Abstracts

Surgery for extra-temporal [amp] non-mesial temporal epilepsy often fails due to epileptic regions that are expressed only after a primary focus is resected. This may represent a network pathophysiologic process. In 69 patients who underwent diagnostic invasive monitoring prior to therapeutic resection, intracranial electrodes were replaced for a 2nd invasive monitoring [amp] further resection. These 3 stage cases were retrospectively reviewed. Efficacy, complications, seizure characterization, rationale for, and findings supporting network behavior is described. All procedures were between 1994 [amp] 2003. Each patient was managed by a multidisciplinary team [amp] had a typical epilepsy surgery evaluation. A 2 stage procedure was initially planned, but data from the initial invasive monitoring compelled the additional diagnostic [amp] therapeutic stage. Engel score was used to describe efficacy. Outcome data at a minimum of 1 yr from surgery was available for 95%. All cases had non-concordant VEEG, MRI, Wada, [amp] physiologic imaging, or had broad VEEG onsets. There were 30 females. The ave age was 25.4 yrs. In 50/69 (72.5%) additional resection was performed at the 3rd operation. For Engel Class 1, 2, 3, [amp] 4 there were respectively 36 (55%), 10 (15%), 13 (20%) [amp] 7 (10%), of 66 cases (3 were unavailable.) In 24 cases 1 lobe was involved, in 33 two lobes, in 11 three lobes [amp] in 1 four lobes. Pathology was cortical dysplasia or neuronal migration defect in 35, nonspecific in 12, gliosis in 8, cavernous hemangioma in 2, old infarct in 2, low grade neoplasm in 1, hamartoma in 1, foreign body reaction in 1, [amp] hippocampal sclerosis in 3. There were 29 right, 36 left, [amp] 4 bilateral seizure onsets. Neural network node combinations defined by invasive monitoring were temporal frontal (23%), frontal parietal (13%), temporal only (10%), [amp] temporal parietal (4%).
Complications were: 2 infections, 1 hydrocephalus, 1 motor/sensory deficit, 2 bone plate resorptions, 1 DVT, [amp] 1 pneumonia. There were no deaths, hemorrhages, subdural hygromas, pathologic ICP, permanent occulo-cranial nerve injuries, significant language or memory deterioration. The population described are difficult cases since 45/69 (65%) had multi-lobar seizure onsets [amp] only 7 had sole temporal onsets, characterizing the population as extra-temporal or temporal plus syndromes. Histopathology included mesial sclerosis in only 3. Three stages can be considered when there is suspicion of complex, multi-lobar, or non-contiguous multi-foci seizure onset. Three stage procedures have acceptable risks. Many of these patients can be sub-grouped into 1 of 5 specific epilepsy networks suggested by Spencer (1). Since the network theory predicts variable nodes of noncontiguous seizure onsets, 3 stage surgery may be a viable treatment option.
(1) Susan Spencer, Neuronal Networks in Human Epilpesy: Evidence of and Implications for Treatment. Epilepsia, 43(3):219-227, 2002.