Neuropsychology and temporal lobe epilepsy

The purpose of this article is to review aspects of the neuropsychology of temporal lobe epilepsy. Evidence will be presented to demonstrate that the cognitive consequences of this focal seizure disorder can be more generalized in nature than expected. Consistent with the extratemporal neurocognitive findings, careful quantitative magnetic resonance imaging volumetrics have shown that structural brain changes may be detected outside the temporal lobes. Many factors can potentially affect cognition and brain structure. We focus on the potential neurodevelopmental impact of early-onset temporal lobe epilepsy on brain structure and cognition positing that this disorder can have both immediate and lifespan implications for cognition and psychosocial status.     

Interictal discourse production in temporal lobe epilepsy

Discourse production was investigated in 16 patients with left-temporal-lobe epilepsy (LTLE) and 17 neurologically normal relatives. Narrative was elicited with the Joanette and Goulet (1989) eight frame "Cowboy" cartoon on three immediately consecutive occasions. Variables of interest were fluency (words/s), speaking time, and word count. While controls produced increasingly concise forms of the story over repetitions, the TLE group became more verbose. Fluency did not not differentiate the groups. These findings suggest that LTLE is associated with macrolinguistic disturbances. The role of capacity limitations and motivational factors in TLE discourse production is discussed.     

Attentional control in patients with temporal lobe epilepsy

Across different studies, patients with temporal lobe epilepsy (TLE) demonstrate impairments on numerous measures of attentional control that are classically associated with frontal lobe functioning. One aspect of attentional control that has not been examined in TLE is the ability to execute two modality‐specific tasks concurrently. We sought to examine the status of dual‐task coordination in TLE. We further examined the cohorts' performance on a range of traditional measures of attentional control. Eighteen TLE patients and 22 healthy controls participated in the study. Dual‐task performance involved comparing the capacity to execute a tracking and a digit recall task simultaneously with the capacity to execute the tasks separately. We also administered measures of: set shifting (odd‐man‐out test), sustained attention (elevator counting), selective attention (elevator counting with distraction), and divided attention (trail making test). We found that the proportional decrement in dual‐task performance relative to single‐task performance did not vary between the groups (TLE = 92.48%; controls = 93.70%), nor was there a significant difference in sustained attention (p > .10). Patients with TLE did demonstrate marked deficits in selective attention (p < .0001), divided attention (p < .01), and set shifting (p < .01). These findings add to the knowledge about cognitive dysfunction in TLE, indicating that impairments in attentional control in TLE tend to be selective. The greatest deficits appear to be on tasks that invoke a high level of processing resources. In contrast, sustained attention is less compromised and the capacity to allocate cognitive resources appears to be normal in patients with TLE.     

Controlled recall of verbal material in temporal lobe epilepsy

This study used a guided process‐dissociation procedure to examine the contribution of controlled and automatic uses of memory to a cued‐recall task in 24 patients with unilateral temporal lobe epilepsy (TLE: 12 left‐sided; 12 right‐sided), and 12 neurotypical controls. In an inclusion task, participants attempted to complete three‐letter word stems using previously studied words, in an exclusion task they aimed to avoid using studied words to complete stems. Patients with left TLE produced fewer target completions under inclusion conditions. Completion rates were not significantly different under exclusion conditions. Estimates derived from process dissociation calculations, confirmed that the cued‐recall deficit in left TLE patients arose entirely from impairment in controlled memory processes. There were no group differences in the estimates of automatic processes. Recognition judgements of stems corresponding to studied words did not differ between the groups. Overall the results support the view that controlled and automatic memory processes are mediated by separable neural systems. Hippocampal and related structures within the left MTL are more important than corresponding right hemisphere structures for the controlled retrieval of verbal material. In contrast, the findings from this study do not suggest that the left and right temporal lobes make a differential contribution to automatic memory processing. The theoretical and clinical relevance of these findings are discussed.     

Metamemory in temporal lobe epilepsy: Self-monitoring of memory functions

Temporal lobe seizure patients and control subjects participated in an investigation of metamemory. The two-part study explored the individual's perception of memory abilities for both encoding and retrieval. Experiment I addressed self-monitoring of encoding through a study of prediction of memory span. Experiment II explored self-monitoring of retrieval through a study of "Feeling of Knowing." The results indicate that left and right temporal lobe seizure patients tend to overestimate their memory capacities, in comparison with normal controls, and that self-monitoring tends to be less accurate for material (verbal or nonverbal) mediated by side of lesion. The potential impact of inaccurate memory monitoring on the memory dysfunction observed in seizure patients is discussed.     

Language after temporal or frontal lobe surgery in children with epilepsy

The purpose of this study was to compare language function in children undergoing temporal or frontal lobe surgery to control intractable epilepsy. Language measures (expressive vocabulary, receptive vocabulary, comprehension, reading, spelling, phonemic fluency, and category fluency) were administered to 9 children with frontal lobe epilepsy (mean age: 10.8 +/- 2.7 years) and 10 children with temporal lobe epilepsy (11.5 +/- 2.6 years). The results indicate that no differences exist, in language function, before and after surgery between children with frontal and temporal lobe epilepsy (all p > or = .05). Children with left hemisphere lesions had significantly lower scores than those with right on category fluency and comprehension, but laterality effects were not seen on the other measures. In both groups, language function was not significantly affected by surgery.     

Confirmatory factor analysis of the WMS-III in patients with temporal lobe epilepsy

Five competing models specifying the factor structure underlying the Wechsler Memory Scale-Third Edition (D. Wechsler, 1997b) primary subtest scores were evaluated in a sample of patients with intractable temporal lobe epilepsy (N = 254). Models specifying separate immediate and delayed constructs resulted in inadmissible parameter estimates and model specification error. There were negligible goodness-of-fit differences between a 3-factor model of working memory, auditory memory, and visual memory and a nested--more parsimonious--2-factor model of working memory and general memory. The results suggest that specifying a separate visual memory factor provides little advantage for this sample--an unexpected finding in a population with lateralized dysfunction, for which one might have predicted separate auditory and visual memory dimensions.     

Ambiguous emotion recognition in temporal lobe epilepsy: The role of expression intensity

The lateralization of emotion processing is currently debated and may be further explored by examining facial expression recognition (FER) impairments in temporal lobe epilepsy (TLE). Furthermore, there is also debate in the literature whether FER deficits in individuals with TLE are more pronounced in the right than in the left hemisphere. Individuals with TLE were tested with an FER task designed to be more sensitive than those classically used to shed light on this issue. A total of 25 right- and 32 left-TLE patients, candidates for surgery, along with controls, underwent an FER task composed of stimuli shown not only at full-blown intensities (100 %), but also morphed to lower-intensity display levels (35 %, 50 %, and 75 %). The results showed that, as compared to controls, right-TLE patients showed deficits in the recognition of all emotional categories. Furthermore, when considering valence, right-TLE patients were impaired only in negative emotion recognition, but no deficits for positive emotions were highlighted in left-TLE patients. Finally, only the right-TLE patients’ impairment was found to be related to the age of epilepsy onset. Our work demonstrates that the FER deficits in TLE span multiple emotional categories and show manifestations dependent on the laterality of the epileptic focus. Taken together, our findings provide the strongest evidence for the right-hemisphere model, but they also partially support the valence model. We suggest that current models are not exhaustive at explaining emotional-processing cerebral control, and further that multistep models should be developed.     

Disruption of spelling-to-sound correspondence mapping during single-word reading in patients with temporal lobe epilepsy

Processing and/or hemispheric differences in the neural bases of word recognition were examined in patients with long-standing, medically-intractable epilepsy localized to the left (N=18) or right (N=7) temporal lobe. Participants were asked to read words that varied in the frequency of their spelling-to-sound correspondences. For the right temporal lobe group, reaction times (RTs) showed the same pattern across spelling-to-sound correspondence conditions as previously reported for normal participants. For the left temporal lobe group, however, the pattern of RTs suggested a greater relative influence of orthographic frequency than rime frequency, such that performance was worse on words whose orthographic body was less frequent in the language. We discuss these results in terms of differences in processing between the two cerebral hemispheres: the results for the right-temporal lobe patients are taken to support connectionist models of reading as described for the dominant (left) hemisphere, while results for the left-temporal lobe patients support a view of the right hemisphere as relatively less sensitive to phonology and relatively more sensitive to orthography.     

Levels of processing with free and cued recall and unilateral temporal lobe epilepsy

This study investigates the role of the temporal lobes in levels-of-processing tasks (phonetic and semantic encoding) according to the nature of recall tasks (free and cued recall). These tasks were administered to 48 patients with unilateral temporal epilepsy (right "RTLE"=24; left "LTLE"=24) and a normal group (n=24). The results indicated that LTLE patients were impaired for semantic processing (free and cued recall) and for phonetic processing (free and cued recall), while for RTLE patients deficits appeared in free recall with semantic processing. It is suggested that the left temporal lobe is involved in all aspects of verbal memory, and that the right temporal lobe is specialized in semantic processing. Moreover, our data seem to indicate that RTLE patients present a retrieval processing impairment (semantic condition), whereas the LTLE group is characterized by encoding difficulties in the phonetic and semantic condition.     

Contributions of volumetrics of the hippocampus and thalamus to verbal memory in temporal lobe epilepsy patients

Recent theories have posited that the hippocampus and thalamus serve distinct, yet related, roles in episodic memory. Whereas the hippocampus has been implicated in long-term memory encoding and storage, the thalamus, as a whole, has been implicated in the selection of items for subsequent encoding and the use of retrieval strategies. However, dissociating the memory impairment that occurs following thalamic injury as distinguished from that following hippocampal injury has proven difficult. This study examined relationships between MRI volumetric measures of the hippocampus and thalamus and their contributions to prose and rote verbal memory functioning in 18 patients with intractable temporal lobe epilepsy (TLE). Results revealed that bilateral hippocampal and thalamic volume independently predicted delayed prose verbal memory functioning. However, bilateral hippocampal, but not thalamic, volume predicted delayed rote verbal memory functioning. Follow-up analyses indicated that bilateral thalamic volume independently predicted immediate prose, but not immediate rote, verbal recall, whereas bilateral hippocampal volume was not associated with any of these immediate memory measures. These findings underscore the cognitive significance of thalamic atrophy in chronic TLE, demonstrating that hippocampal and thalamic volume make quantitatively, and perhaps qualitatively, distinct contributions to episodic memory functioning in TLE patients. They are also consistent with theories proposing that the hippocampus supports long-term memory encoding and storage, whereas the thalamus is implicated in the executive aspects of episodic memory.     

Intelligence profiles in children and adolescents with left temporal lobe epilepsy: relationship to language laterality

Pre- and postoperative IQ profiles were examined in children and adolescents with left temporal-lobe epilepsy. Participants were grouped according to their pattern of speech dominance, as indicated by the intracarotid sodium amobarbital procedure. Seven participants with typical (left hemisphere) speech representation were matched for age and sex to seven participants with atypical (bilateral or right hemisphere) speech representation. The group with atypical speech representation tended to perform worse pre- and postoperatively on several verbal and nonverbal subtests. Unlike Strauss et al.'s (1990) sample, however, the groups did not differ significantly on pre- or postoperative Performance IQ. The results suggest that when individuals are matched closely for age and sex, there may only be limited support for a nonverbal "crowding" hypothesis. Atypical speech representation in children and adolescents with early temporal-lobe epilepsy appears to be associated with lowered performance on both verbal and nonverbal IQ subtests.     

Neurodevelopmental differences in emotional prosody in normal children and children with left and right temporal lobe epilepsy

Emotional prosody, which has been defined as the emotional aspects of speech which communicate pleasure, fear, sorrow, anger, etc., has been demonstrated to be primarily a function of the nondominant hemisphere (typically, the right hemisphere) in adult populations. However, few researchers have addressed the developmental or lateralized nature of emotional prosody in children. In this study, an instrument was developed to measure the receptive aspects of emotional prosody in pediatric populations and administered to normal children ages 6 to 11 years old. An analysis of variance revealed significant age-related differences. Additionally, the instrument was administered to 12 children with right temporal lobe epilepsy and 11 children with left temporal lobe epilepsy. Analysis of variance indicated that there was no significant difference in scores between the left temporal and right temporal lobe groups. However, right temporal epileptic patients scored significantly lower than normal children on all sections of the instrument, suggesting that in children like adults, the right temporal lobe may be dominant with respect to the receptive aspects of emotional prosody.     

Dissociation of short- and long-term face memory: evidence from long-term recency effects in temporal lobe epilepsy

We tested whether memory deficits in temporal lobe epilepsy (TLE) are better described by a single- or dual-store memory model. To this aim, we analyzed the influence of TLE and proactive interference (PI) on immediate and 24-h long-term recency effects during face recognition in 16 healthy participants and 18 right and 21 left non-surgical TLE patients. PI in healthy participants or TLE erased the long-term recency effect, but left the immediate recency effect unaffected. Although the immediate recency effect was still visible in right TLE patients, the number of detected recency items during immediate recognition was decreased in right TLE compared to left TLE. Right TLE was also related to decreased detection of pre-recency items during delayed recognition compared to left TLE, and decreased detection of pre-recency items during immediate recognition under PI. The results show that the temporal lobes are necessary for the long-term recency effect, but not for the immediate recency effect, and thus speak for a dissociation of short- and long-term memory for faces. Right TLE is related to more severe long-term memory deficits than left TLE and is also related to additional short-term memory deficits for faces.     

Remembering versus knowing during face recognition in unilateral temporal lobe epilepsy patients with or without hippocampal sclerosis

Recognition memory involves knowing an item was learned (familiarity) and remembering contextual details about the prior learning episode (recollection). We tested three competing hypotheses about the role of the hippocampus in recollection and familiarity. It mediates either recollection or familiarity, or serves both processes. We further tested whether the left temporal lobe mediates recollection and the right temporal lobe familiarity (modes of processing view), or whether the two temporal lobes mediate remembering material specifically (material specificity view). We investigated 24-h face recognition using the "remember-know" procedure. We studied 23 left and 24 right temporal lobe epilepsy (LTLE/RTLE) patients with and without hippocampal sclerosis (HS+/HS-) and 31 healthy participants. HS+ patients made fewer know responses than HS- patients or healthy participants. RTLE was related to fewer remember responses than LTLE. Our results suggest the hippocampus has a critical role in familiarity. Further, our findings support the material specificity hypothesis of laterality.     

Dichotic listening with forced attention in patients with temporal lobe epilepsy: significance of left hemisphere cognitive dysfunction

Fifty right-handed patients with focal temporal lobe epilepsy were administered a dichotic listening test with consonant-vowel syllables under non-forced, forced right and forced left attention conditions, and a neuropsychological test battery. Dichotic listening performance was compared in subgroups with and without left hemisphere cognitive dysfunction, measured by the test battery, and in subgroups with left and right temporal epileptic focus. Left hemisphere cognitive dysfunction led to more correct responses to left ear stimuli in all three attention conditions, and fewer correct responses to right ear stimuli in the non-forced attention condition. This was probably caused by basic left hemisphere perceptual dysfunction. Dichotic listening was less affected by a left-sided epileptic focus than by left hemisphere cognitive dysfunction. General cognitive functioning influenced dichotic listening performance stronger in forced than in non-forced attention conditions. Larger cerebral networks were probably involved in the forced attention conditions due to the emphasis on conscious effort.