fNIRS-VFT is a complementary tool to 18F-FDG PET in presurgical evaluation of drug-resistant epilepsy: an exploratory study

Abstract Background Accurate presurgical localization of the epileptogenic zone (EZ) and preservation of postoperative cognitive function are two critical but often competing goals in epilepsy surgery for drug-resistant epilepsy (DRE). This study aimed to (1) evaluate the complementary value of functional near-infrared spectroscopy during a verbal fluency task (fNIRS-VFT) alongside interictal 18 F-FDG PET for presurgical EZ localization; and (2) explore whether task-based fNIRS functional connectivity is associated with postoperative verbal memory decline. Methods We retrospectively included patients with DRE who underwent epilepsy surgery between December 2022 and December 2023 and completed both interictal PET and fNIRS-VFT assessments. All analyses in this study were descriptive, and no robust inferential statistical testing was performed. For the localization outcome, concordance between PET hypometabolism and fNIRS-VFT hypoactivation was analyzed in patients who achieved seizure freedom two years after surgery. For the memory outcome, post- vs. pre-operation change of task-based fNIRS functional connectivity was compared between patients with and without clinically significant postoperative verbal memory decline (≥ 4 point reduction on Auditory Verbal Learning Test delayed recall). Results Twenty-eight patients were included in this study. Among 20 patients achieving seizure freedom, PET demonstrated concordant lateralization with the EZ in 14 cases (70%), including 11 with accurate localization; fNIRS-VFT showed concordant lateralization in 9 patients (45%), with 4 accurate localizations including one case in whom PET failed to lateralize the EZ, indicating the complementary value of fNIRS-VFT. Among 19 patients with complete memory data, four showed significant postoperative decline. These patients exhibited more pronounced alterations in task-related functional connectivity, particularly involving frontopolar, dorsolateral prefrontal, premotor, and temporal regions. Conclusions In this exploratory, descriptive analysis, interictal 18 F-FDG PET showed higher concordance with lateralization of surgical areas compared with fNIRS-VFT among patients with DRE. Nevertheless, fNIRS may provide complementary information for cases with inconclusive PET findings. Additionally, changes of task-based fNIRS connectivity may be correlated with postoperative verbal memory alterations. These preliminary findings suggest that fNIRS-VFT may provide complementary information in selected cases, which warrants validation in larger prospective cohorts.