H. Li, X. Shao, F. He, C. He, Y. Yang, Y. Ge, R. Chen, Z. Wang, Y. Gong, X. Long, P. Miao, Y. Ding, S. Wang, et al. (2025). Neural dynamics remodeling induced by ketogenic diet therapy in drug-resistant epilepsy: A functional magnetic resonance imaging study. Epilepsia, 66(9), 3465-3479. https://doi.org/10.1111/epi.18485

PubMed 40464768

[Abstract]
OBJECTIVE: Ketogenic diet therapy (KDT) is a safe and effective intervention for drug-resistant epilepsy (DRE), yet its neural mechanisms remain unclear. This study aimed to explore how KDT-induced changes in brain network dynamics relate to its therapeutic efficacy.

METHODS: Resting-state functional magnetic resonance imaging was performed on 38 DRE patients at both pre- and post-KDT stages, alongside 35 healthy controls. Using energy landscape analysis, we assessed individualized brain network dynamics, including brain states and transitions. Patients were categorized as responders or nonresponders based on seizure reduction rates following KDT. A subset of 14 follow-up patients completed an auditory reaction time (ART) task to evaluate information processing speed. Exploratory analyses examined associations between brain dynamics and KDT outcomes.

RESULTS: Six consistent brain states were identified across healthy controls and patients, categorized into two major states and one intermediate state. KDT normalized abnormal brain dynamics in DRE by reducing the frequency and duration of major states and direct transitions between major states, and by increasing the frequency and duration of the intermediate state and indirect transitions between major states. Notably, the altered frequency of direct transitions was correlated with the rate of seizure reduction. Whereas ART was correlated with rigid dynamics at the pre-KDT stage, post-KDT improvements in ART showed an association with enhanced dynamic flexibility. The pre-KDT frequency of direct transitions emerged as a potential biomarker for predicting KDT responsiveness.

SIGNIFICANCE: This study provides novel insights into the neural mechanisms underlying KDT in DRE, highlighting the significance of monitoring network dynamics and its normalization effects on the brain. The identified biomarker holds promise for facilitating personalized treatment strategies, thereby optimizing therapeutic outcomes for patients with DRE.