Abstracts

Rationale: We have previously shown that PET scanning with AMT is able to identify epileptogenic tubers interictally in patients with tuberous sclerosis complex (TSC). AMT PET is also useful in localizing the epileptic focus in non-TSC patients, particularly those with malformations of cortical development (MCD). It is, however, unclear whether increased AMT uptake in the epileptic focus is specific for particular types of MCD. We addressed this issue by evaluating the clinical and histopathological correlates of AMT PET abnormalities in non-TSC, non-tumor children undergoing resective epilepsy surgery. Methods: Complete clinical, radiological (MRI, AMT PET) and surgical data, including histopathology, were reviewed in 30 children who underwent cortical resection for intractable epilepsy (mean age: 6.5 years; 17 males) between 1998 and 2008 at Children’s Hospital of Michigan. MRI scans were normal in 14 (47%), showed only non-specific changes from a previous surgery in 7 (23%), and suggested MCD in 9 (30%) children. All AMT PET scans were evaluated visually and also using asymmetry indices (AIs) to calculate % increase of AMT uptake in epileptogenic cortex compared to contralateral homotopic cortex. Resected brain specimens were reviewed by a neuropathologist, focusing particularly on brain regions with AMT PET abnormalities. Cortical dysplasia (CD) was classified according to Palmini. Results: Histopathology revealed MCD in 16 (53%) children, including 12 with CD [mild MCD=3; CD type IA=2; CD type IIA=2 and CD type IIB (severe CD with balloon cells)=5]. Polymicrogyria and heterotopias (P&H) were seen in 3 cases and small foci of subependymal heterotopias (SEH) in 1 child. The remaining 14 cases showed normal histopathology with varying degrees of gliosis. Among children with MCD, increased AMT uptake was found in all 5 with CD type IIB (mean AI=10.1%), and 3 with P&H (AI=12.3%), but in none with mMCD and types IA-IIA CD or SEH (TABLE). Whereas all 5 children with CD IIB and 2 with P&H had excellent surgical outcome (Engel class-I), children with milder CD or SEH had variable surgical outcome (Engel class-I=1, II=2, III=4, IV=1). The 14 patients with normal histopathology consisted of 2 groups based on AMT PET findings: 7 patients each with focally increased (AI=15.4%) or normal AMT uptake. Interestingly, while patients with normal pathology and normal AMT PET had better surgical outcome (class I=5; II=2), those with normal pathology but abnormal AMT-PET had normal MRI and poor surgical outcome (class III=4; IV=3). Conclusions: Increased AMT uptake in children with CD may predict type IIB (with balloon cells) and good surgical outcome. Similarities between CD type IIB and cortical tubers in TSC may imply a role of indoleamine 2-3-dioxygenase activation (kynurenine pathway of trptophan metabolism) for increased AMT uptake in these cases. In children with normal MRI and normal histopathology, there is a subgroup with increased AMT uptake and poor surgical outcome. Further characterization of this subgroup is required.