Background and Rationale: Accurate identification of the side of epileptogenic (seizure) focus among temporal lobe epilepsy is vital if surgical intervention (temporal lobectomy) is to be successful. A number of pre-surgical evaluations contribute to the determination of seizure focus including the neurological examination, ictal and inter-ictal EEG, neuropsychological testing, the WADA test, brain imaging (e.g., MRI, SPECT, PET), and speech pathology assessment. Although some studies have consistently found that pre-surgical neuropsychological impairments relate to side of seizure focus, other studies have not. Nevertheless, pre-surgical neuropsychological impairment can provide important information regarding the side of epileptogenic focus in the context of other assessment procedures.

Method: The current study examined the ability of neuropsychological data to predict the side (left vs. right) of temporal lobe resection in patients with intractable temporal lobe epilepsy. Subjects were 66 patients (33 male, 33 female) who were tested pre-surgically as part of an evaluation for possible surgical treatment of intractable epilepsy. They were given a comprehensive neuropsychological battery including the Wechsler Adult Intelligence Scale - R, Wechsler Memory Scale, Halstead-Reitan Neuropsychological Battery (yielding the Halstead Impairment Index), Trails A & B, Controlled Oral Word Association Test, Boston Naming Test, Grip Strength, Finger Tapping Test, and Aphasia Screening. Based on the correlation matrix, scores on selected measures were entered into a logistic regression analysis in an attempt to predict side of epileptogenic focus. A backward stepwise regression model with a p-to-remove =.10 was employed for variable selection.

Results: Patients with left (N=33) or right (N=33) temporal lobe resection did not differ in terms of age, sex, education, or estimated IQ. Using a weighted combination of these scores on a selected number of tests, we were able to correctly classified the side of epileptogenic focus in 55 of 65 cases. Range of scores was -10.9 (left lesion) to +7.25 (right lesion). Five of the ten incorrect classifications had a Logit score close to zero (between -0.9 and +0.9). Overall predictive power was 87.87% and negative predictive power was 81.81%.

Conclusions: The findings suggest that pre-surgical performance on a select number of neuropsychological measures can be predictive of the side of temporal lobe resection. A selective battery of neuropsychological tests may provide sufficient information to make a decision regarding side of temporal lobe resection. However, other considerations (for example, emotional issues, impact of proposed surgery on intellectual function and memory abilities) will still require the administration of other neuropsy- chological tests in addition to those used for predicting the side of the focus.

www.umich.edu/~gusb/predictionINS00.html