1 Integrated Program in Neuroscience, McGill University, Montreal, Quebec, Canada.
2 Sunnybrook Research Institute, University of Toronto, Toronto, Ontario, Canada.
3 Department of Psychiatry, McGill University, Montreal, Quebec, Canada.
4 Department of Neurology & Neurosurgery, Montreal Neurological Institute, McGill University, Montreal, Quebec, Canada.
5 McConnell Brain Imaging Centre, Montreal Neurological Institute, Montreal, Quebec, Canada.
6 Department of Psychology, McGill University, Montreal, Quebec, Canada.
7 Center for Studies in Behavioral Neurobiology, Concordia University, Montreal, Quebec, Canada.

Addictions are thought to be fostered by the emergence of poorly regulated mesocorticolimbic responses to drug-related cues. The development and persistence of these responses might be promoted by altered glutamate transmission, including changes to type 5 metabotropic glutamate receptors (mGluR5s). Unknown, however, is when these changes arise and whether the mGluR5 and mesocorticolimbic alterations are related. To investigate, non-dependent cocaine polydrug users and cocaine-naïve healthy controls underwent a positron emission tomography scan (15 cocaine users and 14 healthy controls) with [¹¹ C]ABP688, and a functional magnetic resonance imaging scan (15/group) while watching videos depicting activities with and without cocaine use. For some drug videos, participants were instructed to use a cognitive strategy to lower craving. Both groups exhibited drug cue-induced mesocorticolimbic activations and these were larger in the cocaine polydrug users than healthy controls during the session's second half. During the cognitive regulation trials, the cocaine users' corticostriatal responses were reduced. [¹¹ C]ABP688 binding was unaltered in cocaine users, relative to healthy controls, but post hoc analyses found reductions in those with 75 or more lifetime cocaine use sessions. Finally, among cocaine users (n = 12), individual differences in prefrontal [¹¹ C]ABP688 binding were associated with midbrain and limbic region activations during the regulation trials. Together, these preliminary findings raise the possibility that (i) recreational polydrug cocaine users show biased brain processes towards cocaine-related cues and (ii) repeated cocaine use can lower cortical mGluR5 levels, diminishing the ability to regulate drug cue responses. These alterations might promote susceptibility to addiction and identify early intervention targets.