Dorsolateral Prefrontal rTMS Symptom Cluster Response Trajectories for Depression

Importance Repetitive transcranial magnetic stimulation (rTMS) is an established intervention for treatment-resistant depression, but only a minority of patients achieve remission, and characteristics associated with response are lacking. Sum-score analyses of the Hamilton Depression Rating Scale (HDRS) obscures how distinct symptom clusters change over time. Objective To identify distinct longitudinal symptom-cluster response trajectories during dorsolateral prefrontal rTMS and to determine baseline characteristics associated with trajectory membership. Design, Setting, and Participants This cohort study used data from the THREE-D randomized clinical trial (conducted from 2013 to 2016 at 3 Canadian academic hospitals), applying group-based multitrajectory modeling to mood, anxiety, insomnia, and somatic symptom cluster scores over 4 weeks of rTMS. Analysis was performed in 2024 to 2026. Participants were adults aged 18 to 65 years with major depressive disorder. Outcome assessors were blinded to treatment allocation. Exposures Once-daily high-frequency (10 Hz) or intermittent theta-burst rTMS to the left dorsolateral prefrontal cortex for 4 to 6 weeks. Main Outcomes and Measures Primary outcomes were sum scores of 4 previously validated HDRS-17 symptom clusters. Trajectories were derived using group-based multitrajectory modeling; baseline characteristics associated with membership were examined with weighted multinomial logistic regression. Results Of 414 randomized participants, 26 were excluded, leaving 388 in the analytic cohort (mean [SD] age, 42.33 [11.48] years; 229 female [59.02%]). Four trajectory groups were identified: optimal response (119 participants [30.67%]); partial response, high anxiety (128 participants [32.99%]); partial response, low anxiety (91 participants [23.45%]); and minimal response (50 participants [12.88%]). Final response and remission rates ranged from 83.48% and 65.22%, respectively, in the optimal response group to 0% in the minimal response group. Compared with the partial response, high anxiety reference, minimal response was associated with younger age (odds ratio [OR], 0.95; 95% CI, 0.92-0.99; P = .01), benzodiazepine use (OR, 2.73; 95% CI, 1.24-6.05; P = .01), lower baseline anxiety (OR, 0.73; 95% CI, 0.60-0.89; P < .001), and higher baseline mood symptoms (OR, 1.35; 95% CI, 1.13-1.61; P < .001). Conclusions and Relevance In this cohort study of adults with treatment-resistant depression, 4 longitudinal symptom-cluster response trajectories were identified, with marked divergence in clinical outcomes. Baseline anxiety severity and benzodiazepine use were associated with trajectory membership and may inform personalized treatment planning, although prospective studies are required.