SPECT and Localized Proton MRS of Epilepsy in Children with Cortical Dysplasia

Purpose: Recently cortical dysplasia (CD) attracted attention because of its importance as a focus for epilepsy and a lesion for surgery. However, the clinical features of children with epilepsy with CD vary in the severity of their symptoms, depending on the type or extent of the CD. We performed single photon emission computed tomography (SPECT) and proton magnetic resonance spectroscopy (MRS) in children with epilepsy with CD to investigate the relation between their clinical and neuroimaging features. Methods: Ten children with epilepsy with distinct CD revealed by previous CT or MRI were studied. The ages examined were from 5 days to 13 years. The types of CD were lissencephaly in one case, diffuse pachygyria in two, schizencephaly in one (extensive CD), focal pachygyria in two, polymicrogyria in two, hemimegalencephaly in one, and nodular heterotopia in one (focal CD). Interictal SPECT using99Tc-ECD was performed in all cases, and ictal SPECT in two cases. Proton MRS was performed in eight cases on a superconductive magnet with a main magnetic field of 1.5 Tesla (Magneton-Siemens Co., Erlangen, Germany). In each case, the regions of interest (2×2×2 cm) were visually selected in the dysplastic cortex, and the contralateral normal or mildly dysplastic cortex. Results: The four cases with extensive CD showed severe mental and motor retardation. All of their epileptic seizures began within a year of life and were atypical and intractable tonic spasms with focal signs. On the other hand, the six cases with focal CD showed mild psychomotor handicaps. They had mainly localization-related epilepsies originating from the region of CD, which tended to be relatively late in onset and have a good prognosis. On the brain perfusion (rCBF) obtained with SPECT, the area with CD exhibited plump and sausage-like images in many cases. The rCBF in the dysplastic cortex varied in each case: hyperperfusion in four cases, hypoperfusion in three, and normoperfusion in three. In a case with hemimegalencephaly, the perfusion pattern showed dynamic changes during the course: diffuse hypoperfusion at 5 days of life, marked hyperperfusion in the dysplastic hemisphere at 15 days, and almost bilaterally symmetrical perfusion at 2 months. In our cases, the SPECT findings were not always related to the intensity of clinical symptoms or the EEG paroxysms. Many MRS findings, even for those cases with extensive CD, showed almost normal patterns for the three peaks (choline, creatine, and NAA). As to the ratio of NAA to choline in the CD cortex, only two cases showed significantly low values (<1.0), and the other two cases exhibited relatively lower values in the CD cortex compared with those in the contralateral cortex. Abnormal peaks, such as for free lipids, were not observed in all cases. There was no relation between the MRS findings and the rCBF with SPECT. Conclusions: In children with epilepsy due to CD, the severity of clinical symptoms, both psychomotor handicaps and epilepsy, is closely related to the extent of the CD. However, their SPECT findings suggest that their pattern of brain perfusion does not directly reflect their clinical features. Interestingly, the proton MRS findings for almost every patient with CD were the normal pattern. This result means that the cerebral lesions of CD have some neuronal activities. Much work needs to be done to establish the relation between the clinical and proton MRS features of children without epilepsy with CD.