Antenatal and Postnatal Brain Magnetic Resonance Imaging in Muscle-Eye-Brain Disease

Background  Muscle-eye-brain disease (MEB) is a rare autosomal recessive disorder characterized by congenital muscular dystrophy, structural eye abnormalities, and type II lissencephaly. Previous reports of brain abnormalities on magnetic resonance images (MRIs) in MEB have been in children older than 1 year.

Objective  To describe serial antenatal and postnatal brain MRIs in a child with MEB.

Design  Case report.

Patient  We report a 2-year-old white boy with genetically confirmed MEB. Antenatal MRIs at 25 and 35 weeks’ gestation showed posterior ventriculomegaly but no cortical dysplasia. A postnatal brain MRI at age 1 week showed frontal cortical dysplasia and abnormal signal intensity within the frontal white matter. A brain MRI at 8 months showed bilateral frontoparietal polymicrogyria. All images demonstrated flattening of the pons and mild hypoplasia of the inferior vermis. The child had no weakness, and muscle involvement was only suspected when the serum creatine kinase level was found to be elevated at age 8 months.

Conclusion  Cortical dysplasia in MEB may not be evident until several postnatal months; therefore, if MEB is suspected, brain MRI performed in the first few months of life should be interpreted with caution.

Muscle-eye-brain disease (MEB) is a rare autosomal recessive disorder characterized by congenital muscular dystrophy, structural eye abnormalities, and cortical neuronal migration disorder, typically type II (cobblestone) lissencephaly. Associated features are abnormal white matter, flattened brainstem, cerebellar hypoplasia, and ventriculomegaly. Patients with MEB usually have muscle weakness and hypotonia at birth or in the first few months after birth. Few patients achieve independent ambulation. The most common ocular features are severe myopia and retinal hypoplasia, and other abnormalities such as cataract or glaucoma may occur. Most have severe mental retardation, absent speech, and epilepsy.¹ The recently identified MEB gene (POMGnT1) encodes the glycosyltransferase O-linked mannose β-1,2-N-acetylglucosaminyltransferase 1.² Following the identification of the gene defect, there has been evidence that the clinical spectrum and distribution of MEB is broader than previously recognized.³ The combination of muscle involvement and cortical dysplasia also occurs in Fukuyama type congenital muscular dystrophy ⁴ and Walker-Warburg syndrome,⁵ the genes for which also encode putative glycosyltransferase enzymes. Although MEB is generally milder than Walker-Warburg syndrome, there is significant clinical and neuroradiological overlap.⁶ More recently, our group reported structural brain involvement in patients with mutations in 2 other putative glycosyltransferases, the FKRP gene^7,8 and the Large gene,⁹ responsible for the forms of congenital muscular dystrophy referred to as MDC1C and MDC1D, respectively.

We report serial antenatal and postnatal brain magnetic resonance images (MRIs) in a 2-year-old white boy with a genetically confirmed diagnosis of MEB. To our knowledge, this is the first report describing early brain MRI findings in MEB.

The patient is the only son of healthy nonconsanguineous parents of Anglo-Irish descent. He has 4 half siblings, all unaffected by MEB, though 3 of them have autism or Asperger syndrome. A paternal uncle also has autism. Ventriculomegaly was documented by antenatal ultrasonography at 20 weeks’ gestation and confirmed by fetal MRI at 25 weeks’ gestation. Ventricular dilatation was most marked in the posterior lateral ventricles (Figure 1A). The third ventricle was also dilated and the fourth, prominent (Figure 1B). The inferior vermis appeared mildly hypoplastic (Figure 1B). Cortical folding was slightly delayed with minimal infolding of the central sulcus (Figure 1B). Fetal MRI 10 weeks later showed that cortical folding had increased. The lateral ventricles remained dilated, the pons flattened, and the vermis mildly hypoplastic (Figure 2 A and B).

The patient was born at 38 weeks’ gestation by Cesarean section for breech presentation. No resuscitation was required. Birth weight and head circumference were in the 25th centile. Routine examination findings in the postnatal ward were normal.

At 1 week, he was feeding well but had suboptimal visual attention and poor quality of movements. Brain MRI confirmed the previous findings and also showed dilatation of the anterior horns of the lateral ventricles. The cerebellar hemispheres appeared small and the cerebellar cortex dysplastic (Figure 3A). The frontal cortex was less folded than expected, and the frontal white matter showed abnormal long T1 and T2 (Figure 3B).

At 2 months, feeding difficulties were noticed. He had just started to smile. He had full conjugate eye movements, although visual following was limited. He had antigravity limb movements but reduced head and axial tone and a mild increase in limb tone. These features persisted at 7 months when Griffith neurodevelopmental testing gave an age equivalent of 4 to 4.5 months.

Brain MRI at 8 months showed obvious cortical dysplasia with areas of polymicrogyria in the frontal, frontoparietal, and anterior temporal lobes. There was abnormal low signal intensity (decreased T2) in the white matter, suggesting disordered myelination and possible areas of ectopic neurons. The pons was flattened, and the vermis remained hypoplastic. There was more marked ventricular dilation but no pachygyria or midline anomalies (Figure 4). Ophthalmological investigations revealed severe myopia (−12 diopters).

His serum creatine kinase level was elevated at 1567 U/L (normal, <200 U/L). Electroencephalography (at age 13 months) and peroneal nerve conduction velocity results were normal.

The results of muscle biopsy performed at age 16 months showed dystrophic changes. Immunohistochemical studies revealed a significant reduction of α-dystroglycan labeling with antibodies VIA4-1 (Upstate biotechnologies, Charlottesville, Va, following the manufacturer's instructions) and IIH6 (gift of K. Campbell, PhD), directed against different glycosylated epitopes of α-dystroglycan, and a mild reduction with an antibody (gift of S. Kroger, PhD ) that recognizes the core protein (Figure 5). There was normal labeling of β-dystroglycan and other sarcolemmal and extracellular matrix proteins with the exception of laminin-α2 chain, which was slightly reduced in some fibers. Immunoblot analysis of skeletal muscle protein extract showed a virtual absence of polypeptide bands corresponding to α-dystroglycan but normal labeling of β-dystroglycan (Figure 6).

Sequencing of the entire coding region of the POMGnT1 gene identified a novel heterozygous missense mutation in 1 allele, a 1373G>C, resulting in an Asp427His, which was inherited from his mother, and a single-nucleotide insertion 542insT creating a frameshift at amino acid 150. This second mutation was inherited from his healthy father; the missense mutation was excluded from 94 healthy controls. FKRP gene mutations were excluded as well following direct sequencing of the entire coding region as described by Brockington et al.⁷

At age 2 years, both the boy’s weight and head circumference were between the 3rd and 10th centile. He had prominent muscles, poor head control, and truncal hypotonia but was able to sit unsupported for a few minutes, to roll, and to bear weight on his legs if held. He had antigravity movements, but the quality of his movements was slow and poor because of the pyramidal involvement. There were mild contractures of the knees and ankles bilaterally, and tendon reflexes were brisk. His severe myopia was partially corrected with glasses. He babbled but had no recognizable words.

Muscle-eye-brain disease was initially described in Finnish patients, but it is now clear that the distribution of the disease is wider than originally thought.^1,3 This is the first report of MEB in a UK patient. Our patient is mildly affected when compared with most patients with MEB and in particular does not have overt muscle weakness. A neuromuscular disorder was suspected only on the discovery of high serum creatine kinase levels, which were measured when his cortical dysplasia was recognized. These findings support our previous suggestion that serum creatine kinase levels should be determined in patients with cortical dysplasia,¹⁰ and this is especially valid if the pattern of brain involvement suggests a cobblestone lissencephaly. Muscle biopsy results showed dystrophic features with reduced immunolabeling of α-dystroglycan. This, together with his clinical features, led us to suspect a diagnosis of MEB, which was confirmed by finding homozygous mutations in the MEB gene, POMGnT1.

The initial MRI in our patient was performed at 25 weeks’ gestation and confirmed the ventriculomegaly detected on ultrasound. This was more marked posteriorly, as is often observed in MEB,¹¹ but such posterior dilatation is nonspecific. Cortical abnormalities may be difficult to identify at this early gestation, as the major sulci are only beginning to form.¹² A second fetal MRI at 35 weeks’ gestation still did not show any obvious abnormality of cortical maturation. While severe lissencephaly, such as that observed in Walker-Warburg syndrome, can easily be seen on fetal MRIs, more discrete polymicrogyria may not be recognized.¹³ Both prenatal MRIs showed vermis hypoplasia and, in retrospect, flattening of the pons. A suggestion of abnormal folding in the frontoparietal cortex was noted for the first time on the early postnatal brain MRI obtained 1 week after birth. The pattern of polymicrogyria, however, was still not obvious and was only fully recognized on the MRI performed at age 8 months, when widespread polymicrogyria affecting the frontal, frontoparietal, and anterior temporal lobes was apparent. The white matter also had abnormal signal intensity. While other authors have reported similar brain changes of cortical dysplasia, cerebellar atrophy, flattened pons, and ventricular dilatation in MEB, none have examined patients younger than 1 year; thus, the age at which these abnormalities become apparent was not known.^1,11,14,15 In addition, the presence of polymicrogyria has been reported most frequently in Fukuyama type congenital muscular dystrophy.^14,15 This case therefore illustrates the evolution of the polymicrogyria in MEB, which may not become evident until several months postnatally. α-Dystroglycan has been implicated in central nervous system development. Defects of cortical layering, brain morphogenesis, and basement membrane expression, resulting in overmigration of neurons beyond the pia mater as observed in type II lissencephaly, have been documented in the Large ^myd mouse, which has abnormal glycosylation of α-dystroglycan because of a mutation in the gene encoding a putative glycosyltransferase, Large,¹⁶ also mutated in MDC1D.⁹ Targeted deletion of dystroglycan in mouse brain results in similar defects.¹⁷ Hypoglycosylation of α-dystroglycan may underlie the observed central nervous system defects in MEB.¹⁸ The cortical migration defect, however, may not be obvious in the first few months of life, and this should be taken into account when assessing children with suspected MEB.

Correspondence: Francesco Muntoni, MD, Dubowitz Neuromuscular Centre, Imperial College, Hammersmith Campus, DuCane Road, London W12 ONN, England (f.muntoni@imperial.ac.uk).

Accepted for Publication: December 5, 2003.

Author Contributions:Study concept and design: Longman, Mercuri, Rutherford, and Muntoni. Acquisition of data: Mercuri, Cowan, Allsop, Brockington, and Toda. Analysis and interpretation of data: Longman, Cowan, Brockington, Jimenez-Mallebrera, Rutherford, Toda, and Muntoni. Drafting of the manuscript: Longman, Mercuri, Cowan, and Muntoni. Critical revision of the manuscript for important intellectual content: Longman, Cowan, Allsop, Brockington, Jimenez-Mallebrera, Rutherford, Toda, and Muntoni. Obtained funding: Muntoni. Administrative, technical, and material support: Longman, Allsop, and Jimenez-Mallebrera. Study supervision: Longman, Mercuri, Cowan, Rutherford, and Toda.

Funding/Support: This study was supported by Centre and Research grants from the Muscular Dystrophy Campaign, London, England; European Community grant QLG1 CT 1999 00870 from the Medical Research Council, London; and the National Specialist Commissioning Advisory Group, London, to the Hammersmith Hospital Neuromuscular Centre, London.

Acknowledgment: We thank Yuko Nakabayashi, BSc, for technical assistance and Kevin Campbell, PhD, and Stephen Kroger, PhD, for the kind gift of the IIH6 and core α-dystroglycan antibodies, respectively.