Resting state connectivity correlates of treatment resistant depression

Introduction

Treatment-resistant depression (TRD) is defined as failure to respond to at least two separate antidepressant treatments, administered with an adequate dose and duration [1]. Approximately one third of individuals with major depressive disorder (MDD) present with TRD, a common condition that severely impairs quality of life worldwide, leading to persistent interpersonal difficulties and substantial social and economic burden. However, the neurobiological basis of TRD remains poorly understood. In this context, neuroimaging studies provide a non-invasive approach to investigate structural and functional abnormalities associated with both TRD and treatment-sensitive depression (TSD), thereby contributing to a more comprehensive understanding of the underlying pathophysiology and to the development of more effective therapeutic strategies. Although numerous studies have focused on MDD, the structural and functional alterations underlying TRD remain insufficiently explored. In light of these considerations, the present study aimed to investigate differences in resting state functional connectivity between TRD and TSD.

Methods

Our study was performed in a sample of 149 patients with a diagnosis of Major Depressive disorder admitted during an ongoing depressive episode to San Raffaele Hospital psychiatric ward. Treatment resistance status was assessed from available clinical documentation: TRD was defined as failure to respond to two separate antidepressant trials administered at adequate dose and duration. All patients underwent 3T MRI acquisition. T1 and resting-state fMRI sequences were acquired and processed through CONN standard preprocessing pipeline. We first tested differences in Intrinsic Connectivity (IC) between TRD and TSD; significant clusters were further tested through seed-based connectivity analysis.

Results

104 patients met TRD criteria (69.8% of the sample). A significant difference between TSD and TRD patients emerged for a cluster encompassing bilaterally the frontal pole and superior frontal gyrus as well as the left paracingulate gyrus (pFWE=0.002, k=235). Post-hoc seed based connectivity analysis revealed (a) reduced rs-FC of the identified cluster with the supramarginal gyrus, post central gyrus (Left: pFWE<0.001, k=382; Right: pFWE<0.0001, k=522) (dorsal attention, salience and sensorimotor network) the left insular and opercular cortexes (pFWE=0.018, k=215) (salience network) as well as the lateral occipital cortex (pFWE=0.039, k=179) in TSD; and (b) reduced rs-FC with the frontal pole and the frontal medial cortex (pFWE<0.0001, k=402) (default mode network) in TRD.

Discussion

Our analysis identified connectivity differences between TRD and TSD in a cluster located in medial frontal and paracingulate corteces, brain regions highly implicated in mood regulation and affective disorders. In TRD these regions exhibited increased connectivity with task-positive regions and reduced connectivity with a large frontal Default Mode Network (DMN) hub. These findings align with models of treatment resistance that emphasize abnormal integration between default mode, salience/attention, and cognitive control networks in TRD [2,3]. The pattern of enhanced coupling with task-positive regions and reduced frontal default mode connectivity in TRD suggests a network-level marker that may help distinguish treatment response profiles and guide targeted neuromodulatory or pharmacological interventions [4,5].

References

1-Sforzini, L., Worrell, C., Kose, M., Anderson, I. M., Aouizerate, B., Arolt, V., … & Pariante, C. M. (2022). A Delphi-method-based consensus guideline for definition of treatment-resistant depression for clinical trials. Molecular psychiatry, 27(3), 1286-1299. 2-De Kwaasteniet, B., Rive, M., Ruhé, H., Schene, A., Veltman, D., Fellinger, L., Van Wingen, G., & Denys, D. (2015). Decreased Resting-State Connectivity between Neurocognitive Networks in Treatment Resistant Depression. Frontiers in Psychiatry, 6. https://doi.org/10.3389/fpsyt.2015.00028. 3-Han, W., Wu, X., Wang, L., Qu, C., Dou, L., Fang, Y., & Sun, P. (2024). Altered brain function in treatment-resistant depression patients: A resting-state functional magnetic resonance imaging study.. Neuroscience letters, 138004 . https://doi.org/10.1016/j.neulet.2024.138004. 4-Lees, T., Woronko, S., Li, M., Scott, J., Kuhn, M., Esfand, S., Bogdanov, M., Boyle, B., Linton, S., Borchers, L., Zhukovsky, P., Miller, C., Bolton, P., Li, S., Meisner, R., & Pizzagalli, D. (2025). Differences in High-Frequency Connectivity Among Large-Scale Functional Networks Linked to Major Depressive Disorder and Treatment-Resistant Depression. Biological Psychiatry Global Open Science, 5. https://doi.org/10.1016/j.bpsgos.2025.100602. 5-Barreiros, A., Breukelaar, I., Prentice, A., Mayur, P., Tomimatsu, Y., Funayama, K., Foster, S., Malhi, G., Arns, M., Harris, A., & Korgaonkar, M. (2024). Intra- and Inter-Network connectivity of the default mode network differentiates Treatment-Resistant depression from Treatment-Sensitive depression. NeuroImage : Clinical, 43. https://doi.org/10.1016/j.nicl.2024.103656.