7 degrees of freedom and p = 0.0037, Cohen’s dz effect size of 1.51, published 95% confidence interval of 4.22 to 14.03 points, Wilcoxon test at 0.0078. The 30.3% decrease is here relative to the group means. The mean score at week twelve remains at 21, that is, still clearly within a symptomatic range.
Response and remission	4/8 on each dimension, 2/8 in depressive remission
FindingRemission is defined as a Montgomery-Åsberg score of 10 or lower at the end of follow-up. The published 95% Wilson intervals run from 0.215 to 0.785 for each of the two response rates and from 0.071 to 0.591 for remission, which by itself sums up what eight patients allow. Recounting the published individual data shows that responders overlap only partially between dimensions, with three patients responding on both. No obsessive-compulsive remission criterion is reported, which reflects an absence of analysis, not an absence of remission.
Differential timing	Mood in weeks 1 to 4, obsessions in weeks 8 to 12
FindingThis is the study’s most useful observation: the drop in mood is largely achieved within the first four weeks, while a clearer improvement in obsessions appears only between weeks eight and twelve, among responders. It rests, however, on trajectories described in eight patients, and no formal test of a difference in kinetics between the two dimensions is reported.
Tolerability	Dissociation 50%, nausea 12.5%, no discontinuation for an adverse effect
FindingNo serious adverse event was documented. Dissociation, reported in four of eight patients, is described as transient, resolving spontaneously within twenty to thirty minutes; one of eight patients had moderate nausea. Across eight patients and twelve weeks, the absence of discontinuation does not allow the frequency of rare events to be estimated.

Critical appraisal

Domain	Risk of bias
Absence of randomization	Structural
FindingInherent to the design. A case series describes a trajectory; it does not test a causal hypothesis and was not built to do so.
Regression to the mean	Expected
FindingInclusion requires very high scores, a Montgomery-Åsberg score of 35 or higher and a Yale-Brown score of 24 or higher. Selecting on extreme values makes a subsequent average decrease expected even without active treatment; the authors themselves acknowledge this mechanism as a plausible contributor to part of the observed reduction.
Co-interventions	Maintained
FindingPrior treatments, including serotonin reuptake inhibitors and antipsychotics, were continued throughout the observation period, which matches real-world practice but prevents attributing the change to the added drug alone. The authors further specify that concomitant psychotherapy was neither standardized nor quantified, and that esketamine dosing could undergo minor adjustments according to clinical judgment.
Unblinding	Unavoidable
FindingDissociation, felt by half the patients, makes the drug identifiable to whoever receives it. Here the question does not arise, since no blinded assessment was conducted. In a future controlled trial, however, this point would condition interpretation.
Single center	Highly specialized
FindingA reference hospital department for OCD. The result reflects the center’s expertise and patient selection as much as the molecule used.
Adequacy of the conclusions	Calibrated
FindingThe authors do not claim demonstrated efficacy and present their own limitations. This analysis does not go beyond what they state.

Level of evidence

Scientific63/100

Editorial75/100

Oxford level of evidence 4, prospective case series. Confidence is reasonable on the description itself, the reported scores, the observed tolerability and the trajectory of these eight patients, whose data collection was planned and whose instruments are validated. It is very low on any attribution of these changes to the administered drug, and the absence of a control group in a population selected on extreme scores is enough to explain why. The temporal gap between the two dimensions is the most worthwhile element to explore further, but it remains an observation in eight patients, without formal testing, and cannot be presented as an established fact. It should also be recalled that the absence of a serious adverse event in eight patients is not a safety argument.

The colleague test

What an experienced colleague would say if shown this study in two minutes, between two consultations.

Eight patients, I know what that’s worth. But these are eight patients I know well, the ones for whom I’ve tried everything and who keep coming back every three months with nothing moving. What stays with me isn’t the percentage, it’s the timeline: if I ever refer someone to this protocol and the OCD hasn’t budged at four weeks, I’ll know not to draw conclusions yet. That’s not much, and it’s already useful.

Translation for practice: no new indication is created here. What the study contributes is a hypothesis about the kinetics of response, worth keeping in mind so as not to judge a protocol run in a specialized setting too soon.

What you can do with this

Do not conclude that there is no effect on obsessions at four weeks when such a protocol is conducted in a specialized hospital setting. The available data, though very limited, suggest a slower kinetics than for mood.
Check the regulatory framework that applies before drawing any conclusion about referral. What this publication supports stops at its own protocol: esketamine was administered in hospital, under supervision, following the antidepressant protocol approved for treatment-resistant depression, in patients with a severe depressive episode; the obsessive-compulsive benefit is an associated observation here, not an indication. A clinical signal is not a regulatory authorization: this Spanish publication says nothing about how esketamine is authorized, prescribed or dispensed in any given country, on-label or off-label. Verify the rules in force in your own jurisdiction, for instance with your national regulatory body (in France, the Haute Autorité de santé), rather than assuming this source settles the question.
If such a protocol is being considered, inform the patient of how frequent transient dissociation was in this series, observed in half the patients.
When presenting these data to a patient or a family, distinguish clearly between the description of an eight-case series and the result of a controlled trial. The difference is not one of degree, it is one of kind.
Keep documenting prior treatment failures precisely, since it is the strict characterization of resistance that makes these patients identifiable for future trials.

Frequently asked questions

Can esketamine be offered to a patient with treatment-resistant OCD?

Not on the basis of this study. An eight-case series with no control group does not establish an indication; at best it justifies a controlled trial. In this series, the drug was administered following the antidepressant protocol approved for treatment-resistant depression, in patients with a severe depressive episode. The publication establishes nothing about any country’s regulatory framework, which should be checked against the rules currently in force before any decision.

Isn’t an effect size of 1.51 considerable?

The figure is high, but it is calculated on eight patients selected for their extreme scores, with no comparator. Under these conditions, an effect size mostly reflects regression to the mean and the homogeneity of a small group, not the efficacy of a drug.

Is the gap between depressive and obsessive response an established finding?

No. It is described from mean trajectories in eight patients, with no statistical test of a difference in kinetics reported. It is an interesting hypothesis, not a result.

Does the absence of treatment discontinuation mean the drug is well tolerated long term?

No. Across eight patients followed for twelve weeks, only frequent, early effects can come to light. Nothing here can be said about rare effects or about tolerability beyond three months.

What should clinicians outside a specialized center take from this?

Mainly the characterization of this population, severe on both dimensions, and the idea that a longer assessment window may be needed on the obsessive-compulsive dimension. In this series, administration took place in a specialized hospital unit, under supervision, and the authors explicitly tie the feasibility they describe to that setting.

Annotated bibliography

Source study. López-Rodríguez S, Segalàs C, Real E, Urretavizcaya M, Bertolín S, Menchón JM, Del Pino Alonso. Repeated intranasal esketamine augmentation in treatment-resistant obsessive-compulsive disorder with comorbid major depressive disorder: a prospective case series. BMC Psychiatry. 2026;26:469. Received August 20, 2025, accepted April 22, 2026, published online April 26, 2026. DOI 10.1186/s12888-026-08119-5. PMID 42036648. Funding: Instituto de Salud Carlos III (PI22/00752 and PI25/01407) and Fundació La Marató de TV3 (grant 202201), co-funded by FEDER funds; institutional support from the CERCA program of the Generalitat de Catalunya; open-access publication partly funded by the Universitat de Barcelona under its agreement with Springer Nature. Conflicts of interest: the authors declare none. The publication includes no supplementary material: the data availability statement specifies that all data generated or analyzed are contained in the published article and that no additional dataset is available.

Regulatory framework. The publication analyzed is Spanish and documents no country’s regulatory framework for intranasal esketamine. The conditions of authorization, prescribing, dispensing and administration, including the status of off-label prescribing, vary by jurisdiction and are to be established against the texts currently in force locally, for example the summary of product characteristics and the opinion of the relevant national health authority (in France, the Haute Autorité de santé). These conditions were not verified here, and no claim in this article rests on them.

Editorial collections