that tabes is a system-disease of the cord in the sense in which he
employs that term. In his fascinating and suggestive thesis on the
systemic affections he cites the observations of Tuczek made on the
posterior sclerosis ensuing from ergotism as confirming his view that
special systems of nerve-fibres are separately vulnerable to special
morbid influences. The remarkable correspondence between the
topographical distribution of the lesion of ergotin tabes described by
Tuczek and the characteristic areas {842} of sclerosis found in the
column of Burdach in ordinary tabes, seems to justify a very different
conclusion. Ergot exerts its morbid influence through a disturbance of
nutrition. Just as it produces gangrene by constriction of the
nutritive vessels of the fingers and of whole extremities, so it
produces interference with nutrition of those nerve-elements in the
cord which are most liable to suffer from general arterial
constriction. These are presumably those portions which are supplied
by the longest and narrowest arterial stems—a condition obtaining in
precisely those parts of the cord which are affected both in ergotin
tabes and in ordinary tabes.
It has also been supposed that the morbid process began in the
posterior roots and crept in with these, thence extending upward.[37]
This view is opposed by the fact that there is no constant
relationship between the root lesion and the cord lesion; the
sclerosis of the root-zones within the columns of Burdach I found to
be absolute in at least one case where the outer nerve-roots were not
distinctly affected.[38]
[Footnote 37: Takacs is the most recent defender of this view.]
[Footnote 38: The root lesion may, like that of the column of Goll,
according to a minority of the interpreters, be a secondary process,
for in ergotin tabes (Tuczek) both are usually intact.]
In typical tabes the sclerotic process begins in a special triangular
field of the posterior column in the lumbar enlargement of the cord.
The innermost of the posterior rootlets run through this field, which
has been designated by the French school as the posterior root-zones,
and the rootlets become exposed to all the vicissitudes which the
conducting tracts are destined to undergo in consequence. Throughout
the remainder of the cord it is the involvement of the root-zone which
indicates the advance of the affection upward. At a time when the
lesion is comparatively slight in the cervical level, manifesting
itself by a slight grayish or reddish streak to the naked eye, a
faint, pale discoloration in chromic-acid specimens, a deeper stain in
carminized, and a pallor in sections stained by Weigert's method, the
affected part in the lumbar region has undergone great shrinkage;
scarcely a single myelin-tube is preserved to receive Weigert's stain
within the affected area; the latter involves nearly the entire field
between the cornua, and absorbs carmine deeply, so that it is
difficult to differentiate the gray and white substance in carminized
sections. A small part of the posterior column which most closely
adjoins the posterior (gray) commissure remains free in all cases; so
also does a small laurel-leaf-shaped field bisected by the posterior
median septum. Examination with higher magnifying powers, aided by the
modern improved methods of preservation and staining, reveals that the
gray substance of the posterior segment of the cord is usually more or
less diseased. In advanced cases the gelatinous as well as the spongy
part of the posterior gray horn, and even the posterior gray
commissure, undergo such considerable atrophy that they may occupy but
half the normal area. In some cases the columns of Clarke appear to be
involved quite early in the disease.[39]
[Footnote 39: Leyden discovered the degeneration in the fibres which
in part make up these columns and correctly traced its origin to the
posterior rootlets. Lissaner and Zacher (_Archiv für Psychiatrie_, xv.
p. 437) confirmed him, the latter finding, as Leyden was the first to
discriminate, that the cells may escape while the fibres are
materially diminished. Krause claims that the columns of Clarke are
constantly affected, but not necessarily in their entire length, the
higher levels appearing quite normal at a time when the lower are
severely involved (_Neurologisches Centralblatt_, 1884, p. 50).]
{843} The triangular field in the lumbar part of the posterior column,
which is one of the typical starting-points of the affection, contains
those ascending nerve-bundles which in their cephalic course
emancipate themselves from the column of Burdach and constitute the
slender columns of Goll. The result is that the degenerative process
creeps up these columns at the same time that it ascends in the
root-zones and deep portions of Burdach's columns. Some authorities
regard this as a mere extension by contiguity;[40] others incline to
consider it a secondary degeneration. It may extend to the medulla
oblongata, becoming lost in the level where the nucleus of the column
of Goll terminates, and is accompanied, at least in those advanced
cases in which the upper extremities are involved, by a comma-shaped
area of degeneration in the adjoining part of the column of Burdach,
which similarly extends into the oblongata and terminates slightly
more cephalad. In typical advanced tabes, therefore, the cross-section
of the cord exhibits a characteristic distribution of the sclerosis in
each level. As this distribution is associated with certain constant
symptoms, it is permissible to attempt bringing certain features of
the lesion in relation with special features of the disease symptoms.
The posterior gray horns and the posterior white columns, together
with other fibre-systems connected with them, are much more
complicated in structural and physiological relations than the
corresponding anterior structures. The relations of the anterior
rootlets to the gray substance, and those of the motor ganglionic
elements to their controlling tracts, are comparatively simple; those
of the posterior roots are very intricate. They run up, in great part,
at an angle to the longitudinal tracts; a few pass in directly, and
still fewer dip to a lower level. The result is that a section of the
cord made in the longitudinal direction through the root-zones, so as
to pass from the root-entry to the anterior commissure, shows the
column and root-fibres to be woven into each other like a plait.
Trabeculæ of connective tissue, dragged in as it were with the
posterior roots, fill up the interstices of this labyrinth. They are
particularly dense in the lowest part of the lumbar enlargement of the
cord, constituting the so-called posterior processi reticulares. It is
reasonable to suppose that the overlapping of ascending and descending
root-fibres, associated with the presence of an extra amount of
connective tissue, imbedded as this fibre-maze is in that part of the
cord which is most distant from its lymphatic emunctories, affords a
favorable soil for slow inflammatory trouble. This is the primary
field of tabic sclerosis, and in it the disease may remain most
intense for years, extending but slowly and with diminishing intensity
upward, hand over hand, as it were, on the natural ladder which the
intertwined fasciculi and their matrix constitute. The longitudinal
tracts which lie in and near the root-zones belong to the so-called
short fibre systems, uniting the segments of higher and lower levels
of {844} the cord with each other. As the sclerotic process ascends it
involves the caudal ends of these systems: they consequently undergo
secondary degeneration, and, shrinking in their turn, affect the
caudal part of the next system above in the same manner. The morbid
process in the column of Burdach may therefore be considered as a
combination of inflammatory and degenerative changes, the inflammatory
products causing a series of short ascending degenerations, and the
vulnerable path thus established being followed by a cirrhotic
condition in which the connective and vascular structures participate
actively. With regard to the reasons for regarding the degeneration of
the column of Goll and that of the comma-shaped field near it as a
secondary process due to the cutting off of its apparent nerve-supply
at the caudal end, and of the posterior nerve-roots or their
provisional terminations, they may be stated in this way: When the
lesion of the primary field is limited to the lower lumbar or sacral
part of the cord, the degeneration of the column of Goll is limited to
its postero-internal part; when the upper lumbar and lower dorsal cord
is involved, the entire tract is affected; and when the cervical
portion is diseased, the supplementary comma-shaped area degenerates.
In other words, the projection tract of the sciatic nerve, as far as
it is represented in Goll's column, suffers in the first, that of the
crural nerves in the second, and that of the brachial nerves in the
third instance. In all advanced cases of tabes the affection of the
column of Goll is in direct proportion to the altitude of the lesion
in the primary field. Symptomatically, it bears an equally constant
relation to the ataxia.[41] No case is on record in which these
columns were totally degenerated without some motor inco-ordination of
the lower extremity having been observed during life; and no case is
recorded in which brachial ataxia had been a marked and persistent
feature in which the comma-shaped area—area of the column of
Burdach—was healthy.
[Footnote 40: It is held by them that the histological character of
the change of the columns of Goll is not different from that in the
column of Burdach. Zacher (_Archiv für Psychiatrie_, xv. p. 435) urges
that it does not resemble true secondary degeneration, beginning in
the vessels and connective substance instead of the nerve-fibres.
Schultze (_ibid._, xiv. p. 386), on the other hand, recognizes a
primary involvement of the nerve-fibres in both of the areas of
fascicular degeneration in tabes. The observation of intact
axis-cylinders by Babinski in the sclerotic fields is in conflict with
the latter's claim, and the various differences of observation and
interpretation seem to be reconcilable only on the assumption that
there are two different modes of origin, both leading to nearly the
same results and occasionally combined in one and the same case.]
[Footnote 41: Krause's case and others show that the ataxia of
movement is not influenced by lesion of the column of Clarke; but we
are not informed as to the static equilibrium of the patients in whose
cords these columns were found intensely affected.]
It is scarcely necessary to seriously consider the suggestion of
Strümpell, that the lesion of the column of Goll is in relation with
the bladder disturbance.[42] Cases are on record by Wolff and others
where this lesion was intense and there was little or no bladder
disturbance.[43]
[Footnote 42: _Archiv für Psychiatrie_, xii.]
[Footnote 43: The column of Goll is not present in those mammals
which, like the porpoise, have no developed hind limbs, but these
animals have urinary bladders.]
If the disease of the column of Goll were a primary systemic affection
independent of the disease of the root-field, it would be difficult to
understand why it, as well as the likelihood of finding a
corresponding degeneration of the direct cerebellar tract, increases
with the extent to which this field is involved. This occurrence
becomes quite clear when we remember that both the direct cerebellar
tract and that of Goll, being centripetal, are under the trophic
dominion of the posterior nerve-roots. The opinion is not distinctly
expressed, but implied in some writings, that the column of Goll
degenerates because of a general transverse cord lesion at a low
level; this is not the case in the tabic cord. There is a difference
in appearance between that part of the primary field which corresponds
to the column of Goll in the lumbar cord and the surrounding sclerosis
in early cases: it is more intensely degenerated, more {845}
homogeneous in appearance, and more evenly stained. The other part of
the triangular field presents a more trabecular appearance. In the
specimen represented in the accompanying figure this is easily
recognized: the darker field corresponds exactly with the ascending
degeneration, which follows compression of the cauda equina,[44] and
is the sciatic equivalent of the column of Goll.
[Illustration: FIG. 31. Trans-section of Upper Lumbar Cord of a
Patient moderately advanced in Tabes: _f_, ventral or anterior
fissure; _g_, caput gelatinosum; _r r′_, entry and deep course of the
posterior rootlets; _d_, degenerated field, including the origin of
the tracts which in higher levels form the columns of Goll; _n_, field
near the posterior commissure which remains free from degeneration,
both in the lumbar and cervical cord; _s_, sharply marked leaf-shaped
field, bisected by posterior septum, which, as claimed by Strümpell
and shown in this case, remains free from disease; _z_, zone of nearly
normal consistency around it; _a_, diseased field, suspected to be
related to the analgesia, not usually affected in early tabes; _v_,
fibres running up and down in front of the gelatinous substance; *,
region where the tabic process sometimes begins.]
[Footnote 44: I have also found that this field corresponds to the
column of Goll in its myelinic development: the lumbar part of this
column—designated as such by Flechsig—is an entirely different tract,
which enjoys a remarkable immunity from disease in tabes.
As illustrating the bearing of the lesion of the column of Goll on the
motor ataxia I may refer to two cases which happen to be related side
by side by Strümpell (_Archiv für Psychiatrie_, xii. p. 737, Cases 1
and 2). As far as the lumbar segment of the cord is concerned, the
distribution of the lesion is similar; but in the one presenting
marked motor ataxia the triangular field was slightly diseased, and
there was no upward extension of the lesion in the column of Goll. In
the other, with marked ataxia, the triangular field was intensely
diseased, and ascending degeneration (?) occurred in the sciatic
fields of the latter.
In support of the view that the comma-shaped area in Burdach's columns
is homologous with the fibres of the column of Goll, it is to be
advanced that degeneration of this field bears the same relation to
symptoms in the distribution of the ulnar nerve that degeneration of
the column of Goll bears to sciatic and crural symptoms. Where the
initial pains and subsequent tactile and locomotor disturbance were
severe, this field was found affected, and most so in the side where
the symptoms had been most intense (Friedreich-Schultze's cured cases,
_Archiv für Psychiatrie_, xii. p. 234). This area has no direct
connection with the root-fields. Secondly, in a primary system disease
of the column of Goll, associated with degeneration of the nucleus of
the column of Goll, described by Scoli, an irregular encroachment of
the column of Burdach was noted. Third, the innermost fibres of the
column of Burdach (those belting the nucleus in the oblongata) have
the same relation to the interolivary layer which the column of Goll
has through its provisional nucleus of termination.
While the evidence of high lesion of the cerebral continuation of the
column of Goll, and, what I regard as its homologue, the comma-shaped
area of Burdach, together with the constant association of marked
degeneration of these columns with motor ataxia, is strong positive
proof of its relation to this symptom, there is equally strong
evidence negativing its relation to any other of the prominent
symptoms of tabes dorsalis. Thus Babesin (_Virchow's Archiv_, lxxvi.
p. 74) found degeneration of the posterior columns limited to the
column of Goll, and the patellar reflex was not destroyed; the
root-fields at the upper lumbar levels were intact. That the columns
of Goll have been found profoundly affected without bladder
disturbance has been stated previously, and constitutes a stronger
argument against Strümpell's view than the frequent observation of
bladder trouble in spinal diseases, along with which these columns may
be entirely free.]
Among the various constituents of the posterior columns which appear
{846} to present a relative immunity to the disease, aside from the
area near the posterior commissure and the laurel-leaf-shaped area of
the posterior septum, Strümpell noted one which is situated at the
periphery of the cord, bordering on the entry-line of the posterior
roots and the inner contour of the posterior roots. In the few cases
where it was found destroyed there was, what is a comparatively rare
thing in moderately severe tabes, complete analgesia. It is not,
however, certain that there is a necessary connection between the
pathological and clinical fact here. The nerve-roots themselves are
involved within the diseased area of the root-zones. The lesion is one
of a kind which, affecting a nerve-trunk, would produce first
irritation of, then impediment to, and ultimately destruction of, its
function. The clinical parallel to this is the occurrence of the
lightning-like pains in the earlier phases of the disease, which are
followed by delayed pain-conduction, and finally by loss of sensation.
Of the rootlets or fibres subservient to the various sensory and
reflex functions mediated by the posterior roots, those which convey
the centripetal impression normally evoking the patellar jerk appear
to be the most vulnerable, or, because of their limited number, the
earliest to be destroyed or compressed, with the result of total
functional paralysis.[45]
[Footnote 45: It is now conceded that, as Westphal claimed, the
patellar jerk is always abolished when the upper lumbar level of the
root-zones (bandelettes externes of Charcot) is involved. Tshirijew
has shown that the translation of the knee-jerk reflex occurs in a
single segment of the rabbit's cord at the homologous level. In
transverse sections a distinct fascicle may be seen coursing from the
innermost root-fibres toward the antero-intermediate cell-group of the
anterior horn; it furnishes a pictorial substratum, if not anatomical
proof, for the patho-physiological observation, and harmonizes with
the fact that it is the innermost rootlets which usually suffer first.
Perhaps the delicacy of this tract accounts for the frequent
disappearance of the jerk in old people as a result of senile
sclerosis.]
With regard to the interpretation of the various tactile sensory
disturbances of tabes and the delayed pain-sense conduction little
positive advance has been made. That the lesion of the root-zones and
gray substance is responsible for them seems to be the general
conclusion of French and German observers. Recent researches have
shown that disease of the peripheral nerves (multiple neuritis) may
produce sensory phenomena which it had previously been customary to
regard as pathognomonic of disturbance of their intraspinal
terminations and continuations. At the same time, we are confounded by
the observation of Erb, that even retardation of conduction of the
pain-sense, which the dictum of Schiff taught us to regard as a sign
of disturbed function of the spinal gray matter, is also produced by
peripheral neuritis.
Immediately adjoining the gelatinous apex of the posterior horn there
is a column of vertical fibres which bear a relation to it resembling
that {847} of the ascending root of the fifth pair in the oblongata to
the tuber cinereum of Rolando. Sclerosis of this column, as well as of
the ascending root in question (Demange), has been found in tabes, and
usually in association with pronounced trophic disturbances. In a
number of cases (Oppenheim, Eisenlohr, and others) where spasmodic
laryngeal crises had been a marked feature during life, lesion of the
floor of the fourth ventricle, or atrophy of the pneumogastric nerve,
or even of its nucleus, was found. In one case with marked gastric
crises I found sclerotic changes of the arteries in the ala cinerea.
With regard to the involvement of the peripheral nerves proper, aside
from the optic, opinion is somewhat divided. The discovery of multiple
neuritis, and the recognition of the fact that it had been erroneously
confounded with tabes, led Dejerine to claim that tabes might be of
peripheral origin. He even proposed to account for the oculo-motor
trouble on the basis of an affection of the oculo-motor nerves. This
explanation has been repudiated by Westphal and sound authorities
generally. That the peripheral nerves are occasionally involved in
tabes was already known to Friedreich in 1863,[46] and later Westphal
found the cutaneous branches in an advanced tabic patient to present
similar changes to those discovered by Friedreich: these findings are
confirmed by his Japanese pupil, Sakaky. The nerve-tubes are
atrophied, the axis-cylinder being often preserved, and the
interneural connective substance is proliferated. But there is no
constant relation between these changes and the symptoms of the
disease; in one of Sakaky's cases the nerves of an extremity which had
been the seat of severe sensory symptoms were entirely normal. The
findings in the peripheral nerves of tabic subjects lose much of their
value in view of their recent discovery in marantic persons[47]
advanced in life, but who had no nervous disease whatever.
[Footnote 46: _Virchow's Archiv_, vol. xxvi. pp. 399-452.]
[Footnote 47: Krause, _Neurologisches Centralblatt_, 1885, p. 53.]
It has been attempted to bring the tabic process in relation to a
supposed primary meningitis. Tabes is a rather common nervous
affection, and primary spinal meningitis is one of the very rarest.
The leptomeninges are found considerably thickened in one out of ten
tabic cords, and those who defend the meningitic origin of the disease
base their theory on this inconstant finding, and allege that in the
cases where it is absent the meningitis has disappeared while the cord
lesion progressed. It is a fatal objection to this view that the part
of the posterior columns immediately adjoining the pia is often quite
free from disease. A meningitic affection, either as an etiological or
a complicating factor, can be admitted in those cases only where there
is a marginal sclerosis.
The changes in the optic nerve resemble those of the white columns of
the cord in their naked-eye and minute character as well as in the
controversial nature of the various interpretations made. When
affected, the nerve is found to be firmer than normal, and discolored;
later it becomes quite gray, and may eventually shrink to two-thirds,
and even less, of its normal diameter. It is generally believed that,
as in the cord, the myelin undergoes wasting before the axis-cylinder
disappears, and that the latter may survive a long time, thus
explaining why the patient may retain his visual power for a
considerable period after the ophthalmoscope determines the existence
of atrophy. No satisfactory explanation {848} has as yet been offered
for the optic-nerve affection of tabes. There is no direct continuity
of the spinal and optic sclerosis. Two theoretical possibilities
suggest themselves. The first is that the lesion of the cord exerts a
remote effect upon the physiological, and through this the structural,
condition of the optic nerve. That such an influence may be exerted is
shown by cases of transverse myelitis low down in the cord, which,
according to Erb and Seguin, were complicated by double optic-nerve
atrophy. The second theory is that the involved part of the cord and
the optic nerve present a similar vulnerability to the same morbid
influences. This is illustrated in some cases of chronic alcoholic and
nicotine poisoning, in ergotism, and in the spinal affections due to
hereditary influences and developmental defects.