Standardized Whole Brain Mapping of Tubers and Subepedymal
Standardized Whole Brain Mapping of Tubers and Subepedymal
Tuberous sclerosis complex is associated with radiologically visible abnormalities of brain structure, principally tubers and subependymal nodules. We reviewed the literature on neuroimaging of tubers and subependymal nodules and found qualitative evidence of bilateral, predominantly frontal distribution of tubers and bilateral, predominantly subcortical distribution of subependymal nodules in prior studies of pediatric samples. We studied 25 high-functioning adults with tuberous sclerosis complex and normal IQ, acquiring both dual spin-echo and fluid-attenuated inversion recovery magnetic resonance imaging sequences to optimize radiologic diagnosis of tubers and nodules. Individual lesion maps were then coregistered in a standard stereotactic space to facilitate construction of lesion density maps and estimation of lesion density in cortical and subcortical regions reliably defined by a parcellated template image. We found the highest frequency of tubers in frontal lobes and the highest density of tubers in parietal regions. There was significant regional variation in tuber density but no significant lateralization of frequently bilateral tubers. Nodules were located predominantly in the caudate nucleus and were not significantly lateralized. Tuber and nodule volumes were significantly positively correlated. Tuber volume was larger, on average, in patients with a lifetime history of epilepsy, but there was no correlation between IQ and these measures of lesion load. Contemporary image processing tools can be used to enhance quantitative, whole brain analysis of lesion load in patients with tuberous sclerosis complex.
Tuberous sclerosis complex is an autosomal dominant cortical dysgenesis syndrome for which two causative genes are known ( TSC1 and TSC2 ). The basic central nervous system pathology of tuberous sclerosis complex includes cortical and subcortical tubers, subependymal nodules, subependymal giant cell astrocytomas, heterotopic dysplastic neurons in white matter, and some ventricular enlargement, although, in most cases, overall brain size is normal. Possible neurologic outcomes of tuberous sclerosis complex include epilepsy, mental retardation, attention-deficit hyperactivity disorder (ADHD), and autism. The number and size of central nervous system lesions, like the behavioral and cognitive presentation, vary greatly for different individuals. Thus, tuberous sclerosis complex is an important genetic model to lead to wider understanding of the neuropathologic correlates of intelligence, ADHD, and autism.
Many studies have attempted to define the neuropathologic correlates of clinical phenotypes by comparing neuroradiologic "lesion load" across subjects. Despite the differing psychopathologic criteria for inclusion in such studies, we were interested in possible commonalities across subjects, perhaps indicating a typical lesion topography of tuberous sclerosis complex. Table 1 presents a summary of studies that reported information on lesion topography in subjects with tuberous sclerosis complex. To obtain the sample sizes needed for these mainly retrospective clinical studies, samples included a wide range of ages (mostly pediatric) and levels of intellectual function. We also note that some prior studies of lesion load, for example, Bolton et al. and Asano et al, are not tabulated here because the results were not reported in a form that allowed direct comparison with the rest of the literature.
Another immediate observation from this summary of previous reports of lesion topography in tuberous sclerosis complex was that a variety of imaging modalities had been used. In adults and older children with tuberous sclerosis complex, fluid-attenuated inversion recovery sequences are currently the most sensitive way to identify tubers ; however, only two studies in the current literature used fluid-attenuated inversion recovery for tuber identification. Although T2-weighted magnetic resonance images (MRIs) identify many intermediate-sized tubers, the tubers are usually identified by a signal change in the subjacent white matter, and on T2-weighted images, the gray matter of tubers is rarely highlighted. Fluid-attenuated inversion recovery images show both the gray and the white matter of a tuber very clearly, which seemed essential when attempting a volumetric analysis of tuber topography.
Despite this heterogeneity in terms of sample characteristics and imaging methodology, several important findings were consistent across the studies summarized in Table 1 First, all three classic, neuroradiologic lesions of tuberous sclerosis complex (tubers, subependymal nodules, and white-matter abnormalities) occurred in approximately 90% or more of the patients studied. Second, when a subject presented with any lesion, there would be more than one lesion in a majority of cases, and the brain would often be affected bilaterally (65-92% of cases). Third, there was a considerable variance in the number of lesions reported in different studies and different individuals (see Table 1 ). This likely reflected not only clinical heterogeneity, for example, owing to variable disease severity, but also differences in both scanning sensitivity (owing to scanner field strength, slice thickness, and MRI protocol) and lesion identification protocols.
To measure the lobar distribution of tubers, studies have, on the whole, employed two reporting styles. The first method reported the percentage of subjects with a tuber affecting each lobe, which provides an estimate of the frequency of lobar involvement by tubers. The second method reported the percentage of the group's tubers present in each lobe, which provided an estimate of tuber density. Tubers occurred most frequently in the frontal (43-100% frequency) and parietal lobes (60-100% frequency); they were less common and had more variable incidence in the occipital (36-80% frequency) and temporal (18-80% frequency) lobes. Tubers were least likely to be present in the cerebellum (12-44% frequency), and there were no reports in the literature of cerebellar tubers in the absence of cortical tubers. Whether tubers were bilateral at the lobar level did not appear to be dependent on their location and was variable for different studies (0-57% of tubers affected lobes bilaterally; see Table 1 for details). Several studies reported the percentage of the group's tubers identified in each lobe, and all of these studies produced a similar profile. The highest density of tubers occurred in the frontal lobe (34-58% of all tubers); then, in order of tuber density, came the parietal, occipital, and temporal lobes; finally, the lowest density of tubers was in the cerebellum (only 1-5% of total tubers affected the cerebellum). Previously, no study had considered whether lobe size or lesion volume affected the lobar density of lesions in tuberous sclerosis complex.
Baron and Barkovich explored correlations between the different types of lesions in tuberous sclerosis complex in a pediatric sample of seven patients (all aged less than 3 months). A positive correlation was found between the number of subependymal nodules, white-matter abnormalities, and tubers. Baron and Barkovich also observed that lesion size can be correlated for the three types of lesions; however, this was not formally tested owing to the small size of the sample.
When we focused solely on the results from studies using fluid-attenuated inversion recovery images, it was interesting to note that the percentage of subjects with a lesion in any lobe was typically somewhat higher than for studies using conventional MRI sequences. Walz and colleagues found no significant relationship between tuber location and autism and noted the high incidence of lesions in each lobe for all individuals. They called for more sophisticated measures of lesion load and more careful assessment of the relationship between neuroradiologic lesions and other neuroimaging markers of tuberous sclerosis complex. Two studies have considered whole brain lesion load and compared it with other neuroimaging variables. The first of these studies used positron emission tomography and MRI to analyze the relationship of autism and epilepsy to structural and functional brain abnormalities in children with tuberous sclerosis complex. There was no significant association between lesion topography and autism; however, there was evidence of functional imbalance in subcortical circuits (including areas without lesions), which could be linked to the severity of autistic symptom scores in tuberous sclerosis complex. The second study concluded that irrespective of tuber number, temporal lobe lesions played a crucial role in the pathogenesis of autistic behavior of some children with tuberous sclerosis complex. Seri and colleagues found that disruption in auditory sensory processing and transient auditory memory storage could play an important role in the pathogenesis of autistic behavior in tuberous sclerosis complex and commented that tubers might constitute a hallmark of a disturbance of cerebral function that is more extensive than suggested by morphologic imaging. As Walz and colleagues suggested, more refined methods of morphometry and/or the combination of multiple imaging modalities might help to resolve inconsistencies in the literature concerning the neurophysiologic correlates of autism in tuberous sclerosis complex.
However, we note also that there might be relevant variation in the instruments used to diagnose autism. For example, both the Walz et al. and Asano et al. studies focused their attention on the presence or absence of autism (diagnosed clinically or with the aid of the Autism Diagnostic Interview), whereas the study by Bolton et al. used both the Autism Diagnostic Interview and the Autism Diagnostic Observation Schedule to assess and diagnose the behavioral syndrome and examined tuber distribution according to the presence or absence of an autism spectrum disorder, which encompasses a broader range of impairments.
Having thus reviewed the literature on the topography of lesions in tuberous sclerosis complex, it seemed that more sensitive and standardized methods of lesion load measurement might be useful to provide a more comprehensive profile of neuropathologic abnormality and to address several questions about lesion topography in tuberous sclerosis complex: (1) Is the reportedly higher frequency of tubers in the frontal lobe just a reflection of the greater size of frontal lobes, that is, what is the lobar distribution of tubers when lobar volumes are appropriately controlled? (2) Is the anatomic distribution of tubers among sublobar brain regions significantly nonrandom? and (3) Is there any evidence of significant lateralization of lesion load in tuberous sclerosis complex?
Tuberous sclerosis complex is associated with radiologically visible abnormalities of brain structure, principally tubers and subependymal nodules. We reviewed the literature on neuroimaging of tubers and subependymal nodules and found qualitative evidence of bilateral, predominantly frontal distribution of tubers and bilateral, predominantly subcortical distribution of subependymal nodules in prior studies of pediatric samples. We studied 25 high-functioning adults with tuberous sclerosis complex and normal IQ, acquiring both dual spin-echo and fluid-attenuated inversion recovery magnetic resonance imaging sequences to optimize radiologic diagnosis of tubers and nodules. Individual lesion maps were then coregistered in a standard stereotactic space to facilitate construction of lesion density maps and estimation of lesion density in cortical and subcortical regions reliably defined by a parcellated template image. We found the highest frequency of tubers in frontal lobes and the highest density of tubers in parietal regions. There was significant regional variation in tuber density but no significant lateralization of frequently bilateral tubers. Nodules were located predominantly in the caudate nucleus and were not significantly lateralized. Tuber and nodule volumes were significantly positively correlated. Tuber volume was larger, on average, in patients with a lifetime history of epilepsy, but there was no correlation between IQ and these measures of lesion load. Contemporary image processing tools can be used to enhance quantitative, whole brain analysis of lesion load in patients with tuberous sclerosis complex.
Tuberous sclerosis complex is an autosomal dominant cortical dysgenesis syndrome for which two causative genes are known ( TSC1 and TSC2 ). The basic central nervous system pathology of tuberous sclerosis complex includes cortical and subcortical tubers, subependymal nodules, subependymal giant cell astrocytomas, heterotopic dysplastic neurons in white matter, and some ventricular enlargement, although, in most cases, overall brain size is normal. Possible neurologic outcomes of tuberous sclerosis complex include epilepsy, mental retardation, attention-deficit hyperactivity disorder (ADHD), and autism. The number and size of central nervous system lesions, like the behavioral and cognitive presentation, vary greatly for different individuals. Thus, tuberous sclerosis complex is an important genetic model to lead to wider understanding of the neuropathologic correlates of intelligence, ADHD, and autism.
Many studies have attempted to define the neuropathologic correlates of clinical phenotypes by comparing neuroradiologic "lesion load" across subjects. Despite the differing psychopathologic criteria for inclusion in such studies, we were interested in possible commonalities across subjects, perhaps indicating a typical lesion topography of tuberous sclerosis complex. Table 1 presents a summary of studies that reported information on lesion topography in subjects with tuberous sclerosis complex. To obtain the sample sizes needed for these mainly retrospective clinical studies, samples included a wide range of ages (mostly pediatric) and levels of intellectual function. We also note that some prior studies of lesion load, for example, Bolton et al. and Asano et al, are not tabulated here because the results were not reported in a form that allowed direct comparison with the rest of the literature.
Another immediate observation from this summary of previous reports of lesion topography in tuberous sclerosis complex was that a variety of imaging modalities had been used. In adults and older children with tuberous sclerosis complex, fluid-attenuated inversion recovery sequences are currently the most sensitive way to identify tubers ; however, only two studies in the current literature used fluid-attenuated inversion recovery for tuber identification. Although T2-weighted magnetic resonance images (MRIs) identify many intermediate-sized tubers, the tubers are usually identified by a signal change in the subjacent white matter, and on T2-weighted images, the gray matter of tubers is rarely highlighted. Fluid-attenuated inversion recovery images show both the gray and the white matter of a tuber very clearly, which seemed essential when attempting a volumetric analysis of tuber topography.
Despite this heterogeneity in terms of sample characteristics and imaging methodology, several important findings were consistent across the studies summarized in Table 1 First, all three classic, neuroradiologic lesions of tuberous sclerosis complex (tubers, subependymal nodules, and white-matter abnormalities) occurred in approximately 90% or more of the patients studied. Second, when a subject presented with any lesion, there would be more than one lesion in a majority of cases, and the brain would often be affected bilaterally (65-92% of cases). Third, there was a considerable variance in the number of lesions reported in different studies and different individuals (see Table 1 ). This likely reflected not only clinical heterogeneity, for example, owing to variable disease severity, but also differences in both scanning sensitivity (owing to scanner field strength, slice thickness, and MRI protocol) and lesion identification protocols.
To measure the lobar distribution of tubers, studies have, on the whole, employed two reporting styles. The first method reported the percentage of subjects with a tuber affecting each lobe, which provides an estimate of the frequency of lobar involvement by tubers. The second method reported the percentage of the group's tubers present in each lobe, which provided an estimate of tuber density. Tubers occurred most frequently in the frontal (43-100% frequency) and parietal lobes (60-100% frequency); they were less common and had more variable incidence in the occipital (36-80% frequency) and temporal (18-80% frequency) lobes. Tubers were least likely to be present in the cerebellum (12-44% frequency), and there were no reports in the literature of cerebellar tubers in the absence of cortical tubers. Whether tubers were bilateral at the lobar level did not appear to be dependent on their location and was variable for different studies (0-57% of tubers affected lobes bilaterally; see Table 1 for details). Several studies reported the percentage of the group's tubers identified in each lobe, and all of these studies produced a similar profile. The highest density of tubers occurred in the frontal lobe (34-58% of all tubers); then, in order of tuber density, came the parietal, occipital, and temporal lobes; finally, the lowest density of tubers was in the cerebellum (only 1-5% of total tubers affected the cerebellum). Previously, no study had considered whether lobe size or lesion volume affected the lobar density of lesions in tuberous sclerosis complex.
Baron and Barkovich explored correlations between the different types of lesions in tuberous sclerosis complex in a pediatric sample of seven patients (all aged less than 3 months). A positive correlation was found between the number of subependymal nodules, white-matter abnormalities, and tubers. Baron and Barkovich also observed that lesion size can be correlated for the three types of lesions; however, this was not formally tested owing to the small size of the sample.
When we focused solely on the results from studies using fluid-attenuated inversion recovery images, it was interesting to note that the percentage of subjects with a lesion in any lobe was typically somewhat higher than for studies using conventional MRI sequences. Walz and colleagues found no significant relationship between tuber location and autism and noted the high incidence of lesions in each lobe for all individuals. They called for more sophisticated measures of lesion load and more careful assessment of the relationship between neuroradiologic lesions and other neuroimaging markers of tuberous sclerosis complex. Two studies have considered whole brain lesion load and compared it with other neuroimaging variables. The first of these studies used positron emission tomography and MRI to analyze the relationship of autism and epilepsy to structural and functional brain abnormalities in children with tuberous sclerosis complex. There was no significant association between lesion topography and autism; however, there was evidence of functional imbalance in subcortical circuits (including areas without lesions), which could be linked to the severity of autistic symptom scores in tuberous sclerosis complex. The second study concluded that irrespective of tuber number, temporal lobe lesions played a crucial role in the pathogenesis of autistic behavior of some children with tuberous sclerosis complex. Seri and colleagues found that disruption in auditory sensory processing and transient auditory memory storage could play an important role in the pathogenesis of autistic behavior in tuberous sclerosis complex and commented that tubers might constitute a hallmark of a disturbance of cerebral function that is more extensive than suggested by morphologic imaging. As Walz and colleagues suggested, more refined methods of morphometry and/or the combination of multiple imaging modalities might help to resolve inconsistencies in the literature concerning the neurophysiologic correlates of autism in tuberous sclerosis complex.
However, we note also that there might be relevant variation in the instruments used to diagnose autism. For example, both the Walz et al. and Asano et al. studies focused their attention on the presence or absence of autism (diagnosed clinically or with the aid of the Autism Diagnostic Interview), whereas the study by Bolton et al. used both the Autism Diagnostic Interview and the Autism Diagnostic Observation Schedule to assess and diagnose the behavioral syndrome and examined tuber distribution according to the presence or absence of an autism spectrum disorder, which encompasses a broader range of impairments.
Having thus reviewed the literature on the topography of lesions in tuberous sclerosis complex, it seemed that more sensitive and standardized methods of lesion load measurement might be useful to provide a more comprehensive profile of neuropathologic abnormality and to address several questions about lesion topography in tuberous sclerosis complex: (1) Is the reportedly higher frequency of tubers in the frontal lobe just a reflection of the greater size of frontal lobes, that is, what is the lobar distribution of tubers when lobar volumes are appropriately controlled? (2) Is the anatomic distribution of tubers among sublobar brain regions significantly nonrandom? and (3) Is there any evidence of significant lateralization of lesion load in tuberous sclerosis complex?
Source...