Abstracts

Rationale: Epilepsy surgery has long been considered an effective, and possibly a last resort treatment for select epileptics who are intractable to medical control. Among these patients, those who are found to be eligible for resective surgery are fortunate in that they can expect to have a better outcome in terms of seizure freedom as compared to their non-resective surgical counterparts. Most institutions utilize EEG as their greatest tool in delineating focal lesions in resective surgical candidates, and employ various imaging modalities to reinforce the finding. We have employed the results of imaging modalities in pediatric patients who received successful epilepsy surgery in order to determine the accuracy of each imaging tools on identifying the epileptic zone. Methods: All pediatric patients who received resective epilepsy surgery in Severance Hospital between October2003 to April2008 were reviewed, and those with outstanding results, namely, who satisfy Class I of Engel’s Classification of postoperative outcome, were selected. Their EEG, MRI, PET, and Subtraction Ictal SPECT Coregistered to MRI(SISCOM) results were tabulated, and their accuracy in locating the epileptic foci, defined as “area that resulted in seizure ablation after resection”, were compared. Concordance with epileptic lesion was defined as an imaging test being able to localize the epileptic foci precisely, or failing that, lateralize the target to a certain hemisphere. Results: 42patients were enrolled (23 temporal lobectomy, 19 extratemporal resections). Seizure focus was delineated primarily by the use of EEG, hence, EEG results showed 100% concordance with the epileptic foci. MRI showed concordance in 84.2% of extratemporal cases, with precise localization of the lesion in all, while it was 91.3% in temporal cases, with localization in 78.1%. PET was particularly useful in localizing temporal lobe lesions, with concordance rate at 91.7% and localization in 72.7%, but its efficiency was lower in extratemporal cases, with concordance in 68.5% only. SISCOM, when successfully acquired, was outstanding, with concordance in 100% of temporal and 92.3% of extratemporal cases. It is, however, inferior to MRI in precise localization of epileptic foci, with localization in 66.7% of temporal and 84.6% of extratemporal cases. Conclusions: As a tool to supplement the electrographic results, imaging modalities are invaluable assets, especially if EEG results are confounding. Our experience shows that MRI is invariably reliable in all cases of focal epilepsies. PET results are as reliable as MRI in lateralizing the epileptic area in temporal lesions, while its efficacy is lower in extratemporal cases. SISCOM can be counted on to localize the foci in extratemporal cases, while its efficacy is much lower in temporal lesions. Thus, in cases of conflicting pre-surgical results, MRI, with supplementary data from PET, would help the physician in establishing correct decision in temporal lobe epilepsies, while utilization of SISCOM and MRI data would be wiser in extratemporal cases.