Network connectivity predicts effectiveness of responsive neurostimulation in focal epilepsy

Abstract

Responsive neurostimulation (RNS) is a promising treatment for drug-resistant focal epilepsy; however, clinical outcomes are highly variable across individuals. The therapeutic mechanism of RNS likely involves modulatory effects on brain networks; however, with no known biomarkers that predict clinical response, patient selection remains empiric. This study aimed to determine whether functional brain connectivity measured non-invasively prior to device implantation predicts clinical response to RNS therapy. Resting-state magnetoencephalography was obtained in 31 participants with subsequent RNS device implantation between August 15, 2014 and October 1, 2020. Functional connectivity was computed across multiple spatial scales (global, hemispheric, and lobar) using pre-implantation magnetoencephalography and normalized to maps of healthy controls. Normalized functional connectivity was investigated as a predictor of clinical response, defined as percent change in self-reported seizure frequency in the most recent year of clinic visits relative to pre-RNS baseline. Area under the receiver operating characteristic curve (AUC) quantified the performance of functional connectivity in predicting responders (≥50% reduction in seizure frequency) and non-responders (<50%). Leave-one-out cross validation was furthermore performed to characterize model performance. The relationship between seizure frequency reduction and frequency-specific functional connectivity was further assessed as a continuous measure. Across participants, stimulation was enabled for a median duration of 52.2 (interquartile range, 27.0-62.3) months. Demographics, seizure characteristics, and RNS lead configurations were matched across 22 responders and 9 non-responders. Global functional connectivity in the alpha and beta bands were lower in non-responders as compared to responders (alpha, pfdr < 0.001; beta, pfdr < 0.001). The classification of RNS outcome was improved by combining feature inputs; the best model incorporated four-features (i.e. mean and dispersion of alpha and beta bands) and yielded an AUC of 0.970 (0.919-1.00). The leave-one-out cross validation analysis of this four-feature model yielded a sensitivity of 86.3%, specificity of 77.8%, positive predictive value of 90.5%, and negative predictive value of 70%. Global functional connectivity in alpha band correlated with seizure frequency reduction (alpha, p = 0.010). Global functional connectivity predicted responder status more strongly, as compared to hemispheric predictors. Lobar functional connectivity was not a predictor. These findings suggest that non-invasive functional connectivity may be a candidate personalized biomarker that has the potential to predict RNS effectiveness and to identify patients most likely to benefit from RNS therapy. Follow-up large-cohort, prospective studies are required to validate this biomarker. These findings furthermore support an emerging view that the therapeutic mechanism of RNS involves network-level effects in the brain.

Responsive Neurostimulation for Seizure Control: Current Status and Future Directions

“…Fan et al looked at functional connectivity across various spatial scales, including global, hemispheric, and lobar, after spectral decomposition of patient ECoG data obtained from RNS. Fan et al demonstrated that increased baseline functional network connectivity was associated with improved clinical outcomes, as measured by percent change in self-reported seizure frequency in the most recent year of clinic visits, compared to prior to RNS device implantation [ 54 ]. More specifically, they identified that increased global functional connectivity in the alpha frequency band was correlated with seizure frequency reduction (P = 0.010).…”

“…RNS currently targets 1–2 foci in order to treat seizures, in approaching epilepsy as a focus-based disorder. However, as described above, studies are now showing that neuromodulatory devices exert their therapeutic effects through network disruption, not focus disruption [ 54 ]. Using this information, we can inform RNS target placement to address key nodes in the epileptogenic network, allowing for improved seizure control.…”

Epileptic Tissue Localization through Skewness-Based Functional Connectivity in the High-Frequency Band of Intracranial EEG

“…For nearly one-third of refractory patients, seizures cannot be fully controlled by antiepileptic drugs [ 2 ]. In order to achieve seizure freedom for those patients, clinical treatment including surgical resection, laser ablation and responsive neurostimulation (RNS) of epileptic tissue are considered during intracranial electroencephalography (iEEG) monitoring [ 3 , 4 , 5 ]. However, effective treatment depends on accurate localization of epileptic tissue, which is a major challenge in clinical practice.…”

Prognostication of responsive neurostimulation system responsiveness using presurgical magnetoencephalography

“…In their recent article in Brain Communications, Fan et al . 7 provide a fascinating and well-constructed analysis of resting-state magnetoencephalography (MEG) data collected from patients who went on to have RNS device implantations and show that FC measures could provide a patient-specific biomarker for providing an a priori expected RNS responsiveness pre-implantation ( Fig. 1 ).…”

“… The main findings of Fan et al . 7 are summarized in this figure. Thirty-one patients underwent MEG ( left ) before implantation of a RNS device.…”