Comparison of SPECT, EEG, CT, MRI, and pathology in partial epilepsy

Abstract

Twenty children with partial epilepsy who had surgery between the ages of 4$\text{1}\text{2}$ months and 18 years were studied preoperatively with electroencephalography (EEG), computed tomography (CT), and technetium-99m hexamethylpropyleneamineoxime ^99mTc-HmPAO single photon emission computed tomography (SPECT; 20 interictal, 4 postictal). Fourteen had magnetic resonance imaging (MRI). All had an epileptiform focus (12 unilateral, 8 predominantly unilateral) on EEG. The combination of interictal and postictal regional cerebral blood flow (rCBF) abnormalities alone correlated with EEG foci in 16 of 20 patients. Interictal rCBF abnormalities correlated with EEG foci in 14 of 20. CT findings correlated with EEG foci in 14 of 20. MRI findings correlated with EEG foci in 13 of 14. Pathology demonstrated tumor in 6, cortical dysplasia in 4, mesial temporal sclerosis in 3, Sturge-Weber in 2, cavernous hemangioma in 1, Rasmussen encephalitis in 1, porencephalic cyst and gliosis in 1, and cysts found at surgery (but normal histology) in 2. Interictal and postictal SPECT, EEG foci, and CT findings each correlated with the pathology site in 17, 19, and 15 patients, respectively. MRI correlated with pathology site in 13 of 14 patients. Postictal and interictal abnormalities of rCBF correlated with EEG and pathology as frequently as CT. In 5 patients with normal CT scans and in 1 with a normal MRI, postictal and interictal rCBF correlated with EEG and pathology results; however, these 6 patients all had abnormalities on CT or MRI. SPECT, therefore, may be considered a valuable additional diagnostic procedure in the evaluation of epilepsy surgery candidates in that it adds to the evidence of abnormality at the involved site. EEG findings and structural neuroradiologic abnormalities, however, most influenced surgical decisions in this series.