Abstracts

Rationale: We aimed to define the clinical, MRI, and pathological characteristics of a cohort of children who had surgery for temporal lobe epilepsy (TLE) associated with focal cortical dysplasia (FCD)Methods: We retrospectively analyzed 23 consecutive children who had a temporal lobectomy for TLE from March 2011 to March 2015. 15 patients had FCD based on pathology and were included. 8 patients were excluded: 4 had tumors, 2 had gray matter heterotopias, 1 had a 1.5T only preoperative MRI and 1 had a prior MCA stroke. All specimens were re-reviewed by a neuropathologist according to the 2011 ILAE classification of FCD. Preoperative epilepsy protocol 3T MRIs from the 15 TLE patients interleaved with 14 “controls” (9 epilepsy patients with normal MRI and 5 patients who had extra-temporal (ET) lobe epilepsy and FCD) were blindly re-reviewed by a pediatric neuroradiologist. MRIs were categorized as 0) normal, 1) presence of grey-white matter blurring only, 2) blurring associated with decreased myelination of the temporal lobe, or 3) marked signal prolongation on T2-FLAIR. We also examined age of onset, seizure duration, age of resection, most common FCD pathology type, and Engel outcome. Location of MRI abnormality (temporal tip vs. non-temporal tip) was compared to MRI results using likelihood ratio. MRI findings were then condensed into two clinical meaningful groups: less evident (0, 1) and more pronounced (2, 3) and these were examined by group (TLE vs. ET FCD). Within the TLE group pathology and MRI findings were compared using likelihood ratiosResults: Mean age of seizure onset was 32.1 months (SD=27.1; range=3-95 months), mean seizure duration was 6.7 years (SD=5.2; range=0.3-17.9 years), and mean age at surgery was 9.6 years (SD=5.5; range=1.6-18.8 years). FCD Type Ib was the most common pathological finding (n=7; 47%) in the TLE group. Engel outcome were: Class I (n=9), Class II (n=2), and Class III (n=1) and unavailable for 3 patients. In the TLE group, the most frequent location of MRI abnormality was the temporal lobe tip (n= 11; 73%). Temporal tip abnormalities on MRI (n=11) were blurring with decreased myelination (n=9) and marked signal prolongation (n=2). Both TLE and ET patients without temporal tip involvement (n=18) did not demonstrate these abnormalities (χ2(4)=38.50, p<0.001). We found an association between more pronounced vs. less evident MRI findings and group for TLE compared to ET lobe FCD cases (n=5) (χ2(1)=5.89, p=0.02). Based on the odds ratio, TLE patients were 16 times more likely to have more pronounced MRI findings than ET FCD cases. Within the TLE group, we did not find a relationship between FCD classification and MRI findingsConclusions: FCD in pediatric TLE is associated with specific MRI findings, including blurring with decreased myelination of the temporal lobe (especially the tip). The MRI did not relate to specific types of FCD with our sample. These findings suggest the importance of conducting a meticulous analysis of the temporal pole myelination patterns using 3T MRI in patients with suspected TLE