Brain Mechanisms of Reinforcement Learning in Obsessive-Compulsive Disorder

Introduction: Obsessive-compulsive disorder is a chronic and highly prevalent neuropsychiatric disorder, with treatment lacking efficacy in a significant proportion of cases. Thus, a more comprehensive understanding of the brain mechanisms underlying OCD is needed. Dysfunction in corticostriatal circuits, which underlie learning of new actions, has been suggested by several studies. However, there is no consensus about how such dysfunction may generate obsessive-compulsive symptoms, nor about the therapeutic potential of invasive or non-invasive neuromodulation techniques targeting the brain areas that are involved. Objectives: To review neuroimaging and behavioral findings related to learning of actions in OCD, and its potential therapeutic relevance. Methods: Non-systematic literature review, using the keywords "obsessive-compulsive disorder", "neuroimaging", "corticostriatothalamic dysfunction model", "instrumental learning", "reinforcement learning", "deep brain stimulation" e "transcranial magnetic stimulation". Results: In reinforcement learning tasks, OCD patients have a deficit in goal-directed or model-based learning, with a tendency towards habitual behavior or model-free learning. On the other hand, the most consistent neuroimaging findings in OCD are low volume and hyperactivity of the orbitofrontal cortex, as well as caudate hyperactivity - and both of these areas are important for reinforcement learning. In fact, rodent studies have demonstrated that the OFC, in particular the lateral OFC, plays an important role in the shift between habitual and goal-directed behavior, suggesting that this may be an important mechanism underlying OCD. Conclusions: Notwithstanding all of the progress in this area, new reinforcement learning tasks are needed for improved understanding of the role of the OFC in OCD, and to explore the behavioral, and potentially also therapeutic effects of neuromodulation of this brain area.