Refractory epilepsy: Comparison of MR imaging, CT, and histopathologic findings in 117 patients

PURPOSE: We compared computed tomography (CT) and magnetic resonance (MR) imaging for the detection of abnormalities underlying epilepsy. MATERIALS AND METHODS: CT and MR imaging findings in 117 patients (56 female, 61 male patients; age at surgery, 12-56 years) who underwent surgery for medically refractory epilepsy were compared with histopathologic findings by using the McNemar and chi(2) statistics. RESULTS: Sensitivities for detecting abnormalities were 95% (104 of 109) for MR imaging and 32% (35 of 109) for CT; specificities were 87% (13 of 15) for MR imaging and 93% (14 of 15) for CT (P <.001 for MR versus histopathologic findings). In the subgroup of 113 patients with solitary findings, MR imaging depicted an abnormality at the surgical site in 86% (n = 97) of 113 patients compared to 28% (n = 32) for CT (P <.001). In this same subgroup, histopathologic findings were predicted by using MR imaging in 88% (n = 99) of 113 patients versus 35% (n = 40) with CT (P <.001). Multiple findings were observed in 3% of CT (three of 117) and 17% of MR (20 of 117) images. CONCLUSION: CT has no role in the diagnostic evaluation of medically refractory epilepsy. Even in patients with medically controlled epilepsy, use of less costly CT instead of MR imaging seems imprudent.