sphincter reflex, with frequent stools. The tumor in this case is
reported at the level of the tenth dorsal vertebra, which would be at
the level of the eleventh dorsal segment. The tumor was tubercular,
and therefore probably meningeal, so that the irritation to the anal
centre, which is in the lumbo-sacral segment, may have been caused by
extension of inflammation along the meninges.
PATHOLOGY.—We present in tabulated form the various kinds of tumors as
found in the fifty cases which have been collected: {1097}
Aneurism . . . . . . . . . 1 | Myxoma . . . . . . . . 2
Cancer (?) . . . . . . . . 2 | Neuroma . . . . . . . 1
Carcinoma . . . . . . . . 1 | Organized blood-clot . 1
Cysticercus and hydatids . 3 | Osteoma . . . . . . . 1
Cyst (Dermoid?). . . . . . 1 | Phlegmon . . . . . . . 1
Fibroma . . . . . . . . . 5 | Psammoma . . . . . . . 2
Glioma . . . . . . . . . . 5 | Sarcoma . . . . . . . 7
Gumma . . . . . . . . . . 5 | Tubercle . . . . . . . 4
Myo-lipoma . . . . . . . . 2 | Unclassified . . . . . 6
It will be observed that the predominance in this list is decidedly in
favor of the sarcomata and structures which are likely to be
associated with or to graduate into them, such as the gliomata,
myxomata, and the psammoma. In one instance the resemblance (Case 15)
to psammoma is referred to by the reporter. The comparatively large
number of unclassified, and the two cases referred to vaguely as
cancer, would probably, on more exact report, have added several more
to the group of the sarcomata. The table shows that next in frequency
come the fibromata and gummata, while the carcinomata have but a
single representative in the group. Tubercular tumors occurred with
comparative frequency, no less than 8 per cent. being recorded.
Although the exact origin of only somewhat more than one-half (29) of
all the tumors is given, it is stated of this fraction that 17 sprang
from the membranes, while of the remainder 8 were located in the cord
itself and 4 in the vertebræ. Of the 3 oases of parasitic invasion, it
is recorded of one (Case 43) that a hydatid cyst was also found in the
liver; and it is probable that in any given case the spinal cord would
not be the only part to suffer. Cobbold's work refers to one case of
hydatid of the spinal cord. Erb refers to 13 cases, all but 2 external
to the dura mater. In Case 43 of the table pains in the back and hip,
simulating rheumatism, were present early.
The dimensions of the spinal tumors of whatever character are never
very great, for the reason that they have but little space in which to
enlarge, and that their presence soon causes such grave changes as to
be incompatible with life. They rarely exceed an inch in their longest
diameter, and not unfrequently are smaller than this. There is usually
about them a more or less marked meningitis and an area of vascular
fulness. Œdema of the membranes is sometimes noted. The substance of
the cord beneath is compressed, atrophied, or softened, and this
softening sometimes extends for a considerable distance both above and
below the neoplasm. These changes were reported in Case 8, in which
there was the addition of an abscess. Secondary degenerations would
probably be found in all cases, unless very recent; and these changes,
following the Wallerian law, would ascend the posterior and descend
the lateral columns. Such degenerations are reported in some of the
cases. The spinal nerves are sometimes compressed and atrophied. Old
or recent hemorrhages are found, as in Case 18. Among the changes
which occur, probably at a late stage, are the formation of cysts,
either large or small, either in the substance of the cord or
consisting simply of a dilatation of the central canal of the cord.
This condition is known as syringo-myelia, and is of exclusive
pathological interest. In Case 1 is recorded, apparently, a
well-marked dilatation of the central canal (hydromyelus), and
cyst-formation is recorded with gliomata, sarcomata, and gummata in
other cases. Caries of the vertebræ is recorded in a number of cases,
and occurred both in cases of gumma (Case 8) and carcinoma (Case 14).
These cases were, however, exceptions to the general rule that the
bony envelope of the cord does not furnish external {1098} evidence of
the location of the tumor. It is worthy of note that the one instance
of phlegmon or inflammatory exudate (Case 20) also presented
infiltration of the tissues of the throat and mediastinal space. A
case of organized blood-clot (Case 39) has been included in the list,
although, properly, a spinal hemorrhage, because it became and acted
as a tumor. The location of the single case of aneurism (Case 48) is
not given. The symptoms were those of tumor in the dorsal spine.
As sequelæ of tumors of the cord may be mentioned especially
bed-sores, which sometimes commit frightful ravages, as in a case (37)
in which the spinal canal was laid open. The bronzing of the skin and
diseased condition of the suprarenal capsules, as recorded in Case 41,
were mere coincidences, and not probably at all connected in
pathological sequence with the spinal lesion. Cystitis, pyelitis, and
pyonephritis are not uncommon in cases of tumor of the cord, just as
they are observed in other compressing and destructive lesions of that
organ. In those cases in which the tumor is the result of a general
taint, as in gummy and tubercular growths, the evidence of this taint
is not usually wanting in other organs; thus in Case 37, of tubercle
of the cord, tubercles were also found in the lungs, bowels, and
uterus. In gummata of the cord it would not be likely to escape
careful inquiry that the patient's history or his body presented
evidence of the disease.
In a case (47) of congenital sacral neuroma amyilinicum the infant was
also hydrocephalic and had a bifid spine—conditions of faulty
development with which the patient cannot long survive.
In one case of psammoma (Case 5) a resemblance to endothelioma is
noted, while in another (Case 15), already referred to, a sarcoma is
said to have resembled a psammoma.
Vascular changes are usually notable. In addition to the congestion
already spoken of, it is recorded in one case (No. 14) that the right
vertebral artery was obliterated. This was a carcinoma which had
partly destroyed one vertebra.
Virchow's case (No. 50) of a stillborn child with a large tumor of the
size of the head of a child of two years, and containing bone, has
some analogies in three cases, referred to by that author, in which
both hair and bone were found.
In several cases a brown or yellow exudate (plastic lymph?) is
mentioned as extending along the cord far beyond the immediate
neighborhood of the tumor.
The histology of tumors of the whole cerebro-spinal axis will be found
described in the article on Tumors of the Brain.
DIAGNOSIS.—The diagnosis of tumors of the spinal cord presents itself
naturally under two heads—the differential, or general, and the local
diagnosis. The conclusions reached in this paper are based on a
careful study of the cases appended, two of which were personal
observations, and the remainder were collected from American, English,
French, and German literature. It cannot be denied that much obscurity
rests upon the diagnosis of tumors of the spinal cord, and that the
doubts expressed by Erb and other writers have much to support them.
It is hoped that the systematized study presented in this table will
do something to dispel this obscurity.
(1) General Diagnosis.—The differential diagnosis has regard first to
{1099} certain general phenomena which are broadly indicative of a
spinal disorder as distinct from a cerebral or peripheral one. Thus,
mental symptoms are absent, or if present are an accompaniment of
tumors high up in the spinal axis, are the results of suffering, or
appear very late in the disease because of progressive weakness.
Briefly stated, the phenomena which point with comparative certainty
to the existence of spinal tumors are symptoms of meningeal irritation
gradually increasing, and symptoms of slow compression of the cord.
These have been sketched at the beginning of Symptomatology. As to
duration, the data in the cases studied were somewhat meagre. The
usual duration is from six months to three years.
The differential diagnosis of spinal tumors will be considered in
reference to the following affections: congestion, hemorrhage,
meningitis (simple and specific), caries, traumatisms, sclerosis,
aneurisms, neuritis, metallic and infectious disorders, and hysteria.
Spinal tumors, it will be recalled, are from constitutional or special
causes, as syphilis, cancer, and tuberculosis. The onset is gradual
and irregular. The duration is comparatively long. The progress is by
irregular advances toward a fatal termination. The symptoms are
inclined to be at first unilateral or local; later, bilateral. Special
symptoms, as paralysis, spasm, sensory and visceral disorders, occur
irregularly as to time. Decubitus and trophic changes are common late
in the history. Reactions of degeneration are often present. Gowers
refers to the fact that two morbid processes often occur, one
consecutive upon the other, as a secondary degeneration or a
hemorrhage, after the establishment of the morbid growth, with
characteristic increase of symptoms.
In spinal congestion a constitutional cause is not likely to be
present. The onset is usually sudden and after exposure. The duration
is shorter than in tumors, and is from a few days to four months. The
disease is stationary for a while; then retrogression of symptoms
toward recovery occurs. The symptoms are more uniformly bilateral, and
motor and other symptoms develop about the same time. Decubitus is
rare. Reactions of degeneration are rare (?). It is desirable that
cases of so-called spinal congestion should be differentiated from the
forms of peripheral neuritis above referred to, the most
characteristic symptom of which appears to be tenderness of
nerve-trunks.
In spinal hemorrhage there is no special history, or a history and
signs of cardiac and vascular degeneration may be present. The onset
is quite sudden and the progress of the case regular. The first
symptoms persist, and secondary degenerations follow, and differ
according to the extent and location of the lesion, but are most
likely to be uniformly bilateral.
In meningitis the symptoms of localized compression are absent. The
girdle symptom is absent. The affection is sometimes curable, and
especially so if it has been of syphilitic origin. The reactions of
degeneration are not marked.
In caries of the spinal vertebræ deformity is rarely absent,
especially if the case has continued a few months. Rigidity of the
muscles of the back is an important symptom, which, however, is
occasionally found with tumor. Jarring of the spinal column by tapping
upon the head or jumping from a chair or stool is more likely to
elicit pain in caries than {1100} in tumors. Strumous symptoms and
evidence of tubercles in the lungs or other organs are often present.
In traumatisms usually a history of the injury can be obtained. The
symptoms are those of caries, myelitis, meningitis, or of combinations
of these, according to the character of the case.
In sclerosis the symptoms are usually those of progressive systemic
affections, with absence of compression symptoms. The duration is
longer. The progress is gradual and more regular.
Aneurisms are only to be distinguished when extra-spinal, causing
erosion and compression.
In neuritis there is the soreness of the nerve-trunk already referred
to, while compression symptoms and visceral disorders are absent. The
motor and sensory symptoms are confined to the area of distribution of
the affected nerve. It is amenable to treatment. In advanced stages
the reactions of degeneration are marked. In the form of general
peripheral neuritis, the existence of which, as a distinct disease, is
being at present claimed, the characteristic symptoms are as yet not
sufficiently determined or the pathology demonstrated by post-mortem
research to admit of much discussion.
In metallic and infectious disorders a history of definite causation
is present. Metallic disorders may present special distinctive signs,
such as lead-line, wrist-drop, etc.
In hysteria a precedent characteristic history is usual. The onset is
often sudden and an emotional element is present. The symptoms are
bilateral and protean. Trophic changes are absent. No reactions of
degeneration are present.
(2) Local Diagnosis.—It may be said of spinal tumors in a modified
sense, as it can be said of brain tumors, that they are not good
pathological experiments for illustrating the functions of the exact
areas which they occupy. The spinal canal has such narrow limits, the
tumor itself soon attains such a relatively large size and causes such
wide vascular engorgement, and the different tracts and systems of the
cord are so closely packed together, that the tumor does not often
invade only one functional area and escape another. Hence the regional
diagnosis presents special and greater difficulties than the diagnosis
of the level of the cord at which the tumor presents itself. A tumor
which destroys the trophic centres for the arm in the anterior cornua
might exert sufficient backward pressure to paralyze the motor tracts
running to the leg; or a cervical tumor, as in Case 5, might produce
symptoms which are almost wholly observed in the legs. It will be
seen, however, by reference to the table, that in Case 3, reported by
Wilks, we have a tumor whose exact anatomical seat could have been
predicted, and which seems to have reproduced almost the upper-arm
paralysis of Remak. When we compare these two cases, in which the
pathological conditions are so similar, it will be observed that the
paralyzed arm is much wasted, which indicates a lesion of its trophic
centre, whereas the affected legs in the other case are irritated by
pressure and by isolation, but are not wasted, because their trophic
centres are far below the point of lesion. The invasion of the trophic
centres, and the accompanying wasting of particular groups of muscles,
especially when this occurs early in the case, with the consequent
reactions of degeneration in these muscles, would furnish very
valuable indications {1101} both as to the region and the level of the
cord involved (Case 29). Unfortunately, the exact observations are
wanting in most of the cases as reported.
M. Allen Starr, in a recent paper,[7] has devoted much labor to the
elucidation of the functions of different segments and regions of the
cord. He demonstrates the existence of groups of cells in the gray
matter, especially in the anterior horns, each of which he believes
constitutes a physiological unit. He affirms that these cell-groups
preside over certain associated movements or combinations of certain
muscles, and, quoting from Spitzka, says that “the nearer a muscle is
to the ventral aspect of an animal the nearer will its nucleus be to
the median line of the cord; and the nearer the muscle is to the
dorsal aspect of the animal the nearer will its nucleus be to the
lateral cornua of the cord. Flexor nuclei are therefore in internal,
extensor nuclei in external and posterior, cell-groups.” The only
light that such a theory throws upon the subject of diagnosis is by
affording a possible explanation of the fact that spastic flexion is
much more common than spastic extension, and may be due to the fact
that the cell-groups for flexion lie deeper and are more protected
than those for extension; and the additional fact above referred to,
that a paralysis of associated muscles or groups of muscles, with
degeneration, as in the types of Remak, would indicate with great
clearness the destruction of the cell-group which presides over them.
Starr, in his article, also tabulates the various reflexes and their
seats in the cord. As this subject is of much importance in any exact
study of spinal-cord diseases, we will state here some of the facts as
given in that article: The neck-pupil reflex (dilatation of the pupil
on irritation of the neck) has its seat from the fourth to the seventh
cervical segment; the elbow-tendon reflex in the fifth and sixth
cervical; the wrist tendons from the sixth to the eighth cervical; the
palmar in the seventh and eighth cervical; the epigastric and
abdominal skin reflexes in the fourth to the eleventh dorsal segments;
the cremasteric reflex in the first to the third lumbar; the patellar
tendon in the second to the fourth lumbar, and bladder and sexual
centres in same; the rectal centre in the fourth lumbar to the third
sacral; the foot-clonus and Achilles-tendon reflex in the first
sacral. A destructive lesion, such as a tumor, at any one of these
points would cause abolition of that particular reflex, and this would
probably occur early in the case. Our table of cases does not present
any such observation, whereas exaggerated reflexes, such as occur from
a compressing lesion above the seat of the excited centre, are
recorded in abundance. Many of these deductions are of course only
possible early in the history of the case, as at a late stage the
secondary degenerations have caused too widespread havoc to admit of
any exact localization. The distinction must also be sought for
between a destructive lesion and the symptoms of irritation which it
may project to distant parts. Fürstner's cases of syringo-myelia,[8]
in which were marked vaso-motor changes, such as pallor, flushings,
copious sweat, and trophic disorders in the integument and its
appendages, seem to show that a lesion in the gray matter just
posterior and external to the central canal is necessary for such
phenomena. Similar vaso-motor changes may be observed in some of the
tabulated cases, as in No. 4, in {1102} which there were islands of
heat and cold in the leg, with a hydromyelia in the cord. Sensory
symptoms are very common in cases of spinal tumor, but they furnish
indications rather of the exact level of the lesion than of its
region.
[Footnote 7: “Localization of the Functions of Spinal Cord,” _Am.
Journ. Neur. and Psych._, 2-3, p. 443.]
[Footnote 8: Quoted by Starr.]
[Illustration: FIG. 45. Diagram of Spinal Column, Cord, and
Nerve-exits (after Gowers).]
[Illustration: FIG. 46. Sarcoma compressing Cervical Cord, Case 17 of
Table (E. Long Fox).]
Before considering briefly the indications which point to the various
levels of the cord as a possible seat of spinal tumor, it will be
necessary also to make plain a few anatomical facts. It must be borne
in mind, first, that the nerve-origins in the cord are never at the
same level as their exits from the spinal canal, or, in other words,
that the spinal segments do not correspond with the bodies of the same
numerical vertebræ, and that there is, in fact, one more cervical
segment than there are cervical vertebræ. The tendency is for the
nerve-trunks to run downward before passing out of the canal, so that
in every instance, without exception, from the medulla oblongata to
the filum terminale the segments of the cord are above the
corresponding vertebral body. This discrepancy increases as we descend
the cord; whereas it is approximately correct to say of the cervical
and dorsal regions that every segment is opposite the vertebral body
which is numerically just above it, this difference becomes much
greater in the lumbar and sacral regions. The cord itself terminates
in the lumbar enlargement which ends opposite the interval between
{1103} the first and second lumbar vertebræ. All the remainder of the
canal is occupied by the descending trunks of the lumbar, sacral, and
coccygeal nerves as they pass to their respective foramina,
constituting the cauda equina. It must be recalled, however, that the
vertebral bodies, lying very deep, cannot serve as guides, but that we
are dependent upon the spinous processes as landmarks in diagnosis.
These again differ in their levels from their respective vertebral
bodies, as they are deflected at somewhat different angles at
different regions of the spine. Gowers has illustrated these facts by
a very graphic wood-cut[9] (Fig. 45), from which the general rule may
be drawn that each vertebral spine is about opposite the spinal
segment which is numerically two places below it; thus the eighth
dorsal spine is opposite the tenth dorsal segment, etc. The
indications afforded by this exact anatomical knowledge have reference
largely to the existence of pain on pressure and to any deformity of
the bony structures. The cases as reported do not indicate that this
method of research has been utilized, and it may possibly be of only
theoretical importance; but it has been considered worthy of reference
as an indication in diagnosis.
[Footnote 9: _Diagnosis of Diseases of Spinal Cord_, p. 6.]
[Illustration: FIG. 47.]
[Illustration: FIG. 48.]
[Illustration: FIG. 49.]
[Illustration: FIG. 50. Sarcoma of Lower Cervical Cord, Case 13 of
Table (Adamkiewicz).]
It will be seen by reference to the table that usually certain general
features in the symptomatology indicate the seat of the lesion. Thus
in tumors of the cervical region pain and stiffness of the neck occur,
while the first appearances of paresis and sensory disturbances are
usually observed in the arms and about the chest. The centres for the
forearm and hand lie in the lower portion of the cervical enlargement;
that for the upper arm, including the supinator longus, in the upper
portion. Mental symptoms are more marked, and in Case 4 several of the
cranial nerves were implicated. In the lower cervical and upper dorsal
region there are symptoms of dyspnœa, fixation of the chest (Nos. 20,
22, 24), and cough. The girdle symptom is an important indication at
any level, {1104} as it is due to irritation of the nerves at the
lowest level of healthy cord just above the transverse lesion. It has
already been discussed under Symptomatology. In many of the dorsal
cases (Nos. 32, 37, 38, 39, and 45) the symptoms are almost entirely
confined to the legs and lower trunk, the arms escaping entirely. The
condition of the bladder is usually given in the table as one of
paralysis; this does not indicate whether automatic evacuation existed
at the beginning of the case; which condition would indicate that the
centre for micturition was below the lesion, and intact. It is
probable that later in these cases the bladder is actually paralyzed
by destruction of its centre in the cord, and this even when the tumor
has been situated some distance above.
[Illustration: FIG. 51. Fibroma of Lower Dorsal Cord, Case 32 of Table
(W. Cayley).]
[Illustration: FIG. 52. Tumor of Cauda Equina, Case 45 of Table (W. W.
Fisher).]