PRIMARY DEMYELINATION AS A NONSPECIFIC

CONSEQUENCE OF A CELL-MEDIATED
IMMUNE REACTION*
BY HENRYK M. WISNIEWSKIt AND BARRY R. BLOOM

(From the Departments of Pathology, Microbiology and Immunology, and Cell Biology, Albert
Einstein College of Medicine, Bronx, New York 10461)

Two types of demyelination are generally recognized and result from myelin
damage by exogenous or endogenous agents . Primary demyelination is character-
ized by segmental myelin damage without primary changes in the axon . In
secondary demyelination both the axoplasm and the myelin undergo degenera-
tion as in ischemic necrosis or axon death consequent upon dystrophic processes
or in Wallerian degeneration . In these conditions, myelin degeneration is a
secondary phenomenon, a sequela to death of the axon . In contrast, in primary
demyelination, myelin damage is the principal, and often only, lesion .
In considering the great number of pathological conditions in which primary
demyelination, i.e . segmental myelin damage without death of the axon, is
observed, two general patterns emerge . In some cases, myelin damage occurs in
the absence of inflammatory cells, for example in lesions produced by diphtheria
toxin, lead poisoning, and metabolic disorders (diabetes) . Of greater interest,
perhaps, are those pathological conditions in which inflammation is the
consistent feature associated with the demyelination. These include the human
demyelinating diseases such as acute disseminated encephalomyelitis, poly-
ganglio-radiculo-neuritis, and possibly multiple sclerosis, and the experimental
situations of experimental allergic encephalomyelitis (EAE)l, experimental
allergic neuritis and distemper encephalomyelitis (1) . While some of these
situations clearly involve viruses as primary etiologic agents, all are associated
with inflammatory type demyelination. However, in a viral demyelinating disease
such as progressive multifocal leukoencephalopathy histiocytic cells appear at
the advanced stage of myelin destruction and are presumed to remove already
damaged myelin (2) . In contrast, in virus-induced demyelinating encephalitis
caused by canine distemper (3), and perhaps in subacute sclerosing panencepha-

* This investigation was supported by grants no . 920-A-3 from the National Multiple Sclerosis
Society, from the Alfred P. Sloan Foundation, and grant AI 07118 from the National Institutes of
Health, U.S . Public Health Service .
$ Present address: MRC Demyelinating Diseases Unit, Newcastle General Hospital, Newcastle upon
Tyne NE4 6BE, England.
'Abbreviations used in this paper: EAE, experimental allergic encephalomyelitis ; PPD, purified
protein derivative ; SSPE, subacute sclerosing panencephalitis .

346 THE JOURNAL OF EXPERIMENTAL MEDICINE - VOLUME 141, 1975

 HENRYK M . WISNIEWSKI AND BARRY R. BLOOM 347

litis (SSPE) in man (4), hematogenous cells of the type seen in EAE appear at
the very early stages of nerve tissue damage . The mechanism of myelin damage
in all of these situations remains unclear. In some, it is possible that segmental
demyelination may be the result of the death of myelinated cells caused by the
infecting organism, e.g . the distemper virus . A second hypothesis holds that
whatever the initial exogenous trigger, inflammatory demyelination represents
destruction of autologous myelin by an autoimmune process.
A further possibility worthy of consideration is that damage to myelin could
occur as a nonspecific consequence of a specific cell-mediated immune reaction
occurring to an antigen in the vicinity of a myelinated nerve. In this situation
both primary (segmental) and secondary demyelination could occur as a
consequence of immunological reactions to a variety of different agents, perhaps
to viral antigens or mycobacterial antigens in neurotuberculosis and leprosy. It is
the purpose of this investigation to ascertain whether myelin can be damaged as
a nonspecific consequence of a specific delayed-type hypersensitivity reaction
directed at a nonnervous tissue antigen, in this case, tubercle bacillus or soluble
products thereof.

Materials and Methods

Both human and animal material were studied. The human specimens were cranial nerves and
samples of pons and medulla taken from two cases of tuberculous meningitis .
The animal material consisted of guinea pigs sensitized with 1 .5 mg killed tubercle bacilli in
Freund's complete adjuvant injected into the foot pads and nuchal muscles as described previously
(5). The animals were used 4-S wk after sensitization and were challenged with one of four sets of
antigens : purified protein derivative (PPD) (125 hg/0 .1 ml) ; Old Tuberculin (1 :200) ; sonicated HRa
(100 lAg/0.1 ml), and living avirulent tubercle bacilli, BCG or R,Rv (10'-10' viable organisms/0.1 ml) .
Under general anesthesia (nembutal) the skull was exposed and 2 mm posterior to the coronal and 2
mm lateral to the sagittal suture a burr hole was made . In the same animals the atlanto-occipitalis
membrane and both sciatic nerves were exposed and 0.1 ml of the antigen to be tested was injected
into the lateral ventricle; 0.2 ml into the cisterna cerebellomedullaris; 0.01 ml into the left peroneal
and tibial branches of the sciatic nerve. The right sciatic nerve served as a control and was injected
with 0.01 ml of saline solution . All injections were carried out with a 0.01 ml of microsyringe
(Hamilton Co ., Inc ., Whittier, Calif.) with a fine (no . 30) gauge needle which, for sciatic nerve
injection, was bent to an angle of 40° to facilitate entry. Injections into the lateral ventricle were
performed using stereotaxic apparatus . Antigen was injected by hand into the cisterna cerebel-
lomedullaris and peroneal and tibial branches of the sciatic nerve . In five guinea pigs, laminectomy at
L-5 or L-6 level of the spinal cord was performed and 100 u1 of PPD (125 fag) was injected bilaterally
into the posterior tracts and subarachnoid space. Each of the antigens used in the study was also
injected into nonsensitized, control guinea pigs at the same time . At 1, 2, 3, 4, and 10 days after
challenge, animals were anesthetized, heparinized, and killed by perfusion with 5% glutaraldehyde in
Sorenson phosphate buffer . Control animals were also sacrificed at the same time intervals . After
perfusion, samples of tissue for light and electron microscopic studies were taken from the following
areas: peroneal and tibial branches of the sciatic nerves, intracranial portion of II, III, V, VI, and VII
cranial nerves, C-1, L-5, L-6, and S-1 spinal cord with anterior and posterior roots, and samples of
brain tissue close to the lateral and IV ventricle and the base of the brain, together with the
leptomeninges. The tissue was postfixed with 2% Dalton's chrome osmium solution, dehydrated in
graded concentrations of ethanol, immersed in propylene oxide, and embedded in epon . 1-um sections
were cut and stained with toluidine blue . Thin sections of selected blocks and areas were cut, stained
with uranyl acetate and lead citrate, and examined in a Siemens electron microscope (Siemens Corp .,
Medical Industrial Div., Iselin, N.J .) . The formalin-fixed human material (III, IV, V, VI, VII cranial
348 DEMYELINATION AND A CELL-MEDIATED IMMUNE REACTION

nerves C-1 roots, and samples of tissue from the base of the brain) were embedded in paraffin for
routine histologic examination (Hematoxylin & Eosin, Bodian, and modification of Speilmeyer
methods), and in plastic for electron microscopic studies according to the methods described above.

Results

Injection of antigen into the subarachnoid space or ventricles of delayed-hyper-

sensitive guinea pigs invariably produced lesions exhibiting primary demyelina-
tion . Tuberculin-sensitized guinea pigs challenged either with Old Tuberculin,
PPD, live or dead tubercle bacilli displayed similar morphological changes . In
the majority of instances the presence of inflammatory cells was limited to the
surface of the ependyma and the subarachnoid space. On occasion, however, cells
aggregated in the Virchow-Robin space and around the entry zone of the cranial
nerves, and penetrated the brain parenchyma . In such areas, demyelinated
axons, axons undergoing demyelination, and myelin-laden macrophages were
observed . By and large the extent of demyelination was proportional to the
degree of inflammation . Axons showing Wallerianlike degeneration (secondary
demyelination) were also present particularly in foci with extensive inflamma-
tion . Peroneal and tibial branches of the sciatic nerves injected with tuberculin or
tubercle bacilli displayed extensive accumulation of inflammatory cells in the
interfascicular connective tissue . Although cells penetrated the perineurium only
rarely, once cells were found among the nerve fibers, demyelination was a
common picture. Control animals revealed only minimal cellular reactions along
the injection tracts, consisting of occasional polymorphonuclear leukocytes and a
few mononuclear cells .
Figs . 1 and 2 show the sciatic nerve with aggregates of myelin-laden
macrophages next to the demyelinated nerve fibers resulting from challenge with
tuberculin in tuberculin-sensitive guinea pigs . Around the vessels (Fig . 1) some
hematogenous cells without myelin debris were also seen . In the affected nerves,
as in the CNS, axons undergoing Wallerianlike degeneration were found .
However, they were seen only in areas of the most extensive inflammation . The
best defined areas of demyelination were observed in guinea pigs injected with
PPD into the posterior columns and the subarachnoid space of the lumbar spinal
cord . In these animals demyelinating plaques were seen outside the traumatic
zone in the vicinity of the subpial vessels with perivascular cuffs of inflammatory
cells. Fig. 3 illustrates the perivascular accumulation of inflammatory cells inter-
mixed with groups of demyelinated nerve fibers . Higher magnification of this
area (Fig . 4) shows naked neurons surrounded by macrophages with myelin de-
bris . Splitting of the myelin sheaths and the presence of dark droplets within the
boundaries of the nerve fibers are indicative of active demyelination (electron
micrographs of such nerve show that these droplets are myelin debris in the cyto-
plasm of macrophages which penetrated with myelin sheath) . Fig. 5 is an electron
micrograph of one such area showing extensive vesicular and netlike disruption
of the myelin sheath adjacent to myelin-laden macrophages. Fig. 6 demonstrates
active stripping of myelin lamellae by the mononuclear cell processes . The nor-
mal locking axoplasm of the nerve fiber which is undergoing active demyelination
should be noted . In areas of active demyelination, the plasma membrane of the
 HENRYK M . WISNIEWSKI AND BARRY R. BLOOM 349

FIGS . 1-4. These are 1 u, toluidine blue-stained sections from challenged tuberculin sensitive
guinea pigs .
FIG. 1 . Overall view of demyelination in sciatic nerve 72 h after challenge with killed,
sonicated tubercle bacilli. V, vessels, surrounded by macrophages and mononuclear cells.
Between the demyelinated axons (arrows) myelin laden macrophages are present. x 200.
FIG. 2 . Macrophages with myelin debris adjacent to naked (arrows) axons. Sciatic nerve as
above. x 560.

mononuclear cells was often "covered" with the disrupted and altered myelin
lamellae giving an appearance of "myelination" of the mononuclear cells (Figs .
7 and 8) .
Human cranial nerves obtained from patients with neurotuberculosis (Fig . 9)
350 DEMYELINATION AND A CELL-MEDIATED IMMUNE REACTION

FIG. 3 . Cuffs of inflammatory cells around the subpial vessels of the spinal cord, posterior
tract in animals 72 h after challenge with PPD. Arrows point to the groups of demyelinated
axons. x 150 .
FIG . 4 . Area of active demyelination from the above. Long arrows indicate nerve fibers
invaded by debris-laden macrophages . Short arrows show axons with splitting of the myelin
sheath . At the E/M level it was found that some splits correspond to areas where vesicular
disruption of the myelin took place; others, where pockets of the invading cell cytoplasmic
processes were interdigitated with the myelin lamellae . Double arrows point to denuded axons
surrounded by macrophages containing myelin debris . x 560.
FIG. 5. Extensive netlike disruption of the myelin sheath around normal-looking axon (A) in
area infiltrated by hematogenous cells. M, mononuclear cells with myelin debris . x 6,000.
FIG. 6. Active stripping of the myelin sheath by a macrophage . Arrow points to fingerlike
process stripping the remaining few myelin lamellae . x 10,000 .
351
352 DEMYELINATION AND A CELL-MEDIATED IMMUNE REACTION

FIG. 7. Mononuclear cell with myelin debris covered with altered myelin lamellae . Outer
myelin lamellae of the nerve fiber in the center appeared also to "fuse" with the membranes of
the macrophage . x 10,000 .
FIG . 8. Higher magnification of the area where the myelin lamellae are in very close
apposition to the plasma membrane of the macrophage . x 43,200 .
 HENRYK M. WISNIEWSKI AND BARRY R. BLOOM 353

were also extensively infiltrated by inflammatory cells . The same area, when
stained for axons (Fig . 10) showed that the majority of them were well preserved.
However, the stain of myelin (Figs . 11 and 12) demonstrated loss of myelin
sheath and active demyelination . Electron microscopic studies of these nerves
(Fig . 13) revealed that in areas infiltrated by the mononuclear cells, the myelin
sheaths had undergone vesicular disruption of the type seen in the experimental

FIGs . 9-13 . These are derived from human autopsy cases with neurotuberculosis .
FIG. 9. Extensive cellular infiltrate in the cranial nerve . Hematoxylin and eosin stain, x 150.
FIG. 10 . Similar area as above stained with Bodian technique to show that the majority of
axons in the affected nerves appear normal . x 150.
354 DEMYELINATION AND A CELL-MEDIATED IMMUNE REACTION

FIG. 11 . This photomicrograph demonstrates that in areas of cellular infiltrates the majority
of the nerve fibers lose their myelin sheath (black longitudinal lines correspond to remnants of
the myelin sheath) . Luxol fast blue and PAS. x 300.
FiG. 12 . 1 ym toluidine blue section showing partially demyelinated nerve fibers (arrows) . x
560.

model . It is interesting to note the presence of some plasma cells here, since they
are invariably observed in SSPE and multiple sclerosis plaques .

Discussion
When tuberculin-sensitive guinea pigs were challenged either with the tubercle
bacillus or with the tuberculin protein either in the central or peripheral nervous
Fig. 13 . Electronmicrograph taken from area of active demyelination . A mononuclear cell,
possibly a plasma cell, is present between two almost completely demyelinated nerve fibers .
Note that the remnants of the myelin sheath shows vesicular and netlike disruption similar to
that seen in experimental animals. x 6,000.
355
356 DEMYELINATION AND A CELL-MEDIATED IMMUNE REACTION

system, an inflammatory reaction ensued and demyelination invariably resulted .

Demyelination appeared to be proportional to the intensity of the inflammatory
reaction . In all cases, it was impossible to distinguish between reactions provoked
by living or sonicated killed tubercle bacilli, Old Tuberculin, or tuberculin PPD.
In the areas exhibiting most pronounced cellular inflammation, especially at
later times after challenge, the tuberculin reaction damaged not only the myelin
sheaths, but also the axons. A difference in intensity of inflammation and
demyelination between nerves and roots was seen, perhaps best explained by the
fact that peripheral nerves have a perineurium known to be a barrier to infection,
and perhaps inflammation (6, 7) . An essentially indistinguishable histopatholog-
ical picture emerged from observations on the brains of two patients who died of
tuberculous meningoencephalitis.
As early as 1932 Burn and Finley (8) reported that living or dead tubercle
bacilli and their products, when placed in direct contact with the leptomeninges
of tuberculous animals initiated a marked clinical and pathological reaction
characterized by weakness, convulsions, and death with intense inflammatory
exudate distributed throughout the subarachnoid space . A similar picture
including perivascular damage was reported by Rich and McCordock (9) as well .
A recent elegant series of histopathologic studies on human neurotuberculosis
presented by Dastur and associates described myelin destruction in the spinal
roots and perivascular demyelination in the white matter of the brain (10-12) .
Demyelination of the roots was attributed to compression (strangulation) (11) by
the leptomeningeal exudate. In the present experiments, a leptomeningeal
exudate was seen, but appeared to be a direct function of the cellular infiltrate in
the inflammatory reaction . However, the parenchymal perivascular demyelina-
tion they postulated to be a consequence of a delayed-type hypersensitivity
reaction probably occurred due either to brain antigens or to tuberculoprotein
antigens in the area (12) . The present studies are a direct confirmation of that
view .
We believe that the demyelination after a tuberculin reaction in the brain may
serve as a useful model for approaching other demyelinating diseases particularly
by permitting study of the purely immunological histopathology . It has most
obvious relevance to the origin of neurologic damage in CNS tuberculosis, and in
the tuberculoid form of leprosy, both brought about largely by cell-mediated
immune reactions. This is dramatically made clear by the excellent morphologic
studies of Sunderland (7), Job (13), Nishiura (14), as well as the more classical
studies of Wade (15) and Kanolkar (16) in leprosy in which a considerable degree
of pathology in this form of the disease appears to be a nonspecific consequence of
the delayed hypersensitivity reaction to the products of the M. leprae . In these
situations as well as in the present study, as a consequence of intensive
inflammatory reaction, there was often secondary demyelination resulting from
both myelin and axon destruction . The mechanism of myelin damage and
sometimes concomitant destruction of the axon remain unclear . Based on such
experimental models as EAE or EA neuritis, a number of mechanisms have been
proposed (17-20) : (a) myelin damage by either circulating or locally produced
antibodies ; (b) direct destruction of myelin by sensitized lymphoid cells; and (c)
myelin damage either by mediators produced by sensitized lymphocytes or
 HENRYK M. WISNIEWSKI AND BARRY R. BLOOM 357

secondarily by activated macrophages influenced by lymphocytes. In the present

model, the histologic picture suggests that the last of these mechanisms is likely
to be most significant . The sequence of events observed was as follows:
perivascular accumulation of hematogenous cells, infiltration of surrounding
parenchyma by cells from the perivascular cuffs, vesicular disruption and
stripping of the myelin lamellae by the hematogenous cells, and phagocytosis of
the myelin sheaths . The morphogenesis is thus similar to that seen in EAE (17),
and similar as well to lesions studied in distemper (3) and SSPE (4) encephalitis,
and suggests that some components of demyelination even in the viral-induced
conditions could result from a similar mechanism (1) . This model suggests that if
the exogenous source of antigen were removed, by appropriate treatment or
immunological means, the degree of neurologic damage would be limited, and in
the case of primary demyelination, could be reversible . This point stands in
contradistinction to demyelination which might occur as a result of an autoim-
mune process, in which even if the initiating antigen were removed, the autoim-
mune process would likely be progressive.
It has not yet been shown in experimental systems that a specific immune
response to an exogenous agent can initiate a secondary autoimmune EAE-like
encephalitis . Clearly in neurologic diseases associated with breakdown of nervous
parenchyma, detection of lymphocytes reactive to encephalitogenic protein
suggests that possibility . It is well known that in the absence of Freund's
adjuvant, for example, it is extremely difficult to induce EAE, e .g . Rivers et al .
(21, 22) had to give 60-120 injections of 5-10 ml brain homogenate to produce
disease . However, the possibility that sensitization to brain antigens may occur
as a result of persistent inflammatory reactions in nervous tissue is currently
under study.

Summary
Primary demyelination occurs in a variety of human and experimental diseases
known to be associated with the presence of inflammatory cells . However, the
mechanism of demyelination remains unclear. The possibility that myelin can be
damaged as a nonspecific consequence of a specific delayed type of hypersensitiv-
ity reaction directed at nonnervous tissue antigens was investigated .
Guinea pigs were sensitized to tuberculin with Freund's complete adjuvant,
and were challenged in the central and peripheral nervous system either with live
or killed sonicated tubercle bacilli, Old Tuberculin, or tuberculin purified protein
derivative (PPD) . Local inflammatory reactions were invariably produced and
primary demyelination was a constant feature of the lesions. The morphological
picture was rather similar to that observed in human neurotuberculosis and
early tuberculoid leprosy, and in experimental allergic encephalomyelitis and
distemper encephalitis in animals . The infiltrates consisted predominantly of
mononuclear cells with some polymorphonuclear cells as well . Vesicular disrup-
tion of the myelin sheath in the immediate vicinity of the inflammatory cells and
stripping of the myelin lamellae by the histiocytes without axonal damage were
the leading features of the lesion . The results indicate that cell-mediated
immune reactions to a variety of nonbrain antigens could be responsible for a
358 DEMYELINATION AND A CELL-MEDIATED IMMUNE REACTION

component of the demyelination seen in some inflammatory demyelinating

conditions, and suggest that this system may serve as a useful model for studying
the immunopathology of demyelinating disease .

We thank Dr.-Robert D . Terry for his support and constructive criticism . The authors appreciate the
expert technical, photographic, and secretarial assistance of Mrs . Carol Fitzgerald, Mr. Larry
Gonzales, Mr. Sidney Gravney, Mrs . Judith Gaffney, Mrs . Lenore Grollman, and Mrs . Roselyne
Schwartz . We express our appreciation to Doctors Marius Valsamis and Mauro Dal Canto for
providing us with the human autopsy material.

Received for publication 9 September 1974.

References
1 . Wisniewski, H . M . 1972 . Patterns of myelin damage resulting from inflammatory and
toxin-induced lesions and their relationship to multiple sclerosis . In Multiple
sclerosis, UCLA Medical Forum . Wolfgram and Andrews, editors . Academic Press
Inc ., New York. 53.
2. Zu Rhein, S. M. 1969 . Association of Papova-virions with a human demyelinating
diseases (Progressive multifocal leukoencephalopathy) . Prog. Med. Virol. 11:185.
3. Wisniewski, H. M., C . S. Raine, and W. J. Kay . 1972. Observation s on viral
demyelinating encephalomyelitis canine distemper . Lab . Invest. 26:589.
4. Herndon, R. M., and L . J. Rubinstein . 1968. Light and electron microscopy
observations on the development of viral particles in the inclusion of Dawson's
encephalitis (subacute sclerosing panencephalitis) . Neurology . 18 :8 .
5. Bloom, B. R., and B. Bennett . 1971 . In, In vitro Methods in Cell Mediated Immunity.
B. R. Bloom and P. R. Glade, editors . Academic Press, Inc., New York. 235 .
6. Kristensson, K., and Y. Olsson . 1973 . Diffusion pathways and retrograde axonal
transport of protein tracers in peripheral nerves . Progress in Neurology . Vol . 1, part 2.
Kerkut and J. W. Phillis, editors . Pergamon Press, Inc ., New York. 85 .
7 . Sunderland, S . 1973 . The internal anatomy of nerve trunks in relation to the neural
lesions of leprosy . Observations on pathology, symptomatology and treatment . Brain .
96 :866 .
8. Burn, C. G., and K. H. Finley . 1932. The role of hypersensitivity in the production of
experimental meningitis . J. Exp . Med . 56:203.
9 . Rich, A., and H . A. McCordock . 1933. The pathogenesis of tuberculous meningitis .
Johns Hopkins Med . J. 52:5.
10 . Dastur, D . K., and P. M. Udani . 1966. The pathology and pathogenesis oftuberculous
encephalopathy . Acta Neuropathol . 6 :311 .
11 . Dastur, D. K ., and N. H. Wadia . 1969. Spinal meningitides with radiculomyelopathy .
Part 2. Pathology and pathogenesis . J. Neurol . Sci. 8:261.
12. Dastur, D. K., and V. S . Lalitha . 1973. The many facets of neurotuberculosis : an
epitome of neuropathology . In Progress in Neuropathology, Vol . II. H. M . Zimmer-
man, editor . Grune & Stratton, Inc., New York. 351.
13. Job, C. K. 1970 . Mycobacterium leprae in nerve lesions in lepromatous leprosy . Arch.
Pathol . 89:195 .
14 . Nishiura, M . 1960. The electron microscopic basis of the pathology of leprosy. Int . J.
Lepr. 28:357 .
15. Wade, H. W. 1934 . Tuberculoid changes in leprosy . III. The pathology of a nerve
abscess . Int . J. Lepr. 2:293.
16. Khanolkar, V. R. 1964. Pathology of leprosy. In Leprosy in Theory and Practice . R. G .
 HENRYK M. WISNIEWSKI AND BARRY R . BLOOM 35 9
Cochrane and T . F. Davey, editors . The Williams and Wilkins Company, Baltimore,
Md . 2nd edition . 125 .
17 . Bornstein, M . B. and Iwanami, H . 1971 . Experimental allergic encephalomyelitis .
Demyelination activity of serum and sensitized lymph node cells on cultured nerve
tissues . J. Neuropath . Exp . Neurol . 30 :240,
18 . Wisniewski, H ., J . Prineas, and C . S. Raine. 1969 . An ultrastructural study of
experimental allergic encephalomyelitis in the peripheral nervous system . Lab .
Invest . 21 :105 .
19 . Arnason, B . G . W ., G . F . Winkler, and N . M . Hadler . 1969 . Cell-mediated
demyelination of peripheral nerve in tissue culture . Lab . Invest . 21 :1 .
20 . Johnson, A. B., H . M . Wisniewski, C. S . Raine, E. H . Eylar, and R. D . Terry . 1971 .
Specifi c binding of peroxidase-labeled myelin basic protein in allergic encephalomye-
litis . Proc . Natl . Acad . Sci . U .S.A . 68:2694 .
21 . Rivers, T . M ., D. M . Sprunt, and G. P. Berry . 1933 . Observation s on attempts to
produce acute disseminated encephalomyelitis in monkeys . J. Exp . Med . 58 :39 .
22 . Rivers, T. M ., and F. F. Schwentker . 1935 . Encephalomyeliti s accompanied by myelin
destruction experimentally produced in monkeys. J. Exp . Med . 61 :689 .