Alexander disease: Combined gene analysis and MRI clarify pathogenesis and extend phenotype

In 1949, W. Steward Alexander reported the neuropathological findings in a 15-month-old boy with severe and progressive developmental defects associated with predominantly frontal megancephaly. The most striking histopathological finding was the presence of large homogeneous bodies in fibrillary astrocytes. These cells surrounded blood vessels and were most prominent in a zone of varying depth beneath the ependyma, were less prominent in subarcuate fibers and absent in gyri. They were most prominent in the frontal and parietal regions and much less so in the occipital lobes. The basal ganglia were also involved. The tracts and nuclei of the brainstem nuclei were involved heavily from midbrain to lower medulla, the cerebellar hemispheres only slightly. Alexander concluded that this was a “non-familial idiopathic sclerosis in which fibrinoid degeneration of fibrillary astrocytes has occurred”. In 1961, Hallervorden re-examined the slides of this case and concluded that the abnormal fibers were Rosenthal fibers (RF), that is intracytoplasmic filamentous inclusions with particular staining characteristics within astrocytes. RF are seen in conditions with long-lasting intensive fibrillary gliosis, such as in astrocytic tumors, glial scars, syringomyelia, tuberous sclerosis and sometimes in Parkinson and Alzheimer Disease, but their accumulation in the case described by Alexander, and similar cases seen since then, is particularly striking. The major chemical components of RF are glial fibrillary acidic protein (GFAP), the small heat shock proteins alpha crystalline and hsp27, and ubiquitin. Sixty-four cases conforming to the clinical and pathological features of the first case were reported by 1964 with a detailed review of 40 such cases provided by Pridmore et al. Friede proposed the eponym Alexander’s Disease in 1964. Until 2001 the nosology of AD was a matter of controversy and diagnosis during life was a challenge. The characteristic and consistently progressive course in infants and young children associated with the striking RF accumulation led to the consensus that it was a genetically determined disorder that involves astrocytes. However, a genetic basis remained unproven because, except for one instance, all cased were sporadic. In-vivo diagnosis required brain biopsy. Adult patients with a progressive disorder associated with RF accumulation remained a nosological challenge, and it was recommended that they be referred to as Rosenthal Fiber encephalopathy rather than adult Alexander disease. These uncertainties were resolved in 2001 by the seminal report of Brenner et al., which demonstrated GFAP mutations in 11 unrelated AD patients with infantile, juvenile and adult phenotypes. More than 30 different pathogenic mutations have now been identified and are updated in the website http:// www.waisman.wisc.edu/alexander. This accomplishment was the result of long range and effective collaboration between investigators in neuropathology, veterinary medicine and genetics. That GFAP is a major component of RF had been demonstrated by neurochemical studies of Goldman et al. and the immunocytochemical studies of Johnson and Bettica. Surprisingly GFAP null mice show only relatively subtle changes. A key step was the demonstration by Messing et al. that mice that overexpress GFAP 10 to 15 fold develop a fatal encephalopathy associated with astrocyte swelling. This finding led to the collaborative search for GFAP mutations in AD patients and the demonstration that they are the cause of the disease. Two articles in this issue provide a wealth of new information about AD with focus respectively on genetics and neuroimaging. The two articles interdigitate closely both in respect to topic and authorship. Taken together they provide exciting new insights about the diagnosis, genetics and pathogenesis of AD and the biology of GFAP. In addition they clearly document and delineate variant milder phenotypes in adolescents and adults that have important implications for differential diagnosis and genetic counseling. Figure 3 in the report by Li et al. shows the phenotype associated with each of 102 pathogenic mutations in AD patients that have been identified. Seventy-two had the infantile phenotype, 22 juvenile, adult and one person was asymptomatic. All of the infantile phenotype patients, 21 of the juvenile, and 4 of the adult phenotype patients were sporadic. Five pedigrees with more than one affected member have been reported. All of the sixteen affected persons in these pedigrees had the adult or the juvenile phenotype. One pedigree had 7 affected members in 3 generations. The phenotype of the infantile forms in the article of Li et al. conformed to those in previous reports. Onset ranged from birth to 1.5 years. Seizures, cognitive defects, macrocephaly bulbar signs and spasticity were present in 92, 82, 62, 58 and 52 percent respectively. Survival ranged from 38 days to 19.5 years. However, in the juvenile and adult onset cases the phenotype was quite different, and in many instances the demonstration of the AD molecular defect came as a surprise. In the juvenile phenotype, macrocephaly was present in only 27% and in the adult cases it was not a feature. Cognitive defects were present in 60% of the juvenile cases but not in the adults. Eye movement disturbances, bulbar and pseudobulbar signs, and ataxia are the most frequent features. Palatal myoclonus has been reported. In the family with 7 affected members dysautonomia and sleep disturbances were prominent features in addition to bulbar signs and ataxia. Magnetic Resonance Imaging studies confirm the differences between the infantile and later onset cases, and are of critical significance for diagnosis. In 2001, van der Knaap et al. established five criteria for the MRI diagnosis of AD (cerebral white matter change with frontal preponderance, periventricular rim with high-signal T1 and low signal on T2 weighted images, abnormalities of basal ganglia and thalami, and in brain stem, and contrast enhancement with a particular distribution) which mirror the histopathology of the infantile onset cases; the MRI alterations in the periventicular rim and the enhancement pattern probably relate directly to Rosenthal fiber accumulation. The report from the same group which appears in this EDITORIALS