Differential Effects of Left Versus Right Mesial Temporal Lobe Epilepsy on Wechsler Intelligence Factors

Differential Effects of Left Versus Right Mesial Temporal LobeEpilepsy on Wechsler Intelligence Factors

Hongkeun KimDaegu University

Sangdoe Yi and Eun Ik SonKeimyung University

Jieun KimCatholic University of Daegu

This study investigates the effects of left versus right mesial temporal lobe epilepsy (MTLE)on Wechsler intelligence factors. In the left MTLE group, the Verbal Comprehension (VC)factor score was significantly lower than the Perceptual Organization (PO) factor score,whereas in the right MTLE group, the PO factor score was significantly lower than the VCfactor score. The VC factor score was significantly lower for the left than the right MTLEgroup, whereas the PO factor score was significantly lower for the right than the left MTLEgroup. Thus, left versus right MTLE was associated with relative deficits in verbal versusnonverbal intelligence, respectively. These findings indicate that lateralized cognitive deficitsin unilateral MTLE patients are not limited to the learning–memory domain but include moreglobal intelligence functions.

Mesial temporal lobe epilepsy (MTLE) may be the mostcommon form of human epileptogenic disorders (Engel,1996). Pre- and postsurgical neuropsychological studies ofMTLE patients have greatly contributed to the current un-derstanding of human brain–behavior relationship. MTLEis frequently associated with memory impairments reflect-ing epileptogenic lesions in the mesial temporal lobe, aregion known to be critical for consolidating new memory.A notable characteristic of the memory deficits in MTLE isthe significant relationship between side of seizure onsetand material-specific memory capacity: Patients with leftMTLE tend to show more severe verbal than nonverbalmemory deficits, whereas patients with right MTLE tend toshow more severe nonverbal than verbal memory impair-ments (e.g., Delaney, Rosen, Mattson, & Novelly, 1980;Fedio & Mirsky, 1969; Helmstaedter, Pohl, & Elger, 1995;Hermann, Wyler, Richey, & Rea, 1987; Kim, Yi, Son, &Kim, 2003; Ladavas, Umilta, & Provinciali, 1979). Al-though the link between right MTLE and nonverbal mem-ory deficits has not been always supported, some well-controlled prior studies (e.g., Abrahams, Pickering, Polkey,& Morris, 1997; Baxendale, Thompson, & van Paesschen,

1998) have provided convincing evidence for the link.These selective verbal or nonverbal memory deficits that aredependent on side of seizure onset provide evidence that theleft and the right mesial temporal regions are specialized forverbal versus nonverbal memory processing, respectively(Milner, 1975).

The aim of the present study was to investigate possibledifferential effects of left versus right MTLE on material-specific intelligence. Although neuropsychological studiesof MTLE have emphasized memory deficits, neuropsycho-logical deficits in MTLE are not limited to the learning–memory domain but extend into more global intelligencefunctions (Hermann, Seidenberg, Schoenfeld, & Davies,1997). For example, the IQ level of MTLE patients is as lowas their memory quotient level (Chelune, Naugle, Luders, &Awad, 1991; Gold et al., 1995; Ivnik, Sharbrough, & Laws,1987; Mayeux, Brandt, Rosen, & Benson, 1980; Selwa etal., 1994). Consistent with global intelligence deficits, thereis a growing body of evidence that brain abnormalities inMTLE, even in well-defined cases of unilateral MTLE, arenot limited to the epileptogenic region but extend intowidespread areas of extrahippocampal temporal and extra-temporal regions. This evidence is known from positionemission tomography (PET; e.g., Arnold et al., 1996;Henry, Mazziotta, & Engel, 1993; Henry et al., 1990; Jokeitet al., 1997; Van Bogaert et al., 2000), single photon emis-sion computed tomography (SPECT; e.g., Avery et al.,2001; Duncan, Patterson, Hadley, Roberts, & Bone, 1996;Yune et al., 1998), quantitative magnetic resonance imaging(MRI; e.g., Briellmann, Jackson, Kalnins, & Berkovic,1998; DeCarli et al., 1998; Jutila et al., 2001; Kuzniecky etal., 1999; Lee et al., 1998; Marsh et al., 1997; Moran,Lemieux, Kitchen, Fish, & Shorvon, 2001; Sisodiya et al.,1997; Specht et al., 1997), and postmortem pathologicalstudies of MTLE patients (Margerison & Corsellis, 1966).The abnormal extratemporal regions, although variable

Hongkeun Kim, Department of Rehabilitation Psychology,Daegu University, Daegu, South Korea; Sangdoe Yi, Departmentof Neurology, Keimyung University, Daegu, South Korea; Eun IkSon, Department of Neurosurgery, Keimyung University, Daegu,South Korea; Jieun Kim, Department of Neurology, Catholic Uni-versity of Daegu, Daegu, South Korea.

This research was supported as a Brain Neuroinformatics Re-search Program sponsored by the Korean Ministry of Science andTechnology.

Correspondence concerning this article should be addressed toHongkeun Kim, Department of Rehabilitation Psychology, DaeguUniversity, 2288 Daemyung-dong, Nam-gu, Daegu 705-714,South Korea. E-mail: [email protected]

Neuropsychology Copyright 2003 by the American Psychological Association, Inc.2003, Vol. 17, No. 4, 556–565 0894-4105/03/$12.00 DOI: 10.1037/0894-4105.17.4.556

556

across MTLE patients, can be extremely widespread, in-cluding the frontal and parietal lobes, the temporal neocor-tex, the basal ganglia, the thalamus, the limbic system, andthe cerebellum. Important for the present purpose, the ex-tratemporal abnormalities are usually more severe on theipsilateral than the contralateral side of the epileptogenicregion. For example, using high-resolution MRI, Lee et al.(1998) found that the mean relative reduction in the tempo-ral lobe volume (excluding the hippocampal formation andthe parahippocampal gyrus) was about 15% in the ipsilateraltemporal lobe and about 7% in the contralateral temporallobe. The more severe ipsilateral abnormalities presumablyreflect an asymmetric disease process associated with uni-lateral MTLE.

Thus, the working hypothesis for the present study wasthat left MTLE is associated with relative verbal intelli-gence deficits, whereas right MTLE is associated with rel-ative nonverbal intelligence impairments. A review of theliterature concerning this lateralized intelligence deficitshypothesis reveals that most of the research has used theWechsler intelligence scales; research using other intelli-gence instruments is virtually nonexistent. Moreover, withfew exceptions, these studies used verbal IQ (VIQ) andperformance IQ (PIQ) as dependent measures.1 Some ofthese studies reported significantly lower VIQ for left thanfor right MTLE patients but no significant difference in PIQbetween the two groups (Blakemore, Ettlinger, & Falconer,1966; Kneebone, Chelune, & Luders, 1997; Loring, Lee, &Meador, 1988). Others found significantly lower PIQ aswell as VIQ for left than for right MTLE patients (Moore &Baker, 1996; Selwa et al., 1994). However, in the greatmajority of these studies, neither VIQ nor PIQ was signif-icantly different between left versus right MTLE patients(Abrahams et al., 1997; Bornstein, Pakalnis, & Drake, 1988;Breier et al., 1997; Fedio & Mirsky, 1969; Hermann et al.,1995; Hermann et al., 1987; Loring, Lee, Martin, & Mea-dor, 1988; Mayeux et al., 1980). Thus, extant literatureprovides poor support for the lateralized intelligence deficitshypothesis.

However, the poor support is rather puzzling in view ofthe fact that MTLE is associated with widespread extratem-poral (and extrahippocampal temporal) lesions that are moresevere on the ipsilateral than the contralateral side of theepileptogenic region. Thus, the prior failures to support thelateralized intelligence deficits hypothesis may reflect, atleast in part, methodological inadequacy rather than trueabsence of lateralized intelligence deficits in MTLE. Forexample, in studies that have used a relatively small samplesize (Abrahams et al., 1997; Blakemore et al., 1966; Born-stein, Pakalnis, & Drake, 1988; Fedio & Mirsky, 1969;Hermann et al., 1987; Loring, Lee, Martin, & Meader, 1988;Loring, Lee, & Meador, 1988; Mayeux et al., 1980), a weakstatistical power may have contributed to the failures. Morerelevant for the present purpose, VIQ and PIQ, which wereused as dependent measures in most prior studies, may notbe ideal measures of verbal and nonverbal intelligence.Numerous factor-analytic studies of the Wechsler intelli-gence scales indicate that a three-factor structure, consistingof a Verbal Comprehension (VC) factor, a Perceptual Or-

ganization (PO) factor, and a Freedom From Distractibility(FD)/Working Memory (WM) factor, best fits the data (e.g.,Dai, Gong, & Zhong, 1990; Waller & Waldman, 1990;Ward, Ryan, & Axelrod, 2000). This is found not only forWechsler intelligence data obtained from nonclinical con-trol subjects but also from various clinical groups (e.g.,Bornstein, Drake, & Pakalnis, 1988; Burgess, Flint, & Ad-shead, 1992; Choi & Kim, 1990; Fowler, Richards, & Boll,1980; Hermann et al., 1995; Lansdell, 1968; Ryan &Schneider, 1986). It is generally recognized that VC and POfactor scores (or index scores) are better measures of verbaland nonverbal intelligence than VIQ and PIQ, respectively(for a review, see Kaufman, 1990). Nevertheless, few priorstudies (Hermann et al., 1995; Ivnik et al., 1987) haveaddressed the lateralized intelligence deficits hypothesis us-ing VC and PO factor scores as dependent measures. Thus,in the present study, VC and PO factor scores were used asdependent measures, with the expectation that these scoresmight reveal certain lateralized material-specific intelli-gence deficits that are not apparent with VIQ and PIQ asdependent measures.

Method

Subjects

The subjects were 71 patients who had ultimately undergonesurgery of the left or right mesial temporal lobe for treatment ofmedically intractable epilepsy of unilateral mesial temporal origin.They were selected from a consecutive series of patients undergo-ing mesial temporal lobe surgery in an epilepsy surgery center inKorea who met the following criteria: (a) completion of the K-WAIS (Yeom, Park, Oh, Kim, & Lee, 1992), the Korean versionof the Wechsler Adult Intelligence Scale—Revised (WAIS–R;Wechsler, 1981) as part of presurgical workup, (b) age greaterthan 15 years, (c) left hemisphere language dominance as deter-mined by intracarotid amobarbital procedure (IAP), (d) full scaleIQ greater than 64, (e) no evidence of space-occupying structurallesions on MRI scanning other than hippocampal atrophy, and (f)seizure-free (Class I) or near seizure-free (Class II) status (Engel,Van Ness, Rasmussen, & Ojemann, 1993) at 1-year postsurgeryfollow-up. The presurgical evaluation included seizure semiology,prolonged interictal and ictal video electroencephalograph (EEG)from scalp/sphenoidal electrodes, MRI scanning, interictalSPECT, IAP, neuropsychological testing, and if necessary, EEGfrom chronically implanted bilateral subdural strip electrodes.Thirty-four subjects had epileptogenic discharges localized in theleft mesial temporal lobe, and 37 in the right mesial temporal lobe.Forty-seven were male, and 24 were female. The mean age (�SD)at the time of study was 29.3 � 6.8 years. The mean age atrecurrent seizure onset was 15.0 � 6.6 years, and the mean seizureduration was 14.8 � 7.4 years. Sixty-nine were right-handed and 2were left-handed, as determined by a 13-item questionnaire (Chap-man & Chapman, 1987). The average full scale IQ on the K-WAISwas 86.9 � 11.0. Five subjects had a history of pure complexpartial seizures (CPS), and 66 subjects had a history of CPSfrequently or occasionally followed by generalized tonic-clonicseizures.

1 In some of these studies, the VIQ and PIQ scores were re-ported as part of clinical characteristics of the sample and were notrelated to the main hypotheses of the studies.

557LEFT VERSUS RIGHT MESIAL TEMPORAL LOBE EPILEPSY

Procedures

Intelligence assessment. The K-WAIS (Yeom et al., 1992)was administered to each patient in the context of a comprehensivepresurgical neuropsychological evaluation. The structure and pat-tern of the K-WAIS are the same as those of the WAIS–R (Wechs-ler, 1981), although some items (e.g., What are the colors in theAmerican flag?) were changed to make them suitable for Koreansubjects. The K-WAIS standardization sample included a totalof 1,396 healthy adult men and women, ranging in age from 16to 64 years. Only 9 of 11 subscales of the K-WAIS, excluding theVocabulary and Picture Arrangement subscales, were administeredto shorten the testing time. The decision to exclude the Vocabularysubscale was based on the fact that its administration time is thelongest among the Wechsler subscales and that it is highly corre-lated with other VC subscales (Kaufman, 1990). The decision toexclude the Picture Arrangement subscale reflected the fact that itsadministration time is relatively long and that the strength of itsloading on the PO factor is comparatively modest (Kaufman,1990).

IAP. All patients underwent IAP as part of presurgical evalua-tion. The IAP was conducted, with the patient supine, immediatelyfollowing angiography. Amobarbital 125 mg in a 10% solution wasinjected into the internal carotid artery using a transfemoral catheterover a 4–5-s interval. In most patients, the side considered for resec-tion was injected first. Left and right hemisphere injections were doneon the same day, with a minimum of 40 min between the twoinjections. Following demonstration of hemiplegia, the patient waspresented with a series of language tasks that lasted approximately60–90 s. Assessment of language functions during the course of theIAP have varied slightly over the years but always included bothreceptive and expressive language tasks. The core tasks includedfollowing a simple command (e.g., “close your eyes”), reading a shortsentence, naming a picture, and repeating a phrase spoken by theexaminer. Only patients with left hemisphere language dominance, asdetermined by the IAP, were included to avoid possible confoundingeffects associated with a shift of language dominance.

Analyses

To test the hypothesis that left and right MTLE have differentialeffects on material-specific intelligence, we conducted three separateanalyses, one involving the VIQ and PIQ, one involving the VC andPO index scores, and one involving the VC and PO factor scores. Inthe analyses involving the VIQ and PIQ, the VIQ was computed onthe basis of the protated sum of scaled scores of the five verbalsubscales, and the PIQ was computed on the basis of the protated sumof scaled scores of the four performance subscales. In the analysesinvolving the VC and PO index scores, the VC index score wasderived from Information, Comprehension, and Similarities subscalescores, and the PO index score was derived from Picture Completion,Block Design, and Object Assembly subscale scores. Each derivationinvolved converting the sum of scaled scores of the three subscalesinto its index score equivalent on the basis of the conversion tablesprovided by Kim (2002). Each index score had a normative mean of100 and a standard deviation of 15. In the analyses involving the VCand PO factor scores, the correlation matrix for the nine subscale rawscores were first subject to a factor analysis, and resulting factorscores were then subject to analyses. The factor-analytic procedureinvolved a principal component analysis, followed by extraction ofthree factors with varimax rotation. The factor scores were z scoreswith a mean of 0 and a standard deviation of 1, and they weremutually uncorrelated, reflecting use of an orthogonal rotation

method. Finally, we also performed analyses of individual subscalescores to explore possible source differences in the results obtainedwith various summary measures.

Results

Demographic and Clinical Variables

The demographic and clinical characteristics of the leftand right MTLE groups are presented in Table 1. Groupdifferences were examined by F tests for continuous vari-ables and by 2 � 2 Fisher exact tests for categorical vari-ables. Statistical analyses revealed no significant differencesbetween the left and right MTLE groups in terms of age,education, gender ratio, handedness, onset age of recurrentseizures, duration of seizures, incidence of febrile convul-sions, 1-year postsurgery outcome (% Class I), or full scaleIQ ( ps � .10). Thus, these demographic and clinical vari-ables may not account for any differences in material-specific intelligence between the left and right MTLEgroups.

VIQ and PIQ

The VIQ and PIQ scores on the K-WAIS were analyzedusing a 2 � 2 (Intelligence Type � Seizure Laterality)analysis of variance (ANOVA). The mean VIQ and PIQscores of the left and right MTLE groups are depicted inFigure 1. In this ANOVA, no main or interaction effectsreached statistical significance ( ps � .10). Within the leftMTLE group, the VIQ and PIQ scores were not significantlydifferent, F(1, 33) � 1. However, within the right MTLEgroup, the PIQ score was significantly lower than the VIQscore, F(1, 36) � 4.24, p � .05. The left versus right MTLEgroups were not significantly different in the VIQ score,F(1, 69) � 1, or the PIQ score, F(1, 69) � 2.22, p � .10.Thus, with the VIQ and PIQ scores as dependent variables,there is little support for the hypothesis that left and rightMTLE are associated with relative verbal and nonverbalintelligence impairments, respectively.

Table 1Demographic and Clinical Characteristics of the Sample

Variable

LeftMTLE

(n � 34)

RightMTLE

(n � 37)

M SD M SD

Age (years) 29.9 6.1 28.7 7.4Education (years) 11.8 2.9 11.5 3.2Gender (% male) 74.0 60.0Handedness (% right-handed) 97.0 97.0Onset age of recurrent seizures (years) 15.2 6.2 14.9 7.1Duration of seizures (years) 15.4 7.5 14.2 7.4Febrile convulsions (%) 41.0 41.01-year postsurgery outcome (% Class I) 68.0 84.0Full scale IQ (K-WAIS) 87.8 12.2 86.0 10.0

Note. MTLE � mesial temporal lobe epilepsy; K-WAIS � K-Wechsler Adult Intelligence Scale (Yeom, Park, Oh, Kim, & Lee,1992).

558 KIM, YI, SON, AND KIM

VC and PO Indices

We analyzed the VC and PO index scores using a 2 � 2(Intelligence Type � Seizure Laterality) ANOVA. The der-ivation of the VC and PO index scores is described above.The mean VC and PO index scores of the left and rightMTLE groups are depicted in Figure 2. The only significanteffect in this ANOVA was an Intelligence Type � SeizureLaterality interaction, F(1, 69) � 10.98, p � .005. Withinthe left MTLE group, the VC index score was significantlylower relative to the PO index score, F(1, 33) � 4.63, p �.05, whereas within the right MTLE group, the PO indexscore was significantly lower compared with the VC indexscore, F(1, 36) � 6.72, p � .05. Between-groups compar-isons showed that the right MTLE group had a significantlylower PO index score than the left MTLE group, F(1,69) � 5.81, p � .05. The left MTLE group had a lower VCindex score than the right MTLE group, but this differencewas not statistically significant, F(1, 69) � 1.07, p � .30.Thus, with the VC and PO index scores as dependentvariables, the results indicate that left and right MTLE areassociated with relative verbal and nonverbal intelligenceimpairments, respectively.

VC, PO, and WM Factors

The correlation matrix for the nine subscales were firstsubject to a factor analysis. Details of this factor-analyticprocedure are presented in the Method section. The vari-max-rotated factor loadings of the three-factor solution areshown in Table 2. Factor loadings greater than .50 wereconsidered significant. With this criterion, the first factorwas a familiar VC factor (i.e., Information, Comprehension,Similarities), and the second factor was a PO factor (i.e.,

Picture Completion, Block Design, Object Assembly). Onthe third factor, the Digit Span and Arithmetic subscalesloaded strongly, at .86 and .66, respectively, whereas theDigit Symbol subscale loaded comparatively weakly, at .32.Thus, the third factor was considered as a WM factor. Thepercent of the total variance attributed to each of the factorswas 29.6 for VC, 22.8 for PO, and 18.5 for WM.

The VC and PO factor scores derived from the factoranalysis were subject to a 2 � 2 (Intelligence Type �

Table 2Factor Loadings, Eigenvalues, and Percentages ofVariance for K-Wechsler Adult Intelligence ScaleSubscales Based on Varimax-Rotated Three-FactorSolution

Subscale

Factor loading

Communality1 2 3

Information .85 .14 .15 .77Comprehension .80 .26 .21 .76Similarities .80 .02 .33 .75Picture Completion .43 .68 �.17 .68Block Design .01 .80 .41 .81Object Assembly .16 .85 .14 .77Digit Span .26 .04 .86 .81Arithmetic .38 .28 .66 .65Digit Symbol .48 .23 .32 .39

Eigenvalue 2.67 2.05 1.66% of variance 29.64 22.81 18.49

Note. Boldface indicates factor loadings greater than .50.

Figure 2. Mean Verbal Comprehension (VC) and PerceptualOrganization (PO) index scores for the left and the right mesialtemporal lobe epilepsy (MTLE) groups. Error bars represent �1standard error of the mean.

Figure 1. Mean verbal IQ and performance IQ for the left and theright mesial temporal lobe epilepsy (MTLE) groups. Error barsrepresent �1 standard error of the mean.

559LEFT VERSUS RIGHT MESIAL TEMPORAL LOBE EPILEPSY

Seizure Laterality) ANOVA. The mean VC, PO, and WMfactor scores of the left and right MTLE groups are depictedin Figure 3. In this ANOVA, the Intelligence Type �Seizure Laterality interaction was significant, F(1,69) � 11.07, p � .05, and no other effects reached statisticalsignificance ( ps � .80). Within the left MTLE group, theVC factor score was significantly lower than the PO factorscore, F(1, 33) � 4.63, p � .05, whereas within the rightMTLE group, the PO factor score was significantly lowerthan the VC factor score, F(1, 36) � 6.84, p � .05. Be-tween-groups comparisons showed that the left MTLEgroup had a significantly lower VC factor score than theright MTLE group, F(1, 69) � 4.48, p � .05, whereas theright MTLE group had a significantly lower PO factor scorethan the left MTLE group, F(1, 69) � 5.82, p � .05. Thus,with the VC and PO factor scores as dependent variables,there is clear evidence that left and right MTLE are asso-ciated with relative verbal and nonverbal intelligence im-pairments, respectively.

With respect to the WM factor, the right MTLE grouphad a significantly lower WM factor score than the leftMTLE group, F(1, 69) � 5.19, p � .05 (see Figure 3).Within the left MTLE group, the WM factor score wassignificantly higher than the VC factor score, F(1,33) � 4.77, p � .05, whereas within the right MTLE group,the WM factor score was significantly lower than the VCfactor score, F(1, 36) � 5.67, p � .05. The WM factor scorewas not significantly different from the PO factor scorewithin the left MTLE group, F(1, 33) � 1, or within theright MTLE group, F(1, 36) � 1. These results indicate thatWM factor performance is more adversely affected by rightthan left MTLE. Thus, when VC, PO, and WM factors wereconsidered simultaneously, left MTLE was associated with

lowered VC factor but relatively preserved PO and WMfactors, whereas right MTLE was associated with depressedPO and WM factors but relatively preserved VC factor.

Subscale Scores

We performed analyses of subscale scores twice, first withthe raw scores and second with the age-corrected scaled scores.General trends and significant differences were similar for thetwo sets of analyses. Therefore, only results obtained with theage-corrected scaled scores are presented in detail. The meansubscale scores for the left and right MTLE groups are shownin Table 3. The left and right MTLE groups were significantlydifferent in only one subscale, Block Design, F(1, 69) � 6.57,p � .05, reflecting lower performance by the right MTLEgroup. Statistical significance aside, the group means of allnine subscales were in the direction consistent with the groupdifferences found with summary measures. Thus, all three VCsubscales were performed less well by the left MTLE group,whereas all three PO subscales were performed less well by theright MTLE group. In addition, both WM subscales (i.e., DigitSpan and Arithmetic) were performed less well by the rightMTLE group. These results suggest that the group differencesfound with various summary measures reflect the differencesof most subscales rather than a small number of particularsubscales.

Individual Patient Prediction

The group results described above indicate a significantassociation between seizure laterality and patterns of VCversus PO discrepancy. However, the group data do notspeak directly to the issue of the clinical utility of VC versusPO discrepancy as a predictor of seizure laterality. Toaddress this issue, we performed three separate classifica-tion studies, using 15-, 23-, and 28-point discrepanciesbetween VC index and PO index as the cut-off criteria. The15-, 23-, and 28-point discrepancies represent the 25th,10th, and 5th percentiles in Kim’s (2002) normative study.

Table 3Group Differences for K-Wechsler Adult IntelligenceScale Subscale Age-Corrected Scaled Scores Between Leftand Right MTLE Patients

Subscale

LeftMTLE

(n � 34)

RightMTLE

(n � 37)

F(1, 69)M SD M SD

Information 7.09 2.61 7.92 2.34 2.00Comprehension 8.59 3.34 8.68 3.13 0.01Similarities 7.47 2.54 8.35 2.29 2.37Picture Completion 8.03 2.56 7.57 2.51 0.59Block Design 8.79 3.05 7.08 2.58 6.57*Object Assembly 8.65 2.14 7.92 2.19 2.00Digit Span 8.21 2.40 7.38 2.53 1.99Arithmetic 8.21 2.73 7.19 2.22 2.99Digit Symbol 7.85 2.39 7.81 2.11 0.01

Note. MTLE � mesial temporal lobe epilepsy.* p � .05.

Figure 3. Mean Verbal Comprehension (VC) and PerceptualOrganization (PO) factor scores for the left and the right mesialtemporal lobe epilepsy (MTLE) groups. Error bars represent �1standard error of the mean. WM � Working Memory.

560 KIM, YI, SON, AND KIM

In each classification, those patients who had a magnitude ofdiscrepancies less than the cut-off criterion were left un-classified (i.e., indeterminate). The classification results areshown in Table 4. In each classification, a large number ofpatients failed to meet the criterion and were left unclassi-fied. Nonetheless, each classification produced an assign-ment of patients that significantly differed from chance (allps � .05). Using the criterion representing the 25th percen-tile, a lateralizing or determinate result was obtained in 32%of patients. In 78% of these, the prediction was correct, andin 22% it was false. Using the criterion representing the 10thpercentile, a lateralizing result was obtained in 13% ofpatients. In 89% of these, the prediction was correct, and in11% it was false. Using the criterion representing the 5thpercentile, a lateralizing result was obtained in 7% of pa-tients, and in 100% of these, the prediction was correct.

Effects of Surgical Outcome

One of the subject selection criteria required that patientshave a favorable surgical outcome (i.e., Class I or II). Thisselection criterion may have inadvertently limited our sam-ple to patients with more focal and perhaps ipsilateral neu-ropathology, reflected in the VC versus PO findings. Thus,there is the possibility that our lateralized VC versus POfindings may reflect the surgical outcome selection crite-rion. With this possibility in mind, we examined VC and POfactor scores for all MTLE patients who met the sameselection criteria outlined in the Method section, with theexception of the surgical outcome criterion. This analysisinvolved 80 patients: 71 with a favorable surgical out-come, 8 with a nonfavorable surgical outcome (i.e., Class IIIor IV), and 1 with no follow-up data because of loss ofcontact. The VC and PO factor scores derived from 80patients were analyzed using a 2 � 2 (Intelligence Type �

Seizure Laterality) ANOVA. The Intelligence Type � Sei-zure Laterality interaction was again significant, F(1, 78)� 9.87, p � .05. Within the left MTLE group, the VC factorscore was significantly lower than the PO factor score, F(1,41) � 4.23, p � .05, whereas within the right MTLE group,the PO factor score was significantly lower than the VCfactor score, F(1, 37) � 8.07, p � .01. Between-groupscomparisons showed that the left MTLE group had a sig-nificantly lower VC factor score than the right MTLE group,F(1, 78) � 6.63, p � .05, whereas the right MTLE grouphad a significantly lower PO factor score than the leftMTLE group, F(1, 78) � 3.96, p � .05. Thus, the lateral-ized VC versus PO findings were apparent even when thesurgical outcome selection criterion was dropped. Thesefindings indicate that the selection criterion is not critical forthe present lateralized VC versus PO findings.2

2 An anonymous reviewer raised the possibility that a congruentVC versus PO discrepancy might be a favorable prognostic signfor surgical outcome. We have addressed this possibility by ex-amining the relationship between the lateralization of VC versusPO difference (correctly lateralized, falsely lateralized, nonlater-alized) and the surgical outcome (favorable, nonfavorable). Thelateralization of VC versus PO difference was derived from pre-diction of seizure laterality using the clinical criterion of a 15-pointor greater difference between the VC index and PO index. Thisanalysis involved 71 patients with a favorable outcome and 8 witha nonfavorable outcome. The proportion of patients with a favor-able outcome was .95, .83, and .89 for the correctly lateralized,falsely lateralized, and nonlateralized groups, respectively. Thesedifferences were not statistically significant, �2(2, N � 79) � 1,p � .50, indicating no association between lateralization of VCversus PO difference and surgical outcome.

Table 4Classification Rates on the Basis of VC Index Versus PO Index Discrepancies

ClassificationActual left temporal

(n � 34)Actual right temporal

(n � 37) �2(2, N � 71)

15-point discrepancy

Predicted left temporal 9 2Predicted right temporal 3 9Indeterminate 22 26 7.68*

23-point discrepancy

Predicted left temporal 6 0Predicted right temporal 1 2Indeterminate 27 35 7.25*

28-point discrepancy

Predicted left temporal 4 0Predicted right temporal 0 1Indeterminate 30 36 5.43*

Note. The 15-, 23-, and 28-point discrepancies between the Verbal Comprehension (VC) indexand the Perceptual Organization (PO) index represent the 25th, 10th, and 5th percentiles, respec-tively, in Kim’s (2002) normative study.* p � .05.

561LEFT VERSUS RIGHT MESIAL TEMPORAL LOBE EPILEPSY

Discussion

Neuropsychological studies of MTLE have concentratedlargely on memory functioning, as recently noted by Her-mann et al. (1997). A major finding in this literature is thatleft MTLE is associated with relative verbal memory defi-cits, whereas right MTLE is associated with relative non-verbal memory impairments (Delaney et al., 1980; Fedio &Mirsky, 1969; Helmstaedter et al., 1995; Hermann et al.,1987; Kim et al., 2003; Ladavas et al., 1979). This findingis thought to reflect the fact that the left mesial temporalregion is more important for verbal learning, whereas theright mesial temporal lobe is more important for nonverballearning. However, less is known about whether material-specific cognitive deficits in MTLE are confined to thelearning–memory domain or extend into more general in-telligence functions. Thus, the present study investigatedwhether left versus right MTLE has differential effects onmaterial-specific intelligence, as determined by VC and POfactor scores derived from K-WAIS. The results show thatwithin the left MTLE group, the VC factor score wassignificantly lower relative to the PO factor score, whereaswithin the right MTLE group, the PO factor score wassignificantly lower compared with the VC factor score.Between-groups comparisons showed that the left MTLEgroup had a significantly lower VC factor score than theright MTLE group, whereas the right MTLE group had asignificantly lower PO factor score than the left MTLEgroup. These findings provide strong evidence that left andright MTLE are associated with relative verbal and nonver-bal intelligence deficits, respectively.

Factor scores are, by nature, hypothetical. Thus, thegroup differences found with the factor scores might becriticized for lack of actuality. However, the differentialeffects of left versus right MTLE on material-specific intel-ligence were also apparent with the VC and PO index scoresas dependent measures. Thus, within the left MTLE group,the VC index score was significantly lower relative to thePO index score, whereas within the right MTLE group, thePO index score was significantly lower compared with theVC index score. Thus, both analyses involving actualisticand hypothetical measures of verbal and nonverbal intelli-gence supported the lateralized intelligence deficits hypoth-esis. In striking contrast to the present results, virtually allprior relevant studies (Abrahams et al., 1997; Blakemore etal., 1966; Bornstein, Pakalnis, et al., 1988; Breier et al.,1997; Fedio & Mirsky, 1969; Hermann et al., 1995; Her-mann et al., 1997; Kneebone et al., 1997; Mayeux et al.,1980; Moore & Baker, 1996; Selwa et al., 1994) failed tofind supporting evidence for the lateralized intelligencedeficits hypothesis. However, the great majority of thesestudies have used VIQ and PIQ as dependent measures.Even in our data, the material-specific intelligence deficitswere not apparent with VIQ and PIQ as dependent mea-sures. Thus, neither VIQ nor PIQ was significantly differentbetween the left and right MTLE groups. This finding aswell as prior failures to support the lateralized intelligencedeficits hypothesis presumably reflects, at least in part, thefact that VIQ and PIQ are relatively poor measures of verbal

and nonverbal intelligence. Thus, the present results under-score the importance of using sensitive measures of verbaland nonverbal intelligence in addressing the lateralized in-telligence deficits hypothesis.

The neural substrates underlying the material-specificintelligence deficits in MTLE may be widespread extratem-poral (and extrahippocampal temporal) lesions that are moresevere on the ipsilateral than the contralateral side of theepileptogenic region, as indicated by numerous reports ofquantitative MRI (e.g., Briellmann et al., 1998; DeCarli etal., 1998; Jutila et al., 2001; Kuzniecky et al., 1999; Marshet al., 1997; Moran et al., 2001; Sisodiya et al., 1997) andfunctional brain-imaging studies (e.g., Arnold et al., 1996;Avery et al., 2001; Duncan et al., 1996; Henry et al., 1990,1993; Jokeit et al., 1997; Van Bogaert et al., 2000; Yune etal., 1998). Consistent with this view, some studies havefound a significant relation between extratemporal abnor-malities in MTLE patients and their intellectual impair-ments (e.g., Arnold et al., 1996; Jokeit et al., 1997; Martinet al., 2000; Rausch, Henry, Ary, Engel, & Mazziotta,1994). For example, Rausch et al. (1994) reported thatrelative reductions in glucose metabolism of the left hemi-sphere and left lateral temporal lobe in MTLE patientscorrelated with a lower VIQ score. However, not all aspectsof the material-specific intelligence deficits in MTLE pa-tients may be a direct result of neurological lesions butsecondary to their material-specific memory impairments.For example, the verbal memory deficits associated with leftMTLE may interfere more with acquisition of verbal thannonverbal knowledge. Thus, the relatively depressed verbalintelligence in left MTLE patients could be, at least in part,secondary to their relatively low verbal memory.

Right MTLE was also associated with relative deficits inthe WM factor compared with left MTLE. This finding wassomewhat surprising because the Digit Span and Arithmeticsubscales are at least partially verbal. However, the presentpattern of findings is not without precedence. Hermann et al.(1995) authored one of the few studies that has comparedthe Wechsler subscale scores of left and right MTLE pa-tients. The data reported in this study indicate that the DigitSpan and Arithmetic subscale scores are nonsignificantlylower for right than for left MTLE patients. Testing post-surgical temporal lobe epilepsy patients, Lansdell (1968)reported that the right temporal group showed poorer free-dom-from-distractibility (FD) performance relative to theleft temporal group. This postsurgical group difference re-ported by Lansdell may also have been present presurgi-cally, although to a lesser extent. The greater impairment onWM tasks in right than left MTLE may be related to righthemisphere dominance for certain attentional functions(Heilman & Van Den Abell, 1980; Roy, Reuter-Lorenz,Roy, Copland, & Moscovitch, 1987). However, in brainlesions other than MTLE, such as cerebral vascular acci-dents and tumors, the Digit Span and Arithmetic subscalestend to be performed less well by patients with left-braindamage than by patients with right-brain damage (e.g.,Black, 1986; De Renzi & Nichelli, 1975; Russell, 1979;Warrington, James, & Maciejewski, 1986; Zillmer,Waechtler, Harris, Khan, & Fowler, 1992). For example,

562 KIM, YI, SON, AND KIM

testing patients with unilateral brain damage of variousetiologies, De Renzi and Nichelli (1975) found that digitrepetition was impaired in patients with left-brain damagebut not in patients with right-brain damage. Thus, the hy-pothesis that WM performance is more adversely affectedby left than right hemisphere lesions may be limited topathological processes associated with unilateral MTLE.Clearly, a larger study including MTLE as well as otherbrain lesion groups, with both cognitive and brain abnor-mality measures, is needed to fully address these complexbut potentially important issues.

The present results indicate that left MTLE is associatedwith lower VC than PO scores, whereas right MTLE isassociated with lower PO than VC scores. These resultssuggest the possible use of VC versus PO difference topredict the side of seizure onset. However, the mean differ-ences in VC versus PO scores were of relatively smallmagnitude for either the left or right MTLE group, suggest-ing their limited clinical utility. Thus, for example, applyingthe clinical criterion of a 15-point difference (i.e., the 25thpercentile) between VC index and PO index to the presentsample provided correct lateralization in only 18 of 71patients (25%), with false lateralization in 5 patients (7%).Forty-eight patients (68%) were considered indeterminateby this criterion. This classification rate, although poor,compares favorably to the classification rates reported inprior studies based on verbal versus nonverbal memorydifference (e.g., Kneebone et al., 1997; Loring, Hermann,Lee, Drane, & Meador, 2000; Naugle, Chelune, Schuster,Luders, & Comair, 1994; Wilde et al., 2001). For example,Wilde et al. (2001) reported that the criterion of a 17-pointdifference between Auditory versus Visual Delayed Indexesfrom Wechsler Memory Scale—III (Wechsler, 1997) pro-vided correct lateralization in 27 of 101 (27%) MTLEpatients, with false lateralization in 12 (12%). Thus, thelateralizing value of verbal versus nonverbal intelligencedifference may be at least as high as the lateralizing value ofverbal versus nonverbal memory difference. Future studiesmay address whether a combination of verbal versus non-verbal intelligence difference measures with verbal versusnonverbal memory difference measures may provide animproved rate of seizure laterality prediction.

In summary, previous studies have established that cog-nitive deficits in MTLE patients are not limited to thelearning–memory domain but include more global intelli-gence functions. The present study extends the findings byshowing that material-specific cognitive deficits in MTLEpatients are also not limited to the learning–memory domainbut include more global intelligence functions. Thus, leftMTLE was associated with a depressed VC factor but arelatively spared PO factor, whereas right MTLE was asso-ciated with lowered PO factor but a relatively preserved VCfactor. The neural substrates underlying these material-spe-cific intelligence impairments may be widespread extratem-poral (and extrahippocampal temporal) lesions that are moresevere on the ipsilateral than the contralateral side of theepileptogenic region. Prior multiple failures to find evidencefor the lateralized intelligence deficits hypothesis may re-flect, at least in part, use of VIQ and PIQ scores, which are

relatively poor indices of verbal and nonverbal intelligence.Consistent with this view, even in our data, lateralizedmaterial-specific intelligence deficits were not apparent withVIQ and PIQ as dependent measures. However, there areprevious investigations (e.g., Hermann et al., 1995) thatused VC and PO scores but that failed to find lateralizedintelligence effects. Thus, certain factors other than choiceof dependent measures, such as certain patient characteris-tics, may also have significant effects on whether lateralityfindings are obtained or not. Addressing what these otherfactors might be may further elucidate the link betweenMTLE and material-specific intelligence deficits.

References

Abrahams, S., Pickering, A., Polkey, C. E., & Morris, R. G.(1997). Spatial memory deficits in patients with unilateral dam-age to the right hippocampal formation. Neuropsychologia, 35,11–24.

Arnold, S., Schlaug, G., Niemann, H., Ebner, A., Luders, H.,Witte, O. W., et al. (1996). Topography of interictal glucosehypometabolism in unilateral mesiotemporal epilepsy. Neurol-ogy, 46, 1422–1430.

Avery, R. A., Zubal, I. G., Studholme, C., Slawski, J., Corsi, M.,Spencer, D. D., et al. (2001). Interictal 99mTc-HMPAO SPECTin temporal lobe epilepsy: Relation to clinical variables. Epilep-sia, 42, 869–874.

Baxendale, S. A., Thompson, P. J., & van Paesschen, W. (1998).A test of spatial memory and its clinical utility in the pre-surgical investigation of temporal lobe epilepsy patients. Neu-ropsychologia, 36, 591–602.

Black, F. W. (1986). Digit repetition in brain-damaged adults:Clinical and theoretical implications. Journal of Clinical Psy-chology, 42, 770–782.

Blakemore, C. B., Ettlinger, G., & Falconer, M. A. (1966). Cog-nitive abilities in relation to frequency of seizures and neuropa-thology of the temporal lobes in man. Journal of Neurology,Neurosurgery, and Psychiatry, 29, 268–272.

Bornstein, R. A., Drake, M. E., & Pakalnis, A. (1988). WAIS–Rfactor structure in epileptic patients. Epilepsia, 29, 14–18.

Bornstein, R. A., Pakalnis, A., & Drake, M. E. (1988). Verbal andnonverbal memory and learning in patients with complex partialseizures of temporal lobe origin. Journal of Epilepsy, 1, 203–208.

Breier, J. I., Brookshire, B. L., Fletcher, J. M., Thomas, A. B.,Plenger, P. M., Wheless, J. W., et al. (1997). Identification ofside of seizure onset in temporal lobe epilepsy using memorytests in the context of reading deficits. Journal of Clinical andExperimental Neuropsychology, 19, 161–171.

Briellmann, R. S., Jackson, G. D., Kalnins, R., & Berkovic, S. F.(1998). Hemicranial volume deficits in patients with temporallobe epilepsy with and without hippocampal sclerosis. Epilep-sia, 39, 1174–1181.

Burgess, A., Flint, J., & Adshead, H. (1992). Factor structure of theWechsler Adult Intelligence Scale—Revised (WAIS–R): Aclinical sample. British Journal of Clinical Psychology, 31,336–338.

Chapman, L. J., & Chapman, J. P. (1987). The measurement ofhandedness. Brain and Cognition, 6, 175–183.

Chelune, G. J., Naugle, R. I., Luders, H., & Awad, I. A. (1991).Prediction of cognitive change as a function of preoperativeability status among temporal lobectomy patients seen at6-month follow-up. Neurology, 41, 399–404.

563LEFT VERSUS RIGHT MESIAL TEMPORAL LOBE EPILEPSY

Choi, M. R., & Kim, Z. S. (1990). Korean Wechsler IntelligenceScale (KWIS) scatter analysis for psychiatric patients comparedwith neurotic patients. Korean Journal of Clinical Psychol-ogy, 9, 192–205.

Dai, X., Gong, Y., & Zhong, L. (1990). Factor analysis of themainland Chinese version of the Wechsler Adult IntelligenceScale. Psychological Assessment, 2, 31–34.

DeCarli, C., Hatta, J., Fazilat, S., Fazilat, S., Gaillard, W. D., &Theodore, W. H. (1998). Extratemporal atrophy in patients withcomplex partial seizures of left temporal origin. Annals of Neu-rology, 43, 41–45.

Delaney, R. C., Rosen, A. J., Mattson, R. H., & Novelly, R. A.(1980). Memory function in focal epilepsy: A comparison ofnon-surgical, unilateral temporal lobe and frontal lobe samples.Cortex, 16, 103–117.

De Renzi, E., & Nichelli, P. (1975). Verbal and non-verbal short-term memory impairment following hemisphere damage. Cor-tex, 11, 341–354.

Duncan, R., Patterson, J., Hadley, D., Roberts, R., & Bone, I.(1996). Interictal temporal hypoperfusion is related to early-onset temporal lobe epilepsy. Epilepsia, 37, 134–140.

Engel, J. (1996). Introduction to temporal lobe epilepsy. EpilepsyResearch, 26, 141–150.

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). NewYork: Raven Press.

Fedio, P., & Mirsky, A. F. (1969). Selective intellectual deficits inchildren with temporal lobe or centrencephalic epilepsy. Neu-ropsychologia, 7, 287–300.

Fowler, P. C., Richards, H. C., & Boll, T. J. (1980). WAIS factorpatterns of epileptic and normal adults. Journal of ClinicalNeuropsychology, 2, 115–123.

Gold, J. M., Blaxton, T. A., Hermann, B. P., Randolph, C., Fedio,P., Goldberg, T. E., et al. (1995). Memory and intelligence inlateralized temporal lobe epilepsy and schizophrenia. Schizo-phrenia Research, 17, 59–65.

Heilman, K. M., & Van Den Abell, T. (1980). Right hemispheredominance for attention: The mechanism underlying hemi-spheric asymmetries of inattention (neglect). Neurology, 30,327–330.

Helmstaedter, C., Pohl, C., & Elger, C. E. (1995). Relationsbetween verbal and nonverbal memory performance: Evidenceof confounding effects particularly in patients with right tem-poral lobe epilepsy. Cortex, 31, 345–355.

Henry, T. R., Mazziotta, J. C., & Engel, J. (1993). Interictalmetabolic anatomy of mesial temporal lobe epilepsy. Archivesof Neurology, 50, 582–589.

Henry, T. R., Mazziotta, J. C., Engel, J., Christenson, P. D., Zhang,J. X., Phelps, M. E., et al. (1990). Quantifying interictal meta-bolic activity in human temporal lobe epilepsy. Journal ofCerebral Blood Flow and Metabolism, 10, 748–757.

Hermann, B. P., Gold, J., Pusakulich, R., Wyler, A. R., Randolph,C., Rankin, G., et al. (1995). Wechsler Adult IntelligenceScale—Revised in the evaluation of anterior temporal lobec-tomy candidates. Epilepsia, 36, 480–487.

Hermann, B. P., Seidenberg, M., Schoenfeld, J., & Davies, K.(1997). Neuropsychological characteristics of the syndrome ofmesial temporal lobe epilepsy. Archives of Neurology, 54, 369–376.

Hermann, B. P., Wyler, A. R., Richey, E. T., & Rea, J. M. (1987).Memory function and verbal learning ability in patients withcomplex partial seizures of temporal lobe origin. Epilepsia, 28,547–554.

Ivnik, R. J., Sharbrough, F. W., & Laws, E. R., Jr. (1987). Effectsof anterior temporal lobectomy on cognitive function. Journal ofClinical Psychology, 43, 128–137.

Jokeit, H., Seitz, R. J., Markowitsch, H. J., Neumann, N., Witte,O. W., & Ebner, A. (1997). Prefrontal asymmetric interictalglucose hypometabolism and cognitive impairment in patientswith temporal lobe epilepsy. Brain, 120, 2283–2294.

Jutila, L., Ylinen, A., Partanen, K., Alafuzoff, I., Mervaala, E.,Partanen, J., et al. (2001). MR volumetry of the entorhinal,perirhinal, and temporopolar cortices in drug-refractory tempo-ral lobe epilepsy. American Journal of Neuroradiology, 22,1490–1501.

Kaufman, A. S. (1990). Assessing adolescent and adult intelli-gence. Boston: Allyn & Bacon.

Kim, H. (2002). Norms for 3-factor interpretation of K-WAIS.Korean Journal of Clinical Psychology, 21, 631–645.

Kim, H., Yi, S., Son, E. I., & Kim, J. (2003). Material-specificmemory in temporal lobe epilepsy: Effects of seizure lateralityand language dominance. Neuropsychology, 17, 59–68.

Kneebone, A. C., Chelune, G. J., & Luders, H. O. (1997). Indi-vidual patient prediction of seizure lateralization in temporallobe epilepsy: A comparison between neuropsychological mea-sures and the intracarotid amobarbital procedure. Journal of theInternational Neuropsychological Society, 3, 159–168.

Kuzniecky, R., Bilir, E., Gilliam, F., Faught, E., Martin, R., &Hugg, G. (1999). Quantitative MRI in temporal lobe epilepsy:Evidence for fornix atrophy. Neurology, 53, 496–501.

Ladavas, E., Umilta, C., & Provinciali, L. (1979). Hemisphere-dependent cognitive performances in epileptic patients. Epilep-sia, 20, 493–502.

Lansdell, H. (1968). The use of factor scores from the Wechsler–Bellevue scale of intelligence in assessing patients with tempo-ral lobe removals. Cortex, 4, 257–268.

Lee, J. W., Andermann, F., Dubeau, F., Bernasconi, A., Mac-Donald, D., Evans, A., et al. (1998). Morphometric analysis ofthe temporal lobe in temporal lobe epilepsy. Epilepsia, 39,727–736.

Loring, D. W., Hermann, B. P., Lee, G. P., Drane, D. L., &Meador, K. J. (2000). The Memory Assessment Scales andlateralized temporal lobe epilepsy. Journal of Clinical Psychol-ogy, 56, 563–570.

Loring, D. W., Lee, G. P., Martin, R. C., & Meador, K. J. (1988).Material-specific learning in patients with partial complex sei-zures of temporal lobe origin: Convergent validation of memoryconstructs. Journal of Epilepsy, 1, 53–59.

Loring, D. W., Lee, G. P., & Meador, K. J. (1988). Revising theRey-Osterrieth: Rating right hemisphere recall. Archives ofClinical Neuropsychology, 3, 239–247.

Margerison, J. H., & Corsellis, J. A. N. (1966). Epilepsy and thetemporal lobes. Brain, 89, 499–530.

Marsh, L., Morrell, M. J., Shear, P. K., Sullivan, E. V., Freeman,H., Marie, A., et al. (1997). Cortical and hippocampal volumedeficits in temporal lobe epilepsy. Epilepsia, 38, 576–587.

Martin, R. C., Sawrie, S. M., Gilliam, F. G., Palmer, C. A., Faught,E., Morawetz, R. B., et al. (2000). Wisconsin card sortingperformance in patients with temporal lobe epilepsy: Clinicaland neuroanatomical correlates. Epilepsia, 41, 1626–1632.

Mayeux, R. M., Brandt, J., Rosen, J., & Benson, D. F. (1980).Interictal memory and language impairment in temporal lobeepilepsy. Neurology, 30, 120–125.

Milner, B. (1975). Psychological aspects of focal epilepsy and itsneurosurgical management. In D. P. Purpura, J. K. Penry, &R. D. Walter (Eds.), Advances in neurology (Vol. 8, pp. 299–321). New York: Raven Press.

564 KIM, YI, SON, AND KIM

Moore, P. M., & Baker, G. A. (1996). Validation of the WechslerMemory Scale—Revised in a sample of people with intractabletemporal lobe epilepsy. Epilepsia, 37, 1215–1220.

Moran, N. F., Lemieux, L., Kitchen, N. D., Fish, D. R., & Shorvon,S. D. (2001). Extrahippocampal temporal lobe atrophy in tem-poral lobe epilepsy and mesial temporal sclerosis. Brain, 124,167–175.

Naugle, R. I., Chelune, G. J., Schuster, J., Luders, H. O., &Comair, Y. (1994). Recognition memory for words and facesbefore and after temporal lobectomy. Assessment, 1, 373–381.

Rausch, R., Henry, T. R., Ary, C. M., Engel, J., & Mazziotta, J.(1994). Asymmetric interictal glucose hypometabolism andcognitive performance in epileptic patients. Archives of Neurol-ogy, 51, 139–144.

Roy, E. A., Reuter-Lorenz, P., Roy, L. G., Copland, S., & Mosco-vitch, M. (1987). Unilateral attention deficits and hemisphericasymmetries in the control of attention. In M. Jeannerod (Ed.),Neurophysiological and neuropsychological aspects of spatialneglect (pp. 25–39). Amsterdam: North-Holland.

Russell, E. W. (1979). Three patterns of brain damage on theWAIS. Journal of Clinical Psychology, 35, 611–620.

Ryan, J. J., & Schneider, J. A. (1986). Factor analysis of theWechsler Adult Intelligence Scale—Revised (WAIS–R) in abrain-damaged sample. Journal of Clinical Psychology, 42,962–964.

Selwa, L. M., Berent, S., Giordani, B., Henry, T. R., Buchtel,H. A., & Ross, D. A. (1994). Serial cognitive testing in temporallobe epilepsy: Longitudinal changes with medical and surgicaltherapies. Epilepsia, 35, 743–749.

Sisodiya, S. M., Moran, N., Free, S. L., Kitchen, N. D., Stevens,J. M., Harkness, W. F. J., et al. (1997). Correlation of wide-spread preoperative magnetic resonance imaging changes withunsuccessful surgery for hippocampal sclerosis. Annals of Neu-rology, 41, 490–496.

Specht, U., May, T., Schulz, R., Rohde, M., Ebner, A., Schmidt,R. C., et al. (1997). Cerebellar atrophy and prognosis aftertemporal lobe resection. Journal of Neurology, Neurosurgery,and Psychiatry, 62, 501–506.

Van Bogaert, P., Massager, N., Tugendhaft, P., Wikler, D., Dam-haut, P., Levivier, M., et al. (2000). Statistical parametric map-ping of regional glucose metabolism in mesial temporal lobeepilepsy. NeuroImage, 12, 129–138.

Waller, N. G., & Waldman, I. D. (1990). A reexamination of theWAIS–R factor structure. Psychological Assessment, 2, 139–144.

Ward, L. C., Ryan, J. J., & Axelrod, B. N. (2000). Confirmatoryfactor analyses of the WAIS–III standardization data. Psycho-logical Assessment, 12, 341–345.

Warrington, E. K., James, M., & Maciejewski, C. (1986). TheWAIS as a lateralizing and localizing diagnostic instrument: Astudy of 656 patients with unilateral cerebral lesions. Neuropsy-chologia, 24, 223–239.

Wechsler, D. (1981). Wechsler Adult Intelligence Scale—Revised.San Antonio, TX: Psychological Corporation.

Wechsler, D. (1997). Wechsler Memory Scale-Third Edition. SanAntonio, TX: Psychological Corporation.

Wilde, N., Strauss, E., Chelune, G. J., Loring, D. W., Martin,R. C., Hermann, B. P., et al. (2001). WMS–III performance inpatients with temporal lobe epilepsy: Group differences andindividual classification. Journal of the International Neuropsy-chological Society, 7, 881–891.

Yeom, Y., Park, Y., Oh, K., Kim, J., & Lee, Y. (1992). K-WechslerAdult Intelligence Scale. Seoul, Korea: Korea Guidance.

Yune, M. J., Lee, J. D., Ryu, Y. H., Kim, D. I., Lee, B. I., & Kim,S. J. (1998). Ipsilateral thalamic hypoperfusion on interictalSPECT in temporal lobe epilepsy. Journal of Nuclear Medi-cine, 39, 281–285.

Zillmer, E. A., Waechtler, C., Harris, B., Khan, F., & Fowler, P. C.(1992). The effects of unilateral and multifocal lesions on theWAIS—R: A factor analytic study of stroke patients. Archivesof Clinical Neuropsychology, 7, 29–40.

Received June 3, 2002Revision received January 8, 2003

Accepted January 27, 2003 �

565LEFT VERSUS RIGHT MESIAL TEMPORAL LOBE EPILEPSY