Predicting craving and relapse using neurobehavioral markers of goals and habits in opioid use disorder

PROJECT SUMMARY
Opioid overdose-related deaths have claimed over 100,000 lives in 2022, necessitating a focus on biomarkers
of recovery and relapse risk. As addiction progresses, the motivational control over drug use is posited to shift
from being goal-directed (i.e., dependent on the drug's hedonic value and the ventral striatum and prefrontal
cortex) to habitual (i.e., cue-triggered, via the dorsal striatum). Rooted in preclinical studies, the evidence for this
cortico-striatal basis in human drug addiction is scarce. The available computational evidence suggests goal-
directed control impairments that are associated with shorter abstinence and relapse in alcohol and
methamphetamine use disorders. No study to date has inspected the neurobiology of motivational control, nor
its putative fluctuations during recovery, in opioid use disorder. This K99/R00 application intersects naturalistic,
longitudinal, and computational methods to inspect the contribution of goal-driven behaviors to opioid (heroin
and fentanyl) addiction severity and recovery, encompassing craving and relapse. In the K99 phase, Aim 1 will
track motivational control daily (using individual smartphones) for eight weeks during inpatient heroin addiction
treatment (with and without fentanyl use), offering the candidate the opportunity to train in naturalistic and
computational methods. Here, we expect impairments to be evident in early treatment but to normalize as a
function of abstinence. Aim 2 will use the expected dynamic shifts in motivational control to inform changes in
craving and relapse, tracked daily by ecological momentary assessments. The candidate will learn to harness
longitudinal and predictive models for translational insights. We expect goal-directed control decreases to predict
elevations in craving and relapse events. Aim 3 in the R00 phase will use a novel functional magnetic resonance
imaging task that estimates choice under drug and nondrug contexts to naturalistically inspect motivational
control cortico-striatal engagement in an independent sample of inpatients with opioid use disorder (inclusive of
heroin and fentanyl). The candidate will gain expertise in translational neuroimaging methods. We expect lower
goal-directed control to be associated with lower prefrontal cortex activity, evident especially in a drug cue
context, in the patient group compared to controls. This proposal will allow the candidate to gain proficiency in
the naturalistic, longitudinal, and computational study of drug addiction. The candidate will be guided by an
excellent mentorship team: Drs. Rita Z. Goldstein (drug addiction neurobiology and its naturalistic study),
Nathaniel Daw (computational modeling of motivational control), Katie Witkiewitz (longitudinal assessments of
addiction treatment outcomes), Dr. Nelly Alia-Klein (clinical assessment and treatment of drug addiction), Hung-
Mo Lin (consultant; biostatistical support), and Vincenzo Fiore (consultant; model-based neuroimaging), and the
state-of-the-art resources at the Icahn School of Medicine at Mount Sinai. The innovative proposal and the
carefully crafted training plan will propel the candidate to an impactful and independent research career,
ultimately yielding insights into improving treatment outcomes for people with drug addiction.

Grant Number: 1K99DA060948-01A1
NIH Institute/Center: NIH
Principal Investigator: Ahmet Ceceli