2022 Articles

Dynamic Resting-State Network Biomarkers of Antidepressant Treatment Response

Kaiser, Roselinde H.; Chase, Henry W.; Phillips, Mary L.; Deckersbach, Thilo; Parsey, Ramin V.; Fava, Maurizio; McGrath, Patrick J.; Weissman, Myrna M.; Oquendo, Maria A.; McInnis, Melvin G.; Carmody, Thomas; Cooper, Crystal M.; Trivedi, Madhukar H.; Pizzagalli, Diego A.

Background: Delivery of effective antidepressant treatment has been hampered by a lack of objective tools for predicting or monitoring treatment response. This study aimed to address this gap, by testing novel dynamic resting-state functional network markers of antidepressant response. Methods: The Establishing Moderators and Biosignatures of Antidepressant Response in Journal Pre-proof Clinical Care (EMBARC) study randomized adults with major depressive disorder to eight weeks of either sertraline or placebo, and depression severity was evaluated longitudinally. Participants completed resting-state neuroimaging pre-treatment and again after one week of treatment (n=259 eligible for analyses). Co-activation pattern analyses identified recurrent whole-brain states of spatial co-activation, and computed time spent in each state for each participant as the main dynamic measure. Multilevel modeling estimated the associations between pre-treatment network dynamics and sertraline response, and between early (pretreatment to one-week) changes in network dynamics and sertraline response. Results: Dynamic network markers of early sertraline response included increased time in network states consistent with canonical default and salience networks, together with decreased time in network states characterized by coactivation of cingulate and ventral limbic or temporal regions. The effect of sertraline on depression recovery was mediated by these dynamic network changes. In contrast, early changes in dynamic functioning of corticolimbic and frontoinsulardefault networks were related to patterns of symptom recovery common across treatment groups. Conclusions: Dynamic resting-state markers of early antidepressant response or general recovery may assist development of clinical tools for monitoring and predicting effective intervention.