Neural Correlates of ECT Response and Relapse in Major Depressive Disorder

Lay Summary

Can imaging reveal how ECT improves major depression and identify those at risk for relapse?

This study will employ resting state fMRI to determine how electroconvulsive therapy exerts a therapeutic effect in patients with major depressive disorder, and to identify biomarkers that predict which patients are likely to relapse following ECT treatment.

Major depressive disorder (MDD) tends to be resistant to current therapies. Electroconvulsive therapy (ECT) can be effective for patients with this form of depression. The success rate is impressive. About 75 percent of patients improve, and the improvements can be striking: during a three to four week course of ECT treatment, most depressive episodes will remit and previously suicidal or psychotically depressed patients will completely recover. But, about 40 percent of patients then relapse within two months. This study is designed to answer two questions to improve outcomes; 1) how does ECT therapy work- how does it change brain function during therapy? Learning this could lead to improvements in ECT, also in the non-invasive treatment called transcranial magnetic stimulation (in which a magnet is placed over the patient’s head to alter electrical activity) and in drug treatments; and 2) are there biomarkers that can identify those patients who are at risk of relapse following ECT therapy, so that more aggressive therapy can be provided following ECT treatment to prevent relapse?

Researchers hypothesize that successful ECT response is associated with normalizing brain activity within and between an overactive limbic/paralimbic network and that sustained ECT response may be contingent upon diminishing signaling between the paralimbic and cortical brain networks. They will test this by comparing resting state fMRI in 20 healthy volunteers and 20 patients with MDD. Resting state fMRI identifies regional network connectivity that occurs in the absence of performing a specific task. Volunteers will be imaged to establish “healthy” resting state network activity levels. Patients’ clinical state and brain imaging will be assessed prior to and after ECT treatment. Investigators will determine whether patients’ post-treatment images more closely resemble those of the healthy volunteers, whether the changes in network activity correlate with their clinical outcomes, and what changes in the network are associated with improved outcomes. Patients will be imaged again at the time of relapse or after three months with no relapse, to identify potential biomarkers that differentiate those who relapse from those who do not.

Significance: The study findings may lead to a better understanding of processes involved in MDD and lead to improved treatment and relapse prevention in at-risk patients.

Abstract

Biomarkers of ECT response and relapse in major depressive disorder

Previous research has demonstrated that normalization of cortical and limbic neural networks is necessary for resolution of depressive episodes, but methodological issues have limited the application of this conceptual framework to electroconvulsive therapy (ECT) response. The first aim of this investigation is to determine the biomarkers of ECT response using resting state functional magnetic resonance imaging (fMRI) data examined with advanced data analysis methods. Resting state fMRI expands the generalizability of imaging investigations to patients with severe functional impairment. We will use a case-control, longitudinal design with independent component analysis (ICA) to identify networks affected in major depressive disorder (MDD) including the subcallosal cingulate gyrus, default mode network, dorsal lateral prefrontal cortex, and the dorsal medial prefrontal cortex. Three complementary ICA measures and seed-voxel correlations will assess within and between network changes associated with ECT response. Our preliminary data suggests that ECT response results in 1) the normalization of the component connectivity and temporal coherence in the subcallosal cingulate and anterior default mode network and 2) diminished temporal correlations between the anterior default mode and the dorsal medial prefrontal cortex. An understanding of the biomarkers associated with ECT response will optimize ECT treatment parameters (e.g., pulse width, stimulus delivery method, number of treatments) and may extend to other less invasive, but currently less effective, approaches such as transcranial magnetic stimulation. The second aim of this investigation is to identify the biomarkers associated with sustained ECT response and relapse in MDD. Our preliminary data suggests that changes in temporal correlations between the anterior default mode and the dorsal medial prefrontal cortex are associated with sustained response in ECT. Hence, the long-term goals of this line of research will be to use post-ECT resting state fMRI to identify patients at risk of relapse more robustly than the existing practice of clinical assessment alone. These particular “at risk” patients may benefit from more aggressive continuation therapies (e.g., continuation ECT and pharmacotherapy or TMS), thereby lowering the rate of post-ECT relapse.