Predictors and correlates of outcome for dorsolateral, dorsomedial, and orbitofrontal rTMS in major depression

Objective: It is increasingly being recognized that depression is clinically heterogeneous in terms of clinical presentation and neuroimaging. Repetitive transcranial magnetic stimulation (rTMS) is a neuroanatomically focal treatment for depression that may be used to probe this heterogeneity. Here, we examine the predictors and correlates of response to rTMS targeting the dorsolateral prefrontal cortex (DLPFC), dorsomedial prefrontal cortex (DMPFC), or orbitofrontal cortex (OFC). It is expected that each rTMS target will be associated with distinctive mechanisms of outcome, centered in different resting-state networks implicated in depression. Method: Resting-state fMRI data was collected in 120 patients with depression before and after receiving rTMS targeting the DLPFC (n=50), DMPFC (n=40) or OFC (n=30). An age- and sex-matched comparator group of 50 healthy controls was included to examine relative connectivity changes following rTMS. Baseline sgACC connectivity to the DLPFC and stimulation site was examined as a predictor of treatment outcome. fMRI predictors and correlates of outcome were also examined with seed-based analyses (using DMPFC and nucleus accumbens as a priori regions of interest). Results: sgACC-DLPFC connectivity was unable predict treatment outcome for any of the rTMS targets, while sgACC to stimulation site connectivity only predicted outcome for DMPFC-rTMS. DLPFC-, DMPFC-, and OFC-rTMS were shown to have differential predictors and correlates using seed-based analyses, suggesting that each has different mechanisms of action. Conclusions: The mechanisms of action for DLPFC-, DMPFC-, and OFC-rTMS are distinctive, and do not universally involve connectivity to the sgACC, but instead involve different resting-state networks, highlighting the heterogeneity of depression.