Abstract

Corticostriatal circuits are central to reward processing and reinforcement learning functions that often become dysregulated in substance use disorder (SUD) and drive compulsive drug use. Human neuroimaging research seeking to identify how corticostriatal circuits become altered in SUD has primarily focused on evaluating connectivity between cortex-striatum node-pairs. Yet, striatal nodes receive appreciable input from many cortical nodes, and the morphological and electrophysiological properties of striatal nodes dictate that combinational features of their multivariate "connectivity profiles" shape their activity more so than any individual cortical node. Here, we introduce an approach for quantifying and statistically evaluating different types of multivariate connectivity profile configuration differences (i.e., aggregate divergence, rank order arrangement, and entropy shift) that may reflect different forms of circuit plasticity, and apply it to nicotine dependent smokers (n=46) as an exemplar SUD. Foremost, we find evidence of significant connectivity profile misconfiguration throughout much of the striatum, suggesting that prior findings of abnormal connections between individual striatal-cortical node-pairs may only represent the "tip of the iceberg" of corticostriatal circuit alteration in nicotine dependence. Moreover, we find that dorsolateral and ventromedial striatum display distinct types of connectivity profile misconfiguration. Whereas dorsolateral striatum almost exclusively displays abnormal rank order arrangement that is present in both the nicotine sated and acutely abstinent states - indicative of a "trait" misconfiguration - ventromedial striatum almost exclusively displays abnormal aggregate divergence that only manifests during acute abstinence - indicative of a "state" misconfiguration. Further, we identify a unique striatal site in the right caudal ventral putamen that displays multiple forms of connectivity profile misconfiguration, where connections with cognitive processing cortical areas overtake those with motor/premotor control cortical areas as the strongest in the connectivity profile, and acute abstinence significantly strengthens this abnormal arrangement. Moreover, the interactive magnitude of these misconfigurations during acute abstinence is significantly linked to dependence severity. Collectively, the present findings underscore the need for increased examination of connectivity profile misconfigurations as a mechanism of SUD etiology and as a potential guide for identifying therapeutic intervention targets.