Neuropsychology 2003, Vol. 17, No.

1, 59 68

Copyright 2003 by the American Psychological Association, Inc. 0894-4105/03/$12.00 DOI: 10.1037/0894-4105.17.1.59

Material-Specic Memory in Temporal Lobe Epilepsy: Effects of Seizure Laterality and Language Dominance
Hongkeun Kim
Daegu University

Sangdoe Yi and Eun Ik Son

Keimyung University

Jieun Kim
Catholic University of Daegu
This study investigated the effects of seizure laterality and language dominance on materialspecic memory in temporal lobe epilepsy (TLE). Left TLE (LTLE) patients with lefthemisphere language dominance (LHLD) showed signicantly higher nonverbal than verbal memory capacity, whereas right TLE patients with LHLD showed signicantly better verbal than nonverbal memory capacity. LTLE patients with non-left-hemisphere language dominance (NLHLD) showed signicantly better verbal memory capacity compared with LTLE patients with LHLD. Thus, selective verbal or nonverbal memory decits that are dependent on side of seizure onset were apparent in patients with LHLD but not in patients with NLHLD. Relative sparing of verbal memory capacity in LTLE patients with NLHLD may reect interhemispheric reorganization of verbal memory function.

Temporal lobe epilepsy (TLE) may be the most common form of human epileptogenic disorders (Engel, 1996). Preor postsurgical neuropsychological studies of TLE patients have greatly contributed to the current understanding of the brain behavior relationship. A classical nding in this eld is that surgical removal of the left mesial temporal structures (i.e., left temporal lobectomy) results in reduction of verbal memory capacity, whereas surgical removal of the right mesial temporal structures (i.e., right temporal lobectomy) diminishes nonverbal (or visuospatial) memory capacity. Consequently, left temporal lobectomy patients show superior nonverbal compared with verbal memory capacity, whereas right temporal lobectomy patients show better ability to learn verbal than nonverbal material. These ndings are consistent with the hypothesis that the learning of verbal material is dependent more on the left than the right mesial temporal structures, whereas the capacity to learn nonverbal material relies more on the right than the left mesial temporal structures (Milner, 1975). The common pathology underlying TLE is mesial temporal sclerosis and other abnormalities in the mesial temporal lobe structures (McMillan, Powell, Janota, & Polkey, 1987; Meencke & Veith, 1991; Wolf et al., 1993). Thus, it

Hongkeun Kim, Department of Rehabilitation Psychology, Daegu University, Daegu, South Korea; Sangdoe Yi, Department of Neurology, Keimyung University, Daegu, South Korea; Eun Ik Son, Department of Neurosurgery, Keimyung University; Jieun Kim, Department of Neurology, Catholic University of Daegu, Daegu, South Korea. This research was supported by a Daegu University research grant in 2002. Correspondence concerning this article should be addressed to Hongkeun Kim, Department of Rehabilitation Psychology, Daegu University, 2288 Daemyung-dong, Nam-gu, Daegu 705-714, South Korea. E-mail: hongkn@daegu.ac.kr 59

may be hypothesized that material-specic memory decits are present not only in postsurgical TLE patients but also in presurgical TLE patients, although to a lesser degree. However, prior investigations of this hypothesis have yielded, at best, mixed results. Only a relatively small proportion of these studies (Delaney, Rosen, Mattson, & Novelly, 1980; Fedio & Mirsky, 1969; Helmstaedter, Pohl, & Elger, 1995; Hermann, Wyler, Richey, & Rea, 1987; La ` davas, Umilta `, & Provinciali, 1979) have conrmed the hypothesis. In the majority of the relevant prior studies, the hypothesis has been only partially conrmed or not supported at all. Thus, certain studies (Bornstein, Pakalnis, & Drake, 1988; Cohen, 1992; Giovagnoli & Avanzini, 1999; Gold et al., 1995; Moore & Baker, 1996; Selwa et al., 1994) have found that the left TLE (LTLE) group shows verbal memory decits relative to the right TLE (RTLE) group but that there is no difference in nonverbal memory capacity between the two. On the other hand, others (Abrahams, Pickering, Polkey, & Morris, 1997; Baxendale, Thompson, & van Paesschen, 1998; Baxendale, van Paesschen, et al., 1998; Hermann, Seidenberg, Schoenfeld, & Davies, 1997; Loring, Hermann, Lee, Drane, & Meador, 2000) have reported that RTLE is associated with poor ability to learn nonverbal material relative to LTLE but that there is no difference in verbal memory capacity between the two. Still other studies (Ellis, Hillam, Cardno, & Kay, 1991; Glowinski, 1973; Hermann, Connell, Barr, & Wyler, 1995; Lendt, Helmstaedter, & Elger, 1999; Loring, Lee, Martin, & Meador, 1988; Mayeux, Brandt, Rosen, & Benson, 1980; Naugle, Chelune, Schuster, Lu ders, & Comair, 1994) have failed to nd a signicant difference in either verbal or nonverbal memory capacity between the LTLE and RTLE groups. One thing that is clear from these variable results is that any difference in material-specic memory capacity between the LTLE and RTLE groups is of relatively subtle magnitude.

60

KIM, YI, SON, AND KIM

Multiple factors may be involved in the lack of reliable material-specic memory differences between the LTLE and RTLE groups. One critical factor may be that not all verbal or nonverbal memory tasks are equally valid measures of hemisphere-dependent memory capacity. Thus, there is evidence that certain nonverbal memory tasks are more sensitive to the effects of RTLE than others (Abrahams et al., 1997; Breier et al., 1996; Giovagnoli, Casazza, & Avanzini, 1995; Helmstaedter et al., 1995; Loring, Lee, & Meador, 1988; Piguet, Saling, OShea, Berkovic, & Bladin, 1994; for a review, see Barr, 1997). The other important factor may be that not all TLE patients may show equally large material-specic memory impairment. For example, in Breier et al.s (1997) study, TLE patients were classied into reading-decient (RD) and non-reading-decient (nonRD) groups. They found that selective decits in verbal or nonverbal memory are evident in the non-RD group but not in the RD group. In Sawrie et al.s (2001) study, unilateral TLE patients were divided into groups with bilaterally normal hippocampal volumes, unilateral atrophy, or bilateral atrophy, on the basis of magnetic resonance imaging (MRI) volumetry. Results showed that verbal memory capacity is signicantly lower in LTLE relative to RTLE only in the group with bilateral atrophy. Thus, at least in certain studies, subject characteristics may have contributed to failures to nd signicant material-specic memory differences between LTLE and RTLE groups. The present study investigated the effects of another subject variable, hemispheric language dominance, on material-specic memory of TLE patients. Data from the intracarotid amobarbital procedure (IAP; Benbadis, Dinner, Chelune, Piedmonte, & Lu ders, 1995; Dinner, 1991; Kurthen et al., 1994; Loring et al., 1990) clearly show that patterns of hemispheric language dominance in TLE patients are highly diverse, although the most common type is left-hemisphere language dominance (LHLD). The diversity is found in both LTLE and RTLE patients, but it is especially pronounced in LTLE patients, reecting interhemispheric reorganization of language following early lefthemisphere damage (Rasmussen & Milner, 1977). It is generally accepted that hemispheric language dominance is an important variable in hemispheric specialization research. Nevertheless, only a few studies have actually addressed the question of whether TLE patients with LHLD versus non-left-hemisphere language dominance (NLHLD) show differential material-specic memory capacity. Helmstaedter, Kurthen, Linke, and Elger (1994) found that LTLE patients with right-hemisphere language dominance (RHLD), despite left mesiotemporal dysfunction, show better verbal than nonverbal memory performance. Rausch, Boone, and Ary (1991) reported that LTLE patients with RHLD do not exhibit verbal memory loss following left temporal lobectomy, which is evident in LTLE patients with LHLD. The authors suggested that in LTLE patients with RHLD, verbal memory as well as general language functions may be supported primarily by the right hemisphere. Glosser, Saykin, Deutsch, OConnor, and Sperling (1995) investigated lateralization of material-specic memory processing in TLE patients using an IAP memory test.

In this study, TLE patients with LHLD demonstrated lefthemisphere specialization for verbal memory and righthemisphere specialization for nonverbal memory. In contrast, TLE patients with NLHLD showed no lateral specialization for either verbal or nonverbal memory processing, indicating interhemispheric reorganization of memory functions. On the basis of these ndings, we hypothesized that selective verbal or nonverbal memory decits that are dependent on side of seizure onset would be apparent in TLE patients with LHLD but not in TLE patients with NLHLD.

Method Subjects
The subjects were 92 patients who ultimately underwent left or right mesial temporal lobe surgery for treatment of medically intractable epilepsy of unilateral temporal origin. They were selected from a consecutive series of patients undergoing temporal lobe surgery in an epilepsy surgery center who met the following criteria: (a) completion of relevant memory measures as a part of presurgical workup, (b) age greater than 15 years, and (c) seizurefree or near seizure-free status (Class I or II; Engel, van Ness, Rasmussen, & Ojemann, 1993) at 1-year postsurgery follow-up. The last criterion was included as a means of establishing an ultimate standard of the correct localizationlateralization of the epileptogenic foci. Of the 92 patients, 48 had left mesiotemporal foci and 44 had right mesiotemporal foci. Fifty-nine were males, and 33 were females. Eighty-three were right-handed, and 9 were left-handed, as determined by a 13-item questionnaire (Chapman & Chapman, 1987). The mean age at the time of testing was 28.9 years (SD 6.8). The mean age of recurrent seizure onset was 13.6 years (SD 6.7). All patients had their seizure onset lateralized to a single mesial temporal lobe in a presurgical evaluation. The presurgical evaluation included seizure semiology, prolonged interictal and ictal video electroencephalograph (EEG) from scalpsphenoidal electrodes, MRI scanning, interictal single photon emission tomography, IAP, neuropsychological testing, and if necessary, EEG from chronically implanted bilateral subdural strip electrodes.

Memory Testing
All patients were administered the ReyKim Memory Test (RKMT; Kim, 1999) as a part of presurgical examination. In the RKMT, verbal memory performance is assessed by the K-Auditory Verbal Learning Test (KAVLT), a Korean version of the Rey Auditory Verbal Learning Test (Rey, 1964), and nonverbal memory performance is assessed by the K-Complex Figure Test (KCFT), a Korean version of the Rey Complex Figure Test (Rey, 1941). The KAVLT required serial learning of a list of 15 unrelated words over ve consecutive trials, each trial followed by immediate recall. After a delay period of 20 min, the patient was again required to recall the 15 words. Following completion of the delayed recall, the patient was presented with a list of 50 words and required to choose a total of 15 words that were in the original list spoken by the examiner. The KCFT was essentially identical to a standard version of the Rey Complex Figure Test (Lezak, 1983). The patient was required to copy the gure as accurately as possible with no time limit imposed. An immediate recall trial was administered following completion of the gure copy. After a delay period of 20 min, the patient was again required to recall the gure. All KCFT productions were scored according to a standard version of the 36-point scoring system (Lezak, 1983).

MATERIAL-SPECIFIC MEMORY IN TEMPORAL LOBE EPILEPSY

61

IAP
All patients underwent IAP as a part of presurgical evaluation. The IAP was conducted, with the patient supine, immediately following angiography. Amobarbital, 125 mg in a 10% solution, was injected into the internal carotid artery using a transfemoral catheter over a 4 5-s interval. In most patients, the side considered for resection was injected rst. Left- and right-hemisphere injections were done on the same day with a minimum of 40 min between the two injections. Following demonstration of hemiplegia, the patient was presented with a series of language tasks that lasted approximately 60 90 s. Assessment of language functions during the course of the IAP has varied slightly over the years but has always included both receptive and expressive language tasks. The core tasks were as follows: following a simple command (e.g., close your eyes), reading a short sentence, naming a picture, and repeating a phrase spoken by the examiner. On the basis of the IAP language test, language dominance was dichotomized into two types, LHLD and NLHLD. A conservative criterion for LHLD was adopted because it was thought to be critical for proper testing of the research hypotheses. Thus, only patients who displayed exclusive left-hemisphere competence in all or nearly all administered language tasks were classied as having LHLD. More specically, there were 61 patients who failed all language tasks following left-hemisphere injection but performed all language tasks correctly following right-hemisphere injection. These patients were classied into the LHLD group. In addition, there were 3 patients who failed all language tasks following left-hemisphere injection and performed all but one language task correctly following right-hemisphere injection or who failed all but one language task following left-hemisphere injection and performed all language tasks correctly following right-hemisphere injection. These 3 patients, considered as having near exclusive left-hemisphere representation of language, were also classied into the LHLD group. All other patients (n 28) were classied into the NLHLD group. The incidence of NLHLD was 50% in LTLE patients but only 9% in RTLE patients. This difference in the incidence of NLHLD between LTLE and RTLE patients was highly signicant (Fishers exact test, p .001).

Statistical Analyses
We composed four groups on the basis of the cross-classication of patients based on language dominance and seizure laterality: LHLDLTLE (n 24), LHLDRTLE (n 40), NLHLD LTLE (n 24), and NLHLDRTLE (n 4). The NLHLDRTLE group was excluded from further analyses because of the extremely small sample size, although the data for this group are presented for descriptive purposes. The data from the remaining three groups were analyzed in two separate analyses of variance (ANOVAs), each restricted to two groups. The rst ANOVA, restricted to the LHLDLTLE and LHLDRTLE groups, was performed to investigate the effects of seizure laterality on material-specic memory in patients with LHLD. The second ANOVA, restricted to the LHLDLTLE and NLHLDLTLE groups, was performed to investigate the effects of language dominance on material-specic memory in LTLE patients. Two types of analyses were performed, one involving individual trial scores and the other involving aggregate measures of individual trial scores. The aggregate measures, referred to as Z scores, were obtained by converting raw scores of all relevant trial scores to z scores (based on the entire study sample) and then averaging the z scores. The KAVLT Z score was derived from the seven learning and memory trials of the KAVLT (Trials I to V, delayed recall, and delayed recognition), and the KCFT Z score was derived from the two memory trials of the KCFT (immediate recall and delayed recall). The analyses of individual trial scores were performed to explore the possible source differences in the aggregate measures. The analyses of the aggregate measures were given primary importance for the following reasons. First, the aggregate measures were inherently more comprehensive than individual trial scores, allowing a more focused test of the research hypotheses. Second, the aggregate measures permitted a direct statistical comparison (i.e., paired t test) of verbal versus nonverbal memory scores within a group, whereas this was impossible for individual trial scores because of noncomparable units of measurement. Finally, the analyses of the aggregate measures involved fewer statistical tests than the analyses of individual trials scores. Thus, the former was less vulnerable to experimentwise Type I error than the latter. The RKMT (Kim, 1999) yields scales for each KAVLT and KCFT trial, which provide normalized standard scores with a normative mean of 10 and a standard deviation of 3. Aggregate measures of these scaled scores were obtained by averaging the scaled scores of all relevant trials. All statistical analyses reported in this article were performed twice: directly on raw scores and Z scores and on the scaled scores and aggregate scaled scores. The pattern of statistical signicance was highly comparable with raw scores and with scaled scores. Thus, for simplicity of presentation, the following description of statistical results will be limited to the analyses of raw scores, unless otherwise noted.

Surgical Pathology
A tailored temporal lobectomy, including both anterolateral temporal resection and amygdalohippocampectomy, was performed on all patients. We routinely varied the extent of hippocampal resections, depending on the intraoperative interictal epileptiform abnormalities on preresection electrocorticography (ECoG) recorded from the hippocampus and parahippocampal gyrus (Son, Howard, Ojemann, & Lettich, 1994). After initial removal, postresection recording of ECoG from the remaining hippocampus and parahippocampal gyrus was repeated to decide whether further hippocampal resection was needed. No subjects had major neurologic complications following surgery. A pathologic diagnosis of the resected mesial temporal lobe was available for 88 of 92 patients as follows: 60 (68%) had mesial temporal sclerosis (with or without other lesions), 4 (5%) had tumors, 11 (13%) had other pathology, and 13 (15%) had no abnormality detected (for more detailed classication of pathology, see Table 1). The proportion of mesial temporal sclerosis in the present sample corresponds well with estimations of mesial temporal sclerosis in other series of temporal lobe surgery (McMillan et al., 1987; Meencke & Veith, 1991).

Results Demographic and Clinical Variables

Demographic and clinical characteristics of the sample are presented in Table 1, separately for the LHLD LTLE, LHLDRTLE, NLHLDLTLE, and NLHLDRTLE groups. Group differences were examined by F tests for continuous variables and by 2 2 Fisher exact tests for categorical variables. Statistical analyses revealed no signicant differences between the LHLDLTLE and LHLD RTLE groups in terms of age, education, gender ratio,

62

KIM, YI, SON, AND KIM

Table 1 Demographic and Clinical Characteristics of the Sample

LHLD LTLE (n 24) Characteristic Age (years) Education (years) Gender (% male) Handedness (% right-handed) Onset age of recurrent seizures (years) Duration of seizures (years) Febrile convulsions (%) 1-year postsurgery outcome (% Class I) MQ (RKMT) FSIQ (KWAIS) Neuropathology (n) MTS MTS with cortical dysplasia MTS with microdysgenesis Cortical dysplasia Microdysgenesis Tumors Arteriovenous malformation Fibrous nodule Schizencephaly Paragonimiasis Astrogliosis No abnormality detected No report available M 30.8 11.5 SD 5.7 3.1 LHLD RTLE (n 40) M 28.6 11.2 63 98 14.0 14.5 38 78 81.1 82.2 18 4 4 1 1 1 2 0 0 0 0 6 3 14.3 12.7 85.9 81.5 14 3 1 1 0 2 0 0 0 0 0 2 1 7.4 7.1 10.5 17.2 67 75 14.8 14.9 90.2 86.2 3 0 0 0 0 0 0 0 1 0 0 0 0 SD 7.2 3.2 NLHLD LTLE (n 24) M 27.8 11.0 63 71 4.9 9.0 12.5 14.5 50 75 16.1 13.6 SD 7.0 2.5 NLHLD RTLE (n 4) M 27.0 10.2 50 75 7.2 11.8 SD 6.6 2.8

71 100 16.2 6.0 14.6 7.1 42 63 83.3 14.2 85.2 13.3 7 2 4 1 1 1 0 1 0 1 1 5 0

Note. LHLD left-hemisphere language dominance; LTLE left temporal lobe epilepsy; RTLE right temporal lobe epilepsy; NLHLD non-left-hemisphere language dominance; MQ memory quotient; RKMT ReyKim Memory Test (Kim, 1999); FSIQ full-scale IQ; KWAIS KWechsler Adult Intelligence Scale (Yeom, Park, Oh, Kim, & Lee, 1992); MTS mesial temporal sclerosis.

handedness, onset age of recurrent seizures, duration of seizures, incidence of febrile convulsions, 1-year postsurgery outcome (percent Class I), memory quotient, full-scale IQ, and incidence of mesial temporal sclerosis ( ps .10). The LHLDLTLE and NLHLDLTLE groups were significantly different in two variables, handedness and onset age of recurrent seizures. The incidence of left-handedness was signicantly higher in the NLHLDLTLE group (Fishers exact test, p .009). The onset age of recurrent seizures was signicantly earlier in the NLHLDLTLE group, F(1, 46) 12.47, p .01. These ndings replicate the corresponding results of several prior studies (e.g., Rasmussen & Milner, 1977; Satz, Strauss, Wada, & Orsini, 1988).

Effects of Seizure Laterality

Means and standard deviations for KAVLT and KCFT performance of the sample are presented in Table 2, separately for the LHLDLTLE, LHLDRTLE, NLHLDLTLE, and NLHLDRTLE groups. Both raw and scaled scores are listed in Table 2. As stated before, the pattern of statistical signicance was highly comparable with raw scores and with scaled scores. Thus, the following report of statistical results is limited to the analyses of raw scores.

The mean Z scores of the four groups are depicted in Figure 1. The Z scores of the LHLDLTLE and LHLDRTLE groups were analyzed using a 2 2 (Material Seizure Laterality) ANOVA. The only signicant effect in this ANOVA was a Material Seizure Laterality interaction, F(1, 62) 15.03, p .001. A priori contrasts using paired t tests indicated that the LHLDLTLE group had a signicantly better KCFT than KAVLT Z score, t(23) 3.38, p .01, one-tailed, whereas the LHLDRTLE group had a signicantly better KAVLT than KCFT Z score, t(39) 1.94, p .05, one-tailed (see Figure 1). Thus, the LHLDLTLE group showed relative verbal memory decit, whereas the LHLD RTLE group showed relative nonverbal memory impairment. The LHLDLTLE group had a signicantly higher KCFT Z score relative to the LHLDRTLE group, F(1, 62) 8.83, p .01. An examination of KCFT trial scores indicated that both immediate, F(1, 62) 10.38, p .01, and delay recall scores, F(1, 62) 6.67, p .05, were signicantly higher for the LHLDLTLE group. The LHLDRTLE group showed a better KAVLT Z score than the LHLDLTLE group, but this did not reach statistical signicance, F(1, 62) 1.27, p .20. No KAVLT trial scores were significantly different between the two groups ( ps .05). Thus, the present sample of LHLDLTLE and LHLDRTLE patients showed a signicant difference in nonverbal memory

MATERIAL-SPECIFIC MEMORY IN TEMPORAL LOBE EPILEPSY

63

Table 2 Memory Scores of the Sample

LHLD LTLE (n 24) Measure Z score KAVLT KCFT KAVLT score Trial I Trial II Trial III Trial IV Trial V Delayed recall Delayed recognition KCFT score Immediate recall Delayed recall Scaled aggregate score KAVLT KCFT Scaled KAVLT score Trial I Trial II Trial III Trial IV Trial V Delayed recall Delayed recognition Scaled KCFT score Immediate recall Delayed recall M 0.26 0.39 4.3 6.2 7.6 7.6 8.5 5.4 11.7 18.2 17.7 6.8 8.8 8.8 7.5 7.0 5.7 6.2 5.0 7.4 8.8 8.8 SD 0.84 0.89 1.7 1.9 2.1 3.1 2.8 3.2 2.5 6.6 5.9 2.6 3.1 2.8 2.8 2.8 4.1 3.5 3.6 3.2 3.2 3.1 LHLD RTLE (n 40) M 0.03 0.31 4.6 6.4 7.6 8.4 9.7 6.8 12.2 12.6 13.6 7.5 6.5 9.2 7.6 7.0 6.7 7.5 6.3 7.9 6.2 6.8 SD 0.77 0.94 1.2 2.2 2.1 2.5 2.5 3.3 2.6 6.8 6.3 2.7 3.3 2.4 3.0 3.2 3.7 3.2 3.8 4.0 3.4 3.3 NLHLD LTLE (n 24) M 0.23 0.13 5.0 7.0 8.5 9.5 10.1 7.3 12.9 16.5 16.0 8.0 7.7 9.0 8.2 8.0 7.9 7.8 6.8 8.5 7.8 7.5 SD 0.69 1.00 1.5 1.6 2.0 2.2 2.2 3.4 1.9 7.0 6.7 2.6 3.4 2.6 2.6 3.1 3.1 3.2 4.1 3.3 3.3 3.7 NLHLD RTLE (n 4) M 0.46 0.00 5.0 7.7 9.0 9.2 10.7 9.7 13.5 16.8 14.0 9.0 7.6 10.0 9.5 8.0 7.5 8.5 9.8 10.0 8.0 7.3 SD 1.18 1.01 2.3 1.8 3.3 2.8 5.2 3.9 2.3 8.4 5.5 2.8 4.0 2.6 1.3 4.1 3.1 5.8 3.0 2.4 4.3 3.6

Note. LHLD left-hemisphere language dominance; LTLE left temporal lobe epilepsy; RTLE right temporal lobe epilepsy; NLHLD non-left-hemisphere language dominance; KAVLT K-Auditory Verbal Learning Test; KCFT K-Complex Figure Test.

capacity in the expected direction but no signicant difference in verbal memory capacity.

Effects of Language Dominance

The Z scores of the LHLDLTLE and NLHLDLTLE groups were analyzed using a 2 2 (Material Type of Language Dominance) ANOVA. The only signicant effect in this ANOVA was a Material Type of Language Dominance interaction, F(1, 46) 6.84, p .05. A priori contrasts using paired t tests indicated that the LHLD LTLE group had a signicantly better KCFT than KAVLT Z score, as described above. In contrast, the NLHLDLTLE group showed no signicant difference in the ability to learn verbal versus nonverbal material, t(23) 1 (see Figure 1). Thus, the LHLDLTLE group showed a relative verbal memory decit, whereas the NLHLDLTLE group showed relatively equal ability to learn verbal versus nonverbal material. The NLHLDLTLE group had a signicantly better KAVLT Z score relative to the LHLDLTLE group, F(1, 46) 5.00, p .05, indicating relatively spared verbal memory capacity. An examination of KAVLT trial scores showed a signicantly higher performance by the NLHLD

Figure 1. Mean Z scores for two types of material in left temporal lobe epileptic patients with left-hemisphere language dominance (LHLDLTLE; n 24), right temporal lobe epileptic patients with left-hemisphere language dominance (LHLDRTLE; n 40), left temporal lobe epileptic patients with non-left-hemisphere language dominance (NLHLDLTLE; n 24), and right temporal lobe epileptic patients with non-left-hemisphere language dominance (NLHLDRTLE; n 4). Error bars represent the standard error of the mean. KAVLT K-Auditory Verbal Learning Test; KCFT KComplex Figure Test.

64

KIM, YI, SON, AND KIM

LTLE group in two trial scores: Trial IV, F(1, 46) 5.61, p .05, and Trial V, F(1, 46) 4.45, p .05. The NLHLDLTLE group had a lower KCFT Z score relative to the LHLDLTLE group, but this was not statistically signicant, F(1, 46) 1. An examination of KCFT trial scores revealed no signicant difference between the NLHLD LTLE and LHLDLTLE groups. Thus, the NLHLDLTLE group showed a signicantly higher verbal memory capacity compared with the LHLDLTLE group, but the two groups were similar in nonverbal memory capacity.

Effects of Etiology
The present study involved both patients with evidence of mesial temporal sclerosis (MTS; n 60) and patients without evidence of mesial temporal sclerosis (MTS; n 28). These two groups of patients were analyzed separately to investigate possible confounding because of inclusion of patients with different pathologic substrates. First, the effects of seizure laterality on material-specic memory were examined separately for LHLDMTS (n 39) and LHLDMTS patients (n 22). For LHLDMTS patients, the Z scores were analyzed using a 2 2 (Material Seizure Laterality) ANOVA. The only signicant effect in this ANOVA was a Material Seizure Laterality interaction, F(1, 37) 8.64, p .01. As may be seen in Figure 2, this interaction reected a signicantly higher KCFT than KAVLT Z score for LTLE patients, t(12) 2.83, p .01, one-tailed, and a trend toward higher KAVLT than KCFT Z score for RTLE patients, t(25) 1.28, p .10, one-tailed. An analogous ANOVA for LHLDMTS patients also

yielded a signicant Material Seizure Laterality interaction, F(1, 20) 5.56, p .05. This interaction reected a signicantly higher KCFT than KAVLT Z score for LTLE patients, t(10) 1.86, p .05, one-tailed, and a trend toward higher KAVLT than KCFT Z score for RTLE patients, t(10) 1.45, p .09, one-tailed (see Figure 2). Thus, LHLDMTS and LHLDMTS patients were similar in the effects of seizure laterality on material-specic memory performance. Second, the effects of language dominance on materialspecic memory were examined separately for LTLE MTS (n 31) and LTLEMTS patients (n 16). For LTLEMTS patients, the Z scores were analyzed using a 2 2 (Material Type of Language Dominance) ANOVA. In this ANOVA, the Material Type of Language Dominance interaction was signicant, F(1, 29) 4.83, p .05, and no other effects reached statistical signicance. As may be seen in Figure 3, this interaction reected a signicantly higher KCFT than KAVLT Z score for LHLD patients, as described above, but no signicant difference between KCFT versus KAVLT Z score for NLHLD patients, t(17) 1. For LTLEMTS patients, the Material Type of Language Dominance interaction failed to reach statistical signicance. However, separate analyses for each language dominance group revealed a signicantly higher KCFT than KAVLT Z score for LHLD patients, as described above, but no signicant difference between KCFT versus KAVLT Z score for NLHLD patients, t(4) 1 (see Figure 3). Thus, LTLEMTS and LTLEMTS patients were similar in the effects of language dominance on material-specic memory performance.

Figure 2. Mean Z scores as a function of material (KAVLT, KCFT) and seizure laterality (LTLE, RTLE). Left-hemisphere language dominant patients with evidence of mesial temporal sclerosis (LHLDMTS; n 39) are depicted in the left panel, and left-hemisphere language dominant patients without evidence of mesial temporal sclerosis (LHLDMTS; n 22) are depicted in the right panel. Error bars represent the standard error of the mean. KAVLT K-Auditory Verbal Learning Test; KCFT K-Complex Figure Test; LTLE left temporal lobe epilepsy; RTLE right temporal lobe epilepsy.

MATERIAL-SPECIFIC MEMORY IN TEMPORAL LOBE EPILEPSY

65

Figure 3. Mean Z scores as a function of material (KAVLT, KCFT) and language dominance (LHLD, NLHLD). Left temporal lobe epileptic patients with evidence of mesial temporal sclerosis (LTLEMTS; n 31) are depicted in the left panel, and left temporal lobe epileptic patients without evidence of mesial temporal sclerosis (LTLEMTS; n 16) are depicted in the right panel. Error bars represent the standard error of the mean. KAVLT K-Auditory Verbal Learning Test; KCFT K-Complex Figure Test; LHLD left-hemisphere language dominance; NLHLD non-left-hemisphere language dominance.

Discussion
The present study investigated material-specic memory in TLE patients, with reference to the effects of seizure laterality and language dominance. The examination of seizure laterality effects was restricted to the LHLDLTLE and LHLDRTLE groups. The Seizure Laterality Group Material interaction was highly signicant, indicating differential patterns of material-specic memory for the LHLDLTLE and LHLDRTLE groups. The LHLDLTLE group showed signicantly higher nonverbal than verbal memory capacity, whereas the LHLDRTLE group showed signicantly better ability to learn verbal than nonverbal material. The LHLDLTLE group had signicantly higher nonverbal memory capacity relative to the LHLDRTLE group. The LHLDRTLE group was superior to the LHLD LTLE group in verbal memory performance, but this difference was not statistically signicant. Thus, the LHLD LTLE group showed relative verbal memory decit, whereas the LHLDRTLE group showed relative nonverbal memory impairment. These results indicate that LTLE may interfere more with verbal than nonverbal memory capacity, whereas RTLE may interfere more with nonverbal than verbal memory capacity. The examination of language-dominance effects was restricted to the LHLDLTLE and NLHLDLTLE groups. The Language Dominance Group Material interaction was signicant, indicating differential patterns of materialspecic memory for the LHLDLTLE and NLHLDLTLE groups. The LHLDLTLE group showed higher capacity to learn nonverbal than verbal material, whereas the NLHLD LTLE group showed no signicant difference in the capacity to learn verbal versus nonverbal material. Moreover, the

NLHLDLTLE group was signicantly superior to the LHLDLTLE group in verbal memory performance, indicating relatively spared verbal memory capacity. Thus, the NLHLDLTLE group did not show relative verbal memory decit, which was evident in the LHLDLTLE group. Separate analyses for MTS and MTS patients were also performed to investigate possible confounding because of inclusion of patients with different pathologic substrates. The results indicated that effects of both seizure laterality and language dominance on material-specic memory performance are highly comparable for MTS versus MTS patients. Thus, the present results are not easily attributed to a heterogeneous patient sample. A plausible hypothesis for relative sparing of verbal memory function in the NLHLDLTLE group is interhemispheric reorganization of verbal memory function. According to this hypothesis, certain early left-hemisphere injury associated with LTLE may cause interhemispheric transfer of verbal memory as well as general language functions to the right hemisphere, which in turn causes relative sparing of the functions. An alternative hypothesis may be intrahemispheric as opposed to interhemispheric reorganization of verbal memory function. Although the intrahemispheric reorganization hypothesis is not entirely implausible (Seidenberg et al., 1997), available evidence from other studies is more consistent with the interhemispheric reorganization hypothesis. An IAP study reported by Helmstaedter et al. (1994) indicates that in LTLE patients with RHLD the right hemisphere is superior to the left hemisphere in verbal memory processing (see also Glosser et al., 1995; Kim & Yi, 1997). Rausch et al. (1991) reported that LTLE patients with RHLD do not show verbal memory decline following

66

KIM, YI, SON, AND KIM

left temporal lobectomy, suggesting that verbal memory in these patients is supported primarily by the right hemisphere. However, the present study involved LTLE patients with NLHLD, which is a distinctly different population from LTLE patients with RHLD. Thus, future research will need to address possible differences in interhemispheric reorganization of verbal memory function in LTLE patients with RHLD versus bilateral language dominance. One weak aspect of the present data is failure to nd a signicant difference in verbal memory capacity between the LHLDLTLE and LHLDRTLE groups. However, the group means were in the predicted direction and related within-groups material-specic effects reached statistical signicance. Thus, the lack of statistical signicance is best attributed to insufcient statistical power in combination with small magnitude of the effect. Lateralized materialspecic effects could be demonstrated in either between- or within-groups comparisons. The within-groups comparison (i.e., comparison of verbal versus nonverbal memory performance within a group) is conceptually at least as important as the between-groups comparison. Moreover, the within-groups comparison should reach statistical signicance more efciently than the between-groups comparison, reecting its higher statistical power. Nevertheless, prior studies of material-specic memory in TLE patients have relied more on the between- than the within-groups comparisons. The low use of the within-groups comparisons may reect the fact that they require equal units of measurement across memory tasks, which may be difcult to fulll unless standardized tasks are used. The present use of aggregate Z scores provides one means that can circumvent this difculty. The present study shows that the NLHLDLTLE group, despite left mesiotemporal dysfunction, does not show relative verbal memory decit. Thus, inclusion of NLHLD LTLE patients may have contributed to certain prior failures to nd a signicant difference in material-specic memory between LTLE and RTLE patients. Consistent with this view, a number of prior studies (Cohen, 1992; Delaney et al., 1980; Ellis et al., 1991; Giovagnoli & Avanzini, 1999; Giovagnoli et al., 1995; La ` davas et al., 1979; Lendt et al., 1999; Mungas, Ehlers, Walton, & McCutchen, 1985; Piguet et al., 1994) failed to control variations in type of language dominance. Even when a study controlled type of language dominance, the study sample may have included some patients with NLHLD. This is suggested by wide differences in the reported incidence of LHLD in the TLE population across epilepsy centers, ranging from 60% to 90% (Chelune et al., 1998; Dinner, 1991; Loring et al., 1990; Zatorre, 1989). These vastly different estimates suggest nontrivially differential criteria for dening LHLD (or NLHLD) among the centers, as noted by many authors (Benbadis et al., 1995; Kurthen et al., 1994; Snyder, Novelly, & Harris, 1990). The present study adopted a conservative criterion for LHLD, because it was thought to be critical for proper testing of the research hypotheses. The signicant differences between the LHLDLTLE and NLHLDLTLE groups in the present study suggest the utility of conservatively dening LHLD.

In conclusion, the present study shows that both seizure laterality and language dominance have signicant effects on material-specic memory in TLE patients. The effects of seizure laterality indicate that LTLE may interfere more with verbal than nonverbal memory capacity, whereas RTLE may interfere more with nonverbal than verbal memory capacity. The effects of language dominance indicate that selective verbal or nonverbal memory decits that are dependent on side of seizure onset are apparent in patients with LHLD but not in patients with NLHLD. However, the difference in verbal versus nonverbal memory capacity was of relatively small magnitude even in patients with LHLD. Thus, the lateralizing value of material-specic memory performance would be, at best, of modest degree. Limitations of the present study include the use of one particular type of verbal and nonverbal memory tasks. Moreover, the two memory tasks were not matched in procedures. Thus, the external validity of the present ndings needs to be investigated using other types of verbal and nonverbal memory tasks. In addition, there are several other possible ways of classifying hemispheric language dominance on the basis of the IAP than the one adopted in the present study. For example, the NLHLD group may be further divided into those with RHLD versus bilateral language dominance. Thus, the effects of language dominance on material-specic memory need to be further explored using other classication criteria.

References
Abrahams, S., Pickering, A., Polkey, C. E., & Morris, R. G. (1997). Spatial memory decits in patients with unilateral damage to the right hippocampal formation. Neuropsychologia, 35, 1124. Barr, W. B. (1997). Examining the right temporal lobes role in nonverbal memory. Brain and Cognition, 35, 26 41. Baxendale, S. A., Thompson, P. J., & van Paesschen, W. (1998). A test of spatial memory and its clinical utility in the presurgical investigation of temporal lobe epilepsy patients. Neuropsychologia, 36, 591 602. Baxendale, S. A., van Paesschen, W., Thompson, P. J., Connelly, A., Duncan, J. S., Harkness, W. F., & Shorvon, S. D. (1998). The relationship between quantitative MRI and neuropsychological functioning. Epilepsia, 39, 158 166. Benbadis, S. R., Dinner, D. S., Chelune, G. J., Piedmonte, M., & Lu ders, H. O. (1995). Objective criteria for reporting language dominance by intracarotid amobarbital procedure. Journal of Clinical and Experimental Neuropsychology, 17, 682 690. Bornstein, R. A., Pakalnis, A., & Drake, M. E. (1988). Verbal and nonverbal memory and learning in patients with complex partial seizures of temporal lobe origin. Journal of Epilepsy, 1, 203 208. Breier, J. I., Brookshire, B. L., Eletcher, J. M., Thomas, A. B., Plenger, P. M., Wheless, J. W., et al. (1997). Identication of side of seizure onset in temporal lobe epilepsy using memory tests in the context of reading decits. Journal of Clinical and Experimental Neuropsychology, 19, 161171. Breier, J. I., Plenger, P. M., Castillo, R., Fuchs, K., Wheless, J. W., Thomas, A. B., et al. (1996). Effects of temporal lobe on spatial and gural aspects of memory for a complex geometric gure. Journal of the International Neuropsychological Society, 2, 535540.

MATERIAL-SPECIFIC MEMORY IN TEMPORAL LOBE EPILEPSY Chapman, L. J., & Chapman, J. P. (1987). The measurement of handedness. Brain and Cognition, 6, 175183. Chelune, G. J., Naugle, R. I., Hermann, B. P., Barr, W. B., Trenerry, M. R., Loring, D. W., et al. (1998). Does presurgical IQ predict seizure outcome after temporal lobectomy? Evidence from the Bozeman Epilepsy Consortium. Epilepsia, 39, 314 318. Cohen, M. (1992). Auditory/verbal and visual/spatial memory in children with complex partial epilepsy of temporal lobe origin. Brain and Cognition, 20, 315326. Delaney, R. C., Rosen, A. J., Mattson, R. H., & Novelly, R. A. (1980). Memory function in focal epilepsy: A comparison of non-surgical, unilateral temporal lobe and frontal lobe samples. Cortex, 16, 103117. Dinner, D. S. (1991). Intracarotid amobarbital test to dene language lateralization. In H. Lu ders (Ed.), Epilepsy surgery (pp. 503506). New York: Raven Press. Ellis, A. W., Hillam, J. C., Cardno, A., & Kay, J. (1991). Processing of words and faces by patients with left and right temporal lobe epilepsy. Behavioral Neurology, 4, 121128. Engel, J. (1996). Introduction to temporal lobe epilepsy. Epilepsy Research, 26, 141150. Engel, J., van Ness, P. C., Rasmussen, T. B., & Ojemann, L. M. (1993). Outcome with respect to epileptic seizures. In J. Engel (Ed.), Surgical treatment of the epilepsies (pp. 609 621). New York: Raven Press. Fedio, P., & Mirsky, A. F. (1969). Selective intellectual decits in children with temporal lobe or centrencephalic epilepsy. Neuropsychologia, 7, 287300. Giovagnoli, A. R., & Avanzini, G. (1999). Learning and memory impairment in patients with temporal lobe epilepsy: Relation to the presence, type, and location of brain lesion. Epilepsia, 40, 904 911. Giovagnoli, A. R., Casazza, M., & Avanzini, G. (1995). Visual learning on a selective reminding procedure and delayed recall in patients with temporal lobe epilepsy. Epilepsia, 36, 704 711. Glosser, G., Saykin, A. J., Deutsch, G. K., OConnor, M. J., & Sperling, M. R. (1995). Neural organization of material-specic memory functions in temporal lobe epilepsy patients as assessed by the intracarotid amobarbital test. Neuropsychology, 9, 449 456. Glowinski, H. (1973). Cognitive decits in temporal lobe epilepsy. The Journal of Nervous and Mental Disease, 157, 129 137. Gold, J. M., Blaxton, T. A., Hermann, B. P., Randolph, C., Fedio, P., Goldberg, T. E., et al. (1995). Memory and intelligence in lateralized temporal lobe epilepsy and schizophrenia. Schizophrenia Research, 17, 59 65. Helmstaedter, C., Kurthen, M., Linke, D. B., & Elger, C. E. (1994). Right hemisphere restitution of language and memory functions in right hemisphere language-dominant patients with left temporal lobe epilepsy. Brain, 117, 729 737. Helmstaedter, C., Pohl, C., & Elger, C. E. (1995). Relations between verbal and nonverbal memory performance: Evidence of confounding effects particularly in patients with right temporal lobe epilepsy. Cortex, 31, 345355. Hermann, B. P., Connell, B., Barr, W. B., & Wyler, A. R. (1995). The utility of the Warrington Recognition Memory Test for temporal lobe epilepsy: Pre- and postoperative results. Journal of Epilepsy, 8, 139 145. Hermann, B. P., Seidenberg, M., Schoenfeld, J., & Davies, K. (1997). Neuropsychological characteristics of the syndrome of mesial temporal lobe epilepsy. Archives of Neurology, 54, 369 376.

67

Hermann, B. P., Wyler, A. R., Richey, E. T., & Rea, J. M. (1987). Memory function and verbal learning ability in patients with complex partial seizures of temporal lobe origin. Epilepsia, 28, 547554. Kim, H. (1999). ReyKim Memory Test. Daegu, Korea: Neuropsychology Press. Kim, H., & Yi, S. (1997). The effect of early versus late onset of temporal lobe epilepsy on hemispheric memory laterality: An intracarotid amobarbital procedure study. Journal of Korean Medical Science, 12, 559 563. Kurthen, M., Helmstaedter, C., Linke, D. B., Hufnagel, L. A., Elger, C. E., & Schramm, J. (1994). Quantitative and qualitative evaluation of patterns of cerebral language dominance: An amobarbital study. Brain and Language, 46, 536 564. La ` davas, E., Umilta ` , C., & Provinciali, L. (1979). Hemispheredependent cognitive performances in epileptic patients. Epilepsia, 20, 493502. Lendt, M., Helmstaedter, C., & Elger, C. E. (1999). Pre- and postoperative neuropsychological proles in children and adolescents with temporal lobe epilepsy. Epilepsia, 40, 15431550. Lezak, M. D. (1983). Neuropsychological assessment. New York: Oxford University Press. Loring, D. W., Hermann, B. P., Lee, G. P., Drane, D. L., & Meador, K. J. (2000). The Memory Assessment Scales and lateralized temporal lobe epilepsy. Journal of Clinical Psychology, 56, 563570. Loring, D. W., Lee, G. P., Martin, R. C., & Meador, K. J. (1988). Material-specic learning in patients with partial complex seizures of temporal lobe origin: Convergent validation of memory constructs. Journal of Epilepsy, 1, 5359. Loring, D. W., Lee, G. P., & Meador, K. J. (1988). Revising the ReyOsterrieth: Rating right hemisphere recall. Archives of Clinical Neuropsychology, 3, 239 247. Loring, D. W., Meador, K. J., Lee, G. P., Murro, A. M., Smith, J. R., Flanigin, H. F., et al. (1990). Cerebral language lateralization: Evidence from intracarotid amobarbital testing. Neuropsychologia, 28, 831 838. Mayeux, R. M., Brandt, J., Rosen, J., & Benson, D. F. (1980). Interictal memory and language impairment in temporal lobe epilepsy. Neurology, 30, 120 125. McMillan, T. M., Powell, G. E., Janota, I., & Polkey, C. E. (1987). Relationships between neuropathology and cognitive functioning in temporal lobectomy patients. Journal of Neurology, Neurosurgery, and Psychiatry, 50, 167176. Meencke, H. J., & Veith, G. (1991). Hippocampal sclerosis in epilepsy. In H. Lu ders (Ed.), Epilepsy surgery (pp. 705715). New York: Raven Press. Milner, B. (1975). Psychological aspects of focal epilepsy and its neurosurgical management. In D. P. Purpura, J. K. Penry, & R. D. Walter (Eds.), Advances in neurology (Vol. 8, pp. 299 321). New York: Raven Press. Moore, P. M., & Baker, G. A. (1996). Validation of the Wechsler Memory ScaleRevised in a sample of people with intractable temporal lobe epilepsy. Epilepsia, 37, 12151220. Mungas, D., Ehlers, C., Walton, N., & McCutchen, C. B. (1985). Verbal learning differences in epileptic patients with left and right temporal lobe foci. Epilepsia, 26, 340 345. Naugle, R. I., Chelune, G. J., Schuster, J., Lu ders, H. O., & Comair, Y. (1994). Recognition memory for words and faces before and after temporal lobectomy. Assessment, 1, 373381. Piguet, O., Saling, M. M., OShea, M. F., Berkovic, S. F., & Bladin, P. F. (1994). Rey gure distortions reect nonverbal

68

KIM, YI, SON, AND KIM Selwa, L. M., Berent, S., Giordani, B., Henry, T. R., Buchtel, H. A., & Ross, D. A. (1994). Serial cognitive testing in temporal lobe epilepsy: Longitudinal changes with medical and surgical therapies. Epilepsia, 35, 743749. Snyder, P. J., Novelly, R. A., & Harris, L. J. (1990). Mixed speech dominance in the intracarotid sodium amytal procedure: Validity and criteria issues. Journal of Clinical and Experimental Neuropsychology, 12, 629 643. Son, E. I., Howard, M. A., Ojemann, G. A., & Lettich, E. (1994). Comparing the extent of hippocampal removal to the outcome in terms of seizure control. Stereotaxic and Functional Neurosurgery, 62, 232237. Wolf, H. K., Campos, M. G., Zentner, J., Hufnagel, A., Schramm, J., Elger, C. E., & Wiestler, O. D. (1993). Surgical pathology of temporal lobe epilepsy: Experience with 216 cases. Journal of Neuropathology and Experimental Neurology, 52, 499 506. Yeom, Y., Park, Y., Oh, K., Kim, J., & Lee, Y. (1992). Korea Wechsler Adult Intelligence Scale. Seoul, Korea: Korea Guidance. Zatorre, R. J. (1989). Perceptual asymmetry on the dichotic fused words test and cerebral speech lateralization determined by the Carotid Sodium Amytal Test. Neuropsychologia, 27, 1207 1219.

recall differences between right and left foci in unilateral temporal lobe epilepsy. Archives of Clinical Neuropsychology, 9, 451 460. Rasmussen, T., & Milner, B. (1977). The role of early left-brain injury in determining lateralization of cerebral speech functions. Annals of the New York Academy of Sciences, 299, 355369. Rausch, R., Boone, K., & Ary, C. M. (1991). Right-hemisphere language dominance in temporal lobe epilepsy: Clinical and neuropsychological correlates. Journal of Clinical and Experimental Neuropsychology, 13, 217231. Rey, A. (1941). Lexamen psychologique dans les cas dencephalopathie traumatique [Psychological examination for traumatic encephalopathy]. Archives de Psychologie, 28, 286 340. Rey, A. (1964). Lexamen clinique en psychologie [The clinical examination in psychology]. Paris: Press Universitaire de France. Satz, P., Strauss, E., Wada, J., & Orsini, D. L. (1988). Some correlates of intra- and interhemispheric speech organization after left focal brain injury. Neuropsychologia, 26, 345350. Sawrie, S. M., Martin, R. C., Gilliam, F., Knowlton, R., Faught, E., & Kuzniecky, R. (2001). Verbal retention lateralizes patients with temporal lobe epilepsy and bilateral hippocampal atrophy. Epilepsia, 42, 651 659. Seidenberg, M., Hermann, B. P., Schoenfeld, J., Davies, K., Wyler, A., & Dohan, F. C. (1997). Reorganization of verbal memory function in early onset left temporal lobe epilepsy. Brain and Cognition, 35, 132148.

Received October 23, 2001 Revision received June 27, 2002 Accepted June 27, 2002