A serotonergic biobehavioral signature differentiates cocaine use disorder participants administered mirtazapine

Abstract

Cocaine use disorder (CUD) patients display heterogenous symptoms and unforeseeable responses to available treatment approaches, highlighting the need to identify objective, accessible biobehavioral signatures to predict clinical trial success in this population. In the present experiments, we employed a task-based behavioral and pharmacogenetic-fMRI approach to address this gap. Craving, an intense desire to take cocaine, can be evoked by exposure to cocaine-associated stimuli which can trigger relapse during attempted recovery. Attentional bias towards cocaine-associated words is linked to enhanced effective connectivity (EC) from the anterior cingulate cortex (ACC) to hippocampus in CUD participants, an observation which was replicated in a new cohort of participants in the present studies. Serotonin regulates attentional bias to cocaine and the serotonergic antagonist mirtazapine decreased activated EC associated with attentional bias, with greater effectiveness in those CUD participants carrying the wild-type 5-HT_2CR gene relative to a 5-HT_2CR single nucleotide polymorphism (rs6318). These data suggest that the wild-type 5-HT_2CR is necessary for the efficacy of mirtazapine to decrease activated EC in CUD participants and that mirtazapine may serve as an abstinence enhancer to mitigate brain substrates of craving in response to cocaine-associated stimuli in participants with this pharmacogenetic descriptor. These results are distinctive in outlining a richer "fingerprint" of the complex neurocircuitry, behavior and pharmacogenetics profile of CUD participants which may provide insight into success of future medications development projects.

Fig. 1. The initially fully connected model, visualized with the BrainNet Viewer (http://www.nitrc.org/projects/bnv/) [71].

The six DCM nodes are illustrated as gold spheres, and the dark red lines with arrows represent the endogenous connectivities. The All-Words driving input (green line with arrow) affected all six nodes, and the modulator (e.g., mirtazapine CW-modulator) affected all endogenous connectivities. Whether the driving input effect to a node (or the modulatory effect on an EC) was different from zero was determined based on the posterior probability (see Results). ACC, anterior cingulate cortex; MOFC, medial orbitofrontal cortex; INS, insula; HIPP, hippocampus; PCC, posterior cingulate cortex; PUT, putamen. The left side of this figure aligns with the left brain hemisphere.

Fig. 2. The brain activation used to constrain the a priori-selected L-ACC (in medial left view) and R-MOFC (in medial right view) DCM node is depicted using the Surf Ice software (https://www.nitrc.org/plugins/mwiki/index.php/surfice:MainPage, posted by Dr. Chris Rorden).

The brain activation clusters shown were identified by a SPM second level one-sample t-test analysis for the placebo scan and for all participants (CW minus NW > 0), with cluster-defining threshold t = 2.4 and uncorrected two-tailed cluster level p < 0.05. Scale on the color bar represents voxel t values. The nodes (spheres) are larger than the exact ones for demonstration purpose.

Fig. 3. The brain activation used to constrain the a priori-selected R-putamen (in anterior cut view), L-INS (in superior cut view), L-PCC (in medial left view), and R-hippocampus (in medial right view) DCM nodes.

These brain activations were found by a SPM second level t-test comparison of the two genotype groups, with cluster-defining threshold t = 2.4 and uncorrected two-tailed cluster level p < 0.05. This analysis found that, during the mirtazapine scan, the participants with the wild-type HTR2Cexhibited greater cocaine word-elicited activation than those with the HTR2C SNP. Scale on the color bar represents voxel t values. The nodes (spheres) are larger than the exact ones for demonstration purpose.

Fig. 4. The mean CW modulatory change of the L-ACC → R-hippocampus (HIPP) EC is illustrated for the HTR2C SNP subgroup (red line), wild-type HTR2C subgroup (green line), and all participants (combined HTR2C SNP and wild-type HTR2C subgroups) (blue line) upon placebo and mirtazapine scans.

The differences between the mirtazapine and placebo scans (mirtazapine minus placebo) were all reliable (PP = 1) with 0.2028 Hz, −0.5327 Hz, and −0.2390 Hz for the HTR2C SNP, wild-type HTR2C, and all participants, respectively [, , –74].