have been, in our experience, practically as constant. In all of the
eleven acute cases which we have been able to examine, there has
been more or less extensive involvement of the lymphoid tissue and
of parench)rmatous organs. The lymphoid tissue throughout the
body appears to react to the virus. The Peyer's patches of the
intestine and the mesenteric lymph glands show perhaps the most
marked acute swelling. The mucosa over the Peyer's patches is,
Digitized by
Google
Francis W. Peabody, Oeorge Draper, and A. B. Dochez. 23
however, unaffected. There is also definite, and sometimes pro-
nounced enlargement of the substernal, bronchial, cervical, axillary,
and inguinal lymph glands and of the tonsils. The spleen is fre-
quently somewhat enlarged, and on section the Malpighian cor-
puscles stand up in raised, pale, obviously translucent nodules. The
thymus shows changes identical with those in the lymphoid tissue
elsewhere. On histological examination, one is struck by the resem-
blance of the lesions found to those described by Mallory^ in.
typhoid fever. The reaction is, in general, the same throughout the
lymphoid tissue, regardless of its location. On histological exami-
nation some of the lymphoid nodules may present a normal appear-
ance, but the majority consist of a zone of lymphocytes surrounding
a more or less sharply circumscribed pale center (figure 12). High
magnification shows the center of the lobule to consist chiefly of
large endothelial cells with oval vesicular nuclei. These cells are
similar to the cells lining the lymph sinuses, but most of them are
larger, more swollen, and take the stain very lightly. Sometimes
the nuclei look like pale shadows, and the outline of the protoplasm
is so faint that it can scarcely be distinguished. Where they are
closely packed, the individual cells appear to be fused together to
form compact masses. The better preserved of these cells are
markedly phagocytic and frequently contain many particles of
necrotic cells. These cell inclusions are surrounded by a lightly
stained halo and are apparently situated in vacuoles in the proto-
plasm. Scattered throughout the center of the nodule are many
broken down cells and granular fragments of necrotic nuclei. The
cells which are going to pieces are for the most part lymphoc)rtes,
but the endothelial cells also seem to swell up and finally disin-
tegrate. In areas with extensive necrosis there is often an invasion
by polymorphonuclear leucocytes.
In the lymph sinuses there are also large numbers of the same
phagocytic endothelial cells. Many of them are of great size and
contain necrotic fragments of nuclei, whole lymphocytes, or num-
bers of red blood corpuscles. In the l)miph sinuses, there is exten-
sive proliferation of the endothelial cells, as is evidenced by the
frequency with which mitotic figures are found. Numbers of ne-
"Mallory, Jour, Exper, Med., 1898, iii, 611.
Digitized by
Google
24 A Clinical Study of Acute Poliomyelitis.
erotic cells are met with in the lymph sinuses, but, in general,
necrosis is a more prominent feature in the centers of the lymphatic
nodules, and proliferation in the lymph sinuses.
Among the parenchymatous organs, "cloudy swelling," such as
has been frequently described, is usually met with. In the liver,
however, there are other more striking changes, and these again
resemble in character those that have been described by Mallory and
.others in typhoid fever. They are for the most part sharply circum-
scribed areas, ranging in size from lesions which consist of one or
two cells to others which include nearly one eighth of a liver lobule,
in which degeneration of liver cells and infiltration of lymphoid
cells and polymorphonuclear cells have taken place (figure 13).
The number of these necroses is very variable. They may be rather
difficult to find, or there may be many of them in a single low power
field. The lesions are apparently closely associated with the blood
vessels, and while chiefly with the portal vein, they are also asso-
ciated with central or sublobular vessels. The section is, however,
sometimes cut so that the relation to the blood-vessels is not directly
evident. The lesions are round or oval in cross section, but there
may be long finger-like projections, extending out from the main
area, and involving one or two columns of liver cells. It is striking
that the liver cells directly adjacent to the infiltrated areas are usu-
ally completely spared, but occasionally one finds that some of them
show a homogeneous quality of the protoplasm, and an exaggerated
affinity for eosin, which suggest a beginning involvement through a
peripheral spread of the lesion.
The character of the lesion varies, of course, with its size and age,
but in general it is very constant. In the earlier lesions one finds
one or two disintegrating liver cells, with homogeneous, hyaline,
pink-staining protoplasm, and occasionally with irregular, distorted
nuclei. In the older lesions, even more apparent than the necrotic
liver cells is the evidence of an early and rapid proliferation.
Scattered through practically all the lesions are small groups of liver
cell nuclei, frequently a nest of three or four nuclei surrounded by
infiltrated tissue, or several nuclei apparently beginning to pro-
liferate within a degenerated liver cell body. The protoplasm around
these dividing nuclei may appear as a faint pink halo, or it may be
Digitized by
Google
Francis W. Peabody, George Draper, and A. B. Dochez. 25
quite impossible to make out any protoplasm at all. In older lesions,
the protoplasm is better defined and the cells have frequently-
arranged themselves circularly, in acini, or in double columns. This
arrangement of cells simulates that in normal bile ducts, but it is
found in association with nearly all the lesions, irrespective of their
position in the lobule, although most frequently adjacent to the
portal spaces. According to the position of the infiltrative lesion,
much or little fibrillated tissue appears among the degenerating and
infiltrating cells. The greater amount of fibrillated tissue occurs in
and adjacent to the portal spaces, where all the lesions of an early
cirrhosis may be displayed. That this condition is of the nature
of a beginning cirrhosis is further indicated by the microscopic de-
pressions of the capsule of the organ, where the lesions extend to-
ward and reach that structure. However, the process probably is
not permanent, since it is so young and cellular that it can readily
undergo retrogression ; while the infiltrated uodules elsewhere in the
lobules are surely easily subject to restoration. One case has indeed
come to autopsy, the child dying of a laryngeal diphtheria two
months after the onset of poliomyelitic paralysis affecting both legs.
In the liver of this case were remains of the portal infiltrative lesions
described and, in addition, younger necrotic and infiltrated lesions
among the columns of liver cells. The former probably had been
greater and were diminishing; it is uncertain whether the latter
were residues of the poliomyelitic infection or the result of the
recent diphtheritic process. That these changes in the lymphoid
tissues and in the liver are, in fact, a part of the reaction of the body
to the virus of poliomyelitis, would seem to be made certain by the
fact that exactly similar lesions may be found in the organs of
monkeys which have been experimentally infected with the disease
(Flexner27).
The demonstration of such a widespread reaction to the virus is
wholly in line with recent clinical and epidemiological advances
which tend to recognize acute poliomyelitis as a general infection.
The disease must be regarded as a generalized process which affects
parench5miatous organs, lymphoid tissue, and more especially the
nervous system. It is possible that two distinct effects of the disease
^Flexner, Folio serolog., 1911, vii, iioi.
Digitized by
Google
26 A Clinical Study of Acute Poliomyelitis.
on the organism should be differentiated. On the one hand is the
general toxemic process which affects organs throughout the body,
but which apparently acts mildly. On the other hand is the local
process in the spinal cord, producing death by destruction of the
nerve cells controlling respiration. The anatomical findings which
point to an action of the virus of poliomyelitis on the lymphoid
tissue throughout the body may well be correlated with the results
obtained by the inoculation of emulsions of lymphoid organs into
monkeys. Not long after the disease was first transferred to
monkeys by the intracerebral inoculation of filtrates of the spinal
cord, the virus was proved, in the same manner, to be present in a
mesenteric lymph gland. Since then similar positive results have
been obtained with still other lymphatic glands in the monkey, and
with emulsions of the tonsils in man and the monkey.
These observations are not, however, of the same significance and
importance. What they tend to indicate is that the living virus may
come to rest for a time in organs outside the central nervous system,
to which it is conveyed by the blood. It does not establish the fact
that the visceral lesions outside the nervous system are caused
directly by the virus, rather than by some secondary toxic substance
produced in the course of its proliferation. Since the only means at
present available to demonstrate the presence of the virus is the in-
oculation of monkeys, the tests made are too few to determine how
widespread in the organs the virus really is. It is established that
it is present in the central nervous system, even when it cannot be
demonstrated in the viscera generally. But the finding of the virus
in the tonsils and nasal mucosa with as great constancy as in the
nervous system indicates that these organs play a part in the con-
veyance of the virus into and away from the central nervous system.
The virus is regularly present in them in fatal and doubtless in non-
fatal cases of poliomyelitis in children, and it has been demonstrated
in the nasal mucosa, the tonsils, and even in the nasal mucous mem-
brane in infected monkeys; it finds its way into the nasal mucous
membrane, even when injected into the peritoneal cavity. The evi-
dence, therefore, is strong that, as was first pointed out by Flexner,
the upper respiratory mucous membrane provides for both the
ingress and egress of the virus of the disease, through which infec-
Digitized by
Google
Francis W. Peabody, Oeorge Draper, and A. R. Dochez. 27
tion is produced and the renewal of the cause of the disease main-
tained.
SYMPTOMATOLOGY.
CLINICAL CLASSIFICATION OF CASES.
The Study of any disease is simplified if the various clinical forms
in which it may appear can be grouped together into a few definite
types. While no two cases of any disease are ever exactly similar,
one probably finds variety of g^oss clinical manifestation most strik-
ingly illustrated in the diseases of the nervous system, and a patho-
logical process which may extend more or less continuously through
brain, medulla, pons, spinal cord, and spinal ganglia, or which may
be localized in any part of this system may produce signs and
symptoms of very diverse character. Such are the possibilities that
are present in poliomyelitis, and it is no wonder that the clinical
classification of cases has been the subject of much discussion. The
most generally accepted system has been that of Wickman, who
recognizes eight forms: (i) the spinal, poliomyelitic form, (2) the
cases simulating Landry's paralysis, (3) the bulbar or pontine form,
(4) the encephalitic form, (5) the ataxic form, (6). the neuritic
form, (7) the meningeal form, and (8) the abortive cases. While
such a classification certainly covers all possible cases and makes it
easy to place most cases in their appropriate groups, it does not ap-
pear to us to be wholly satisfactory. It is based neither on patho-
logical anatomy, nor on clinical symptomatology, but on a mixture
of the two. The spinal, the bulbar, and the encephalitic forms are
anatomical forms ; the neuritic, the meningeal, and the ataxic forms
are essentially symptomatic forms. A more satisfactory classifica-
tion would be either anatomical or clinical. Such a classification
would simplify a confusion arising from the use of two systems, for
it is impossible to draw any hard and fast line between some of
the groups. Thus most neuritic forms are essentially of the spinal
type, and most meningeal cases are either spinal or bulbar. The
ataxic form Zappert regards as based on a single, not especially
prominent symptom, rather than on an anatomical foundation. The
cases simulating Landry's paralysis are in the main instances of the
spinal type in which the process advances, usually to end in death
Digitized by
Google
28 A Clinical Study of Acute Poliomyelitis.
from involvement of the respiratory muscles. Cases that are
classed among the meningeal or neuritic forms are so classed be-
cause they show some of the clinical manifestations of meningitis
or of neuritis, rather than because they have a pathology which is
essentially different from other bulbar or spinal cases. These types
of the disease are, in fact, not true types at all. They merely repre-
sent some of the variations in symptomatology which may be found
more or less developed in association with either the abortive, the
bulbar, the spinal, or the cerebral types of the disease, and their
enumeration as definite forms is complicating and at times confus-
ing. As a result of this duplicating and overlapping of sub-
divisions, the picture of the disease, as a whole, is made much more
complicated and obscure than is necessary.
More simple and, at the same time, more practical is the classifica-
tion proposed by Miiller. Adopting, as far as possible, an ana-
tomical basis, he proposes the four following classes : ( i ) the spinal
form, (2) the bulbar form, (3) the cerebral form, and (4) the
abortive cases. In any system of classification, the abortive cases
must, of course, form a distinct type. In considering the cases,
however, which result in paralysis, one would seem to get a clearer
outline of the disease process, if the cases are looked at from the
point of view of pathological physiology and divided into two main
groups: (i) those cases in which the upper motor neurone is
primarily affected, and (2) the larger group of cases in which the
lower motor neurone is involved. The first group is the cerebral
form of Miiller, and the second group is composed of his spinal and
bulbar forms. The two latter forms it is wiser to consider together,
as in both the lesion is essentially the same, a lower motor neurone
lesion involving, on the one hand, the nuclei of the pons or medulla
and, on the other hand, the anterior horn cells, the result being in
either instance a flaccid paralysis with subsequent atrophy. The
two types which are thus based on similar pathological lesions are
also inseparable clinically, for a somewhat larger number of cases
occurs with cranial nerve involvement, in association with spinal
lesions, than occurs with cranial nerve involvement alone. The
third group of cases in Miiller's classification is the cerebral type, the
same group to which Wickman gives the name encephalitic. This
Digitized by
Google
Francis W. Peabody, George Draper, and A. B. Dochez. 29
is a fairly definite pathological entity, at least in so far as the most
prominent lesion apparently involves the upper motor neurone.
While the bulbospinal type is characterized by flaccid paralyses, this
cerebral type is characterized by spastic paralyses or by other evi-
dence of a lesion above the lower neurone. The exact location of
the pathological processes in these cases is not wholly clear, but it
seems probable that the upper neurone may be attacked in either or
both of two places. Some few cases are probably due to lesions
in the cortex with involvement of much larger areas than are com-
monly found. Other cases, the majority, seem to depend on in-
volvement of the pyramidal tracts, either high up or in the cord.
That this class of case, originally described by Striimpell as polioen-
cephalitis acuta is very uncommon, but that it is in reality a form
of infantile paralysis, has been made quite certain by epidemiological
studies which show the spastic type occurring in the same epidemic
and in the same house with the flaccid types, by the occasional
occurrence of both spastic and flaccid paralyses in the same patient,
as well as by the pathological studies of Harbitz and Scheel.
It would seem, then, that the best appreciation of acute polio-
myelitis, from the clinical point of view, is obtained if one recognizes
three groups of cases. The first group consists of the abortive
cases, cases of infection which never become paralyzed. The second
or cerebral group contains the rare cases in which involvement of
the upper motor neurone with resulting spastic paralysis is the chief
characteristic. The third or bulbospinal group is much larger and
comprises all cases with lesions in the lower motor neurone and
flaccid paralyses.
Such a classification is, of course, open to the objection that many
cases are not purely of one type, either anatomically or clinically.
Thus most bulbospinal cases show some pathological foci in the
brain, and cerebral cases may show lesions extending into the cord
(Harbitz and Scheel). Then from the clinical point of view a
certain number of bulbospinal cases develop sjrmptoms as, for in-
stance, ataxia, which may well depend on an upper motor neurone
lesion, either in the brain or cerebellum, in the pyramidal tracts, or
in Clarke's columns. When one recognizes how diffusely the patho-
logical process extends through the whole central nervous system.
Digitized by
Google
30 A Clinical Study of Acute Poliomyelitis.
it becomes evident that it is impossible to make any hard and fast
classification which shall be strictly applicable to all cases. The
best one can do is to attempt to reconcile the chief clinical symptoms
with the predominant anatomical lesion. In the following sections,
we shall consider the s)miptomatology and course of each of these
types in detail.
PREDISPOSING CAUSES.
Acute poliomyelitis is a disease which shows very definite sea-
sonal variations in its incidence. The records of epidemics in many
countries show that it occurs during the summer and reaches its
maximum in the late summer and early autumn. In the Swedish
epidemic of 1905, 86 per cent, of the cases had their onset between
July and October, and 35 per cent, during August. The New York
epidemic of 1907 showed a very similar curve, reaching its maxi-
mum, however, in September. Other epidemics show some slight
variations in duration and in the period at which the highest point
is reached, but that the disease is distinctly one of summer and fall
has been generally accepted. There are a few well substantiated
epidemics which prove that poliomyelitis may also occur during cold
weather. Wickman cites one epidemic in the north of Sweden, last-
ing through the winter, and reaching its maximum in April and
May. He also reports three closely related epidemics, occurring in
neighboring sections of the country, the first (eighteen cases) lasted
from June to October, the second (twenty-seven cases), from July
to December, and the third (sixty-two cases), from the end of
September to February, with its maximum in November and De-
cember. These winter epidemics are of interest in relation to
attempts that are being made to account for the spread of the disease
by some intermediary insect host. Stress has been laid on the fre-
quent occurrence of epidemics of poliomyelitis in unusually hot, dry
years, but this relationship is certainly far from constant.
It has been generally noticed that acute poliomyelitis is a disease
of open country, rather than of cities. Among our own cases, a
relatively larger number of children came from the suburbs and
surrounding country than from the densely populated tenement
district in which the hospital stands. Moreover, a considerable
Digitized by
Google
Francis W. Peabody, George Draper, and A. B. Dochez. 31
proportion came from families which are in comparatively well-
to-do circumstances and in which the children enjoy every comfort
and care. Pneumonia, measles, and other acute diseases are often
mentioned as predisposing causes of poliomyelitis, but in our experi-
ence the absence of any history of previous illness has been much
more noticeable. In general, they had been perfectly healthy
children.
Childhood is essentially the age which is most susceptible to polio-
myelitis, but that adults are by no means immune is shown by Wick-
man's statistics. More than one fifth of his cases were persons over
fifteen years old. He mentions one case in a man of forty-six
years, and instances of the disease occurring in even older persons
are on record. In the New York epidemic of 1907, the youngest
case was two weeks old. Three cases seen by us were three months
old, and fifteen cases were between six and twelve months. As in
the 1907 epidemic in New York, the majority of our cases were
between the ages of one and three years. The figures obtained by
Miiller agree rather remarkably with ours in respect to age incidence.
Of his cases 96 per cent., and of ours 97 per cent, were in the first
decade, while 90 per cent, of his, and 89 per cent, of ours were
below the age of five. Perhaps the age most liable to infection is
the latter half of the second year. Both sexes are almost equally
susceptible to the disease, the number of males being slightly greater
than the number of females.
TABLE OF AGE
INQDENCE.
Month*.
Vwn.
Under
6
N
I
r
I
I
I
X
I
X
1?
k
i
M
r
Over
as
Total.
Rockefeller In-
stitute Hos-
pital, Z9ZX..
3
15
62
221
29
180
18
13
6
2B
4
18
II
a
II
I
7
3
14
5
I
z
157
New York
Epidemic,
1907
6a
106 63
3
729
Wickman,
Sweden,
1905
183
214
179
2a9
aao
1.025
Digitized by
Google
32 A Clinical Stu^y of Acute Poliomyelitis.
THE PRODRCaiiAL PERIOD.
In most typical cases of infantile paralysis, the course of the
disease is fairly constant. A period of incubation is followed first
by a period of prodromal symptoms, then by an acute stage with
paralysis, and finally by a stage of retrogression. The duration of
the incubation period is variable. In the experimental disease in
monkeys, Flexner has found that the time elapsing between inocula-
tion and the onset of paralysis is from three or four days to thirty-
three days, the average being eight or nine days. Prodromal symp-
toms are of short duration in the monkey, rarely lasting over twenty-
four hours. With a virus that has been passed through several
generations of monkeys, and has thus become adapted to its new
host, the period of incubation is much more constant than it is
on the first transfer from man. Such a fixed virus causes paralysis
with great regularity in seven or eight days. In the disease in
man it is obviously much more difficult to determine the exact
length of the incubation period. Thus, even when two cases
appear in one family, it is often impossible to be certain at what
time the second case became infected by the first, or even whether
both were not infected at the same time from a common source.
Wickman considers the incubation period as being from one to four
days, his opinion being based on the interval between the dates of
onset of the disease in two persons of the same family. Our own
observations of families in which two cases have occurred would
lead us to a nearly similar conclusion, but we believe that these are
in all probability simultaneous infections from a common source.
At any rate, previous contact had been so intimate that infection
might have taken place at any time during the incubation period of
the first case. In one family at least, two children became para-
lyzed on the same day. Miiller has seen six instances in which the
length of the incubation period could be determined more accurately,
and he concluded that it varies between five and ten days. The
average incubation period was about a week, thus practically the
same as is found in the experimentally produced disease.
Following the incubation period and preceding the onset of