Abstracts

In the surgical candidate with intractable complex partial seizures, a neurocognitive profile that lateralizes to the suspected temporal lobe (TL) provides confirmatory evidence of the primary seizure focus. In some patients, however, there is cognitive evidence of involvement of the contralateral hemisphere. The etiologies of the collaborative lateralized cognitive changes have been well addressed; however, little information is known regarding the cognitive changes in the contralateral temporal lobe. Our research addresses a) the relationship of bilateral cognitive changes in patients with documented lateralized temporal lobe onset to presence and type of initial precipitating injury (IPI) and b) the surgical prognostic value of the bilateral neurocognitive deficits. Subjects were 316 surgical candidates with intractable epilepsy whose primary seizure onset was localized to the left (140) or right TL(176). Independent variables were presence or absence of IPI and type of IPIs (i.e., prolonged seizure, brief seizure, or a nonseizure event). The preoperative neurocognitive data included Verbal IQ (VIQ), Performance IQ (PIQ), the Boston Visual Confrontation Naming Score, and a Hard Word-Pairs Verbal Learning score from the WMS. Seizure outcome after surgery was based on a 1-4 rating scale, with R1 being seizure free, R2 fewer than 6 seizures per year, R3 between 0.5 and 2 seizures per month, and R4 more than 2 seizures per month. The group of patients with LTL seizures, as compared with the group of with RTL seizures, had lower VIQs (p[lt].02), Naming scores (p[lt].002), and Verbal Learning scores (p[lt].0001). In the LTL group, neither the presence or absence of an IPI or the type of IPI related to the cognitive scores. However, in the RTL group, a prolonged seizure as the type of IPI related to lower Verbal Learning scores , but not to the Naming scores or IQ scores. In RTL patients, an IPI of a prolonged seizure was more predictive of lower Verbal Learning scores than either brief seizures (p=.05) or nonseizure injuries (p[lt].03). Also, in the RTL group, using a logistic regression model, Verbal Learning scores had stronger predictive value of surgical outcome than a history of a prolonged seizure. Lower Verbal Learning scores in RTL patients were associated with a lower probability of surgical seizure control (p[lt].03). A history of a prolonged seizure as the IPI is associated with the presence of poorer verbal learning scores in patients with RTL seizures. This finding suggests that the precipitating seizure event resulted in bilateral hippocampal damage. The extent of verbal learning loss in the RTL indicates the degree of contralateral LTL involvement and relates to a poorer surgical seizure outcome. (Supported by RO1NS31277 (RR); RO1NS38992(GWM); PO5NS02808(GWM))