Dissociating executive prefrontal substrates following capsulotomy in obsessive-compulsive disorder

Obsessive-compulsive disorder (OCD) is characterized by widespread executive function impairments linked to disrupted fronto-striatal circuits. Ablative therapies such as capsulotomy show promise efficacy in treatment-refractory OCD, yet their effects on executive neural substrates remain poorly understood. This study explores how OCD and capsulotomy influence executive prefrontal function. Twenty-three post-capsulotomy OCD patients, thirty OCD controls, and thirty-two health controls (HC) were recruited in the study. Post-capsulotomy patients were recruited at least 6 months following surgery to allow for post-operative stabilization. Participants completed three executive function tasks assessing distinct cognitive domains: the Extra-Dimensional Intra-Dimensional (EDID) task testing set-shifting, the N-back task testing working memory and flanker task measuring conflict processing. The EDID and N-back tasks were administered during concurrent task-based fMRI, while the flanker task provided only behavioral measures. In the EDID task, OCD capsulotomy patients demonstrated greater post-error flexibility towards ID versus ED shifting relative to both OCD controls and healthy controls, despite exhibiting more ED errors. OCD controls showed less shifts after errors compared to HC, which was not revealed in the capsulotomy group. OCD patients showed less neural differentiation between ED versus ID in lateral and mesial prefrontal regions relative to HC. Capsulotomy was associated with decreased pre-supplementary motor area activity to ED shifts compared to HC, indicating impairments in neural reactivity to ED shifting. No evidence supported the influence of capsulotomy in the n-back or the flanker task. Specifically, both OCD groups showed impaired working memory performance at high memory load along with dysfunction in the whole frontoparietal network. In the flanker task, impairment in dissociating congruent and incongruent conditions were found in both OCD groups but not in the HC group. These findings suggest capsulotomy specifically remediates error-monitoring and behavioral flexibility through mesial prefrontal remodeling, while leaving working memory and conflict processing deficits intact. This dissociation implies that cognitive flexibility may represent a state-dependent process amenable to intervention, whereas working memory and conflict processing impairments may reflect trait markers of OCD vulnerability. These results have clinical implications on relative safety profile of capsulotomy and patient selection for precision neuromodulation.