Adult-onset Alexander disease with unusual inflammatory features and a novel GFAP mutation in two patients

Alexander disease (AxD), which was first described by a New Zealand pathologist, Dr W. S. Alexander, is a rare, usually fatal neurodegenerative leukodystrophy caused by dominant mutations in the gene for glial fibrillary acidic protein (GFAP), the major type III intermediate filament of astrocytes in the central nervous system (CNS) [1, 2]. Patients are classified into two distinct disease categories depending on the distribution of the lesions and clinical presentation, with type I cases being early-onset and type II disease occurring at all ages [3, 4]. Later-onset patients commonly show features of type II disease with insidious signs of hindbrain dysfunction such as ataxia, palatal myoclonus, dysphagia, dysphonia and autonomic dysfunction, and spasticity [3, 4]. The purpose of this case report is to document the NM_002055.5(GFAP): c.592C>T (p. Arg198Trp) variant as a novel adult-onset AxD-associated pathogenic mutation and discuss the unusual clinico-pathological features in two patients with this hitherto unreported GFAP variant. The first patient was diagnosed post mortem following tissue sampling. In addition to the characteristic Rosenthal fibres (RFs), a florid inflammatory component including B lymphocytes and plasma cells was noted, a finding that has not been previously reported in any of the animal models or human cases [3]. The second patient was diagnosed following the identification of characteristic MRI features. Patient 1: A 37-year-old man developed progressive bulbar symptoms and four-limb weakness over 10 years. MRI of the brain showed expansion of the medulla with T2 hyperintensity and ill-defined heterogeneous enhancement extending to the right cerebellar peduncle (Figures 1A and S1M). No evidence of diffusion restriction or volume loss to suggest neurodegeneration was noted. No abnormalities were noted in the cerebrum or ventricles. Imaging features were suggestive of an inflammatory process or a low-grade glioma. CT scan of the chest showed bilateral nodular opacities in the upper lobes of the lungs. CSF analysis showed a normal white cell count, and protein, immunoglobulin and light chain levels. CSF cytology was negative for tumour cells, and there were two or more oligoclonal bands that were absent in the serum with a mildly elevated IgG index (ratio 0.8) indicative of intrathecal synthesis. Other investigations were non-contributory, with no evidence to support the existence of a comorbid disorder. Based on clinical and imaging features of a mass-like lesion in the brainstem as well as lung lesions, he was treated for a presumptive diagnosis of neurosarcoidosis. However, he died following an episode of severe breathlessness. Post mortem examination of the brain showed only a circumscribed area of discolouration in the medulla oblongata corresponding to the known brainstem lesion. The remaining brain parenchyma appeared normal. No diagnostic abnormality was noted in any of the other organs including the lungs. The cause of death was presumed to be central respiratory failure. Histopathological examination of the lesion in the medulla oblongata showed rarefaction and pallor of brain parenchyma due to loss of myelin, associated with extensive accumulation of brightly eosinophilic RFs, which appeared more intense in a perivascular, subependymal and subpial location (Figure 1B). Scattered hypertrophic astrocytes were also seen in the background. There was diffuse involvement of the medulla including the inferior olivary nucleus (IO), medullary tegmentum and cortico-spinal tract. In addition, there was a florid inflammatory infiltrate composed of a mixed population of mature T and B lymphocytes as well as an abundant polyclonal population of plasma cells (Figures 1C–H and S1I–L). An IgG4 stain was negative. No atypical lymphoid cells were seen to suggest a lymphoproliferative disorder. Sections of the frontal cortex, cerebellum and midbrain were normal. Features were consistent with localised/type II AxD. Subsequent sequencing of the GFAP gene from paraffin-embedded tissue from the lesion in the medulla showed a germline mutation, heterozygous for the NM_002055.5(GFAP): c.592C>T (p. Arg198Trp) likely pathogenic variant consistent with a histological diagnosis of AxD. Testing was performed on formalin-fixed paraffin-embedded (FFPE) tissue; however, the variant allele frequency of 46% (VAF) was consistent with a germline variant. The NM_002055.5(GFAP): c.592C>T (p. Arg198Trp) variant in exon 3 is a novel variant and is not annotated in any of the major genetic database registries. There was no family history of AxD. Genetic referral for relevant family members has since been initiated. Patient 2: A 53-year-old male presented with features suggestive of restless leg syndrome (RLS) at night. Suspicion of RLS, provoked by iron deficiency (ferritin 23 ng/mL), was raised. Clinical examination however showed bilateral Babinski signs and diffuse hyperreflexia. No ataxia, sensory disturbance or other abnormalities were present. The patient had no urinary urgency or dysphagia. An MRI revealed ‘tadpole’ atrophy of the medulla oblongata (Figure 2A), reminiscent of adult-onset AxD. Other areas of the brain including the cerebrum showed normal white matter, ventricles and the absence of dark periventricular rim (Figure 2B). Genetic analysis corroborated an NM_002055.5(GFAP): c.592C>T (p. Arg198Trp) mutation, identical to patient 1. In silico analysis predicted a disease-causing mutation. One year later, he developed dysphagia but remained ambulatory. Speech therapy was started. Unfortunately, he was not eligible for an ongoing trial with antisense oligonucleotide therapy. No other family members had similar symptoms. AxD-associated variants in the GFAP gene located on chromosome 17 are typically single-nucleotide missense mutations, located throughout the coding sequence, which cause a toxic gain of function for the encoded protein [1, 5]. Approximately 95% of patients harbour mutations in the GFAP gene, and no other genetic causes are known [1, 5, 6]. More than 100 different mutations have been reported [1, 5]. Mutations result in the accumulation of intermediate filaments within the astrocytes, which in turn lead to large-scale molecular changes such as stress kinase activation, proteasome inhibition and changes in cell morphology due to the inability to complete cytokinesis and mitosis resulting in astrocyte dysfunction and white matter degeneration [6, 7]. Given these features, AxD can be considered as an astrocytopathy leading to leukodystrophy [6, 7]. In 2001, MRI criteria were proposed for the diagnosis of AxD: (1) extensive cerebral white matter abnormalities with a frontal preponderance, either in the extent of the white matter abnormalities, the degree of swelling, the degree of signal change or the degree of tissue loss (white matter atrophy or cystic degeneration); (2) presence of a periventricular rim of decreased signal intensity on T2-weighted images and elevated signal intensity on T1-weighted images; (3) abnormalities of the basal ganglia and thalami, either in the form of elevated signal intensity and some swelling or of atrophy and elevated or decreased signal intensity on T2-weighted images; (4) brain stem abnormalities, in particular involving the midbrain and medulla; and (5) contrast enhancement involving one or more of the following structures: ventricular lining, periventricular rim of tissue, white matter of the frontal lobes, optic chiasm, fornix, basal ganglia, thalamus, dentate nucleus and brain stem structures [8, 9]. Fulfilment of four of the five criteria was required for an MRI-based diagnosis [8, 9]. Since the analysis of the GFAP gene has become available for diagnostic confirmation, it has become clear that the MRI phenotype is broad [8, 9]. In families with autosomal dominant adult-onset disease, atrophy of the medulla and spinal cord can be the only or most important MRI findings, whereas no or minimal white matter abnormalities are present in the supratentorial region [9]. Quantitative evaluation of brain stem atrophy using MRI can be quite helpful. A sagittal diameter of the medulla oblongata <9.0 mm, medulla oblongata/midbrain <0.60 and sagittal medulla oblongata/pons <0.46 showed high sensitivity and specificity, suggesting that these parameters promote the accuracy of diagnosing AxD, even when a GFAP gene test is not available [10, 11]. Tumour-like lesions of the brainstem identified in patient 1 may be the only or predominant manifestation of AxD [10, 11]. This scenario can lead to misdiagnosis as an inflammatory lesion or low-grade glioma. Despite the range of variants and heterogeneity in presentation, the distinctive histopathological feature of the AxD is the widespread deposition of RFs in the cytoplasm of astrocytes in a subpial or perivascular distribution combined with variable loss of myelin [1]. RFs are ubiquitinated protein aggregates composed of the intermediate filament protein GFAP and a variety of proteins such as α B-crystallin and heat shock protein 27, cJun, ubiquitin, proteasome 20S, vimentin, nestin, plectin and synemin [2]. The role of neuroinflammation in human AxD is an area of mounting interest since Olabarria et al. demonstrated that mutations in GFAP are sufficient to trigger a profound inflammatory response in mouse models and human AxD CNS tissues [12]. The cellular stress caused by the accumulation of GFAP likely leads to the production of inflammatory molecules such as CXCL10 and CCL2, microglial activation and T-cell infiltrates [6, 12]. Despite the high levels of CXCL10 and CCL2, T-cell infiltration was mild and no B cells were found [12]. The presence of a florid B cell and plasma cell inflammatory infiltrate as seen in our case is of great interest since such a finding has not been reported in animal or human models as yet. It is a matter of speculation whether this finding is associated with the novel GFAP mutation akin to florid inflammation noted in the muscle in patients with limb-girdle muscular dystrophy such as dysferlinopathy [13]. The localised nature of the inflammatory cell infiltrate in the medulla is not supportive of a concomitant viral infection or other autoimmune conditions. The possibility of adaptive immune response is a consideration, especially since the role of B cells and humoral responses in neurodegenerative conditions, such as Alzheimer's disease, is an area of active research because of the potential targeted therapeutic implications [14]. Oxidative stress due to traumatic brain injury, infection or excessive alcohol consumption could act as a potential inducer of AxD pathology in individuals carrying GFAP mutations that have been asymptomatic until adulthood [2]. AxD is an autosomal dominant disease of variable penetrance. In most cases, the mutations occur spontaneously. Molecular genetic testing to identify heterozygous pathogenic variants in GFAP includes a combination of gene-targeted testing (single-gene testing and multigene panel) and comprehensive genomic testing (exome sequencing) depending on the phenotype [15]. A correct genetic diagnosis of AxD may have clear therapeutic implications. Long-lasting GFAP suppression after a single treatment with GFAP-targeted ASO (antisense oligonucleotides), with clearance of RFs and reversal of cellular pathology, was observed in proof-of-concept studies in AxD mouse and rat models [1, 2, 16]. Intrathecal administration of ION373 (Zilganersen), an antisense medicine targeting GFAP mRNA, is currently being investigated in an ongoing phase 1–3 trial [2, 16]. The identification of a novel NM_002055.5(GFAP): c.592C>T (p. Arg198Trp) variant in two unrelated individuals with clinico-pathological features consistent with adult-onset AxD suggests that this finding should be regarded as an AxD-associated pathogenic variant. Furthermore, a florid inflammatory component rich in B cells and plasma cells, as observed in our case, should prompt consideration of research into the role of humoral immunity in the pathogenesis of AxD, which remains an enigmatic disease despite decades of progress. Fouzia Ziad and Gert Cypers were responsible for the original draft preparation, data collection, curation and final draft preparation; Matthew Phillips, Satish Yadavraj, Piet Vanhoenacker, Arne Hostens, Duncan Lamont and Thomas Robertson were responsible for the supervision and final draft preparation. All authors approved the final version of the manuscript. The authors gratefully acknowledge Dr Albee Messing, VMD PhD, Waisman Center & Dept. of Comparative Biosciences. University of Wisconsin-Madison, Dr Chloe Stutterd FRACP. Clinical Geneticist, Victorian Clinical Genetics Service, Dr Patrick Yap, Clinical Geneticist, Auckland District Health Board and Dr Glenn Francis, Pathologist and Medical Director, Genomics for Life, Queensland Australia for guidance with genetic testing for AxD. The authors declare no funding or conflict of interest for this article. Informed consent was obtained for this publication. The peer review history for this article is available at https://www.webofscience.com/api/gateway/wos/peer-review/10.1111/nan.12927. The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions. Data S1. Supporting Information. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.