Putative Mechanisms of Electroconvulsive Therapy in Treatment-Resistant Schizophrenia Examined Using Magnetic Resonance Imaging

Abstract

Background

The neural mechanisms of electroconvulsive therapy (ECT) in refractory schizophrenia remain elusive. In the current study, we aimed to identify magnetic resonance imaging (MRI)–derived structural (cortical/subcortical volumes) and functional (resting-state connectivity) brain changes after ECT and their associations with clinical response.

Methods

We used an inductive (whole-brain, hypothesis-free) approach to examine structural and functional brain changes and their association with clinical response (positive symptom reduction) in clozapine-refractory schizophrenia (n = 30) after ECT (median 8 sessions). Furthermore, a deductive approach was used to compare baseline whole-brain MRI data from clozapine-refractory patients (n = 31) to data from clozapine responders (n = 23), thereby identifying regions of interest unique to clozapine-refractory schizophrenia. Changes in these regions of interest post-ECT and their association with clinical response were then examined.

Results

The inductive approach identified volumetric enhancement in the bilateral amygdalae (Cohen’s d = 0.4), which was significantly associated with clinical response (β = −0.01, p = .003). The deductive approach identified posterior cerebellar hyperconnectivity as being unique to clozapine-refractory schizophrenia (d = 1.57), which was associated with baseline positive symptoms (r = 0.36, p = .04). Following ECT, there was a significant reduction in posterior cerebellar hyperconnectivity (d = −0.86), and this reduction was significantly associated with clinical response (β = 0.42, p = .002). Increased hippocampal and frontal volumes, frontoparietal connectivity, and reduced sensorimotor connectivity were also observed but were unrelated to clinical response.

Conclusions

ECT may drive clinical improvement in refractory schizophrenia by increasing amygdala volumes and reducing posterior cerebellar connectivity. Randomized sham-controlled trials can confirm these findings in the future.