Linking inflammatory and structural neuroimaging biomarkers in treatment-resistant depression: Current evidence and emerging hypotheses

Abstract

Treatment-resistant depression (TRD) represents a major clinical challenge within major depressive disorder (MDD) and is currently defined primarily by inadequate response to antidepressant treatment rather than by underlying biological mechanisms. Despite extensive research, the absence of validated biomarkers continues to limit diagnostic precision, hamper cross-study comparability, and constrain the development of mechanism-based interventions. Growing evidence suggests that immune–inflammatory dysregulation and structural brain alterations are both prominent features of TRD and may represent interrelated dimensions of its pathophysiology. Over the past two decades, a number of studies have reported peripheral immune abnormalities in TRD, including alterations in acute-phase proteins, inflammatory cytokines, immune cell composition, and neuroendocrine–immune interactions. Elevated C-reactive protein (CRP), interleukin-6 (IL-6), tumor necrosis factor-α (TNF-α)–related signaling, and complement components have been observed in subsets of TRD patients; however, findings remain heterogeneous and non-linear across studies, suggesting that these markers may reflect inflammatory heterogeneity rather than a uniform immune phenotype. Importantly, responder–non-responder comparisons indicate that treatment resistance may not simply be associated with greater inflammatory burden, but rather with alterations in the resolution of immune activation. In addition, some studies suggest that clinical improvement is accompanied by partial normalization of central immune pathways, whereas persistent non-response has been associated with sustained inflammatory activity. In parallel, structural neuroimaging studies have reported reductions in hippocampal and caudate volumes, grey matter alterations, and white matter abnormalities in TRD, although many of these findings overlap with those observed in MDD more broadly. Emerging evidence further suggests potential associations between immune markers and neuroimaging features, including links between cytokines, CRP, kynurenine pathway metabolites, and brain structure.

Taken together, these findings are broadly consistent with models proposing that inflammation-related processes, altered neurotrophic signaling, glucocorticoid dysregulation, and disrupted immune–brain interactions may contribute to reduced neuroplasticity and treatment resistance. Overall, the literature points to TRD as a biologically heterogeneous condition, potentially encompassing inflammation-sensitive and inflammation-insensitive subgroups. However, current evidence remains insufficient for formal biological stratification. Future integrative approaches combining peripheral and central measures with longitudinal and interventional designs may help advance biologically informed classification and treatment strategies in TRD.