Title : Somatosensory auras in focal Epilepsy: lateralising value and association with temporal lobe Epileptogenic zones in a surgically validated cohort
Abstract:
Aims: Somatosensory auras (SAs) are classically localised to the parietal lobe, yet evidence from surgically validated cohorts remains limited. This study aims to characterise the semiology of SAs and evaluate their lateralising value and association with epileptogenic zone (EZ) localisation in a surgically validated cohort of patients with focal epilepsy.
Methods: A retrospective analysis was conducted of 774 patients undergoing presurgical evaluation for focal epilepsy at a tertiary centre between 2012 and 2024. 64 patients with SAs and a definitive EZ localisation were identified. SAs were characterised via structured clinical documentation obtained during initial consultation and classified by subtype, anatomical distribution and laterality. EZ localisation was determined by multimodal consensus using EEG, MRI and PET imaging, and corroborated by histopathology and postoperative outcomes (Engel Classification). Associations between aura features and EZ localisation were analysed using Fisher’s exact and binomial tests.
Results: 64 patients (8.3%) reported SAs. The most common aura subtypes were numbness (23%) and pain (19%), with the upper limb being the most frequent anatomical distribution (28%). Among patients with unilateral SAs, 74% lateralised contralaterally (p = 0.01). Bilateral SAs were associated with temporal lobe localisation (67%; Fisher’s exact test, p = 0.044, OR 3.05). The EZ was most commonly localised to the temporal lobe (52%), followed by the frontal lobe (17%) and parietal lobe (9%). In MRI-negative cases (n = 17), 67% demonstrated histopathological abnormalities, yet Engel Class I outcomes were achieved in 39/47 (83%) of patients with available outcome data, with no significant difference between lesional and non-lesional cases (p = 0.668).
Conclusion: SAs are not specific to parietal lobe epilepsy and were frequently associated with temporal lobe EZs in this surgically selected cohort. Unilateral SAs provide reliable lateralising information, while bilateral SAs were associated with temporal lobe localisation, potentially reflecting network propagation. These findings support a network-based model of seizure semiology and highlight the importance of detailed aura assessment within a multimodal presurgical evaluation framework.
