Toward Precision Medicine for Smoking Cessation: Developing a Neuroimaging-Based Classification Algorithm to Identify Smokers at Higher Risk for Relapse. Issue 8 (14th November 2019)

• Record Type: Journal Article
• Authors: Frank, David W; Cinciripini, Paul M; Deweese, Menton M; Karam-Hage, Maher; Kypriotakis, George; Lerman, Caryn; Robinson, Jason D; Tyndale, Rachel F; Vidrine, Damon J; Versace, Francesco
• Is Part Of: Nicotine & tobacco research. Volume 22:Issue 8(2020)
• Page Start: 1277
• Page End: 1284
• Publication Date: 2019-11-14
• Subjects: Nicotine -- Periodicals; Tobacco -- Research -- Periodicals; Tobacco habit -- Periodicals; Nicotine -- Periodicals; Tobacco -- Periodicals; Smoking -- Periodicals; 613.85
• DOI: 10.1093/ntr/ntz211
• Languages: English
• ISSNs: 1462-2203

Abstract

Abstract: Introduction: By improving our understanding of the neurobiological mechanisms underlying addiction, neuroimaging research is helping to identify new targets for personalized treatment interventions. When trying to quit, smokers with larger electrophysiological responses to cigarette-related, compared with pleasant, stimuli ("C > P") are more likely to relapse than smokers with the opposite brain reactivity profile ("P > C"). Aim and Method: The goal was to (1) build a classification algorithm to identify smokers characterized by P > C or C > P neuroaffective profiles and (2) validate the algorithm's classification outcomes in an independent data set where we assessed both smokers' electrophysiological responses at baseline and smoking abstinence during a quit attempt. We built the classification algorithm applying discriminant function analysis on the event-related potentials evoked by emotional images in 180 smokers. Results: The predictive validity of the classifier showed promise in an independent data set that included new data from 177 smokers interested in quitting; the algorithm classified 111 smokers as P > C and 66 as C > P. The overall abstinence rate was low; 15 individuals (8.5% of the sample) achieved CO-verified 12-month abstinence. Although individuals classified as P > C were nearly 2.5 times more likely to be abstinent than smokers classified as C > P (12 vs. 3, or 11% vs. 4.5%), this result was nonsignificant, preliminary, and in need of confirmation in larger trials. Conclusion: These results suggest that psychophysiological techniques have the potential to advance our knowledge of the neurobiological underpinnings of nicotine addiction and improve clinical applications. However, larger sample sizes are necessary to reliably assess the predictive ability of our algorithm. Implications: We assessed the clinical relevance of a neuroimaging-based classification algorithm on an independent sample of smokers enrolled in a smoking cessation trial and found those with the tendency to attribute more relevance to rewards than cues were nearly 2.5 times more likely to be abstinent than smokers with the opposite brain reactivity profile (11% vs. 4.5%). Although this result was not statistically significant, it suggests our neuroimaging-based classification algorithm can potentially contribute to the development of new precision medicine interventions aimed at treating substance use disorders. Regardless, these findings are still preliminary and in need of confirmation in larger trials.