Therapeutic DBS for OCD Suppresses the Default Mode Network

Abstract

Deep brain stimulation (DBS) of the anterior limb of the internal capsule (ALIC) is a circuit-based treatment for severe, refractory obsessive-compulsive disorder (OCD). The therapeutic effects of DBS are hypothesized to be mediated by direct modulation of a distributed cortico-striato-thalmo-cortical network underlying OCD symptoms. However, the exact underlying mechanism by which DBS exerts its therapeutic effects still remains unclear. In five participants receiving DBS for severe, refractory OCD (3 responders, 2 non-responders), we conducted a DBS On/Off cycling paradigm during the acquisition of functional MRI (23 fMRI runs) to determine the network effects of stimulation across a variety of bipolar configurations. We also performed tractography using diffusion-weighted imaging (DWI) to relate the functional impact of DBS to the underlying structural connectivity between active stimulation contacts and functional brain networks. We found that therapeutic DBS had a distributed effect, suppressing BOLD activity within regions such as the orbitofrontal cortex, dorsomedial prefrontal cortex, and subthalamic nuclei compared to non-therapeutic configurations. Many of the regions suppressed by therapeutic DBS were components of the default mode network (DMN). Moreover, the estimated stimulation field from the therapeutic configurations exhibited significant structural connectivity to core nodes of the DMN. Based upon these findings, we hypothesize that the suppression of the DMN by ALIC DBS is mediated by interruption of communication through structural white matter connections surrounding the DBS active contacts.

Keywords: DTI; OCD; anterior limb of the internal capsule; deep brain stimulation; default mode network; fMRI; tractography.

FIGURE 1

Structural and functional characterization of DBS configurations. (A) Reconstruction of DBS leads for five subjects. Leads for separate patients are in distinct colors. Blue leads indicate treatment responders; orange leads indicate non‐responders. Therapeutic electrode contacts shown in red. (B) Example of tractography derived from diffusion imaging seeding the estimated volume of tissue activation for a therapeutic bipolar contact configuration for a single representative subject. (C) Design of DBS cycling On versus Off paradigm during fMRI acquisition for different stimulation configurations. (D) fMRI BOLD changes with DBS On versus Off for same subject and configuration in (B). Suppression of BOLD with DBS On versus. Off is depicted in blue while activation with DBS On versus Off is depicted in red. Color bar indicates percentage change in BOLD signal. p < 0.05; OLSQ.

FIGURE 2

BOLD response differences between therapeutic and non‐therapeutic DBS. Group comparison using linear mixed effects model of BOLD response between DBS On and Off in therapeutic (n = 6 runs, 3 subjects) versus nontherapeutic (n = 17 runs, 5 subjects) DBS configurations. Activations are in red and suppressions in blue. Color bar indicates percentage change in BOLD signal. dmPFC, Dorsomedial prefrontal cortex; OFC, Orbitofrontal cortex; PCUN, Precuneus; PCC, Posterior cingulate cortex; STN, Subthalamic nucleus. p < 0.05; LME.

FIGURE 3

Network Impact of Therapeutic and Non‐Therapeutic DBS. Comparison of average BOLD changes within canonical resting‐state networks for therapeutic (blue) and non‐therapeutic (red) configurations. ***p < 5.0 × 10⁻³, *p < 0.05; permutation test (one‐sided with Bonferroni correction).

FIGURE 4

Structural Connectivity with Therapeutic Electrode Configurations. (A) Percentage of total streamlines from the estimated volume of activated tissue to functional network parcels for the therapeutic configurations (n = 6 runs, 3 subjects). (B) Resting state networks ordered based on increasing fraction of streamline counts from the estimated volume of activated tissue for the therapeutic configurations. *p < 0.05, permutation test (one‐sided).