In the phase 3 SUSTAIN-3 long-term extension (1,148 patients, 3,777 patient‑years), intermittent intranasal esketamine plus an oral antidepressant showed no new safety signals and an adverse event profile consistent with earlier studies. Depression severity improved during induction and was generally maintained during optimisation/maintenance, with remission in 35.6% at induction and about 48–50% at week 112/maintenance endpoint.
Papers cited by this study that are also in Blossom
Importance
The rates of relapse and suicide risk are higher in treatment-resistant depression (TRD) vs non-treatment-resistant major depressive disorder. Even among patients with TRD who initially respond, the majority (70%) relapse within 6 months.
Objective
To evaluate the long-term safety and efficacy of esketamine nasal spray, combined with an oral antidepressant, in patients with TRD.
Design
Phase 3, open-label, single-arm long-term extension study (SUSTAIN-3) conducted from June 2016 to December 2022.
Setting
Outpatient.
Participants
Adults with TRD who participated in ≥1 of 6 phase 3 “parent” studies continued esketamine by either entering a 4-week induction phase followed by an optimization/maintenance phase of variable duration (n = 458) or directly entering the optimization/maintenance phase of SUSTAIN-3 (n = 690), based on their individual response to study drug at the endpoint of the parent study.
Interventions
Intranasal esketamine dosing was flexible, twice-weekly during induction and individualized to depression severity during optimization/maintenance (weekly, every-other-week, or every-4-weeks), under direct supervision by site staff.
Main Outcomes and Measures
To assess the long-term safety of esketamine. Efficacy endpoints included the change in depressive symptoms, assessed by the Montgomery-Åsberg Depression Rating Scale (MADRS).
Results
A total of 1148 patients were enrolled. Total exposure to esketamine was 3777 cumulative patient-years. Mean (median, range) exposure to esketamine in SUSTAIN-3 was 42.9 (45.8, range 0-79) months. The most common adverse events were headache (36.9%), dizziness (33.9%), nausea (33.6%), dissociation (25.5%), nasopharyngitis (23.8%), somnolence (23.1%), dysgeusia (20.2%), and back pain (20.0%). During the study, 5.3% and 6.4% of participants discontinued due to lack of efficacy or adverse event, respectively. Nine participants died: COVID-19-related (n = 3), pneumonia (n = 2), and completed suicide, myocardial infarction, multiple injuries, unknown cause (n = 1 each). The mean MADRS total score decreased during induction, and this reduction persisted during optimization/maintenance (mean [SD] change from baseline-to-phase endpoint of each phase: induction: −12.8 [9.73]; optimization/maintenance: + 0.2 [9.93]). A total of 35.6% of participants were in remission at the induction endpoint, and 48.5% and 49.6% at week 112 and optimization/maintenance endpoint, respectively.
Conclusions and Relevance
In the SUSTAIN-3 final dataset, no new safety signals were identified during long-term treatment with intermittently-dosed esketamine, combined with oral antidepressant, and improvement in depression generally persisted among participants who remained on maintenance treatment. These results add to the accumulated evidence on TRD treatment with esketamine.
Treatment-resistant depression (TRD) is associated with markedly elevated rates of relapse, disability, and suicide risk relative to non-treatment-resistant major depressive disorder. Even among patients with TRD who achieve an initial response to standard antidepressants, approximately 70% relapse within six months, and longer periods of inadequate treatment are associated with progressively worsening prognosis. The unmet need for treatments providing sustained efficacy in TRD has long been recognised. Esketamine nasal spray — the S-enantiomer of ketamine — demonstrated rapid antidepressant effects in short-term trials and received regulatory approval for TRD, but long-term safety and maintenance efficacy data were required to inform sustained clinical use. The SUSTAIN-3 study aimed to evaluate the long-term safety and efficacy of esketamine nasal spray combined with an oral antidepressant in adults with TRD over an extended treatment period, encompassing both an induction phase and a prolonged optimisation and maintenance phase.
SUSTAIN-3 (NCT02782104) was a phase 3, open-label, single-arm, multicentre, long-term extension study conducted between June 2016 and December 2022 across 222 sites in 27 countries. Participants were adults with TRD who enrolled from six phase 3 parent studies. The study consisted of a 4-week induction phase (where applicable) followed by an optimisation and maintenance phase of variable duration; esketamine was administered via nasal spray at in-person clinic visits. Safety assessments included adverse events, vital signs, clinical laboratory measures, and suicidality monitoring. Efficacy was assessed using the clinician-rated Montgomery-Asberg Depression Rating Scale (MADRS) and the Sheehan Disability Scale (SDS). A total of 1,148 adult patients were enrolled.
The mean duration of esketamine treatment in SUSTAIN-3 was 42.9 months (median 45.8, range 0–79 months), with nearly two-thirds of participants remaining in the study for three years or longer. The most common adverse events occurring in 20% or more of participants were headache, dizziness, nausea, dissociation, nasopharyngitis, somnolence, dysgeusia, and back pain. The vast majority of adverse events on dosing days (97%) resolved on the same day. Serious adverse events were reported in 18.8% of participants; those related to depressive symptoms were most frequent. Depressive symptoms assessed by MADRS decreased substantially during the induction phase and were maintained through optimisation and maintenance. Clinical response rates were 55.9% at the end of the induction phase and 60.9% at the end of the optimisation and maintenance phase. Functional outcomes assessed by the SDS also improved and were maintained throughout the study period.
Given the high relapse rate reported in previous studies of patients with treatment-resistant depression (TRD), there is clinical interest in longer-term safety and efficacy data with esketamine. In SUSTAIN-3, an open-label single-arm study, 1,148 adults with TRD received intermittently-dosed esketamine nasal spray plus oral antidepressant for up to 6.5 years (3,777 cumulative patient-years). No new safety concerns were identified versus prior short-and long-term (~1 year) studies. Most adverse events were transient, occurring and resolving on a dosing day. Depressive symptoms improved during the 4-week induction phase; improvement generally persisted during the long-term optimization/maintenance phase ( Few participants discontinued across phases due to lack of efficacy (5.3%) or adverse event (6.4%). Within the context of an open-label, single-arm design, the long-term safety data and durability of treatment effect observed in SUSTAIN-3 add substantially to the body-of-evidence supporting the treatment of TRD using esketamine nasal spray. Downloaded fromby guest on 07 May 2025
Treatment-resistant depression (TRD) is associated with a higher rate of relapse and higher risk of suicide compared to non-treatment-resistant major depressive disorder (MDD).Even among patients with TRD who initially respond to standard antidepressants, the majority (~70 %) relapse within 6 months.The longer patients remain unsuccessfully treated, the worse their prognosis.The unmet need for new TRD treatment options with sustained efficacy and safety has been noted for decades. In this regard, esketamine nasal spray, in conjunction with an oral antidepressant, was approved for TRD by the United States Food and Drug Administrationand the European Medicines Agency,followed by approvals from >75 health authorities worldwide. These approvals were based on phase 2/3 short-term studies of TRD patientsas well as a relapse prevention studyand an openlabel, 1-year safety study.The SUSTAIN-3 trial (NCT02782104) was conducted to assess longer-term safety (including cognitive, hepatic, and renal safety) and sustained efficacy of esketamine nasal spray, combined with an oral antidepressant, in patients with TRD. Reported herein are the final results of SUSTAIN-3, which support and extend the findings from the previously published interim data analyses.In that manuscript we presented evidence for safety and durability of treatment effect across a total of 2769 cumulative patient-years; the current manuscript presents data from an additional 1008 patient-years of treatment with esketamine nasal spray.
Create a free account to open full-text PDFs.
Martinotti, G., Vita, A., Fagiolini, A. et al. · Journal of Affective Disorders (2022)
Papers in Blossom that reference this study
Mallevays, M., Fuet, L., Danon, M. et al. · MedRvix (2026)
Samalin, L., Rothärmel, M., Mekaoui, L. et al. · Journal of Psychiatric Research (2026)
Reif, A., Anıl, Y. A., Bitter, I. et al. · European Neuropsychopharmacology (2026)
SUSTAIN-3 provides the most extensive long-term safety and efficacy dataset yet generated for any rapidly acting antidepressant in TRD, with mean treatment exposure approaching four years and a subset of patients treated for over six years. The safety profile observed over this extended period was broadly consistent with that established in shorter-term trials, and the persistence of functional improvement through the optimisation and maintenance phase supports the durability of esketamine's antidepressant effects with continued intermittent administration. The study was conducted partly during the peak of the COVID-19 pandemic, a period during which in-person clinic visits were required for all esketamine dosing — a logistical challenge that may have influenced adherence and retention. The open-label design without a placebo arm limits interpretation of efficacy data, though the scale and duration of the dataset substantially advance the evidence base for long-term esketamine use in clinical practice.
SUSTAIN-3 demonstrates that esketamine nasal spray combined with an oral antidepressant can be administered safely and with maintained antidepressant efficacy over treatment durations of up to 79 months in adults with TRD. The adverse event profile remained consistent with that observed in short-term trials, with most dosing-day events resolving the same day. Response rates were sustained through the maintenance phase, supporting esketamine as an option for long-term management of treatment-resistant depression.
The study protocol and its amendments were approved by Institutional Review Board (United States) or Independent Ethics Committee (all other locations) at each site (Methods S1, online Supplementary Material). The study was conducted in accordance with ethical principles of the Helsinki Declaration. All individuals provided written informed consent before entering the study.
SUSTAIN-3 (NCT02782104), a phase 3, open-label, single-arm, multicenter, long-term extension study, was conducted between June 2016 and December 2022 at 222 sites in 27 countries/locations. The study consisted of two phases: a 4-week induction phase (if applicable) and an optimization/maintenance phase of variable duration (these phases are defined under the Study Drug section below).
Participants in 6 phase 3 "parent" studies of esketamine enrolled into the 4-week induction phase or the long-term optimization/maintenance phase of SUSTAIN-3 based on their clinical status at parent study end. Participants who did not meet the criteria for response at parent study end enrolled into the induction phase of SUSTAIN-3, and responders at the end of induction progressed to the optimization/maintenance phase. Participants who met the criteria for response at parent study end, or
Adverse events and other safety assessments (i.e., hematology, chemistry, urinalysis, vital signs, electrocardiogram) were monitored throughout the study. Adverse events of special interest included events such as blood pressure elevation and respiratory depression, which were monitored during each dosing visit using vital sign measures that included pulse oximetry; liver enzyme elevation via laboratory testing; and events related to renal disorder (cystitis, dysuria, pollakiuria, nephrolithiasis, micturition urgency, urinary incontinence, hematuria). Urine was screened every 12 weeks for illicit drugs (i.e., barbiturates, methadone, opiates, cocaine, phencyclidine, amphetamine/methamphetamine). The Columbia-Suicide Severity Rating Scale 16 (C-SSRS) was used to evaluate potential suicidal ideation and behavior, the Modified Observer's Assessment of Alertness/Sedation (MOAA/S) scale to evaluate post-dose sedation, and the Clinical Global Assessment of Discharge Readiness (CGADR) to evaluate participants' discharge readiness from study sites following dosing, based on their overall clinical status. Cognition was assessed using the Cogstate computerized test batteryand the Hopkins Verbal Learning Test-Revised(HVLT-R). Descriptions of these safety measures appear in the online Supplementary Material (Methods S2).
The clinician-rated efficacy measures used to rate the severity of depressive symptoms and depressive illness, respectively, were the Montgomery-Åsberg Depression Rating Scale(MADRS) and the CGI-S.Patient-rated outcome measures used for participants to rate their depressive symptoms and sociooccupational disability, respectively, were the Patient Health Questionnaire 9-item
The number (percentage) of participants with adverse events, serious adverse events, and adverse events leading to discontinuation of study drug were summarized by standardized Medical Dictionary for Regulatory Activities (MedDRA) preferred term, version 25.1. Study drug exposure was reported as cumulative person-years. The number of patients who attempted suicide, the number who died by suicide, and the number who died by any cause were each reported per 100 patient-years (exposure-adjusted rate), a standardized measure of risk, allowing comparison of the SUSTAIN-3 data to data from studies of other TRD treatments. Analyses of cognitive data were conducted by age group (i.e., <65 years old and ≥ 65 years old). Descriptive statistics, including means and mean change from baseline scores, were calculated A c c e p t e d M a n u s c r i p t for cognition tests; in addition, the difference in each participant's performance on any cognitive measure from baseline of the optimization/maintenance phase was characterized based on the Reliable Change Index (RCI), with an absolute value of RCI ≥ 1.96 (i.e., -1.96 indicates test performance was worse than at baseline and 1.96, better than at baseline) considered to be a meaningful change from the baseline score for a test. Cohen's d was calculated to further evaluate the magnitude of any cognitive changes. Descriptive statistics (mean and SD or SE) were provided for other safety parameters. Efficacy data were summarized by descriptive statistics. Efficacy endpoints include: change from baseline in depressive symptoms (MADRS; PHQ-9); proportion of participants who achieved response (defined as ≥50% improvement from baseline) and remission (MADRS score ≤12; PHQ-9 score <5); overall severity of illness (CGI-S); and change from baseline in functioning and associated disability (SDS score); proportion of participants who achieved response (SDS scores ≤4 for each item, ≤12 for the total score) and remission (SDS ≤2 for each item score, ≤6 for the total score).In efficacy analyses for the induction phase, baseline was the last observation prior to or on the start date of induction phase for participants who entered SUSTAIN-3 at the induction phase. In efficacy analyses for the optimization/maintenance phase, baseline was the last observation prior to or on the start date of the optimization/maintenance phase.
A total of 1,148 adult patients with TRD were enrolled into SUSTAIN-3 (Figure). The number of participants enrolled into SUSTAIN-3 is summarized by parent study and entry point in Supplementary Table. Participants' mean (SD) age was 49.6 (12.28) years at baseline and two-thirds were female. Demographics at baseline are shown in Table.
Mean exposure to esketamine in SUSTAIN-3 was 42.9 months (median 45.8, range 0-79 months), with 728 (63.4%) treated for ≥36 months, 496 (43.2%) for ≥48 months, and 322 (28.0%) for ≥60 months (Supplementary Figure); total exposure was 3,777 cumulative patient-years. The mean (range) cumulative duration of intermittent esketamine treatment during the parent and SUSTAIN-3 studies combined was 48.3 (0-88) months, with 991 (86.3%) participants treated for ≥12 months, 867 (75.5%) for ≥24 months, 777 (67.7%) for ≥36 months, 582 (50.7%) for ≥48 months, 432 (37.6%) for ≥60 months, and 111 (9.7%) for ≥72 months; total exposure was 4,247 cumulative patient-years. During maintenance, most participants received 84 mg (64.4%) as their final esketamine dose, while 33.0% received 56 mg and 2.5% received 28 mg. The most frequent dosing interval was weekly (~40-50% of participants for most of their dosing); every-other-week was ~35-45%, and every-4-week dosing was ≤~25% (Supplementary Figure). The most common concomitant oral antidepressants taken by participants in the optimization/maintenance phase were duloxetine (36.7%), venlafaxine (27.7%), escitalopram (26.2%), sertraline (18.6%), and bupropion (11.2%) (Supplementary Table). The most common antipsychotics used in the phase were quetiapine (7.4%) and aripiprazole (3.8%), and the most common mood stabilizer was lithium (2.3%).
The most common adverse events (≥20%) were headache, dizziness, nausea, dissociation, nasopharyngitis, somnolence, dysgeusia, and back pain (Table). For adverse events occurring on an esketamine dosing day, most events (76769/79118 [97%]) resolved the same day. Serious adverse events were reported for 18.8% (216/1148) of participants (Supplementary Table). Serious adverse events related to depression (MedDRA terms: depression, major depression, persistent depressive disorder, affect liability) occurred in 23 (2.0%) participants. Serious adverse events related to suicidality occurred in 28 (2.4%) participants, including suicide attempt (15, 1.3%), suicidal ideation (11, 1.0%), depression suicidal (1, 0.1%), and completed suicide (1, 0.1%). Of all serious adverse events, investigators considered one event (blood pressure [BP] diastolic increased) to be very likely related, 4 events (arrhythmia, spontaneous abortion, hypertensive emergency, loss of consciousness) possibly related, and the remainder of events doubtfully or not related to esketamine. All 5 events considered very likely or possibly related to esketamine resolved (i.e., participants recovered from the events without sequalae). There were 9 (0.8%) deaths, none considered related to esketamine (Results S1, online Supplementary Material). Of the 1,148 participants, 74 (6.4%) discontinued esketamine due to adverse events (Figure). The most common adverse events leading to discontinuation included: elevated BP (8 [0.7%]), including BP increased (n=6), systolic BP increased (n=1), and hypertensive emergency (n=1), all in participants with). Dissociation, including reports of perceptual disturbances, derealization, or depersonalization, was reported as an adverse event during at least one dosing session for 25.5% of participants (Table), in 10.4% (365/3526) and 4.4% (5720/130976) of dosing sessions during the induction and optimization/maintenance phases, respectively. Almost all (99.8%) dissociation events occurred and resolved on the day of dosing across study phases. Thirty-six participants (3.1%) had a dissociation event(s) that persisted beyond 2 hours, the occurrence of which did not increase over long-term treatment (Supplementary Figure). Few participants (n=5) experienced a dissociation event that led to withdrawal of study drug (Supplementary Table). There were no serious adverse events of dissociation. Sedation was reported as an adverse event during at least one dosing session for 8.2% (94/1148) of participants, in 3.0% (107/3526) and 1.0% (1310/130976) of dosing sessions during the induction and optimization/maintenance phases, respectively. The majority (99.6%) of sedation events occurred on a dosing day and resolved the same day. There were no serious adverse events or adverse events of sedation that led to withdrawal of study drug. No case of treatment-related interstitial/ulcerative cystitis was reported. Urinary tract infections were reported in 181 (15.8%) participants. Other adverse events (incidence ≥1%) related to a renal disorder included dysuria (3.0%), cystitis (2.4%), pollakiuria (2.4%), nephrolithiasis (1.7%), micturition urgency (1.5%), urinary incontinence (1.5%), and hematuria (1.3%). A minority (7.9%) of participants experienced ≥1 hepatic adverse events, the most common being gamma glutamyl transferase increased (2.5%), alanine aminotransferase increased (1.9%), hepatic enzyme increased (1.6%), hepatic steatosis (1.5%), cholelithiasis (1.4%), and aspartate aminotransferase increased (1.3%). The occurrence of hepatic adverse events did not increase over time (Supplementary Figure). Few participants discontinued esketamine due to hepatic adverse events (Supplementary Table).
The greatest mean (SD) change in systolic BP and diastolic BP from predose to a scheduled postdose timepoint (n>10 participants with data) were all at the 40-minute postdose assessment (induction phase: +9. 5 Incidence rates of increased BP events were generally similar at visits throughout both the induction and optimization/maintenance phases (Supplementary Figure). An adverse event related to increased BP was reported in 3.6% (126/3526) and 1.0% (1312/130976) of dosing sessions during the induction and optimization/maintenance phases, respectively. Most (≥95%) increased BP events, including the event of hypertensive emergency, occurred and resolved on the day of dosing. Postdose oxygen saturation (SpO2) levels <93% at ≥2 consecutive assessments were reported for few participants (induction phase: 5/458 [1.1%]; optimization/maintenance phase: 9/1110 [0.8%], including 1 participant in both phases). Of all dosing sessions with postdose SpO2 measurements, 0.1% and <0.1% of dosing sessions in the induction and optimization/maintenance phases, respectively, had postdose SpO2 levels <93% at ≥2 consecutive assessments. Four (0.3%) participants had oxygen saturation decreased reported as a treatment-emergent adverse event, none requiring concomitant treatment or intervention. These incidences were participants reporting no events of suicidal ideation or behavior increased over time during the induction phase, from 79.9% at baseline to 93.2% at induction endpoint (Supplementary Figure). Throughout the optimization/maintenance phase, >90% of participants reported no events of suicidal ideation or behavior (Supplementary Figure).
Efficacy results for the induction phase are reported elsewhere, 13 but are also reported below (from the final dataset) to facilitate comparison to efficacy results for the optimization/maintenance phase.
Depressive symptoms, assessed by MADRS, decreased during induction (Figureat the induction (55.9%) and optimization/maintenance (60.9%) phase endpoints (Figure).
Functioning and associated disability, assessed by SDS score, improved during induction; improvement was maintained during optimization/maintenance (mean [SD] change from the baseline to endpoint of each phase: induction, -6.4 [7.13]; optimization/maintenance, -0.
The long-term safety and efficacy of esketamine nasal spray, combined with an oral antidepressant, was evaluated in the global SUSTAIN-3 study of TRD. The mean duration of intermittent esketamine treatment in this trial was 42.9 months (range up to 79 months). Nearly two-thirds of study participants continued in the SUSTAIN-3 study for 3 years or longer, despite that esketamine could only be administered at in-person clinic visits and the research period overlapped the peak of the COVID-19 pandemic. Importantly, few participants discontinued due to lack of efficacy (5.3% across phases) or an adverse event (6.4% across phases). The safety profile of esketamine, with continuous intermittent dosing for up to 6.5 years in SUSTAIN-3 (3,777 cumulative patient-years) aligns with its established safety profile reported in studies of shorter exposure.Consistent with the safety observations in the 32-week ESCAPE-TRD trial,most adverse events reported in SUSTAIN-3, including events of dissociation and sedation, were transient, occurring and resolving on a dosing day. Long-term exposure to esketamine yielded no new concerns related to increased BP, renal disorders, lower urinary tract symptoms, or suicidality. Additionally, there was no evidence of impaired cognition, interstitial or ulcerative cystitis, or hepatotoxicity. Safety findings have been confirmed in real-world, naturalistic settings,with no evidence of abuse, misuse, or withdrawal within the context of the current delivery model for esketamine nasal spray.In surveillance of realworld data, Dartreported an increasing trend for misuse/abuse of ketamine, but not of esketamine. Few (2%) participants had serious adverse event(s) related to depression that required hospitalization during this long-term study of esketamine, as also was the case (0.6%) in the ESCAPE-TRD trial.These data compare favorably to a prior report of patients with TRD receiving standard-of-care treatments, wherein 8% of patients required hospitalization during a much shorter (6-month) period of study.Incidences of suicide attempts, suicide death, and all-cause mortality are higher among patients with TRD than in the general MDD population.In SUSTAIN-3, 16 patients (1.4%) attempted suicide (0.423 per 100 patient-years), 1 died by suicide (0.0265 per 100 patient-years), and 9 died by any cause (allcause mortality rate 0.238 per 100 patient-years). For context, the rate of non-fatal suicidal behavior was 4.66 per 100 patient-years and rate of suicide death was 0.14-0.47 per 100 patient-years in studies assessing several other TRD treatments.The literature-reported rate of all-cause mortality for TRD ranged from 0.79 31 to 4.6per 100 patient-years. Of note, the rates of suicide-related behavior and death, and of all-cause mortality observed in SUSTAIN-3 may reflect the effectiveness of maintenance treatment with intermittently-dosed esketamine plus daily antidepressant; the rates may also have been influenced by the entrance criteria, as the parent studies from which participants were recruited had excluded volunteers who reported suicidal ideation with intent or plan within the past 6 months or a history of suicidal behavior within the past year, and by the close clinical follow-up that was provided as a result of participating in this clinical trial. Cognition remained stable over time for participants <65 years and participants ≥65 years of age, based on performance on most cognitive tests. Of note, decline on measures of higher cognitive unclear considering the relatively small sample size and the intraindividual variability in performance, which has been observed as a factor in slowing of processing speed/reaction time in longitudinal studies of older, healthy participantsand of mood disorder patients,including patients with MDD aged ≥65.In the absence of a control arm, it cannot be determined whether slowing in reaction time/processing speed is the result of study drug(s), or intrinsic processes among older and/or depressed individuals that contribute to increasing intraindividual variability of performance over time, which in turn result in slowing reaction time/processing speed. The absence of negative effects of esketamine on any aspect of memory, learning, executive function, and working memory in both subgroups makes it unlikely that the subtle changes in simple and choice reaction time reflect the effects of treatment with esketamine. Measures of depressive symptoms and functioning improved during the first 4 weeks of treatment (induction phase) and were sustained for up to 6.5 years with intermittently-dosed esketamine plus daily antidepressant (during the optimization/maintenance phase). Half of participants were in (MADRS-Downloaded fromby guest on 07 May 2025 A c c e p t e d M a n u s c r i p t defined) remission at the optimization/maintenance phase endpoint. Our findings are consistent with and extend those of SUSTAIN-2, in which a 47% remission rate was observed at 12 months.The long-term efficacy of esketamine for TRD is further supported by sustained improvements in clinician-assessed severity of illness (CGI) and patient-reported measures of depressive symptoms (PHQ-9) and functionality (SDS). In SUSTAIN-3, the percentage of participants who achieved functional remission (defined as SDS ≤6) was 22.9% (89/388) and 37.0% (401/1084) at the induction and optimization/maintenance phase endpoints, respectively. The latter finding is consistent with that reported in the ESCAPE-TRD trial, in which 37.0% of esketamine-treated participants achieved functional remission at 32 weeks.In addition, for a study spanning up to 6.5 years in patients with TRD, the proportions of participants who discontinued during optimization/maintenance due to lack of efficacy (4.7%), worsening depression (0.6%), suicide ideation (0.2%), or suicide attempt (0.1%) are relatively low when considered within the context of prior literature reports. Most participants maintained clinical stability with the 84 mg or 56 mg esketamine dose, administered either weekly or every-other-week. The every-4-week dosing interval was used more frequently over time during the optimization/maintenance phase, from about 10% during the first 6 months to about 20% by study end. This suggests that some participants may be maintained with longer intervals between doses during long-term treatment. While these figures may be relevant to clinical practice, research study participants typically have more time and access to clinical staff than patients in the community who are treated outside of research settings. These differences should be considered when applying the dosing and scheduling of participants in the trial to community-based clinical care. Several factors may limit the generalizability of our findings, including the open-label and single-arm design, potential bias of participants electing to proceed from a parent study into this extension study, exclusion of participants with significant psychiatric co-morbidities, lack of racial heterogeneity, and level of care participants received (in addition to esketamine), which generally exceeded that in realworld settings. Participants received esketamine and clinical care in SUSTAIN-3, with most participants staying in the study; thus, the SUSTAIN-3 results cannot determine the optimal duration of treatment in clinical practice. The study design did not allow extending the dosing interval beyond 4 weeks, as may be observed in real world practice. Additional analyses are warranted to further assess disease characteristics and their course over time, impact of comorbidities, and the optimal dosing regimen. In summary, no new safety signals were identified during long-term treatment with intermittently-dosed esketamine, combined with daily antidepressant, and improvement in depression generally persisted among participants who remained on maintenance treatment. Taken together, these data from SUSTAIN-3 add to the evidence supporting the safety of long-term use of intermittently-dosed esketamine for treating TRD. A c c e p t e d M a n u s c r i p t Other reasons (each ≤1%, e.g., investigator/sponsor decision, employment/school, personal reasons). c A participant was considered to have completed the study if they actively participated in the induction or optimization/maintenance phase until esketamine was approved in the respective country and accessible through the local healthcare system funding or until end of December 2022, whichever was earlier.