| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, University of California at San Francisco, 505
Parnassus Ave., L358, San Francisco, CA 94143-0628.
2 Department of Neurology, University of California at San Francisco, San
Francisco, CA.
3 University of California at San Francisco, School of Medicine, San Francisco,
CA.
Received January 28, 2005;
accepted after revision October 10, 2005.
ARRS 2005 Executive Council Award Winner
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Brain MR images of adult and pediatric patients from our institution and from a national support organization (the ACC Network) were retrospectively evaluated for the type and severity of commissural anomalies and the presence and type of other structural abnormalities.
RESULTS. Of 142 cases that were reviewed, 82 patients had agenesis of the corpus callosum (ACC), while 60 had hypogenesis of the corpus callosum (HCC). Of the overall cohort, almost all had reduced white matter volume outside the commissures, the majority had malformations of cortical development (most commonly heterotopia or abnormal sulcation), many had noncallosal midline anomalies (including abnormal anterior or hippocampal commissures and interhemispheric cysts and lipomas), and several patients had abnormalities of the cerebellum or brainstem. Sixty-six patients had Probst bundles, which were more common in patients with ACC than in those with HCC. Probst bundles were present in all four patients who had ACC or HCC but no other midline, cortical, or posterior fossa anomalies.
CONCLUSION. Isolated commissural anomalies were rare in the populations of patients examined. Most cases of ACC and HCC were associated with complex telencephalic, diencephalic, or rhombencephalic malformations. Reduced cerebral hemispheric white matter volume and malformations of cortical development were seen in more than half of the patients, suggesting that many commissural anomalies are part of an overall cerebral dysgenesis. ACC and HCC appear to lie along a dysgenetic spectrum, as opposed to representing distinct disorders.
Keywords: abnormality • anomaly • brain • corpus callosum • MRI
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The corpus callosum is the principal supratentorial cerebral commissure (interhemispheric white matter tract). It has traditionally been separated into four segments: the rostrum, genu, body, and splenium (Figs. 1A, 1B, and 1C). Two smaller interhemispheric connections are located near the corpus callosum: the anterior commissure, which runs between the amygdaloid nuclei, and the hippocampal commissure or psalterium, which connects the fornices and blends with the ventral aspect of the callosal splenium. Prior work has shown that anomalies of the corpus callosum, including agenesis (absence) and hypogenesis (presumed later-formed segments missing but earlier-formed segments present, sometimes referred to as "partial agenesis of the corpus callosum"), may be associated with a variety of other brain malformations [1, 2, 5, 8-10].
|
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Images from the 82 remaining MRI studies were evaluated by two
neuroradiologists and a pediatric neurologist. Further exclusion criteria
included inadequate scan quality; commissural dysmorphisms interpreted as
being secondary, such as those from congenital hydrocephalus, Chiari II
malformations (most which were associated with open neural tube defects), or
large destructive injuries to the brain; holoprosencephaly, in which
commissural anomalies are believed to result from incomplete induction of
associated regions of brain; and those that are part of known multiple
congenital anomaly (MCA) syndromes (e.g., Aicardi's syndrome) for which
underlying genetic causes have been identified previously
[11,
12]. These excluded patients
(except those with inadequate scans) are considered as distinct groups and
will be evaluated in separate studies. Additional reasons for exclusion were
identification of a normal corpus callosum on MRI and presence of a normal
corpus callosum other than a nonvisualized rostrum, which can indicate a true
absence or merely an inability to detect that small structure on an MR image
obtained with a sagittal slice thickness of 
4 mm.
The scans of included patients were evaluated for type and severity of commissural anomalies; presence and type of interhemispheric cyst; presence and type of malformations of cortical development; distortions of the cerebral ventricles; anomalies of white matter (normal or reduced volume, location of reduced volume, and state of myelination); presence or absence of Probst bundles (collections of axons running anteroposteriorly lateral to the lateral ventricles, thought to represent axons that would have crossed the midline if a callosum were present); presence of microcephaly; and anomalies of the cerebellum, brainstem, orbits, and olfactory apparatus. With regard to qualitative white matter volume, the three expert observers subjectively compared the size of the centrum semiovale to the overall size of the brain and reached a consensus as to whether the volume of white matter was qualitatively normal, increased, or mild, moderately, or severely decreased as compared with age-matched control subjects (from our institution's database of normal brain MRI examinations) who were imaged for reasons other than developmental delay. The cerebellar size and folial pattern and the size and location of the fourth ventricle were similarly qualitatively assessed to determine whether foliation was normal and whether segmental or overall hypoplasia was evident. In the category labeled abnormalities of sulcation were included cortical malformations, such as pachygyria and polymicrogyria, as well as oversulcation (defined as the presence of too many sulci in the absence of identification of the sulci normally present in that region). Similarly, other qualitative and quantitative anomalies were determined by consensus of the three observers. A subset of scans from our hospital was also reviewed by a neuroradiologist blinded to the first consensus interpretation. The findings of this additional evaluation were compared with the consensus findings in order to calculate kappa statistics for interobserver consistency (MacKappa [computer software], Watkins MW).
Eighty-seven additional cases were obtained from an ongoing prospective cohort of individuals with anomalies of the corpus callosum. Families learned of the study through the National Organization of the Corpus Callosum and the ACC Network, national family support organizations for individuals with ACC. Membership in these organizations is based on self-referral by patients or the families of patients who have been diagnosed with agenesis or abnormalities of the corpus callosum. Seizures and developmental delay are the most common reasons these patients initially came to clinical attention. MRI examinations of patients in the ACC Network were subjected to the same exclusion criteria and analyses as outlined for patients from our institution, leaving 85 for inclusion. ACC Network patients had MRI examinations of highly variable quality, at times limiting their evaluation.
|
|
|
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The patients with ACC and those with HCC were compared. The mean ages (± SD) of ACC and HCC patients were different (p = 0.04), being 4.0 ± 7.1 years (range, 1 day-39 years) and 7.8 ± 13.7 years (range, 2 days-68 years), respectively. Forty-seven percent of ACC patients and 37% of HCC patients were female (p = 0.78). ACC patients constituted a lesser proportion of patients (28%) from our tertiary care hospital cohort than from the ACC Network cohort (57%; p < 0.001).
Interhemispheric cysts were similar in frequency between the two groups (17% of ACC patients and 10% of HCC patients; p = 0.23), as were interhemispheric lipomas (2% in both ACC and HCC patients). Abnormalities of the anterior commissure (71% ACC, 67% HCC; p = 0.65) and hippocampal commissure (95% ACC, 88% HCC; p = 0.1) were similar in frequency between the two groups. The presence of any cortical malformation (51% ACC, 55% HCC; p = 0.65) and of gray matter heterotopia (29% ACC, 21% HCC; p = 0.28) was not significantly different in frequency between ACC and HCC patients.
Abnormal cerebral ventricles were more frequently present in ACC patients (96%) than in HCC patients (83%, p < 0.01). Abnormalities of the orbits (6% ACC, 13% HCC; p =0.14), pituitary (13% ACC, 19% HCC; p =0.31), state of white matter myelination (25% ACC, 22% HCC; p = 0.62), cerebellar hemispheres (18% both ACC and HCC), cerebellar vermis (30% ACC, 33% HCC; p = 0.71), and brainstem (22% ACC, 27% HCC; p =0.51) were similar in frequency between the ACC and HCC patients. Abnormalities of the olfactory sulci (33% ACC, 15% HCC; p =0.03) were more frequent in ACC patients. Reductions in extracallosal white matter volume (98% ACC, 92% HCC; p = 0.1) and the presence of moderately or severely reduced extracallosal white matter volume (69% ACC, 58% HCC; p = 0.16) differed for the two groups, but that difference did not reach statistical significance. The most statistically marked difference between ACC and HCC patients was the frequency of lateral callosal bundles of Probst (59% ACC, 30% HCC; p <0.001).
Comparison was also made between patients examined at our tertiary referral university hospital with those who underwent MRI examinations at outside hospitals and were enrolled in our prospective study through the ACC Network. These groups differed both demographically and anatomically. The mean age (± SD) of patients examined at our hospital was 8.3 ± 15.2 years (range, 2 days-68 years) versus 4.3 ± 3.6 years (range, 1 day-19 years) for patients examined elsewhere and enrolled through the ACC Network (p < 0.01). Sixty-nine percent of patients referred through the ACC Network had absence of the corpus callosum versus 40% of our patients (p < 0.001); the remaining patients in both groups had callosal hypogenesis. The ACC Network prospective cohort patients were more likely to be young children than those examined at our university hospital.
Our hospital's patients were more likely to have abnormal anterior commissures (78% tertiary hospital, 63% ACC Network; p = 0.05), cortical malformations (64% tertiary hospital, 45% ACC Network; p = 0.027), abnormal ventricles (82% tertiary hospital, 95% ACC Network; p = 0.013), moderately to markedly reduced white matter volume (76% tertiary hospital, 56% ACC Network; p = 0.016), abnormal (delayed or incomplete) myelination (32% tertiary hospital, 17% ACC Network; p = 0.024), and anomalies of the cerebellar vermis (42% tertiary hospital, 24% ACC Network; p = 0.027) and brainstem (34% tertiary hospital, 17% ACC Network; p = 0.023). Probst bundles, however, were far more common among ACC Network prospective cohort patients (66%) than among patients from our hospital (18%, p < 0.001).
ACC and HCC patients were also categorized by the presence of Probst bundles, midline anomalies, malformations of cortical development, and other anomalies (primarily rhombencephalic and diencephalic anomalies and known multiple congenital anomaly syndromes, such as Aicardi's syndrome). Callosal agenesis and hypogenesis were usually accompanied by other abnormalities. Only three patients had isolated ACC without other anomalies, and only one patient had isolated HCC without other anomalies; all four of these patients had Probst bundles. Midline anomalies (primarily cysts, lipomas, anomalies of the anterior or hippocampal commissures) were the most common supratentorial abnormalities. Whereas the majority (22/25) of patients with ACC and other midline anomalies had Probst bundles, the majority of patients (13/21) with HCC and other midline anomalies did not have Probst bundles. Patients with rhombencephalic or diencephalic abnormalities (14/61) were less likely to have identifiable organized bundles of Probst than patients without such anomalies (53/81).
A subset of scans from our university hospital was also reviewed by a neuroradiologist blinded to the first consensus interpretation. Comparison of this observer's findings with the consensus results showed a high degree of interobserver consistency, with kappa statistics for interobserver agreement ranging from 0.89 to 1.0 for examined parameters. The kappa statistic for agreement was 0.89 when categorizing extracallosal supratentorial white matter volume as normal versus reduced to any degree. Similarly, when white matter was categorized into two groups—normal or mildly reduced versus moderately or markedly reduced—the kappa statistic for agreement remained 0.89.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
An abnormality or the absence of cerebral commissures was almost always accompanied by other brain malformations. The type, number, and severity of associated anomalies, however, differed between the retrospective cohort from our university hospital and the prospective ACC Network cohort. These differences underscore the selection biases inherent in drawing patients from a tertiary referral center and from a support group for individuals or families of individuals with a specific anomaly. Our university patients tended to have more complex and extensive anomalies, but fewer had ACC and fewer had identifiable Probst bundles; neonates with multiple congenital anomalies were more prevalent in our university cohort than in the ACC Network cohort. The relative lack of such patients in the ACC Network may reflect early mortality (patients not surviving long enough for their parents to join a support group) or that patients with multiple anomalies may be categorized principally by a diagnosis other than their callosal anomaly (e.g., parents of patients with lissencephaly associated with callosal hypogenesis might seek association with a support group for lissencephaly instead of the ACC Network). Despite the relative lack of neonates in the ACC Network cohort, on average patients were younger in this group than in our university's cohort. To our knowledge, this is the first documentation in the literature that cohorts of patients examined for the presence of a specific congenital malformation might differ only because of the source of referral.
Beyond the difference between patients from the ACC Network and our university hospital, there is also an inherent selection bias toward patients with clinical symptoms. The relative dearth of patients with isolated ACC or HCC may reflect either rare prevalence of isolated callosal anomalies or that people with isolated callosal anomalies do not have clinical symptoms that bring them to the attention of physicians. Neuropsychologic testing of patients with callosal agenesis has previously revealed subtle deficits of higher cognitive function not evident on routine clinical examinations [20, 21]. Another study reported that the only individuals with normal neurodevelopment were those with isolated callosal hypogenesis [22]. Only a large population-based MRI screening study could sort this out, and that task is beyond the scope of our current study.
An abnormality or the absence of cerebral commissures was rarely isolated in our study. Barkovich and Norman [2] examined 68 adult and pediatric patients with brain anomalies using MRI. Of these, 32 had ACC or HCC, but only one had isolated ACC, similar to the current study. Byrd et al. [9] later performed a retrospective and prospective analysis of 105 children with ACC in which all participants were examined on X-ray CT, 60 also on MRI, and 15 also on sonography. The 105 patients were sorted into seven nonexclusive groups (10 patients were counted twice), including 26 with isolated ACC. Of note, however, in the "isolated" ACC group, eight of 26 had Aicardi's syndrome, which always includes heterotopia and polymicrogyria among other characteristic features [23-25]. The inability to detect associated brain anomalies in these patients likely reflects the known lack of sensitivity of CT and sonography in detecting malformations of cortical development [7]. This limitation is compounded by a lack of sensitivity and specificity of CT in the analysis of the corpus callosum itself.
Midline anomalies, including interhemispheric cysts and lipomas, were present in several patients in this study and have been noted on prior studies of patients with commissural anomalies. Byrd et al. [9] found that 35 of 105 children with callosal agenesis had interhemispheric cysts, of which 31 communicated with the ventricles (type 1) and four did not communicate (type 2). In our study, 20 patients (14%) with ACC had interhemispheric cysts, of which 11 communicated with the ventricles and nine did not. Approximately 3% of the patients in the current study and in the prior study of Byrd et al. [9] had interhemispheric lipomas.
Malformations in cortical development are frequently associated with commissural anomalies. Gray matter heterotopia was the most common malformation of cortical development in our study, present in 29% of subjects with ACC and 21% of those with HCC. This compares with two of 68 patients with heterotopia in the study of Barkovich and Norman [2] and 18 of 105 patients with "migration disorder" in the study of Byrd et al. [9]. In the current study, 17 patients had polymicrogyria and 35 had abnormal sulcation, with similar frequencies in the ACC and HCC cohorts (Table 1).
|
Classic lissencephaly, cobblestone lissencephalies, polymicrogyria, schizencephaly, and heterotopia have all been described previously in conjunction with anomalies of the cerebral commissures. The current study documents the frequency of such associated anomalies and reveals that many patients with ACC or HCC have abnormal gyral patterns that do not fit into one of the classic categories (30 of 35 with abnormal sulcation, as described earlier and in Table 1). Beyond the expected eversion of the cingulum and radial orientation of paramedian gyri that routinely accompany callosal agenesis, abnormalities of sulcation ranged from overly shallow olfactory sulci to frank hemispheric dysplasia. We speculate that abnormal sulcation may reflect a more generalized developmental disorder of the cerebral white matter because white matter has been postulated to contribute to normal sulcation and, by definition, commissural anomalies are abnormalities of white matter [26]. The abnormality of sulcation may possibly be associated with the decreased volume of white matter so commonly found in our subjects.
Given that the corpus callosum is the principal transverse white matter tract connecting the cerebral hemispheres, it is not surprising that reduced extracallosal supratentorial white matter volume was identified in nearly all patients. In fact, only one of 81 patients with ACC and five of 59 patients with HCC were found to have normal white matter volume. Reductions in extracallosal white matter may represent a primary dysplasia or hypogenesis, with fewer axons forming during development, or a secondary regression, possibly due to retraction of axons that do not find their way across midline to synapse with their homologues and thereby gain the neurotrophic support necessary for survival [27]. Alternatively, the number of axons in white matter tracts that do not cross midline could be similar, but have relatively less myelin than is seen in normal commissures. The normalappearing signal intensity of the hemispheric white matter in our patients makes this possibility unlikely. Diffusion tensor MRI is an emerging tool for mapping white matter tracts in vivo [28]; in the next phase of this project, patients with commissural anomalies will undergo diffusion tensor MR tractography to better assess the presence, location, and organization of white matter tracts.
The lateral callosal bundles of Probst indent the superomedial aspect of the lateral ventricles and are postulated to represent the axons that would have crossed the midline [29, 30]. Our study shows that Probst bundles are about twice as common in patients with ACC (59%) as in patients with HCC (30%). This may reflect a genuine biologic difference or may simply be a result of the technique used to analyze the brains: The human eye can more easily detect a long, wide, continuous, well-organized Probst bundle running the entire length of the lateral ventricle (ACC) than a narrower tract running only a short distance (HCC). Byrd et al. [9] also commented on 26 children with ostensibly isolated ACC, remarking that "in children with true ACC, Probst bundles are present." Whereas Probst bundles were uncommon in patients with anomalies beyond the midline and malformations of cortical development (principally malformations of the diencephalon and rhombencephalon), all three of the isolated ACC cases and one isolated HCC case had Probst bundles, perhaps portending a more organized brain overall. Further studies comparing MR appearance, including diffusion tensor tractography, with clinical outcome will address this important question.
Acknowledgments
We would like to acknowledge the help of Mari Wakahiro, study coordinator
for our institution's Brain Development Research Program. We also thank Gary
and Kathy Schilmoeller, founders of the ACC Network, for their invaluable
assistance in alerting ACC families to our research. We would also like to
thank Ying Lu for his help with statistical analysis.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
