OBJECT.: The authors conducted a study to correlate histopathological features, MRI findings, and postsurgical outcomes in children with cortical dysplasia (CD) by performing a novel resection site-specific evaluation.
Methods: The study cohort comprised 43 children with intractable epilepsy and CD. The MR image review was blinded to pathology but with knowledge of the resection location. An MRI score (range 0-7) was calculated for each resection region based on the number of imaging features of CD and was classified as "lesional" or "nonlesional" according to all imaging features. Outcome was determined using the International League Against Epilepsy (ILAE) scale. The determination of pathological CD type was based on the ILAE 2011 consensus classification system, and the cortical gliosis pattern was assessed on GFAP staining.
Results: There were 89 resection regions (50 ILAE Type I, 29 Type IIa, and 10 Type IIb). Eleven (25.6%) of 43 children had more than one type of CD. The authors observed MRI abnormalities in 63% of patients, characteristic enough to direct resection (lesional) in 42%. Most MRI features, MRI score ≥ 3, and lesional abnormalities were more common in patients with Type II CD. Increased cortical signal was more common in those with Type IIb (70%) rather than Type IIa (17.2%) CD (p = 0.004). A good outcome was demonstrated in 39% of children with Type I CD and 72% of those with Type II CD (61% in Type IIa and 100% in Type IIb) (p = 0.03). A lesional MRI abnormality and an MRI score greater than 3 correlated with good outcome in 78% and 90% of patients, respectively (p < 0.03). Diffuse cortical gliosis was more prevalent in Type II CD and in resection regions exhibiting MRI abnormalities. Complete surgical exclusion of the MRI abnormality was associated with a better postoperative outcome.
Conclusions: This study provides a detailed correlation of MRI findings, neuropathological features, and outcomes in children with intractable epilepsy by using a novel resection site-specific evaluation. Because 25% of the patients had multiple CD subtypes, a regional analysis approach was mandated. Those children with lesional MRI abnormalities, Type II CD, and surgical exclusion of the MRI abnormality had better outcomes. Type II CD is more detectable by MRI than other types, partly because of the greater extent of associated gliosis in Type II. Although MRI findings were correlated with the pathological CD type and outcome in this study, the majority of patients (58%) did not have MRI findings that could direct surgical therapy, underscoring the need for improved MRI techniques for detection and for the continued use of multimodal evaluation methods in patient selection.
Keywords: CD = cortical dysplasia; EEG = electroencephalography; FSE = fast spin echo; GFAP = glial fibrillary acidic protein; ILAE = International League Against Epilepsy; MEG = magnetoencephalography; MRI; NeuN = neuronal nuclei; SISCOM = subtraction ictal SPECT coregistered to MRI; SPECT = single-photon emission computed tomography; children; cortical dysplasia; epilepsy; histopathology; surgical outcomes.