Transdiagnostic Brain-Behavior Profiling to Enhance Cognitive Behavioral Therapy Response

Major Depressive Disorder (MDD) and generalized Social Anxiety Disorder (gSAD) are pervasive major public health problems. These disorders are characterized by emotion dysregulation, an inability or inefficiency to regulate negative and positive affect as reflected in common and disorder-specific symptoms (e.g., attentional bias to negative stimuli, excessive/inappropriate negative thoughts, hyperarousal, anhedonia, emotional blunting). Such dysregulation is believed to result from an imbalance between top-down 'emotion regulating' (ER) frontal nodes central in inhibitory control of bottom-up subcortical 'emotion-generating' (EG) nodes in a Fronto-Limbic Affect Regulation and Emotional Salience (FLARES) network. Therefore, successful treatment would be expected to 'normalize' neurofunctional disturbances in the FLARES network, which can be measured with fMRI and more distal units of brain function -- event-related potentials (ERPs) from electroencephalography, startle potentiation from electromyography (EMG), neurocognitive performance, and use of regulation strategies in daily life via self-report. The overarching objective of the proposed study is to understand how, when, and where CBT works and for whom to tailor treatment to improve clinical outcome. Without precisely identified "targets" and "predictors" of change, CBT response will continue to be unpredictably varied with few achieving meaningful clinical improvement placing them at risk for relapse and recurrence. Our proposal builds on published data from our lab and others and Preliminary Data which shows FLARES function, as assayed with fMRI, ERPs, EMG, and behaviors, is sensitive to change following CBT. Importantly, both baseline fMRI and non-fMRI units of brain-behavioral measures predict CBT response better than baseline clinical measures. Such knowledge can lead to more precise interventions aimed at capitalizing on 'strengths' or improving 'deficits' that may each exist before CBT and/or explain why CBT does not work for some patients. The dual development of fMRI ('mechanistic') and non-fMRI ('pragmatic') predictors and indices of therapeutic change is aimed at advancing precision medicine while increasing the clinical utility of 'biomarkers' in the outpatient setting. With this objective, we propose to employ well-validated paradigms to test ER and EG in the context of negative stimuli, reward processes, and fear systems in MDD and gSAD to delineate common and disorder-specific mechanisms of change and predictors of CBT outcome. We will enroll 200 patients: 100 MDD (without comorbid gSAD), 100 gSAD (without comorbid MDD) and randomize them to 12 weeks of manualized CBT or 12 weeks of 'placebo' psychotherapy (supportive therapy) (1:1 ratio). Multiple units of FLARES function will be collected in all patients before (Week 0), during (midway/Week 6) and after treatment (Week 12) to ascertain CBT 'dose' effects, and in 40 healthy controls for comparison. Pre-CBT predictors based on binary (responder/non-responder status) and continuous (extent of change) outcomes will be examined midway (Week 6), immediately after treatment (Week 12), and at 6-month follow-up.