(57) A novel pathway-activation model of deep brain stimulation for obsessive-compulsive disorder to study anxiogenic effects

Introduction: Deep brain stimulation (DBS) of the anterior limb of the internal capsule (ALIC) is an effective therapy for severe obsessive-compulsive disorder (OCD). However, 30-40% of implanted patients remain treatment-refractory due to imprecise targeting¹. Current computational models of DBS can help improve targeting but often lack patient specificity, relying on normalized diffusion tractography and simplified isotropic electric field properties. Another challenge is identifying stimulation parameters to optimize benefits while minimizing adverse effects such as anxiety². To address these limitations, we developed a patient-specific modeling approach using a recent diffusion tractography algorithm to generate a retrospective individualized biophysical model of ALIC DBS for OCD. Utilizing this model, we conducted a preliminary study to elucidate white matter pathways within the ALIC associated with stimulation-induced anxiogenic effects.

Methods: A pathway activation model was generated for one OCD patient implanted with bilateral ALIC DBS leads (Medtronic 3391). The model analyzed four monopolar configurations (6 mA, 150 µs, 130 Hz) computationally and clinically. Clinical notes identified four key settings: three linked to mood improvements alone, while one linked to both mood improvements and anxiogenic effects. The modeling system included: 1) an individualized 3D reconstruction of ALIC pathways3 and DBS electrode positions from post-operative CT registered to pre-operative 7T MRI, 2) finite element models of the DBS-induced electric field using tissue conductivity from diffusion MRI, and 3) multicompartment axon models. Pathway activation was predicted based on myelinated axon responses to the applied electric field.

Results: Pathway activation results showed relatively low percent activation of ALIC white matter connected to the dorsolateral prefrontal cortex (PFC), dorsomedial PFC, ventrolateral PFC, and dorsal anterior cingulate for the anxiogenic setting compared to non-anxiogenic settings. However, low activation of these pathways when ramping up stimulation through other contacts did not induce anxiogenic effects.

Conclusion: This study introduces a novel patient-specific computational model of ALIC DBS for OCD, incorporating individualized imaging data for a more detailed and biologically accurate representation. The data suggest that other pathways, not currently modeled, near the anterior commissure may underlie the anxiogenic effects in this case study. The study shows how ALIC DBS is likely to activate a broader range of pathways beyond those connected with the prefrontal cortex and further shows how the patient-specific pathway activation modeling approach, which does not require a normalized brain atlas, can be useful for understanding clinical outcomes of ALIC DBS for OCD patients.