Beck A1, Seo S2, Mohr J2, Wüstenberg T1, Heinz A1, Obermayer K2
1 Department of Psychiatry and Psychotherapy, Charité – Universitätsmedizin Berlin, Campus Mitte, Germany
2 Neural Information Processing Group, Department of Electrical Engineering and Computer Science, Technische Universität Berlin, and Bernstein Center for Computational Neuroscience Berlin, Germany
One major characteristic of alcohol use disorders is an increased attention allocation towards alcohol-related stimuli as well as a heightened neuronal response associated with this so-called cue reactivity. Here, a growing body of imaging studies focussing on alcohol-dependent patients have identified brain areas robustly activated by alcohol-associated cues, e.g. the medial prefrontal cortex (MPFC) and the adjacent anterior cingulate cortex (ACC) as well as the ventral striatum, including the nucleus accumbens. Moreover, these heightened alcohol cue-induced brain responses were associated with an increased prospective risk of relapse.
However, most of these studies are based on group comparisons. More recently, there is a novel approach using different mathematical strategies (such as machine learning techniques) to individually predict treatment outcome (i.e. the individual relapse risk). Indeed, there is evidence that an individual prediction of future relapse from imaging measurement is possible and even outperforms prediction from clinical measurements.
This individual prediction of treatment outcome based on neuroimaging endophenotypes may help to target specific interventions via identifying different risk groups. Patients showing neuronal patterns associated with a lower relapse risk might best profit from treatment approaches focusing on rehabilitation and stabilization. Contrary, in subjects with a heightened relapse risk, treatment approaches based on the extinction of conditioned behaviour might work most effectively. Here, training to habitually reject alcohol cues might lower neuronal cue reactivity as well as reduce the risk to relapse.