Brain 2005 128(11):2477-2479; doi:10.1093/brain/awh669
© The Author (2005). Published by Oxford University Press on behalf of the Guarantors of Brain. All rights reserved. For Permissions, please e-mail: journals.permissions@oxfordjournals.org
From the Archives
The pathological anatomy of disseminated sclerosis. By C. P. Symonds, M.D. OXON., M.R.C.P. Lond., Assistant Physician for Nervous Diseases, Guy's Hospital. (From the Wards of Guy's Hospital, and the Pathological Laboratories of Guy's Hospital and Lambeth Infirmary). Brain 1924: 47; 36–56 and Observatons on the histopathology of the cerebral lesions in disseminated sclerosis. By J. G. Greenfield and Lester S. King. (From the Pathological Laboratory of the National Hospital, Queen's Square, London). Brain 1936: 59; 445–458.
As a young man, (Sir) Charles Symonds—generally held as the outstanding diagnostician of his generation—wrote up the pathological findings in a single case of disseminated sclerosis admitted to Guy's Hospital under the care of Dr Arthur Hurst. Symonds and Drs Greenfield and King in their paper come straight to the point: despite an extensive literature, histopathological evidence forming the basis for ideas on the pathogenesis of disseminated sclerosis remains controversial, and several important features have been largely ignored.
For Symonds, the matter boils down to whether or not there is invariable lymphocytic and plasma cell infiltration in the lesions of disseminated sclerosis. If so, the disease can be considered to be of infective origin. But the available literature does not allow a firm conclusion since the histological identification of infiltrating cells is not made with sufficient clarity, and many cases are not convincing examples of disseminated sclerosis based on clinical and pathological evidence for lesions having developed in different places and at different times.
Mrs H.M. had relapsing and remitting symptoms affecting the brain and spinal cord. She died during a brain stem episode just under 2 years from onset, at the age of 39 years. At autopsy, lesions were present throughout the central nervous system. These show the expected regional distribution of disseminated sclerosis. Typically, at the centre of each is a dilated venule with a distended perivascular sheath containing a layer of lymphocytes and plasma cells, and also globular cells filled with fat granules interlaced by the dense strands of giant astrocytes. Beyond, the tissue is pale and vacuolated, but also contains fat granule cells and giant glia. At the periphery are more astrocytes than fat granule cells, the myelin sheaths showing different stages of degeneration as normal appearing tissue is reached. Some lesions show intense perivascular lymphocyte infiltration in the absence of fat granule cells. In others, the converse is true. For those lesions that are intensely gliotic, appearances differ between the brain, where there is astrocyte proliferation, and the spinal cord, which shows dense thickets of glial processes without an apparent increase in the number of cells. Thus, lesions of different ages are present and the histological criteria for diagnosis are met. The microscopic anatomy of the brain and brainstem closely matches that already described by James Dawson (The histology of disseminated sclerosis. Transactions of the Royal Society of Edinburgh 1916: 50; 517–740)—invariably a perivenular lesion, with accumulation of fat granule cells but also often showing lymphocyte infiltration extending to the edge of lesions when these are judged to be of recent origin (Figs 1
–3): ‘vascular engorgement and perivascular infiltration with lymphocytes and plasma cells occur in the acute phases of this disease’. And taking a stance that reflected thinking at the time, it follows that ‘it may be deduced that the cause of disseminated sclerosis is in all probability due to micro-organisms’. But none were found. Symonds's only comment on axons is to point out that these are often displaced by the fat granule cells, and so may appear to be reduced in number, but his impression was of continuity across lesions despite loss of the myelin sheaths (Fig. 3).
View larger version (125K):
[in this window]
[in a new window]
|
Fig. 1 Perivascular infiltration with lymphocytes. Focus in pons. Celloidin section, stained by Busch modification of Marchi method. Counterstained with carmine. Note the absence of fat granules.
|
|
View larger version (98K):
[in this window]
[in a new window]
|
Fig. 2 Perivascular infiltration with lymphocytes at the margin of a demyelinated area. Focus in pons. Celloidin section, stained by Kulschitzky-Pal method. Counterstained with carmine.
|
|
View larger version (93K):
[in this window]
[in a new window]
|
Fig. 3 Relatively normal axis cylinders running through the demyelinated area. Frozen section stained by Bielschowzky's method for axis cylinders.
|
|
Lester King was able to visit the Pathological Laboratory of
the National Hospital, Queen's (
sic) Square with support from
a Moseley Travelling Fellowship awarded by Harvard University
Medical School. With Dr J. G. Greenfield, he studied 125 plaques
identified in thirteen autopsy cases of disseminated sclerosis,
and applied a panel of stains to frozen sections: Scharlach
R and haematoxylin to show cells, myelinated fibres and lipoids;
Victoria blue for neuroglial fibres; Hortega's double impregnation
for neurofibrils; Hortega's method for connective tissue; the
Weil stain for myelin; and, wherever necessary. Gros, da Fano,
Perdrau and modified Cajal impregnations. The descriptions are
organized by pathological features.
The presence of cell-bound or free lipoid reflects the presence, or not, of demyelination. ‘Gitter cells’ are seen gathering at the periphery of the earliest lesions. Already, they contain large ‘Markball’ accumulations or smaller droplets of lipid, derived from myelin, and are also present as free material in the extracellular spaces where intact myelin sheaths appear to have been transformed without the direct mediation of cellular activity. Free lipoid represents detached myelin undergoing varying stages of chemical decomposition to fatty acids and cholesterol—the amount increasing with age of the lesion, except in cortical plaques. Greenfield and King sense that the patterns of staining they observe with Scharlach R—light yellow to dull rose, brick-red and brilliant light-red—reflect the shifting chemical composition of myelin undergoing ‘katabolism’, but they appreciate that new experimental methods are needed to confirm this interpretation. Next, there is a perivascular accumulation of lymphocytes, plasma cells and monocytes containing many fat droplets. These are mesodermal cells of blood and bone marrow origin—macrophages rather than tuberous or amoeboid microglia. Here in the perivascular space, the swollen, fat-laden ‘gitter cells’ now seem to disgorge their contents and new cells take up the lipoid. Later, the fat disappears as young microglia accumulate in the perivascular spaces. This is the conventional perivascular inflammatory and demyelinating plaque, known to observers for half a century. But the next feature had been a more closely guarded secret.
`In the demyelinating diseases, certain morphological changes in the axis cylinders are well known, especially diffuse oedematous swelling of the entire fibre, and local fusiform thickenings and node formation'. Largely ignored in the English literature, German and French neuropathologists had described loops, swellings, buttons and ball formations, variously interpreting these as regenerating or degenerating nerve fibres. Doinokow had not managed to study cerebral plaques and his descriptions related only to the spinal cord (Über De- und Regenerationserscheinungen an Achsenzylindern bei der Multiplen Sklerose. Zeitschrift Für die Gesamte Neurologie und Psychiatrie 1915: 27; 151–178). But King and Greenfield had better luck. They observed many fine calibered nerve fibres terminating in small loops with transition to a solid club, bombs, whorls and occasional cast-off isolated balls (Figs 4–6). Some, observed in shadow plaques, were assumed to be regenerating nerve fibres. Most were considered to be degenerative. Of these, the loops were especially common in cerebral plaques by comparison with previous descriptions of the spinal cord. Nonetheless, the authors do not agree with Tracy Putnam (Studies in multiple sclerosis. Archives of Neurology and Psychiatry 1936; 35: 1289–1308) that axonal pathology distinguishes multiple sclerosis from other demyelinating diseases. They consider him to have over-estimated its frequency, taking a 20% reduction in nerve fibre intensity as the yardstick of severe destruction. Gently, they dismiss Putnam's findings on the basis of imperfect technique, in which aniline blue staining of celloidin fixed samples was erroneously assumed to have identified neurofibrils whereas in fact it stains axoplasma: better to have used ammoniacal silver stains and frozen sections, Dr Putnam.
View larger version (183K):
[in this window]
[in a new window]
|
Fig. 4 Degenerative swellings ("torpedoes") on the axones (sic) of Purkinje cells in a cerebellar plaque. Hortega double impregnation for neurofibrils. x160.
|
|
View larger version (103K):
[in this window]
[in a new window]
|
Fig. 5 Mixed degeneration and regeneration of axis cylinders. The large ‘bomb’ is a retraction ball, from which fine lateral fibrils are sprouting. Hortega double impregnation for neurofibrils.x560.
|
|
View larger version (156K):
[in this window]
[in a new window]
|
Fig. 6 Terminal loops, the type of regenerative phenomenon most frequent in plaques in cerebral white matter. Hortega double impregnation for neurofibrils. x1200.
|
|
On gliosis, or proliferation of astrocytes and fibrous glia,
Drs King and Greenfield rehearse the prevailing English/American
versus French/German difference of opinion whereby this aspect
of the histopathology is considered to be a late or an early
feature, respectively. Indeed, batting for the French team,
Charcot had considered multiple sclerosis to be primarily a
disorder of astrocytes. But despite working at the headquarters
of British neurology, two lines of evidence also suggested to
Greenfield and King that there is indeed a simultaneous and
independent effect of the disease process on myelin sheaths
and on astrocytes: gliosis is observed early in the course and
at the same time fat-laden ‘gitter’ cells are in
abundance; and dense astrocytic proliferation may be seen in
lesions that display scant demyelination.
Finally, they describe and ponder the nature of two other cell types seen in the lesions of disseminated sclerosis. Rod cells had been described in the previous literature, but not at a time when their origins as microglia were understood. Although keeping the same company, the appearance of rod cells in the lesions of multiple sclerosis is distinct from phagocytes and ‘gitter’ cells. Spindle cells seem different: neither microglia nor fibroblasts, as previously designated, rather, they represent undifferentiated mesodermal cells. Their function and potential are unknown, although Greenfield and King feel instinctively that these are cells capable of repairing the lesions. Not for another 50 years did techniques become available that defined the phenotype and properties of remyelinating cells, enabling their distribution and performance in the lesions of multiple sclerosis subsequently to be characterized.
Alastair Compston
Cambridge
CiteULike 
Connotea 
Del.icio.us What's this?
