Abstracts

RATIONALE: There have been numerous attempts to establish a valid neuropsychological measure of nonverbal memory that is sensitive to the effects of right temporal lobe dysfunction in presurgical subjects. Figural reproduction measures, including the Rey Complex Figure and the Wechsler Memory Scale designs, have been found to relatively insensitive in this regard. More positive findings have been reported in studies using measures of figural learning over repeated trials. The purpose of this study was examine the use of a commercially measure of visuospatial learning in a sample of epilepsy patients undergoing neuropsychological testing as part of a comprehensive presurgical workup.
METHODS: The Brief Visuospatial Memory Test [ndash] Revised (BVMT-R) was administered to 47 subjects (22 male, 25 female) undergoing VEEG monitoring. All subjects were found to have electrophysiologic evidence of [dsquote]probable[dsquote] unilateral temporal lobe onset, as determined by video EEG recordings of ictal and interictal epileptiform abnormalities. Primarily left side abnormalities (LTL) were detected in 25 subjects. Right side abnormalities (RTL) were identified in 22 subjects. The mean age of the sample was 35.1 years (SD = 13.5). The mean age at onset of epilepsy was 22.4 years (SD = 12.7). Mean years of education was 13.9 years. Full Scale IQ for the sample was 94.1. The LTL and RTL groups did not differ on any of these variables.
The BVMT-R consists of 6 geometric designs represented on an 8[dsquote] x 11[dsquote] page. The designs are shown for a 10 second period, followed by an immediate reproduction of the designs in their correct location on the page. This procedure is repeated over three learning trials. Subjects are asked to reproduce the designs for a fourth time, approximately 30 minutes following the learning trials. This is followed by yes/no recognition testing, with 6 distractor foils. All reproductions are scored according to standardized criteria. The test includes 6 repeatable forms. The same form (Form 1) was used for all subjects in this study.
RESULTS: Differences between mean BVMT-R scores from the LTL and RTL groups were assessed with a series of independent t-tests. No group differences were found in analyses of any of the learning, recall, or recognition measures. Receiver operating characteristic (ROC) curve analyses were used to determine empirical cutoff scores for maximal separation of the LTL and RTL groups. Cutoff scores below16.5 (out of a possible 36) over the three learning trials identified RTL subjects with 32% sensitivity and 68% specificity. Delayed recall scores lower than 6 (out of 12) were associated with 23% sensitivity and 68% specificity.
CONCLUSIONS: The results of this study indicate that LTL and RTL subjects exhibit similar patterns of performance on the learning, delayed recall, and recognition trials of the BVMT-R. No differences in scores were obtained in comparison of group means or in examination of rates of impairment using empirically derived cutoffs. The findings suggest that the BVMT-R, like many other tests, lacks sensitivity to detecting the nonverbal memory deficits reported in patients with right temporal lobe epilepsy.