The late positive potential predicts contingency management outcomes for cocaine use disorder.

Contingency management (CM) relies on motivation and reward processing and is among the most effective treatments available for cocaine use disorder (CUD), yet success rates do not meet desired clinical efficacy. To improve treatment-individual fit, we investigated the late positive potential (LPP), a measure of incentive salience to cues, as a predictor of treatment response. Treatment-seeking adults with CUD (N = 58) completed an EEG task where they viewed pleasant, unpleasant, cocaine, and neutral images. Participants received CM and counseling for four weeks. Urine was collected thrice weekly. We defined incentive salience to cues by A) cluster analyses on LPP amplitudes across the range of emotional images, B) cocaine-LPP relative to pleasant-LPP, and C) LPP to individual categories (pleasant, unpleasant, cocaine), relative to neutral. Treatment outcomes were: 1) response status (abstinent vs. non-abstinent at week four), and 2) Treatment Effectiveness Score (TES), a count of total negative urines. Bayesian generalized linear modeling was used to predict treatment response as a function of LPP amplitude, quantifying the posterior probability of an effect. Those with a greater LPP to cocaine images when controlling for neutral were more likely to respond to CM treatment (OR=1.93, PP=99.4 %) and had a higher TES (IRR=1.22, PP=90.9 %). These results held for the other analyses (cluster analysis and cocaine controlling for pleasant). LPPs to pleasant and unpleasant cues were not consistently associated with CM outcomes. These results suggest that heightened cocaine reactivity may represent a neuroaffective state that could indicate greater propensity for CM treatment response in CUD.