Altered insula connectivity under MDMA

CONCLUSION

The present study employed a network theory approach, the Intrinsic Connectivity Contrast (ICC), to elucidate the impact of MDMA on voxelwise functional brain connectivity. The ICC allows functional connectivity analysis to proceed without the need to first geometrically define a priori ROIs, thus extending the scope of previous efforts. We observed alterations in functional connectivity between the placebo and MDMA condition in many clusters overlapping with the Harvard-Oxford Atlas (Supplementary Information S1), notably including the right insular cortex (Figure). All further analyses used the insula cluster revealed by the ICC as a seed ROI because it was selected by a prior application of the ICC, decreased right insular connectivity was strikingly absent from previous fMRI analyses applied to this data set, and the insula plays a crucial role in both subjective feelings of the bodyand anxiety. When we examined the functional network membership of the right insula cluster using seed-to-voxel functional connectivity in both the MDMA and placebo condition (Figureand), we observed a network corresponding to the salience networkupon visual inspection. A contrast between conditions (MDMA4placebo) revealed decreased connectivity between the right insula seed ROI, the bilateral anterior insula and portions of the DLPFC (ie, the middle frontal gyrus). We confirmed that decreased salience network connectivity following MDMA administration was not specific to our choice of insula ROI, by conducting an ROI-to-ROI analysis with independent coordinates reported for three hubs of the salience network: the left AI, the right AI and the dACC. Of note is the decreased connectivity between the right AI seed and left AI, but not the right AI seed and dACC, even when the search space for differences in statistically significant correlations was reduced from the whole brain to only three salience network ROIs (Figure). This suggests that altered right insula connectivity under MDMA might be specific to certain hubs of the salience network, leaving other connections between hubs functionally uncompromised., have previously demonstrated increased connectivity in the salience network (dACC and dorsolateral prefrontal cortex) in association with pre-scan anxiety, interpreted as evidence of trait anxiety coded to some degree in the functional architecture of the salience network (2007). Furthermore, alterations in insular functional connectivity following MDMA administration have never been explored despite converging evidence implicating a functional network anchored in the insula in subjective feelings of the bodyand anxiety. As such, we endeavored to investigate how differences in right insula functional connectivity under MDMA (compared with placebo) might be related to individual differences in trait anxiety. Interestingly, we found that greater decreases in right insular connectivity within a dorsolateral prefrontal region-the superior frontal gyrus-in individuals who had greater trait anxiety. The cluster of deactivation that negatively correlated with levels of trait anxiety overlapped and extended anteriorly beyond the boundaries of the salience network we functionally defined. Although speculative, it is intriguing to consider whether the observed correlation between baseline trait anxiety and MDMA-induced decreases in salience network connectivityspecifically between the right insula and superior frontal gyrus-might reflect the action of MDMA on a putative functional network coding trait anxiety. This would suggest a candidate mechanism (ie, pharmacologically attenuated salience network connectivity) for the therapeutic role of MDMA in clinical syndromes with anxiety as a core component (eg, PTSD). In line with such a possibility, meta-analyses of the functional neuroimaging of anxiety have found heightened insula activation in PTSD, Social Anxiety Disorder and Specific Phobia compared with controls, and in PTSD compared with both trauma-naïve and traumaexposed controls. Increased salience network connectivity has also been documented in a host of anxiety disorders, notably including PTSD. Furthermore, propranolol-another pharmacological agent investigated as an adjunct for PTSD therapy-decreases connectivity within the salience network. Thus, it is worth considering whether action at the salience network might be an important component of the mechanism by which MDMA operates in MDMA-assisted psychotherapy for PTSD. It should be noted that MDMA is not considered generally anxiolytic, as it has been shown to increase self-reported state anxiety. Thus, the acute therapeutic effect of MDMA may specifically target neural substrates coding trait, rather than state, anxiety (ie, the salience network). Future studies should specifically investigate this possibility, examining variations in the relationship between altered neural connectivity under MDMA and state vs trait anxiety, both acutely and in the long term. Based on the unique phenomenology of MDMA, we hypothesized that unusual bodily sensations might be related to dysregulated interoceptive processingserved by a body awareness network anchored in the right anterior insula. Thus, we investigated the correlation between self-reports of unusual bodily sensations and MDMAinduced decreases in right insula-anchored salience network connectivity. Our results revealed decreased right insular connectivity with the right SFG and anterior mid-cingulate cortex (aMCC). While decreased connectivity with the right SFG was also associated with greater levels of trait anxiety, decreased connectivity with the aMCC, a region implicated in body awareness, was uniquely related to awareness of unusual bodily sensations. As interoceptive awareness has been found to mediate the relationship between trait anxiety and the intensity of unpleasant feelings, our results raise the possibility that MDMA might attenuate unpleasant feelings-and conversely produce its characteristic positive feelings-by decreasing the integrity of a neural network important for encoding trait anxiety and supporting interoceptive awareness. It has previously been demonstrated that inductions of sad mood increase functional connectivity between the right insula and cingulate cortex, and our result supports the possibility that MDMA disrupts the functional circuitry supporting certain negative affective states by compromising the interoceptive network that supports awareness of such feelings. Furthermore, there is a developing theoretical basisto suggest that compromised functional integrity of this network might simultaneously give rise to subjective beliefs of unusual bodily sensations typical of the MDMA experience. Converging evidence of the salience network's importance in schizophreniaand altered salience network connectivity following dopaminergic challenge (ie, connectivity is increased by L-dopa and decreased by haloperidol;may initially seem at odds with our demonstration that MDMA alters insula/salience network connectivity. Human research suggests that serotonin mediates the majority of physiological and psychosocial effects of MDMA, while the effect of dopaminergic (D2) blockade was primarily limited to attenuating euphoria. However, a complex interaction between neuromodulators is likely to mediate salience network functioning. Indeed, recent work has suggested that epigenetic modification (ie, increased methylation) of the serotonin transporter gene (SLC6A4) promoter region predicts heightened insula reactivityand salience network connectivity. This phenomenon also has relevance for clinical syndromes, such as PTSD, as SLC6A4 methylation has been linked to unresolved trauma. Thus, an emerging relationship between epigenetic modulation of the serotonin transporter gene, stress/anxiety-related clinical pathology, and the salience network suggests that we should look beyond the confines of the dopaminergic system when thinking about salience network functioning. Future neuroimaging studies incorporating an acute MDMA challenge plus pretreatment with selective dopaminergic (eg, haloperidol, D2) and/or serotonergic (eg, ketanserin, 5-HT2A) receptor antagonists would help to elucidate the respective contributions of these neuromodulatory systems on the behavior of specific brain networks in both healthy individuals and those with anxiety/ trauma-related disorders (eg, PTSD). Previous resting-state studies examining functional brain networks under MDMA have not implicated the insula, and studies that have demonstrated insula involvement have not directly examined functional networks. One reason for this could be that ventromedial prefrontal cortex, hippocampus and amygdala functional connectivity with the insula is not altered under MDMA, if functional coupling between these regions is even expected in the first place. Indeed, of these anatomical regions, only the dorsal portion of the amygdala is thought to belong to the salience network. Furthermore, we did not find altered insula connectivity with any of these regions under MDMA in the present study. Roseman et al, selected 10 independent components of interest corresponding to functional network labels from the BrainMap database (6 components from the analysis were determined to be non-neural noise), but none were identified as corresponding to the salience network (2014) and only inter-network connectivity was examined. The only significant inter-network alteration under MDMA did not involve the insula. Finally, Gamma et al used positron emission tomography, and although they found a constellation of activations and deactivations under MDMA that they infer reflects a functional network, no formal network analysis approach was applied. Therefore, none of these approaches were able to identify altered insulaanchored functional network connectivity under MDMA due to their respective methodological constraints. Thus, our study represents the first synthesis of these lines of research, demonstrating decreased insula/salience network functional connectivity under MDMA and linking this to baseline levels of trait anxiety and changes in interoception. In light of recent hypotheses suggesting an 'insular view of anxiety', further understanding of how MDMA affects the insula might be crucial to elucidating the neurobiological underpinnings of re-emerging interest in MDMA as a therapeutic adjunct to psychotherapy in the treatment of anxiety disorders including PTSD, see Amoroso and Workman, 2016 for a preliminary meta-analysis).