Ketamine-related neural changes in treatment-resistant depression: A multimodal synthesis of fMRI and PET studies.
Nesreen Sedeek, Carley Rivers, Lucas Williamson, Ayoub Asadi, John G Grundy
Journal of affective disorders September 1, 2026 Peer reviewed DOI: 10.1016/j.jad.2026.121891 via PubMed
Summary
Ketamine has rapid antidepressant effects in some patients with treatment-resistant depression, but neuroimaging results vary due to differences in study methods. A multimodal synthesis of fMRI and PET studies indicates that ketamine effects are often found in subcortical areas and vary across cortical regions like the prefrontal and anterior cingulate. Network analyses suggest involvement of default-mode, ventral attention, and visual systems, highlighting the need for future studies to better connect these findings to clinical outcomes.
Study at a glance
| Design | multimodal synthesis |
|---|---|
| Population | adults with treatment-resistant depression |
| Key finding | Ketamine-related effects were frequently reported in subcortical regions and showed distributed effects across cortical systems. |
Abstract
Ketamine produces rapid antidepressant effects in a subset of patients with treatment-resistant depression, yet neuroimaging findings have been difficult to integrate because studies differ in imaging modality, analytic approach, task context, and post-infusion timing. To address this, we conducted a multimodal synthesis of functional magnetic resonance imaging (fMRI) and positron emission tomography (PET) studies of ketamine in adults with treatment-resistant depression, integrating region-level inspection and functional network mapping to evaluate patterns across heterogeneous designs. The findings suggest that ketamine-related effects were frequently reported in subcortical regions, alongside more distributed and context-dependent effects across cortical systems, including prefrontal and anterior cingulate regions. Network-level summaries further suggested involvement of default-mode, ventral attention, and visual systems. Given variability in imaging modality, task state, and scan timing, these results should be interpreted as hypothesis-generating and motivate future harmonized multimodal studies designed to directly link circuit-level changes to molecular mechanisms and clinical response.