Epilepsy & Behavior 44 (2015) 225–233

Contents lists available at ScienceDirect

Epilepsy & Behavior journal homepage: www.elsevier.com/locate/yebeh

Verbal memory and literacy outcomes one year after pediatric temporal lobectomy: A retrospective cohort study Suncica Lah a,b,⁎, Mary Lou Smith b,c,d,⁎⁎ a

Department of Psychology, University of Sydney, Sydney, NSW, Australia ARC Centre of Excellence in Cognition and Its Disorders, Australia Department of Psychology, University of Toronto Mississauga, Mississauga, ON, Canada d Neurosciences and Mental Health Program, Hospital for Sick Children, Toronto, ON, Canada b c

a r t i c l e

i n f o

Article history: Received 2 October 2014 Revised 28 December 2014 Accepted 30 December 2014 Available online 13 March 2015 Keywords: Episodic memory Semantic memory Reading Spelling Child epilepsy surgery Temporal lobe epilepsy

a b s t r a c t Objective: In children with temporal lobe epilepsy (TLE), temporal lobectomy (TL) is a treatment of choice for those children with seizure that are difficult to control with medication. Semantic memory is dependent on functional integrity of the temporal lobes and is thought to be critical for development of literacy skills. However, little is known about semantic memory and literacy outcomes post-TL in children. Method: In this retrospective cohort study, 40 children with TLE were administered tests of memory and literacy pre-TL and 1 year post-TL in one hospital between 1996 and 2011. Results: One year post-TL, 60% of the children became seizure-free. A significant decline was found in one aspect of semantic memory (naming) in children who underwent left TL. In addition, a significant drop was also evident in one aspect of literacy (reading accuracy), irrespective of the side of surgery. These declines were related neither to each other nor to epilepsy variables including seizure outcome. Conclusions: This is the largest pediatric outcome study of memory and literacy skills to date and shows that TL is associated with a risk of a mild drop in specific aspects of semantic memory (naming, following left TL) and reading accuracy, while other areas of memory and literacy remain unchanged. © 2015 Elsevier Inc. All rights reserved.

1. Introduction Children with temporal lobe epilepsy (TLE) are at risk of memory deficits (see [1] for review) and academic underachievement [2–6]. Deficits in episodic memory, new learning, and recall of information after delays have been well documented in this patient population (i.e., [7–9]). In contrast, deficits in semantic memory, the recall of factual information, general knowledge, naming, and word meaning have come to light only recently [10,11] Interestingly, episodic memory and semantic memory relate differently to literacy skills in children with TLE [12]; while semantic memory has strong relations with a range of literacy skills (reading comprehension, reading accuracy, and spelling accuracy), episodic memory has weak relations with reading and spelling accuracy and none with reading comprehension. This finding of semantic, rather than episodic, memory playing a prominent role in literacy skills is consistent with the adult neuropsychological literature.

⁎ Correspondence to: S. Lah, School of Psychology, The University of Sydney, NSW 2006, Australia. Tel.: +61 2 9351 2648; fax: +61 2 9036 5223. ⁎⁎ Correspondence to: M.L. Smith, Department of Psychology, University of Toronto Mississauga, Mississauga, ON L5L 1C6, Canada. Tel.: +1 905 828 3960; fax: +1 905 569 4326. E-mail addresses: [email protected] (S. Lah), [email protected] (M.L. Smith).

http://dx.doi.org/10.1016/j.yebeh.2014.12.040 1525-5050/© 2015 Elsevier Inc. All rights reserved.

For example, patients with semantic dementia who have impaired semantic memory, but relatively preserved episodic memory, often (although not always [13–15]) present with reading difficulties, [16–23]. Turning our attention to cognitive models of reading, we note that leading cognitive models of reading, the Dual Route/Dual Route Cascaded (DR: [24,25]; DRC: [26,27]) model and the Parallel Distributed Processing (PDP: [28,29]) model, both include semantic, but not episodic, nodes. Nevertheless, according to the DRC model, reading aloud can, but does not have to, involve the semantic system. By contrast, according to the PDP model, the semantic system is involved when reading aloud, especially when reading infrequent irregular words. Taken together, findings that semantic memory plays a role in reading of children with TLE and is associated with reading difficulties in patients with semantic dementia are primarily consistent with the PDP model. Finally, it is of interest to observe the adult neuropsychology literature, findings of which suggest that whilst different brain structures underpin semantic and episodic memory, common brain structures support semantic memory and reading. While episodic memory is primarily dependent on the mesial temporal structures and the hippocampus in particular (e.g., [30–32]), semantic memory and reading are both related to the integrity of the temporal neocortex [33,34]. Many children with unilateral onset TLE will undergo a temporal lobectomy (TL), as their epilepsy is often intractable to pharmacological,

226

S. Lah, M.L. Smith / Epilepsy & Behavior 44 (2015) 225–233

but responsive to surgical, treatment [35]. Typically, TL involves resection of the hippocampus (which is critical for episodic memory) and removal of the anterior temporal neocortex (which is purported to represent a semantic memory hub [36]). In adults, a resection of a dominant temporal lobe has been associated with a decline in verbal episodic memory (see [37] for review). Findings have been less consistent in children than in adults. For example, in a review of outcomes after TL [38], eight of the thirteen studies found no significant change in memory, four studies uncovered a deterioration (with mixed findings regarding the relation to laterality), and one study documented an increase in verbal memory (after right TL). Episodic memory outcomes, however, may not be representative of semantic memory outcomes, as we found that episodic memory and semantic memory could be impaired independently of one another presurgery in children with TLE [10]. A recent review of research with adults revealed a significant decline on one task of semantic memory, confrontational naming, measured with the Boston Naming Test [39] in 19 of the 21 studies [40]. In eleven of these studies, predictors of the decline were examined. Late age at onset (examined in 6 studies) and the absence of hippocampal sclerosis (3 studies) were associated with the decline. The one study which examined the impact of these two factors concurrently found the latter to be a more significant predictor of semantic memory outcome [41,42], which is surprising, as the anterior temporal neocortex, rather than the hippocampus, is thought to be critical for semantic memory [36]. As children considered for TL are likely to have earlier age at epilepsy onset compared with adults, a risk of semantic memory decline may be lower in children relative to adults. Indeed, Blanchette and Smith [43] found no change in semantic memory (recall of factual information and word meaning) post-TL in 10 children and no effect of side of resection, but no confrontational naming test was used. Jambaque and colleagues [7] found a significant improvement after TL (including hippocampectomy) in verbal episodic memory, as well as in naming, but no changes in recall of factual information or word meaning in 20 children. The relations between improvements in verbal episodic memory and verbal semantic memory (naming) were not examined in that study. The increase in either aspect of verbal memory was not related to the age at seizure onset or side of surgery. Nevertheless, while the increase in episodic memory was greater in children who had evidence of previous hippocampal damage, the increase in semantic memory was greater in children who were free of previous hippocampal damage. This dissociation raises a possibility that changes in semantic memory and episodic memory were independent of one another. Similarly, dissociation between episodic memory and semantic memory was found in a study by Smith and Lah [10] that involved children with TLE who were candidates for surgery; the factor analysis revealed two factors (semantic and episodic), and children could be impaired in either episodic memory or semantic memory alone. As semantic memory has been found to be closely related to literacy skills in children with TLE [12], in children with reading disability [44], and in typically developing children [44,45], it is likely that semantic memory outcome will be related to literacy skills postsurgery. On the one hand, the adult neuropsychological literature indicates that the resection of the temporal neocortex could place children at risk of a decline in the semantic memory and literacy skills. On the other hand, elimination of seizures and reduction of antiepileptic medication and their side effects may increase school attendance and result in children being more receptive to learning, which could promote development of literacy skills. However, no changes in reading and spelling skills (rather than an increase) and no impact of side of resection were found in the studies that involved children who underwent surgery for epilepsy [43,46]. We found no study that examined both semantic memory and episodic memory as well as academic skills in children with TLE pre-TL to post-TL. In this study, we aimed to (i) investigate presurgical to postsurgical changes in verbal memory (semantic and episodic) and literacy

(reading accuracy, reading comprehension, and spelling) and to (ii) examine relations between changes in memory and literacy. We predicted that (i) changes in verbal episodic and semantic memory would not be related and that (ii) changes in semantic memory, but not in episodic memory, would be related to changes in literacy. In addition, we explored whether epilepsy factors contributed to changes in memory and literacy. 2. Methods 2.1. Participants The study included 40 school-age children and adolescents who underwent unilateral temporal lobe surgery for treatment of intractable TLE (as determined by prolonged video-EEG monitoring, MRI, and, in some cases, magnetoencephalography and PET) at the Hospital for Sick Children (Toronto, ON, Canada) between 1996 and 2011. Over 90% of the operations were done by the same neurosurgeon; however, the surgical approach was tailored to the individual case, with the amount of resection determined by factors such as the size of the epileptogenic zone, the site and extent of structural lesions, and the proximity to language areas. The hippocampus was spared in 10 patients and was resected in 30 patients. The resection of the hippocampus was complete in 25 patients and incomplete in 4 patients. In one patient, it was impossible to determine whether the hippocampal resection was complete or not. Inclusion criteria were as follows: (1) seen for clinical neuropsychological assessments presurgery and 1 year postsurgery, (2) free of other medical diagnoses/treatments (such as stroke or cranial radiation) that could impact their neuropsychological functions, (3) underwent assessment of at least one literacy skill using the Wechsler Individual Achievement Test ([47,48]: WIAT or WIAT-II) or the Wide Range Achievement Test ([49]: WRAT-3), and (4) had performance IQ (PIQ) above 69. Verbal IQ was not used as an inclusion variable because one of the dependent variables, the semantic memory measure (see below),was taken from the verbal scale of the intelligence test. In 33 of the 40 participants, laterality of speech representation was established using fMRI, magnetoencephalography, or the intracarotid amobarbital test. Thirty children were found to have left-hemisphere speech representation, and three children had bilateral (two participants with left TLE and one participant with right TLE) hemisphere speech representation. All children who did not undergo language laterality testing (n = 7) were righthanded. 2.2. Materials and procedures The study was approved by the Research Ethics Board of the Hospital for Sick Children, Toronto, Canada. Episodic memory was assessed by the delayed recall score on a word list recall task in which learning of words is assessed over five trials and recall is demanded following a 20-minute delay. In our previous study on memory in children with TLE, we found that this measure had high loading on the episodic memory factor and no loading on the semantic memory factor [10]. Two measures of semantic memory were used, both of which have been shown previously to load on a semantic memory factor, but not on an episodic memory factor, in children with TLE [10]. Word knowledge was assessed with the Vocabulary subtest of the Wechsler Intelligence Scale, and naming was evaluated with one of three standardized measures (see Table 1) that required children to name pictures of objects. Reading accuracy, reading comprehension, and spelling accuracy were assessed using standardized tests of academic skills. Table 1 presents the standardized tests that were administered presurgery and 1 year postsurgery in these various domains. Age-scaled scores obtained from the manuals were converted to z-scores. Because the data were collected for clinical purposes over an extended time period, different or updated test versions were administered. There were no significant differences in performance across these versions (p N .05; see

S. Lah, M.L. Smith / Epilepsy & Behavior 44 (2015) 225–233

227

Table 1 Neuropsychological instruments: comparison of scores obtained across tests measuring the same construct. Domain

Memory: semantic Episodic Literacy: reading Spelling

Tests used

Vocabulary Naming Delayed Recall Accuracy Comprehension Accuracy

Tests of significance

WISC-III, WISC-IV, WAIS-III, WPPSI-III WASI EOWPVT, BNT, EVT CAVLT, CVLT WIAT, WIAT-II, WRAT, WJ-III WIAT, WIAT-II, WRAT, WJ-III WIAT, WIAT-II, WRAT, WJ-III

Presurgery

Postsurgery

F(4, 39) = 0.41, p = .80, h2 = .44 F(2, 35) = 1.64, p = .21, h2 = .09 t(36) = −0.64, p = .52, d = .24 F(3, 35) = 0.87, p = .47, h2 = .08 F(3, 29) = 1.02, p = .40, h2 = .11 F(3, 34) = 0.20, p = .90, h2 = .02

F(3, 39) = 1.43, p = .25, h2 = .11 t(30) = 1.31, p = .20, d = .89 t(37) = −0.89, p = .38, d = −.31 F(3, 34) = 0.19, p = .90, h2 = .02 t(16) = 1.52, p = .15, d = .61 F(3, 32) = 1.33, p = .28, h2 = .12

BNT: Boston Naming Test [39]; CAVLT: Children’s Auditory Verbal Learning Test [50]; CVLT: California Verbal Learning Test [51]; EOWPVT: Expressive One-Word Picture Vocabulary Test [52]; EVT: Expressive Vocabulary Test [53]; WAIS-III: Wechsler Adult Intelligence Scale Third Edition [54], WASI: Wechsler Abbreviated Scale of Intelligence [55]; WIAT: Wechsler Individual Achievement Test [47] and WIAT II: Wechsler Individual Achievement Test – Second Edition [48]; WISC-III: Wechsler Intelligence Scale for Children – Third Edition [56] and WISC IV: Wechsler Intelligence Scale for Children - Fourth Edition [57]; WPPSI-III: Wechsler Preschool and Primary Scale of Intelligence – Third Edition [58]; WRAT: Wide Range Achievement Test [49]; WJIII: Woodcock-Johnson- Third Edition [59].

Table 1); thus, scores were collapsed across tests measuring the same construct. 2.3. Statistical analysis Two-tailed tests were undertaken, with α b .05 considered to be significant. T-tests and chi-squared (χ2) tests examined the differences between the group with LTLE and the group with RTLE on demographic and clinical variables. One-sample t-tests compared scores to the normative means presurgery as well as postsurgery. A set of two-way (laterality (RTLE vs. LTLE) × time (presurgery vs. postsurgery)) repeated measures analyses of variance (ANOVAs) or covariance (ANCOVA; variables that differed between the groups with TLE and were related to outcomes used as covariates) examined prechanges to postchanges on tests of memory and literacy. For memory and literacy variables on which significant changes presurgery to postsurgery were found, two additional separate sets of ANCOVA were run: (i) seizure status (seizure-free vs. not seizure-free) × time (presurgery vs. postsurgery) and (ii) hippocampal status (hippocampus resected vs. hippocampus spared), which examined relations between these changes with epilepsy variables. Finally, change scores (differences between z-scores obtained postsurgery relative to presurgery) were calculated on memory and literacy variables on which significant changes were found presurgery to postsurgery. Correlations were employed to examine (i) whether change scores in memory relate to change scores in literacy and (ii) relations between change score in memory and literacy with epilepsy variables (age at seizure onset, age at surgery, proportion of life with epilepsy at surgery and a change in the number of antiepileptic medications from presurgery to postsurgery). 3. Results 3.1. Demographic and clinical characteristics Demographic and clinical information is presented in Table 2. The only background or clinical variable on which the two groups (left TLE onset vs. right TLE onset) differed was PIQ, which was significantly lower in the group with left TLE (LTLE) relative to the group with right TLE (RTLE) both presurgery and postsurgery. There was no significant change in PIQ from presurgery to postsurgery (p = .70, d = −.04). Scores of children with LTLE and those of children with RTLE on tests of memory and literacy (please see Fig. 1) were compared with the normative means using one-sample t-tests. In the domain of semantic memory (Naming and Vocabulary tests), children with LTLE scored significantly below the normative means (presurgery: t(21) = − 3.68, p = .001, d = − 1.61 and t(21) = − 3.62, p = .002, d = − 1.58, respectively; postsurgery: t(17) = − 6.14, p b .001, d = −2.98 and t(21) = −3.36, p = .003, d = −1.47, respectively). Children with RTLE scored significantly below the norms on (i) Naming

presurgery (t(13) = −2.98, p = .01, d = −1.66) but not postsurgery (t(13) = − 1.35, p = .20, d = − 0.75) and on (ii) Vocabulary postsurgery (t(17) = −2.81, p = .01, d = −1.36) but not presurgery (t(17) = −1,75, p = .10, d = −0.85). In the domain of episodic memory (Delayed Recall), the difference between the norms and the scores obtained by the LTLE group was significant only postsurgery (t(21) = −3.64, p = .002, d = −1.59), but not presurgery (t(21) = −1.19, p = .25, d = −0.52). The difference between the norms and scores obtained by the RTLE group was not significant (presurgery: t(15) = − 1.26, p = .23, d = − 0.65; postsurgery: t(16) = −0.69, p = .50, d = −0.34). Comparison of scores obtained by children with epilepsy with the norms on the literacy tests indicated significantly reduced Reading Accuracy, Reading Comprehension, and Spelling scores in children with LTLE only postsurgery (t(19) = − 2.44, p = .02, d = − 1.12; t(16) = − 2.41, p = .03, d = − 1.20; and t(17) = − 2.14, p b .05, d = − 1.04, respectively; presurgery: t(19) = − 1.64, p = .12, d = − 0.75; t(16) = − 1.79, p = .09, d = − 0.90; and t(18) = − 2.01, p = .06, d = − 0.95, respectively). In children with RTLE, scores on the Reading Accuracy and Reading Comprehension tests did not differ from the norms (presurgery: t(15) = − .31, p = .76, d = − 0.16 and t(12) = − .84, p = .42, d = − 0.49, respectively; postsurgery: t(14) = − 1.53, p = .15, d = − 0.82 and t(13) = − 1.54, p = .15, d = −0.85, respectively). Scores on the Spelling tests were significantly below the norms postsurgery (t(14) = −2.23, p = .04, d = −1.19) but not presurgery (t(15) = −.54, p = .60, d = −0.28). 3.2. Presurgery to postsurgery changes in memory and literacy To determine when the PIQ, on which significant differences were found between the group with LTLE and the group with RTLE, needed to be used as a covariate in the two-way (laterality (left TL vs. right TL) × time (presurgery vs. postsurgery)) repeated measures ANOVAs, we ran Pearson's correlations between presurgical PIQ and scores obtained on tests of memory and literacy presurgery and postsurgery. Significant correlations were found between the PIQ and at least one (either presurgery or postsurgery) of the scores on the Vocabulary (presurgery: r = .40, p = .01; postsurgery: r = .40, p = .01), Delayed Recall (presurgery: r = .03, p = .85; postsurgery: r = .34, p = .03), Reading Comprehension (presurgery: r = .44, p = .01; postsurgery: r = .31, p = .08), and Spelling (presurgery: r = .33, p = .05; postsurgery: r = .23, p = .25) tests. For these tests for which the correlation was significant, PIQ was used as a covariate in subsequent between-groups analyses. No significant correlations were found between the PIQ and the score obtained on tests of Naming (presurgery: r = .27, p = .12; postsurgery: r = .33, p = .07) or Reading Accuracy (presurgery: r = .30, p = .07; postsurgery: r = .23, p = .18). In the domain of semantic memory, the laterality (left TL vs. right TL) × time (presurgery vs. postsurgery) repeated measures ANOVAs

228

S. Lah, M.L. Smith / Epilepsy & Behavior 44 (2015) 225–233

Table 2 Demographic and clinical characteristics. Left temporal (n = 22)

Right temporal (n = 18)

Test of significance

Sex (M/F), n Age at seizure onset: mean (SD), years Handedness (L/R), n

7/15 9.00 (4.65) 3/19

10/8 7.15 (3.76) 1/17

χ2(1) = 1.42 t(38) = −1.36 χ2(1) = 0.10

Presurgery Seizure frequency, n 1 or more per day 1 or more per week 1 or more per month b1 per month Clusters Antiepileptic drugs: no, mono, poly, n Antiepileptic drugs: mean (SD) Age at assessment: mean (SD), years Performance IQ: mean (SD)

9 6 4 1 2 0/10/12 1.59 (0.60) 13.64 (2.89) 91.77 (13.95)

9 2 5 1 1 1/9/8 1.44 (0.71) 13.01 (3.74) 101.11 (13.81)

14.79 (2.99) .40 (.27) 5.79 (4.25) 6/16

13.66 (3.77) .45 (.28) 6.50 (4.72) 4/14

6 3 1 3 3 2 2 0 2

9 1 1 2 2 2 0 1 0

14/8 1/10/11 1.55 (0.80) 15.91 (3.00) 1.12 (0.18) 92.09 (13.02)

10/8 1/12/5 1.28 (0.67) 14.69 (3.76) 1.03 (0.16) 101.89 (15.54)

Surgery Age at surgery: mean (SD), years Proportion of life with epilepsy: mean (SD), years Duration of seizure disorder: mean (SD), years Hippocampus spared (Y/N), n Histopathological diagnosis, n Tumor Mesial temporal sclerosis (MTS) Gliosis MTS + cortical gliosis Dual pathology other Neuronal malformation Vascular malformation Tuberous sclerosis No detectable pathology Postsurgery Seizure frequency: seizure-free, not seizure-free, n Antiepileptic drugs: no, mono, poly, n Antiepileptic drugs: mean (SD) Age at postsurgery assessment: mean (SD), years Time since surgery: mean (SD), years Performance IQ:mean (SD)

on the Naming test showed no significant main effects for laterality (F(1, 27) = 3.67, p = .07, η2p = .12) and time (F(1, 27) = 2.46, p = .13, η2p = .08), but the interaction was significant (F(1, 27) = 4.42, p = .05, η2p = .14) (see Fig. 1). According to the post hoc analyses, presurgery, the difference between the group with LTLE and the group with RTLE on the Naming test was not significant (p = .44, d = .28). Postsurgery, the group with LTLE named significantly fewer items correctly relative to the group with RTLE (p = .001, d = 1.29). The within-group pairwise comparisons showed a significant decline from presurgery to postsurgery in the Naming score in the group with LTLE (t(17) = − 2.51, p = .02, d = .44) but no significant change in the group with RTLE (t(10) = −0.61, p = .55, d = −.09). On the Vocabulary test (ANCOVA), none of the effects were significant: laterality (F(1, 37) = 0.02, p = .88, η2p b .01); time (F(1, 37) = 0.12, p = .74, η2p b .01); and interaction (F(1, 37) = 0.06, p = .81, η2p b .01). In the domain of episodic memory (Delayed Recall), ANCOVA findings were nonsignificant (main effect of laterality and main effect of time and interaction: Fs(1, 35) = 0.97, 2.73, and 2.81; ps = .33, .11, and .10; η2p = .03, .07, and .07, respectively). With respect to literacy, on the Reading Accuracy test, the ANOVAs revealed a significant main effect of time (F(1, 30) = 4.20, p = .049, η2p = 1.23) shown by the post hoc testing to be due to significantly lower postsurgery relative to presurgery scores. The effects of laterality (F(1, 30) = .29, p = .59, η2p = .01) and interaction (F(1, 30) = .61, p = .044, η2p = .02) were not significant. Finally, ANCOVAs showed that there were no significant main effects of laterality, time, nor their interaction for tests of (i) Reading Comprehension (Fs(1, 23) = b0.01, b0.01, and 0.08; ps = .97, .99, and .78, respectively, with all η2ps b .01) and (ii)

p-Value

Effect size

Overall sample (n = 40)

.20 .18 .61

φ = .24 d = .44 φ = .13

17/23 8.17 (4.32) 4/36

χ2(4) = 2.07

.72

φ = .23

χ2(2) = 1.47 t(38) = −0.72 t(38) = −0.61 t(38) = 2.12

.48 .48 .55 .04

φ = .19 d = .23 d = .19 d = −.67

18 8 9 2 3 1/19/20 1.53 (0.64) 13.36 (3.27) 95.98 (14.86)

t(38) = −1.06 t(38) = .56 t(38) = .50 χ2(1) = .00 χ2(8) = 6.67

.30 .58 .62 1.00 .57

d = .33 d = −.18 d = −.16 φ = −.06 φ = .41

14.23 (3.36) .42 (.28) 6.11 (4.42) 10/30 15 4 2 5 4 2 1 2

χ2(1) = .04 χ2(2) = 2.05 t(38) = −1.13 t(38) = −1.15 t(38) = −1.76 t(38) = 2.17

.75 .36 .27 .26 .09 .04

φ = −.08 φ = .23 d = .37 d = .36 d = .53 d = −.68

24/16 2/22/16 1.43 (0.75) 15.35 (3.38) 1.08 (0.17) 96.50 (14.86)

Spelling Accuracy (Fs(1, 27) = 0.04, 0.09, and 1.40; ps = .85, .77, and .25, η2p b .01, b. 01, and =.05, respectively). Given that group by time interactions and main effects of group were not significant, scores of the group with LTLE and the group with RTLE were collapsed for each test in all subsequent analyses, except for the Naming test on which the interaction was significant.

3.2.1. Seizure freedom As seizure freedom is the ultimate goal of epilepsy surgery, we examined whether memory and literacy outcomes were related to being seizure-free postsurgery (see Table 3). To determine whether and when to use PIQ as a covariate, we compared presurgery and postsurgery PIQs of patients who were seizure-free with PIQs of patients who were not seizure-free postsurgery. The between-groups differences in PIQ were not significant either presurgery (seizure-free: M = 94.96, SD = 14.99; not seizure-free: M = 97.50, SD = 14.06; t(38) = −0.54, p = .59, d = −.18) or postsurgery (seizure-free: M = 96.42, SD = 14.95; not seizure-free: M = 96.63, SD = 15.23; t(38) = − 0.43, p = .97, d = − .01). Thus, PIQ was not used as a covariate. Analyses of variance (see Table 3) found no significant main effects of seizure status (seizure-free vs. not seizure-free) or interactions in the combined group for any of the tasks. Because of the significant decline in naming in the group with LTLE, but not in the group with RTLE, the effect of seizure outcome on this task was examined with respect to laterality. Neither seizure outcome nor its interaction with laterality was significant.

S. Lah, M.L. Smith / Epilepsy & Behavior 44 (2015) 225–233

Semantic Memory: Naming

Semantic Memory: Vocabulary

Episodic Memory: Word Recall

Literacy: Reading Accuracy

229

Literacy: Reading Comprehension

Literacy: Spelling .

0 -0.5

Mean Z Score

-1 -1.5 -2 -2.5

LTLE Pre-surgery LTLE Post-surgery

-3 -3.5 Semantic Memory: Naming

Semantic Memory: Vocabulary

Episodic Memory: Word Recall

Literacy: Reading Accuracy

Literacy: Reading Comprehension

Literacy: Spelling .

0 -0.5

Mean Z Score

-1 -1.5 -2 -2.5 RTLE Pre-surgery

-3

RTLE Post-surgery

-3.5 Fig. 1. Mean z-scores obtained by children with left (top) or right (bottom) temporal lobe epilepsy presurgery and postsurgery.

3.2.2. Hippocampus spared vs. hippocampus resected Comparison of patients whose hippocampus was resected (n = 30) with patients whose hippocampus was spared (n = 10) showed no difference between PIQs either presurgery (hippocampus spared: M = 99.10, SD = 18.44; hippocampus resected: M = 94.93, SD = 13.13; t(38) = − 0.78, p = .051, d = 0.26) or postsurgery (hippocampus spared: M = 98.10, SD = 17.85; hippocampus resected: M = 95.97, SD = 14.04; t(38) = 0.39, p = .46, d = 0.13). Analyses of variance found no significant changes related to hippocampal resection status from presurgery to postsurgery on any of the memory or literacy tests (see Table 4).

reading accuracy (r = .38, p = .47) nor with epilepsy variables: age at seizure onset (r = .38, p = .47), age at surgery (r = .38, p = .47), proportion of life with epilepsy at surgery (r = −.33, p = .18), or a change in the number of antiepileptic medications (r = .01, p = .98). Similarly, a decline on the Reading Accuracy test was not related to any of the epilepsy variables: age at seizure onset (r = .09, p = .62), age at surgery (r = − .17, p = .36), proportion of life with epilepsy at surgery (r = −.10, p = .60), or change in the number of antiepileptic medications (r = .07, p = .71).

3.2.3. Correlations between outcome variables on which significant prechanges to postchanges were found and epilepsy variables The correlational analyses (r) showed that in the group with LTLE, a decline on the Naming test was associated neither with a change in

The objectives of this study were to (i) investigate presurgical to postsurgical changes in verbal memory (semantic and episodic) and literacy skills (reading accuracy, reading comprehension, and spelling) and to (ii) examine relations between changes in memory and literacy.

4. Discussion

230

S. Lah, M.L. Smith / Epilepsy & Behavior 44 (2015) 225–233

Table 3 Mean z-scores (standard deviations) obtained presurgery and postsurgery: seizure-free and not seizure-free patients postsurgery. Group: seizure freedom postsurgery Seizure-free (n = 24)

Memory Semantic

Episodic

Literacy Reading

Spelling

Vocabulary (n = 40) Naming (LTLE) (n = 18) Naming (RTLE) (n = 11) Delayed Recall (n = 38)

Accuracy (n = 32) Comprehension (n = 26) Accuracy (n = 19)

ANOVA Not seizure-free (n = 16)

Group

Time

Interaction

F(1, 38) = 0.30, p = .59, η2p = .01 F(1, 16) = 0.58, p = .45, η2p = .04 p = .74, η2p = .01

F(1, 38) = 0.48, p = .49, η2p = 01 F(1, 16) =0.50, p = .83, η2p b .00 F(1, 9) = 0.04, p = .85, η2p b .00 F(1, 36) = 0.27, p = .61, η2p = .01 F(1, 30) = 0.01, p = .92, η2p b .00 F(1, 24) = 0.86, p = .36, η2p = .03 F(1, 28) = 0.06, p = .81, η2p b .00

Presurgery

Postsurgery

Presurgery

Postsurgery

−0.56 (0.83)

−0.72 (0.91)

−0.48 (0.99)

−0.50 (0.92)

−1.50 (1.73)

−2.27 (1.90)

−2.07 (2.50)

−3.00 (1.71)

−0.83 (1.14)

−0.76 (1.22)

−1.16 (1.56)

−1.01 (1.36)

−0.29 (1.37)

−0.78 (1.70)

−0.42 (0.97)

−0.63 (1.20)

F(1, 36) b 0.01, p = .98, η2p b .00

F(1, 38) = 0.78, p = .38, η2p = .02 F(1, 16) = 5.99, p = .03*, η2p = .27 F(1, 9) = 0.26, p = .62, η2p = .03 F(1, 36) = 1.73, p = .20, η2p = .05

−0.11 (.86)

−0.27 (.81)

−0.51 (.98)

−0.68 (.84)

−0.31 (1.02)

−0.55 (1.48)

−0.70 (.83)

−1.40 (1.62)

−0.26 (.98)

−0.36 (.89)

−0.62 (84)

−0.78 (.73)

F(1, 30) = 1.71, p = .20, η2p = .05 F(1, 24) = 1.88, p = .18, η2p = .07 F(1, 28) = 1.58, p = .22, η2p = .05

F(1, 30) = 3.44, p = .07, η2p = .10 F(1, 24) = 3.62, p = .07, η2p = .13 F(1, 28) = 1.11, p = .30, η2p = .04

LTLE: left temporal lobe epilepsy; RTLE: right temporal lobe epilepsy. ⁎ p b .05.

Our study revealed significant declines in one aspect of semantic memory (naming, in children who underwent left TL) and reading accuracy but no change in other areas of memory and literacy postsurgery. The decrements in naming and in reading accuracy were related neither to each other nor to epilepsy variables. The postoperative decline in confrontation naming by the group with LTLE has not previously been reported in studies on pediatric epilepsy surgery. To our knowledge, only one other study has examined naming before and after temporal lobe resection in children [7]; in that study, naming ability improved after surgery, a change not associated with the laterality of the excision. One difference between that study and the present study was that all of Jambaqué et al.'s cases were seizure-free after surgery; however, seizure outcome does not seem to explain the differences in findings, as task performance in our study was not related to postoperative seizure status. The current findings are, however, highly similar to the results of research with adults, in which naming declines after left TL are common (see review [40]). Preoperative electrical cortical stimulation has identified naming sites throughout the left lateral and the basal temporal lobe in both adults and children [60,61].

With respect to episodic memory, tested by delayed recall of a word list learned over multiple trials, the findings are consistent with other studies examining outcomes of TL in children. For similar word listlearning tasks, the majority of research has found no group change in learning, delayed recall, or recognition after either left TL or right TL [7,62–69]. One study [70] did report a significant decline, but the sample size in that study was quite small (n = 9). Gleissner et al. [71] demonstrated an initial decline in verbal recall, evident three months after surgery, but with improvement to the preoperative baseline by one year after surgery, the typical time of follow-up in the majority of studies, and that used in the present study. The absence of laterality effects in the preoperative performance on the semantic and episodic memory tasks is also consistent with other pediatric studies examining surgical outcomes [62–68]. Research on verbal memory performance in nonoperated samples with TLE has also found limited evidence of laterality effects [8,11,30,63,67,72–76], with the exception of one finding of poorer confrontation naming in children with left mesial temporal sclerosis [11], a result that should be replicated due to the small number of participants (n = 5) in that group.

Table 4 Mean z-scores (standard deviations), hippocampus spared, and hippocampus excised.

Memory Semantic

Episodic

Literacy Reading

Spelling

Vocabulary (n = 40) Naming (LTLE) (n = 18) Naming (RTLE) (n = 11) Delayed Recall (n = 38)

Accuracy (n = 32) Comprehension (n = 26) Accuracy (n = 40)

Group: hippocampus status postsurgery

ANOVA

Hippocampus spared (n = 10)

Hippocampus excised (n = 30)

Group

Time

Interaction

Presurgery

Postsurgery

Presurgery

Postsurgery

−0.80 (1.04)

−0.63 (.97)

−0.43 (.83)

−0.63 (.90)

−1.89 (1.28)

−2.12 (1.55)

−1.68 (2.41)

−2.83 (1.92)

−1.68 (.62)

−1.54 (.75)

−0.80 (1.52)

−0.69 (1.37)

0.31 (1.05)

−0.25 (1.49)

−0.57 (1.22)

−0.89 (1.52)

F(1, 38) = 0.36, p = .56, η2p = .01 F(1, 16) = 0.07, p = .79, η2p b .00 F(1, 9) = 1.01, p = .34, η2p = .10 F(1, 36) = 3.49, p = .07, η2p = .09

F(1, 38) = 0.02, p = .89, η2p b .00 F(1, 16) = 3.90, p = .07, η2p = .20 F(1, 9) = 0.31, p = .59, η2p = .03 F(1, 36) = 2.31, p = .14, η2p = .06

F(1, 38) = 2.50, p = .12, η2p = .06 F(1, 16) = 1.76, p = .20, η2p = .10 F(1, 9) = 0.01, p = .94, η2p b .00 F(1, 36) = 0.17, p = .68, η2p = .01

−0.28 (1.20)

−0.33 (1.13)

−0.24 (.80)

−0.44 (.71)

−0.33 (1.14)

−1.08 (2.19)

−0.52 (.89)

−0.79 (1.25)

−0.22 (1.24)

−0.27 (1.14)

−0.45 (.81)

−0.61 (.72)

F(1, 30) = 0.01, p = .92, η2p b .00 F(1, 24) = 0.01, p = .92, η2p b .00 F(1, 28) = 0.71, p = .41, η2p = .03

F(1, 30) = 1.87, p = .18, η2p = .06 F(1, 24) = 3.79, p = .06, η2p = .14 F(1, 28) = 0.54, p = .47, η2p = .02

F(1, 30) = 0.76, p = .39, η2p = .03 F(1, 24) = 0.81, p = .38, η2p = .03 F(1, 28) = 0.14, p = .72, η2p = .01

LTLE: left temporal lobe epilepsy; RTLE: right temporal lobe epilepsy.

S. Lah, M.L. Smith / Epilepsy & Behavior 44 (2015) 225–233

The task used to assess episodic memory in the current study, delayed recall of a word list learned over multiple trials, has been commonly employed for assessment of episodic memory in clinical practice. Nevertheless, this task is somewhat at odds with the more contemporary conceptualization of episodic memory and hippocampal function. Episodic memory is currently conceptualized as the ability to recollect unique, personally experienced events of a known temporality that are rich in contextual details [77,78]. The retrieval and reexperiencing of these episodic memories are proposed to be hippocampusdependent indefinitely [79]. To our knowledge, only one study examined episodic memory in children with unilateral TLE using such a task [9]. The study involved 21 children with TLE, of which 7 had undergone surgery, and 24 healthy control children. Compared with controls, children with TLE recalled significantly fewer episodic details. Children with TLE who underwent surgery did not differ from children with TLE who did not undergo surgery in recall of episodic details. This study, however, did not follow up children presurgery to postsurgery. With respect to episodic memory tasks, Saling [80] has argued that tasks which involve formation of novel arbitrary associations and have little semantic loading,” … can be regarded as proximal to the neurobiological mechanisms of mesial TLE, and therefore …” more sensitive to the mesial temporal lobe dysfunction than tasks that do not involve formation of novel arbitrary association (like a task used in our study). Nevertheless, a study that employed learning of novel arbitrary associations in children with TLE to examine episodic memory found no significant change from presurgery to postsurgery, just like our study that did not use such a task [73]. The postoperative decline in reading accuracy found in our study is at odds with the finding of no significant change in the only previous study that examined changes in reading following TL in children [43]. Nevertheless, inspection of the scores reveals the same trend: mean scores on the reading tests were lower post-TL relative to pre-TL. This earlier study, however, involved a much smaller number of children who underwent TL (n = 10) compared with our study (n = 32), with the decline in reading scores failing to reach statistical significance possibly because of the limited statistical power, an issue recognized by the authors themselves. In contrast to reading accuracy, reading comprehension did not change significantly from presurgery to postsurgery. Contrary to our expectations, the decline in verbal semantic memory was not related to the decline in reading accuracy. This finding could be due to several reasons. First, the decrement in semantic memory was limited to naming; it did not involve the ability to describe the meaning of different words. Such a pattern of performance suggests compromised access to the semantic system. It is possible that the printed words on the reading tests facilitated access to the semantic system and word retrieval; hence, the drop in confrontational naming did not impact reading accuracy. Second, while semantic memory is an integral part of the two prominent cognitive models of reading, one of these models, the DRC model [26,27], allows for accurate reading to occur even when semantic memory is compromised and can easily account for the findings of our study. While a decline in reading would be expected in association with a decrease in semantic skills according to the PDP model [28,29], such a decline in reading would be selective; it would involve irregular (especially infrequent), but not regular, words resulting in surface dyslexia [81]. Materials used for assessment of reading accuracy in our study, while normed and standardized, did not provide different scores for regular and irregular words. Thus, it is possible that the majority of the words included in these materials were regular, in which case the reading accuracy would not be compromised by changes in the semantic system according to either of the models. Third, it is possible that reading became more effortful, slower, and less fluent postsurgery and that such declines in reading speed and fluency, which have not been assessed in the current study, were associated with a decline in semantic memory. Postsurgical seizure outcome did not impact on performance of any of the measures. Furthermore, changes in either naming or reading

231

accuracy scores in the group with LTLE were not related to age at surgery, proportion of life with epilepsy, number of antiepileptic medications, nor changes in episodic verbal memory or in vocabulary, the other measure of semantic memory. These findings are consistent with previous investigations of surgical outcome in children [7,62,65, 82]. There is evidence that cognitive deficits are apparent at or before the time of seizure onset, suggesting that the underlying neural abnormality may be more influential in the genesis of the cognitive abnormalities than the seizures themselves [83]. Several limitations of our study should be noted. This retrospective study spans 15 years of clinically collected data using tests that do not necessarily examine memory and literacy according to current theories. Moreover, different forms of tests were used to assess the same construct, as test versions were updated during the time frame of the study. The comparison of scores obtained on these different tests, however, failed to reveal statistically significant differences. While our retrospective clinical approach has enabled us to include a large number of children, the sample itself remains relatively heterogeneous, although all children had unilateral temporal lobe epilepsy. This heterogeneity is a common feature of pediatric surgical samples. In addition, we used the normative data, rather than a control group. To address these limitations, large, prospective, multicenter studies that include control groups and theoretically developed materials are needed. Despite these limitations, this study is novel, clinically significant, and theoretically relevant. It examines changes in different aspects of memory and literacy from pre-TL to post-TL in children. It is the first study to examine relations between changes in memory and literacy skills from pre-TL to post-TL in children. The study reveals that, as a group, children who undergo TL are at risk of a slight decline in semantic memory (naming, following left TL) and reading (accuracy, unrelated to the side of TL), which are independent of one another. This lack of relationship between declines in semantic memory and reading is primarily consistent with the DRC model of reading; nevertheless, this finding should be viewed as preliminary. More refined methodology needs to be used to examine predictions arising from the two (DRC and PDP) models adequately. The brain basis and epilepsy factors contributing to these changes postchildhood TL remain to be established. Acknowledgment This work was supported by an Academy of the Social Sciences in Australia (ASSA) International Science Linkages Scheme grant (SL and MLS), an Australian Research Council Centre of Excellence in Cognition and its Disorders Cross Program grant, and the University of Sydney Thompson Fellowship (both SL). We thank Ms. Carly Black for her assistance with the preparation of manuscript and Mr. Klajdi Puka and Ms. Monique Tremblay for their assistance with the data collection. Conflict of interest There is no conflict of interest to declare. References [1] Smith ML, Direnfeld E. Memory in children with epilepsy. In: Zeman A, Kapur N, Jones-Gotman M, editors. Epilepsy and memory. Oxford: Oxford University Press; 2012. p. 102–13. [2] Camfield PR, Gates R, Ronen G, Camfield C, Ferguson A, MacDonald GW. Comparison of cognitive ability, personality profile, and school success in epileptic children with pure right versus left temporal lobe EEG foci. Ann Neurol 1984;15(2):122–6. [3] Chaix Y, Laguitton V, Lauwers-Cances V, Daquin G, Cances C, Demonet J-F, et al. Reading abilities and cognitive functions of children with epilepsy: influence of epileptic syndrome. Brain Dev 2006;28(2):122–30. [4] Temple C. Surface dyslexia in a child with epilepsy. Neuropsychologia 1984;22(5): 569–76. [5] Vanasse CM, Beland R, Carmant L, Lassonde M. Impact of childhood epilepsy on reading and phonological processing abilities. Epilepsy Behav 2005;7(2):288–96. [6] Vanasse CM, Beland R, Jambaque I, Lavoie K, Lassonde M. Impact of temporal lobe epilepsy on phonological processing and reading: a case study of identical twins. Neurocase 2003;9(6):515–22.

232

S. Lah, M.L. Smith / Epilepsy & Behavior 44 (2015) 225–233

[7] Jambaque I, Dellatolas G, Fohlen M, Bulteau C, Waiter L, Dorfmuller G, et al. Memory functions following surgery for temporal lobe epilepsy in children. Neuropsychologia 2007;45:2850–62. [8] Nolan MA, Redoblado A, Lah S, Sabaz M, Lawson J, Cunningham AM, et al. Memory function in childhood epilepsy syndromes. J Paediatr Child Health 2004;40:20–7. [9] Gascoigne MB, Smith ML, Barton B, Webster R, Gill D, Lah S. Autobiographical memory in children with temporal lobe epilepsy. J Int Neuropsychol Soc 2014;19(10):1076–86. [10] Smith ML, Lah S. One declarative memory system or two? The relationship between episodic and semantic memory in children with temporal lobe epilepsy. Neuropsychology 2011;25(2):634–44. [11] Rzezak P, Guimaraes C, Fuentes D, Guerreiro MM, Valente KD. Episodic and semantic memory in children with mesial temporal sclerosis. Epilepsy Behav 2011;21(3): 242–7. [12] Lah S, Smith ML. Semantic and episodic memory in children with temporal lobe epilepsy: do they relate to literacy skills? Neuropsychology 2014;28:113–22. [13] Blazely AM, Coltheart M, Casey BJ. Semantic impairment with and without surface dyslexia: implications for models of reading. Cogn Neuropsychol 2005;22(6): 695–717. [14] Gerhand S. Routes to reading: a report of a non-semantic reader with equivalent performance on regular and exception words. Neuropsychologia 2001;39(13):1473–84. [15] Cipolotti L, Warrington EK. Semantic memory and reading abilities: a case report. J Int Neuropsychol Soc 1995;1:104–10. [16] Brambati SM, Ogar J, Neuhaus J, Miller BL, Gorno-Tempini ML. Reading disorders in primary progressive aphasia: a behavioral and neuroimaging study. Neuropsychologia 2009;47(8–9):1893–900. [17] Bub D, Cancelliere A, Kertesz A. Whole-word and analytic translation of spelling to sound in a non-semantic reader. In: Patterson K, Marshall J, Coltheart M, editors. Surface dyslexia. London: Lawrence Erlbaum; 1985. p. 15–34. [18] Glosser G, Grugan P, Friedman RB. Semantic memory impairment does not impact on phonological and orthographic processing in a case of developmental hyperlexia. Brain Lang 1997;56(2):234–47. [19] Jefferies E, Lambon Ralph MA, Jones R, Bateman D, Patterson K. Surface dyslexia in semantic dementia: a comparison of the influence of consistency and regularity. Neurocase 2004;10(4):290–9. [20] McCarthy RA, Warrington EK. Phonological reading: phenomena and paradoxes. Cortex 1986;22(3):359–80. [21] Patterson K, Hodges JR. Deterioration of word meaning: implications for reading. Neuropsychologia 1992;30(12):1025–40. [22] Warrington EK. The selective impairment of semantic memory. Q J Exp Psychol 1975;27(4):635–57. [23] Woollams AM, Ralph MAL, Plaut DC, Patterson K. SD-squared: on the association between semantic dementia and surface dyslexia. Psychol Rev 2007;114(2):316–39. [24] Morton J, Patterson K. A new attempt at an interpretation or an attempt at a new interpretation. In: Coltheart M, Patterson K, Marshall J, editors. Deep dyslexia. London: Routledge and Kegan Paul; 1980. p. 91–118. [25] Coltheart M. Acquired dyslexias and normal reading. In: Malatesha R, Whitaker H, editors. Dyslexia: a global issue Lancaster. UK: Martinus Nijhoff Publishers; 1984. p. 357–73. [26] Coltheart M, Rastle K, Perry C, Langdon R, Ziegler J. DRC: a dual route cascaded model of visual word recognition and reading aloud. Psychol Rev 2001;108(1): 204–56. [27] Coltheart M, Curtis B, Atkins P, Haller M. Models of reading aloud: dual-route and parallel-distributed-processing approaches. Psychol Rev 1993;1000(4):589–608. [28] Plaut DC, McClelland JL, Seidenberg MS, Patterson K. Understanding normal and impaired word reading: computational principles in quasi-regular domains. Psychol Rev 1996;103(1):56–115. [29] Seidenberg MS, McClelland JL. A distributed, developmental model of word recognition and naming. Psychol Rev 1989;96(4):523–68. [30] Gonzalez LM, Anderson VA, Wood SJ, Mitchell LA, Harvey AS. The localization and lateralization of memory deficits in children with temporal lobe epilepsy. Epilepsia 2007;48:124–32. [31] Smith ML, Milner B. Right hippocampal impairment in the recall of spatial location: encoding deficit or rapid forgetting. Neuropsychologia 1989;27:71–81. [32] Helmstaedter C, Grunwald T, Lehnertz K, Gleißner U, Elger CE. Differential involvement of left temporolateral and temporomesial structures in verbal declarative learning and memory: evidence from temporal lobe epilepsy. Brain Cogn 1997; 35(1):110–31. [33] Brambati SM, Ogar J, Neuhaus J, Miller BL, Gorno-Tempini ML. Reading disorders in primary progressive aphasia: a behavioral and neuroimaging study. Neuropsychologia 2009;47:1893–900. [34] Brambati SM, Rankin KP, Narvid J, Seeley WW, Dean D, Rosen HJ, et al. Atrophy progression in semantic dementia with asymmetric temporal involvement: a tensor based morphometry study. Neurobiol Aging 2009;30:103–11. [35] Duchowny M, Levin B, Jayakar P, Resnick T, Alvarez L, Morrison G, et al. Temporal lobectomy in early childhood. Epilepsia 1992;33(2):298–303. [36] Patterson K, Nestor PJ, Rogers TT. Where do you know what you know? The representation of semantic knowledge in the human brain. Nat Rev Neurosci 2007;8(12): 976–87. [37] Lee TMC, Yip JTH, Jones-Gotman M. Memory deficits after resection from left or right temporal lobe in humans: a meta-analytic review. Epilepsia 2002;43:283–91. [38] Lah S. Neuropsychological outcome following focal cortical removal for intractable epilepsy in children. Epilepsy Behav 2004;5:804–17. [39] Kaplan E, Goodglass H, Weintraub S. The Boston Naming Test. Boston: E. Kaplan & H. Goodglass; 1976. [40] Ives-Deliperi VL, Butler JT. Naming outcomes of anterior temporal lobectomy in epilepsy patients: a systematic review of the literature. Epilepsy Behav 2012;24:194–8.

[41] Davies KG, Bell BD, Bush AJ, Hermann BP, Dohan FC, Jaap AS. Naming decline after left anterior temporal lobectomy correlates with pathological status of resected hippocampus. Epilepsia 1998;39:407–19. [42] Hermann B, Davies KG, Fioley K, Bell B. Visual confrontation naming outcome after standard anterior temporal lobectomy with sparing versus resection of the superior temporal gyrus: a randomized prospective clinical trial. Epilepsia 1999;40:1070–6. [43] Blanchette N, Smith ML. Language after temporal or frontal lobe surgery in children with epilepsy. Brain Cogn 2002;48:280–4. [44] Nation K, Snowling MJ. Semantic processing and the development of wordrecognition skills: evidence from children with reading comprehension difficulties. J Mem Lang 1998;39(1):85–101. [45] Dixon P, LeFevre J-A, Twilley LC. Word knowledge and working memory as predictors of reading skill. J Educ Psychol 1988;80(4):465–72. [46] Smith ML, Elliott IM, Lach L. Cognitive, psychosocial, and family function one year after pediatric epilepsy surgery. Epilepsia 2004;45:650–60. [47] Wechsler D. Wechsler Individual Achievement Test. San Antonio, TX: The Psychological Corporation; 1992. [48] Wechsler D. Wechsler individual achievement test. second ed. San Antonio, TX: The Psychological Corporation; 2002. [49] Jastak JF, Jastak SR. The Wide Range Achievement Test. Wilmington DE: Guidance Associates of Delaware; 1976. [50] Talley JL. Children's Auditory Verbal Learning Test. Psychological Assessment Resources: Odessa, FL; 1990. [51] Delis DC, Kramer JH, Kaplan E, Ober BA. The California Verbal Learning Test. San Antonio, TX: Psychological Corporation; 1987. [52] Gardner MF. Expressive One-Word Picture Vocabulary Test. Novato, California: Academic Therapy Publications; 1997. [53] Williams KT. Expressive Vocabulary Test. Circle Pines, Minn.: American Guidance Service; 1997 [54] Wechsler D. Wechsler Adult Intelligence Scale - Third Edition. San Antonio, TX: The Psychological Corporation; 1997. [55] Wechsler D. Wechsler Abbreviated Scale of Intelligence. San Antonio, TX: The Psychological Corporation; 1999. [56] Wechsler D. Wechsler Intelligence Scale for Children - Third Edition. San Antonio, TX: The Psychological Corporation; 1991. [57] Wechsler D. Wechsler Intelligence Scale for Children - Fourth Edition. NCS Pearson: San Antonio, TX; 2003. [58] Wechsler D. Wechsler Preschool and Primary Scale of Intelligence - Third Edition. San Antonio, TX: The Psychological Corporation; 2002. [59] Woodcock RW, McGrew KS, Mather N. Woodcock-Johnson Test - Third Edition. Itasca, IL: Riverside Publishing Company; 2001. [60] Hamberger MJ, Seidel WT, McKahn GM, Goodman RR. Hippocampal removal affects visual but not auditory naming. Neurology 2010;74:1488–93. [61] Kadis DS, Iida K, Kerr EN, Logan WJ, McAndrews MP, Ochi O, et al. Intrahemispheric reorganization of language in children with medically intractable epilepsy of the left hemisphere. J Int Neuropsychol Soc 2007;13:505–16. [62] Smith ML, Elliott IM, Lach L. Memory outcomes after pediatric epilepsy surgery: objective and subjective perspectives. Child Neuropsychol 2006;12:151–64. [63] Kuehn SM, Keene DL, Richards PMP, Ventureyra ECG. Are there changes in intelligence and memory functioning following surgery for the treatment of refractory epilepsy in childhood? Childs Nerv Syst 2002;18:306–10. [64] Lee GP, Westerveld M, Blackburn LB, Park YD, Loring DW. Prediction of verbal memory decline after epilepsy surgery in children: effectiveness of Wada memory asymmetries. Epilepsia 2005;46:97–103. [65] Lendt M, Helmstaedter C, Elger CE. Pre and postoperative neuropsychological profiles in children and adolescents with temporal lobe epilepsy. Epilepsia 1999;40:1543–50. [66] Lewis DV, Thompson Jr RJ, Santos CC, Oakes WJ, Radtke RA, Friedman AH, et al. Outcome of temporal lobectomy in adolescents. J Epilepsy 1996;9:198–205. [67] Mabbott DM, Smith ML. Material-specific memory in children with temporal and extratemporal lobectomies. Neuropsychologia 2003;41:995–1007. [68] Meyer FB, Marsh WR, Laws Jr ER, Sharbrough FW. Temporal lobectomy in children with epilepsy. J Neurosurg 1986;64:371–6. [69] Szabó CA, Wyllie E, Stanford LD, Geckler C, Kotagal P, Comair YG, et al. Neuropsychological effect of temporal lobe resection in preadolescent children with epilepsy. Epilepsia 1998;39:814–9. [70] Williams J, Griebel ML, Sharp GB, Boop FA. Cognition and behavior after temporal lobectomy in pediatric patients with intractable epilepsy. Pediatr Neurol 1998;19: 189–94. [71] Gleissner U, Sassen R, Schramm J, Elger C, Helmstaedter C. Greater functional recovery after temporal lobe epilepsy surgery in children. Brain Cogn 2005;128:2822–9. [72] Cormack F, Vargha-Khadem F, Wood SJ, Cross JH, Baldeweg T. Memory in paediatric temporal lobe epilepsy: effects of lesion type and side. Epilepsy Res 2012;98:255–9. [73] Gonzalez LM, Mahdavi N, Anderson VA, Harvey AS. Changes in memory function in children and young adults with temporal lobe epilepsy: a follow-up study. Epilepsy Behav 2012;23:213–9. [74] Hershey T, Craft S, Glauser TA, Hale S. Short-term and long-term memory in early temporal lobe dysfunction. Neuropsychology 1998;12:52–64. [75] Schouten DM, Hendriksen JGM, Aldenkamp AP. Performance of children with epilepsy on the Rey–Osterrieth complex figure test: is there an effect of localization or lateralization? Epilepsy Res 2009;83:184–9. [76] Guimarães CA, Li LM, Rzezak P, Fuentes D, Franzon RC, Augusta Montenegro M, et al. Temporal lobe epilepsy in childhood: comprehensive neuropsychological assessment. J Child Neurol 2007;22:836–40. [77] Levine B, Svoboda E, Hay J, Winocur G, Moscovitch M. Ageing and autobiographical memory: dissociating episodic from semantic retrieval. Psychol Aging 2002;17: 677–89.

S. Lah, M.L. Smith / Epilepsy & Behavior 44 (2015) 225–233 [78] Tulving E. Episodic memory: from mind to brain. Annu Rev Psychol 2002;53:1–25. [79] Nadel L, Moscovitch M. Memory consolidation, retrograde amnesia and the hippocampal complex. Curr Opin Neurobiol 1997;7(2):217–27. [80] Saling MM. Verbal memory in mesial temporal lobe epilepsy: beyond material specificity. Brain Cogn 2009;132(3):570–82. [81] Castles A, Coltheart M. Varieties of developmental dyslexia. Cognition 1993;47(2): 149–80.

233

[82] Oitment C, Vriezen E, Smith ML. Everyday memory in children after resective epilepsy surgery. Epilepsy Behav 2013;28:141–6. [83] Jackson DC, Dabbs K, Walker NM, Jones JE, Hsu DA, Stafstrom CE, et al. The neuropsychological and academic substrate of new/recent-onset epilepsies. J Pediatr 2013; 162:1047–53.