Inter-ictal network of focal epilepsy and effects of clinical factors on network activity

The Effectiveness of Vagus Nerve Stimulation in Drug-Resistant Epilepsy Correlates with Vagus Nerve Stimulation-Induced Electroencephalography Desynchronization

Introduction: VNS is an adjunctive neuromodulation therapy for patients with drug-refractory epilepsy. The antiseizure effect of VNS is thought to be related to a diffuse modulation of functional connectivity but remains to be confirmed.

Aim: To investigate electroencephalographic (EEG) metrics of functional connectivity in patients with drug-refractory epilepsy treated by vagus nerve stimulation (VNS), between VNS-stimulated “ON” and nonstimulated “OFF” periods and between responder (R) and nonresponder (NR) patients.

Methods: Scalp-EEG was performed for 35 patients treated by VNS, using 21 channels and 2 additional electrodes on the neck to detect the VNS stimulation. Patients were defined as VNS responders if a reduction of seizure frequency of ∼50% was documented. We analyzed the synchronization in EEG time series during “ON” and “OFF” periods of stimulation, using average phase lag index (PLI) in signal space and phase-locking value (PLV) between 10 sources. Based on graph theory, we computed brain network models and analyzed minimum spanning tree (MST) for responder and nonresponder patients.

Results: Among 35 patients treated by VNS for a median time of 7 years (range 4 months to 22 years), 20 were R and 15 were NR. For responder patients, PLI during ON periods was significantly lower than that during OFF periods in delta (p = 0.009), theta (p = 0.02), and beta (p = 0.04) frequency bands. For nonresponder patients, there were no significant differences between ON and OFF periods. Moreover, variations of seizure frequency with VNS correlated with the PLI OFF/ON ratio in delta (p = 0.02), theta (p = 0.04), and beta (p = 0.03) frequency bands. Our results were confirmed using PLV in theta band (p < 0.05). No significant differences in MST were observed between R and NR patients.

Conclusion: The correlation between VNS-induced interictal EEG time-series desynchronization and decrease in seizure frequency suggested that VNS therapeutic impact might be related to changes in interictal functional connectivity.

Effects of antiepileptic drugs on electroencephalography (EEG): Insights and applicability

OBJECTIVE

The objective of the study was to assess the effects of antiepileptic drugs (AEDs) on posterior alpha rhythm and determine whether they produce pathological slow waves in patients with epilepsy.

METHODS

Outpatient electroencephalographs (EEGs) in alert patients were selected. The three compared groups include 1) patients with an interested AED (monotherapy or combined with other AEDs); 2) patients with AEDs other than the interested AED; and 3) patients who did not take AEDs. Outcomes were frequency of posterior alpha rhythm and presence of generalized continuous (CSWs) and generalized intermittent slow waves (ISWs). Analysis of variance was used to assess which AED was associated with slower posterior alpha rhythm. Chi-square and logistic regression were employed to assess association and odds ratio (OR) with 95% confidence interval (CI) between pathological generalized slow waves and each AED.

RESULTS

Among 1050 EEG tracings, 638 EEGs met our criteria. Electroencephalographs requested because of cognitive decline and psychiatric symptoms were excluded, leaving 616 EEGs for analysis. Four hundred thirty-seven patients received at least one AED, whereas the remaining 179 patients did not take AED. Conventional AEDs [carbamazepine (CBZ), p = 0.024; phenobarbital (PB), p = 0.013; phenytoin (PHT), p = 0.001] except valproic acid (VPA) were associated with slower alpha frequency. Carbamazepine [adjusted OR: 5.74 (95% CI: 2.07, 15.94)] and PB [adjusted OR: 2.58 (95% CI: 1.15, 5.78)] were significantly associated with generalized ISWs. None were associated with generalized CSWs.

CONCLUSIONS

Phenytoin, CBZ, and PB are associated with slower posterior alpha frequency. The latter 2 AEDs also produced pathological generalized ISWs. Valproic acid, benzodiazepines, and new-generation AEDs are not associated with either outcome. The presence of generalized ISWs in patients taking CBZ or PB should be cautiously interpreted since there can be drug effects. Association with cognitive side effects of these slow waves should be further studied.