Pathology in resected areas of FDG PET hypometabolism in pediatric epilepsy patients with focal cortical dysplasia type II

Abstract

Introduction: [18F]fluorodeoxyglucose positron emission tomography (FDG-PET) frequently reveals hypometabolism extending beyond MRI-visible focal cortical dysplasia (FCD). Whether surrounding hypometabolic areas harbour pathological cells remains unclear. We compared histopathology between the lesion epicentre and surrounding hypometabolic borders in children undergoing epilepsy surgery. Methods: Fourteen children with intractable, extra-temporal focal epilepsy (mean age 9.0{+/-}5.0 years; 9 female) were retrospectively reviewed. Resections encompassed the visually-apparent MRI signal abnormalities as well as areas of surrounding FDG-PET hypometabolism when safely feasible. Multiple pathological specimens were obtained from the epicentre and surrounding hypometabolic areas. Overall, 136 specimens were analyzed: 64 epicentre (median 4{+/-}3/patient) and 72 border (median 5{+/-}4/patient). Results: Pathology was identified in 75% of epicentre specimens (59% with frank FCD (fFCD) IIa/b, 16% with dysmorphic neurons only (DNO)). Border specimens showed pathology in 62% (31% fFCD IIa/b, 31% DNO). We fitted a Bayesian logistic mixed model with pathology as outcome variable, location as predictor, and subject as a random effect. Compared to negative pathology, the log-odds of fFCD in the epicentre was 1.00 (credible interval (CI) 0.32, 1.77) and -1.25 in the border (CI -2.17, -0.40). The log-odds of DNO vs negative pathology was non-significant in both locations. Exploratory analyses did not show significant associations between pathology and degree of regional hypometabolism or volume. All patients achieved Engel Ia status at one-year follow-up with no long-term neurological deficits. Conclusion: These findings suggest a spatial difference in distribution of histopathology, with fFCD concentrated in the epicentre and DNO present in both the epicentre and hypometabolic borders. Thus, FDG-PET may be used to better detect the histopathological borders of FCD type II, and the high seizure-freedom rate presented here supports the inclusion of these surrounding hypometabolic regions in the surgical resection (when safe to do so), improving the likelihood of removing potentially epileptogenic cells.