Background: Predicting relapse vulnerability can inform level-of-care and personalized substance use treatment. Few reliable predictors of relapse risk have been identified from traditional clinical, psychosocial, and demographic variables. However, recent neuroimaging findings highlight the potential prognostic import of brain-based signals, indexing the degree to which neural systems have been perturbed by addiction. These proposed "neuromarkers" forecast the likelihood, severity, and timing of relapse but the reliability and generalizability of such effects remains to be established.
Methods: Activation likelihood estimation was used to conduct a preliminary quantitative, coordinate-based meta-analysis of the addiction neuroprediction literature; specifically, studies wherein baseline measures of regional cerebral blood flow were prospectively associated with substance use treatment outcomes. Consensus patterns of activation associated with relapse vulnerability (greater activation predicts poorer outcomes) versus resilience (greater activation predicts improved outcomes) were specifically investigated.
Results: Twenty-four eligible studies yielded 134 foci, representing 923 subjects. Consensus activation was identified in right putamen and claustrum (p < .05, cluster-corrected) in relation to positive and negative treatment outcomes - likely reflecting variability in measurement context (e.g., task, sample characteristics) across datasets. A single cluster in rostral-ventral anterior cingulate cortex (rACC) was associated with relapse resilience, specifically (p < .05, cluster-corrected); no significant vulnerability-related clusters were identified.
Conclusions: Right putamen activation has been associated with relapse vulnerability and resilience, while increased baseline rACC activation has been consistently associated with improved treatment outcomes. Methodological heterogeneity within the existing literature, however, limits firm conclusions and future work will be necessary to confirm and clarify these results.
Keywords: Activation likelihood estimation; Addiction; Meta-analysis; Neuroimaging; PET; Relapse; Substance use disorders; Treatment outcome; fMRI.
Published by Elsevier B.V.