A pilot PET/MR study of hippocampal, diffusivity and brain metabolism changes following electroconvulsive therapy

Objectives: Major depressive disorder (MDD) imparts a massive burden to patients and society through its severity and non-responsiveness to treatments. Electroconvulsive therapy (ECT) is effective for treatment resistant depression, although its mechanisms of action are unclear. Recently, it has been put forth that ECT exerts neuroplastic effects in the amygdala and hippocampus. We used simultaneous PET/MR to examine a variety of local and global changes in the brain induced by ECT.

Methods: PET/MR was performed before and after bilateral ECT in 6 medicated MDD patients. Hamilton Depression Rating Scale (HDRS) was used to quantify symptom severity. PET utilized FDG to estimate metabolic rate of glucose (MRGlu); MRI acquisition consisted of MPRAGE and diffusion tensor imaging (DTI), the latter was used to calculate mean diffusivity (MD).

Results: Only one patient did not remit. Although gross hippocampal volume did not change, increases were observed in region IV and the dentate gyrus (p=0.054, all participants) with treatment. Only remitters showed decreased MD in the left hippocampus (p=0.04 for remitters, p=0.15 otherwise) these changes were linearly correlated with HDRS reduction (Pearson ρ=0.88;p=0.02). MRGlu, normalized by cerebellum uptake, generally increased throughout the brain between scans. Whereas PET data were only available for four remitters, significant differences following treatment were observed in prominent midbrain regions. In particular, the dorsal raphe (p=0.038); hippocampus (p=0.052); putamen (p=0.026); and insula (0.0029) showed markedly higher MRGlu post-treatment. Discussion The PET data showed generally increased metabolism throughout the brain following ECT treatment. Despite the small number of remitters included in this analysis, significant differences were observed after treatment in the midbrain. The MRGlu increase in the raphe is of particular interest as the region provides the primary serotonergic innervation for most other brain structures.Although coming short of statistical significance, the MRGlu increase in the hippocampus of responders provides a basic level of support for the emergent theory of neuroplasticity in the hippocampus mediating the therapeutic effects of ECT. Both PET and MRI data implicated the hippocampus, and the potential observation of subfield volume increases along with the correlation between MD reduction in the left hippocampus and ECT efficacy suggests the hippocampus warrants further investigation in studies of ECT.

Conclusion: ECT engenders widespread changes in brain metabolism, with increases noted throughout midbrain structures widely implicated or hypothesized as serving roles in MDD pathophysiology. The synergy of PET and MRI findings in the hippocampus warrants its investigation in larger cohorts which might allow for greater statistical resolution of changes in the region. Research Support: Dana Foundation and the Brain and Behavior Research Foundation