FormalPara Key Points

This subgroup analysis evaluated efficacy and safety outcomes in patients with treatment-resistant depression who relapsed while taking or after discontinuing esketamine nasal spray (ESK) plus an oral antidepressant during SUSTAIN-1 and received a second induction and maintenance treatment with ESK in SUSTAIN-3.

Patients with treatment-resistant depression who previously responded to ESK and experienced a relapse benefitted from receiving a second induction and maintenance treatment with ESK.

ESK was generally well tolerated and no new safety signals were identified.

1 Introduction

Major depression is one of the leading causes of disability worldwide, affecting approximately 8.9 million adults with medication-treated major depressive disorder (MDD) in the United States, including approximately 2.8 million adults (30.9%) who exhibit treatment-resistant symptoms [1]. Individuals with treatment-resistant depression (TRD) often experience prolonged depressive episodes, which can lead to treatment discontinuation prior to reaching recovery [2]. These prolonged or recurrent depressive episodes can result in difficulties with social and occupational function, decline of physical health, suicidal thoughts, and increased health care utilization [3, 4]. While there is no universal definition, TRD is commonly defined as the absence of a response to two or more oral antidepressants (ADs) of adequate dose and duration [4, 5]. Response and remission are commonly used outcome criteria for the treatment of depression with the goal of achieving remission because these outcomes can be predictors of long-term stability [5, 7]. Therefore, identifying and optimizing treatment options for patients with TRD represents a substantial unmet need [4, 8].

Esketamine nasal spray (ESK) is a noncompetitive N-methyl-d-aspartate (NMDA) receptor antagonist approved by the United States Food and Drug Administration, as well as health authorities in other countries, in conjunction with an oral AD, for the treatment of TRD in adults and for the treatment of depressive symptoms in adults with MDD with acute suicidal ideation or behavior [9, 10]. Several real-world studies have demonstrated the benefit of ESK treatment in adult patients with TRD [11, 12]. However, in some cases, after a reduction in frequency of ESK dosing or treatment discontinuation, patients may experience a relapse or a worsening of depressive symptoms [8]; in these scenarios, clinicians may consider restarting treatment with ESK if clinically appropriate. A recent study found that a subset of patients ≥65 years of age, who were nonresponders to ESK treatment after completing the double-blind induction (IND) phase of a randomized 4-week efficacy study and subsequently entered the IND phase of an open-label, long-term, safety and efficacy study of ESK (SUSTAIN-2), received clinical benefit from a second IND of ESK treatment [13]. Understanding the outcomes of patients with TRD who previously responded to ESK, experienced a relapse, and subsequently received a second IND and maintenance treatment with ESK is of clinical interest due to the complex nature of treating patients with TRD [3, 4].

The SUSTAIN-3 study (ClinicalTrials.gov identifier NCT02782104) is a long-term, open-label, phase III study evaluating the efficacy, safety, and tolerability of ESK [14,15,16]. The SUSTAIN-3 study provides an opportunity to assess a subset of patients who relapsed in a previous study (SUSTAIN-1; NCT02493868) and were eligible to receive a second induction in SUSTAIN-3 [8]. Patients in SUSTAIN-1 could have experienced relapse on oral AD + placebo (PBO) or ESK + AD (e.g., following treatment discontinuation or a reduction in ESK dosing frequency). Overall, the SUSTAIN-1 study demonstrated that continued treatment with ESK+AD resulted in a significant and clinically meaningful delay in relapse compared with AD+PBO in patients with TRD. This subgroup analysis of the ongoing SUSTAIN-3 study evaluated efficacy and safety outcomes in patients who relapsed while taking or after discontinuing ESK during SUSTAIN-1 and received a second induction followed by maintenance treatment with ESK.

2 Methods

2.1 Study Design and Treatment

The SUSTAIN-1 relapse prevention study had five phases: a 4-week screening phase, a 4-week open-label (OL) IND phase, a 12-week optimization phase, a variable-duration maintenance phase, and a 2-week follow-up phase. After the 4-week IND phase and 12-week optimization phase (in which all patients received ESK+AD), patients who achieved stable remission or stable response were randomly assigned 1:1 to continue ESK treatment or switch to PBO. ESK dosing of 56 or 84 mg in SUSTAIN-1 was flexible, twice weekly for 4 weeks in IND (with dosing frequency subsequently individualized based on depression severity in the optimization and maintenance phases), and in conjunction with an oral AD. Full study design and inclusion and exclusion criteria for SUSTAIN-1 are published elsewhere [8].

SUSTAIN-3 is an ongoing, multicenter, phase III, OL extension study to evaluate the efficacy, safety, and tolerability of ESK, given in conjunction with an oral AD, in adults with TRD. Data collected from June 21, 2016, to December 1, 2020, were used in this interim analysis with patients having up to 48 months of treatment at data cutoff. The study comprises two phases: a 4-week IND phase and a variable-duration optimization/maintenance (OP/M) phase (Fig. 1). Dosing of 56 or 84 mg of ESK in SUSTAIN-3 was flexible, twice weekly for 4 weeks in IND. During the OP/M phase, ESK was self-administered as a flexible-dose regimen of 56 or 84 mg, with adjustments according to efficacy and tolerability based on the Clinical Global Impression-Severity (CGI-S) score. Patients could enter the IND phase of SUSTAIN-3 from SUSTAIN-1 if they attained stable remission or response status with ESK after an initial 16 weeks of treatment and subsequently experienced relapse in the maintenance phase of SUSTAIN-1 (defined as Montgomery–Åsberg Depression Rating Scale [MADRS] [17] total score ≥22 for two consecutive assessments separated by 5–15 days and/or hospitalization for worsening depression or any other clinically relevant event).

Fig. 1
figure 1

SUSTAIN-3 study design. aStable remission was defined as MADRS total score ≤12 for at least 3 of the last 4 weeks of the optimization phase; stable response was defined as ≥50% reduction in the MADRS total score from baseline in each of the last 2 weeks of the optimization phase but does not meet criteria for stable remission. bBased on CGI-S and tolerability. AD antidepressant, CGI-S Clinical Global Impression-Severity, ESK esketamine nasal spray, OL open label, PR-AD+PBO previously relapsed on an oral AD plus placebo, PR-ESK+AD previously relapsed on ESK plus an oral AD

Full size image

2.2 Patient Population

Adult patients 18–64 years of age determined to be medically stable on the basis of a physical examination, vital signs, and 12-lead electrocardiogram performed on the day of the first treatment session and according to the investigator’s judgment and knowledge of the participant’s stability in SUSTAIN-1 were eligible. Patients were also required to meet the Diagnostic and Statistical Manual of Mental Disorders—Fifth Edition (DSM-5) criteria for recurrent MDD, without psychotic features, confirmed by the Mini International Neuropsychiatric Interview. Patients were not eligible if the benefits versus risk of continued ESK treatment was not favorable in the opinion of the investigator or if the patient had suicidal ideation with intent to act since the last study visit in SUSTAIN-1. Patients who had a neurodegenerative disorder, evidence of mild cognitive impairment, ongoing evidence of uncontrolled hypertension, or a positive test result for drugs of abuse on the day of the first treatment session (pre-dose) were excluded. Patients were also excluded if they had a current or prior DSM-5 diagnosis of a psychotic disorder, bipolar or related disorders, obsessive compulsive disorder (current only), intellectual disability, autism spectrum disorder, borderline personality disorder, antisocial personality disorder, histrionic personality disorder, or narcissistic personality disorder.

2.3 Efficacy and Safety Assessments

Clinician- and patient-reported disease severity measures, including the MADRS and the Patient Health Questionnaire 9-item (PHQ-9) [18], were administered. The MADRS was assessed by site clinicians at baseline, on days 8, 15, 22, and 28 of the IND phase, and every 4 weeks during the OP/M phase. The PHQ-9 was completed by patients at baseline, on days 15 and 28 of the IND phase, and every 4 weeks during the OP/M phase. The proportions of patients who attained response or remission status were determined with both MADRS and PHQ-9 for day 28 of the IND phase and weeks 24 and 52 of the OP/M phase. For MADRS, response was defined as ≥50% improvement in total score from baseline, and remission was defined as a score ≤12 [19, 20]. For PHQ-9, response and remission were defined as ≥50% improvement in total score from baseline and total score <5 (normal), respectively, which are commonly used clinical definitions [21,22,23,24]. Efficacy outcome results are summarized descriptively. The incidence of treatment-emergent adverse events (TEAEs) and serious adverse events (SAEs) were also evaluated.

3 Results

3.1 Patients

This analysis included a total of 96 patients who had relapsed during the SUSTAIN-1 study and subsequently entered the IND phase in SUSTAIN-3; most patients (98%) continued to the OP/M phase of SUSTAIN-3. Of these patients, 32 had been taking ESK+AD and 64 had been taking AD+PBO at time of relapse in the maintenance phase of SUSTAIN-1; these patient groups are henceforth referred to as PR-ESK+AD (previously relapsed on ESK plus an oral AD) and PR-AD+PBO (previously relapsed on an oral AD plus PBO after receiving ESK). Mean [SD] disease severity scores at baseline of the second IND phase (SUSTAIN-3) were consistent with moderate to severe disease for MADRS (PR-ESK+AD: 26.5 [9.13]; PR-AD+PBO: 29.2 [6.53]) and PHQ-9 (PR-ESK+AD: 12.3 [4.65]; PR-AD+PBO: 13.7 [5.00]). Mean age at baseline was 46.0 and 44.5 years for PR-ESK+AD and PR-AD+PBO, respectively; 75.0% and 67.2% were female, and 90.6% of patients in each group were White (Table 1).

Table 1 Baseline demographics and disease characteristics in SUSTAIN-3
Full size table

3.2 Prior Exposure to Study Medication

Following randomization, during the maintenance phase of SUSTAIN-1, the median extent of exposure to ESK was 9.1 weeks in the PR-ESK+AD group and the median exposure to PBO was 3.8 weeks in the PR-AD+PBO group. The median duration between relapse in SUSTAIN-1 and the beginning of the second IND in SUSTAIN-3 was 22.0 days and 16.0 days in the PR-ESK+AD group and PR-AD+PBO group, respectively. Most patients (71.9%) in the PR-ESK+AD group were taking 84 mg of ESK at the end of the SUSTAIN-1 optimization phase. Prior to relapse, most patients were dosed weekly in both the PR-ESK+AD group (66.7%) and the PR-AD+PBO group (68.4%).

3.3 Clinician- and Patient-Rated Depressive Symptoms

Substantial improvements in depressive symptoms were observed over the second 4-week IND phase in both groups, with mean (SD) changes from baseline in MADRS total scores of −14.6 (9.84) and −19.1 (9.16), respectively, for PR-ESK+AD and PR-AD+PBO; these improvements were maintained over the course of the OP/M phase (Fig. 2). A similar trend was observed for PHQ-9 scores, with mean (SD) changes from baseline of −5.3 (4.92) and −8.2 (5.77) for PR-ESK+AD and PR-AD+PBO, respectively, with improvements maintained over the OP/M phase (Fig. 3). Similar MADRS and PHQ-9 results were also observed in the combined PR-ESK+AD and PR-AD+PBO groups, with mean (SD) changes from baseline over the second 4-week IND phase of −17.6 (9.59) for MADRS and −7.3 (5.65) for PHQ-9 (Supplemental Fig. 1, see electronic supplementary material [ESM]). The PR-ESK+AD group had numerically higher mean MADRS and PHQ-9 total scores (last observation carried forward [LOCF]) at the end of the 4-week IND phase (IND endpoint) and at weeks 24 and 52 in the OP/M phase compared with the PR-AD+PBO group (Table 2).

Fig. 2
figure 2

Mean MADRS total scores (LOCF) during the reinduction and optimization/maintenance phases of SUSTAIN-3 among patients who previously experienced relapse in SUSTAIN-1. IND induction, LOCF last observation carried forward, MADRS Montgomery-Åsberg Depression Rating Scale, OP/M optimization/maintenance, PR-AD+PBO previously relapsed on an oral antidepressant plus placebo, PR-ESK+AD previously relapsed on esketamine  nasal spray plus an oral antidepressant

Full size image
Fig. 3
figure 3

Mean PHQ-9 total scores (LOCF) during the reinduction and optimization/maintenance phases of SUSTAIN-3 among patients who previously experienced relapse in SUSTAIN-1. IND induction, LOCF last observation carried forward, OP/M optimization/maintenance, PHQ-9 Patient Health Questionnaire 9-item, PR-AD+PBO previously relapsed on an oral antidepressant plus placebo, PR-ESK+AD previously relapsed on esketamine nasal spray plus an oral antidepressant

Full size image
Table 2 Mean MADRS and PHQ-9 total scores (LOCF)
Full size table

3.4 Response and Remission

In the PR-ESK+AD group, 71.9% and 62.5% of patients achieved response and remission based on MADRS, respectively, at the IND phase endpoint; response and remission rates were 37.5% and 40.6% at week 24 in the OP/M phase and 40.6% and 37.5% at week 52 in the OP/M phase, respectively (Table 3). Corresponding response and remission rates based on PHQ-9 scores were 46.9% and 21.9% at the IND phase endpoint, 31.3% and 28.1% at week 24 in the OP/M phase, and 31.3% and 28.1% at week 52 in the OP/M phase.

Table 3 Proportion of patients in SUSTAIN-3 who attained response or remission status according to MADRS and PHQ-9 scores (LOCF)
Full size table

In the PR-AD+PBO group, 73.4% and 60.9% of patients achieved response and remission based on MADRS, respectively, at the IND phase endpoint; response and remission rates were 69.4% and 58.1% at week 24 in the OP/M phase and 54.8% and 41.9% at week 52 in the OP/M phase, respectively (Table 3). Corresponding response and remission rates based on PHQ-9 scores were 63.5% and 53.1% at the IND phase endpoint, 57.4% and 41.9% at week 24 in the OP/M phase, and 49.2% and 33.9% at week 52 in the OP/M phase.

3.5 Safety

During the IND and OP/M phases, 58.3% and 83.3% of patients experienced a TEAE, respectively; the most common TEAEs in the IND phase were dissociation (17.7%), dizziness (13.5%), vertigo (13.5%), and dysgeusia (12.5%). The most common TEAEs in the OP/M phase included dissociation (27.1%), headache (25.0), dizziness (21.9%), vertigo (20.8%), and somnolence (18.8%; the full list of the most common [≥10%] TEAEs is shown in Table 4). No patients experienced an SAE during the IND phase; 12 patients (12.8%) experienced an SAE during the OP/M phase. The most common SAEs were psychiatric disorders (3.1%; including affect lability, depressed mood, depression, suicidal ideation, and major depression); neoplasms benign, malignant, and unspecified (2.1%; including prostate cancer and esophageal squamous cell carcinoma); and nervous system disorders (2.1%; including loss of consciousness and metabolic encephalopathy). Overall, no new safety signals were identified from this analysis.

Table 4 Most common TEAEs (occurring in ≥10% of patients) during the IND and OP/M phases of SUSTAIN-3
Full size table

During the OP/M phase, four patients (4.3%) withdrew from the study due to a TEAE (one patient who experienced a TEAE of major depression in SUSTAIN-1 withdrew from SUSTAIN-3 due to major depression; no other patients had overlapping TEAEs in both studies). No patients withdrew from the IND phase due to a TEAE. Overall, rates of discontinuation during the OP/M phase were similar among groups (PR-ESK+AD: 28.1%; PR-AD+PBO: 22.6%) over a period of up to 52 weeks of follow-up. Other reasons for discontinuation during the OP/M phase included withdrawal by subject (n = 3, 3.2%), lack of efficacy (n = 2, 2.1%), loss to follow-up (n = 1, 1.1%), and death (n = 1, 1.1%; patient died from COVID-19 disease, which was not considered related to ESK treatment by the investigator).

4 Discussion

Patients with TRD who previously responded to treatment with ESK and relapsed experienced improvement in depressive symptoms and functioning during a second course of IND therapy; this improvement was sustained during maintenance treatment. Patients appeared to benefit from a second IND of ESK treatment regardless of whether they experienced relapse after discontinuation of ESK (PR-AD+PBO) or while continuing treatment with ESK (PR-ESK+AD). Overall, ESK was well tolerated with no new safety signals identified. No patients discontinued due to a TEAE during a second IND; however, four patients discontinued due to TEAEs during the OP/M phase. The overall rates of discontinuation during the OP/M phase were similar between groups (PR-ESK+AD: 28.1%; PR-AD+PBO: 22.6%) over a period of up to 52 weeks of follow-up. To provide context, a long-term open-label study of vortioxetine for up to 52 weeks reported that 310/834 (37.2%) patients discontinued treatment early for a variety of reasons [25].

The patient trajectories in this study (i.e., discontinuing ESK treatment or relapsing while on ESK) could be reflective of real-world clinical scenarios. Approximately one-third of the patients included in this analysis had relapsed while receiving ESK on a maintenance schedule during SUSTAIN-1, a scenario that could occur outside of a clinical trial setting. The other two-thirds relapsed after discontinuing ESK, which would likely occur in clinical settings if patients opted not to continue ESK on a maintenance schedule.

In SUSTAIN-1, both patient groups received an initial 16 weeks of ESK treatment, were randomly assigned to either continue on ESK treatment or switch to PBO treatment, and subsequently relapsed during the maintenance phase. Notably, patients from SUSTAIN-1 had a similar trend in mean MADRS scores over time compared with the subgroup of patients in this analysis, suggesting that a second course of ESK IND and maintenance treatment is comparably beneficial to patients [8]. Additionally, when examining the combined PR-ESK+AD and PR-AD+PBO groups, substantial improvements were observed over the second 4-week IND phase, with mean (SD) changes from baseline of −17.6 (9.59) for MADRS and −7.3 (5.65) for PHQ-9. Overall, these findings provide evidence to help guide clinical decision making regarding the potential benefit of a second ESK IND and treatment course in both the PR-ESK+AD group and the PR-AD+PBO group.

Several limitations should be noted when considering this analysis. First, SUSTAIN-3 was an open-label study with no comparator group. Second, the generalizability of the study findings may be limited by the exclusion of participants with significant psychiatric or medical comorbidities or substance use disorder. Third, there may be an inherent variability associated with PHQ-9 scores as they are patient reported. Finally, the small sample size of the subgroups may limit the statistical power and interpretation of results. Of note, the results in the PR-ESK+AD group were driven by a very small number of patients (n = 32).

5 Conclusions

This subgroup analysis found that ESK was generally well tolerated and patients with TRD who previously responded to ESK and experienced a relapse benefited from receiving a second induction and maintenance treatment with ESK. These findings support the potential benefit of a second ESK IND and treatment course in patients with TRD who relapsed while taking or after discontinuing ESK and can help guide clinical decision making in the future.