Neuronal complexity tracks changes of epileptic activity and identifies epilepsy patients independent of interictal...

To date, the identification of objective biomarkers of neural epileptic activity (EA) remains challenging. We therefore investigated whether neuronal complexity could serve as an interictal electroencephalographic measure of EA, independent of interictal epileptiform discharges (IEDs). By tapering anti-seizure medication (ASM) during video-EEG (electroencephalography) monitoring (VEM), we studied whether changes in neuronal complexity could reliably indicate the increase in EA and identify patients with epilepsy.

The study included 27 patients with unilateral mesial temporal lobe epilepsy (TLE) and 24 control patients with non-epileptic episodes (NEEs) only, each undergoing ASM reduction during VEM. Thirteen additional patients undergoing intracranial recordings during VEM were included to study the relation of surface EEG complexity to intracranial IED. Neuronal complexity was quantified using sample entropy. Delta power served as a control parameter. Receiver-operating characteristic (ROC) analysis was used to evaluate diagnostic performance.

As ASM was reduced, patients with epilepsy showed a significant decrease in neuronal complexity over consecutive days (p = .0008). In contrast, patients with NEE showed no significant change in neuronal complexity (p = .78). Delta power in contrast increased and did not differ significantly between patients with TLE and patients with NEE (p = 1). ROC analysis demonstrated that neuronal complexity effectively distinguished between patients with epilepsy and patients with NEE (area under the curve [AUC] = .76), whereas delta power performed at chance level (AUC = .5). Analysis of simultaneously recorded surface and intracranial EEG showed that hippocampal IEDs are followed by an increase in surface EEG delta power (p = 1.8 × 10^-18) without any significant change in complexity (p = .39).

An increase in EA caused by ASM reduction resulted in a loss of neuronal complexity in surface EEG recordings of patients with epilepsy, independent of IEDs. These findings suggest that neuronal complexity could serve as a potential biomarker to differentiate between epilepsy patients and those with NEEs only. This holds promise for improving the clinical evaluation of EA in epilepsy, addressing the limitations of seizure frequency and IED identification.