Resting-state connectivity of the amygdala predicts response to cognitive behavioral therapy in obsessive compulsive disorder

Highlights

• We investigated voxel-wise whole-brain resting-state connectivity in OCD patients.

• Functional data was acquired prior to an 8-week cognitive behavioral therapy (CBT).

• Connectivity of the basolateral amygdala predicted the outcome of CBT.

Abstract

Background

Obsessive–compulsive disorder (OCD) is a psychiatric disorder which is characterized by recurrent intrusive thoughts (obsessions) and ritualized, repetitive behaviors or mental acts (compulsions). The gold standard for the treatment of OCD is cognitive behavioral therapy (CBT) with exposure and response prevention. This is the first study exploring the predictive value of resting-state functional connectivity for the outcome of CBT.

Methods

We assessed whole-brain resting-state functional connectivity in a group of 17 un-medicated OCD inpatients prior to CBT compared to 19 healthy controls using functional magnetic resonance imaging. The graph theoretical metric degree centrality served as indicator for altered voxel-wise whole-brain functional connectivity. The relative change in the Yale-Brown Obsessive Compulsive Scale (YBOCS) score was used to evaluate treatment outcome.

Results

The degree centrality of the right basolateral nuclei group of the amygdala was positively correlated with the response to subsequent CBT. OCD patients showed a lower degree centrality of the superficial amygdala (bilateral).

Conclusions

Our results suggest that two different sub-regions of the amygdala and their respective neural networks are affected in OCD: the superficial amygdala and networks related to evaluation of reinforcers and risk anticipation and the basolateral amygdala which is implicated in fear processing. The diminished CBT response in patients showing a lower degree centrality of the basolateral amygdala reflects a deficient fear circuit in these patients which may impact fear extinction as a core mechanism of exposure-based CBT.

Introduction

Obsessive–compulsive disorder (OCD) is characterized by recurrent intrusive thoughts (obsessions) and ritualized, repetitive behaviors or mental acts (compulsions) (Abramowitz, Taylor, & McKay, 2009; Cooper, 2001). Cognitive behavioral therapy (CBT) with exposure and response prevention is an effective first-line therapy for OCD (Rosa-Alcazar, Sanchez-Meca, Gomez-Conesa, & Marin-Martinez, 2008). Convergent experimental evidence from neuroimaging studies associated dysfunctional orbitofronto-striatal loops with the pathophysiology of OCD. Early PET studies at rest and during symptom provocation consistently found hypermetabolism in the orbitofrontal cortex and the striatum (Baxter et al., 1987, Baxter et al., 1990, Baxter et al., 1988, Rauch et al., 1994, Simon et al., 2010; Swedo et al., 1989). These findings suggested dysfunctional cortico–striato–thalamo–cortical (CSTC) pathways to be involved in this disorder which led to the fronto–striatal model of OCD (Chamberlain et al., 2008, Menzies et al., 2008, Saxena et al., 1999, Saxena et al., 1998, Saxena and Rauch, 2000). Recent findings indicate the involvement of brain regions as the amygdala, hippocampus, anterior cingulate, dorsolateral prefrontal and the parietal cortex as reviewed by Menzies et al. (2008) and Milad and Rauch (2012). Furthermore, the CSTC model of OCD disregards the role of the amygdala and hippocampus and their interaction with the frontal cortex in mediating fear and anxiety (Milad et al., 2013, Milad and Rauch, 2012). The role of the amygdala and the hippocampus in fear extinction is of particular interest as CBT with exposure to fear-provoking stimuli and situations may be viewed as a clinical application of fear extinction (Milad and Quirk, 2012). The amygdala has also been implicated in other OCD findings as risk aversion, an exaggerated response to threat, and a diminished response to reward (Admon et al., 2012).

Few studies and to our knowledge only one whole-brain approach (Olatunji et al., 2014) investigated the predictive value of neuroimaging data for the outcome of CBT. There are no respective studies using functional connectivity for outcome prediction. Olatunji et al. (2014) found that the activity in the temporal pole and amygdala during symptom provocation were associated with better outcome. Brody et al. (1998) showed in an early ROI-based PET-study that higher OFC metabolism predicted a better therapy response. Saxena et al. (2009) investigated the effects of a brief intensive CBT on brain glucose metabolism in OCD and found an increase in right dorsal anterior cingulate cortex activity (dACC) which correlated with the extent of improvement in OCD symptoms. The authors related this observation to the role of the dACC in reappraisal and suppression of negative emotions. Fullana et al. (2013) found that the cortical thickness of the rostral anterior cingulate cortex (ACC) was inversely associated with CBT outcome. Schwartz, Stoessel, Baxter, Martin, & Phelps (1996) found that behavior therapy responders had significant bilateral decreases in caudate glucose metabolic rates that were greater than those seen in poor responders to treatment. Using Proton-MRS, we could show that the concentration of myo-Inositol, a compound in the lateral OFC (but not in the rostral ACC and striatum) was highly predictive for the outcome of subsequent CBT in OCD (Zurowski et al., 2012). O’Neill et al. (2013) found metabolic effects of CBT treatment in the (pregenual) ACC, particularly in terms of glutamatergic and N-acetyl compounds.

Shin et al. (2014) investigated the effects of pharmacological treatment (selective serotonin reuptake inhibitors; SSRI) on topological properties of brain networks in OCD. Prior to treatment, OCD patients showed a decreased global clustering and as a consequence a decreased small-world parameter, which was restored after sixteen weeks of treatment. Zhang et al. (2011), however, found an increased global clustering in OCD patients. One notable difference between the two analyses was the choice of the network nodes. This illustrates the importance of voxel-based whole-brain approaches and the comparison of ROI- and voxel-based analyses.

In a previous study (Göttlich, Krämer, Kordon, Hohagen, & Zurowski, 2014b) we used the same data set to investigate differences in brain network organization between healthy controls and OCD patient at the scale of network modules. OCD patients showed decreased limbic and increased fronto–parietal connectivity. Here, we investigated if resting-state functional connectivity allows predicting the success of CBT applying a graph theoretical approach. We used the degree centrality as a marker for functional connectivity. The degree centrality of a node (here a voxel) within a network is defined as the number of connections to the rest of the network (Bullmore and Sporns, 2009). The degree centrality has been used successfully as a marker for altered resting state functional connectivity in Alzheimer’s disease (Buckner et al., 2009), Parkinson’s disease (Göttlich et al., 2013), bilateral vestibular failure (Göttlich et al., 2014a) and Obsessive Compulsive Disorder (Beucke et al., 2013, Hou et al., 2014). Beucke et al. (2013) found an abnormally high degree centrality in the orbitofrontal cortex which positively correlated with symptom severity.

To our knowledge, this is the first study to evaluate the predictive value of resting-state functional connectivity for therapy outcome applying CBT. Therefore, we followed an exploratory approach studying whole-brain voxel-wise functional connectivity.