The overarching aim is to compare the effects of ventral capsule/ventral striatum (VC/VS) and subthalamic nucleus (STN) deep brain stimulation (DBS) in the same participants. Investigators will test the hypothesis, grounded in cognitive neuroscience, that DBS at both sites is better than either site alone for treating the symptom dimensions of obsessive compulsive disorder (OCD). Specifically, Investigators will employ novel cognitive paradigms and neurophysiological measures of cortical synaptic function to test the hypothesis that VS/VC and STN DBS have different mechanisms of action and that alleviation of OCD symptoms is mediated by improvement in mood/anxiety with VS/VC DBS and by directly interrupting obsessions and compulsions with STN DBS. Investigators will additionally determine whether adjunctive cognitive behavioural therapy (CBT) enhances the response to DBS by providing the cognitive and behavioural skills to optimise symptom management and daily function.

The VS/VC and STN are both 'stations' in the neural circuitry thought to be dysfunctional in OCD. The pre-eminent neurobiological model of OCD implicates abnormalities of orbitofrontal cortex (OFC), anterior cingulate cortex (ACC) and ventral striatum. These structures are integral to limbic cortico-striatal-thalamo-cortical circuitry whereby cortical outputs synapse successively in ventral striatum, ventral pallidum and mediodorsal nucleus of thalamus before projecting back to cortex; within this system, the pallidum projects to thalamus either directly or indirectly via the subthalamic nucleus. It is therefore possible that the observed clinical improvement in OCD with DBS is due to high frequency electrical stimulation disrupting abnormal activity in the limbic circuit and that this can be obtained at either of the two target sites. However, the ventral striatum and STN are thought to have separate functions, important in different ways for modulating information processing in the cortico-striatal-thalamo-cortical circuitry essential for the control of behaviour. Thus, although both VS/VC and STN DBS affect the same neural circuit, their mechanism of action may be quite different and stimulation at each of these locations may have unique effects on the symptom dimensions of OCD.

Evidence for different clinical effects comes from findings that VS/VC DBS produced early and sometimes dramatic mood elevation and anxiety reduction prior to improvement in obsessions and compulsions whereas STN DBS, while diminishing obsessions and compulsions, had no effect on emotion. This observation is relevant because mood and anxiety are significant symptom dimensions in OCD and increases or decreases in the severity of anxiety or depression are generally accompanied by parallel changes in the severity of obsessions/compulsions. The ventral striatum, as well as being the first target of OFC/ACC cortical outputs, receives inputs from amygdala and midbrain dopaminergic neurones, which together provide the emotional and motivational impetus for goal directed behaviour. The ventral striatum is therefore in a unique position, via its involvement in limbic cortico-striato-thalamo-cortical circuitry, to influence the cognitive and motor processing in parallel circuits which is preparatory for action selection and ultimately manifest in thoughts and actions. Denys and colleagues proposed that VS/VC DBS influences OCD symptoms via its effect on anxiety and mood. Taking this further, Investigators hypothesize that stimulation of ventral striatal neurones influences reinforcement learning and emotion processing thereby improving mood and reducing anxiety and resulting in a decrease in the intensity of obsessions and compulsions.

STN DBS on the other hand may have a more direct effect on obsessions and compulsions. The limbic STN, like the ventral striatum, is in a unique position to influence behavioural outputs of cortico-striatal-thalamo-cortical circuitry. This is because the STN receives a direct projection from right inferior frontal cortex which acts as a 'hyperdirect pathway', activation of which overrides neural processing in the direct cortico-striato-pallido-thalamic pathway. This results in the inhibition of ongoing, planned acts and thoughts. Investigators therefore hypothesise that, via this inhibitory mechanism, stimulation of STN neurones enables OCD patients to interrupt their repetitive thoughts and actions resulting in a decrease in obsessions and compulsions.

Accordingly, DBS at both sites, because of different mechanisms of action, is predicted to be clinically more effective than stimulation of either site alone. It also follows that if there are two mechanisms mediating OCD improvement, this should be evident in different patterns of cognitive and neurophysiological effects following DBS. In the proposed study Investigators will test these predictions clinically and experimentally by using paradigms designed to distinguish the two hypothetical mechanisms of action.

The main aim of CBT is to enable OCD patients to obtain greater control over behaviour so that they can inhibit their repetitive thoughts and acts and switch to more meaningful alternatives. One abnormal mechanism contributing to such cognitive inflexibility is thought to be blunting of the value of positive and negative reinforcement normally used to guide responses and dependent on intact OFC-striatal neural circuitry. Dysfunction of this cognitive control system has been shown in first-degree clinically unaffected relatives of OCD patients thus confirming this mechanism as a neurobiological phenotype of OCD 15.It is therefore hypothesised that DBS of VS/VC will improve the processing of reinforcing stimuli and that this will result not only in elevation of mood and reduction in anxiety but also in increased cognitive flexibility. Investigators will test this by measuring the effectiveness of VS/VC DBS on clinical measures of mood, anxiety, obsessions and compulsions and relate these to changes in reinforcement learning and emotional processing using cognitive tasks sensitive to OCD which measure the ability to respond appropriately to reinforcing contingencies.

Attentional processes are also thought to contribute to impaired behavioural control in OCD. Being able to stop or inhibit planned acts when environmental circumstances change depends on the function of the right inferior cortical-STN hyperdirect pathway and can be operationalised by measuring the stop signal reaction time (SSRT). The finding of specific prolongation of SSRT in OCD patients and their unaffected first-degree relatives suggests that this cognitive abnormality is also a phenotype of OCD. Although the neural substrate of this mechanism does not involve the limbic cortico-striato-pallido-thalamic circuit, recent evidence suggests that neural systems outside this circuitry, involving the right inferior frontal cortex, are impaired in OCD. Investigators therefore hypothesise that STN DBS allows inhibition of repetitive thoughts and acts in OCD by facilitating the action of the hyperdirect pathway from the right inferior frontal cortex to STN. Investigators will test this by measuring the effectiveness of STN DBS on obsessions and compulsions and relate these to changes in SSRT and performance on other tasks sensitive to right inferior frontal cortex function which measure the ability to inhibit pre-potent tendencies.

OCD symptoms and cognitive impairment may be mediated by abnormal cortical excitability due to aberrant long term depression (LTD) synaptic plasticity and/or reduced gamma-aminobutyric acid (GABA) mediated intracortical inhibition. Transcranial magnetic stimulation (TMS) can be used to measure these neurophysiological processes safely in patients undergoing DBS and both abnormalities have been shown to normalise following DBS in other disorders. Investigators will use TMS to assess these neural processes before and after DBS in OCD patients and test for whether there are differential changes from the two DBS sites and whether combined VS/VC and STN DBS produces more vigorous improvements compatible with our hypothesis concerning the superior clinical effects of this condition. Investigators will measure intracortical inhibition not only at rest but also during the anticipation of reinforcement, previously shown to affect motor cortex excitability and which Investigators predict will be specifically affected by VS/VC DBS.

Inclusion Criteria:

1. Participants must have undergone intensive treatment and have demonstrable treatment resistance as defined by:

   1. At least two serotonin reuptake inhibitors (SRI) for a minimum of 12 weeks at optimal British National Formulary (BNF) doses.
   2. Augmentation of SRI treatment with antipsychotic drugs administered at maximally tolerated doses or by extending the selective serotonin reuptake inhibitor (SSRI) dose beyond BNF limits.
   3. Two trials of CBT of at least 10 hours.
   4. Failed inpatient treatment for OCD.

2. Participants must be older than 20 years of age.

3. Confirmation of a primary diagnosis of obsessive compulsive disorder (International Classification of Diseases (ICD 10) F42.0-F42.9).

4. Duration of OCD of at least 10 years.

5. At least 2 years of unremitting symptoms despite intensive psychopharmacological and psychological treatment or failure to sustain, over a 3 month period, a response to inpatient psychological treatment by at least 33% with accompanying optimised pharmacological therapy.

6. A minimum score of 32 on Yale Brown Obsessive Compulsive Scale (YBOCS) thus constituting profound illness and a maximum score of 50 on the Diagnostic and Statistical Manual of Mental Disorders (DSM IV) General Assessment of Function Scale (GAF).

7. Ability to provide sustained informed consent.

Exclusion Criteria:

1. Current diagnosis of Substance misuse (ICD10 F10-F19)
2. Current diagnosis of Organic brain syndrome (ICD10 F00-F09)
3. Current diagnosis of Adult personality disorder (ICD10 F60-F69)
4. Current diagnosis of Pervasive developmental disorder (ICD10 F84)
5. Current diagnosis of Schizophrenia (ICD10 F20-F29)
6. Current diagnosis of Bipolar disorder (ICD 10 F30-31)
7. Contraindications to neurosurgery
8. Pregnancy