Abstracts

Single-photon emission computed tomography (SPECT) is a valuable tool in the localization of focal onset epilepsy. However, radiotracer injection latency is a difficult to control variable which may dramatically affect the interpretability of these images. We have investigated the effect of radiotracer injection latency on the localizing and lateralizing sensitivity of ictal-interictal SPECT images. Scans obtained from patients injected during seizures, based on video/EEG review, were defined as ictal, while those obtained after seizures were post-ictal. We used SPM2 to analyze ictal-interictal SPECT image pairs from 28 consecutive temporal lobe, and 16 extra-temporal epilepsy patients. Seizure localization was based on concordance of EEG, MRI and clinical data. For each patient, ictal-interictal SPECT differences were compared to scan pairs obtained from 14 healthy normal subjects using SPM2. We investigated the localizing value of increased perfusion based on the lobe containing the largest number of voxels with significant hyperperfusion. Our investigation of hypoperfusion found that it was not helpful for localizing the lobe of seizure onset, so we instead focused our studies on its sensitivity in lateralizing the hemisphere of seizure onset. In 28 patients with mesial temporal lobe epilepsy, we obtained 7 ictal and 21 post-ictal injections; in 16 patients with extra-temporal epilepsy we obtained 7 ictal and 9 post-ictal injections. With true ictal SPECT injections, perfusion increases correctly localized the lobe of seizure onset in 6 of 7 temporal lobe epilepsy patients (86%), and in 6 of 7 extra-temporal lobe patients (86%). Considering only scans obtained at post-ictal injection times, perfusion increases correctly localized only 5 of 21 (24%) temporal lobe patients, and 1 of 9 (11%) in extra-temporal patients. For post-ictal injections, we found that the hemisphere with greater hypoperfusion correctly lateralized the side of seizure onset in 17 of 21 temporal lobe patients (81%) and 7 of 9 extra-temporal patients (78%). These results demonstrate that ictal-interictal SPECT analysis is highly sensitive for localizing both temporal and extra-temporal epilepsy. However, post-ictal injections have very poor localizing value based on conventional image interpretation. Nevertheless, when total hypoperfusion is compared between the two hemispheres, more extensive hypoperfusion can be helpful for lateralizing the side of seizure onset even with late post-ictal injections. (Supported by Epilepsy Foundation of America Merritt-Putnam Clinical Research Fellowship)