Brain Functional Connectivity in Alcohol Use Disorder

Overview

Due to the relapsing nature of alcoholism, excessive alcohol consumption represents a significant cost to US society ($249 billion in 20101). About 64% of those entering treatment will relapse within one year. New interventions targeting the underlying brain biomarkers of relapse vulnerability hold significant promise in reducing this critical public health problem. Using resting functional magnetic resonance imaging (fMRI) we have identified brain biomarkers that support long-term abstinence and brain biomarkers that predict relapse. Our data point to specific brain biomarkers that index higher relapse vulnerability at 11 weeks of abstinence. Many individuals, however, have already relapsed by this time. It is unknown whether these biomarkers can be identified earlier during the recovery period. We need to investigate whether this biomarker of relapse vulnerability can be identified during earlier stages of abstinence. Earlier identification of this biomarker will give valuable information for timely targeted interventions (e.g. closer monitoring, longer stay in treatment program, neuromodulation), increasing the chances of maintaining abstinence. The overall objective of this study is to identify biomarkers of relapse during early abstinence (2-3 weeks of abstinence). A secondary objective is to evaluate whether non-imaging measures such as craving6 and executive function7 add value to prediction models. Findings from this proposal will provide insight into the neurobiology of relapse vulnerability that will inform new treatment strategies needed to improve treatment outcome.

BACKGROUND: Due to the relapsing nature of alcoholism, excessive alcohol consumption represents a significant cost to US society ($249 billion in 2010). About 64% of those entering treatment will relapse within one year. Development of new and improved treatments that could be personalized to maximize the chance of maintaining abstinence in the first year will require advancement in understanding the behavioral and neural mechanisms underlying vulnerability to relapse during early abstinence. New interventions targeting the underlying brain biomarkers of relapse vulnerability hold significant promise in reducing this critical public health problem. Using resting functional magnetic resonance imaging (fMRI) we have identified brain biomarkers that support long-term abstinence and brain biomarkers that predict relapse. Our cross-sectional and longitudinal findings provide evidence that higher functional connectivity (FC), particularly between nucleus accumbens (NAcc) and dorsolateral prefrontal cortex (DLPFC) is a potential brain biomarker that supports abstinence. Resting NAcc-DLPFC FC is graded depending on abstinence length, with higher FC in long-term abstinent alcoholics (7 years of abstinence) than controls and intermediate FC in short-term abstinent alcoholics (11 weeks of abstinence). Further, lower NAcc-DLPFC FC at 11-weeks of abstinence can be a predictor of subsequent relapse (with 74% accuracy). Our data point to specific brain biomarkers that index higher relapse vulnerability at 11 weeks of abstinence. Many individuals, however, have already relapsed by this time and were not included in the studies above. It is unknown whether these biomarkers can be identified earlier during the recovery period. We need to investigate whether this biomarker of relapse vulnerability can be identified during earlier stages of abstinence. Earlier identification of this biomarker will give valuable information for timely targeted interventions (e.g. closer monitoring, longer stay in treatment program, neuromodulation), increasing the chances of maintaining abstinence. PURPOSE OF THE STUDY: Our long-term goal is to use identified brain and behavioral biomarkers to facilitate the development of new, personalized treatments that will support enduring abstinence in addiction. The overall objective of this study is to identify biomarkers of relapse during early abstinence (2-3 weeks of abstinence). A secondary objective is to evaluate whether non-imaging measures such as craving and executive function add value to prediction models. Findings from this study will provide insight into the neurobiology of relapse vulnerability that will inform new treatment strategies needed to improve treatment outcome. SPECIFIC AIMS: SA 1: To evaluate prediction accuracy of FC measures in individuals with alcohol use disorder (AUD) during early (2-3 weeks) abstinence. To evaluate whether non-imaging measures (e.g. craving, executive function) add value to FC prediction models we will test prediction models using (i) only imaging variables (FC during task and rest fMRI), (ii) only non-imaging variables, and (iii) a combination of imaging and non-imaging variables as predictors of time to relapse. Hypothesis: Based on compensatory mechanisms hypothesis, we expect that subsequent abstainers will have stronger NAcc-DLPFC FC when compared to subsequent relapsers and controls. Pilot data suggests that a combination of imaging and non-imaging variables will have high prediction accuracy. SA 2. To expand our examination of NAcc-DLPFC FC beyond resting state we will use the Reversal Learning (RL) task, which involves executive functioning and reward processing. Hypothesis: Pilot data suggests that subsequent abstainers will show higher task-related NAcc FC than relapsers and controls. Exploratory Aim: We will explore whether the strength of resting NAcc-DLPFC FC changes after performing the RL task. We will collect rest fMRI data before and after RL task performance in the scanner. Hypothesis: Based on previous findings of increased task-evoked activity in these regions, we hypothesize that patients will show increased resting NAcc-DLPFC FC after performing the RL task. SIGNIFICANCE: Addiction treatment outcomes are poor. Alcohol use disorder (AUD) in the U.S. remains an important public health problem with an estimated total annual cost to society of $249.5 billion. The chronic and relapsing nature of AUD is a major obstacle to successful recovery. We lack tools to subtype in alcohol use disorder (AUD). Subtyping in clinical populations can be an effective approach to model targeted interventions. For example, targeted treatment for breast cancer patients based on individual tumor estrogen receptor typing (biomarker) has improved outcomes from a 25 to 77% survival rate; there has not been a similar improvement in outcome for AUD treatment. We need to identify biomarkers in AUD that predict relapse and guide the development and selection of treatment based on subtypes to increase the chances of achieving long-term abstinence.