Cortical Network Activity Predicts IGF-1 Treatment Response in Rett Syndrome

Abstract

Background: Rett Syndrome is a neurodevelopment disorder associated with mutations in the gene MECP2, which is involved in the development and function of cortical networks. This dysfunction results in specific electrophysiological abnormalities which are associated with clinical features. Insulin-like growth factor 1 (IGF-1) has been shown to ameliorate the symptoms of Rett in animal models and in early clinical trials, but it remains unclear whether IGF-1 treatment impacts the underlying network architecture in patients with Rett. Methods: In this study, we performed clinical assessment and resting-state EEG recordings in eighteen patients with Rett, nine of which were treated with IGF-1. We repeated the assessment at baseline and twelve months following treatment. Network measures were derived using statistical modelling techniques based on inter-electrode coherence measures. Results: We demonstrate that IGF-1 treatment is associated with alterations in network measures, and that there are differences in network architecture associated with clinical response. Further, we show that network measures capture heterogeneity within the Rett Syndrome population that is not evident clinically, and that there are electrophysiological differences between treatment responders and non-responders prior to IGF-1 treatment. Further, we assessed the ability of these computational biomarkers to predict treatment response. Using derived network measures, we trained a support vector machine model and we demonstrate that pre-treatment network measures can predict treatment response in unseen patient data with 100% accuracy (100% sensitivity & 100% specificity) in this small group. Conclusions: These results further underline the importance of network pathology in this disorder and highlights the potential for approaches using these techniques to better characterise disease and allow more targeted treatment.