Abstracts

Rationale: T2 hyperintense signal changes in corpus callosum have previously been described in epilepsy patients. Little is known about the mechanisms producing these changes. It is also unknown whether these signal changes have any particular relationship to epilepsy types, seizure frequency or any antiepileptic medications. We characterized the clinical features of these patients and analyzed the radiological lesion seen on MRI in order to find predictors of their occurrence and reversibility. Methods: We did a retrospective search of children with epilepsy with MRI signal changes in the corpus callosum between 2004-2010. Presence of tumor or other brain lesions including ADEM, which could produce similar signal change was considered an exclusion criterion. Demographic information, seizure type, seizure frequency and anticonvulsants were then ascertained from the database. Time gap between the last seizure and the MRI was also noted. Lesions were divided based on their orientation, apparent diffusion coefficient (ADC) values and whether there was associated corpus callosum thinning. MRI signal changes were analyzed sequentially in T1, T2, FLAIR, DWI and contrast. The locations and sizes of the lesion were described and available follow up scans were reviewed. Results: Twelve children (age range -3 months to 15 years; median- 6 years ; M:F-1:3) met the inclusion criteria. 8 had generalized seizures; 2 had occipital lobe epilepsy; 1 had temporal lobe seizures; 1 had neonatal seizures. The seizure frequency was variable. Four had an explosive and prolonged onset of seizures. In 6 of the 10 patients, a prolonged seizure was observed in the week prior to the MRI. There was no relationship to any particular antiepileptic medications (Phenytoin, Phenobarbital, Carbamazepene, Valproate, Oxcarbazepine Lamotrigine, Levetiracetam & Zonisamide). The MRI lesion was in the splenium of corpus callosum in 10 of the 12 children and in the genu in the other two. It was usually hypo or isointense on T1 and hyperintense on T2 & FLAIR. There was no contrast enhancement. The ADC value of the lesions at presentation was variable. Eight of the twelve children had follow-up MRI, and four of them had complete resolution of the lesion. The splenial lesions with an oval shape and lack of corpus callosum thinning were the most likely lesions to resolve on follow up imaging. Conclusions: MRI lesions in the corpus callosum may be associated with both generalized or focal epilepsy. The frequency and duration of the seizures immediately before the MRI may play a role. Lesions are not related to any particular antiepileptic medication and may be reversible in about 50% over time.