Anticipatory reward processing among cocaine-dependent individuals with and without concurrent methadone-maintenance treatment: Relationship to treatment response

Abstract

Background: Cocaine dependence among opioid-dependent methadone-maintained individuals is a significant public health problem and is particularly challenging to treat. The neurobiology of this clinically complex population has not been previously assessed using fMRI.

Methods: fMRI data from cocaine-dependent, methadone-maintained (CD-MM) patients (n=24), cocaine-dependent (CD) patients (n=20) and healthy comparison (HC) participants (n=21) were acquired during monetary incentive delay task performance. All patients were scanned prior to treatment for cocaine dependence. Between-group differences in anticipatory reward and loss processing were assessed using whole-brain ANOVAs in SPM12 (pFWE<0.05). Correlations between durations of abstinence during treatment and BOLD responses within the insula and caudate were also explored.

Results: Main effects of diagnostic group, primarily involving decreased BOLD responses among CD-MM patients in comparison to HCs, were observed during anticipatory reward and loss processing within regions of posterior cingulate cortex, precuneus, inferior frontal gyrus and dorsolateral prefrontal cortex. BOLD responses within the right caudate were negatively associated with percentage of cocaine-negative urines during treatment among CD-MM patients, but not among non-methadone-maintained CD patients.

Conclusions: These data suggest neurofunctional differences that may be related to treatment outcomes for behavioral therapies between cocaine-dependent individuals with and without methadone-maintenance treatment. These findings may relate to differences in treatment efficacies and to the elevated relapse rates observed in methadone-maintained populations.

Keywords: Cocaine; Heroin; Medication; Methadone; Opioids; Polysubstance use.

Figure 1

Findings from whole-brain ANOVA and groupwise analyses of BOLD responses during anticipatory reward and loss processing Panel A (top) shows main effects of diagnostic group on BOLD responses during anticipatory reward/loss processing (voxel-level p=.001, pFWE<.05, df=63). Panel A (bottom) shows mean±standard error BOLD signals for diagnostic groups separately within selected clusters identified in the whole-brain ANOVA. Panel B (top) shows findings from whole-brain t-tests of healthy comparison (HC) versus cocaine-dependent, methadone-maintained (CD-MM) participants (voxel-level p=.001, pFWE<.05, df=43). Blue – light blue indicates reduced activity among CD-MM participants. Panel B (bottom) shows associations between BOLD signal within the PCC/precuneus clusters identified in the whole-brain t-tests and daily methadone dose among CD-MM participants. IFG=inferior frontal gyrus; DLPFC= dorsolateral prefrontal cortex; PCC=posterior cingulate cortex

Figure 2

Association between caudate activity during anticipatory win processing and abstinence during treatment among methadone-maintained, cocaine-dependent individuals Scatterplot shows the negative associations between pretreatment right caudate (defined anatomically using the automatic anatomical labeling (AAL) atlas in SPM12) activation during anticipatory win processing and percentage of cocaine-negative urines during treatment for cocaine dependence among cocaine-dependent, methadone-maintained patients. No significant associations between durations of abstinence during treatment and caudate activity were observed among non-methadone-maintained, cocaine-dependent patients. Statistical comparisons indicated significant differences in r-values between patient groups (Z=−1.82, p=.034). CBT=computer-based cognitive behavioral therapy; TAU=treatment-as-usual