Reinforcement Learning and Obsessive-compulsive Disorder: Exploring the Role of the Orbitofrontal Cortex

INTRODUCTION:

Obsessive-compulsive disorder (OCD) is the fourth most common psychiatric disorder, characterized by recurrent and intrusive obsessive thoughts and/or compulsions that are distressing, time-consuming , and a cause of significant impairment. OCD has a lifetime prevalence of 1-3%, and is associated with a reduced life expectancy due to natural and unnatural causes. Current treatments, based on a combination of pharmacotherapy and psychotherapy, are ineffective in at least 50% of cases, with rTMS of the prefrontal cortex having recently received FDA clearance as an adjunctive treatment. Yet, OCD pathophysiology remains largely unknown , and it is expected that a better understanding of OCD 's biological mechanisms would contribute to a better prevention and treatment of the disorder.

Current neurobiological models for the disorder highlight the role of cortico-striatal-thalamo-cortical (CSTC) loops and hyperactivity of the orbitofrontal cortex (OFC), and it has been hypothesized that corticostriatal dysfunction in OCD may occur as an end-consequence of immune and inflammatory dysregulation that leads to structural and functional changes in these circuits. This pathophysiological mechanism was first proposed based on the association between group A β-hemolytic streptococcal (GAS) infection and the abrupt onset of OCD in children, described by Swedo and colleagues (1998) as a distinctive nosologic entity named Pediatric Autoimmune Neuropsychiatric Disorders Associated with Streptococcal infection (PANDAS). This post-infectious neuropsychiatric syndrome was hypothesized to occur due to a cross-reaction between anti-streptococcal antibodies and the basal ganglia, causing CSTC circuit dysfunction. In addition to GAS, other pathogens have since been suggested as contributing to the development of OCD, albeit no definitive causal associations have been established. Interestingly, childhood adversity, an environmental risk factor for OCD, has also been associated with chronic immune dysfunction in adulthood, namely with increased C-reactive protein, interleukin-6, and tumor necrosis factor (TNF)-α, as well as with increased volume of the right OFC in adult patients with OCD.

While the specific psychopathological consequences of OFC dysfunction in OCD remain unclear, in recent years research in rodents has shown that the OFC, particularly the lateral OFC, is crucial in controlling transitions between automatic, repetitive stimulus-response behaviors (such as habitual behavior in operant-learning tasks, or model-free RL), and behaviors that reflect the acquisition, by the agent, of a predictive model of the consequences of each action (such as goal-directed behavior, or model-based RL). Evidence suggests that patients with OCD may have an abnormal bias towards habitual behavior during operant learning tasks, with subjects having a higher difficulty in developing model-based reinforcement learning strategies. This bias has been proposed to favor the emergence of ritualized compulsive behaviors typical of OCD, and is regarded as a marker of corticostriatal dysfunction that is central to the disorder's pathophysiology.

In this case-control study, the investigators propose an integrative experimental approach to test if structural, functional , and metabolic changes in corticostriatal circuits of patients with OCD correlate with performance in an RL task, as well as with peripheral blood markers of immune dysfunction and early-life environmental risk factors previously implicated in the disorder (e.g., infectious insults and childhood adversities).

STUDY OBJECTIVES:

In this study, the investigators aim to compare patients with OCD and healthy volunteers to assess the association between OCD and:

• Behavioral performance on a RL task;
• Markers of immune dysfunction/autoimmunity in peripheral blood;
• Exposure to proposed environmental risk factors, including infectious agents and childhood adversities;
• Structural and functional brain connectivity patterns, particularly between the OFC and striatum;
• Concentrations of brain metabolites.

The investigators also intend to explore associations of the outcome variables with clinical variables such as obsessive-compulsive symptom severity and age of onset.

Additionally, for patients enrolled in this study who will undergo rTMS for clinical purposes at our clinical center, the investigators aim to assess if there is an association between the behavioral, neuroimaging, and immunological variables and treatment response. Specifically, the investigators will test 1) if post-treatment changes in behavioral and neuroimaging measures correlate with treatment response, and 2) if baseline behavioral, neuroimaging, and immune measures predict response to treatment. Specifically, the investigators will test:

• if rTMS treatment will alter performance in the RL task and if such changes correlate with response to treatment;
• if rTMS modulates organizational patterns in large-scale neural networks and in the functional and structural connectivity between OFC and striatum, and if these changes correlate with response to treatment or to changes in performance of the RL task;
• if rTMS leads to changes in metabolite concentrations in the OFC and striatum and if these are associated with response to treatment or with changes in performance of the RL task;
• if environmental risk factors and baseline peripheral markers of immune dysfunction, brain structural and functional connectivity correlates, and baseline concentration of metabolites in the OFC and striatum predict the outcome of rTMS both in terms of OCD symptom change and in terms of RL task performance.

PROCEDURES:

For this study, the investigators will recruit 115 age- and sex-matched subjects with OCD and 115 control volunteers. After signing informed consents, participants will complete a thorough clinical assessment, including confirmation of psychiatric diagnoses using the appropriate modules of the Structured Clinical Interview for DSM-5, and rating of obsessive-compulsive symptoms using the Yale-Brown Obsessive-Compulsive Scale II, the gold-standard structured clinical interview for OCD assessment. Other clinical variables of interest will be assessed using the following instruments:

• Mini International Neuropsychiatric Interview (M.I.N.I.) - Portuguese or English version. The version used for this study does not include the module for Psychopathy.
• Yale-Brown Obsessive-Compulsive Scale-II - Portuguese or English version
• State-Trait Anxiety Inventory - Portuguese or English version
• Beck Depression Inventory-II - Portuguese or English version
• Childhood Trauma Questionnaire - Portuguese or English version
• Perceived Stress Scale - Portuguese or English version
• Barratt Impulsiveness Scale - Portuguese or English version
• Edinburgh Handedness Inventory - Portuguese or English version After completing the clinical assessment, participants will proceed to blood collection for measurement of peripheral blood markers of immune dysfunction/autoimmunity. This will be followed by performance of a RL task on a computer, and lastly by a magnetic resonance imaging session. The protocol for the latter will comprise a T1 sequence for structural assessment, resting-state BOLD images (rs-fMRI) of the whole brain, Diffusion Weighted Imaging (DWI) to measure diffusivity of water molecules within the brain, and brain tissue concentration of metabolites using Magnetic Resonance Spectroscopy (MRS) analysis of two voxels, one in the left OFC and the other in the left caudate.

For the subgroup of patients enrolled in the study that will undergo rTMS for clinical purposes at our Champalimaud Clinical Programme, the investigators will propose a follow-up visit after completion of the treatment. In this visit, participants will be asked to repeat the clinical assessment, RL task , and the MRI protocol as described before. The clinical TMS protocol comprises 30 sessions of excitatory rTMS (20 Hz) over the medial prefrontal cortex, as defined by international guidelines, using a protocol that includes individualized symptom provocation before each stimulation session, to elicit a moderate level of obsessional distress, as reported by patients. This subgroup will also complete all the procedures of experiment 1 prior to initiating the rTMS treatment protocol. After the last session of rTMS treatment, up to one week after the treatment has ended, these participants will repeat the RL task and the MRI protocol as described for baseline.

DATA COLLECTION:

• Clinical assessment: Clinical interview and psychometric instruments (self-report questionnaires and semi-structured interviews);
• Behavioral assessment: Participants will perform a RL task in the form of a computer game;
• Blood sample collection: A blood sample will be collected, processed on the same day, and stored in a -80ºC freezer;
• MRI protocol: MRI sequences will be acquired on a 3T Phillips scanner with a 32-channel head coil. The imaging protocol will include T1-weighted sequence using a 3-dimensional fast field echo (FFE); resting-state BOLD echo-planar imaging (EPI); DTI EPI sequence with b values of 0, 1000 , and 2000 s/mm2. For each b value, 64 equally spaced gradient directions will be sampled. Magnetic Resonance Spectroscopy (MRS) with a Point-Resolved Spectroscopy (MEGA-PRESS) pulse sequence will be used to acquire signals from hydrogen protons to measure spectra from metabolites in the OFC and striatum. The total scan time will be around 50 minutes.