- Top of page
Purpose: This study compared frontotemporal brain volumes in children with childhood absence epilepsy (CAE) to age- and gender-matched children without epilepsy. It also examined the association of these volumes with seizure, demographic, perinatal, intelligence quotient (IQ), and psychopathology variables.
Methods: Twenty-six children with CAE, aged 7.5–11.8 years, and 37 children without epilepsy underwent brain magnetic resonance imaging (MRI) scans at 1.5 Tesla. Tissue was segmented, and total brain, frontal lobe, frontal parcellations, and temporal lobe volumes were computed. All children had IQ testing and structured psychiatric interviews. Parents provided seizure, perinatal, and behavioral information on each child.
Results: The CAE group had significantly smaller gray matter volumes of the left orbital frontal gyrus as well as both left and right temporal lobes compared to the age- and gender-matched children without epilepsy. In the CAE group these volumes were related to age, gender, ethnicity, and pregnancy complications but not to seizure, IQ, and psychopathology variables. In the group of children without epilepsy, however, the volumes were related to IQ.
Conclusion: These findings suggest that CAE impacts brain development in regions implicated in behavior, cognition, and language. In addition to supporting the cortical focus theory of CAE, these findings also imply that CAE is not a benign disorder.
Recent voxel-based morphometry studies have demonstrated increased anterior thalamic volume (Betting et al., 2006b) and gray matter concentration in the superior mesiofrontal region in adults with absence seizures (Betting et al., 2006a), but thalamic atrophy and reduction of subcortical gray matter volume, increase in right temporal and frontal lobe white matter volumes, and reduced gray matter volume in the thalamus of adolescents and young adults with childhood absence epilepsy (CAE) (Chan et al., 2006; Pardoe et al., 2008). Given possible confounding of long-term effects of duration of the disorder and chronic use of antiepileptic drugs (AEDs) on brain maturation, the study presented herein compared brain volumes in 7.5–11.8-year-old children with CAE to those of age- and gender-matched children without epilepsy.
The study focused on volumes of the frontal lobe, frontal lobe parcellations (i.e., dorsolateral prefrontal gyrus, orbital frontal gyrus, and inferior frontal gyrus), and temporal lobe for four reasons. First, electroencephalography (EEG) studies demonstrate that absence seizures begin with discrete, often unilateral, spikes in the dorsolateral frontal or orbital frontal regions and evolve to engage orbital frontal and mesial frontal regions as well as the temporal lobes during the repeating spike and wave cycles (Holmes et al., 2004; Tucker et al., 2007). Second, the results of imaging studies described previously imply involvement of these brain regions in CAE (Betting et al., 2006a,b; Chan et al., 2006). Third, ongoing development of the frontal and temporal lobes during childhood and adolescence (Sowell et al., 2003b; Gogtay et al., 2004) underscores the importance of examining brain volumes in both these regions in CAE. Fourth, involvement of the orbital frontal gyrus, inferior frontal gyrus, and dorsolateral prefrontal gyrus in behavior/emotions (Sowell et al., 2003a), cognition (Alvarez & Emory, 2006), and language (Szaflarski et al., 2006) suggests that structural abnormalities in these brain regions might be related to the difficulties children with CAE have in these areas of functioning (Williams et al., 1996; Mandelbaum & Burack, 1997; Ott et al., 2001; Pavone et al., 2001; Caplan et al., 2008).
On the basis of prior volumetric (Betting et al., 2006a; Chan et al., 2006) and EEG findings (Holmes et al., 2004; Tucker et al., 2007), albeit in older CAE subjects, we predicted abnormal total brain, frontotemporal, dorsolateral prefrontal gyrus, and orbital frontal gyrus volumes in the CAE group compared to the group without epilepsy. Because age, gender, and IQ (Reiss et al., 1996; Lenroot & Giedd, 2006; Wilke et al., 2007) are related to brain volumes in children without epilepsy, we explored whether the CAE group differed from the nonepilepsy group in the relationship of their brain volumes to these variables.
Within the CAE group, we determined if abnormal frontotemporal volumes were associated with seizure variables and with perinatal complications due to evidence for increased pregnancy complications in pediatric epilepsy (Sidenvall et al., 2001). Finally, the high rate of attention deficit hyperactivity disorder (ADHD) and anxiety disorders in CAE (Caplan et al., 2008), structural and functional abnormalities in the frontal lobe of children without epilepsy who have these psychiatric disorders (De Bellis et al., 2002; Sowell et al., 2003a; McClure et al., 2007), and increased frontal lobe volumes in children with new-onset epilepsy with ADHD (Hermann et al., 2007) underlie the rationale for investigating if psychopathology variables would be associated with CAE frontotemporal volumes.
- Top of page
This first study of frontotemporal volumes in children with CAE demonstrates significantly smaller gray matter volumes of the left orbital frontal gyrus and both temporal lobes compared to age- and gender-matched nonepilepsy children. The lack of an association of these volumetric abnormalities with seizure variables in the CAE group, similar to older CAE patients (Betting et al., 2006a; Chan et al., 2006), implies a role for the neuropathology underlying CAE on brain development from childhood.
These findings have important functional, theoretical, and developmental implications. From the functional perspective, involvement of the frontal and temporal lobes in cognition (Alvarez & Emory, 2006), language (Szaflarski et al., 2006), and behavior/emotions (De Bellis et al., 2002; Sowell et al., 2003a; McClure et al., 2007) suggest that abnormalities in these brain regions might contribute to the reported deficits in these skills in children with CAE (Williams et al., 1996; Mandelbaum & Burack, 1997; Pavone et al., 2001; Henkin et al., 2003, 2005; Caplan et al., 2008). Furthermore, these structural brain abnormalities challenge the theory that CAE is a “benign” disorder.
Among the study’s developmental implications, smaller gray matter volumes in the orbital frontal gyrus and temporal lobes of the CAE group but no significant volumetric differences in the mesial and dorsolateral prefrontal regions, unlike the adult and adolescent studies (Betting et al., 2006a; Chan et al., 2006), might reflect the younger age of our subjects. In addition, the association of smaller orbital frontal gray matter volumes with perinatal complications could imply a specific developmental vulnerability of this brain region in CAE. In support of this explanation, there was no relationship between perinatal difficulties and smaller orbital frontal gyrus gray matter volumes in the control group despite similar rates of pregnancy problems in the CAE and nonepilepsy groups. Furthermore, epidemiologic studies have shown that children whose mothers had pregnancy problems are at increased risk of unprovoked seizures (Sidenvall et al., 2001) and epilepsy (Whitehead et al., 2006).
A third developmental implication is the dissociation in the relationship of frontotemporal volumes with age in the CAE group but with IQ in the normal group. Similar findings in youth who have recent-onset seizures (Hermann et al., 2006) and chronic complex partial seizures (Daley et al., 2007) suggest abnormal maturation patterns in the brains of children with epilepsy with average intelligence compared to children without epilepsy. Although the mean IQ of the nonepilepsy group in our study was significantly higher than that of the CAE group, the relationship between IQ and frontotemporal volumes is similar to other findings in typically developing children (Reiss et al., 1996).
Supporting the left orbital frontal gyrus and bilateral temporal lobe gray matter volume findings in the CAE group, Holmes and colleagues in their source analysis of dense-array, 256-channel scalp EEG studies demonstrated that the onset of absence seizures could be unilateral in the orbital frontal or dorsolateral prefrontal region (Holmes et al., 2004) and that the slow waves of the discharges were restricted to frontotemporal networks, particularly ventromedial frontal networks (Tucker et al., 2007). In fact, localization of CAE spikes (Holmes et al., 2004; Tucker et al., 2007) and abnormal gray matter volumes in the orbital frontal gyrus, together with evidence for the regulatory inhibitory role of this brain region on thalamocortical circuitry via the nucleus reticularis of the thalamus [See review in (Holmes et al., 2004)] bolster the cortical focus theory of CAE [See review in (Meeren et al., 2005)]. Moreover, the lack of an association between orbital frontal gyrus gray matter volumes and seizure variables implies that the underlying neuropathology, not seizures, influences brain development in CAE.
Involvement of the orbital frontal region might also play a role in the comorbidities of CAE (Williams et al., 1996; Mandelbaum & Burack, 1997; Pavone et al., 2001; Henkin et al., 2003, 2005; Caplan et al., 2008). Based on their review of evidence that the orbital frontal region inhibits the thalamoreticular nucleus, mainly in the rostral pole and its association with the limbic, midline, and intralaminar thalamic nuclei, Holmes et al. (2004) suggested that this region regulates alertness, arousal, and motivation. Impairment of these essential functions might, in turn, contribute to the difficulties children with CAE have with cognitive, linguistic, and behavioral/emotional functioning (Williams et al., 1996; Mandelbaum & Burack, 1997; Pavone et al., 2001; Henkin et al., 2003, 2005; Caplan et al., 2008).
Of note, we found no significant differences in the frontal lobe parcellation volumes of the CAE children with and without ADHD in the current study and in children with complex partial seizures (Daley et al., 2007), despite high rates of ADHD in both disorders (Caplan et al., 2004, 2008). Interestingly, children with recent-onset epilepsy with ADHD have significantly larger gray matter volumes of the frontal lobe (Hermann et al., 2007), whereas ADHD children without epilepsy have reduced volume (Castellanos et al., 2002; Durston et al., 2004; Shaw et al., 2006) and cortical thickness (Sowell et al., 2003a) of the frontal lobe. These discrepant findings underscore the need to use similar volumetric and morphometric methods to determine if ADHD in epilepsy involves specific neural circuits and how these vary by epilepsy syndrome and chronicity.
Generalizability of the study findings is restricted by sample-related, study design, and statistical limitations. Regarding the sample, the high rate of uncontrolled seizures in 73% of the CAE subjects reflects the study’s inclusionary criterion that each CAE subject had to have had at least one seizure in the year prior to participation in the study. Of note, 61.5% of study subjects in Chan et al. (2006) and 25% in Betting et al. (2006a) also had uncontrolled seizures. In addition, lack of information on which study subjects had oral or eyelid myoclonia, as well as inclusion of three CAE subjects with generalized tonic–clonic convulsions, could predict an unfavorable prognosis (Grosso et al., 2005) or different course (Incorpora et al., 2002) of CAE. However, removal of these three children from the analyses did not change our findings.
Despite significantly higher IQ scores in the normal group, these were not “supernormal” children, as evident from the previously described similar relationship between IQ and frontotemporal volumes in other normal children (Reiss et al., 1996). Because we controlled for the effects of IQ, the IQ differences between the CAE and subjects without epilepsy did not account for the significantly smaller CAE volumes. Yet, we cannot rule out an association of the volume findings with psychopathology, given the heterogeneity of the epilepsy group in terms of different psychiatric diagnoses and the exclusion of control subjects with a psychiatric diagnosis.
These sample-related limitations underscore the need to replicate our findings on larger CAE samples. As for the study’s data analyses, although we computed multiple statistical tests, they were hypothesis driven, and we reported findings with a p-value below 0.05. Finally, given the study’s cross-sectional design, our developmental conclusions need to be confirmed by a prospective study.
In conclusion, frontotemporal structural abnormalities with volume reduction of the left orbital frontal gyrus and bilateral temporal lobe gray matter suggest that CAE impacts brain development. Localization of these abnormalities in brain regions implicated in behavior/emotions, cognition, and language, also suggest a possible biologic basis for the comorbidities of CAE. In addition to supporting the cortical focus theory of CAE, these findings provide further evidence that CAE is not a benign disorder.