Abstracts

Rationale: PET has been shown to be a sensitive, though not specific, diagnostic test in relation to the decision to operate or the localization of the epileptogenic zone. Some authors have suggested that it is more economical to forego routine PET administration in favor of selective use in cases where MRI is non-localizing or there are discordant imaging and neurophysiological findings. The purpose of the present retrospective analysis was to investigate the role of PET as part of the decision to perform invasive monitoring prior to temporal lobectomy in patients with epilepsy. The possibility of differential influence based on the side of surgery was also examined. Methods: The medical records of 63 temporal lobectomy patients consecutively referred to the LSUHSC Epilepsy Center of Excellence from 1994 to 2005 were reviewed. A breakdown of patients by side of surgery (right or left) and then MRI results (localizing or normal) was used to determine the relative role of PET in the decision to perform invasive monitoring prior to temporal lobectomy and investigate possible differences in applicability based on the side of surgery. Results: In RTL patients, 76% had a localizing MRI and 37% of these had invasive monitoring. The percentage of patients that had PET in the non-invasive subgroup was 33.3% compared with 43% of the invasive sub-group. In LTL patients, 87.5% had a localizing MRI and 54% of these had invasive monitoring. Only 31% of the non-invasive sub-group had PET compared with 67% of the invasive sub-group. In RTL patients with a normal MRI (6 patients; 24%), all patients (100%) had invasive monitoring but only half had a PET (2 right abnormalities, 1 normal). In LTL patients with a normal MRI (4 patients; 12.5%), half (50%) had invasive monitoring and only one-fourth had a PET (1 left abnormality). In RTL, a higher percentage of patients had invasive monitoring when the MRI was non-localizing (100% non-localizing vs. 37% localizing). In LTL, approximately the same percentage of patients had invasive when the MRI was non-localizing (50%) as when it was localizing (54%). Conclusions: Our data suggests that PET has minimal influence on the decision to perform invasive monitoring prior to temporal lobectomy regardless of MRI findings or side of surgery. When the MRI is non-localizing, the decision to perform invasive monitoring may show a bias by side with a right ictal EEG focus requiring invasive procedures more often than a left ictal EEG focus. However, the small number of patients with a non-localizing MRI makes interpretation difficult.