Article Text
Statistics from Altmetric.com
The study of Mathuranath et al 1 of corticobasal degeneration (CBD) and its overlap with frontotemporal dementia (FTD) contributes interesting information to a controversial area of neurodegeneration. Unfortunately it also introduces a potentially confusing histopathological diagnostic error. The authors found oligodendroglial inclusions which they considered to be glial cytoplasmic inclusions and claimed that these, diagnostic hallmark of multiple system atrophy, “have been described in other neurodegenerative diseases, including CBD”. This is clearly not the case.
To support their view the authors refer to two papers2 3reporting, among other cytoskeletal abnormalities, tau positive inclusions in oligodendroglial cells. A letter, published earlier in this Journal,4 but not quoted in the current paper,1 has also claimed that glial cytoplasmic inclusions are not exclusive to multiple system atrophy. The evidence is now overwhelming that they are. What the authors of this1 and the other papers2-4 have described are indeed tau positive oligodendroglial inclusions, but they are not the same as glial cytoplasmic inclusions.5Oligodendroglial inclusions, chiefly coiled bodies, undoubtedly occur in various neurodegenerative diseases, including CBD, but their morphology and molecular pathology are different from those of glial cytoplasmic inclusions. Whereas glial cytoplasmic inclusions immunostain with α-synuclein6 and only with unphosphorylated tau antibodies,7 the oligodendroglial inclusions seen in CBD and other neurodegenerative diseases are α-synuclein negative and give positive reaction with both phosphorylated and unphosphorylated tau antibodies. This basic difference has been recognised by a new classification of neurodegenerative disorders: CBD is one of the tauopathies, whereas multiple system atrophy is an α-synucleinopathy. Thus glial cytoplasmic inclusions remain the most consistent and reliable diagnostic hallmarks of multiple system atrophy and do not occur in other neurodegenerative diseases.
References
Xuereb and Hodges reply
The topic of how to apply the term “glial cytoplasmic inclusions” and their specificity to a particular disorder is clearly controversial and in a state of evolution. We used glial cytoplasmic inclusions in a non-specific way to indicate simply the presence of cytoplasmic inclusions in glial cells. We found these inclusions in subcortical oligodendrocytes. Glial inclusions were initially described in multiple system atrophy by Papp et al in 1989;1-1 their paper brought glial cellular pathology in neurodegenerative disease to the attention of neuropathologists. Cytoplasmic inclusions in glial cells have since been reported in various neurodegenerative diseases. The label “glial cytoplasmic inclusions” and the initials “GCIs” used in the general sense cannot, therefore, properly be regarded as pathognomonic of any single disease entity. Indeed, a neuroscientist without neuropathological training could conceivably misdiagnose tissue as coming from a case of multiple system atrophy if that tissue contained oligodendroglial cytoplasmic inclusions (in silver preparations or ubiquitin immunohistochemistry). Our paper does not, we would argue, contain a “histopathological diagnostic error” as suggested by Lantos.
On the other hand, Lantos' criticism that we made no mention of recent discoveries of α-synuclein involvement in the biology of MSA is justified. In the discussion, we should have emphasised the fact that glial cytoplasmic inclusions in multiple systems atrophy, and so far only in multiple systems atrophy, are indeed α-synuclein-positive and phosphorylated tau negative, whereas the opposite is true for the glial cytoplasmic inclusions of CBD and related tauopathies. Lest the future should see α-synuclein positive glial cytoplasmic inclusions identified in some other disease, it is well to emphasise that the diagnosis of multiple systems atrophy depends on the clinical history and distribution pattern of neurodegeneration (which determines the clinical phenotype), and the presence of α -synuclein positive glial cytoplasmic inclusions is valuable confirmatory evidence in this context.
We also appreciate Lang's comments on the paper and are very pleased that the cumulative Canadian experience mirrors the conclusions of our paper in that language and/or other cognitive disturbances are a virtually universal feature of CBD and in many cases may indeed be the mode of presentation as highlighted by the recent paper by Grimeset al 1-2 which appeared after the submission of our manuscript.