The value of fast ripples in identifying extra-temporal epileptogenicity and predicting surgical outcomes in temporal plus epilepsy: A retrospective SEEG cohort study.

OBJECTIVE

To investigate the value of interictal fast ripples (FRs; 250-500 Hz) recorded by stereoelectroencephalography (SEEG) in identifying extra-temporal epileptogenic sites and predicting surgical outcomes in temporal plus epilepsy (TPE), using a clinically driven, neuroanatomically based categorization of patients and recording contacts.

METHODS

We performed a retrospective cohort study of 21 consecutive patients with electroclinically defined TPE who underwent SEEG monitoring and subsequent resective surgery at our center between January 2021 and June 2023. Interictal ripples (80-250 Hz) and FRs (250-500 Hz) were identified during stable non-rapid eye movement (NREM) sleep by experienced epileptologists using semi-automatic detection followed by blinded visual review. Each recording contact was assigned to one of four neuroanatomical sub-regions on the basis of post-implantation CT-MRI fusion: mesial temporal, lateral temporal, insular, or orbitofrontal cortex. Patients were categorized by the neuroanatomical distribution of the SEEG-defined SOZ into a mesial-restricted TPE group (SOZ confined to mesial temporal structures with rapid extra-temporal propagation, n = 5) and a multilobar TPE group (SOZ involving the temporal lobe plus at least one extra-temporal structure at onset, n = 16). Surgical resection was tailored to each patient based on the SEEG-defined SOZ, with FR-rich extra-temporal sites considered for inclusion when functional risk allowed. Patients were grouped post hoc by the actual resection extent: Group A (temporal resection alone, n = 5) and Group B (temporal plus extra-temporal satellite resection, n = 16). Outcomes were assessed using the Engel classification at ≥ 24 months.

RESULTS

A total of 412 contacts were analyzed. The FR positivity rate was 82.4% (34/41) at SOZ contacts versus 12.3% at non-SOZ contacts (P < 0.001). In the mesial-restricted TPE group, FRs were strongly concentrated in the mesial temporal region (positivity 92.3%), with very low rates in insular (FR density 0.2 ± 0.1 events/min) and orbitofrontal contacts (0.1 ± 0.1 events/min). In the multilobar TPE group, the mesial temporal FR density was comparable (4.5 ± 1.0 vs. 4.8 ± 1.2 events/min, P = 0.452), but FR density was markedly higher in the insula (3.1 ± 0.8 events/min, P < 0.001) and the orbitofrontal cortex (2.4 ± 0.7 events/min, P = 0.003). The Engel Class Ia rate was 81.3% (13/16) in Group B versus 20.0% (1/5) in Group A (OR=0.15, 95% CI: 0.04-0.58, P = 0.006). Focal cortical dysplasia (47.6%) was the most common postoperative histopathological finding.

CONCLUSIONS

When patients with TPE are categorized by the neuroanatomical distribution of the SOZ, interictal FRs co-localize with SOZ contacts and concentrate in the extra-temporal sub-regions (insula, orbitofrontal cortex) that define multilobar TPE. Inclusion of FR-rich extra-temporal sites in the tailored resection was associated with markedly improved postoperative seizure freedom. These findings support FRs as a clinically useful electrophysiological marker of extra-temporal epileptogenicity in TPE, while confirming that formal demonstration of a functional epileptogenic network will require dedicated spatiotemporal connectivity analyses, which we propose as the next step.