Accelerated Resting-State fMRI-Guided Theta Burst Stimulation to the Right DLPFC for Late-Life Depression: A Pilot Study

Abstract

<h3>Introduction</h3> Repetitive transcranial magnetic stimulation (rTMS) is FDA-approved for treatment of major depressive disorder; however, its efficacy in late-life depression (LLD) may be reduced due to age-related changes (see Figure 1). rTMS of the dorsolateral prefrontal cortex (DLPFC) has been shown to affect mood control via modulation of the subgenual cingulate cortex (SgCC). The area within the DLPFC having maximum anticorrelated activity to the SgCC is currently believed the most efficacious target for rTMS for depression. While trials of accelerated neuronavigated rTMS to the left DLPFC for depression have shown large effects for depression,¹ there has not yet been a validation of this technique specifically for LLD. Additionally, no studies have tested accelerated intermittent theta burst stimulation (iTBS) to the right DLPFC for depression, although it may have benefit for both depression and anxiety.² We proposed a pilot study of accelerated resting-state functional MRI (rsfMRI)-guided iTBS for LLD and hypothesized that this intervention would result in a significant improvement in depressive and generalized anxiety disorder symptoms. <h3>Methods</h3> A total of 18 elderly patients (ages 50-79 years) with a clinical diagnosis of moderate to severe major depressive disorder of at least 6 months' duration were recruited from local clinics. Each patient underwent rsfMRI and behavioral assessment prior to treatment. Change in depressive symptoms was measured by the Inventory of Depressive Symptoms (IDS-30-C). Change in anxiety symptoms was measured by the Generalized Anxiety Disorder-7 (GAD-7). Each participant's unique stimulation target was found by preprocessing resting-state data within Analysis of Functional NeuroImages (AFNI). A connectivity map of the bilateral SgCC as defined by the Brainnetome atlas was computed and smoothed to 12mm. The maximally anti-correlated voxel within the DLPFC search region (Brodmann areas 9 and 46) was located. A 3mm sphere around the target voxel was warped to subject space and exported to a neuronavigation system. Each participant received 45 sessions of iTBS to their fMRI target with a Magventure MagPro X100 stimulator with Cool-B70 coil (see Figure 2). Stimulation parameters were: 5Hz burst frequency, 50Hz intraburst frequency, 2 second train duration, 8 second intertrain interval, 60 trains/session, 5 sessions/day, 50 minute intersession interval, for 9 days totaling 81,000 pulses. After the 15th and 45th treatments, participants repeated clinical and imaging assessments (see Figure 3). Depression scores were elicited by phone interview at one and three months after treatment. Means and standard deviations of depressive and anxious symptoms were calculated for each visit. Repeated measures analysis of variances (ANOVA) was used to assess change in symptoms over time. Changes in connectivity between nodes in the default mode network (DMN), salience network (SN), and frontoparietal network (FPN) were analyzed with independent components analysis (ICA) in Statistical Parametric Mapping (SPM). <h3>Results</h3> Mean age of the 18 patients (2M, 16F) was 63 ± 7.4 years. Mean duration of depression was 166 ± 154.4 months. Nine patients had comorbid GAD and five patients had comorbid posttraumatic stress disorder. Four of the 18 patients had previously undergone ECT. Mean depression (IDS-30-C) scores decreased throughout the study protocol (Visit 1: 39.31 ± 9.24; Visit 2: 30.74 ± 10.44; Visit 3: 20.79 ± 8.13) and remained below pre-treatment values at one month (19.84 ± 15.63) and three months (20.34 ± 14.15) (F(4,28) = 7.15, p = .0004, ??_p² = 0.5) (see Figure 4). Mean GAD-7 scores (N=18) also decreased with treatment (Visit 1: 9.95 ± 5.89; Visit 2: 9.58 ± 5.28; Visit 3: 6.53 ± 5.00) (F(2,36) = 7, p < .006, ??_p² = 0.28) (see Figure 5). Preliminary analysis of changes in connectivity from Visit 1 to Visit 3 indicate significantly increased within-network connectivity in the DMN (p = .003), and decreased between-network connectivity between the DMN and FPC (p = .011). <h3>Conclusions</h3> In this single-arm, uncontrolled pilot study, accelerated rsfMRI-guided iTBS to the right DLPFC in patients with LLD was effective in reducing both symptoms of depression and generalized anxiety. Connectivity changes in DMN and FPN network components were also observed. Future work should confirm these effects in a larger sample size and controlled study design. <h3>Secondary Rationale</h3> Reason for Late-Breaking Status: Data collection was continued beyond the regular submission deadline to attain greater statistical power to inform study conclusions. <h3>This research was funded by</h3> UNM Successful Aging Pilot Grant, Mind Research Network Center of Biomedical Research Excellence (COBRE) Pilot Grant (NIH/NIGMS 5P30GM122734-03)