Alexander Disease
A number sign (#) is used with this entry because Alexander disease (ALXDRD) is caused by heterozygous mutation in the gene encoding glial fibrillary acidic protein (GFAP; 137780) on chromosome 17q21.
DescriptionIn decreasing order of frequency, 3 forms of Alexander disease are recognized, based on age of onset: infantile, juvenile, and adult. Younger patients typically present with seizures, megalencephaly, developmental delay, and spasticity. In older patients, bulbar or pseudobulbar symptoms predominate, frequently accompanied by spasticity. The disease is progressive, with most patients dying within 10 years of onset. Imaging studies of the brain typically show cerebral white matter abnormalities, preferentially affecting the frontal region (Gorospe et al., 2002). All 3 forms have been shown to be caused by mutations in the GFAP gene.
Clinical FeaturesThis disorder, first described by Alexander (1949), is characterized clinically by development of megalencephaly in infancy accompanied by progressive spasticity and dementia. The features are similar to those of Canavan disease (271900).
Gorospe et al. (2002) reported 12 genetically confirmed cases of Alexander disease. Seven of the 12 had onset in infancy (range 2-18 months), with seizures being the most common presenting sign, followed by failure to thrive and delayed motor development. Five patients had juvenile onset (between 5 and 9 years) and presented with variable symptoms ranging from asymptomatic (2 patients) to linear growth failure, excessive sleepiness and vomiting. Patients in both groups showed megalencephaly, bulbar or pseudobulbar signs, spasticity, cognitive deficits and developmental delay. In addition, all patients showed diffuse and symmetric white matter abnormalities in the frontal regions of the brain. Gorospe et al. (2002) suggested that GFAP gene analysis be included in the diagnostic evaluation of patients presenting with frontal leukoencephalopathy by MRI.
Bassuk et al. (2003) reported an infant with Alexander disease who presented with poor feeding on the first day of life, followed by emesis and weight loss. MRI on day 21 of life showed signal abnormalities in the frontal lobes and basal ganglia. Over the next 12 days, the patient became increasingly somnolent and hypotonic, and developed seizures on day 33. Magnetic resonance spectroscopy (MRS) performed 14 days apart demonstrated an interval 2-fold increase in the lipid/lactate peak over the right basal ganglia. Over the course of 25 days, head growth increased from the 50th to the 75th percentile. The child died on day 38 from prolonged seizures and respiratory failure. Mutation analysis detected a heterozygous mutation in the GFAP gene (137780.0012). Bassuk et al. (2003) commented on several unusual aspects of the case, including the rapid clinical decline, the rapid head growth, and the demonstration of progressive lactate elevation in the brain by MRS.
Stumpf et al. (2003) reported a family with an autosomal dominant adult form of Alexander disease. The clinical phenotype varied in severity, but the pattern of evolution was similar in all affected members. Although sleep disturbances and dysautonomia, primarily constipation, began in childhood, the major neurologic features began in the third or fourth decade of life. Features included bulbar signs, ataxia, and pyramidal signs. All patients also had mild dysmorphic features, including progressive kyphosis, arched palate, and short neck. MRI of the older patients showed atrophy of the medulla without signal abnormalities. A mutation in the GFAP gene (137780.0013) was identified in all affected members.
Li et al. (2005) reported detailed clinical features of 44 patients with Alexander disease, including 26 with infantile onset, 15 with juvenile onset, and 3 with adult onset. The most common features among the patients with infantile onset included seizures (92%), cognitive defects (82%), macrocephaly (62%), bulbar signs (62%), ataxia (58%), and spasticity (52%). The phenotype of juvenile- and adult-onset cases was less severe. Features of juvenile patients included bulbar signs (73%), cognitive defects (60%), spasticity (53%), ataxia (47%), seizures (27%), and macrocephaly (20%). None of the patients with adult onset had macrocephaly, seizures, or cognitive defects. There was a suggestion of male predominance for the disorder.
Sreedharan et al. (2007) reported an unusual case of a 38-year-old woman with Alexander disease. She presented with a 2-year history of progressive reading difficulty with oscillopsia, slurring dysarthria, choking, and stumbling. Past medical history was significant for endocrine disturbances with an episode of amenorrhea, hypothyroidism, depression, and hypothermic episodes associated with ataxia, facial twitching and drowsiness. Physical examination showed torsional nystagmus and palatal, tongue, and jaw tremor. She had symptomatic microcoria, mild left arm dysmetria, ataxia, and lower limb hyporeflexia. Brain MRI showed brainstem atrophy and symmetric signal changes in the medulla and cerebellum. Her father reportedly had microcoria but refused participation. Genetic analysis identified a heterozygous GFAP mutation in the proband. Sreedharan et al. (2007) commented that this case represented an unusually slowly progressive form of adult-onset Alexander disease.
Pathologic Findings
Histologically, Alexander disease is characterized by Rosenthal fibers, homogeneous eosinophilic masses which form elongated tapered rods up to 30 microns in length, which are scattered throughout the cortex and white matter and are most numerous in the subpial, perivascular and subependymal regions. These fibers are located in astrocytes, cells that are closely related to blood vessels. Demyelination is present, usually as a prominent feature. A few cases have had hydrocephalus (Alexander, 1949). Rosenthal fibers are commonly found in astrocytomas, optic nerve gliomas and states of chronic reactive gliosis, but they are especially conspicuous in Alexander disease. Herndon et al. (1970) expressed the view that Rosenthal fibers found in this situation are the result of degenerative changes in the cytoplasm and cytoplasmic processes of astrocytic glial cells.
Iwaki et al. (1989) found that alpha-B-crystallin (CRYAB; 123590) accumulates in the brain in Alexander disease.
DiagnosisVan der Knaap et al. (2001) proposed specific MRI criteria for the diagnosis of Alexander disease: extensive, symmetric white matter abnormalities with frontal preponderance; periventricular signal changes; basal ganglia and thalamic signal changes; brainstem lesions; and contrast enhancement of multiple areas throughout the brain.
Van der Knaap et al. (2005) reported 9 patients with Alexander disease confirmed by genetic analysis who had atypical MRI features. Alexander disease was not the initial diagnosis for any of the patients, and none of the patients met the MRI-based criteria proposed by van der Knaap et al. (2001). MRI in 8 patients showed predominantly posterior fossa lesions, especially multiple tumor-like brainstem lesions. One patient had asymmetric frontal white matter abnormalities and basal ganglia abnormalities. Van der Knaap et al. (2005) concluded that DNA diagnostics is warranted in patients with atypical MRI features that are only suggestive of Alexander disease.
Van der Knaap et al. (2006) reported 7 patients with genetically confirmed Alexander disease who had no or inconspicuous cerebral white matter abnormalities and no or minimal contrast enhancement on brain MRI. All had juvenile disease onset with signs of brainstem or spinal cord dysfunction, including bladder and gait disturbances. MRI findings were predominantly signal changes or atrophy of the medulla and spinal cord. Four patients had a kind of 'garland' along the ventricular wall. Van der Knaap et al. (2006) concluded that Alexander disease is not invariably a leukoencephalopathy, and that patients with later onset of the disorder may have more unusual phenotypic variation.
InheritanceWohlwill et al. (1959) described a sibship of 9, of whom 1 sister and 3 brothers had large heads called hydrocephalic and died at ages 4, 5, 6 and 3, respectively. Alexander disease was proven histologically in the last. Although this sibship suggested possible autosomal recessive inheritance, all the molecular genetic evidence favors autosomal dominant inheritance, i.e., de novo heterozygous mutations as the cause. In vitro studies of one such mutation in the GFAP gene causing Alexander disease (137780.0003) found that the mutant protein accumulates into Rosenthal fibers by a pathway that involved filament aggregation and the association of alpha-B-crystallin and HSP27 (602195). The data confirmed that the effects of the specific GFAP mutation are dominant; in the heterozygote mutant, the gene product was dominant over wildtype GFAP in coassembly experiments (Der Perng et al., 2006).
Molecular GeneticsMutations were found in the infantile form of Alexander disease by Brenner et al. (2001), in the juvenile form by Sawaishi et al. (2002), and in the adult form by Namekawa et al. (2002).
Brenner et al. (2001) identified de novo, heterozygous mutations in the GFAP gene in 10 of 11 patients with Alexander disease (137780.0001-137780.0005). Rodriguez et al. (2001) likewise identified de novo heterozygous missense GFAP mutations in exon 1 or exon 4 of 14 of 15 patients who were candidates for Alexander disease on the basis of suggestive neuroimaging abnormalities. These included patients without macrocephaly. Affected sibs whose parents were unaffected, including 1 family with neuropathologically proved Alexander disease (Wohlwill et al., 1959), could represent autosomal recessive transmission or germinal mosaicism for a dominant mutation. Therefore, Rodriguez et al. (2001) suggested that after the birth of a patient with Alexander disease with a de novo GFAP mutation, prenatal diagnosis should be proposed for all subsequent pregnancies. It remained to be determined whether the heritable dominant forms of Alexander disease described in 2 families, both of which had late onsets after age 25 years (Howard et al., 1993; Schwankhaus et al., 1995), also had GFAP mutations as the cause.
Of 13 patients with MRI white matter abnormalities consistent with Alexander disease, 12 were found by Gorospe et al. (2002) to have GFAP mutations. Four of the 9 changes identified were novel mutations.
Li et al. (2006) determined that the paternal chromosome carried the GFAP mutation in 24 of 28 unrelated cases of Alexander disease analyzed, suggesting that most mutations occur during spermatogenesis rather than in the embryo. No effect of paternal age was observed.
In 13 unrelated Italian patients with Alexander disease, including 8 with the infantile, 2 with the juvenile, and 3 with the adult form, Caroli et al. (2007) identified 11 different mutations in the GFAP gene (see, e.g., 137780.0005), including 4 novel mutations. Ten mutations occurred in the rod domains and 1 in the tail domain.
Karp et al. (2019) reported a patient with adult-onset Alexander disease in whom, after excluding mutation in the GFAP-alpha isoform, they identified heterozygosity for a missense mutation (c.1289G-A, R430H) in exon 7A of the GFAP-epsilon isoform. The authors noted that the same mutation in GFAP-epsilon had been identified by Melchionda et al. (2013) in a brother and sister half-sib pair with adult onset of the disorder in whom mutation in GFAP-alpha had been excluded. The brother also had a mutation (c.2566C-T, P856S) in the HDAC6 gene (300272).
Animal ModelHagemann et al. (2009) noted that Rosenthal fibers in the complex astrocytic inclusions characteristic of Alexander disease contain GFAP, vimentin (VIM; 193060), plectin (PLEC1; 601282), ubiquitin (UBB; 191339), HSP27, and alpha-B-crystallin. CRYAB regulates GFAP assembly, and elevation of CRYAB is a consistent feature of Alexander disease; however, its role in Rosenthal fibers and disease pathology is not known. In a mouse model of Alexander disease, Hagemann et al. (2009) showed that loss of Cryab resulted in increased mortality, whereas elevation of Cryab rescued animals from terminal seizures. When mice with Rosenthal fibers induced by overexpression of GFAP were crossed into a Cryab-null background, over half died at 1 month of age. Restoration of Cryab expression through the GFAP promoter reversed this outcome, showing the effect was astrocyte-specific. Conversely, in mice carrying an Alexander disease-associated mutation and in mice overexpressing wildtype GFAP, which, despite natural induction of Cryab also died at 1 month, transgenic overexpression of Cryab resulted in a markedly reduced CNS stress response, restored expression of the glutamate transporter Glt1 (SLC1A2; 600300), and protected these animals from death.