Real-World Safety of Esketamine Nasal Spray: A Comprehensive Analysis Almost 5 Years After First Approval

Depression is highly prevalent and many patients do not achieve sustained remission with available oral antidepressants, leaving a sizeable proportion with treatment-resistant depression. Esketamine nasal spray, approved in the United States initially in March 2019 for treatment-resistant depression and later for acute suicidal ideation or behaviour, has been shown in clinical trials to produce rapid antidepressant effects when administered under supervision. Nonetheless, questions remain about its safety in routine clinical practice, including concerns about sedation, dissociation, increases in blood pressure, suicidality, and potential for abuse or misuse, especially given differences between esketamine and racemic ketamine. Johnston and colleagues set out to provide a comprehensive real-world safety assessment of esketamine nasal spray in the United States by analysing 58 months of postapproval data. The study aimed to describe patient characteristics, dosing and utilisation patterns, the incidence of pre-specified adverse events of interest (sedation, dissociation, increased blood pressure), serious adverse events, suicidality, and reports suggestive of abuse or misuse using data from the REMS programme and Janssen’s US Global Medical Safety database covering March 5, 2019 to January 5, 2024. This large-scale surveillance effort is intended to characterise the safety profile of esketamine in outpatient settings under REMS safeguards.

The analysis combined two primary U.S. data sources: patient monitoring forms submitted to the esketamine Risk Evaluation and Mitigation Strategy (REMS) programme and adverse event reports held in Janssen’s U.S. Global Medical Safety (US-GMS) database. The evaluation window spanned from the date of FDA approval (March 5, 2019) through January 5, 2024. Interventional clinical trial reports were excluded. REMS data consisted of solicited patient monitoring forms returned by certified outpatient sites within 7 days of each treatment session; the US-GMS database included spontaneous and solicited reports from multiple sources and was restricted to U.S. reports for comparability. Cases were defined at the treatment-session level (one record per treatment session per patient), so multiple cases could arise for a single patient. Valid US-GMS cases required minimal identifying information and an adverse event or special reporting situation. Because identifiers are sometimes absent in spontaneous reports, the US-GMS dataset may contain duplicate cases. The REMS dataset provided counts of treatment sessions; the cumulative number of U.S. sessions used as the denominator was 1,486,213. Key outcomes included solicited adverse events of special interest (sedation, dissociation, and increased blood pressure), serious adverse events, suicidality-related events, and reports suggestive of drug abuse/misuse. Increased blood pressure was defined as a systolic rise of ≥20 mm Hg reaching ≥180 mm Hg and/or a diastolic rise of ≥15 mm Hg reaching ≥105 mm Hg at 40 minutes after administration versus pre-administration values; when baseline blood pressure was missing, absolute thresholds at 40 minutes were used. Treatment phases were defined as induction (sessions 1–8), early maintenance (sessions 9–12), and late maintenance (session 13 onward). Descriptive statistics summarised patient demographics, dosing patterns, and adverse event incidence, expressed as events per 100 patients or per 100 administrations; patient‑years exposure was computed from summed treatment-day ranges for fatality incidence estimates.

As of January 5, 2024, 58,483 outpatients had at least one esketamine treatment session, totalling 1,486,213 treatment sessions and 37,862.7 patient‑years of exposure. Most patients were aged 26–55 years (64.3%), female (61.1%), and 35.6% resided in the Southern U.S. A substantial proportion continued treatment: 43,908 patients (75.1%) received ≥9 sessions and 35,257 (60.3%) received ≥13 sessions. The mean number of sessions per patient was 25.4 (SD=29.7; range 1–420) and the median was 16, indicating a skewed distribution. Dosing patterns showed that over 80% of patients began at the 56‑mg dose, with the majority transitioned to 84 mg by session 6 (>80%). Older patients (>75 years) were less likely to receive 84 mg. Across 1,486,213 sessions, REMS data reported one or more solicited safety events (sedation, dissociation, or increased blood pressure) in 764,600 sessions (51.4%). Solicited rates per session were 34.7% for sedation, 41.0% for dissociation, and 0.9% for increased blood pressure. Median time to resolution for solicited sedation and dissociation events was 70 and 60 minutes, respectively. Event rates were highest at the first treatment session and declined slightly across induction and early maintenance phases. For example, sedation events per 100 administrations were 39.0 at session 1 versus 37.0 during induction and 35.3 during early maintenance; dissociation showed 46.9 at session 1 versus 43.5 (induction) and 41.2 (early maintenance); increased blood pressure was 1.6 at session 1 versus 1.2 and 0.9, respectively. Serious adverse events were uncommon: REMS recorded 2,096 serious adverse events in 1,184 treatment sessions (<0.1% of sessions). The US-GMS search identified 2,742 serious cases (0.18%) and 3,985 serious events (0.27%). Fatality data in US-GMS included 308 cases referred to life‑threatening or fatal events (0.02% of sessions) and 228 deaths; the estimated incidence of death was 0.6 per 100 patient‑years. The GMS medical team did not consider any deaths to be related to esketamine treatment, although some cases could not be fully assessed. Per REMS, the most common serious-event terms during sessions were vomiting, increased blood pressure, and nausea; per US-GMS the most common serious events (including outside-session occurrences) were suicidal ideation, hospitalization, hypertension, and vomiting. Suicidality-related reports in US-GMS comprised 131 fatal or life‑threatening events, including 70 completed suicides, 30 reports of suicidal ideation, and 21 suicide attempts. Estimated incidences were 0.18 completed suicides and 0.06 suicide attempts per 100 patient‑years. Latency information was available for 52 of the 70 completed suicides; 26 occurred within 1 week of the most recent dose. Other events in US-GMS included 976 cases of increased blood pressure (349 serious), 193 cardiac arrhythmia/heart‑rate events (140 serious), 210 reports suggestive of drug abuse/misuse (110 serious), and two nonserious cases of cystitis.

Johnston and colleagues interpret these pooled REMS and US-GMS data as indicating a real-world safety profile for esketamine nasal spray that is consistent with clinical-trial experience and product labelling. Solicited sedation and dissociation were relatively common immediately after dosing but tended to resolve within approximately 1–2 hours and decreased modestly across subsequent treatment sessions. Serious adverse events were rare, occurring in <0.1% of treatment sessions in REMS and 0.18% in US-GMS, and no new safety signals were identified across almost 5 years of postmarketing data. The authors note several possible explanations for the observed decline in event rates over time: physiological adaptation or increasing tolerance to expected adverse effects, survivor bias whereby patients who tolerate treatment continue, and improved adverse-event management by clinicians as experience accumulates. Regarding mortality and suicidality, the reported rates of death and completed suicide were described as consistent with or lower than background estimates for patients with treatment-resistant depression from prior observational studies and meta-analyses; however, temporal patterns were not definitive and some cases lacked sufficient information for causal assessment. The report emphasises that REMS-driven administration within certified health‑care settings provides a regulated framework likely to mitigate misuse risk compared with off‑label ketamine, which lacks comparable distribution controls. Cystitis, reported with chronic recreational ketamine use, was rare in this dataset (two nonserious events). The authors acknowledge limitations of the analysis: missing or illegible REMS form fields, the solicited nature of REMS reporting which may inflate frequencies of certain events, variability in reporter training and adherence to serious‑event reporting standards, absence of efficacy measures in REMS, potential discrepancies and duplication between REMS and US-GMS records, and incomplete observation for patients who began treatment close to the data‑lock date. Taking these caveats into account, the investigators conclude that the cumulative postmarketing experience through January 5, 2024 remains aligned with the established safety profile of esketamine nasal spray and that no new safety concerns have emerged.