Out of control: loosing oneself in compulsivity

Because of my interest in addiction and treatment possibilities for addiction, I started working at the AMC with a project about deep brain stimulation (DBS) as treatment for therapy resistant addiction in 2009. DBS was already successfully used in obsessive-compulsive disorder (OCD) at the AMC at that time. In some of the OCD-patients unintended changes in behavior were observed after DBS, such as increased impulsivity and hypomania. Moreover, addicted patients with OCD stopped smoking or drinking after DBS treatment. These case reports together with promising animal research about DBS in addiction indicated that DBS could be a promising treatment for addiction and therefore a study was initiated to investigated the efficacy of DBS in addiction.
Unfortunately the recruitment for the DBS addiction study was difficult. This opened up the possibility to investigate other but related questions. First of all, it was unclear which neural changes were induced by DBS in OCD patients. Furthermore, I wanted to explore the concept of compulsivity. Compulsivity is related to the feeling of loss of control over repetitive destructive behavior and plays a key role in OCD and addiction. However, there is no consensus about the definition of compulsivity, hampering research in this field. Last, I was interested in risk aversion in OCD. Since OCD patients are more inclined to prevent or avoid danger, risk aversion may play a role in the pathophysiology of OCD.

Definition of compulsive behavior
The essence of compulsive behavior is the feeling that one 'has to' perform a specific act. In the definition that we proposed in chapter 2, compulsive behavior results from an urge to improve one's internal state, which is modulated by the evaluation of one's current internal state and the anticipation of a negative outcome. For instance you will be more inclined to compulsive behavior when you feel stressed or when you think something bad might happen. This definition assumes affective drives at the core of compulsive behavior since its aimed at reducing or preventing a negative state. This focus on affect in compulsive behavior differs from a recent theory that views compulsive behaviors as maladaptive habits resulting from the repeated association between an action and stimulus; this stimulus-action association will become more automatic through repetition and the outcome of the behavior and motivation will play an decreasingly smaller role. We also believe that habituation can play a role in the development of compulsive behavior but we do not exclude affective (motivational) aspects. In chapter 2 we pose the hypothesis that repeated coupling between an affective state and an action can result in motivational habits which in turn can lead to compulsive behavior. Therefore we suggest an interactive role between affective drives and habituation in compulsive behavior. We believe that this combination renders the behavior insensitive to its negative consequences and makes compulsive behavior so difficult to change.  
This definition needs further investigation to see whether it accurately encompasses compulsive behaviors observed in OCD, addiction and other disorders. With a clear definition of compulsive behavior, it will be possible to develop questionnaires and neuropsychological tests to investigate a possible common neuropsychological endophenotype underlying compulsive behavior across disorders.  

OCD patients are not more risk-avoidant than healthy controls
In chapter 3 we show that OCD patients do not differ from healthy controls in their attitude towards risk. However, there are differences between subgroups of OCD patients: OCD patients with doubt or checking symptoms are more risk-averse than other OCD patients (chapter 3). Moreover, there is an opposite relation between neural activity in the insula and risk aversion during risk processing in OCD patients compared to healthy controls: a positive correlation between insula activity and risk aversion in patients versus a negative correlation in healthy controls. These results indicate that for a specific subgroup of OCD patients risk aversion, associated with increased insula activation during risk processing, may contribute to the persistence of the disorder.
Our hypothesis suggesting that the insula signals urgency to avoid risk in OCD patients and to take risk in healthy controls is consistent with the notion that the insula is sensitive to salient signals. Also in addiction the insula has been proposed to play a key role in motivation of behavior and salience attribution. In sum, the insula seems a particularly interesting brain region to investigate for its involvement in the persistence of maladaptive behavior in both OCD and addiction.

DBS in psychiatry: efficacy and behavioral side effects
The application of DBS in psychiatry seems promising but remains investigational, specifically for major depressive disorder, Tourette syndrome and addiction. Only for OCD there are multiple controlled trails in larger groups of patients and for OCD the estimated response rate is about 50% (). The most commonly observed side effect is hypomania, which seems to happen more often in the ventral striatal area (17.8%) including the NAc, and is more prevalent in men (chapter 5). In chapter 6 we report two OCD patients with nucleus accumbens chapter 4(NAc) DBS, showing that high electric potential may in some OCD patients lead to impulsive behavior. Still little is known about the effects of NAc DBS on impulsivity and more research is needed. Both increases in impulsivity and hypomania can increase the risk for relapse in addiction. Therefore extra caution is warranted for signs of impulsivity and hypomania in these patients when increasing stimulation amplitude.

DBS normalizes frontostriatal activity in OCD 
Chapter 7 describes the effects of NAc DBS on the frontostriatal circuitry in OCD patients. We measured brain activity with functional magnetic resonance imaging (fMRI) and electroencephalography (EEG) when patients were stable on DBS treatment and after one week without DBS stimulation. The fMRI experiments showed that DBS normalized activity in the nucleus accumbens (around the DBS target region) during reward anticipation and functional connectivity between the nucleus accumbens and the prefrontal cortex. This change in frontostriatal connectivity was correlated with DBS-related changes in OCD symptoms. Furthermore, EEG measurements showed that excessive slow EEG fluctuations in response to symptom provoking stimuli were reduced by DBS. Our results suggest that DBS interferes with disruptive unwanted behavioral patterns in OCD by interrupting a pathological frontostriatal loop. These findings are also relevant for DBS in addiction since abnormalities of the frontostriatal network overlap between OCD and addiction.

DBS for addiction appears promising but is not feasible in practice 
We conclude in chapter 8 that the nucleus accumbens is the most promising DBS target area for addiction, with the most studies and most consistent encouraging results, based on a review of the animal and human literature. Therefore we started a DBS pilot-study investigating NAc DBS as treatment for therapy resistant heroin and cocaine addiction in 2010. The positive results of the first patient in this study are described in chapter 9. In addition, EEG recordings in the NAc show that the DBS contact points most responsive to addiction related stimuli are also the most effective contact points for reducing drug use during the course of DBS. This indicates that results from deep brain EEG recordings may help to shorten the optimization period of finding the right stimulation parameters at least with regard to the most promising stimulation target(s).
In chapter 10 we describe the serious problems we encountered with the inclusion of patients for the DBS addiction pilot study. After three years of recruitment, only 23 patients were referred resulting in the inclusion of only two patients. A comparison with a similar OCD DBS trial showed that the number of addiction referrals was relatively small and that many more addiction patients left the study before surgery. A structured telephone interview with the patients that had shown initial interest but aborted the trial before surgery revealed that the invasiveness of the procedure was the main reason for non-participation. Possible explanations for the differences in participation rates between the OCD and the addiction group include: (1) differences between the perceived burden of the disease (more consistently high in OCD vs more fluctuating in addiction) (2) differences in perceived pathogenesis of the disorder: the debate whether addiction is a (brain) disease is more persistent and intense than for OCD (3) more social and physical problems in addiction that increase the barrier to apply for and comply with DBS treatment.

Clinical implementations and conclusion
Currently there is interest in new ways of classifying psychopathology based on dimensions of observable behavior and neurobiological measures that transcend current diagnostic (DSM) categories. Compulsivity could be such a dimension of behavior that transcends diagnostic boundaries and fits well into this new vision on psychopathology. If compulsivity is indeed a consistent behavioral dimension with similar associated neural mechanisms across different disorders,  we can find new therapeutic avenues to specifically target compulsivity and its neural mechanisms. For these purposes existing and new brain modulation techniques such as DBS will become increasingly more relevant. However, during our research we encountered striking differences in the demand for and attitude towards DBS in OCD and addiction. Therefore, to fully understand compulsive behaviors and to effectively treat them it remains important not to lose sight of  the context in which these behaviors take place and how patients view these new treatments and not exclusively focus on behavioral symptom dimensions.