Dynamic Functional Connectivity Predicts Treatment Response to Electroconvulsive Therapy in Major Depressive Disorder

Hossein Dini, Mohammad S.e. Sendi, Jing Sui, Zening Fu, Randall Espinoza, Katherine L. Narr, Shile Qi, Chris Abbott, Sanne J.h. van Rooij, Patricio Riva‐Posse, Luis Emilio Bruni, Helen S. Mayberg, Vince D. CalhounFrontiers in Human Neuroscience 2021 · DOI

Abstract

Background: Electroconvulsive therapy (ECT) is one of the most effective treatments for major depressive disorder (MDD). Recently, there has been increasing attention to evaluate the effect of ECT on resting-state functional magnetic resonance imaging (rs-fMRI). This study aims to compare rs-fMRI of depressive disorder (DEP) patients with healthy participants, investigate whether pre-ECT dynamic functional network connectivity (dFNC) is associated with eventual ECT outcomes, and explore the effect of ECT on brain network states.

Method: Resting-state fMRI data were collected from 119 DEP patients and 61 healthy controls (HC). Twenty-four independent components from the default mode network (DMN) and cognitive control network (CCN) were extracted using group-independent component analysis (GICA). A sliding window approach estimated pre- and post-ECT dFNC, followed by k-means clustering to identify three distinct brain states. The primary metric was the "occupancy rate" (OCR)—the percentage of time spent in each state.

Results: DEP patients showed significantly lower OCR in State 2 (characterized by higher CCN-DMN connectivity) compared to HCs. Pre-ECT OCR in State 1 (more negative CCN-DMN connectivity) predicted clinical improvement (HDRS changes). Post-ECT analysis revealed that treatment significantly increased the time patients spent in the "healthy" State 2.

Conclusion: dFNC features show promise as predictive biomarkers for ECT outcomes and identify a potential neural mechanism for the treatment's efficacy.

Introduction

Major Depressive Disorder (MDD) is a leading cause of global disability. While Electroconvulsive Therapy (ECT) is highly effective for treatment-resistant cases—with remission rates near 75%—its precise neural mechanisms remain elusive. Furthermore, clinicians lack reliable tools to predict which patients will respond to the weeks-long procedure.

Traditional neuroimaging often relies on "static" functional connectivity, which averages brain activity over several minutes. This study argues that static averages mask the brain's highly dynamic nature, potentially missing the subtle "state-switching" patterns that characterize mental health. By analyzing dynamic functional network connectivity (dFNC), researchers sought to determine if the "flickering" nature of brain connectivity could serve as a biomarker for treatment success.

Methods

The researchers employed a sophisticated pipeline to capture temporal brain dynamics:

  1. Group Independent Component Analysis (GICA): 24 nodes were extracted from the Cognitive Control Network (CCN) and the Default Mode Network (DMN).
  2. Sliding Window Approach: Connectivity was calculated within 40-second windows sliding across the time series to observe second-by-second evolution.
  3. K-means Clustering: Thousands of windows were grouped into three distinct "brain states."
  4. Occupancy Rate (OCR): The percentage of time each participant spent in each state was calculated.
  5. Traveled Distance: A novel metric measuring the "dynamism" or fluidity of state-switching.
Figure 1: The dFNC analysis pipeline, from GICA to statistical analysis of occupancy rates.

Results and Discussion

The "Healthy" State and ECT Reset

The study identified State 2 as a pattern characterized by relatively high connectivity between the CCN and DMN. Healthy controls spent significantly more time in this state than depressed patients (FDR corrected $p = 0.015$). Crucially, ECT was found to increase the time DEP patients spent in State 2 (FDR corrected $p = 0.03$), suggesting that the therapy effectively "resets" the brain's ability to maintain healthy configurations.

Predictive Power of Pre-Treatment States

Pre-treatment occupancy in State 1 (characterized by negative or weak connectivity between CCN and DMN) served as a significant predictor of clinical improvement. Patients who spent less time in this low-connectivity state before treatment showed the greatest reduction in Hamilton Depression Rating Scale (HDRS) scores ($R = 0.23, p = 0.03$).

Figure 5: Comparison of estimated brain states and their occupancy rates before (A) and after (B) ECT.

Reduced Dynamism in Depression

The "traveled distance" metric revealed that depressed brains are less "fluid," switching between patterns less frequently than healthy brains ($p = 0.04$). This suggests that MDD may be characterized by a degree of neural rigidity that ECT helps to alleviate.

Conclusion

These findings suggest that dFNC features, particularly those involving the CCN and DMN, are promising predictive biomarkers for ECT outcomes. While limitations such as site variability and participant drowsiness exist, these dynamic features could eventually be integrated into machine learning models to identify likely responders before treatment begins, optimizing clinical care for treatment-resistant depression.

Summary updated

Abstract

<b>Background:</b> Electroconvulsive therapy (ECT) is one of the most effective treatments for major depressive disorder. Recently, there has been increasing attention to evaluate the effect of ECT on resting-state functional magnetic resonance imaging (rs-fMRI). This study aims to compare rs-fMRI of depressive disorder (DEP) patients with healthy participants, investigate whether pre-ECT dynamic functional network connectivity network (dFNC) estimated from patients rs-fMRI is associated with an eventual ECT outcome, and explore the effect of ECT on brain network states. <b>Method:</b> Resting-state functional magnetic resonance imaging (fMRI) data were collected from 119 patients with depression or depressive disorder (DEP) (76 females), and 61 healthy (HC) participants (34 females), with an age mean of 52.25 (<i>N</i> = 180) years old. The pre-ECT and post-ECT Hamilton Depression Rating Scale (HDRS) were 25.59 ± 6.14 and 11.48 ± 9.07, respectively. Twenty-four independent components from default mode (DMN) and cognitive control network (CCN) were extracted, using group-independent component analysis from pre-ECT and post-ECT rs-fMRI. Then, the sliding window approach was used to estimate the pre-and post-ECT dFNC of each subject. Next, k-means clustering was separately applied to pre-ECT dFNC and post-ECT dFNC to assess three distinct states from each participant. We calculated the amount of time each subject spends in each state, which is called "occupancy rate" or OCR. Next, we compared OCR values between HC and DEP participants. We also calculated the partial correlation between pre-ECT OCRs and HDRS change while controlling for age, gender, and site. Finally, we evaluated the effectiveness of ECT by comparing pre- and post-ECT OCR of DEP and HC participants. <b>Results:</b> The main findings include (1) depressive disorder (DEP) patients had significantly lower OCR values than the HC group in state 2, where connectivity between cognitive control network (CCN) and default mode network (DMN) was relatively higher than other states (corrected <i>p</i> = 0.015), (2) Pre-ECT OCR of state, with more negative connectivity between CCN and DMN components, is linked with the HDRS changes (R = 0.23 corrected <i>p</i> = 0.03). This means that those DEP patients who spent less time in this state showed more HDRS change, and (3) The post-ECT OCR analysis suggested that ECT increased the amount of time DEP patients spent in state 2 (corrected <i>p</i> = 0.03). <b>Conclusion:</b> Our finding suggests that dynamic functional network connectivity (dFNC) features, estimated from CCN and DMN, show promise as a predictive biomarker of the ECT outcome of DEP patients. Also, this study identifies a possible underlying mechanism associated with the ECT effect on DEP patients.
BibTeX
@inproceedings{dini2021dynamic,
  author = {Hossein Dini and Mohammad S.E. Sendi and Jing Sui and Zening Fu and Randall Espinoza and Katherine L. Narr and Shile Qi and Chris Abbott and Sanne J.H. van Rooij and Patricio Riva‐Posse and Luis Emilio Bruni and Helen S. Mayberg and Vince D. Calhoun},
  title = {{Dynamic Functional Connectivity Predicts Treatment Response to Electroconvulsive Therapy in Major Depressive Disorder}},
  booktitle = {Frontiers in Human Neuroscience},
  year = {2021},
  doi = {10.3389/fnhum.2021.689488},
}
Figure