FROM THE CENTRE OF NEUROSURGERY, U.A.R.

ARMED FORCES HOSPITAL CAIRO,

EGYPT.

BIOMECHANICS OF THE CERVICAL SPINAL CORD

Relief of contact pressure on and overstretching of the spinal cord

by

ALF BREIG and AHMED FOUAD EL-NADI

The choice of conservative or surgical measures in acute and chronic

spinal cord injury will be determined largely according to the concept of
how detrimental forces act mechanically on the cord. Apart from compression,
which has always been regarded of importance, workers of today tend to
emphasize the significance of pathologically increased tension. Many models
have been designed as aids to the understanding of the problem but up to
now the lack of knowledge of the mechanical properties and biomechanical
behaviour of the hindbrain/spinal cord tissue tract has constituted a considerable
handicap. As new light has been thrown in recent years on some basically
important features, this may perhaps enable the action of compressive and
tensile forces to be seen in a more appropriate mechanical connection.
A fairly common mechanism of damage in processes which encroach upon
the width of the spinal canal is contact pressure on and overstretching of the
spinal cord. This has considerable practical implications. For all stages of
treatment of spinal cord lesions - from first aid, transport and operation to
rehabilitation - further traumatic damage to the nervous tissue may be
Submitted fur publication 16 August 1964.

602

Downloaded from acr.sagepub.com at MCMASTER UNIV LIBRARY on June 29, 2015

 BIOMECHANICS OF THE CERVICAL SPINAL CORD 603

Fig. I. Extreme functional positIOns of a

normal cervical spine in relation to the poste-
rior fossa. In ventroflexion (shaded) the cer-
vical canal is elongated while in dorsal ex-
tension it is shortened (unshaded), the change
in both instances being greater on the dorsal
than the ventral contour.

avoided by correct support of the spine so that the dura, cord and nerve roots
become slack. The biomechanical basis of the necessary measures is outlined
in this paper.
The earlier view that the mechanical effect on the cord and nerve roots main-
ly consisted of compression has led to many neurosurgical and orthopaedic
operations being regarded as more or less exclusively decompressive in their
effects, irrespective of whether they consisted in laminectomy, division of
dentate ligaments or removal of protrusions on the ventral contour of the
spinal canal. Compression of the cervical cord must be considered when the
cervical canal has become narrow as a result of acute or chronic pathologic
changes. As will be shown below, however, mechanical conditions conducive
to compression of the cord are fairly unusual and occur only when the canal
is markedly narrowed. A proper understanding of the mechanical process
will enable the surgeon to predict the outcome of his conservative or operative
measures.
A study has been made of the mechanical action of protrusions in the cervical
canal and their effect on the cord in extreme functional positions. The results
obtained may serve as a basis for determining the type of damage to the cer-
vical cord most likely to result from this type of mechanical trauma.

Downloaded from acr.sagepub.com at MCMASTER UNIV LIBRARY on June 29, 2015

604 AI.F BREW AND AHMED FOUAD EL-NADI

,
,,

Fig. 2. Longitudinal sections of two hindbrains with the cervical cords (fixed in situ) in relation to
the foramen magnum and the upper cervical canal: in full ventrollexion (left) and in full dorsal
extension (right).

The mechanics of the hindbrain and the cervical cord are governed by the
mechanics of the skeleton.
In ventroflexion of the cervical spine, the functional axes of curvature of the
intervertebral and atlanto-occipital joints are on the ventral side of the spine,
so that its dorsal contour (Fig. I) will be slightly elongated. The clivus will
in addition rotate forwards, which leads to a correspondingly increased distance
between the occipital bone and the arch of the atlas.
In dorsal extension of the cervical spine the functional axes of curvature
lie on the dorsal side of the spine and it is now the dorsal contour that undergoes
the greater shortening. The clivus is rotated backwards and the angle between
its upper contour and the posterior contour of the odontoid process decreases,
often to 180 0 or less. The distance between the occipital bone and the arch of

Downloaded from acr.sagepub.com at MCMASTER UNIV LIBRARY on June 29, 2015

 BIOMECHANICS OF THE CERVICAL SPINAL CORD 605

Fig. 3. Hindbrain and cervical cord from the

front. A rhomboid arachnoid veil lies on the
ventral aspect of the medulla and the upper
part of the cervical corel. (From KEY & RETZlUS,
Table 1, Fig. 2.)

the atlas also decreases, perhaps to zero. In extreme posItIOns, the change
in length of the ventral and dorsal contours is greater in the cervical than in
other parts of the spinal canal. In consequence, it is the cervical dura and cord
that display the greatest variations in length. As a result of the shortening of
the cervical canal iii dorsal extension of the cervical spine with the head
tilted backwards, the cervical dura and cord undergo axial compression
and hence shortening and slackening. Since in this position the shortened
cervical canal is often strongly curved backwards there may be a backward
curvature of the axis of the pons, the medulla oblongata and the upper cervical
cord (Fig. 2).
Measurements by means of indicators have shown that the floor of the
fourth ventricle and the dorsal contour of the medulla are shortened in this
position and that this leads to slackening of the hindbrain. Confirmation has
been obtained from autopsy specimens and at operations. The resulting increase
in mobility of the hindbrain and cervical cord is evident from the fact that these
structures can be raised dorsad or moved laterally until the slackened dentate
ligaments and nerve roots in the posterior fossa and cervical canal are pulled

Downloaded from acr.sagepub.com at MCMASTER UNIV LIBRARY on June 29, 2015

606 ALF BREIG AND AHMED FOUAD EL-NADI

Fig. 4. Tomographic gas myelograms of lower cervical spine and

upper thoracic spine (left) and of upper cervical spine (right).
The cord follows the curvature of the upper thoracic canal, whereas
in the ventroflexed cervical canal it follows a straight path across
the sligh t lordosis.

taut. This mobility may be exploited during operation in the region of the
cervical cord and medulla or further rostrally in the region of the pons, as
well as in the rest of the spinal canal, to obtain an easier approach to the
ventral contour of the posterior fossa and the cervical, thoracic and upper
lumbar canal without risk of injury to the hindbrain and cord.
The cord in its relaxed state may be deformed, for instance by surgical
instruments, without impairing its conductivity. When the dura is sufficiently
thin and translucent it is possible to follow the axial movements of the cord,
medulla and cerebellar tonsils. These movements are synchronous with res-
piration and increase in amplitude with the degree of dorsal extension.
During a particularly deep breath, the slack medulla and cord may be thrown
dorsad by a fluid wave. Conversely, during ventroflexion these movements
diminish until they are no longer visible in the extreme position. Synchronous
pulsation movements have seldom been observed with the naked eye.
The intervertebral canals leave the cervical canal on its ventrolateral
contour and in ventrolateral directions.

Downloaded from acr.sagepub.com at MCMASTER UNIV LIBRARY on June 29, 2015

 BIOMECHANICS OF THE CERVICAL SPINAL CORD 607

The attachments of the dentate ligaments to the dura in the cervical canal
are located slightly dorsal to the exit of the nerve roots and slightly nearer
the anterior than posterior aspect. The most rostral ligaments arise in the
lateral bands of the cord and run obliquely backwards and upwards to their
dural attachments in the posterior fossa where they are located dorsal to the
site of entry of the vertebral arteries (Fig. 3). When the cervical dura is pulled
taut in ventroflexion, the tension exerted by the dentate ligaments is transferred
axially and radially to the cord so that the cord becomes stabilized in the
canal and kept away from the odontoid process by the most cranially situated
dentate ligaments. The rhomboid arachnoid veil (Fig. 3) described by KEY &
RETZIUS (1875) appears to have the same function. If also the head is bent
forwards, the hindbrain will also be pulled taut, elongated and stretched
between its fixation at the mesencephalon and the taut cord. The angle be-
tween the upper contour of the clivus and the dorsal contour of the odontoid
process will then increase, and the supplementary ventral angle will decrease.
The occipital bone moves away from the atlas arch and the arches of the cer-
vical vertebrae will spread. The result will be a marked elongation of the
atlanto-occipital part of the upper cervical canal on its dorsal aspect, and
correspondingly of the dorsal contour of the medulla and the floor of the fourth
ventricle, the roof of which becomes more obtuse, while the choroid plexus
bulges less into the ventricle.
It has been demonstrated at autopsy that dorsal extension of the cervical
spine in the prone position causes the cord to sag to the ventral aspect of the
cervical canal. Dorsal extension in the supine position produces slackening of
the dura and the dentate ligaments and allows the cord to come in contact
with the dorsal aspect of the cervical canal. Stretching of the dura, the liga-
ments and the spinal cord in ventroflexion, however, lifts the cervical cord away
from the dorsal aspect of the canal so that it may come in contact with any
protrusion present on the ventral aspect.
The attachments of the dentate ligaments to the dura in the region of the
thoracic canal are located more dorsally. Gas myelography will therefore
usually reveal the cord to follow the normal dorsal convexity of the thoracic
canal nearer its dorsal than its ventral aspect (Fig. 4).
The elastic stretching and relaxation of the tissue tract results in a slight
axial displacement of the dura and the cord within the cervical canal. This
displacement differs slightly for the two structures; consequently there is a
slight displacement of the cord within the dura (SMITH 1956, BREIG 1964).
These displacements, being relatively small, are of minor importance as
compared to the large movements associated with elongation and shortening,
and will therefore be disregarded in the present connection.

Downloaded from acr.sagepub.com at MCMASTER UNIV LIBRARY on June 29, 2015

608 ALF BREIG AND AHMED FOUAD EL-NADI

a b
Fig. 5. Case I. Median sagittal tomographic gas myclograms in dorsal ex-
tension and full ventrofiexion of cervical spine. a) Dorsal extension. The
protrusion at C4-C5 bulges into the cervical cord. b) Ventrofiexion. The
cord is still in contact with protrusion in (a) but now the protrusion at C5-C6
bulges into the cord.

Material alld Jvlethods. Gas myelography was performed in 6 male patients

with funicular signs suggestive of cervical myelopathy. The examinations were
conducted over a period of 5 months from October 1962 to March 1963. The
roentgenograms were taken in full ventroflexion and dorsal extension of the
cervical spine with the object of ascertaining which protrusions came into
contact with the cord and thus were probably responsible for the symptoms
and signs.
General anaesthesia was used only in one uncooperative patient. Premedica-
tion consisted of intramuscular administration of 25 mg Pethidin and 25 mg
Phenergan 15 minutes prior to the examination.
The patient was held in the lateral position with the cervical spine parallel
to the roentgen table, the head end of which was lowered about 15° to 20°.
After combined cisternal and lumbar puncture, air was injected slowly through
the lumbar needle and the cerebrospinal fluid was allowed to drip from the
cisternal needle. When all the fluid had been removed, and the subarachnoidal
air pressure had been raised to between 200 and 300 mm H 2 0 by air insufflation

Downloaded from acr.sagepub.com at MCMASTER UNIV LIBRARY on June 29, 2015

 BIOMECHANICS OF THE CERVICAL SPINAL CORD C09

a b a b
Fig. 6. Case 2. Median sagittal tomographic gas Fig. 7. Case 3. Median sagittal tomographic
myclograms in dorsal extension and full ventro- gas myelograms in dorsal extension and full
flexion of cervical spine. a) Dorsal extension. The ventroflexion of cervical spine. a) Dorsal ex-
protrusions at C3-C4, at C4-C5 and at C5-C6 tension. Fairly wide clearance between the
bulge into the corel. b) Ventroflexion (in this case protrusion C3-C4 and the cord. b) Ventro-
pathologically reduced). A well defined air column flexion. Contact between protrusion and cord
lies along the ventral contour of the corel. except present for a short distance.
at the C3-C4 and C5-C6 levels.

under manometer control, high kilovolt exposures of the cervical spine were
made in extreme positions, with the head first bent backwards and then for-
wards, the chin drawn in and the cervical spine ventroflected. For most of the
patients at least five special tomograms were exposed in each of these extreme
positions, with the beam perpendicular to the cervical column. The entire
width of the cervical canal at the suspected level of the intervertebral space
of the cervical or cervicodorsal regions was thus covered. In an attempt to
locate a probable protrusion more accurately, the sequence and distances of
the five tomograms for the cervical and cervicodorsal regions were fixed at,
respectively, I cm and 0.5 cm to the right of the midline, in the midline, and
0.5 cm and I cm to the left of the midline. It was considered that this would
demonstrate any intraforaminal, dorsolateral or dorsomedial protrusion.

Results
All subjects (Figs 5 to 10, Cases I to 6) exhibited more or less advanced
degeneration of the disks with marked reduction of the intervertebral spaces
and osteophytes at the upper and lower margins of the bodies at various

39-663004. Acta Radiologica Diogllo,i, Vol. 4 (/91;6)

Downloaded from acr.sagepub.com at MCMASTER UNIV LIBRARY on June 29, 2015

610 ALF BREW AND AHMED FOUAD EL-NADI

a Fig. 8 b a Fig. 9 b

Fig. 8. Case 4. Median sagittal tomographic

gas myclograms in dorsal extension and full
ventroflexion of cervical spine. a) Dorsal ex-
tension. The protrusion at C3-C+ bulges into
the cord. b) Ventroflexion. There is still con-
tact but over a smaller area.

Fig. 9. Case 5. Median sagittal tomographic

gas myclograms in dorsal extension and full
ventroflexion of cervical spine. a) Dorsal ex-
tension. Contact at the protrusion C+-C5
and bulging of the protrusion C5-C6 into
the corel. b) Ventl'Oflexion. There is a narrow
air space between the protrusion at C+-C5
and the cord, but now the protrusion at C5-
C6 bulges farther into the cord (cf fig. 5).

Fig. 10. Case 6. Median sagittal tomographic

gas myclograms in dorsal extension and full
ventroflexion of cervical spine. a) Dorsal ex-
tension. The protrusions at C3-C+, at C+--C5
and C5-C6 bulge into the corel. b) Ventro-
b flexion. A clear air space is evident along the
a Fig. 10 b ventral contour of the cord except at a more
laterally situated small area of contact at thc
C1--C5 level and a distinct air space along the dorsal contour of the vertebral bodies.

levels of the cervical spine. These osteophytes, together with the disk protrusions,
sometimes bulged far into the cervical canal. The flexibility of the cervical
spine, as reflected in its extreme positions in the films, was considerably
less than in the normal subject. It was nevertheless retained to the extent
that the bulging of the dura into the lumen of the canal was more marked in

Downloaded from acr.sagepub.com at MCMASTER UNIV LIBRARY on June 29, 2015

 BIOMECHANICS OF THE CERVICAL SPINAL CORD 611

Fig. 11. Diagrammatic representa-

tion of gas myc10grams of Case 4
depicting the effect of the protru-
sion on the cervical cord in full
dorsal extension (left) ancl in full
ventroflexion (right).

dorsal extension than when the spine was in its normal position owing to the
axial compression and ensuing relaxation of the dura and ligaments. This is
particularly well illustrated in Case I.
The relclxation of the cervical cord in dorsal extension was evident in all
the tomograms as a wavy ventral and dorsal contour. When the patient is in
lateral position with the head end of the table tilted, the caudal part of the
cervical cord sags onto the sloping anterior contour of the canal and comes into
contact first with the largest of the protrusions and then with any others. Since
in this position the effect of gravity on the cord is similar to that in the fully
prone position, it will be referred to simply as the 'prone' position. As several
of the roentgenograms show, the slackening may be so marked that the cord
even drops onto the dorsal surface of the vertebral bodies, that is, onto a fairly
large area of the dura. The increase in number of contact areas in dorsal
extension is alse due to the more marked bulging of the protrusions in this
position. When the head is bent backwards, the hindbrain together with the
cranial attachments of the dentate ligaments are also slackened, and in con-
sequence the pons and the medulla fall on to the clivus and the tip of the
odontoid process. Examples are provided in all cases excepting Case 3, in
which the foot end of the roentgen table was probably tipped so that the
slackened hindbrain under the influence of gravity fell dorsad.
A 'differenee in phase' between the undulations on the ventral and dorsal
contours produces the wavy pattern of the slackened cord. The reason for this

Downloaded from acr.sagepub.com at MCMASTER UNIV LIBRARY on June 29, 2015

612 ALF BREIG AND AHMED FOUAD EL-NADI

Fig. 12. Dura stripped from the cer-

vical canal to reveal the functional
behaviour of the flaval ligaments in
full dorsal extension and ventroflexion.
After incision of the dura around the
foramen magnum it was stripped from
the surrounding attachments in the
~
epidural space by means of long scissors,
and severed at C7. ~..-4

is not clear but may possibly derive from the oblique force exerted by the
nerve roots. The wavy dorsal contour is obviously not caused by the folding
of the dura over the ligamenta flava, since in all cases except in Case 2 (perhaps
because of arachnoidal adhesions) there was a clear air stratum in the subarach-
noidal space between the dorsal contour of the cervical cord and that of the
cervical canal.
The cross-section of the cord normally increases in dorsal extension; in the
present subjects this was not marked, though it was just evident in Cases I to 4
at the level of the atlas. Ventroflexion causes stretching of the dura and
surrounding structures, including the dentate ligaments, and hence the spinal
cord, and the bulging of the disks, the dura and the flaval ligaments into the
lumen of the canal is distinctly diminished. Moreover, the lower part of the
cervical cord in particular is 'raised' from the anterior surface of the canal.
The stretching is manifest on the straight ventral and dorsal contours of the
cervical cord (Cases I, 3, 4, 5, 6); the area of contact between the cord and
the protrusion(s) is thus reduced. Ventroflexion of the spine was pathologically
restricted in Case 2, and as the cord was not much stretched its anterior con-
tour retained its wavy form.
Stretching causes the cord to move away from the anterior surface of the

Downloaded from acr.sagepub.com at MCMASTER UNIV LIBRARY on June 29, 2015

 BIOMECHANICS OF THE CERVICAL SPINAL CORD 613
Table
Appearances in dorsal extension of the cervical spine (Cases 1 to 6)

Case Decrease in Shortening Slackening of hindbrain and cervical cord -

angle between of the upper approximation of anterior contour of:
clivus and cervical
odontoid canal·· Pons to Medulla to Medulla to Medulla to
process· (per cent) clivus tip of odon- dorsal contour anterior
toid process of odontoid contour of
process cervical
canal

I 16 28 Evident Evident Evident Marked

2 8 20 Just visible Just visible Just visible Marked
3 10 27 Separation Separation Separation
C3-C4
4 II 25 Evident Evident Evident Marked
5 16 24 Just visible Just visible Evident Marked
6 8 16 Evident Evident Just visible Evident
* Since the clivus plane, the usual reference in measuring the angle between base of the skull and
cervical canal, could not be recognized in all films, the dura on the upper surface of the clivus and
on the dorsal contour of the odontoid process was used as reference.
• ':' Since all the arches were not identifiable in the lower part of the cervical spine, only shortening
between the occipital bone and the lower border of the arch of the epistropheus was measured.

canal and follow a straight course through the lordosis of the upper cervical
canal. This is clearly demonstrated in the films (excepting in those of Case 2)
by the good air filling of the subarachnoid spaces. Cases 1, 3, 4, 5, and 6
prove that the dorsal contour of the cord then approaches the posterior contcur
of the cervical canal at the level of the arches of the axis and the third vertebra.
A distinct air layer in the subarachnoidal space on the dorsal contour of the
cervical cord, both in full dorsal extension and ventroflexion, is however
always present. This rules out the possibility of compression of the cord in all
positions of the cervical spine in the cases referred to, because, for a body to be
compressed, it must lie between two rigid surfaces that are approaching one
another.
A stretched cervical cord loses contact with the anterior contour of the
canal between spondylotic protrusions (Fig. 11). Conversely, dorsal extension
in the prone position produces an increase in the area of contact between the
cord and any protrusion, owing to more marked bulging of the dura and disk
and slackening of the cervical cord.

Downloaded from acr.sagepub.com at MCMASTER UNIV LIBRARY on June 29, 2015

614 ALF BREIG AND AHMED FOUAD EL-NADI

Discussion
Functional roentgenograms. The prone and supine pOSltlOnS cause the upper
part of the cervical cord to be displaced ventrad and dorsad, respectively
(jIROUT 1956). In addition, ventroflexion and dorsal extension cause the pons
respectively to approach and separate from the clivus, and the cisterna pontis
to narrow and widen. As no account is taken of the position of the head and
cervical spine it is necessary to define the conditions more precisely: the dis-
placement of the cervical cord, ventrad in the prone and dorsad in the supine
position, takes place only from the erect position to full dorsal extension of the
cervical spine. The displacements of the pons towards the clivus in ventro-
flexion and away from it in dorsal extension take place only in the supine
position or with the body erect and the cervical spine extended. In the latter
instance the cord sags under gravity onto the dorsal surface of the cervical
canal. Moreover, only in the supine position is there a widening of the cisterna
pontis in dorsal extension.
The cervical cord, according to DECKER (1957), is not rigidly attached to
the canal. This is true so far as the anatomy is concerned but as regards function
the cervical cord in full ventroflexion is immovable in both axial and radial
directions.
It is obvious from our results that it is possible to predict fairly accurately
the mechanical behaviour of the hindbrain and the cervical cord in various
positions of the cervical spine and body postures.
Evaluation of gas myelograms of the cervical cord must take account of
the distance of the vertex of the protrusion from the anterior surface of the
cord and the posterior contour of the canal. It can furthermore usually be
noted whether the protrusion just makes contact with the cervical cord or
bulges into it. These features are of doubtful value in the clinical assessment
unless the position of the body and the degree ofventroflexion or dorsal exten-
sion of the cervical spine are considered and clearly defined. This is obvious
in view of the difference in size of the protrusions, the stretching of the spinal
cord tract and position of the latter in various postures. Hence, the pathogenet-
ic significance of any protrusion can be assessed accurately only in full ventro-
flexion. Irrespective of the posture, only in this position can an examination
reveal which protrusion becomes a hypomochlion to exert contact pressure
on the stretched cervical cord and overstretch it. A protrusion that does not
come in contact with the cord in full ventroflexion is of no pathogenic signif-
icance. Conclusive evidence of this fact is provided by the presence of a gas
strip between the ventral contour of the cervical canal and the cord. As the
roentgenograms show, stretching of the spinal cord tissue tract in full ventro-

Downloaded from acr.sagepub.com at MCMASTER UNIV LIBRARY on June 29, 2015

 BIOMECHANICS OF THE CERVICAL SPINAL CORD 615

flexion may result in loss of contact with a relatively large protrusion to which
a traumatizing action may have been ascribed. On the other hand, a small
protrusion may still be the cause of neurologic signs, even if a collection of gas
in the subarachnoidal space, in positions other than full ventroflexion, may
suggest that this is not the case. Confirmation must therefore be obtained from
functional roentgenograms in full ventroflexion.
More than one protrusion may of course function as a hypomochlion; if located
laterally, only part of the anterior contour of the cord can come into contact
with it. Tomographic films are therefore required.
Explanation of unilateral symptoms may lie in a lateral contact pressure.
In all cases in this series, excepting Case 2, gas myelograms in full ventro-
flexion demonstrated a clear strip of gas in the subarachnoidal space between
the dorsal contour of the cervical cord and the canal. According to the above
definition, the possibility of any compression of the cervical cord in this
position can therefore be eliminated. The absence of such a strip in intermediate
positions or dorsal extension is not simply conclusive evidence of compression.
The conditions that may give rise to true compression of the cord are discussed
below.

Detrim.ental forces
Only forces set up by physiologic movements will be discussed.

Contact pressure and axial tension. In full ventroflexion, when the pia is drawn
out, the nervous tissue of the cervical spinal cord is under axial and horizontal
tension.
In myelograms, a layer of gas is seen along the free dorsal contour of the
cervical cord whereas its ventral contour makes close contact with the protru-
sion. This latter exerts a pressure on the contact area of the stretched dura and
cervical cord (O'CONNEL 1956, REID 1960); this contact pressure diminishes
with the third power of the distance to the free dorsal surface of the cervical
cord where it is zero (BREW & MARIONS 1963). Hence, one of the main forces
causing cervical myelopathy is axial tension.
The type of cervical cord injury due to contact pressure and overstretching
was previously ascribed to compression of the cord (KAHN 1947) and considered
as an anterior spinal cord injury (SCHNEIDER 1962).
Contact pressure and overstretching as a cause of traumatic damage to the
cord is frequently confirmed by clinical experience. Patients suffering from cer~
vical myelopathy due to spondylotic protrusions not seldom experience in-
creasing spasticity and sometimes exhibit Lhermitte's sign in ventroflexion.

Downloaded from acr.sagepub.com at MCMASTER UNIV LIBRARY on June 29, 2015

616 ALF BREIG AND AHMED FOUAD EL-NADl

Neurologic signs by disturbance of the long motor and sensory pathways are
rarely elicited in dorsal extension, however. In fact, urinary incontinence and
spasticity are often relieved, as has been observed when the cervical spine has
been immobilized in slight dorsal extension by osteosynthesis, thus reducing
tension in the cord.
Since increased axial tension is set up also in the normal range of ventro-
flexion of the cervical spine it may be looked upon as causing continuous and
unavoidable trauma. It is produced mainly by disk protrusions and long-
standing changes in the bone. In the latter instance, muscular tension due
to (weak) pain reflex is rarely noticed and, therefore, no effective protection
exists against extreme movements of the head and cervical spine. Only in
acute derangement of the cervical spine is effective blocking evident, and then
it is difficult to assess the extent to which this may be due to subluxation,
inflammatory oedema or the pain reflex itself. Nor is there any such protection
against further damage by ventroflexion movements in the case of tumours lying
on the ventral contour of the cervical canal (foramen magnum). That dele-
terious axial tension is not observed more frequently in spondylotic protrusions
is due to the degenerative shortening of the spine, slackening of the dura
(CLARK & ROBINSON 1956, REID 1960, BREIG & MARIONS 1963) and, as seen
from the present material, also to restricted mobility of the cervical spine.
Attempts have been made to analyse the type of damage to the cervical
cord from an examination of histologic changes. This is a difficult problem,
for overlapping contact pre~sure, axial tension and radial tension, transmitted
also by the dentate ligaments, all exist within the cord substance. Another
complicating factor is the non-uniformity of the mechanical properties of the
cord. That the grey matter around the central canal is more readily displaced
axially than the white matter in the border zones of the cord was established
by MCVEIGH (1923) in his cord crush experiments. This would explain the
more severe damage to central areas of the cord (Bucy, HEIMBURGER & OBER-
HILL 1948; MAIR & DRUCKMAN 1953). This phenomenon obviously plays a
still more decisive role in central cervical cord injury (SCHNEIDER 1962) due
to true compression. In spite of these difficulties, BEDFORD, BOSANQ,UET &
RUSSEL (1952) concluded from their histologic observations that when spondy-
lotic protrusions are present the cord is most likely to be subjected to a tensile
force.

Compression. True compression of the cord occurs when the canal is narrowed
by an expansive process or by displacement of a fractured vertebra or arch
which not only bulge into the cord but press it against the opposite wall of the
canal. According to WOLF, KHILNANI & MALIS (1956), compression of the

Downloaded from acr.sagepub.com at MCMASTER UNIV LIBRARY on June 29, 2015

 BIOMECHANICS OF THE CERVICAL SPINAL CORD 617

Fig. 13. Dcmonstration of thc ccrvical canal in thc cadavcr in

dorsal cxtcnsion aftcr rcmoval of the' cord: bcforc rcmoval of thc
dura (lcft) and aftcr rcmoval of thc dura (right). (Thc lcvel of
thc barium mcdium (top right) indicatcs thc truc horizontal.)

cervical cord occurs when the anteroposterior diameter of the canal (measured
fi'om the lower border of a vertebra to the base of the opposite arch) is less
than 12 mm. Such a compression would seem to occur especially in dorsal
extension when big ventral protrusions are present and folding of the £laval
ligaments is accentuated.
The impressions in the contrast medium evident in the cervical canal in
dorsal extension are due primarily to more marked folding of the dura (Figs
12, 13 and 14). These folds around the canal are fairly soft and can within
limits be displaced axially as a stationary wave. The impressions in the contrast

Downloaded from acr.sagepub.com at MCMASTER UNIV LIBRARY on June 29, 2015

618 ALF BREIG AND AHMED FOUAD EL-NADI

.' I,

Fig. 14. Relationship in 10 subjccts bctwccn thc contour

of thc fold cd dura and thc bulging fJaval Iigamcnts in
dorsal cxtcnsion. In onc·of the cases, the extcnt of the
bulging only of the fJavalligamcnts is depicted.

Downloaded from acr.sagepub.com at MCMASTER UNIV LIBRARY on June 29, 2015

 BIOMECHANICS OF THE CERVICAL SPINAL CORD 619

medium will therefore be due chiefly to the folds of the dura and to a smaller
extent to the flaval ligaments, the bulging of which differs widely from one
case to another. Both the dural folds and the flaval ligaments are normal
structures. The dural folds can obviously be ruled out as a factor in the com-
pression mechanism because they are in themselves yielding. More space
is often available for the cord within the canal than is depicted by gas or
positive contrast myelography. Moreover, it should be borne in mind that in
dorsal extension the cord is fully slackened and thus can adapt itself to any
changes in shape of the cervical canal and the soft tissues, a property that in
some measure constitutes a protective mechanism against compression.
There is no doubt, however, that in processes encroaching on the canal
the cord must ultimately be compressed between the pathologic protrusion
and the opposite wall of the canal, which can consist of part of the arches and/or
the flaval ligaments. This is most likely to happen in full extension when the
canal is narrowed by bulging disks and ligaments and the cord is widened.
In fact, not only in the above cases with a wide canal but also in those in
which the canal is greatly narrowed by pathologic protrusions, a strip of gas
is invariably visible along the dorsal contour of the cord in slight to full
ventro£lexion. This is conclusive evidence that compression of the cervical
cord can never occur during ventroflexion, since in this position the cord
diameter is narrowed by stretching and the canal is widened by elastic stretch-
ing of the ligaments and the dura.
True compression of the cervical cord by spondylotic protrusions in full ex-
tension may be caused by a combination of conditions: a wide canal and
extremely large protrusions; a wide canal, large protrusions and bulging
£laval ligaments; an extremely shallow canal and small protrusions; a shallow
canal, small protrusions and bulging flavalligaments.
Whether a relation invariably exists between the diameter of the cord and
the lumen of the canal remains to be determined. If the cord should be dis-
proportionately large or the lumen of the canal disproportionately small,
the presence of even a small protrusion would give rise not only to contact
pressure on, and overstretching of, the cord in full ventroflexion but also
to its compression in extension. No means of demonstrating compression, for
example with a contrast medium, has existed to date, so that the presence of
repeated compression trauma might be inferred from (1) indirect evidence,
(2) absence of a gas strip between canal and cord in the neutral position of
the cervical spine, or (3), when a positive contrast medium is used, possibly
from an observed arrest in this position. Whereas the deleterious action of the
forces is clearly manifest in full ventroflexion in the gas myelogram, the
mechanism is no longer apparent between neutral position and full dorsal

Downloaded from acr.sagepub.com at MCMASTER UNIV LIBRARY on June 29, 2015

620 ALF BREIG AND AHMED FOUAD EL-NADI

G'x
(Pressure)

P
GY
(Traction)

L x
a b c
Fig. 15. The mechanical stress on the cervical cord under different static and dynamic conditions:
true compression (squeezing) (a), contact pressure and axial tension due to bending (b) and sudden
repulsion by a blow (c). The static forces obtaining in (a) and (b) have been discussed elsewhere
(BREIG & MARIONS 1963). Under the dynamic conditions prevailing in (c) a sudden trauma of the
cord, due for instance to a bulging disk and/or a simultaneous forward bending of the spine, would
cause a sharp displacement of the cord tissue. The rapid movement (acceleration) of the cord
tissue within the pia would set up an axial tension which might result in tearing. Conditions in
(a) and (c) are statically similary. Compare the tensile stress.

extension. Only from further narrowing of the canal and widening of the
cord through axial compression can it be inferred that the cord is subjected
to a deleterious (horizontal) compressive force that ultimately must lead to
dislocation of nervous tissue in axial directions (see Fig. 15).
When the canal is greatly narrowed, irrespective of its cause, and the cord,
being entrapped within a segment of the cervical canal, can no longer escape
trauma in dorsal extension through the yield of the slackened nervous tissue,
double trauma may occur. In ventroflexion, the cervical cord is damaged
by contact pressure and overstretching, and in dorsal extension it is injured
by true compression. A somewhat analogous situation occurs for a cervical
nerve root in the lateral section of a critically narrowed intervertebral foramen
in full dorsal extension. In contrast to this, the foramen in full ventroflexion
is widened and the nerve root is damaged by contact pressure on and over-
stretching by a ventral protrusion. In the case of ventral protrusions into a

Downloaded from acr.sagepub.com at MCMASTER UNIV LIBRARY on June 29, 2015

 BIOMECHANICS OF THE CERVICAL SPINAL CORD 621

Fig. 16. The Breig adjustable head

rest (manufactured by AB Stille-
Werner, Stockholm). As the cer-
vical spine is elongated during ven-
troflexion and shortened during
dorsal extension the position of the
head rest is adjusted automatically
so that the head and spine follow
a normal pattern of movement.

callal of normal width, it is therefore necessary only to avoid ventroflexion

movements, whereas marked ventroflexion and dorso-extension must be
avoided in the case of protrusions into a greatly narrowed canal.

Alleviation oj tension ill the damaged cord. When an artificial tensile stress is
added to the increased axial tension of the spinal cord tissue set up by pathologic
processes, the limiting strength of the cervical cord (the axons) may be suddenly
exceeded, with inevitable aggravation of the cord damage. In the usual
traction treatment for spinal injury, such as compressive fractures, fracture
of the odontoid process and disk hernia, prime consideration must be given
to the mechanical effect on the spinal cord tissue. In a case of damage to the
cervical cord, therefore, an attempt must be made at the earliest possible
moment to reduce the tension in the cord by reclining thc head. It follows,
moreover, that in the reduction of a spi/tal fracture and the treatment of a
disk protrusion, the cervical spine should be placed in slight dorsal extension.
In fact, certain types of fractures may be reduced spontaneously merely by
placing the neck on a soft round bolster which induces this position. Five
cases of fractures of the odontoid process have been treated successfully in
this way. In cases of damage to the cervical spinal cord, the primary neurologic
signs are often so prominent as to mask secondary damage to the cord.
As regards surgical treatment it is self-evident that the best results are
obtained when it is directed to the cause of the damage, for instance by
extirpation of bone fragments, protrusions, or expansive lesions. Such opera-
tions are greatly facilitated if the cervical cord and nerve roots are slackened
by placing the cervical spine in dorsal extension, a position that may be

Downloaded from acr.sagepub.com at MCMASTER UNIV LIBRARY on June 29, 2015

622 ALF BREIG AND AHMED FOUAD EL-NADI

effectively obtained by means of a special head support (Fig. 16). The effect
of other surgical procedures - commonly but often erroneously referred to as
decompression - that have been found to be of practical value is often not
decompression, or at least not only decompression. For instance, laminectomy,
which is decompressive in effect if there is true compression, often results also
in a change in shape of the cervical spine or restriction of its range of mobility.
The possibility of overstretching of the spinal cord and the nerve roots is
thus often reduced. It can in fact be deduced from these sequelae oflaminecto-
my - demonstrated best in postoperaLive films in extreme ventroflexion posi-
tion - that there must be a limit to the degree of stretching of the spinal cord
and the nerve roots.
To give an example: decompression is considered to be the effective factor
in SCOVILLE'S (1961) bilateral facetectomy. Laminectomy, as the first step
in the course of this operation, is of course, the usual means of eliminating any
compression. However, from the widening and backward bulging of the dura
that invariably results after laminectomy also in the absence of compression,
it cannot be inferred that compression was present before laminectomy. Nor
does this feature indicate a dorsal displacement of the whole dura, the anterior
part of which is of course still attached to Trolard's fascia and the anterior
longitudinal ligament. The bulging of the dura is simply due to the hydro-
static pressure of the cerebrospinal fluid. It is therefore more marked in the
seated than in prone posture. In Scoville's operation bilateral facetectomy is
performed after laminectomy. In effect, bilateral facetectomy would appear to
eliminate the supporting action ofthe intervertebral joints so that the vertebrae
are tipped into dorsal extension, with consequent slackening of the dura, cord
and nerve roots.
Restriction of the range of mobility of the cervical spine can be obtained
also by osteosynthesis of the cervical spine, a technique introduced by CLO-
WARD (1958) with the object of restoring the intervertebral distance, with or
without previous extirpation of bony or disk protrusions. This measure, too,
often results in relaxation of the cord aad the nerve roots. Essentially, the
effect of this method, in which two adjacent vertebrae are fixed, often at an
angle to each other, is either to eliminate the possibility of overstretching or
to produce a permanent slackening of the dura and cord and heace a reduction
of the tension.
The beneficial effect oa the hindbrain-spinal cord tract of any spinal
operation that changes the functional relation between soft tissues and the
skeletal framework can be properly understood only through a fundamental
study of the biomechanical functioning of the system. From such an analysis,
priaciples for both conservative and operative treatment might be drawn up.

Downloaded from acr.sagepub.com at MCMASTER UNIV LIBRARY on June 29, 2015

 BIOMECHANICS OF THE CERVICAL SPINAL CORD 623

Acknowledgetnent
This work was supported by a grant to one of the authors (E. B.) from Karolinska Insti-
tutet, Stockholm, Sweden.

SUMMARY
Examinations of six patients with myelopathy due to protrusions in the cervical region
by gas myelography in prone position and full ventroflexion and dorsal extension are described.
Many so-called compression injuries to the cervical cord are due to contact pressure and
overstretching as a result of tensile forces caused by protrusions. By placing the cervical spine
in extension, a reduction in tension is obtained and provides optimal conditions for recovery.

ZUSAMMENFASSUNG
Sechs Patienten mit Myelopathie infolge von Protrusionen in der Zervikalregion wurden
mittels Gasmyelographie in Bauchlage bei optimaler Flexion und Extension der Halswirbel-
saule untersucht. Viele sogenannte Kompressionsschaden des Halsmarks entstehen durch
Kontaktdruck und Uberstreckung als Folge von Zugkraften, welche von den Protrusionen
erzeugt werden. Lagerung der Halswirbelsaule in Extensionsstellung reduziert die Spannung
und schafft eine der wichtigen Voraussetzungen ftir eine Wiederherstellung.

RESUME
Description de I'examen par myelographie gazeuse en procubitus et flexion et extension com-
pU:tes de 6 malades atteints de myelopathie due a des saillies de la face anterieure du canal
rachidien dans la region cervicale. Beaucoup de lesions de la moelle cervicale attribuees a
la compression sont dues a une pression par contact et a un etirement resultant de tension
causee par ces protrusions. L'extension du rachis cervical diminue la tension et donne les
meilleures conditions pour la guerison.

REFERENCES
BEDFORD P. D., BOSANQ.UET F. D. and RUSSEL W. R.: Degeneration of the spinal cord asso-
ciated with cervical spondylosis. Lancet 263 (1952), 55.
BREIG A.: Zur Biomechanik des Zentralnervensystcms. Fortschr. Neurol. Psychiat. 1 (1959),
46.
Biomechanics of the central nervous system. Almqvist & Wicksell, Stockholm 1960.
Dehnungsverschiebungen von Dura und Rtickenmark im Spinalkanal. Fortschr. Neurol.
Psychiat. 4 (1964), 195.
Biomechanics of the spinal cord in kyphosis and kyphoscoliosis. Acta neurol. scand. 40
(1964), 196.
and MAIHONS 0.: Biomechanics of the lumbosacral nerve roots. Acta radiol. I (1963), 1141.
Bucy P. C., HEIMBURGER R. F. and OBERHILL H. R. 0.: Compression of the cervical spinal
cord by herniated intervertebral discs.]. Neurosurg. 5 (1948),471.
CLARK E. and ROBINSON P. K.: Cervical myelopathy, a complication of cervical spondylosis.
Brain 79 (1956), 483.

Downloaded from acr.sagepub.com at MCMASTER UNIV LIBRARY on June 29, 2015

624 ALF BREIG AND AHMED FOUAD EL-NADI

CLOWARD R.: The anterior approach for removal of ruptured cervical discs.]. Neurosurg. 15
(1958),602.
DECKER K.: Enzephalographie am Bildwandler. Fortschr. Rontgenstr. 87 (1957), 707.
]IROUT ].: Changes in the size of the subarachnoid spaces after the insufflation of air. Acta
radiol. 46 (1956),81.
KAHN E. A.: The role of the dentate ligaments in spinal cord compression and the syndrome
oflateral sclerosis.]. Neurosurg. 4 (1947), 191.
KEY A. und RETZIUS G.: Studien in der Anatomie des Nervensystems. Folio 166. P. A.
Norstedt & Saner, Stockholm 1875.
MAIR W. G. P. and DRUCKMAN R.: The pathology of spinal cord lesions and their relation
to the clinical features in protrusion of cervical intervertebral discs. Brain 76 (1953), 70.
l\1CVEIGH]. F.: Experimental cord crushes; with a special reference to the mechanical factors
involved and subsequent changes in the areas of the cord affected. Arch. Surg. 7 (1923),573.
O'CONNEL]. E. A.: The place of surgery in the treatment of cervical spondylosis. Proc. roy.
Soc. Med. 49 (1956),202.
REID]. D.: Effects of flexion-extension movements of the head and spine upon the spinal cord
and nerve roots.]. Neurol. Neurosurg. Psychiat. 23 (1960),214.
SCHNEIDER R. C.: Surgical indications and contraindications in spine and spinal cord trauma.
Clinical Neurosurgery. Proc. Congress Neurological Surgeons, Chicago 1960. Williams &
Wilkins Co, Baltimore 1962.
SCOVILLE W. B.: Cervical spondylosis treated by bilateral facetectomy and laminectomy.
]. Neurosurg. 18 (1961),423.
SMITH C. G.: Changes in length and position of the segments of the spinal cord with changes
in posture in the monkey. Radiology 66 (1956), 259.
SPILLANE]. D. and LLOYD H. T.: The diagnosis oflesions of the spinal cord in association with
"osteoarthritic" disease of the cervical spine. Brain 75 (1952), 177.
WOLF B. S., KHILNANI M. and MALIS L.: The sagittal diameter of the bony cervical spinal
canal and its significance in cervical spondylosis.]. Mt Sinai Hosp. 23 (1956),283.

Downloaded from acr.sagepub.com at MCMASTER UNIV LIBRARY on June 29, 2015