Neuropsychological tests with lateralizing value in patients with temporal lobe epilepsy: Reconsidering material-specific theory

Transcription

1 Seizure (2005) 14, Neuropsychological tests with lateralizing value in patients with temporal lobe epilepsy: Reconsidering material-specific theory Toni Raspall, Marta Doñate, Teresa Boget, Mar Carreño, Antonio Donaire, Rolando Agudo, Núria Bargalló, Jordi Rumià, Xavier Setoain, Luis Pintor, Manel Salamero * Hospital Universitari Clinic, Servei Psicologia Clinica, Villarroel 140, Barcelona, Spain Received 14 September 2004; received in revised form 19 August 2005; accepted 19 September 2005 KEYWORDS Temporal lobe epilepsy; Neuropsychology; Cognitive impairment; Memory and naming Summary Purpose: To assess the ability of neuropsychological tests to determine the side of seizure onset for preoperative assessment in patients with drug-resistant temporal lobe epilepsy. Methods: Twenty-nine consecutive patients diagnosed with temporal lobe epilepsy (TLE), in whom the epileptogenic focus was clearly identified and localized to either the right or left hemisphere. Patients underwent a full neuropsychological assessment as part of their pre-surgical investigation, including the Boston Naming Test (BNT) and a variety of Wechsler Adult Intelligence Scale-Third Edition (WAIS-III) and Wechsler Memory Scale-Third Edition (WMS-III) subtests. Two multivariate analyses of variance were carried out to assess differences on memory and language measures between groups according to side of epileptogenic focus. Binary logistic regression analysis was performed to find the sets of tests that best predicted the side of seizure onset (determined by EEG and MRI). Results: Memory multivariate analysis of variance failed to show significant differences between the right- and left-sided groups. Among language measures, only the BNT revealed significant differences between the groups. The neuropsychological measures that best predicted the side of seizure onset were the BNT and Visual Reproduction II. Conclusions: Language measures predict the side of seizure focus better than memory measures. The results of this study in a sample of drug-resistant temporal lobe epilepsy patients challenge the memory material-specific theory for the side of seizure focus. # 2005 BEA Trading Ltd. Published by Elsevier Ltd. All rights reserved. * Corresponding author. Tel.: ; fax: address: salamero@clinic.ub.es (M. Salamero) /$ see front matter # 2005 BEA Trading Ltd. Published by Elsevier Ltd. All rights reserved. doi: /j.seizure

2 570 T. Raspall et al. Introduction Approximately 30 40% of epileptic patients are resistant to antiepileptic drugs to some degree. 1 Advances in diagnostic techniques such as videoelectroencephalography and magnetic resonance imaging and the development of a range of therapeutic surgical procedures have meant that a growing number of patients with diverse syndromes or epileptogenic foci can benefit from surgery. 2 Temporal lobe epilepsy is listed in the International League Against Epilepsy (ILAE) s International Classification of Epilepsy and Epileptic Syndromes, 3 as a condition in which focal seizures originate in a region of the temporal lobe. Patients with temporal lobe epilepsy are the largest group in surgical series 4,5 and also have the best seizure outcome. 6,7 Neuropsychological assessment is an essential component of the preoperative evaluation of patients with medically intractable epilepsy who are candidates for epilepsy surgery. 8 The assessment aims to determine the site of cerebral dysfunction by obtaining functional information to complement the findings of other exploratory techniques. 9 Furthermore, it provides a baseline performance for use as a reference point for evaluating cognitive surgical outcome. 10 Mesial temporal structures are especially vulnerable to epileptogenesis 11 and, given the region s involvement in mnemonic abilities, research with patients with temporal lobe epilepsy has made a significant contribution to the development of neuropsychological models of memory. 12 Memory disturbances have consistently been found in patients with temporal lobe seizures because of the relationship between mesial temporal structures and memory processes 13 and moreover, several studies have reported a material-specific memory impairment in these patients according to the side of epileptogenic focus: left temporal (LT) lobe seizures appear to be related to a deficit in verbal memory processes and right temporal (RT) lobe seizures to visuospatial memory impairment Several studies have reported evidence of the material-specific theory for verbal memory and drug-resistant LT seizures 17,18 and for RT seizures and visual memory tests. 8,12,19 However, other investigators have failed to detect differential impairment in verbal and visuospatial tasks according to seizure laterality or only report differences when considering delayed memory measures. 28 Moreover, naming is the subjective complaint most frequently reported by patients with epilepsy. 29 The vast majority of studies report that a naming ability deficit is especially prevalent in patients with temporal seizures originating in the language-dominant hemisphere. 17,30 37 Saykin et al. 34 and Schefft et al. 36 found poorer naming performance in patients with LT seizures than in patients with RT seizures. Other authors 37,38 observed a similar deterioration in the two groups with regard to normative data, or found no differences. 39 So one would expect to find scores on the memory subtests to differ according to the type of sensory stimulus involved. The material-specific theory predicts that patients with LT seizures will perform worse on verbal memory measures than patients with RT seizures, and better on visual memory measures, and that patients with LT seizures will perform worse on naming tasks. The present study has two aims: first, to compare cognitive performance in different cognitive domains (declarative memory and language) in patients with drug-resistant temporal seizures, focusing on the influence of hemisphere of the epileptogenic region (left, right); second, to identify which subtests in the neuropsychological battery best predict the epileptogenic hemisphere of patients in this sample. Methods Subjects The sample consisted of 29 consecutive patients diagnosed with drug-resistant temporal lobe epilepsy (TLE) in whom the epileptogenic focus was clearly identified and localized to either the right or left hemisphere. Patients underwent a full neuropsychological assessment as part of their pre-surgical evaluation for epilepsy surgery at the Epilepsy Unit of the Hospital Clínic i Provincial of Barcelona between 2002 and Inclusion criteria were: age 18 years or over, intellectual ability average or above (FIQ > 80), right-handedness, and a well-established seizure origin (demonstrated by video-eeg and neuroimaging). No patients suffered from neurodegenerative disease or expansive tumor. Basic clinical and demographic data are shown in Table 1. Patients were grouped according to side of epileptogenic region (left versus right), determined by ictal EEG and neuroimaging findings. Seizure onset was left temporal (LT) in 12 (41.4%) and right temporal (RT) in 17 (58.6%). Twenty (68.9%) patients exhibited signs of hippocampal sclerosis (hippocampal atrophy and high signal intensity) on magnetic resonance. In the LT group, 9 (75%) exhibited signs of hippocampal sclerosis ipsilateral to the epileptogenic hemisphere. In the RT group, 11

3 Lateralizing value in patients with temporal lobe epilepsy 571 Table 1 Clinical and demographic characteristics of the sample according to side of seizure. Temporal epileptogenic region LH (n = 12) RH (n = 17) Statistical test P-value Gender Women 7 8 x 2 = P = 0.55 Men 5 9 Age 38.2 (10.3) 39.1 (10.9) F = P = Age at start of regular seizures 13.5 (8.8) 17.7 (14.7) Mann Whitney U = 96 P = 0.79 Etiology of seizures Cryptogenic 1 1 Fisher s exact test P = Hippocampal sclerosis 9 11 Others 2 5 Education (years) 10(2.2) 9.8(2.4) F = P = IQ (WAIS-III) a 97.47(8.75) 98.32(7.81) F = P = a IQ-score (mean = 100 and standard deviation = 15). (64.7%) had right hippocampal sclerosis; in one the sclerosis was bilateral, though the right hippocampus (ipsilateral to the epileptogenic hemisphere) was more affected, presenting atrophy and high signal intensity. Of the patients without hippocampal sclerosis, five showed signs of temporal dysplasia (three RT and two LT) and two non-expansive tumoral etiology. The estimated IQ by using the Vocabulary, Block Design, Similarities and Digit Symbol subtests of the WAIS-III did not show differences between groups according to side of epileptogenic region (F = 0.076; P = 0.785). Neuropsychological assessment As a part of the preoperative assessment, a comprehensive neuropsychological test battery was administered, comprising tests covering the cognitive domains of declarative memory and language. The neuropsychological profile is grouped in Table 2 according to areas examined. We included certain subtests of the Wechsler Memory Scale-Third Edition, 40 the most recent revision of the original Wechsler Memory Scale. 41. The Wechsler Memory Scales are among the most commonly used tests for neuropsychological assessment in surgical centers for epilepsy, 10 but the results of the few studies that have using the entire scale have shown a limited ability to detect laterality dysfunction in preoperative patients with temporal lobe epilepsy. 42,43 In addition, administration of the scale is time-consuming, and Cañizares et al. 44 among others have demonstrated that abbreviated forms of the WMS-R, 45 the previous version of the WMS-III, are reliable and time-efficient for estimating memory indexes in patients with drug-resistant partial epilepsy. Our neuropsychological battery therefore included the following subtests from the WMS-III: Logical Memory I and II, Faces I and II, and Visual reproduction I and II. We also included the subtests of Vocabulary, Block Design, Digit Symbol and Similarities of the Wechsler Adult Intelligence Scale-Third Edition (WAIS- III), 46 the Rey Auditory-Verbal Learning Test (RAVLT), 47 the Boston Naming Test (BNT), 48 and the Animal Naming for semantic fluency. Statistical analysis Two multivariate analysis of variance (MANOVA) with a main factor side of the epileptogenic hemisphere (left versus right) were performed in order Table 2 Neuropsychological tests constituting the neuropsychological profile assessing each domain of cognitive functioning. Cognitive domain Neuropsychological tests Intelligence Vocabulary a, Similarities a, Block Design a, Digit Symbol a Declarative memory Logical Memory I (immediate) and II (delayed) b Faces I (immediate) and II (delayed) b Visual Reproduction I (immediate) and II (delayed) b RAVLT Total and delayed recall c Language Boston Naming Test (BNT) d Semantic Category (Animals) e a Wechsler Adult Intelligence Scale-Third Edition subtests (WAIS-III; Wechsler, 1997). b Wechsler Memory Scale-Third Edition subtests (WMS-III; Wechsler, 1997). c Rey Auditory/Verbal Learning Test (RAVLT 47 ). d Boston Naming Test (BNT 48 ). e Semantic Category (Category Naming Test; Morris et al., 1989).

4 572 T. Raspall et al. to analyse declarative memory and language domains. Scores on the Vocabulary subtest and age were used as covariates in the model. In the first MANOVA, we studied several declarative memory measures: RAVLT, Logical Memory, Visual Reproduction and Faces, including immediate and delayed recall. To study the language domain, BNT and semantic fluency were used as dependent variables in a MANOVA method. Finally, a multivariate binary logistic regression analysis with forward conditional stepwise method was performed to determine which neuropsychological variables best predict the side of seizure onset. MANOVA served for two purposes: first, to consider and control the intercorrelation between variablesthatistobeexpectedinalargebattery such as this; second, to avoid type 1 errors in the statistical decision process. The use of MANOVA reduced the probability of obtaining significant results by chance. Though the difference on the estimated IQ score was not significant, this score was used as a covariate in the model, in order to control the wellknown influence of intellectual capacity on memory and language cognitive domains. We also included age as a covariate to control its effect on the performance of some tests. Significant MANOVA effects were subsequently analysed by univariate analyses of variance (ANOVAs). Raw scores were used to increase the measure s sensitivity. Statistical analyses were performed with SPSS/PC+ version Results Clinical and demographic characteristics of the sample The analysis of the categorical variables did not differ significantly from those expected in terms of gender (x 2 = 0.358; P = 0.55) and the presence of hippocampal sclerosis (Fisher s exact test P = 0.694). The groups were therefore homogeneous for these variables. In order to compare groups in terms of age, age of epilepsy onset and education, several factor analyses of variance (ANOVA) with one main factor (hemisphere of the epileptic region) for quantitative variables were performed. Groups were homogeneous for age (F = 0.055; P = 0.816) and education (F = 0.018; P = 0.894). Age of epilepsy onset presented no significant differences according to the side of epileptic region (Mann Whitney U = 96; P = 0.790). Table 3 Descriptive data of scores obtained from neuropsychological tests according to side of seizure onset. Neuropsychological test battery Main results Temporal epileptogenic region (n = 29) LH (n = 12) RH (n = 17) IQ a (8.75) (7.81) Logical Memory I b 36.9 (12.0) 36.8 (10.1) RAVLT Total learning c 43.5 (9.3) 46.8 (8.9) Faces I b 35.1 (3.7) 35.3 (4.5) Visual Reproduction I b 78.0 (10.3) 80.8 (13.8) Logical Memory II b 20.6 (8.7) 22.0 (7.6) RAVLT (delayed recall) d 8.0 (2.7) 8.7 (2.9) Faces II b 35.3 (3.4) 35.8 (4.0) Visual Reproduction II b 66.2 (21.2) 57.0 (14.5) Animals e 14.8 (3.9) 15.7 (3.6) BNT f 44.5 (7.8) 50.9 (4.1) Data are described in raw scores, with mean and standard deviation in parentheses. a Estimated by using Vocabulary, Block Design, Similarities and Digit Symbol subtests of the WAIS-III (mean 100 and SD 15). b WMS-III subtests. c Sum of words correctly recalled over five trials (range between 0-75). d Words correctly recalled on delayed recall trial. e Total words. f Words correctly named. Performance of the groups on the subtest battery was average in general, with some results lowaverage. Descriptive data are displayed on Table 3. Declarative memory Various measures of declarative memory were studied in the first multivariate analysis of the variance: RAVLT, Logical Memory, Visual Reproduction and Faces, including immediate and delayed recall. Multivariate contrasts were statistically significant for the covariate of the model estimated IQ (Pillai s Trace = 0.536; F = 2.596; P = 0.044). The multivariate effect of side of epileptogenic region (right versus left) and the covariate age did not reach statistical significance (Pillai s Trace = 0.411; F = 1.571; P = 0.202; and Pillai s Trace = 0.464; F = 1.946; P = 0.115, respectively). Before interpreting the univariate results, we checked the homogeneity of variance between the groups for each of the dependent variables. All were homogeneous, except for Logical Memory I subtest (F = 7.367; P = 0.011), so we opted to use a non-parametric post hoc statistical analysis. Regarding the univariate results of the MANOVA, significant results were obtained for the models corresponding to immediate and delayed visual

5 Lateralizing value in patients with temporal lobe epilepsy 573 memory (Visual Reproduction I F = 4.313, P = 0.014; Visual Reproduction II F = 3.296, P = 0.037; Faces II F = 4.208, P = 0.015) and total RAVLT (F = 3.783, P = 0.023). Performance on Faces II, Visual Reproduction I and II were explained solely by the covariate estimated IQ (Faces II F = , P = 0.005; Visual Reproduction I F = 7.791, P = 0.010; Visual Reproduction II F = 4.101, P = 0.054). Total RAVLT was explained by the covariate estimated IQ (total RAVLT F = 4.778, P = 0.038), although it showed a trend toward statistical significance of the effect of the covariate age (F = 3.520, P = 0.072). The models of Logical Memory II (F = 2.491, P = 0.083), Faces I (F = 0.792, P = 0.510), and AVLT delayed recall (F = 1.767, P = 0.179) did not reach statistical significance. The Logical Memory I subtest was unsuitable for parametric testing and so nonparametric analysis was carried out. No significant effect was found for the side of epileptogenic region (Mann Whitney U = 101; P = 0.965). Language Multivariate contrasts showed a significant effect of the covariate estimated IQ on semantic category and naming (Pillai s Trace = 0.333; P = 0.008) and of the side of seizure (Pillai s Trace = 0.243; P = 0.035). The covariate age was not significant (Pillai s Trace = 0.040; P = 0.615). Levene s contrast, performed prior to the interpretation of univariate analyses, showed heterogeneity between group variances with regard to the variable naming (F = 8.23; P = 0.008). A non-parametric analysis was therefore carried out, demonstrating a significant effect of the side of seizure origin (Mann Whitney U = 44; P = 0.010) on the Boston Naming Test, LT seizure patients performing worse than their RT counterparts. The univariate model of semantic fluency was significant (F = 4.282; P = 0.014), explained solely by the covariate estimated IQ (F = ; P = 0.002); side of seizure did not reach statistical significance (F = 0.263; P = 0.612). Logistic regression analysis Binary logistic regression analysis was performed to establish whether the seizure foci (determined by EEG and MRI) could be predicted by the sets of neuropsychological variables (of memory and language) used in the preceding MANOVAs. We used a hierarchical procedure in which estimated IQ and age were entered in the first block, and the neuropsychological scores in the second. In our predictive model, estimated IQ and age were considered as possible confounding variables. As they were forced to enter at the beginning, we achieved statistical control over their possible influences. Regression analysis indicated that estimated IQ (t score = 0.082; P = 0.775; df = 1) and age (t score = 0.059; P = 0.807; df = 1) were not significant predictors of the hemisphere of the epileptogenic focus. BNT was the first variable to appear in the model and was therefore the best to predict the side of epileptogenic focus (correctly classifying 69% of the cases). Its predictive capacity was bigger in patients with RT lobe seizures (82.4%) than LT patients (50%). The next variable that increased the predictive value was Visual Reproduction II which, together with BNT, increased the model s predictive power to 82.8% (75% of the cases correctly lateralized to the LT group and 88.2% to the RT group). Among the remaining variables with greatest predictive value were Visual Reproduction I, although it did not increase the total percentage (which remained at 82.8% of the total sample). The fourth variable introduced was the Faces I subtest, which, together with the three previous variables correctly classified 89.7% (91.7% of cases in the LT group and 88.2% in the RT group). In the last step of the method, and to predict the subjects in this sample with an effectiveness of 100%, the total AVLT subtests (sum recalled over five trials) were included. Analyses of the relationships between these five variables, with the main objective of predicting the side of the epileptogenic focus, show that subjects with RT seizures scored better than LT subjects on the BNT (b = 0.229), Visual Reproduction I (b = 0.204) and total RAVLT (b = 5.807), but worse on Visual Reproduction II (b = 0.070) and Faces I (b = 0.437). Discussion The only subtest to present significant differences according to the epileptogenic hemisphere was the BNT, on which LT seizure focus patients performed worse overall than RTsubjects. These results broadly agree with those obtained in most previous studies, which suggest the existence of a slight but frequent deterioration of these functions in patients with an epileptogenic lesion in the language-dominant temporal lobe. 17,30 36 However, some studies have obtained similarly poor performance in the two groups with regard to normative data 37,38 or did not find differences. 39 This diversity of results could be explained by the differences in sample size and composition or by the criteria used to group subjects. In contrast, no significant differences were found between LTand RT patients in encoding and memory

6 574 T. Raspall et al. retrieval processes for either verbal or visual information; the degree of efficiency obtained by the two groups was similar, in general average and lowaverage. Our data from the various memory subtests used did not therefore confirm the material-specific theory for mnemonic performance. None of the memory subtests included in the neuropsychological assessment could differentiate on their own between patients performance according to the side of seizure focus. These results are consistent with those of previous investigations which did not find significant differences between RT and LT epileptogenic patients on visual memory tests 20,23,26,27 or on verbal memory tests ,24,25 However, other studies found significant differences between these patients. 8,17,19 When predicting the side of seizure focus with the whole battery, the neuropsychological subtest that was overall the best predictor was the Boston Naming Test (BNT), which assesses visual confrontation naming ability in response to pictures presented visually through 60 black and white pictures which vary in difficulty according to the frequency of their verbal label. This neuropsychological test, together with the Visual Reproduction delayed subtest (VR-II, from the WMS-III), correctly classified 75% of patients with LT seizures and 88.2% of patients with RT seizures. This better ability to lateralize right hemispheric dysfunction using WMS-III subtests corroborates the results of previous studies. 42,43 There are several possible reasons for the lack of a differential effect of side of seizure region on mnemonic tests. First, if we had used WMS-III primary indexes, we might conceivably have obtained differences between groups according to the side of focus. However, the results of the few studies published to date suggest that the ability of verbal and visual memory indexes of the WMS-III to detect laterality dysfunction is limited in preoperative patients with temporal lobe epilepsy, especially in cases with left temporal epilepsy, 42,43 though in fact closer examination of the primary results obtained in the study by Baker et al. 43 reveals significant differences on Logical Memory I and II subtests (P < 0.05) between patients with LTand RTseizures. Moreover, some of the subtests in the Visual Memory Index of the WMS-III have been subject to criticism. In a recent study, Dulay et al. 49 assessed the extent to which different cognitive abilities contribute to performance on the Family Pictures task in 125 patients evaluated for epilepsy surgery. Their results show that this task relies heavily on auditory-verbal based cognitive abilities, as well as on visual memory, and may in fact better represent a general measure of memory performance. These results raise questions about the appropriateness of including performance on this task in the calculations of the WMS-III Visual Index scores. Nor is the Visual Reproduction subtest of previous versions of the Wechsler Memory Scale immune from criticism. Studies comparing performance on Visual Reproduction (from the WMS-R) between patients with LT and RT seizures only rarely show differences between them, even when sample sizes are large enough to detect small size effects, suggesting that performance on this task from the WMS-R depends more on visuoperceptive and visuoconstructive functions than on memory functions per se. 27 The lack of specificity and validity of the visual subtest of the WMS-R, in which other cognitive functions seem to predominate above memory ability, could have reduced its ability to discriminate between temporal groups. Second, most of the evidence in support of the material-specific theory comes from studies on postsurgical samples. Neuropsychological assessment performed after epilepsy surgery may confound the potential deleterious effect on cognition of surgery-induced damage with a preexisting deficit related to structural or functional dysfunction secondary to seizure activity. In this regard, while some research has found specific differences on memory performance according to side of epileptogenic region, 16,50 other studies have failed to find consistent differences in presurgical samples 24 or report only weak trends. 51 Third, patients with temporal lobe epilepsy constitute a heterogeneous population in terms of clinical and electroencephalographic characteristics, neuroimaging findings, and also neuropsychological aspects, The presence of neurological risk factors in infancy as well as the prolonged dysfunction of cerebral tissue due to epileptiform activity, with or without macroscopic structural damage, may have led to an atypical intra- or inter-hemispheric reorganization of cerebral functions. 55,56 Sample size and differences in the neuropsychological tests used may account for the divergences in the results. Finally, we should note a number of limitations of this study. The sample size was quite small, especially the LT seizure group. This prevented us from splitting it into subgroups according to the presence of hippocampal sclerosis and analysing the possible effect on mnemonic performance. Moreover, we were unable to include bilateral temporal seizures due to a shortage of eligible patients. Another limitation is the lack of control of the effects of medication, since high doses and polytherapy are prevalent in the refractory population, 57 although these factors are believed to influence mnemonic processes in temporal seizure patients regardless of the side of the epileptogenic focus.

7 Lateralizing value in patients with temporal lobe epilepsy 575 Among the strong points of this study is the fact that all patients underwent an exhaustive presurgical assessment for epilepsy, the fact that the sample was homogeneous for handedness, and that subjects had average estimated IQ, thus ensuring the high internal validity of the design. The use of MANOVA as a statistical procedure reduced the probability of obtaining significant results by chance. In conclusion, patients with drug-resistant temporal lobe epilepsy show a very similar memory performance, both in encoding and retrieval processes, regardless of the epileptogenic hemisphere. Furthermore, patients with LT seizures perform worse on visual confrontation naming tasks. These results support the hemispheric laterality theory for language but challenge the material-specific theory on memory performance in preoperative temporal lobe epilepsy patients. Only when we consider together some tests, we can predict the side of epileptogenic hemisphere in an acceptable way in this population of TLE patients. Therefore, in our study, we needed at least two tests (BNT and Visual Reproduction II) to achieve a correctly lateralized epileptogenic hemisphere percentage of almost 83% of the sample, RT patients scoring better than LT on the BNT, but worse on Visual Reproduction II. The findings draw attention to the need to design neuropsychological batteries whose global interpretation is sensitive enough to detect any slight difference in encoding and retrieval memory processes for patients with LT and RT seizures. Acknowledgements The study was supported in part by a Fourth Division grant of the University of Barcelona awarded to Toni Raspall. References 1. Aicardi J, Shorvon SD. Intractable epilepsy. In: Engel Jr J, Pedley T, editors. Epilepsy. Philadelphia: Lippincott-Raven Publishers; p Engel Jr J, Cascino GD, Shields WD. Surgically remediable syndromes. In: Engel Jr J, Pedley T, editors. Epilepsy. Philadelphia: Lippincott-Raven Publishers; p Comission on Classification and Terminology of the International League Against Epilepsy. Proposal for classification of epilepsies and epileptic syndromes. Epilepsia 1985;26 (Suppl. 3): Duchowny MS, Harvey AS, Sperling MR, Williamson PD. Indications and criteria for surgical intervention. In: Engel Jr J, Pedley T, editors. Epilepsy. Philadelphia: Lippincott-Raven Publishers; p Wieser HG, Burcet J, Russi A. Indicaciones del tratamiento quirúrgico de la Epilepsia. Revista de Neurologia 2000;30: Engel Jr J, Shewmon DA. Who should be considered a surgical candidate? In: Engel Jr J, editor. Surgical treatment of the epilepsies. New York: Raven Press; p Berg AT, Vickrey BG. Outcome measures. In: Engel Jr J, Pedley T, editors. Epilepsy. Philadelphia: Lippincott-Raven Publishers; p Jones-Gotman M. Localization of lesions by neuropsychological testing. Epilepsia 1991;32(Suppl. 5):S Rausch R. Factors affecting neuropsychological and psychosocial outcome of epilepsy surgery. In: Lüders H, editor. Epilepsy surgery. New York: Raven Press Ltd.; p Jones-Gotman M, Smith ML, Zatorre RJ. Neuropsychological testing for localizing and lateralizaing the epileptogenic region. In: Engel Jr J, editor. Surgical treatment of the epilepsies. 2nd ed. New York: Raven Press, Ltd.; p Ben-Ari Y. Cell death and synaptic reorganizations produced by seizures. Epilepsia 2001;42: Helmstaedter C, Grunwald K, Lehnertz K, Gleissner U, Elger CE. Differential involvement of left temporolateral and temporomesial strucutres in verbal declarative learning and memory: evidence from temporal lobe epilepsy. Brain Cogn 1997;35: Squire LR, Zola SM. Structure and function of declarative and non-declarative memory systems. Proc Natl Acad Sci USA 1996;93: Milner B. Psychological aspects of focal epilepsy and its neurosurgical management. Adv Neurol 1975;8: Binnie CD, Kastelijn-Nolst Teninté DG, Smit AM, Wilkins AJ. Interaction of epileptiform EEG discharges and cognition. Epilepsy Res 1987;1: Helmstaedter C, Pohl C, Hufnagel A, Elger CE. Visual learning deficits in nonresected patients with right temporal lobe epilepsy. Cortex 1991;27: Sass KJ, Sass A, Westerveld M, Lencz T, Novelly RA, Kim JH, et al. Specificity in the correlation of verbal memory and hippocampal neuron loss: dissociation of memory, language, and verbal intellectual ability. J Clin Exp Neuropsychol 1992; 14: Helmstaedter C, Elger CE. Cognitive consequences of twothirds anterior temporal lobectomy on verbal memory in 144 patients: A three-month follow-up study. Epilepsia 1996; 37: Gleissner U, Helmstaedter C, Quiske A, Elger CE. The performance-complaint relationship in patients with epilepsy: A matter of daily demands? Epilepsy Res 1998;32: Powell GE, Polkey CE, McMillan T. The new Maudsley series of temporal lobectomy. I: Short-term cognitive effects. Br J Clin Psychol 1985;24: Ivnik RJ, Sharbrough FW, Laws ERJ. Effects of anterior temporal lobectomy on cognitive function. J Clin Psychol 1987; 43: Miller LA, Muñoz DG, Finmore M. Hippocampal sclerosis and human memory. Arch Neurol 1993;50: Snitz BE, Roman DD, Beniak TE. Efficacy of the Continuous Visual Memory Test in lateralizaing temporal lobe dysfunction in chronic complex-partial epilepsy. J Clin Exp Neuropsychol 1996;18: Thompson PJ, Trimble MR. Neuropsychological aspects of epilepsy. In: Grant I, Adams KM, editors. Neuropsychological assessment of neuropsychiatric disorders. New York: Oxford University Press; p Seidenberg M, Hermann BP, Dohan FCJ, Wyler AR, Perrine A, Schoenfeld J. Hippocampal sclerosis and verbal encoding

8 576 T. Raspall et al. ability following anterior temporal lobectomy. Neuropsychologia 1996;34(7): Baxendale SA. The role of the hippocampus in recognition memory. Neuropsychologia 1997;35: Barr WB, Chelune GJ, Hermann BP, Loring DW, Perrine K, Strauss E, et al. The use of figural reproduction tests as measures of nonverbal memory in epilepsy surgery candidates. J Int Neuropsychol Soc 1997;3: Sass KJ, Buchanan CP, Kraemer S, Westerveld M, Kim JH, Spencer DD. Verbal memory impairment resulting from hippocampal neuron loss among epileptic patients with structural lesions. Neurology 1995;45: Thompson PJ, Corcoran R. Everyday memory failures in people with epilepsy. Epilepsia 1992;33(Suppl. 6):S Mayeux M, Brandt J, Rosen J, Benson DF. Interictal memory and language impairment in temporal lobe epilepsy. Neurology 1980;30: Hermann BP, Wyler AR, Somes G. Language function following anterior temporal lobectomy. J Neurosurg 1991;74: Davies KG, Maxwell RE, Jennum P, Dhuna A, Beniak TE, Destafney E, et al. Language function following subdural grid-directed temporal lobectomy. Acta Neurol Scand 1994;90: Davies G, Maxwell RE, Beniak TE, Destafney E, Fiol ME. Language function after temporal lobectomy without stimulation mapping of cortical function. Epilepsia 1995;36: Saykin AJ, Stafiniak P, Robinson LJ, Flannery KA, Gur RC, O Connor MJ, et al. Language before and after temporal lobectomy: specificity of acute changes and relation to early risk factors. Epilepsia 1995;36: Ogden-Epker M, Cullum CM. Quantitative and qualitative interpretation of neuropsychological data in the assessment of temporal lobectomy candidates. Clin Neuropsychol 2001; 5: Schefft BK, Marc Testa S, Dulay MF, Privitera MD, Yeh HS. Preoperative assessment of confrontation naming ability and interictal paraphasia production in unilateral temporal lobe epilepsy. Epilepsy Behav 2003;4: Hermann BP, Wyler AR. Effects of anterior temporal lobectomy on language function: a controlled study. Ann Neurol 1988;23: Langfitt JT, Rausch R. Word-finding deficits persist after left anterotemporal lobectomy. Arch Neurol 1996;53: Stafiniak P, Saykin AJ, Sperling MR, Kester DB, Robinson LJ, O Connor MJ, et al. Acute naming deficits following dominant temporal lobectomy: prediction by age at 1st risk for seizures. Neurology 1990;40: Wechsler D. Wechsler Memory Scale. 3rd ed. San Antonio, TX: The Psychological Corporation; Wechsler D. A standarized memory scale for clinical use. J Psychol 1945;19: Wilde N, Strauss E, Chelune GJ, Loring DW, Martin RC, Hermann BP, et al. WMS-III performance in patients with temporal lobe epilepsy: group differences and individual classification. J Int Neuropsychol Soc 2001;7: Baker GA, Austin NA, Downes JJ. Validation of the Wechsler Memory Scale-III in a population of people with intractable temporal lobe epilepsy. Epilepsy Res 2003;53: Cañizares S, Boget T, Salamero M, Rumià J, Elices E, Arroyo S. Reliability and clinical usefulness of the short-forms of the Wechsler memory scale (revised) in patients with epilepsy. Epilepsy Res 2000;41: Wechsler D. Manual for the Wechsler Memory Scale-Revised. New York: The Psychological Corporation; Wechsler D. Wechsler Adult Intelligence Scale. 3rd ed. San Antonio, TX: The Psychological Corporation; Rey A. L examen clinique en psychologie. Paris: Presses Universitaires de France; Kaplan EF, Goodglass H, Weintraub S. The Boston Naming Test. 2nd ed. Philadelphia: Lea & Febiger; Dulay MF, Schefft BK, Testa SM, Fargo JD, Privitera M, Yeh HS. What does the family pictures subtest of the Wechsler Memory Scale-III measure? Insight gained from patients evaluated for epilepsy surgery. Clin Neuropsychol 2002;16: Giovagnoli AR, Avanzini G. Learning and memory impairment in patients with temporal lobe epilepsy: relation to the presence, type and location of brain lesion. Epilepsia 1999;40: Sunderland A. Neuropsychological aspects of epilepsy. Curr Opin Psychiatry 1993;6: Glasser GH. Natural history of temporal lobe-limbic epilepsy. In: Engel J, editor. Surgical treatment of the epilepsies. New York: Raven Press; p Wieser HG, Willamson P. Ictal semiology. In: Engel J, editor. Surgical tretament of the epilepsies. New York: Raven Press; p Paradiso S, Herman BP, Robinson RG. The heterogeneity of temporal lobe epilepsy: neurological, neuropsychological and psychiatric aspects. J Nerv Ment Dis 1995;183: Satz P, Strauss E, Wada J, Orsini DL. Some correlates of intra and interhemispheric speech organization after left focal brain injury. Neuropsychologia 1988;26: Holloway V, Gadian DG, Vargha-Khadem F, Porter DA, Boyd SG, Connelly A. The reorganization of sensoriomotor function in children after hemispherectomy. A functional MRI and somatosensory evoked potential study. Brain 2000;123: Thompson PJ. Antiepileptic drugs on memory. Epilepsia 1992; 33(Suppl. 6):S37 40.