Pearls & Oy-sters: Adult-Onset Alexander Disease With Transient Swelling of the Medulla Oblongata

Discussion

Alexander disease is classified according to the age at onset as infantile (IOAD), juvenile (JOAD), or AOAD, each of which has its own peculiar imaging characteristics.⁵ MRI features of IOAD include cerebral white matter abnormalities in the frontal lobe and signal abnormalities indicating swelling or atrophy of the basal ganglia and thalami and periventricular rim. AOAD is characterized by brainstem lesions; abnormal signals in the white matter, basal ganglia, and thalamus are less frequent. JOAD may present with imaging features similar to those of IOAD or may have brainstem lesions as in AOAD.

Atrophy, from the medulla oblongata to the superior cervical cord, is a well-known feature of AOAD in MRI, and the sagittal diameter of the medulla oblongata (<9.0 mm) is useful for the diagnosis of the disease, with high sensitivity and specificity.^6,7 However, in this case, the medulla oblongata did not atrophy to less than 9 mm, even 13 years after the onset of the disease. It was rather swollen in the early stages of the disease. We compared the sagittal diameter of the medulla oblongata in this case with the diameter of the cases in the control group included in a previous study.⁷ The medullary diameter (mean ± SD) of the control group was 12.3 ± 1.5 mm and the Z scores at each age (45, 50, 55, and 58 years) in this case were 3.13, 1.33, −0.80, and −2.00, respectively. In IOAD, swelling of the white matter and basal ganglia is sometimes observed. This is thought to be related to hyperplasia of astrocytes as well as the accumulation of Rosenthal fibers.⁸ The same pathophysiology may be assumed for the swelling in this case.

The disease course of this case provides insight into the classification of Alexander disease. Given the paucity of reports on long-term follow-up before and after the onset of AOAD, the medulla oblongata may be transiently swollen in the early stages of this disease. A previous study reported a similar course in a case of JOAD and made an interesting hypothesis about the phenotype of Alexander disease: that the brainstem atrophy of AOAD is a sequential condition from JOAD.⁹ Another study statistically classified Alexander disease into 2 groups: those with white matter lesions (type I) and those with brainstem lesions (type II).¹⁰ They proposed that type II includes a wide age range from infantile to adult-onset cases. This may be related to the fact that brainstem lesions in Alexander disease progressively change with patient age.

We listed brainstem encephalitis and Alexander disease as the differential diagnoses in this case. Other differential diagnoses included sarcoidosis, neuromyelitis optica, and brain tumors, such as gliomas. Inflammatory and immune mechanisms were ruled out because of a lack of response to steroids. A brain tumor was ruled out because of the lack of increased uptake on FDG-PET. AOAD might be a possible differential diagnosis in cases considered as refractory brainstem encephalitis.

This patient initially presented with hiccups and vocal cord paralysis, and eventually, apnea during the process of atrophy of the swollen medulla oblongata. There are several reports on cases with AOAD that resulted in obstructive or central respiratory failure.^11,12 As in our case, one study showed vocal cord paralysis on laryngoscopy in a case of AOAD, suggesting that there are other similar cases.¹¹ Therefore, we recommend that patients with AOAD should undergo laryngoscopy and polysomnography for assessment.

In this case, we found a GFAP variant (c. 972 G > A) that had not been registered in any database. Based on the guidelines proposed by the American College of Medical Genetics and Genomics, this variant fulfills one moderate criterion (“The absence of the variant in population databases”) and 2 supporting criteria (“multiple lines of computational evidence supporting a deleterious effect on the gene” and “missense variant in a gene with a low rate of benign missense variation and in which missense variants are a common mechanism of disease”).¹³ This would only qualify for a “variant of unknown significance (VUS).” Although genetic analysis of the parents was not performed in this case, there was no family history of the disease. Based on this fact, if we judge this variant to be a de novo variant, it satisfies the “likely pathogenic” classification. Even if the variant only qualifies as VUS under the guidelines, the clinical picture of this case is consistent with Alexander disease. This is paradoxically important for clinicians. It suggests that even among genetic variants that are not considered highly pathogenic by the guidelines, there may be pathogenic variants that are buried because there are not enough data to classify them as such. Hence, it is important for clinicians to provide personalized medicine based on the clinical picture, not just on guidelines.