Abstracts

Rationale: High-frequency oscillations below 200 Hz (ripples) are promising interictal biomarkers observable in most patients with drug resistant epilepsy (DRE), which can be recorded with both intracranial EEG (iEEG) and non-invasive scalp EEG. Yet, the surgical value of ripples has been long debated since they can be observed even in healthy subjects, and thus, their generating area may include non-epileptogenic regions too. To untangle this debate, we previously showed with both iEEG and scalp EEG recordings that interictal ripples propagate across large brain regions; the onset of this propagation can delineate the EZ better than areas of spread (Tamilia et al. 2018, 2021). Yet, it is still unknown whether ripple propagation is a neurophysiological phenomenon observed exclusively in children with DRE or it can also be seen in typically developing (TD) children. Here, we aim to map ripples and their propagation with noninvasive methods from TD children and children with DRE and assess differences and similarities between these electrophysiological markers. We hypothesize that ripple propagation is a network brain mechanism used to transfer information flow and that this mechanism is impaired in patients with DRE.

Methods: We analyzed portions of HD-EEG data from 93 children with DRE (mean age: 13.0 yr.; 50 male) and 65 healthy TD controls (mean age: 11.1 yr.; 31 male) recruited in a two-center, non-interventional, prospective study. We mapped the spatiotemporal propagation of ripples in the source domain with an in-house algorithm based on the wavelet based Maximum Entropy on the Mean (wMEM) inverse solution (Fig 1). For each propagation, we computed characteristic features and defined two zones: the onset and spread. We finally compared features between the two groups (DRE vs. healthy) and between onset and spread.

Results: Ripples were seen in 39 children with DRE and 29 TD children. Ripple propagation was reconstructed in the source space for 74% of DRE patients vs. 55% of TD children (Fig 2A). No difference between DRE and TD was observed in the percentage of ripple propagation over the ripple events, the number, and the rate of propagation (Fig 2B-D). TD showed shorter propagation (45 vs. 61 ms, p=0.03) and a smaller spatial extent than DRE (3 vs. 4 scouts, p=0.02; Fig 2E-H). In TD, no clear difference between onset and spread were observed for area, time latency and distance (Fig 2I-K). In DRE, the onset was more focal than the spread (0.9 vs. 1.2 cm², p=0.02), as well as slower (13 vs. 20 ms, p=0.036; Fig 2I-J). Finally in TD, the onset was faster than the spread (4.4 vs. 2.0 m/s, p=0.029; Fig 2L).

Conclusions: Our data provide first evidence that ripple propagation exists in both TD children and children with DRE; yet these phenomena are characterized by different features which can help in differentiating pathological vs. physiological ripples. Moreover, our findings suggest that ripple propagation is involved in information transfer inside the brain and that this phenomenon is disrupted in children with DRE. This noninvasive mapping may help during the presurgical evaluation by reducing the inclusion of areas generating physiological ripples.

Funding: RO1NS104116-01A1 by NINDS.