There is a critical need for more effective treatments for depression, which currently affect 20% of Americans during our lifetimes. Brain stimulation treatments, including repetitive transcranial magnetic stimulation (rTMS), represent the front line of innovative approaches by directly targeting and correcting specific dysfunctional brain networks. A core dysfunctional network in major depressive disorder is the fronto-parietal central executive network (CEN), a network critical for decision making and cognitive control. The CEN includes the dorsolateral prefrontal cortex (dlPFC), the target of FDA-cleared rTMS treatment for depression. rTMS to the dlPFC is thought to improve depression by modulating local dlPFC excitability and enhancing downstream CEN connectivity. However, our ability to probe the CEN and study this potential mechanism on an individual basis is critically lacking, likely contributing to suboptimal rTMS remission rates (20-40%). We hypothesize that the CEN connectivity is weakened in depression and can be maximally modulated by individualizing localization.
To test this hypothesis, in a single rTMS 'dose,' prospective, randomized, double-blind, cross-over design with 50 depressed patients, we will prospectively compare the strength, duration, and specificity of CEN modulation after a single session of dlPFC rTMS. These participants will be 18-65 years old and require a current major depressive disorder diagnosis assessed by Structured Clinical Interview for DSM5 (SCID-I62), with a PHQ9\>10. Exclusion criteria includes contraindications for MRIs (e.g. implanted metal), history of head trauma with loss of consciousness, history of seizures, neurological or uncontrolled medical disease, active substance abuse, a history of suicide attempt in the past year, psychotic or bipolar disorders, a prior history of ECT or rTMS failure, and medications that substantially reduce seizure threshold (e.g., bupropion, clozapine).
Following the diagnostic session, participants will undergo a 30-minute MRI session to record structural brain data. For the following sessions, dlPFC will be targeted for each session using different methods and 10Hz dlPFC rTMS will be applied guided by (a) individualized CEN targeting, (b) structural MRI, (c) standard scalp targeting. For each condition, a single session of rTMS at standard parameters (10Hz, 5s on, 10s off, 3,000 total pulses, 15 min duration) will be performed and changes in CEN connectivity will be quantified using pre/post dlPFC-stimulated parietal TMS-evoked potentials (TEPs). The dlPFC will be targeted for rTMS using three methods: (a) MRI-guided with individual CEN optimization using TEPs, (b) MRI-guided alone, and (c) standard scalp targeting (Beam F3 method99). Additionally, a fourth session of sham rTMS will be applied to control for off-target effects. We hypothesize that while each active rTMS method (condition a-c) will suppress the p30 of the TEP in the CEN, optimized CEN localization using individual TEPs (condition a) will induce the strongest and most specific change in the CEN for the longest duration. Our primary outcome will be parietal p30 CEN modulation directly following rTMS. Secondary outcomes will assess parietal p30 changes in the parietal node of the CEN during rTMS (quantifying the p30 after the last pulse in each stimulation train) as well as 15 and 30 min following rTMS. We will also assess pre/post rTMS behavioral changes in attention with a standard continuous performance task and working memory using an N-back task, both of which have been implicated in the CEN and depression100,101. rTMS sessions will be triple-blinded to operator, participant, and statistician. rTMS sessions will be separated by at least two days to remove potential lasting effects \>24 hours, and rTMS session order will be randomized and counterbalanced to reduce any potential bias.
Findings from this study will provide the basis for a clinical trial comparing rTMS treatment outcome using this personalized targeting approach against standard rTMS.
