(Radiology. 1999;213:400-406.)
© RSNA, 1999
Cree Leukoencephalopathy: Neuroimaging Findings1
Ibrahim A. Alorainy, MD,
Yves G. Patenaude, MD,
Augustin M. O'Gorman, MD,
Deborah N. Black, MD and
Kathleen Meagher-Villemure, MD
1 From the Department of Medical Imaging, Montreal Children's Hospital, Quebec, Canada (I.A.A., Y.G.P., A.M.O.); the Department of Neurology, Central Vermont Hospital, Montpelier (D.N.B.); and the Institut Universitaire de Pathologie, Lausanne, Switzerland (K.M.V.). Received November 9, 1998; revision requested December 29; revision received February 2, 1999; accepted April 29. Address reprint requests to I.A.A., Department of Radiology, College of Medicine, King Saud University, P.O. Box 9047, Riyadh 11413, Saudi Arabia.
 |
Abstract
|
|---|
PURPOSE: To describe the computed tomographic (CT) and magnetic resonance (MR) imaging findings in Cree leukoencephalopathy.
MATERIALS AND METHODS: The authors retrospectively reviewed the medical records and neuroimaging studies in 12 infants with Cree leukoencephalopathy (CT in 12 infants, MR in six). The diagnosis was established clinically in six patients and at autopsy in the other six.
RESULTS: At CT, extensive, diffuse, and symmetric hypoattenuation was seen in the cerebral and cerebellar white matter in all 12 patients. Hypoattenuation was also seen in the corpus callosum in 11 (92%), internal capsule in 10 (83%), globus pallidus in nine (75%), brainstem in nine (75%), and thalamus in four (33%). The caudate nucleus and putamen were spared. On T2-weighted MR images in six patients, the cerebral and cerebellar white matter, including the subcortical arcuate fibers, was hyperintense as were the internal capsule, corpus callosum, corticospinal tracts, and globus pallidus. The thalamus was affected in four (67%) patients, pons in five (83%), and medulla in four (33%). The caudate nucleus and putamen were not affected.
CONCLUSION: Cree leukoencephalopathy causes striking symmetric and diffuse involvement of the cerebral and cerebellar white matter and brainstem with sparing of the caudate nucleus and putamen.
Index terms: Brain, CT, 10.1211 • Brain, diseases, 10.872 • Brain, MR, 10.12141 • Infants, central nervous system, 10.872 • Infants, newborn, central nervous system, 10.872
|
Introduction
|
|---|
Cree leukoencephalopathy is a rare and fatal disease that is different from the currently known leukencephalopathies and has been described as occurring exclusively in northern Quebec and Manitoba, Canada (1). The clinical, laboratory, and pathologic findings have been documented in 14 patients as have computed tomographic (CT) findings in eight (1). Age at presentation is around 6 months. The symptoms are hypotonia or spasticity followed by seizures, eye deviation, and abnormal posturing. There is usually a history of preceding viral illness. Affected infants progress to a debilitated state and die within days to months from the onset of clinical symptoms. No biochemical, enzymatic, or electrophysiologic abnormalities have been identified. The pathologic examination shows diffuse attenuation of cerebral myelin with minimal sudanophilia, absence of intracellular storage products, and absence of inflammation (1).
To our knowledge, the magnetic resonance (MR) imaging findings in Cree leukoencephalopathy have not been described. The purpose of this study was to describe the CT and MR imaging features in this disease and compare them to those in other myelin disorders with similar imaging findings.
|
MATERIALS AND METHODS
|
|---|
Patients
From January 1989 through September 1997, 12 consecutive patients (five male and seven female infants; median age, 7 months; age range, 5–11 months) with subsequently proved Cree leukoencephalopathy were referred to our institution for imaging evaluation and further treatment. All patients were referred from Cree Indian villages on the east coast of James Bay in northern Quebec. The medical records were reviewed (I.A.A., D.N.B.) for age, sex, and family history and the clinical symptoms, progress, and outcome. The diagnosis was established by means of autopsy and histopathologic examination in six patients and clinical and laboratory findings in the other six patients. Autopsy was performed 12–60 hours after death. Among the six patients who underwent autopsy, CT and MR studies were available in three and only CT studies in three.
CT and MR Imaging Studies
Three neuroradiologists (I.A.A., Y.G.P., A.M.O.) retrospectively reviewed the neuroimaging studies by consensus for distribution and degree (absent, moderate, severe) of changes in attenuation and signal intensity on brain CT and MR images, respectively. Six patients underwent one CT and one MR imaging examination of the head during hospitalization. MR imaging was performed to further characterize the changes seen at CT and to look for more findings that might help explain the clinical symptoms. The other six patients underwent only CT, also once during their hospital stay. In this group of patients, MR imaging was not performed because either the clinical status of the patient did not allow it or the family decided, in view of a positive family history, to take the child home after they learned the CT and laboratory results.
Nonenhanced and contrast material–enhanced CT (CT Hi 9800 or CTI, GE Medical Systems, Milwaukee, Wis) were performed in three patients, only contrast-enhanced CT in one, and only nonenhanced CT in eight. At CT, sections were 10 mm thick and the table increment was 10 mm in 11 of the patients and were 7 and 7 mm, respectively, in one patient. Exposure factors were 100–140 mAs and 120 kVp.
MR imaging was performed on a 1.5-T system (Signa Advantage, GE Medical Systems) within 6 days after the CT examination. Nonenhanced T1-weighted images (repetition time msec/echo time msec = 400–500/10–14 with two signals acquired) were obtained in the sagittal and axial planes. Fast spin-echo T2-weighted imaging (3, 100–3,100–3,800/100–102; one or two signals acquired; echo train length, eight) was performed in the axial and coronal planes (only axial images were obtained in one patient). For both T1- and T2-weighted studies, section thickness was 4–5 mm with 1-mm intersection gap, and the matrix was 256 x 256 or 256 x 192. Gadolinium-enhanced MR imaging (gadopentetate dimeglumine, Magnevist; Berlex Laboratories, Wayne, NJ) was performed in only patient 5.
|
RESULTS
|
|---|
The clinical data and studies performed are summarized in Table 1. All patients, except one with mild developmental delay, were completely healthy before the onset of disease. Prior to development of neurologic symptoms, 10 patients had flulike illness or gastrointestinal symptoms, and one had otitis media. One patient presented with seizures without preceding symptoms. Positive family history of Cree leukoencephalopathy was present in nine of the 12 patients. Death occurred 12 days to 10 months after presentation (median survival, 17.5 days). Time between the onset of neurologic symptoms and CT was 2–7 days. Symmetric and bilateral abnormal findings were demonstrated at CT and MR imaging in all patients. The frequency and distribution of attenuation changes at CT and signal intensity abnormalities at T2-weighted imaging are listed in Tables 2 and 3, respectively.
View this table:
[in this window]
[in a new window]
|
TABLE 2. Distribution and Frequency of Low Attenuation in Different Brain Structures at CT in 12 Patients with Cree Leukoencephalopathy
|
|
View this table:
[in this window]
[in a new window]
|
TABLE 3. Distribution and Frequency of Abnormal High Signal Intensity on T2-weighted MR Images in Different Brain Structures in Six Patients with Cree Leukoencephalopathy
|
|
CT Findings
In all 12 patients, the lobar and cerebellar white matter was bilaterally and symmetrically hypoattenuating (Fig 1a, 1b) with notable sparing of the caudate nucleus and putamen. The body of the normal caudate nucleus appeared hyperattenuating relative to the adjacent abnormally hypoattenuating lobar white matter (Fig 1a). In the four patients with thalamic low attenuation, there was preferential involvement of the ventral lateral and ventral posterior thalamic nuclei (Fig 2a). The brainstem was difficult to assess owing to beam-hardening artifacts, but in nine patients definite involvement of brainstem white matter was present. At nonenhanced CT in all patients, calcification or hemorrhage was not seen. At contrast-enhanced CT (Hypaque-M, Sanofi Winthrop, Markham, Ontario, Canada, or Omnipaque, Nycomed, Brampton, Ontario) in four patients, no abnormal enhancement was seen.
View larger version (167K):
[in this window]
[in a new window]
|
Figure 1a. Patient 1.
Images obtained after seizures and rapid neurologic deterioration following upper respiratory tract infection. Cree leukoencephalopathy was confirmed at autopsy. (a) Nonenhanced axial CT scan at the level of the body of the caudate nucleus depicts diffuse and symmetric hypoattenuation in the white matter in both cerebral hemispheres. The body of the caudate nucleus (arrows) is unaffected and appears of higher attenuation than usual because of the lower attenuation of the adjacent white matter. (b) Nonenhanced axial CT scan at the level of the basal ganglia shows remarkable reduction in the attenuation of the globus pallidus (arrows). The hypoattenuation in the frontal white matter crosses through the genu of the corpus callosum (arrowheads). (c) Nonenhanced axial T1-weighted MR image (400/14) at the level of the midbrain shows normal red nuclei (arrows) that appear hyperintense due to substantial reduction in the signal intensity of the surrounding midbrain parenchyma. The white matter of the frontal and temporal lobes (*) is hypointense. (d) Axial T2-weighted MR image (3,500/102) at the same level as c shows diffuse increased signal intensity in the midbrain (open arrows). The red nuclei (black straight arrows) are unaffected and appear more hypointense than usual because of high signal intensity in the surrounding midbrain parenchyma. The white matter of the frontal and temporal lobes is diffusely hyperintense with involvement of the subcortical arcuate fibers (black curved arrows). (e) Axial T2-weighted MR image (3,500/102) demonstrates diffuse high signal intensity in the medulla oblongata (straight arrows) with sparing of the inferior olives (curved arrows).
|
|
View larger version (174K):
[in this window]
[in a new window]
|
Figure 1b. Patient 1.
Images obtained after seizures and rapid neurologic deterioration following upper respiratory tract infection. Cree leukoencephalopathy was confirmed at autopsy. (a) Nonenhanced axial CT scan at the level of the body of the caudate nucleus depicts diffuse and symmetric hypoattenuation in the white matter in both cerebral hemispheres. The body of the caudate nucleus (arrows) is unaffected and appears of higher attenuation than usual because of the lower attenuation of the adjacent white matter. (b) Nonenhanced axial CT scan at the level of the basal ganglia shows remarkable reduction in the attenuation of the globus pallidus (arrows). The hypoattenuation in the frontal white matter crosses through the genu of the corpus callosum (arrowheads). (c) Nonenhanced axial T1-weighted MR image (400/14) at the level of the midbrain shows normal red nuclei (arrows) that appear hyperintense due to substantial reduction in the signal intensity of the surrounding midbrain parenchyma. The white matter of the frontal and temporal lobes (*) is hypointense. (d) Axial T2-weighted MR image (3,500/102) at the same level as c shows diffuse increased signal intensity in the midbrain (open arrows). The red nuclei (black straight arrows) are unaffected and appear more hypointense than usual because of high signal intensity in the surrounding midbrain parenchyma. The white matter of the frontal and temporal lobes is diffusely hyperintense with involvement of the subcortical arcuate fibers (black curved arrows). (e) Axial T2-weighted MR image (3,500/102) demonstrates diffuse high signal intensity in the medulla oblongata (straight arrows) with sparing of the inferior olives (curved arrows).
|
|
View larger version (170K):
[in this window]
[in a new window]
|
Figure 1c. Patient 1.
Images obtained after seizures and rapid neurologic deterioration following upper respiratory tract infection. Cree leukoencephalopathy was confirmed at autopsy. (a) Nonenhanced axial CT scan at the level of the body of the caudate nucleus depicts diffuse and symmetric hypoattenuation in the white matter in both cerebral hemispheres. The body of the caudate nucleus (arrows) is unaffected and appears of higher attenuation than usual because of the lower attenuation of the adjacent white matter. (b) Nonenhanced axial CT scan at the level of the basal ganglia shows remarkable reduction in the attenuation of the globus pallidus (arrows). The hypoattenuation in the frontal white matter crosses through the genu of the corpus callosum (arrowheads). (c) Nonenhanced axial T1-weighted MR image (400/14) at the level of the midbrain shows normal red nuclei (arrows) that appear hyperintense due to substantial reduction in the signal intensity of the surrounding midbrain parenchyma. The white matter of the frontal and temporal lobes (*) is hypointense. (d) Axial T2-weighted MR image (3,500/102) at the same level as c shows diffuse increased signal intensity in the midbrain (open arrows). The red nuclei (black straight arrows) are unaffected and appear more hypointense than usual because of high signal intensity in the surrounding midbrain parenchyma. The white matter of the frontal and temporal lobes is diffusely hyperintense with involvement of the subcortical arcuate fibers (black curved arrows). (e) Axial T2-weighted MR image (3,500/102) demonstrates diffuse high signal intensity in the medulla oblongata (straight arrows) with sparing of the inferior olives (curved arrows).
|
|
View larger version (170K):
[in this window]
[in a new window]
|
Figure 1d. Patient 1.
Images obtained after seizures and rapid neurologic deterioration following upper respiratory tract infection. Cree leukoencephalopathy was confirmed at autopsy. (a) Nonenhanced axial CT scan at the level of the body of the caudate nucleus depicts diffuse and symmetric hypoattenuation in the white matter in both cerebral hemispheres. The body of the caudate nucleus (arrows) is unaffected and appears of higher attenuation than usual because of the lower attenuation of the adjacent white matter. (b) Nonenhanced axial CT scan at the level of the basal ganglia shows remarkable reduction in the attenuation of the globus pallidus (arrows). The hypoattenuation in the frontal white matter crosses through the genu of the corpus callosum (arrowheads). (c) Nonenhanced axial T1-weighted MR image (400/14) at the level of the midbrain shows normal red nuclei (arrows) that appear hyperintense due to substantial reduction in the signal intensity of the surrounding midbrain parenchyma. The white matter of the frontal and temporal lobes (*) is hypointense. (d) Axial T2-weighted MR image (3,500/102) at the same level as c shows diffuse increased signal intensity in the midbrain (open arrows). The red nuclei (black straight arrows) are unaffected and appear more hypointense than usual because of high signal intensity in the surrounding midbrain parenchyma. The white matter of the frontal and temporal lobes is diffusely hyperintense with involvement of the subcortical arcuate fibers (black curved arrows). (e) Axial T2-weighted MR image (3,500/102) demonstrates diffuse high signal intensity in the medulla oblongata (straight arrows) with sparing of the inferior olives (curved arrows).
|
|
View larger version (147K):
[in this window]
[in a new window]
|
Figure 1e. Patient 1.
Images obtained after seizures and rapid neurologic deterioration following upper respiratory tract infection. Cree leukoencephalopathy was confirmed at autopsy. (a) Nonenhanced axial CT scan at the level of the body of the caudate nucleus depicts diffuse and symmetric hypoattenuation in the white matter in both cerebral hemispheres. The body of the caudate nucleus (arrows) is unaffected and appears of higher attenuation than usual because of the lower attenuation of the adjacent white matter. (b) Nonenhanced axial CT scan at the level of the basal ganglia shows remarkable reduction in the attenuation of the globus pallidus (arrows). The hypoattenuation in the frontal white matter crosses through the genu of the corpus callosum (arrowheads). (c) Nonenhanced axial T1-weighted MR image (400/14) at the level of the midbrain shows normal red nuclei (arrows) that appear hyperintense due to substantial reduction in the signal intensity of the surrounding midbrain parenchyma. The white matter of the frontal and temporal lobes (*) is hypointense. (d) Axial T2-weighted MR image (3,500/102) at the same level as c shows diffuse increased signal intensity in the midbrain (open arrows). The red nuclei (black straight arrows) are unaffected and appear more hypointense than usual because of high signal intensity in the surrounding midbrain parenchyma. The white matter of the frontal and temporal lobes is diffusely hyperintense with involvement of the subcortical arcuate fibers (black curved arrows). (e) Axial T2-weighted MR image (3,500/102) demonstrates diffuse high signal intensity in the medulla oblongata (straight arrows) with sparing of the inferior olives (curved arrows).
|
|
View larger version (148K):
[in this window]
[in a new window]
|
Figure 2a. Patient 4.
Images obtained after flulike infection followed by abnormal eye movements and progressive neurologic deterioration. Clinical and laboratory findings were consistent with Cree leukoencephalopathy. (a) Nonenhanced axial CT scan shows triangular hypoattenuation in the region of the lateral ventral thalamic nucleus bilaterally (solid arrows). The internal capsules are also hypoattenuating, especially the posterior limbs (open arrows). (b) Axial T2-weighted MR image (3,200/102) at about the same level as a more clearly demonstrates the selective involvement of the lateral ventral thalamic nucleus (curved white arrows) on both sides. The globus pallidus (open arrows) and internal capsule, especially the posterior limbs (short white arrows), are also affected bilaterally, while the caudate nucleus (straight black arrows) and putamen (long white arrows) are unaffected. The lobar white matter including the subcortical arcuate fibers (curved black arrows) is markedly hyperintense. (c) Coronal T2-weighted MR image (3,200/102) shows high signal intensity along the course of the corticospinal tracts (white arrows) in the midbrain and pons. The corpus callosum (straight black arrows) is involved to the same degree as is the lobar white matter. The lateral aspects of the thalamus (curved black arrows) are also affected.
|
|
View larger version (156K):
[in this window]
[in a new window]
|
Figure 2b. Patient 4.
Images obtained after flulike infection followed by abnormal eye movements and progressive neurologic deterioration. Clinical and laboratory findings were consistent with Cree leukoencephalopathy. (a) Nonenhanced axial CT scan shows triangular hypoattenuation in the region of the lateral ventral thalamic nucleus bilaterally (solid arrows). The internal capsules are also hypoattenuating, especially the posterior limbs (open arrows). (b) Axial T2-weighted MR image (3,200/102) at about the same level as a more clearly demonstrates the selective involvement of the lateral ventral thalamic nucleus (curved white arrows) on both sides. The globus pallidus (open arrows) and internal capsule, especially the posterior limbs (short white arrows), are also affected bilaterally, while the caudate nucleus (straight black arrows) and putamen (long white arrows) are unaffected. The lobar white matter including the subcortical arcuate fibers (curved black arrows) is markedly hyperintense. (c) Coronal T2-weighted MR image (3,200/102) shows high signal intensity along the course of the corticospinal tracts (white arrows) in the midbrain and pons. The corpus callosum (straight black arrows) is involved to the same degree as is the lobar white matter. The lateral aspects of the thalamus (curved black arrows) are also affected.
|
|
View larger version (174K):
[in this window]
[in a new window]
|
Figure 2c. Patient 4.
Images obtained after flulike infection followed by abnormal eye movements and progressive neurologic deterioration. Clinical and laboratory findings were consistent with Cree leukoencephalopathy. (a) Nonenhanced axial CT scan shows triangular hypoattenuation in the region of the lateral ventral thalamic nucleus bilaterally (solid arrows). The internal capsules are also hypoattenuating, especially the posterior limbs (open arrows). (b) Axial T2-weighted MR image (3,200/102) at about the same level as a more clearly demonstrates the selective involvement of the lateral ventral thalamic nucleus (curved white arrows) on both sides. The globus pallidus (open arrows) and internal capsule, especially the posterior limbs (short white arrows), are also affected bilaterally, while the caudate nucleus (straight black arrows) and putamen (long white arrows) are unaffected. The lobar white matter including the subcortical arcuate fibers (curved black arrows) is markedly hyperintense. (c) Coronal T2-weighted MR image (3,200/102) shows high signal intensity along the course of the corticospinal tracts (white arrows) in the midbrain and pons. The corpus callosum (straight black arrows) is involved to the same degree as is the lobar white matter. The lateral aspects of the thalamus (curved black arrows) are also affected.
|
|
MR Imaging Findings
Among the six patients who underwent MR imaging, striking and diffuse high signal intensity was seen on T2-weighted images in the lobar white matter in a symmetric distribution in four patients (patients 1–4) and less severe signal intensity changes were seen in two patients (patients 5 and 6). Similarly, all patients had severe cerebellar white matter involvement. In all patients, the anterior and, more conspicuously, posterior limbs of the internal capsule showed bilateral high signal intensity on T2-weighted images (Fig 2b). The globus pallidus was also affected in all patients (Fig 2b), although involvement was moderate in three. As on the CT scans, the MR images depicted sparing of the caudate nucleus and putamen in all patients (Fig 2b).
High signal intensity was identified bilaterally along the course of the corticospinal tract in the midbrain, pons, and medulla oblongata in all patients (Figs 2c, 3a). The red nuclei were affected in two patients, both of whom had diffuse parenchymal changes in the midbrain. In the other two patients with midbrain changes, the red nuclei were normal. In patient 1, the red nuclei were normal, but midbrain parenchyma was severely affected to the extent that the red nuclei appeared hyperintense to the midbrain on T1-weighted images and hypointense to the midbrain on T2-weighted images (Fig 1c, 1d). Similar but less marked findings were seen in patient 6.
View larger version (181K):
[in this window]
[in a new window]
|
Figure 3a. Patient 5.
Images obtained after fever, ocular deviation, and irritability followed by progressive neurologic deterioration. The infant died 5 months after the onset of symptoms. Cree leukoencephalopathy was confirmed at autopsy. (a) Axial T2-weighted MR image (3,800/102) shows high signal intensity in the corticospinal tracts (black arrows), more on the left side with diffuse hyperintensity in the tegmentum (open arrows). The white matter hyperintensity in the temporal lobes extends to the subcortical arcuate fibers. (b) Nonenhanced left parasagittal T1-weighted MR image (450/10) depicts reversal of the signal intensity pattern in the pons, with the tegmentum more hypointense than the basis pontis. The hypointensity in the tegmentum extends superiorly into the midbrain and inferiorly into the medulla oblongata (solid arrows). Hypointensity is also seen along the course of the left corticospinal tract (open arrow).
|
|
View larger version (178K):
[in this window]
[in a new window]
|
Figure 3b. Patient 5.
Images obtained after fever, ocular deviation, and irritability followed by progressive neurologic deterioration. The infant died 5 months after the onset of symptoms. Cree leukoencephalopathy was confirmed at autopsy. (a) Axial T2-weighted MR image (3,800/102) shows high signal intensity in the corticospinal tracts (black arrows), more on the left side with diffuse hyperintensity in the tegmentum (open arrows). The white matter hyperintensity in the temporal lobes extends to the subcortical arcuate fibers. (b) Nonenhanced left parasagittal T1-weighted MR image (450/10) depicts reversal of the signal intensity pattern in the pons, with the tegmentum more hypointense than the basis pontis. The hypointensity in the tegmentum extends superiorly into the midbrain and inferiorly into the medulla oblongata (solid arrows). Hypointensity is also seen along the course of the left corticospinal tract (open arrow).
|
|
The pons was diffusely involved in five patients, in three of whom (patients 1, 5, and 6) the posterior aspect of the pons was more severely affected than was the anterior aspect (Fig 3). High signal intensity was demonstrated in the region of the medulla oblongata in four patients, but the inferior olives were spared in two of them (Fig 1e). In the same four patients, high signal intensity was seen in the thalamus. In two of them, involvement of the ventral lateral and ventral posterior thalamic nuclei was pronounced (Fig 2b, 2c). In the other two, thalamic involvement was moderate and patchy with no nuclear predilection. In general, T1-weighted images depicted low signal intensity at the sites of abnormally high signal intensity depicted on T2-weighted images. T1 shortening was not encountered.
In the patient who underwent gadolinium-enhanced MR imaging, abnormal parenchymal or meningeal enhancement was not seen. Because of the diffuse and extensive nature of the signal intensity changes in white matter, it was difficult to ascertain the degree of myelination. None of the patients had hydrocephalus or macrocephaly. Apart from frontoparietal atrophy in patients 1, 4, and 5, there was no cortical gray matter abnormality.
Autopsy and Histopathologic Findings
In six patients, neuropathologic examination revealed gross loss of myelin in the white matter of the cerebrum, cerebellum, and brainstem (Fig 4a, 4b). At histologic examination within the poorly myelinated regions, rare macrophages were seen that contained sudanophilic material and a severe astrogliosis (Fig 4c). There was no inflammation or calcified deposits. The meninges and cerebral cortex were unremarkable. The changes were variable from one patient to another in the spinal cord, with no myelin loss in the younger infants and poorly myelinated tracts in the older infants. The cervical and thoracic regions seemed more affected in these patients.
View larger version (122K):
[in this window]
[in a new window]
|
Figure 4a. The pathologic findings in Cree leukoencephalopathy. (a)Photograph of the gross specimen. Coronal section of the right hemisphere through the thalamus and hippocampus shows a translucent zone in the white matter (arrowhead) with relative sparing of the arcuate fibers (arrows). (b) Photomicrograph of the same region as in a shows a rarefied myelinated zone (arrow). (Hematoxylin-eosin stain; original magnification, x40.) (c) Photomicrograph shows severe loss of myelin sheaths with rare macrophages (arrowhead) and numerous reactive astrocytes (arrow). (Hematoxylin-eosin stain; original magnification, x400.)
|
|
View larger version (147K):
[in this window]
[in a new window]
|
Figure 4b. The pathologic findings in Cree leukoencephalopathy. (a) Photograph of the gross specimen. Coronal section of the right hemisphere through the thalamus and hippocampus shows a translucent zone in the white matter (arrowhead) with relative sparing of the arcuate fibers (arrows). (b) Photomicrograph of the same region as in a shows a rarefied myelinated zone (arrow). (Hematoxylin-eosin stain; original magnification, x40.) (c) Photomicrograph shows severe loss of myelin sheaths with rare macrophages (arrowhead) and numerous reactive astrocytes (arrow). (Hematoxylin-eosin stain; original magnification, x400.)
|
|
View larger version (138K):
[in this window]
[in a new window]
|
Figure 4c. The pathologic findings in Cree leukoencephalopathy. (a) Photograph of the gross specimen. Coronal section of the right hemisphere through the thalamus and hippocampus shows a translucent zone in the white matter (arrowhead) with relative sparing of the arcuate fibers (arrows). (b) Photomicrograph of the same region as in a shows a rarefied myelinated zone (arrow). (Hematoxylin-eosin stain; original magnification, x40.) (c) Photomicrograph shows severe loss of myelin sheaths with rare macrophages (arrowhead) and numerous reactive astrocytes (arrow). (Hematoxylin-eosin stain; original magnification, x400.)
|
|
|
DISCUSSION
|
|---|
Cree leukoencephalopathy differs from the known diseases of cerebral myelin. To our knowledge, this severe white matter disease has been described exclusively in Native American Indian infants in northern Quebec and Manitoba. Although most patients are Cree Indians, as the name denotes, Cree leukoencephalopathy has been reported in non–Cree Indians who have linguistic and genetic similarities to and possible intermarriage with Cree Indians (1). The disease probably has a genetic basis, possibly autosomal recessive, although an environmental agent or toxin has not been excluded (1). The true incidence and prevalence of the disease are not known.
Black et al (1) described in detail the clinical and laboratory findings in Cree leukoencephalopathy. The disease is 100% fatal, and the presentation is acute or subacute after a viral illness around the age of 6 months. Classically, there is absence of the enzymatic, biochemical, immunologic, and electrophysiologic abnormalities known in other leukoencephalopathies.
Cree encephalitis, another familial neurologic disorder described in children from the same geographic area, is thought to be due to viral meningoencephalitis (2). Both Cree leukoencephalopathy and Cree encephalitis occur in highly inbred and closely knit native communities, and both have familial background (genetic basis). However, Cree encephalitis, in addition to white matter disease, is associated with systemic immunologic abnormalities and intracranial calcifications and has a slower course and relatively better prognosis. Death in cases of Cree encephalitis is due primarily to immunodeficiency (2). To our knowledge, the MR imaging findings in Cree encephalitis have not yet been described.
Diffuse involvement of the white matter of the cerebral hemispheres and cerebellum with involvement of the globus pallidus and preservation of the caudate nucleus and putamen is highly suggestive of Canavan disease (3). On the other hand, the pattern of spread of white matter changes in Canavan disease is centripetal, starting in the subcortical arcuate fibers and ultimately involving the deeper white matter (3). This is the opposite of what we noticed in patients 5 and 6, in whom the subcortical white matter was less involved than was the more central white matter. Furthermore, the clinical, biochemical, and pathologic findings in our patients are different from those in patients with Canavan disease. The subcortical arcuate fibers are also prominently involved in Alexander disease, but there is usually a frontal predominance, tendency to form cysts, and ventricular enlargement (4). On the other hand, very advanced cases may not demonstrate frontal predilection, and the findings could be similar to those encountered in our patients. The clinical symptoms of the infantile form of Alexander disease could also be similar to that of Cree leukoencephalopathy except for lack of familial background in the former. Unlike findings at pathologic examination in patients with Cree leukoencephalopathy, pathologic findings in patients with Alexander disease reveal innumerable Rosenthal fibers throughout the brain and spinal cord.
The clinical symptoms of our patients are also incompatible with those for maple syrup urine disease. In the neonatal form of maple syrup urine disease, MR images depict selective involvement of the cerebellar white matter and the dorsal aspect of the brainstem with milder effect on lobar white matter (5). Among our five patients with brainstem involvement, preferential involvement of the dorsal aspect of the pons was seen in three and diffuse changes in the other two. The imaging findings in older patients with maple syrup urine disease are similar to those in patients with Canavan disease (6).
In our patients, the symmetry of white matter changes made acute disseminated encephalomyelitis improbable. Patients with severe acute disseminated encephalomyelitis, however, may lack asymmetry in the distribution of lesions. The cortical involvement, mass effect, and enhancement on both CT and MR images that have been described in patients with acute disseminated encephalomyelitis (7–9) were not encountered in our patients. The prognosis for patients with acute disseminated encephalomyelitis is much better than that for patients with Cree leukoencephalopathy (8). Pathologically, the two conditions are dissimilar. In cases of acute disseminated encephalomyelitis, demyelination involves the brain and spinal cord and is distinctly perivascular in distribution with associated inflammatory changes (8). Unfortunately, our patients did not undergo imaging of the spinal cord. At autopsy in our six patients, however, the white matter tracts in the upper two-thirds of the spinal cord were involved in one patient and minimal changes were seen in the spinal cord in two others. The spinal cord was free of disease in the remaining three patients. Inflammatory changes were absent in both the brain and spinal cord. Imaging findings in patients with acute disseminated encephalomyelitis, advanced stages of metachromatic leukodystrophy, and globoid cell leukodystrophy (Krabbe disease) may be similar to those in our patients. However, metachromatic leukodystrophy and Krabbe disease are clinically and pathologically different from Cree leukoencephalopathy (1). Moreover, the areas of high attenuation described at CT in patients with globoid cell leukodystrophy (10) were not present in our patients.
The imaging findings in our patients cannot be explained as depicting cerebrovascular disease. Preservation of cortex and lack of vascular territory distribution exclude arterial causes. Absence of mass effect, hydrocephalus, and hemorrhage argue against thrombosis of the superficial or deep cerebral venous systems (11). At neuroimaging in our patients, abnormality in the cerebral veins or dural sinuses was not seen. In addition, both venous systems were found to be patent at autopsy, and petechial hemorrhages were not encountered at histopathologic examination.
The Cree population in James Bay is exposed to methylmercury through dietary consumption of contaminated local fish (12). However, the clinical, pathologic, and MR imaging findings in methylmercury intoxication are different from those in our patients. Unlike Cree leukoencephalopathy, methylmercury poisoning, whether congenital or acquired, is progressive in nature, and symptoms include sensory and visual abnormalities, ataxia, tremors, and cognitive impairment (13–16). Pathologic findings include neuronal loss, cortical atrophy, and gliosis, which is most pro- nounced in the the paracentral and parieto-occipital regions (15). MR images depict enlargement of the cerebellar folia and parietal and occipital sulci with no white matter abnormality (15,16).
We believe our study had two shortcomings. First, our patients were far from neuroimaging facilities at the time of onset of clinical symptoms, and they were in grave neurologic status by the time they were referred to our institution. Thus, neuroimaging studies were not obtained immediately after the onset of the neurologic deterioration, which might in part explain the extensive nature of abnormalities revealed at CT and MR imaging. Second, we were not able to assess the temporal evolution of the neuroimaging findings because of the rapidly deteriorating clinical course or lack of imaging follow-up after the parents returned the infants to their homes in northern Quebec. We cannot explain the selective involvement of the ventral lateral and ventral posterior thalamic nuclei in two cases, the predominant tegmental involvement in three cases, the involvement of the red nuclei in two cases, or the sparing of the olives in two cases.
CT was comparable to MR imaging for the depiction of changes in the lobar and cerebellar white matter and corpus callosum. Moderate changes in the internal capsule, globus pallidus, and thalamus, however, were not depicted at CT. Because of the beam-hardening artifacts at CT and the superior tissue contrast and multiplanar capability at MR imaging, brainstem involvement was depicted far more clearly on MR images. As is true for other white matter diseases, T2-weighted sequences are more sensitive than T1-weighted sequences in the identification of lesions.
In conclusion, the hallmark of Cree leukoencephalopathy at neuroimaging is the extensive, diffuse, and symmetric white matter abnormality that involves the cerebrum, cerebellum, and brainstem with sparing of the caudate nucleus and putamen. Since radiologic findings in advanced stages of other white matter diseases may be similar, neuroimaging findings are complementary to clinical findings in establishing the diagnosis of Cree leukoencephalopathy. Further research is necessary to establish the etiology of this disease.
|
Footnotes
|
|---|
Author contributions: Guarantor of integrity of entire study, I.A.A.; study concepts, I.A.A., Y.G.P., A.M.O.; study design, I.A.A.; definition of intellectual content, I.A.A., Y.G.P., A.M.O.; literature research, I.A.A., D.N.B.; clinical studies, D.N.B., K.M.V.; data acquisition, I.A.A., D.N.B.; data analysis, I.A.A., Y.G.P., A.M.O.; manuscript preparation, I.A.A.; manuscript editing, A.M.O., D.N.B.; manuscript review, A.M.O., Y.G.P., D.N.B., K.M.V.
|
References
|
|---|
-
Black DN, Booth F, Watters GV, et al. Leukoencephalopathy among Native Indian infants in northern Quebec and Manitoba. Ann Neurol 1988; 24:490-496.[Medline]
-
Black DN, Booth F, Watters GV, et al. Encephalitis among Cree children in northern Quebec. Ann Neurol 1988; 24:483-489.[Medline]
-
Van der Knaap MS, Valk J. Canavan's disease. In: Van der Knaap MS, Valk J, eds. Magnetic resonance of myelin, myelination, and myelin disorders. 2nd ed. Heidelberg, Germany: Springer-Verlag, 1995; 216-219.
-
Van der Knaap MS, Valk J. Alexander's disease. In: Van der Knaap MS, Valk J, eds. Magnetic resonance of myelin, myelination, and myelin disorders. 2nd ed. Heidelberg, Germany: Springer-Verlag, 1995; 259-264.
-
Brismar J, Aqeel A, Brismar G, Coates R, Gascon G, Ozand P. Maple syrup urine disease: findings on CT and MRI scans of the brain in 10 infants. AJNR 1990; 11:1219-1228.[Abstract]
-
Uziel G, Savoiardo M, Nardocci N. CT and MRI in maple urine syrup disease. Neurology 1988; 38:486-488.[Abstract/Free Full Text]
-
Caldemeyer KS, Smith RR, Harris TM, Edwards MK. Gadolinium enhancement in acute disseminated encephalomyelitis. J Comput Assist Tomogr 1991; 15:673-675.[Medline]
-
Caldemeyer KS, Smith RR, Harris TM, Edwards MK. MRI in acute disseminated encephalomyelitis. Neuroradiology 1994; 36:216-220.[Medline]
-
Van der Meyden CH, de Villiers JFK, Middlecote BD, Terblanche J. Gadolinium ring enhancement and mass effect in acute disseminated encephalomyelitis. Neuroradiology 1994; 36:221-223.[Medline]
-
Sasaki M, Sakuragawa N, Takashima S, Hanaoka S, Arima M. MRI and CT findings in Krabbe disease. Pediatr Neurol 1991; 7:283-288.[Medline]
-
Yuh WTC, Simonson TM, Wang A, et al. Venous sinus occlusive disease: MR findings. AJNR 1994; 15:309-316.[Abstract]
-
Girard M, Noel F, Dumont C. Varying mercury exposure with varying food sources in a James Bay community. Arctic Med Res 1996; 55:69-74.[Medline]
-
McKeown-Eyssen GE, Ruedy J, Neims A. Methyl mercury exposure in northern Quebec. II. Neurologic findings in children. Am J Epidemiol 1983; 118:470-479.[Abstract/Free Full Text]
-
Inskip MJ, Piotrowski JK. Review of the health effects of methylmercury. J Appl Toxicol 1985; 5:113-133.[Medline]
-
Davis LE, Kornfeld M, Mooney HS, et al. Methylmercury poisoning: long-term clinical, radiological, toxicological, and pathological studies of an affected family. Ann Neurol 1994; 35:680-688.[Medline]
-
Korogi Y, Takahashi M, Shinzato J, Okajima T. MR findings in seven patients with organic poisoning (Minamata disease). AJNR 1994; 15:1575-1578.[Abstract]
TOP
Abstract
Introduction
