Atypical Language Representation in Patients with Chronic Seizure Disorder and Achievement Deficits with Magnetoencephalography

Transcription

1 Epilepsia, 46(4): , 2005 Blackwell Publishing, Inc. C 2005 International League Against Epilepsy Atypical Language Representation in Patients with Chronic Seizure Disorder and Achievement Deficits with Magnetoencephalography Joshua I. Breier, Edwardo M. Castillo, Panagiotis G. Simos, Rebecca L. Billingsley-Marshall, Ekaterina Pataraia, Shirin Sarkari, James W. Wheless, and Andrew C. Papanicolaou Vivian L. Smith Center for Neurologic Research, Department of Neurosurgery, Division of Clinical Neurosciences, Department of Neurology, and Department of Pediatrics, The University of Texas Health Science Center at Houston, Texas, U.S.A. Summary: Purpose: To characterize the relation between hemispheric asymmetries in language-specific brain activity and reading/spelling achievement by using magnetoencephalography (MEG). Methods: Patients (n = 83) with medically intractable complex partial seizures of either left- or right-hemisphere origin were classified as having reading and/or spelling deficits (RS) or as not impaired (NI) by using standard achievement tests. All patients had undergone noninvasive functional mapping of receptive language cortex by using MEG as part of a preoperative seizure surgery evaluation. Results: RS patients with left-hemisphere seizure onset exhibited relatively greater activation and earlier onset of late, language-specific MEG activity in posterior temporal and inferior parietal areas of the right as compared with the left hemisphere than did NI patients. These findings also were evident on an individual basis and were independent of global intellectual abilities. Conclusions: Reading and spelling achievement deficits in patients with complex partial seizures of left-hemisphere origin are associated with atypical language organization, possibly secondary to reorganization of language function to righthemisphere areas that are not as efficient as homotopic areas in the left hemisphere in supporting reading and spelling functions. Key Words: Magnetoencephalography Epilepsy Learning disabilities Functional imaging. Individuals with epilepsy are at increased risk for academic achievement deficit (1 4). The etiology of achievement deficits in this group is likely multifactorial. Behavioral problems and parental attitudes (5), gender, adaptive functioning problems, and negative personal attitude (1), parental education (6), and anticonvulsant medications (AEDs) (7 9) have all been found to be related to achievement deficits in epilepsy patients. A role for the seizures themselves and/or their substrate in achievement deficits also is suggested by correlations between academic performance and variables such as frequency, severity (1,10), and types (11) of seizures, as well as side of (12) and age at (5) seizure onset. However, the presence and role of a specific neurophysiologic abnormality in individuals with epilepsy and achievement deficits are not yet clear. All patients in the current study had previously undergone noninvasive functional imaging by using magnetoencephalog- Accepted December 7, Address correspondence and reprint requests to Dr. J.I. Breier at University of Texas Health Science Center/Houston, Department of Neurosurgery/Division of Clinical Neurosciences, 1333 Moursund Street, Ste. H114, Houston, TX 77030, U.S.A. Joshua.I.Breier@uth.tmc.edu raphy (MEG) as part of a clinical evaluation protocol for potential seizure surgery. This functional brain-imaging method has been shown to provide valid, topographically specific, and accurate maps of receptive language cortex (13 18). All patients exhibited activation of posterior temporal and inferior parietal areas in the left hemisphere known to be involved in auditory comprehension as well as reading and spelling (19 22). Activation of homotopic areas of the right hemisphere also was present in most cases to varying degrees. The goal of the present study was todetermine whether a relation existed between patterns of activation of these areas and reading and spelling skills. Although language deficits have been associated with left-hemisphere seizure onset (11,12), few functional imaging studies focused on the relation between profiles of brain activity and academic skills in individuals with epilepsy. Previous studies evaluated the relation between atypical language organization as determined by the Wada test (23) and various cognitive domains. The Wada test involves unilateral injection of a brief-acting anesthetic such as sodium amobarbital for the purpose of 540

2 LANGUAGE ORGANIZATION AND ACHIEVEMENT 541 examination of the language and memory competence of the isolated hemisphere. Results from such studies have often emphasized the negative effects of atypical language organization, as indicated by greater right-hemisphere language competence, on nonverbal functions (24 27). Loring et al. (25) found, in a relatively large group of >500 patients with left-hemisphere seizure onset, that individuals with predominantly right-hemisphere language representation exhibited a reduction in visuospatial, but not language, skills compared with those with left-hemisphere language dominance. A similar pattern of cognitive dysfunction was observed in patients with bilateral language representation who had also experienced a presumed shift in handedness. This phenomenon, which has been termed crowding, is hypothesized to occur when reorganization of language function occurs in the right hemisphere in areas that might have been originally destined to subserve nonverbal functions (24 28). Strauss et al. (26) also found significant effects of atypical language organization on nonverbal skills with limited effects on linguistic functions in patients with left hemisphere onset seizures. Billingsley and Smith (29), however, found that when children and adolescents with early-onset temporal-lobe epilepsy of left-hemisphere origin were matched closely for age and gender, atypical speech representation was associated with lowered performance on both verbal and nonverbal intellectual abilities. Few studies have assessed the relation between atypical language organization and academic skills in groups of patients with intractable epilepsy. In the current study, we identified patients as having achievement deficits based on patterns of performance on standard tests of reading and spelling. As in previous studies by our group (30,31) and others (32,33), we used a cutoff rather than a discrepancy definition of academic achievement deficiency. A discrepancy approach has the disadvantage of classifying individuals with lower IQ scores as being nondisabled (34,35), although they have a significant functional deficit. This is a particular problem in groups with chronic epilepsy in which the distribution of IQ scores may be skewed toward lower values (36). Both adults and children were included in the study. Whereas this resulted in a wide age range, problems with basic academic skills have been shown to persist into adulthood, with little change in the major phenotypic characteristics over time (37 39). We hypothesized that increased abnormality in the profile of activation observed during the MEG imaging studies, characterized by greater relative activation of right as compared with left posterior superior temporal and inferior parietal areas (including Wernicke s area) known to be involved in language function, would be associated with deficits in reading and spelling achievement within the group with left-, but not right-, hemisphere seizure onset. METHODS Subjects This was a retrospective study of 83 patients aged 9 to 54 years (mean, 27.2; SD, 12.9), including 43 female and 40 male subjects, who were evaluated at the Texas Comprehensive Epilepsy Program at the University of Texas Health Science Center at Houston. This represented a consecutive series of patients who had undergone the MEG language-mapping protocol between 1997 and Evaluation included, for each patient, continuous 24-h video telemetry-eeg monitoring, magnetic resonance imaging, neuropsychological testing, and the Wada procedure. The relations between MEG and Wada findings are reported elsewhere (16). Previous studies from our center and others reported a close correspondence between estimates of hemispheric dominance for language derived from MEG testing and those determined by using the Wada procedure (16). In addition, single-photon emission computed tomography, positron emission tomography, MEG, and intracranial electrodes were used to aid identification of the epileptogenic zone in many cases. With these methods, 60 patients were identified as having complex partial seizures emanating from the left hemisphere, and 23, from the right. Seven participants were lefthanded as determined by using the Edinburgh handedness inventory (40). The preponderance of patients with lefthemisphere seizure onset is in agreement with the findings of other larger scale studies (e.g ). All patients had English as their primary language by history. None had a history of prior neurosurgical resection. All participants received a standardized test of intellectual functioning. Adults were administered either the Wechsler Adult Intelligence Scale Revised (44) or the Wechsler Abbreviated Scale of Intelligence (45). Children ages 16 and younger were administered either the Wechsler Abbreviated Scale of Intelligence or the Wechsler Intelligence Scale for Children III (46). Patients with Full-Scale IQ scores <70 were not included the study. All participants also were given standardized tests of reading and spelling. Adults were administered the Wide Range Achievement Test 3 (47). Children were given the Basic Reading subtests of the Woodcock Johnson III (48) and the spelling subtest of the Wide Range Achievement Test 3. Standard scores on the subtests were averaged to form a composite for each patient, and patients were placed into the reading/spelling-disabled (RS) group on the basis of having a composite score 90, with at least one of the subtests being below this cutoff (13). These criteria are similar to those used in other studies (32,33,50 52). With these methods, 31 patients were placed in the RS group. The remainder were placed into a not-impaired (NI) group. Group means on demographic and seizure variables are presented as a function of side of seizure onset in Table 1. IQ and achievement test scores are presented in

3 542 J. I. BREIER ET AL. TABLE 1. Demographic data for achievement deficit groups by side of seizure onset Left Hemisphere Onset Right Hemisphere Onset NI RS NI RS Number of Patients Age (years) M SD Females (%) 46% 52% 69% 50% Left handed (%) 7% 5% 23% 0% Age at Onset (years) M SD Duration of Seizure Disorder (years) M SD Etiology Symptomatic (N) Cryptogenic (N) Highest School Grade (years of school) M SD NI, no academic deficiency; RS, Reading and/or Spelling deficient; p <.01. Table 2. Group differences on continuous measures were evaluated by using analysis of variance (ANOVA) with group membership (NI, RS) as the independent variable. Group differences on categorical variables were evaluated by using Fisher s exact test. Different numbers of asterisks ( ) indicate significant group differences. Stimuli and tasks Patients were given a recognition memory task for spoken words and event-related fields (ERFs) were recorded to each word stimulus. The word list consisted of 90 abstract English nouns with scores of 3.0 on the Paivio Concreteness scale (53). Word frequency ranged from very frequent (AA) to nine occurrences per million (54). A native speaker of English with a flat intonation produced TABLE 2. Intellectual and achievement test data for achievement deficit groups Full Scale IQ (SS) M SD Verbal IQ (SS) M SD Performance IQ (SS) M SD Reading (SS) M SD Spelling (SS) M SD SS, Standard Score; NI, no academic defiency; RS, Reading and/or Spelling deficient; p <.01; p < NI RS auditory stimuli (duration between 300 and 750 ms; mean, 450 ms). Stimuli were digitized with a sampling rate of 22,000 Hz and 16-bit resolution and delivered binaurally via two 5 m-long plastic tubes terminating in ear inserts. Intensity was 80 db SPL at the patient s outer ear. Thirty words from each list were used as targets, and the remaining 60, as distractors, forming six blocks of trials. No significant differences were found between the target and distractor lists in either word concreteness or frequency. The target stimuli were repeated in every block (in a different random order each time) mixed, with 10 new distractors. The target stimuli were presented for study immediately before the MEG scan. Stimulus presentation parameters were identical during the actual recording and study sessions. Stimuli were presented with a variable interstimulus interval ( s). Patients were asked to lift their index fingers whenever they detected a repeated word. The responding hand was varied randomly across patients. During the entire testing session, the patient was asked to keep his or her eyes open, fixating on a dark dot placed on the ceiling at eye level, to reduce eye movements or blinks and prevent ERF contamination by rhythmic activity (typically in the alpha band), which can seriously interfere with the accurate detection of task-related brain activity. MEG data acquisition and analysis All patients were tested with a whole-head neuromagnetometer (Magnes 2500; 4D Neuroimaging, San Diego, CA, U.S.A.) equipped with 148 magnetometer sensors and housed in a magnetically shielded room designed to reduce environmental magnetic noise that might interfere with biologic signals. The typical recording session required the patient to lie motionless on a bed with his or her

4 LANGUAGE ORGANIZATION AND ACHIEVEMENT 543 head inside the helmet-like device for 15 min. The signal was recorded with a bandpass filter set at 0.1 and 50 Hz, and digitized for 950 ms (254-Hz sampling rate) including a 150-ms prestimulus period. The next step involved visual inspection of single-trial ERF segments to identify those contaminated by (a) eye- or head-movement related magnetic artifacts, or (b) epileptogenic activity. Movement artifacts were defined as magnetic flux deflections >3 picotesla (pt) peak-to-peak in the recordings from magnetometer sensors placed just above the eyes. As it is possible to mistake small blinks or lateral eye movements with magnetic deflections caused by brain activity, the surface distribution of magnetic flux associated with these deflections was frequently taken into account. ERF epochs that contained epileptiform events (spikes or sharp waves) were similarly excluded from further analyses. The recordings were then filtered offline with a bandpass between 0.1 and 20 Hz and subjected to an adaptive filtering procedure that is part of the 4-D Neuroimaging signalanalysis package. These steps are necessary to minimize the amount of low-frequency magnetic noise that is typically present in MEG recordings. At least 140 ERF artifact-free epochs were used to calculate two averaged waveforms (one for trial blocks 1 3 and a second reflecting magnetic activity obtained during trial blocks 4 6). The intracranial generators, or activity sources, of the averaged ERFs were modeled as single equivalent current dipoles (ECDs) and fitted at successive 4-ms intervals (55). For a given time, the source-fitting algorithm was applied to the magnetic flux measurements obtained from a group of sensors, always including both magnetic flux extremes. Source computation was restricted to latency periods during which a single pair of magnetic flux extremes dominated the left and/or the right half of the head surface. The algorithm used in this study searched for the source that was most likely to have produced the observed magnetic field distribution at a given time. Source solutions were considered satisfactory if they were associated with a correlation coefficient of 0.9 between the observed and the best predicted magnetic field distribution. After dipole modeling was performed, the estimated activity sources associated with the late components of the ERFs (>200 ms after stimulus onset) were examined in detail. Contiguous activity sources from each of the two (split-half) ERFs were compared according to their (a) degree of latency overlap and (b) spatial proximity. With this procedure, we identified reproducible clusters of activity sources, which in previous studies have been verified to overlap anatomically with Wernicke s area (17,18). The number of consecutive late activity sources reflects the amount of time neuromagnetic activity took place in each area during processing of the word stimuli. This measure has been shown to be an excellent index of the degree of engagement of language-specific cortices in each hemisphere in language tasks (16). Hemispheric asymmetries of language-specific magnetic activity were determined by using a laterality index (LI), which was calculated according to the formula: (R L)/(R + L), where R represents the number of acceptable late activity sources in the right hemisphere, and L, the corresponding number in the left hemisphere. In addition, the earliest latency when late activity sources were first noted in languagespecific cortices was examined. In previous studies, this measure was found to be a significant predictor of behavioral performance in language tasks (57). To determine the anatomic regions corresponding to each activity source, source locations, which were initially computed in reference to the MEG cartesian coordinate system mentioned earlier, were coregistered on T 1 -weighted, magnetic resonance images (MRIs) (TR, 13.6 ms; TE, 4.8 ms; recording matrix, pixels, 1 excitation, 240-mm field of view, and 1.4-mm slice thickness) obtained from each participant. Transformation of the MEG coordinate system into MRI-defined space was achieved with the aid of three lipid capsules inserted into the ear canals and attached to the nasion, which were easily visualized on the MRIs, by using the MRI Overlay tool, which is part of the 4-D Neuroimaging software. A standard MRI atlas of the human brain (58) served as a reference for the identification of the cerebral structures where sources were localized. RESULTS All patients exhibited late activity sources (after the resolution of the N1m, 200 ms after stimulus onset) in the posterior portions of the left and/or right temporal lobes, including superior temporal gyrus (often extending into the superior temporal sulcus and the middle temporal gyrus, and occasionally into the supramarginal and angular gyri). To examine the temporal course of late, languagespecific MEG activation during the language task within these areas, the number of sources, averaged across the two split-halves of the data, were determined for successive 100-ms bins beginning after the resolution of the N1m for each patient. Means for the achievement-deficit groups with left-hemisphere seizure onset are presented in Fig. 1a for the left hemisphere and in Fig. 1b for the right hemisphere. Group means for patients with right-hemisphere seizure onset are presented in Fig. 1c and d. These trends were analyzed by using a multivariate approach to a repeated measures design with Epoch ( ms, ms,..., ms) and Hemisphere (left, right) as within-subjects variables and Group membership (RS, NI) as the between-subjects variable. This analysis was performed separately for the left- and righthemisphere seizure-onset groups. Performance IQ was included as a covariate to control for group differences in intellectual function. Within the left seizure-onset group,

5 544 J. I. BREIER ET AL. FIG. 1. Mean number of language-specific magnetoencephalography activity sources observed in the superior temporal gyrus, superior temporal sulcus, middle temporal gyrus, and supramarginal and angular gyri in the (a) left or (b) right hemisphere for patients with left-hemisphere seizure onset, and the (c) left or (d) right hemisphere of patients with right-hemisphere seizure onset. NI, no academic deficiency (solid lines); RS, reading and/or spelling deficient (dotted lines). a significant Group Hemisphere interaction was found: F(1, 57) = 4.13; p < Follow-up analyses, performed separately for each achievement group, indicated a significant effect of Hemisphere for the NI group, F(1, 38) = 34.75; p < In contrast, the effect of Hemisphere was not significant for the RS group (p > 0.1). Analyses for patients with right-hemisphere seizure onset indicated a significant effect of Hemisphere [F(1, 20) = 5.46, p < 0.03] but no effects of achievement-deficit Group (p > 0.35). The mean number of activity sources in each hemisphere across epochs and the onset latency for this activa- tion for each achievement-deficit group with left and right seizure onset are presented in Table 3. Both measures are significant predictors of behavioral performance in language tasks (56). As these analyses indicate, whereas all four groups exhibit greater left- as compared with righthemisphere activation, this asymmetry is attenuated in RS patients with left-hemisphere seizure onset. As mentioned earlier, group effects on the onset latency of late language-specific activation were evaluated by using a repeated measures approach to a within-subjects design with Hemisphere (left, right) as the within-subjects variable and Group (NI, RS) as the between-subjects variable. TABLE 3. Group Means on MEG measures NI Left Hemisphere Right Hemisphere Left Hemisphere Right Hemisphere Left Hemisphere Seizure Onset #ofactivity Sources M SD Onset Latency (ms post stimulus onset) M Right Hemisphere Seizure Onset SD #ofactivity Sources M SD Onset Latency (ms post stimulus onset) M SD RS

6 LANGUAGE ORGANIZATION AND ACHIEVEMENT 545 FIG. 2. Averaged waveforms (bottom) for asingle patient and isofield contour maps (top) obtained at specific times (indicated in milliseconds after the onset of the stimulus) for the same patient. In the waveform, upward deflections indicate outgoing magnetic flux, whereas downward deflections indicate inward magnetic flux. The separation between adjacent isofield lines in the contour maps is 10 ft; dotted lines, inward magnetic flux; solid lines, regions of outward magnetic flux. Although analyses did not reach significance for either the right- (p > 0.6) or left-hemisphere onset groups (p > 0.12), the typical asymmetry in onset latency, with earlier onset in the left as compared with right hemisphere for late, language-specific MEG activation, is reversed in the RS group with left-hemisphere seizure onset. The averaged waveforms and examples of isofield contour maps obtained at specific times for a single patient are presented in Fig. 2. The coregistered MEG-MRI scans for patients with typical profiles of language-specific MEG activity from the NI and RS groups are presented in Fig. 3a and 3b, respectively. The RS patient exhibits a more bilateral profile of activation in posterior language areas than the NI patient. The relation between language representation and academic achievement as a continuous variable To evaluate the relation between academic achievement and language organization on an individual basis, the reading/spelling achievement composite score was regressed on the MEG LI. Again, Performance IQ was used as a covariate to control for effects of general intellectual function on results. A significant relation was found between the two measures within the left-hemisphere group [F(1, 57) = 5.18, p < 0.03] but not within the right-hemisphere group (p > 0.45). The mean of the spelling and reading scores is plotted as a function of the MEG LI in Fig. 3. Academic skills tend to decrease as the MEG LI becomes more positive, indicating a greater degree of atypical (more relative activity in the right hemisphere) hemispheric asymmetry in language-specific brain activity. FIG. 3. Maps of magnetoencephalography (MEG) activation (yellow circles) for patients with left-hemisphere seizure onset and (a) no academic achievement deficit and (b) global academic achievement deficits. (Left hemisphere is on the left side of the page. MEG activation sources extending over the width of 1 cm are projected onto the same sagittal image.)

7 546 J. I. BREIER ET AL. FIG. 4. The relation between the magnetoencephalography laterality index and the mean of the achievement for patients with left-hemisphere seizure onset. The relation between global intellectual abilities and patterns of language-specific MEG activity Within the left hemisphere onset group, no significant correlations were noted between Full-Scale (p > 0.09), Verbal (p > 0.11), or Performance (p > 0.2) IQ scores and the MEG LI. DISCUSSION The MEG task used in the present study has been shown to provide a reliable and valid index of the relative engagement of areas of the brain specialized for receptive language function (16). Therefore current findings suggest that deficits in reading and spelling achievement are associated with atypical language organization in patients with chronic seizure disorder of left-hemisphere onset. This abnormality appears to be characterized by greater relative engagement of the right hemisphere during language processing. Consistent with group findings, examination of the relation between academic achievement and language laterality on an individual basis indicated that the probability of poor academic achievement increased significantly with increasing lateralization of receptive language representation to the right hemisphere across patients with left-hemisphere seizure onset. Previous studies in patients with chronic seizure disorder regarding the relation between atypical language organization, as indexed by the Wada test, and cognition have often emphasized the negative effects of reorganization of language cortex to the right hemisphere on nonverbal functions, with relative preservation of verbal functions (24 28). Few studies, however, have assessed academic skills in this group. In the current study, we found evidence for a reduction in reading and spelling skills to be related to abnormal patterns of activation of posterior temporal and inferior parietal language areas known to be involved in these skills. Although evidence was found for an association between a reduction in academic skills and atypical language organization, unlike Loring et al. (25), we did not find a strong relation between atypical language organization and Performance IQ. As indicated earlier, Loring et al. (25) found that the crowding effect was specific to patients who had sustained a lesion severe enough to cause either a complete shift of language function to the right hemisphere or bilateral representation with a presumed shift of hand dominance from right to left. Current findings suggest no effect of atypical language representation on Performance IQ when the language shift is not complete, but only one patient exhibited what was interpreted as a complete shift of language function to the right hemisphere, and only seven patients with left-hemisphere seizure onset were left-handed. As this was a small number of patients, they were included with the larger group, and no effort was made to address the effect of handedness on the relation between academic achievement and language representation. In a larger group, this factor could be included in analyses, and the negative effects of atypical language organization on global nonverbal abilities might emerge. It is not surprising that relations between reading and spelling skills and the degree of atypical language representation were limited to patients with left-hemisphere seizure onset. Language function is represented within the left hemisphere in the vast majority of neurologically intact right-handed controls (58,59), and a number of studies have found a significant increase in bilateral and/or righthemisphere representation of language function in patients with epilepsy of left-hemisphere origin (16,29,60 63), presumably secondary either to the seizures themselves or to the seizure substrate. Seizure onset in the right hemisphere would, therefore, not be expected to affect language organization except in the most unusual cases in which the right hemisphere was genetically predetermined to be dominant for language function. The consistent relation between atypical language representation and achievement deficits does support the hypothesis that either the seizure substrate and/or seizures themselves result in a neurophysiologic abnormality that affects the acquisition of academic skills in individuals with chronic seizure disorder of left-hemisphere origin. The relatively earlier onset of activity in the right hemisphere in this group suggests that this abnormality may be secondary to reorganization of language function to right-hemisphere areas that are not as efficient as homotopic areas in the left hemisphere in supporting reading and spelling functions. We used a cutoff definition of reading disability that is similar to that used in studies by our group (30,31) and others (32,33). This type of approach is consistent with suggestions that reading ability may be normally distributed in populations without neurologic involvement (64). In addition, a discrepancy approach has the disadvantage of classifying individuals with lower IQ scores as being nondisabled (34,35), although they may have a

8 LANGUAGE ORGANIZATION AND ACHIEVEMENT 547 significant functional deficit. The utility of the cutoff approach in populations with chronic epilepsy is highlighted by the finding that only four participants in the current study would have been classified as having reading disability by using a standard discrepancy approach of one standard deviation between IQ and reading achievement scores, although a much larger number clearly have a significant reading deficit. Furthermore, findings were independent of intellectual function, again providing support for the current approach. Individuals with seizure disorder experience achievement deficits because of reduction in the quality and quantity of academic experience and alterations in parental and teacher expectations (5,65,66). All of the patients in the current study, however, had chronic seizures, no significant differences were seen among groups in highest school grade achieved, group differences in MEG data were independent of Performance IQ, and the relations between measures of verbal, nonverbal, and global IQ and atypical language organization were not significant. However, significant differences appeared among achievement-deficit groups in IQ scores. In addition, although the areas activated during MEG imaging are known to be involved in reading and spelling function, the MEG method used in this study was developed to image receptive language cortex. Therefore further research is necessary to determine to what extent the current findings generalize to specific cognitive functions and more homogeneous groups of patients. Acknowledgment: This study was supported in part by NINDS grant NS37941 to Andrew C. Papanicolaou, Ph.D., and by the Vivian L. Smith Center for Neurologic Research, Houston, Texas. REFERENCES 1. Austin JK, Huberty TJ, Huster GA, et al. Academic achievement in children with epilepsy or asthma. Dev Med Child Neurol 1998;40: Bailet LL, Turk WR. The impact of childhood epilepsy on neurocognitive and behavioral performance: a prospective longitudinal study. Epilepsia 2000;41: Farwell JR, Dodrill CB, Batzel LW. Neuropsychological abilities of children with epilepsy. Epilepsia 1985;26: Williams J, Phillips T, Griebel ML, et al. Factors associated with academic achievement in children with controlled epilepsy. Epilepsy Behav 2001;2: Seidenberg M, Beck N, Geisser M, et al. Academic achievement of children with epilepsy. Epilepsia 1986;27: Mitchell WG, Chavez JM, Lee H, et al. Academic underachievement in children with epilepsy. J Child Neurol 1991;6: Aldenkamp AP. Cognitive side effects of antiepileptic drugs. In: Jambaque I, Lassonde M, Dulac O, eds. Neuropsychology of childhood epilepsies. New York: Kluwer Academic/Plenum, 2001: Bourgeois BF. Relationship between anticonvulsant drugs and learning disabilities. Semin Neurol 1991;11: Bourgeois BF. Antiepileptic drugs, learning, and behavior in childhood epilepsy. Epilepsia 1998;39: Austin JK, Huberty TJ, Huster GA, et al. Does academic achievement in children with epilepsy change over time? Dev Med Child Neurol 1999;41: Hermann BP, Seidenberg M, Schoenfeld J, et al. Neuropsychological characteristics of the syndrome of mesial temporal lobe epilepsy. Arch Neurol 1997;54: Butterbaugh G, Olejniczak P, Roques B, et al. Lateralization of temporal lobe epilepsy and learning disabilities, as defined by disabilityrelated civil rights law. Epilepsia 2004;45: Breier JI, Simos PG, Wheless JW, et al. Language dominance in children as determined by magnetic source imaging and the intracarotid amobarbital procedure: a comparison. J Child Neurol 2001;16: Breier JI, Simos PG, Zouridakis G, et al. Language dominance determined by magnetic source imaging: a comparison with the Wada procedure. Neurology 1999;53: Papanicolaou AC, Simos PG, Breier JI, et al. Magnetoencephalography (MEG): a method comparable to the Wada procedure for language laterality assessment. Epilepsia 2001;42(S7): Papanicolaou AC, Simos PG, Castillo EM, et al. Magnetocephalography: a noninvasive alternative to the Wada procedure. J Neurosurg 2004;100: Simos PG, Breier JI, Maggio WW, et al. Atypical temporal lobe language representation: MEG and intraoperative stimulation mapping correlation. Neuroreport 1999;10: Simos PG, Papanicolaou AC, Breier JI, et al. Localization of language-specific cortex by using magnetic source imaging and electrical stimulation mapping. J Neurosurg 1999;91: Pugh KR, Mencl WE, Jenner AR, et al. Functional neuroimaging studies of reading and reading disability (developmental dyslexia). Ment Retard Dev Disabil Res Rev 2000;6: Shaywitz SE, Shaywitz BA, Pugh KR, et al. Functional disruption in the organization of the brain for reading in dyslexia. Proc Natl Acad Sci USA 1998;95: Simos PG, Breier JI, Fletcher JM, et al. Cerebral mechanisms involved in word reading in dyslexic children: a magnetic source imaging approach. Cereb Cortex 2000;10: Simos PG, Breier JI, Wheless JW, et al. Brain mechanisms for reading: the role of the superior temporal gyrus in word and pseudoword naming. Neuroreport 2000;11: Wada J, Rasmussen T. Intracarotid injection of sodium amytal for the lateralization of cerebral speech dominance: experimental and clinical observations. J Neurosurg 1960;17: Gleissner U, Kurthen M, Sassen R, et al. Clinical and neuropsychological characteristics of pediatric epilepsy patients with atypical language dominance. Epilepsy Behav 2003;4: Loring DW, Strauss E, Hermann BP, et al. Effects of anomalous language representation on neuropsychological performance in temporal lobe epilepsy. Neurology 1999;53: Strauss E, Satz P, Wada J. An examination of the crowding hypothesis in epileptic patients who have undergone the carotid amytal test. Neuropsychologia 1990;28: Teuber HL Why two brains? In: Worden F, ed. The neurosciences: third study program. Cambridge: MIT Press, 1974: Satz P, Orsini DL, Saslow E, et al. The pathological left-handedness syndrome. Brain Cogn 1985;4: Billingsley R, Smith ML. Intelligence profiles in children and adolescents with left temporal lobe epilepsy: relationship to language laterality. Brain Cogn 2000;43: Breier JI, Brookshire BL, Fletcher JM, et al. Identification of side of seizure onset in temporal lobe epilepsy using memory tests in the context of reading deficits. J Clin Exp Neuropsychol 1997;19: Breier JI, Fletcher JM, Wheless JW, et al. Profiles of cognitive performance associated with reading disability in temporal lobe epilepsy. J Clin Exp Neuropsychol 2000;22: Joanisse MF, Manis FR, Keating P, et al. Language deficits in dyslexic children: speech perception, phonology, and morphology. J Exp Child Psychol 2000;77: Manis FR, Mcbride-Chang C, Seidenberg MS, et al. Are speech perception deficits associated with developmental dyslexia? J Exp Child Psychol 1997;66: Fletcher JM, Espy KA, Francis DJ, et al. Comparisons of cutoff and regression-based definitions of reading disabilities. J Learn Disabil 1989;22:334 8.

9 548 J. I. BREIER ET AL. 35. Fletcher JM, Francis DJ, Rourke BP, et al. The validity of discrepancy-based definitions of reading disabilities. J Learn Disabil 1992;25: Holmes GL. Do seizures cause brain damage? Epilepsia 1991;32:S Bruck M. Persistence of dyslexic s phonological awareness deficits. Dev Psychol 1992;28: Schonhaut S, Satz P. Prognosis for children with learning disabilities: a review of follow-up studies. In: Rutter M, ed. Developmental neuropsychiatry. New York: Guilford Press, 1999: Spreen O. Prognosis of learning disability. J Consult Clin Psychol 1988;56: Oldfield RC. The assessment and analysis of handedness: the Edinburgh inventory. Neuropsychologia 1971;9: Doherty MJ, Walting PJ, Morita DC, et al. Do nonspecific focal EEG slowing and epileptiform abnormalities favor one hemisphere? Epilepsia 2002;43: Gatzonis SD, Roupakiotis S, Kambayianni E, et al. Hemispheric predominance of abnormal findings in electroencephalogram (EEG). Seizure 2002;11: Labar D, Dilone L, Solomon G, et al. Epileptogenesis: left or right hemisphere dominance? Preliminary findings in a hospital-based population. Seizure 2001;10: Wechsler D. Manual for the Wechsler Adult Intelligence Scale. revised. San Antonio, Tex: The Psychological Corporation, Wechsler D. Wechsler Abbreviated Scale of Intelligence (WASI). San Antonio, Tex: The Psychological Corporation, Wechsler D. Manual for the Wechsler Intelligence Scale for Children III. San Antonio, Tex: The Psychological Corporation, Wilkinson GS. Examiners manual: Wide Range Achievement Test 3. Austin, Tex: PRO-ED, Woodcock RW, McGrew KS, Mather N, et al. Johnson III technical manual. Circle Pines, MN: American Guidance Services, Fletcher JM, Shaywitz SE, Shankweiler DP, et al. Cognitive profiles of reading disability: comparisons of discrepancy and low achievement definitions. JEdPsychol 1994;86: Post YV, Swank PR, Hiscock M, et al. Identification of vowel speech sounds by skilled and less skilled readers and the relation with vowel spelling. Ann Dyslex 1999;49: Shaywitz BA, Shaywitz SE. Learning disabilities and attention disorders. In: Swaiman K, ed. Principles of pediatric neurology. St. Louis: CV Mosby, 1994: Stanovich KE, Siegel LS. Phenotypic performance profile of children with reading disabilities: a regression-based test of the phonological-core variable-difference model. J Ed Psychol 1994;86: Paivio A, Yuille JC, Madigan SA. Concreteness, imagery, and meaningfulness values for 925 nouns. J Exp Psychol 1968;suppl 25: Thorndike EL, Lorge I. The teacher s wordbook of 30,000 words. New York: Teacher s College, Bureau of Publications, Sarvas J. Basic mathematical and electromagnetic concepts of the biomagnetic inverse problem. Phys Med Biol 1987;32: Simos PG, Breier JI, Fletcher JM, et al. Brain mechanisms for reading words and pseudowords: an integrated approach. Cereb Cortex 2002;12: Damasio H. Human brain anatomy in computerized images. New York: Oxford University Press, Basic S, Hajnsek S, Poljakovic Z, et al. Determination of cortical language dominance using functional transcranial Doppler sonography in left-handers. Clin Neurophysiol 2004;115: McKeever WF, Seitz KS, Krutsch AJ, et al. On language laterality in normal dextrals and sinistrals: results from the bilateral object naming latency task. Neuropsychologia 1995;33: Brazdil M, Zakopcan J, Kuba R, et al. Atypical hemispheric language dominance in left temporal lobe epilepsy as a result of the reorganization of language functions. Epilepsy Behav 2003;4: Janszky J, Jokeit H, Heinemann D, et al. Epileptic activity influences the speech organization in medial temporal lobe epilepsy. Brain 2003;126: Liegeois F, Connelly A, Cross JH, et al. Language reorganization in children with early-onset lesions of the left hemisphere: an f MRI study. Brain 2004;6: Pataraia E, Simos PG, Billingsley-Marshall RL, et al. Reorganization of language-specific cortex in patients with lesions or mesial temporal epilepsy. Neurology 2004;63: Shaywitz SE, Escobar, MD, Shaywitz BA, et al. Evidence that dyslexia may represent the lower tail of a normal distribution of reading ability. N Engl J Med 1992;326: Austin JK. Childhood epilepsy: child adaptation and family resources. J Child Adolesc Psychiatry Ment Health Nurs 1988;1: Austin JK, McDermott N. Parental attitude and coping behaviors in families of children with epilepsy. J Neurosci Nurs 1988;20: Booth J, Burman D, Van Santen F, et al. The development of specialized brain systems in reading and oral language. Neuropsychol Dev Cogn Sect C Child Neuropsychol 2001;7: