Resting-State Functional Metrics Associated with Seizure Recurrence After Antiseizure Medication Withdrawal

Antiseizure medication (ASM) withdrawal after sustained seizure freedom requires a balanced estimate of recurrence risk, yet objective imaging correlates remain uncertain. This prospective observational protocol compared prewithdrawal resting-state functional magnetic resonance imaging (fMRI) in 30 patients with epilepsy and 22 healthy controls (HC). All patients completed 12 months of follow-up after complete withdrawal; 16 developed recurrence (RG) and 14 remained seizure-free (NRG). Seizure-onset zones were temporal (n = 13), frontal (n = 9), generalized (n = 4), parietal/occipital (n = 2), and undetermined focal (n = 2). Metric-specific branches generated amplitude of low-frequency fluctuation (ALFF), fractional ALFF (fALFF), and regional homogeneity (ReHo) maps. Functional connectivity and graph metrics used the automated anatomical labeling atlas; network-based statistic (NBS) testing used 5,000 permutations. Models included group, age, and sex, with Holm correction for pairwise contrasts. RG showed higher left gyrus rectus fALFF than HC and NRG and higher left orbital frontal fALFF and right supplementary motor area ReHo than HC. NRG showed higher middle cingulate ALFF than HC, although this finding was less stable in sensitivity analyses. No global or nodal network metric differed among groups, and no functional connection survived NBS family-wise correction. These results associate seizure recurrence with localized prewithdrawal functional differences while providing no evidence for a corrected large-scale connectivity signature. The findings support regional hypothesis generation but do not establish an individual prediction tool.​.