Exploring fMRI-neurofeedback to reduce craving-related brain activity in people with cannabis use disorder: A feasibility study

Introduction Cannabis Use Disorder (CUD) can come with intense cravings, which can trigger compulsive use/relapse, underscored by anterior cingulate cortex (ACC) hyperactivity during cannabis cue-reactivity functional neuroimaging (fMRI) tasks. ACC activity can be reduced during real-time fMRI-based neurofeedback (fMRI-neurofeedback) in participants who use substances other than cannabis, but remains untested in CUD. We investigated the feasibility of fMRI-neurofeedback to reduce ACC activity and subjective craving in 10 participants with moderate-to-severe CUD. Methods and materials A cue-reactivity fMRI task identified participants’ voxels-of-interest (VOI) within the ACC (ACC_VOI). During two identical runs of 7 Tesla fMRI-neurofeedback, participants were instructed to use self-selected strategies to reduce ACC_VOI activity while receiving real-time feedback via a digital craving bar (ACC_VOI activity) when viewing cannabis images (down-regulation), and to simply look at neutral images (non-regulation). We examined the change in real-time percentage signal change (PSC) of the ACC_VOI during down-regulation versus non-regulation , and post-hoc (offline) whole-brain activation changes. Results Analysis of real-time data revealed that down-regulation versus non-regulation was associated with reductions in ACC_VOI PSC (p ' 0.05). Further, whole-brain analysis reported down-regulation versus non-regulation was associated with lower activity in the ACC, posterior insula, and hippocampus; and greater activity in the anterior insula and middle frontal gyrus (p ' 0.05, FWE-corrected). Subjective craving did not change pre-to-post fMRI-neurofeedback and did not correlate with ACC_VOI PSC or whole-brain brain activity (p ' 0.05). Conclusion fMRI-neurofeedback may reduce the activity of regions associated with craving in CUD. This effect appeared to be dissociable from subjective craving. These observations need to be confirmed in larger samples with placebo-controlled conditions.