Brain structural effects of treatments for depression and biomarkers of response: a systematic review of neuroimaging studies. Issue 2 (20th January 2020)

• Record Type: Journal Article
• Authors: Enneking, Verena; Leehr, Elisabeth J.; Dannlowski, Udo; Redlich, Ronny
• Is Part Of: Psychological medicine. Volume 50:Issue 2(2020)
• Page Start: 187
• Page End: 209
• Publication Date: 2020-01-20
• Subjects: Antidepressant treatment, -- biomarker, -- cognitive behavioural therapy, -- depression, -- electroconvulsive therapy, -- MRI, -- neuroimaging, -- pharmacotherapy, -- psychotherapy, -- review; Psychiatry -- Periodicals; Medicine and psychology -- Periodicals; Clinical psychology -- Periodicals; 616.89
• DOI: 10.1017/S0033291719003660
• Languages: English
• ISSNs: 0033-2917

Abstract

Abstract: Antidepressive pharmacotherapy (AD), electroconvulsive therapy (ECT) and cognitive behavioural therapy (CBT) are effective treatments for major depressive disorder. With our review, we aim to provide a systematic overview of neuroimaging studies that investigate the effects of AD, ECT and CBT on brain grey matter volume (GMV) and biomarkers associated with response. After a systematic database research on PubMed, we included 50 studies using magnetic resonance imaging and investigating (1) changes in GMV, (2) pre-treatment GMV biomarkers associated with response, or (3) the accuracy of predictions of response to AD, ECT or CBT based on baseline GMV data. The strongest evidence for brain structural changes was found for ECT, showing volume increases within the temporal lobe and subcortical structures – such as the hippocampus–amygdala complex, anterior cingulate cortex (ACC) and striatum. For AD, the evidence is heterogeneous as only 4 of 11 studies reported significant changes in GMV. The results are not sufficient in order to draw conclusions about the structural brain effects of CBT. The findings show consistently that higher pre-treatment ACC volume is associated with response to AD, ECT and CBT. An association of higher pre-treatment hippocampal volume and response has only been reported for AD. Machine learning approaches based on pre-treatment whole brain patterns reach accuracies of 64–90% for predictions of AD or ECT response on the individual patient level. The findings underline the potential of brain biomarkers for the implementation in clinical practice as an additive feature within the process of treatment selection.