Using the Neuroscience of Fear Extinction for Anxiety Reduction

The best available treatment for social anxiety disorder is exposure therapy; however, 25% of socially anxious patients do not respond to an adequate course of exposure therapy and it is unclear why. Prior attempts to identify non-responders using clinical and demographic features have been largely unsuccessful, highlighting the need to examine constructs that are more closely tied to the mechanism of treatment (i.e., extinction learning and recall) and the organ of dysfunction (i.e., the brain). The neurobiology of extinction learning and recall is well understood from decades of animal and pre-clinical laboratory work, which has highlighted the importance of the amygdala, dorsal anterior cingulate cortex (dACC), and ventromedial prefrontal cortex (vmPFC). However, this knowledge has not been leveraged to improve exposure therapy response, despite the assumption that response relies on extinction learning and its successful recall. Thus, a critical long-term goal is to improve exposure therapy response by tailoring therapy based on the neurobiological profile of each patient. This project addresses that goal by directly linking neurobiological profiles of extinction learning and recall with clinical symptoms and therapy response. A major objective of this project is therefore to build a mechanistic predictive model of exposure therapy response based on the neurobiology of extinction learning and recall. To accomplish this goal, the investigators will recruit 80 adults with social anxiety disorder who will be randomized to 10 sessions of exposure-focused therapy or waitlist. The primary clinical outcome measure will be the Liebowitz Social Anxiety Scale (LSAS), a validated and widely used measure that assesses anxiety and avoidance symptoms. Pre-therapy, participants will also undergo an experimental protocol for extinction learning and recall. Participants will first view a neutral abstract image repeatedly paired with a loud aversive noise, and another image that is never paired (fear acquisition phase). Following this, participants will view the same images without aversive consequences (extinction learning phase). Better extinction learning will be defined as greater reductions in skin conductance within the extinction learning phase. Brain activation during extinction learning will be assessed in the amygdala, dACC, and vmPFC. Finally, participants will view the same images without aversive consequences one week later (extinction recall phase). Better extinction recall will be defined as less skin conductance during extinction recall relative to fear acquisition. The central hypothesis is that greater activation in the vmPFC during extinction learning will predict both extinction recall and therapy response over and above symptom severity.