Alexander disease is a rare neurological disorder that was first described in young children who presented with megalencephaly, retarded physical and mental development, and seizures. The hallmark feature of the disease is a predominantly frontal leukodystrophy combined with widespread deposition of Rosenthal fibres in astrocytes. Rosenthal fibres are ubiquitinated protein aggregates composed of the intermediate filament protein glial fibrillary acidic protein (GFAP) and small stress proteins, such as αB-crystallin and heat shock protein 27. The identification of abundant Rosenthal fibres in older patients prompted a new classification to include juvenile and adult forms, even though these older patients differed clinically from the infantile cases. Whether the juvenile and adult forms are truly variants of the same disease has engendered considerable debate, but genetic studies have now provided the necessary evidence to resolve this dispute. Transgenic mice that constitutively overexpress GFAP die at an early age and show marked formation of Rosenthal fibres. Prompted by these results, we investigated GFAP as a candidate gene and found that 12 of 13 patients with Alexander disease carried heterozygous GFAP mis-sense mutations. 1 Of those 12, 11 were infantile cases. The other, with an age at onset of 10 years who died at 48 years, was not easily classified as either juvenile or adult. Parental pairs who were available for testing were negative, implying a de novo origin of the GFAP mutations. Other researchers subsequently confirmed these findings and reported examples of GFAP mutations in juvenile Alexander disease. A recent review summarises these findings, which indicate that Alexander disease primarily results from heterozygous mutations in GFAP, which are likely to be gain of function mutations with nearly 100% penetrance. 2 However, whether there was truly an adult-onset Alexander disease remained unsettled. Rosenthal fibres are observed in some adults who are asymptomatic or have entirely unrelated diseases. Reports of Alexander-type illnesses in adults cover a wide range of neurological phenotypes. Two recent reports from Japan now clarify this issue. Okamoto and colleagues3 studied a family in which the mother (in her sixth decade of life) and one of her adult offspring (fourth decade of life) developed palatal myoclonus, spastic paraparesis, and atrophy of the caudal brainstem and spinal cord. Another adult offspring showed subtle neurological signs in the fourth decade of life. Although no tissue was available to confirm the presence of Rosenthal fibres, all three individuals shared the same novel V87G mutation in GFAP. A second family with adult onset Alexander disease has been described by Namekawa and co-workers. 4 Two brothers, the only children of unaffected non-consanguineous parents, both developed an adultonset progressive gait disorder—characterised by spasticity—in association with MRI findings of atrophy in the medulla and cervicothoracic spinal cord. One of the brothers was in his early 30s at onset and died at age 53; autopsy confirmed the presence of Rosenthal fibres. The second brother showed onset at age 48 with milder spasticity but similar MRI findings. Neither had palatal myoclonus. Both brothers proved heterozygous for a novel R276L GFAP mutation. The absence of this mutation in 78 Japanese controls indicates that this mutation causes disease. Hence an adult form of Alexander disease caused by GFAP mutations does indeed exist.

However, the study by Namekawa and colleagues also raises several questions. How could two brothers carry the same mutation? Since both of their …