Neuroimaging Biomarkers at Baseline Predict Electroconvulsive Therapy Overall Clinical Response in Depression: A Systematic Review.

Abstract

Major depressive disorder is a frequent and disabling disease and can be treated with antidepressant drugs. When faced with severe or resistant major depressive disorder, however, psychiatrists may resort to electroconvulsive therapy (ECT). Although very effective, the response falls short of 100%. A recent meta-analysis established clinical and biological predictive factors of the response to ECT. We decided to explore neuroimaging biomarkers that could be predictors of the ECT response. We performed a systematic literature review up to January 1, 2018, using a Boolean combination of MeSH terms. We included 19 studies matching our inclusion criteria. Lower hippocampal, increased amygdala, and subgenual cingulate gyrus volumes were predictive for a better ECT response. Functional magnetic resonance imaging also found that the connectivity between the dorsolateral prefrontal cortex and posterior default-mode network is predictive of increased efficacy. Conversely, deep white matter hyperintensities in basal ganglia and Virchow-Robin spaces, medial temporal atrophy, ratio of left superior frontal to left rostral middle frontal cortical thickness, cingulate isthmus thickness asymmetry, and a wide range of gray and white matter anomalies were predictive for a poorer response. Our review addresses the positive or negative predictive value of neuroimaging biomarkers for the ECT response, indispensable in a personalized medicine dynamic. These data could reduce the risk of nonresponders or resistance with earlier effective management. It might also help researchers elucidate the complex pathophysiology of depressive disorders and the functioning of ECT.