Comparing transcranial magnetic stimulation and esketamine treatment response trajectories in resistant depression.

In a real-world comparison of two FDA-approved treatments for treatment-resistant depression, intranasal esketamine led to faster improvement than repetitive transcranial magnetic stimulation (rTMS). Over 90 days, esketamine patients responded a median of 36 days versus 49 days for rTMS, and suicidal ideation resolved more quickly (median 9 vs. 26 days). However, by about 90 days, overall response and remission rates were similar between the groups (68.8% and 45.2% for esketamine; 59.4% and 40.1% for rTMS), suggesting a difference in speed rather than ultimate effectiveness. For rTMS, slower response was predicted by comorbid anxiety and benzodiazepine use, while former tobacco use predicted faster response. No such predictors were found for esketamine.

Study at a glance

Abstract

Repetitive transcranial magnetic stimulation (rTMS) and intranasal esketamine are FDA-approved for treatment-resistant depression (TRD), yet comparative real-world data on response trajectories and predictors of outcomes remain limited. A retrospective analysis was performed using electronic medical records from UC San Diego Health. Adults with TRD treated with rTMS (n = 279) or intranasal esketamine (n = 93) between 2017 and 2025 were included. The primary outcome was clinical response (≥50% PHQ-9 reduction). Time-to-response was assessed using inverse probability treatment weighted (IPTW) Cox models; Kaplan-Meier curves and log-rank tests provided descriptive comparisons. Covariates included age, trauma history, anxiety comorbidity, benzodiazepine use, tobacco use history, BMI, and baseline symptom severity. Secondary outcomes included remission (PHQ-9 < 5) and suicidal ideation (SI). Esketamine demonstrated earlier response over 90 days (RMST difference = -11.94 days) and faster time-to-response in the IPTW Cox model (HR = 1.62, p = 0.005); KM estimates showed median response at 36 vs. 49 days (p = 0.0096) with convergence by ∼90 days. Cumulative response and remission rates were numerically higher for esketamine (68.8% / 45.2%) than rTMS (59.4% / 40.1%), supporting a speed-of-response difference rather than superior overall efficacy. SI improved more rapidly with esketamine (median 9 vs. 26 days; p = 0.001). In the rTMS cohort, comorbid anxiety (HR = 0.69, p = 0.039) and benzodiazepine use (HR = 0.73, p = 0.046) predicted slower response, while former tobacco use predicted faster response (HR = 1.31, p = 0.006). No significant predictors emerged for esketamine. Esketamine was associated with earlier observed antidepressant and anti-suicidal improvement than rTMS. Baseline factors, including benzodiazepine use, may help inform expectations regarding rTMS response trajectory.