J&J Medical Connect
SPRAVATO®

(esketamine)

J&J Medical Connect

Connect with us

  • Products

This information is intended for US healthcare professionals to access current scientific information about J&J Innovative Medicine products. It is prepared by Medical Information and is not intended for promotional purposes, nor to provide medical advice.

SPRAVATO® (esketamine)
Medical Information

Baseline Demographics and Clinical Characteristics Associated With Response or Remission

Last Updated: 06/10/2026

SUMMARY


Baseline Characteristics Associated With Response/Remission1-14 
Study
Baseline Characteristics Associated With Outcomes
Himedan et al1 
When SPRAVATO was compared with PBO:
  • Higher disease severity (MADRS >37), higher number of prior AD failures (≥3): X Response, Remission (relative to MADRS ≤37 or 2 prior failures)
Maia et al2
(ESCAPE-TRD)
  • Overweight/obesity: ✓ Remission (relative to underweight/normal)
  • ≥3 prior antidepressant failures: ✓ Remission (relative to 2 prior failures)
  • Age, sex, baseline MADRS severity, number/duration of episodes:  X Remission
Turkoz et al3 
(TRANSFORM-1 and TRANSFORM-2)
In patients who received SPRAVATO+oral AD:
  • Younger age category (18-44y): ✓ Response, Remission (relative to 45-64y)
  • Employment: ✓ Response, Remission (relative to unemployment)
  • GAD-7 total score <10: ✓ Response, Remission (relative to ≥10)
Borentain et al4 
(TRANSFORM-1 and TRANSFORM-2)
When SPRAVATO+oral AD was compared with PBO+oral AD:
  • Sleep disturbance: X Response,  Remission
Daly et al5 
(TRANSFORM-2)
When SPRAVATO+oral AD was compared with PBO+oral AD:
  • Comorbid anxiety symptoms/disorder: X Response, Remission
Elmaadawi et al6 
  • Absence of psychiatric comorbidity: ✓ Response  
  • Borderline personality disorder: X Response
  • Greater number of prior AD failures: X Response
  • Age: X Response, Remission
  • Sex: ✓ Response in male (relative to female on MADRS only), X Remission
Clemens et al7 
  • Medicaid insurance: X Remission (relative to commercial insurance)
  • Higher augmentation burden: X Remission
  • Age, sex, race, and psychiatric comorbidity: X Remission
Ansari et al8 
  • Obesity (BMI ≥30 kg/m²): ✓ Response (relative to non-obese)
  • BMI (continuous), age, sex: X Response,  Remission
Molero et al9 (INTEGRATE)
  • Patients with a hospitalization during the current depressive episode, including each additional hospitalization: X Response
  • Benzodiazepine use: X Response
  • Demographics and other clinical characteristics (including age, sex, age at first depressive episode, previous depressive episodes, duration of current depressive episode, suicide attempts during the current episode, employment status, medical comorbidities, substance use, psychotic symptoms in the current episode, and family history of mood disorder): X Response
Patarroyo-Rodriguez et al10 
Based on total QIDS-SR:
  • Hypersomnia: ✓ Response (relative to insomnia)
  • Atypical neurovegetative symptoms: ✓ Response
  • Insomnia phenotypes: X Response
Miller et al11 
  • Depressed mood, anhedonia, feeling of failure: ✓ Response
  • Sleep difficulties: associated with nonresponse
  • Suicidal ideation: associated with nonresponse
  • Age, sex, treatment route: X Response
Pettorruso et al12 
Higher scores on the following were associated with a higher likelihood of response:
  • Inner tension (MADRS item 3): ✓ Response at 1 month
  • Inner tension (MADRS item 3), pessimistic thoughts (MADRS item 9), anhedonia (MADRS item 8), reported sadness (MADRS item 2), concentration difficulties (HAM-D item 5), hyperthymic temperament, feelings of guilt (HAM-D item 2), fears (HAM-A item 3), and cardiovascular anxiety (HAM-A item 9): ✓ Response at 3 months
  • Anhedonia (MADRS item 8), anxiety (BPRS item 2), comorbidity with bipolar disorder, motor tension (BPRS item 19), and emotional blunting (BPRS item 16): ✓ Remission at 3 months
  • See detailed summary below for factors associated with a lower likelihood of response or remission
Cavalcanti et al13 
  • High perceived stress: X Remission
✓ = Association with increased response or remission ratesX = Depending on the study, factors were explicitly evaluated and reported as either not associated with response or remission in some studies or was associated with a decrease in response or remission in other studies. Please read the details of the studies below.
AD, Antidepressant; BMI, body mass index; BPRS, Brief Psychiatric Rating Scale; GAD-7, Generalized Anxiety Disorder 7-item scale; HAM-A, Hamilton Anxiety Rating Scale; HAM-D, Hamilton Depression Rating Scale; MADRS, Montgomery–Åsberg Depression Rating Scale; PBO, placebo; QIDS-SR, Quick Inventory of Depressive Symptomatology–Self-Report; TRD, Treatment resistant depression.

CLINICAL DATA

POST HOC ANALYSES
Studies
Results
Himedan et al (2026)1 conducted a post hoc subgroup analysis of a phase 4, randomized, double‑blind, PBO‑controlled, multicenter study15 including 378 adults with TRD randomized to receive SPRAVATO nasal spray (56 mg or 84 mg) or PBO twice weekly for 4 weeks following an antidepressant washout period.
The analysis evaluated associations between baseline disease severity (≤37 vs >37) and number of prior unsuccessful antidepressant treatments (2 vs ≥3) with treatment outcomes (change in MADRS total score, response [≥50% reduction], and remission [MADRS ≤12] at day 28).
Overall, 52.6% of patients had a baseline MADRS total score ≤37, and 59.3% had 2 prior treatment failures.
  • Compared to PBO, SPRAVATO demonstrated improvements (LS mean difference [95% CI]) in MADRS total score at Day 28 in patients with the following baseline MADRS scores:
    • Score ≤37: SPRAVATO 56mg (-3.7 [-7.72, 0.28]); SPRAVATO 84mg (-7.9 [-11.59 to -4.13)
    • Score >37: SPRAVATO 56mg (-6.4 [-10.39 to -2.42]); SPRAVATO 84mg (-5.3 [-9.41 to -1.27])
  • Based on baseline MADRS scores, the response and remission rates at day 28 were as follows:
    • MADRS ≤37: 27.5%, 36.7%, and 18.8% for SPRAVATO 56 mg, 84 mg, and placebo; 22.5%, 26.5%, 9.4%, respectively
    • MADRS >37: 33.3%, 20.0%, and 11.2% for SPRAVATO 56 mg, 84 mg, and placebo; 14.3%, 17.5%, 5.6%, respectively.
  • By number of prior antidepressant treatment failures, response and remission rates at day 28 were as follows:
    • 2 prior failures: 32.7% and 36.2% (SPRAVATO 84 mg and 56 mg) vs 15.7% (PBO); 25.5% and 23.4% vs 8.3%, respectively.
    • ≥3 prior failures: 23.5% and 22.9% (SPRAVATO 84 mg and 56 mg) vs 14.3% (PBO); 17.6% and 11.4% vs 6.5%, respectively.
  • Regardless of baseline disease severity (MADRS ≤37 vs >37) and number of prior antidepressant failures (2 vs ≥3), patients treated with SPRAVATO achieved higher response and remission rates compared to PBO.
  • Limitations: The post hoc/subgroup analysis from a controlled monotherapy trial included restrictive eligibility criteria, limiting generalizability, particularly to those with multiple comorbidities; multiple subgroup comparisons were conducted without reported multiplicity adjustment.
Maia et al (2025)2 conducted an analysis of the randomized, open‑label, rater‑blinded phase 3b ESCAPE‑TRD16 trial including 676 adults with TRD randomized to SPRAVATO+SSRI/SNRI (n=336) or QUE-XR+SSRI/SNRI (n=340).
The analysis evaluated associations between baseline patient characteristics and time to remission, defined as MADRS ≤10, using Cox proportional hazards models. Ratios of HRs demonstrated association between
baseline characteristics and relative benefit
of SPRAVATO+SSRI/SNRI vs QUE-XR+SSRI/SNRI on achievement of remission.
  • Remission occurred in significantly more patients who received adjunctive SPRAVATO compared to those in the adjunctive QUE-XR group (overall HR 1.71; 95% CI: 1.40-2.09).
  • Relative to QUE XR, remission in the SPRAVATO arm was:
    • 47% higher in overweight/obese versus underweight/normal-weight patients (ratio of HRs 1.47; 95% CI: 0.95-2.27).
    • 32% higher in patients with ≥3 vs 2 previous treatment failures (ratio of HRs 1.32; 95% CI: 0.87-2.00).
  • Age at diagnosis, sex, baseline MADRS total score (>34 or ≤34), episode duration, and number of episodes showed little to no association with remission of SPRAVATO vs QUE-XR.
  • Limitations: The exploratory subgroup analyses reported associations rather than causal predictors, thereby limiting generalizability; adjustment for multiplicity across multiple baseline subgroup comparisons was not reported.
Turkoz et al (2023)3 reported a post hoc pooled analysis of two randomized, double‑blind, active‑controlled, phase 3 short-term trials (TRANSFORM‑117 and TRANSFORM‑218), which included 518 adults with TRD treated with SPRAVATO+oral AD (n=310) or PBO NS+oral AD (n=208) for 4 weeks.
Predictors of response and remission were evaluated using univariate and multivariate logistic regression models. Response was defined as ≥50% MADRS reduction, and remission as MADRS ≤12 at day 28. Only the stepwise multivariate results are reported here.
  • Adjunctive SPRAVATO treatment was associated with higher odds of response (OR 1.68; 95% CI: 1.14-2.47) and remission (OR 1.55; 95% CI: 1.04-2.32) versus PBO.

Regardless of treatment group:
  • Younger age (18-44 vs 45-64 years) was associated with higher likelihood of response (OR 1.56; 95% CI: 1.05-2.33) and remission (OR 1.62; 95% CI: 1.07-2.45).
  • Employment was associated with higher odds of response (OR 1.78; 95% CI: 1.17-2.70) and remission (OR 2.11; 95% CI: 1.35-3.30).
  • Fewer prior antidepressant failures (1-2 vs ≥3) were associated with response (OR 1.68; 95% CI: 1.14-2.49) and remission (OR 1.65; 95% CI: 1.09-2.49).
  • Remission was more likely in patients with GAD7 <10 compared to those with clinically noteworthy anxiety symptoms (GAD7 ≥10) (OR 1.66; 95% CI: 1.05–2.62).

Within the SPRAVATO+oral AD group:
  • Employment (vs unemployment; OR 2.51; 95% CI: 1.44-4.36), a lower number of major depressive episodes including the current episode (2-5 vs >10; OR 4.33; 95% CI: 1.33-14.06), a lower baseline GAD-7 total score (<10 vs ≥10; OR 2.40; 95% CI: 1.25-4.61) were factors associated with higher odds of response.
  • Younger age (18-44 vs 45-64; OR 3.08; 95% CI: 1.59-5.96), employment (vs unemployment; OR 2.58; 95% CI: 1.43-4.67), and a lower baseline GAD-7 total score (<10 vs ≥10; OR 2.81; 95% CI: 1.51-5.21) were factors associated with higher odds of remission.
  • Limitations: The post hoc pooled analysis may have limited generalizability (eg, restrictive inclusion/exclusion criteria), combined data from studies with differing designs/doses, and did not include multiplicity adjustment.
Borentain et al (2021)4 reported a post hoc analysis of pooled data from two randomized, double-blind Phase 3 trials (TRANSFORM-117  and TRANSFORM-218) evaluating SPRAVATO+oral AD vs PBO NS+oral AD in adults with TRD.

The analysis assessed whether baseline sleep disturbance, measured using MADRS item 4, was associated with antidepressant response or remission over a 4-week treatment period. Associations were assessed using multiple logistic regression models.
Response was defined as a ≥50% reduction in MADRS total score from baseline, while remission was defined as a MADRS total score ≤12.
  • At baseline, 65.3% (369/565) had moderate-to-severe sleep disturbance (item 4 score, 4-6).
  • Across treatment groups, a 1‑point increase in baseline MADRS item 4 score, which indicates worsening sleep disturbance, was associated with a 23% higher probability of response (OR 1.23; 95% CI: 1.08-1.40; P=0.001) and 5% greater probability of remission (OR 1.05, 95% CI: 0.93-1.20; P=0.437) at day 28.
  • Regardless of presence/severity of sleep disturbance, a higher proportion of patients treated with SPRAVATO+oral AD achieved response (OR 2.05; 95% CI: 1.40-3.02; P<0.001) and remission (OR 1.81; 95% CI: 1.23-2.66; P=0.003) compared to those treated with PBO+oral AD.
  • Limitations: The post hoc pooled subgroup analysis was exploratory in nature; multiplicity adjustment was not reported; sleep disturbance was assessed using only MADRS item 4; the findings may have limited generalizability due to study exclusion criteria.
Daly et al (2021)5 conducted a post hoc analysis of the randomized, double‑blind, PBO‑controlled phase 3 TRANSFORM‑2 trial,18 which evaluated SPRAVATO+oral AD vs PBO NS+oral AD over 4 weeks in adults with TRD.
Associations between baseline comorbid anxiety (diagnosed by Mini International Neuropsychiatry Inventory assessment or GAD-7 score ≥10) and antidepressant outcomes were assessed using multiple logistic regression models.
Response was defined as ≥50% reduction in MADRS total score, and remission as MADRS ≤12, both at day 28.
  • At baseline, 72.6% (162/223) of patients met criteria for clinically noteworthy comorbid anxiety and had higher mean MADRS scores and longer episode duration than those without anxiety.
  • At day 28, mean MADRS scores decreased significantly from baseline in both anxiety subgroups treated with SPRAVATO+oral AD [mean change from baseline (SD): with anxiety,  -21.0 (12.51), P<0.001; without anxiety, -22.7 (11.98), P<0.001].
  • Response rates at day 28 were numerically higher with SPRAVATO+AD than AD+PBO in patients with anxiety (65.3% vs 54.2%; OR 1.59, 95% CI: 0.81-3.11) and without anxiety (79.3% vs 46.4%; OR 4.42, 95% CI: 1.38-14.19).
  • Remission rates at day 28 were also numerically higher with SPRAVATO+AD in patients with anxiety (47.2% vs 33.3%; OR 1.79, 95% CI: 0.91-3.51) and without anxiety (65.5% vs 25.0%; OR 5.70, 95% CI: 1.81-17.97).
  • No significant treatment‑by‑anxiety interaction was observed for response or remission, indicating baseline comorbid anxiety did not moderate treatment outcomes.
  • Limitations: The post hoc subgroup analysis did not use pre-specified criteria of anxiety, the study was not designed to evaluate comorbid anxiety, and potential confounding factors (eg, concomitant medications) were not fully accounted for; multiplicity adjustment was not reported.
RETROSPECTIVE STUDIES
Elmaadawi et al (2025)6 conducted a chart review including 200 adults with TRD treated with IV KET (n=98) or SPRAVATO (n=102), evaluating baseline demographic, clinical, and comorbidity factors associated with antidepressant outcomes.
Primary outcomes were MADRS and PHQ-9 total scores over 4 weeks of treatment. General linear models were used to assess predictors of depression score change.
  • Mean MADRS scores decreased for both IV KET    (-10.8; P<0.001) and SPRAVATO (-6.2; P<0.001).
  • Mean PHQ-9 scores decreased for both IV KET    (-5.0; P<0.001) and SPRAVATO (-2.8; P<0.001).
  • Response (≥50% improvement from baseline) and remission (≤10) rates based on the MADRS were similar between the two groups.
  • Patients with no psychiatric comorbidity were associated with the largest improvement from baseline in depressive symptoms (mean MADRS change: -22.0; 95% CI: -30 to -14).
  • Patients with borderline personality disorder were associated with minimal symptom improvement (mean MADRS change: -3.7; 95% CI: -10 to 3).
  • Patients with comorbid anxiety and PTSD were associated with slightly greater improvement     (mean MADRS change: -7.3 and -9.0, respectively); patients with multiple comorbidities had a mean MADRS improvement of -11.4.
  • Patients with increasing number of failed prior antidepressant trials were strongly associated with reduced improvement in MADRS (approximately -1.3 points per additional prior medication, P<0.0001) and PHQ-9 (P<0.001) following IV KET or SPRAVATO treatment.
    • In the SPRAVATO group, those with a very high number of prior treatment failures (eg, >5) resulted in minimal MADRS improvement compared to those with 2-3 previous failures for example.
  • Treatment modality was a significant predictor on the MADRS but not on the PHQ-9. Model-adjusted mean MADRS change was -11.98 for IV KET and -6.80 for SPRAVATO (P=0.03).
  • Age was not associated with treatment response, while sex differences were observed only in clinicianrated outcomes, with males showing slightly greater MADRS improvement (MADRS reduction: -7.9 in women vs -10.3 in men; P=0.03).
  • Limitations: The retrospective realworld design may have introduced selection bias and unmeasured confounders, and the simplified comorbidity categorization may not have captured clinical heterogeneity.
Clemens et al (2025)7 conducted a cohort study to evaluate factors associated with remission in adults with TRD (without psychotic features) initiating SPRAVATO.
Adults with TRD initiating SPRAVATO, with baseline PHQ-9 scores ≥10 and ≥6 months of follow-up, were included. Remission was defined as a PHQ-9 score <5 during follow-up.
Claims between January 2016 to June 2023 were sourced from the Komodo Research Database and PHQ-9 scores from the Komodo Clinical Observations database.
Multivariable Cox proportional hazards model was used to evaluate factors associated with remission. Patients who did not achieve remission during the follow-up period were censored at the end of follow-up.
  • Among 184 patients (mean baseline PHQ‑9 score, 18.4; mean age, 43.1), age, sex, race/ethnicity, and baseline comorbidities were not associated with remission.
  • Medicaid insurance (vs commercial) was associated with a 69% lower likelihood of achieving remission (HR 0.31; 95% CI: 0.11-0.88; P=0.028).
  • Each additional non‑antidepressant augmentation agent at baseline was associated with a 27% lower chance of remission (HR 0.73; 95% CI: 0.56-0.94; P=0.016).
    • Agents included anticonvulsants, non-benzodiazepine anxiolytics, psychostimulants, second-generation antipsychotics and thyroid hormones, as well as lithium and atomoxetine.
  • Limitations: The retrospective claims‑based design may have introduced misclassification bias (eg, medication adherence not confirmed); included missing race data; patient‑reported outcomes could have been susceptible to recall bias; results may have been subjected to residual confounding; multiplicity adjustment was not reported.
Ansari et al (2025)8 conducted a single‑center real‑world study of 190 adults with TRD treated with SPRAVATO (8 treatments each ≤7 days apart), evaluating whether baseline BMI was associated with clinical response or remission using generalized linear mixed‑effects models.
Response was defined as ≥50% improvement in MADRS and QIDS-SR-16, and remission was defined as MADRS ≤12 and QIDS-SR-16 ≤10.
Response and Remission Based on MADRS
  • Obese patients (BMI ≥30 kg/m²) were more likely to achieve response than those considered non-obese (BMI <30 kg/m²) (adjusted RR 1.63; 95% CI: 1.04-2.54; P=0.033). There was also a numerical (but not significant) trend for remission.
  • Response occurred in 36.4% of obese patients vs 26.1% underweight/normal and 18.2% overweight patients.
  • Remission rates were numerically higher in obese patients (22.7%) than in underweight/normal (15.9%) and overweight patients (9.1%).
  • There was no significant linear association between BMI as a continuous variable and response (RR, 1.02; P=0.163) and remission (RR, 1.01; P=0.634) nor was there a significant association between weight and response.
  • With increasing age, patients were less likely to respond to SPRAVATO when controlled for BMI and gender; however, there was no relationship between age and remission.

Response and Remission Based on QID-SR-16
  • No statistical significance was found between QID-SR-16 and response or remission.
  • No significant association was found between weight and response.
  • Limitations: The retrospective singlecohort design, lack of data on dropouts from treatment, and absence of detailed data on comorbidities and concurrent medications may have limited robustness; the relatively smaller sample size that included predominantly White patients may have limited generalizability to other populations; multiplicity adjustment for multiple analyses was not reported.
Molero et al9 conducted an observational, multicenter cross-sectional study (INTEGRATE) including 189 adults with TRD treated with SPRAVATO+oral AD for ≥4 weeks.

Treatment response was evaluated at 4 weeks based on clinical judgment.
Logistic regression analyses were used to explore baseline demographic, clinical, and treatment-related characteristics associated with response.
  • Hospitalization during the current depressive episode was associated with a reduced likelihood of response at 4 weeks (OR 0.52; 95% CI: 0.27-0.99; P=0.049).
  • Each additional hospitalization during the current episode reduced the probability of response by 33% (OR 0.67; 95% CI: 0.48-0.94; P=0.019).
  • Patients who had been hospitalized during the current depressive episode needed more time to achieve the first response (11.3 days vs 8.3 days, P=0.019).
  • A higher number of concomitant psychotropic medications at baseline was associated with reduced probability of response (OR 0.76; 95% CI: 0.58-0.99; P=0.044).
  • Benzodiazepine use was associated with reduced likelihood of response (OR 0.33; 95% CI: 0.13-0.85; P=0.021), although response rates remained high in both users (82.3%) and non‑users (92.9%).
  • Psychiatric comorbidity showed a marginally lower probability of response (OR 0.55; 95% CI: 0.28-1.07; P=0.08) but was not statistically significant.
  • Most baseline demographic and clinical characteristics (including age, sex, age at first depressive episode, previous depressive episodes, duration of current depressive episode, suicide attempts during the current episode, employment status, medical comorbidities, substance use, psychotic symptoms in the current episode, and family history of mood disorder) were not significantly associated with response at 4 weeks.
  • Limitations: The retrospective real‑world exploratory analysis in a chronic/complex population required confirmation in larger samples, and multiple baseline variables were evaluated without reported multiplicity adjustment.
Patarroyo-Rodriguez et al (2024)10 conducted a secondary analysis of a retrospective cohort study,19 including 52 adults with TRD treated with IV KET (n=37) over 6 treatments or SPRAVATO (n=15) over 8 treatments, to explore the differences in response rates among different sleep phenotypes.
Baseline sleep phenotypes were defined using the Quick Inventory of Depressive Symptomatology Self Report (QIDS-SR) based on a score >1 on the sleep domains of early insomnia, middle insomnia, late insomnia, and hypersomnia.
Associations between insomnia or hypersomnia and response, number of treatments to response, and % change in total QIDS-SR were evaluated using univariate and multivariate logistic regression, Cox proportional hazard models, and linear regression, adjusting for age, BMI, treatment, sex, and baseline modified total QIDS-SR (total score without sleep domain) as covariates. Pearson’s correlation analyses were conducted between change in sleep scores with changes in modified total QIDS-SR.
  • Compared to those without hypersomnia, patients with baseline hypersomnia had:
    • Higher response rates (70.0% vs 37.5%; P=0.04) based on total QIDS-SR compared with patients without hypersomnia.
    • Required fewer treatments to achieve response (HR 2.4; 95% CI: 1.09-5.30; P=0.03).
    • Greater reduction in depressive symptoms (22.7 percentage points greater reduction in QIDS‑SR; P=0.008).
    • After adjusting for covariates, baseline hypersomnia was still associated with response and a greater reduction in percent change in total QIDS-SR (P≤0.02).
    • No significant relationships were found for other sleep phenotypes (P≥0.12).
  • Correlation was strongest between those with middle insomnia and changes in modified total QIDS-SR score (r=0.54; P<0.001), followed by hypersomnia (r=0.50; P<0.001) and late insomnia (r=0.40; P<0.40; P=0.003). There was low correlation with early insomnia (P=0.12).  
  • Subgroup with neurovegetative symptoms of atypical depression (NVSAD, defined as >1 on hypersomnia and >0 on increased appetite item on the QIDS-SR); 15.4% of the cohort showed higher likelihood of response (87.5% vs 43.2%; P=0.05) and greater QIDS-SR reduction (24.3 points; P=0.04) compared to those without NVSAD.
  • Limitations: The observational design without a PBO control, small sample size, combined IV ketamine and SPRAVATO data, and potential overlap of sleep phenotypes may have limited interpretation; multiplicity adjustment was not reported.
Miller et al (2024)11 conducted an exploratory secondary analysis of a single-centered, retrospective real-world study of 120 veterans with TRD treated with SPRAVATO (n=99) or IV KET (n=20) or IM KET (n=1) across repeated treatment sessions. Most completed 8 treatment sessions.  
Baseline depressive symptoms were assessed using the PHQ-9, and treatment response was modeled using symptom-level trajectories and machine-learning classifiers. The goal was to build predictive models that could confidently predict response and non-response (treatment failure) in a subset of patients.
  • During repeated KET/esketamine sessions, PHQ‑9 total scores and all individual items improved with depressed mood improving faster than tiredness/low energy (P=0.001).
  • In predictive models, higher baseline anhedonia (item 1), depressed mood (item 2), and feeling of failure (item 6) were associated with response; the best baseline‑symptom model achieved ~60% classification accuracy.
  • Baseline sleep difficulties and suicidal ideation were consistently associated with non‑response, with sleep difficulty favoring non‑response in 100% of high‑confidence models and suicidal ideation in 98.7%.
    • Lower scores on remaining PHQ-9 items were less strongly associated with non-response.
  • Threshold‑optimized models, created to improve predictive confidence in these predictive models, identified 22% of patients as likely non‑responders with high confidence (negative predictive value 96.4%), while no models were detected to predict treatment response at a 90% predictive value.
  • Age, sex, and treatment route (intranasal vs IV KET) showed no significant association with response.
  • Limitations: The observational/realworld analysis evaluating multiple predictors may have been subject to confounding; it lacked a control group and detailed information on patients’ treatments, and the use of regression modeling with threshold tuning may have resulted in lower specificity; multiplicity adjustment was not reported.
Pettorruso et al12 analyzed data from a multicenter, retrospective, realworld study20 including 149 patients with TRD treated with intranasal esketamine, using machinelearning models to evaluate baseline clinical predictors of response (≥50% MADRS reduction) and remission (MADRS <10) at one month and three months.
Depression at baseline were measured by different scales, including MADRS, HAM-D-21, HAM-A, and BPRS.
  • Response rates increased from 30.2% at one month to 57.0% at three months, and 38.9% of patients achieved remission by three months.
  • The overall accuracy of the model predicting response at 1 month and 3 months was 68.5% (P<0.005) and 66.3% (P<0.005), respectively.
    • Early response (1 month): Inner tension (MADRS item 3) was the only positive predictor, while benzodiazepine use, self-neglect (BPRS item 13), emotional-withdrawal (BPRS item 17), retardation (HAM-D item 8), somatic concerns (BPRS item 1), motor retardation (BPRS item 18), hypocondrias (HAM-D item 15), suicidality (BPRS item 4), OCD symptoms (HAM-D item 21), somatic anxiety (HAM-A item 8), and paranoid symptoms (HAM-D item 20) were associated with a lower likelihood of response.
    • Later response (3 months): inner tension (MADRS item 3), pessimistic thoughts (MADRS item 9), anhedonia (MADRS item 8), reported sadness (MADRS item 2), concentration difficulties (HAM-D item 5), hyperthymic temperament, feelings of guilt (HAM-D item 2), fears (HAM-A item 3), and cardiovascular anxiety (HAM-A item 9) were associated with higher likelihood of response; previous rTMS failure was associated with lower likelihood of response.
    • There were variables, such as somatic anxiety, that demonstrated mixed patterns.
  • The accuracy of the model predicting remission at 3 months was 68.6% (P<0.005).
    • Higher likelihood of remission: Anhedonia (MADRS item 8), anxiety (BPRS item 2), comorbidity with bipolar disorder, motor tension (BPRS item 19), and emotional blunting (BPRS item 16)
    • Lower likelihood of remission: Higher baseline suicidality (BPRS item 4), obsessive compulsive symptoms (HAM-D item 21), greater number of prior depressive episodes, and previous rTMS failure.
    • Variables such as depressed mood (HAM-D item 1), pessimistic thoughts (MADRS item 9), feelings of guilt (BPRS item 5), psychic and somatic anxiety (HAM-D item 10&11), demonstrated mixed patterns in predicting remission
  • Limitations: The retrospective multicenter realworld design with heterogeneous assessments and attrition may have limited generalizability, and multiple baseline predictors were evaluated without reported multiplicity adjustment.
COHORT STUDY
Cavalcanti et al (2025)13 reported a historical observational cohort study at a single academic center to evaluate whether baseline perceived stress predicted remission following KET or SPRAVATO treatment in adults with TRD.
Baseline stress was measured using the Perceived Stress Scale (PSS), and depressive symptoms were assessed using QIDS-SR. Patients received either IV KET or SPRAVATO during the acute treatment phase. Remission was defined as QIDS-SR ≤5, and response was defined as an improvement of at least 50%.
PSS is a 10-item (0-40 points) self-rated tool that assesses areas such as lack of control, nervousness, stress, confidence, coping, irritability, anger, and feel overwhelmed. A score of ≥27 indicated high stress while a lower score indicated low-to-moderate stress.
  • Among 39 patients, 66.7% had high baseline perceived stress (PSS ≥27).
  • Each 5‑point increase in baseline PSS score was associated with a 60% reduction in odds of remission (OR 0.40; 95% CI 0.14-0.81; P=0.009), independent of baseline depression severity.
  • Among patients who achieved remission, those with high baseline perceived stress required more treatments to achieve remission compared with those with low-moderate stress (median 3 vs 1 treatments; P=0.04).
    • Each 5-pt increase in the baseline PSS score was associated with a 56% increase in the number of treatments needed for remission (IRR, 1.56; P=0.04).
  • Limitations: The singlecenter cohort design without a PBO control, continuation of concomitant psychotropic medications as they initiated ketamine, and small sample size limited statistical power and generalizability, and exploratory analyses may not have been adjusted for multiplicity.
Abbreviations: AD, antidepressant; AE, adverse event; BMI, body mass index; BPRS, Brief Psychiatric Rating Scale; CI, confidence interval; GAD-7, Generalized Anxiety Disorder 7-item scale; HAM, Hamilton Anxiety Rating Scale; HAM-D-21, 21item Hamilton Depression Rating Scale; HR, hazard ratio; i.m., intramuscular; i.v., intravenous; KET, ketamine; MADRS, Montgomery–Åsberg Depression Rating Scale; MDD, major depressive disorder; NS, nasal spray; NVSAD, neurovegetative symptoms of atypical depression; OR, odds ratio; PBO, placebo; PHQ-9, Patient Health Questionnaire-9; PSS, Perceived Stress Scale; QIDS-SR, Quick Inventory of Depressive Symptomatology–Self-Report; QUE-XR, quetiapine extended release; RR, risk ratio; rTMS, repetitive transcranial magnetic stimulation; SNRI, serotonin–norepinephrine reuptake inhibitor; SSRI, selective serotonin reuptake inhibitor; TRD, treatment-resistant depression.

Literature Search

A literature search of MEDLINE®, EMBASE®, BIOSIS Previews®, and DERWENT Drug File (and/or other resources, including internal/external databases) pertaining to this topic was conducted on 07 April 2026.

Studies with <30 patients were excluded along with review articles. This SRL included only studies with the primary objective of evaluating the association between baseline/clinical characteristics and the effect on efficacy/effectiveness, including response and remission.

 

References

Show
1 Himedan M, Lopena O, Lim L, et al. Efficacy of esketamine nasal spray monotherapy for treatment-resistant depression: post hoc analysis by baseline disease severity and prior treatment. Poster presented at: Neuroscience Education Institute (NEI) Spring Congress; May 1-3, 2026; Florida, USA.  
2 Oliveira-Maia A, Baune B, Bartova L, et al. Patient characteristics associated with relative benefit of esketamine nasal spray versus quetiapine extended release on achieving remission in ESCAPE-TRD study. Poster presented at: European College of Neuropsychopharmacology (ECNP) Congress; October 11-14, 2025; Amsterdam, Netherlands.  
3 Turkoz I, Nelson JC, Wilkinson ST, et al. Predictors of response and remission in patients with treatment-resistant depression: A post hoc pooled analysis of two acute trials of esketamine nasal spray. Psychiatry Res. 2023;323(suppl 37019044):115165.  
4 Borentain S, Williamson D, Turkoz I, et al. Effect of sleep disturbance on efficacy of esketamine in treatment-resistant depression: findings from randomized controlled trials. Neuropsychiatr Dis Treat. 2021;17:3459-3470.  
5 Daly EJ, Turkoz I, Salvadore G, et al. The effect of esketamine in patients with treatment-resistant depression with and without comorbid anxiety symptoms or disorder. Depress Anxiety. 2021;38(11):1120-1130.  
6 Elmaadawi AZ, Naha I, Prabhudesai S, et al. Personalizing ketamine therapy: Real-world predictors of response to IV ketamine and intranasal esketamine in treatment-resistant depression. Psychiatry Res. 2025;354:116821.  
7 Clemens K, Zhdanava M, Voegel A, et al. Factors associated with reaching remission on esketamine among patients with treatment-resistant depression in the United States. Poster presented at: Anxiety and Depression Association of America (ADAA) Annual Conference; April 3-5, 2025; Las Vegas, Nevada, USA.  
8 Ansari M, Rhee TG, Santucci MC, et al. Does BMI matter when treating depression with esketamine? A retrospective analysis of real-world data. J Affect Disord. 2025;381:22-28.  
9 Molero P, Ibañez A, de Diego-Adeliño J, et al. Variables associated with antidepressant response to esketamine nasal spray treatment in clinical practice: INTEGRATE study. Poster presented at: 37th European College of Neuropsychopharmacology (ECNP) Congress; September 21-24, 2024; Milan, Italy.  
10 Patarroyo-Rodriguez L, Pazdernik V, Vande Voort JL, et al. Hypersomnia as a predictor of response to intravenous ketamine/intranasal esketamine in treatment resistant depression. J Affect Disord. 2024;349:272-276.  
11 Miller EA, Afshar HT, Mishra J, et al. Predicting non-response to ketamine for depression: An exploratory symptom-level analysis of real-world data among military veterans. Psychiatry Res. 2024;335:115858.  
12 Pettorruso M, Guidotti R, d’Andrea G, et al. Predicting outcome with Intranasal Esketamine treatment: A machine-learning, three-month study in Treatment-Resistant Depression (ESK-LEARNING). Psychiatry Res. 2023;327:115378.  
13 Cavalcanti S, Pazdernik VK, Vande Voort JL, et al. Baseline perceived stress as a predictor of ketamine/esketamine treatment response in treatment-resistant depression. J Affect Disord. 2025;383:283-289.  
14 Lima Constantino J, Godschalk M, van Dalfsen JH, et al. Demographic and clinical predictors of response and remission in the treatment of major depressive disorder with ketamine and esketamine: A systematic review. Psychiatry Res. 2025;345:116355.  
15 Janik A, Qiu X, Lane R, et al. Esketamine monotherapy in adults with treatment-resistant depression a randomized clinical trial. JAMA Psychiatry. 2025;82 (9):877-887.  
16 Reif A, Bitter I, Buyze J, et al. Esketamine nasal spray versus quetiapine for treatment-resistant depression. N Engl J Med. 2023;389(14):1298-1309.  
17 Fedgchin M, Trivedi M, Daly EJ, et al. Efficacy and safety of fixed-dose esketamine nasal spray combined with a new oral antidepressant in treatment-resistant depression: results of a randomized, double-blind, active-controlled study (TRANSFORM-1). Int J Neuropsychopharmacol. 2019;22(10):616-630.  
18 Popova V, Daly EJ, Trivedi M, et al. Efficacy and safety of flexibly dosed esketamine nasal spray combined with a newly initiated oral antidepressant in treatment-resistant depression: a randomized double-blind active-controlled study. Am J Psychiatry. 2019;176(6):428-438.  
19 Singh B, Kung S, Pazdernik V, et al. Comparative effectiveness of intravenous ketamine and intranasal esketamine in clinical practice among patients with treatment-refractory depression: an observational study. J Clin Psychiatry. 2023;84(2):22m14548.  
20 Martinotti G, Dell’Osso B, Lorenzo GD, et al. Treating bipolar depression with esketamine: safety and effectiveness data from a naturalistic multicentric study on esketamine in bipolar versus unipolar treatment-resistant depression. Bipolar Disord. 2023;25(3):233-244.  

Would you like to clear and leave your conversation? Message history will be lost.