Dynamic functional connectivity links with treatment response of electroconvulsive therapy in major depressive disorder

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

TL;DR

This study reveals that the dynamic functional connectivity of brain networks, specifically the Default Mode Network (DMN) and Cognitive Control Network (CCN), can predict the effectiveness of Electroconvulsive Therapy (ECT) in treating major depressive disorder. By analyzing how these networks shift over time, researchers identified patterns that can forecast treatment outcomes even before ECT begins, offering a potential new biomarker for personalized treatment.

What Problem Does This Paper Address?

Major Depressive Disorder (MDD) is a significant global health challenge, and while Electroconvulsive Therapy (ECT) is a highly effective treatment, it doesn't work for everyone. Currently, predicting who will respond best to ECT before treatment is difficult. Traditional functional connectivity (FC) studies often treat brain network interactions as static. However, brain activity is inherently dynamic, and understanding these dynamic changes might hold the key to predicting treatment response. This paper aims to bridge this gap by investigating how dynamic functional network connectivity (dFNC) of key brain networks (DMN and CCN) changes with ECT and whether these dynamic patterns can predict ECT outcomes in patients with depression.

Methods

The researchers employed a sophisticated analysis pipeline using resting-state functional magnetic resonance imaging (rs-fMRI) data. Here's a breakdown of their approach:

  • Independent Component Analysis (ICA): They first used group ICA to identify independent components (ICs) representing the DMN and CCN from the rs-fMRI data of both depressed patients and healthy controls.
  • Dynamic Functional Network Connectivity (dFNC): A sliding window approach was applied to the time-series data of these ICs to capture how the connectivity between them changes over time. This resulted in a dynamic measure of functional connectivity.
  • K-means Clustering: To characterize the different temporal states of brain network interactions, they applied k-means clustering to the dFNC data. This grouped the dynamic connectivity patterns into distinct "states."
  • Occupancy Rate (OCR): For each participant, they calculated the proportion of time spent in each of these identified states. This metric, the Occupancy Rate (OCR), quantifies how frequently a particular brain network configuration is engaged.
  • Statistical Analysis: They compared OCR values between depressed patients (DEP) and healthy controls (HC) and also correlated pre-ECT OCR with changes in depression severity (measured by the Hamilton Depression Rating Scale - HDRS) to assess predictive power.

Datasets

The study utilized resting-state fMRI data from:

  • 119 patients diagnosed with depression (DEP): These patients underwent a series of ECT treatments.
  • 61 healthy control (HC) participants: This group served as a baseline for comparison.

The data was collected from two different sites: the University of New Mexico (UNM) and the University of California Los Angeles (UCLA).

Key Insights and Ideas

  • Brain Dynamics as a Predictor: The core idea is that the way brain networks interact over time, not just their average connection strength, holds crucial information about treatment response.
  • DMN and CCN as Key Players: The study highlights the importance of the Default Mode Network (DMN) and Cognitive Control Network (CCN) in understanding depression and its treatment.
  • ECT Alters Brain Dynamics: The research demonstrates that ECT not only improves symptoms but also significantly alters the dynamic functional connectivity patterns in the brain.
  • Pre-ECT Brain State Predicts Outcome: The most exciting insight is that specific dynamic brain states before ECT can predict how much a patient's depression will improve.

Key Results

  • Differences in Brain States: Before ECT, healthy controls (HC) spent significantly more time in a specific brain state (State 2) characterized by higher CCN/DMN connectivity compared to depressed patients (DEP).
  • ECT Shifts Brain States: After ECT, this pattern reversed, with DEP patients spending more time in State 2, suggesting ECT reconfigures brain dynamics towards a more "healthy" state.
  • Predictive Power of State 1: The time spent in a particular pre-ECT brain state (State 1), characterized by lower CCN/DMN connectivity, was a significant predictor of HDRS change. Specifically, spending more time in State 1 before ECT was associated with less improvement in depression symptoms (i.e., less HDRS change). This suggests that this state might indicate a poorer response to ECT.
  • Increased Brain Dynamism Post-ECT: The "traveled distance" metric, which quantifies the overall dynamism of brain network states, was significantly lower in DEP patients compared to HC before ECT. After ECT, this difference diminished, indicating that ECT increases brain dynamism in depressed individuals.

Limitations and Future Work

  • Eye State Ambiguity: The study acknowledges that they did not directly measure whether participants' eyes were open or closed during scanning, which can influence rs-fMRI signals. Future work could incorporate methods to assess and control for this.
  • Site Differences: Data was collected from two different sites with variations in imaging protocols and treatment numbers. While the researchers attempted to control for these, it remains a potential confound.
  • HDRS Subjectivity: The Hamilton Depression Rating Scale (HDRS) relies on interviewer assessment, which can introduce variability, especially across different sites.
  • Future Directions: The authors suggest that dFNC features, particularly those related to CCN and DMN connectivity, show promise as predictive biomarkers for ECT outcomes. This could lead to more personalized treatment approaches for depression. Further research could explore these dynamic patterns in more detail to refine prediction models and understand the underlying mechanisms of ECT's therapeutic effects.

Summary updated

Abstract

Abstract Background Electroconvulsive Therapy (ECT) is one of the most effective treatments for major depressive disorder (DEP). There is recently increasing attention to evaluate ECT’s effect on resting-state functional magnetic resonance imaging (rs-fMRI). This study aims to compare rs-fMRI of DEP patients with healthy participants, investigate whether dynamic functional network connectivity network (dFNC) estimated from rs-fMRI predicts the ECT outcome, and explore the effect of ECT on brain network states. Method Resting-state fMRI data were collected from 119 patients with depression or DEP (76 females), and 61 Healthy (HC) participants (34 females) with an age mean of 52.25 (N=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 participant. 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 individual spends in each state, 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, number of treatment, and site. Finally, we evaluated the effectiveness of ECT by comparing pre-and post-ECT OCR of DEP and HC participants. Results The main findings include: 1) DEP patients had significantly lower OCR values than the HC group in a state, where connectivity between CCN and DMN was relatively higher than other states (corrected p= 0.015), 2) Pre-ECT OCR of state, with more negative connectivity between CCN and DMN components, predicted the HDRS changes (R=0.23 corrected p=0.03). This means that those DEP patients who spend 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 spend in state 2 (corrected p=0.03). Finally, we found ECT increases the total traveled distance in DEP. Conclusion Our finding suggests that dFNC features, estimated from CCN and DMN, show promise as a predictive biomarker of the ECT outcome of DEP patients. Also, this study identified a possible underlying mechanism associated with the ECT effect in DEP patients.
BibTeX
@misc{sendi2021dynamic,
  author = {Mohammad S.E. Sendi and Hossein Dini 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 Helen S. Mayberg and Vince D. Calhoun},
  title = {{Dynamic functional connectivity links with treatment response of electroconvulsive therapy in major depressive disorder}},
  year = {2021},
  doi = {10.1101/2021.03.31.437958},
  howpublished = {bioRxiv},
}
Figure