Predicting response to electroconvulsive therapy combined with antipsychotics in patients with schizophrenia using insula-related functional connectivity: an interoceptive theory-driven analysis

Vollständiger Abstract

Worum geht es in dieser Arbeit?

Abstract Background Variability in electroconvulsive therapy (ECT) response among patients with schizophrenia and the risks of ineffective treatment limit its clinical use. Identifying reliable predictors to accurately forecast patient response to ECT continues to be a major challenge. This study aimed to investigate whether insula-related functional connectivity features (FCs), grounded in interoceptive dysfunction theory, can accurately predict ECT-antipsychotics (APs) combination outcomes in schizophrenia and identify reliable predictors and potential neural mechanisms of ECT-APs. Methods Forty-five patients with schizophrenia, either treatment-resistant or in an acute exacerbation, underwent MRI and clinical assessments at baseline and after 4-weeks of ECT-APs treatment in this single-center proof-of-concept study. Based on interoceptive dysfunction theory, baseline insula-related FCs were extracted to train a machine learning model for predicting the change in Positive and Negative Syndrome Scale scores (PANSS%) within a nested cross-validation framework. We also identified the functional connectivity patterns of responders and explored potential neural targets of ECT-APs. Results Cross-validation of the optimal model revealed that 16 baseline insula-related FCs significantly predicted post-ECT-APs changes in PANSS(%) scores (r[predicted vs. observed] = 0.93, p < 0.001). These features involved FCs between the insula (ventral anterior and posterior) and large-scale networks encompassing frontal, temporal, limbic, thalamic, and cerebellar regions. Among them, three FCs showed significant post-ECT-APs changes that were associated with symptom improvement, suggesting dual roles as predictors and potential therapeutic targets involving salience-auditory-interoceptive circuits. Others, such as insula-cerebellar Crus II and insula-precuneus connections, distinguished responders and non-responders but less modifiable by ECT-APs. Conclusions These findings provide preliminary evidence for the predictive value of insula-related functional connectivity in ECT-APs response; future multi-center studies are needed to confirm generalizability. Clinical trial number Not applicable.