Small Animal Spinal Disorders (Second Edition)

Dedication
Professor Ian R. Griffiths, Glasgow University Medical School

and
Professor Joe N. Kornegay, College of Veterinary Medicine, University of Missouri
For their inspiration, encouragement and guidance
For Elsevier:
Commissioning Editor: Joyce Rodenhuis
Senior Development Editor: Zo A Youd
Project Manager: Ailsa Laing
Designer: Andrew Chapman
2005, Elsevier Limited. All rights reserved.
Copyright 1994 Times Mirror International Publishers Limited

Copyright 2000 Harcourt Publishers Limited
2004 Elsevier Limited. All rights reserved.
No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any
form or by any means, electronic, mechanical, photocopying, recording or otherwise, without either
the prior permission of the publishers or a licence permitting restricted copying in the United
Kingdom issued by the Copyright Licensing Agency, 90 Tottenham Court Road, London W1T 4LP.
Permissions may be sought directly from Elseviers Health Sciences Rights Department in
Philadelphia, USA: phone: (1) 215 238 7869, fax: (1) 215 238 2239, e-mail: healthpermissions
@elsevier.com. You may also complete your request on-line via the Elsevier Science homepage
(http://www.elsevier.com), by selecting Customer Support and then Obtaining Permissions.
First edition 1994

Second edition 2005
ISBN 0723432090
British Library Cataloguing in Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloging in Publication Data
A catalog record for this book is available from the Library of Congress
Notice
Knowledge and best practice in this field are constantly changing. As new research and experience
broaden our knowledge, changes in practice, treatment and drug therapy may become necessary
or appropriate. Readers are advised to check the most current information provided (i) on procedures
featured or (ii) by the manufacturer of each product to be administered, to verify the recommended
dose or formula, the method and duration of administration, and contraindications. It is the
responsibility of the practitioner, relying on their own experience and knowledge of the patient, to
make diagnoses, to determine dosages and the best treatment for each individual patient, and to
take all appropriate safety precautions. To the fullest extent of the law, neither the publisher nor the
editors assumes any liability for any injury and/or damage.
The Publisher
Printed in China
Preface
The second edition of Small Animal Spinal Disorders

now expands considerably on the foundations of neuroanatomy, clinical neurology and basic neurosurgery
provided in the first edition. The primary aim of the
second edition remains to assist students, general practitioners and specialists in the diagnosis and understanding of the latest neurosurgical procedures used
to manage dogs and cats with spinal disease. The book
has now been completely rewritten and reorganized
for 2005. Most of the pre-1994 literature has been
reviewed again so that we do not forget lessons learned
previously and also so that this knowledge can be
placed in proper context with more recent information. Many exciting contributions have been made in
the last 10 years, but not surprisingly veterinary neurosurgery still tends to suffer from limited case numbers,
with follow-up periods often less than a residencytraining period in duration. Extensive reference has
therefore been made to the human neurosurgical literature, both in an attempt to gain a different perspective
and to stimulate a continued reassessment of our current state of knowledge. Despite obvious differences
and limitations, the greater maturity and collective
knowledge of the human specialty provides important
insights into veterinary neurosurgery.
In addition to being completely rewritten, the second edition now contains 50% more figures. A much
greater emphasis has now been placed on the advanced
imaging techniques of CT and MRI; the chapters on
functional anatomy and diagnostic aids in particular
reflect these changes. Advanced imaging is also used
liberally in the other chapters to illustrate surgical
anatomy, pathology and complications.
Another new innovation emphasizes the problem of
postoperative complications. A specific focus is the
wide variety of complications that can occur during
the management of spinal disorders. Complications are
generally divided into intraoperative, early postoperative and late postoperative. These are listed in tabular
format and are also discussed and illustrated where
possible in the main text. Furthermore, prognostic
information continues to be derived mainly from the
literature rather than from anecdote, at least where
information is available. Physiotherapy now receives
four pages of special emphasis in the postoperative care
chapter. An additional change is the separation of the
procedures section from the main chapter, so that photographs of surgical techniques no longer interrupt the
text. Finally, at the end of most chapters, key issues for
future investigation have been identified, to highlight
areas where we need to focus our creative energy and
future research efforts.
Additional information relating to this text and to
neurosurgery in general has been made available on the
Internet at www.vetneurosurgery.com. This includes
direct links that provide abstract information to many
of the citations used in the second edition, along with
further information on CT and MRI anatomy. It also
provides many links to other websites that contain
information aimed at owners, referring veterinarians
and specialists.
There will come a time when veterinary neurosurgery becomes a specialty recognized in its own right,
just as it is in human healthcare. Until that time it will
fall upon individuals, mainly from the European and
North American specialties of neurology and surgery,
to advance our knowledge one step at a time. This book
aims to bring all of the various quite disparate sources
of information together, in order to provide a comprehensive summary of the current state of knowledge of
veterinary neurosurgery.
Vancouver and Hertfordshire 2005
Nick Sharp
Simon Wheeler
List of Abbreviations
ACE
b.i.d.
CDV
CK
CMC
CNS
COX
CSM
CSF
CT
DISH
DVT
ECG
EMG
FCE
FeLV
FIP
FIV
GI
GME
IM
angiotensin-converting enzyme
bis in die (twice daily)
canine distemper virus
creatine kinase
cerebello medullary cistern
central nervous system
cyclooxygenase
cervical spondylomyelopathy
cerebrospinal fluid
computed tomography
disseminated idiopathic skeletal hyperostosis
deep vein thrombosis
electrocardiogram
electromyography
fibrocartilaginous embolism
feline leukemia virus
feline infectious peritonitis
feline immunodeficiency virus
gastrointestinal
granulomatous meningoencephalomyelitis
intramuscular
IV
LMN
MPSS
MR
MRI
NSAID
OCD
PCV
PgE
PO
PTE
RBC
SQ
STIR
TMJ
TSH
UMN
UTI
VW
WBC
intravenous
lower motor neuron
methylprednisolone sodium succinate
magnetic resonance
magnetic resonance imaging
non-steroidal anti-inflammatory drug
osteochondritis dissecans
packed cell volume
prostaglandin E
per os
pulmonary thromboembolism
red blood cells
subcutaneous
short tau inversion recovery
temporomandibular joint
thyroid-stimulating hormone
upper motor neuron
urinary tract infection
von Willebrand
white blood cell
Functional anatomy
Skeleton 6
Vertebrae 6
Articulations 9
Blood supply 14
Vertebral column 14
17
Further reading
17
Knowledge of functional anatomy is important both

for understanding the neurological examination and for
performing spinal surgery. This chapter concentrates
on clinically relevant points of anatomy and physiology,
including surgical landmarks. Radiographs, CT scans and
MRIs have been used for illustration, and the reader is also
directed to the normal radiographic anatomy illustrated in
Chapter 4. For more detail see References, page 17.
c
a
b
e
NERVOUS TISSUE
Spinal cord
Nervous tissue 1
Spinal cord 1
Relationship of spinal cord segments to
vertebrae 2
Cauda equina 3
Meninges 3
Cerebrospinal fluid 4
Spinal cord white matter tracts 5
Spinal cord nerve fibers and the effect of
compression 6
References
Chapter
The spinal cord lies within the vertebral canal, fitting

snugly within the thoracolumbar spine, but with more
space in the cervical spine. The residual space is filled
with epidural fat. The spinal cord extends from the
caudal limit of the brainstem at the foramen magnum
to the caudal lumbar vertebrae, terminating in the sixth
lumbar vertebra (L6) in most dogs, and in L7 in cats,
with some variations.
The spinal cord is divided into segments:
Cervical C1C8.
Thoracic T1T13.
Lumbar L1L7.
Sacral S1S3.
Caudal (variable number).
The spinal cord is wider at the cervical and lumbar intumescences (segments C6T2 and L4S3 respectively),
from which the lower motor neurons (LMN) to the
thoracic and pelvic limbs arise. These segments contain
the cell bodies for the LMNs, also known as ventral horn
cells, thus the spinal cord is thicker in these areas.
The spinal cord is composed of central gray matter
and peripheral white matter (1.1). A dorsal sulcus and
ventral fissure, lined by pia mater, divide the spinal
cord into two halves. Dorsal and ventral nerve roots
exit the spinal cord at each segment and join to form
the segmental spinal nerves. There are eight cervical
segments, but seven cervical vertebrae. The C1 spinal
nerves leave through the lateral foramina in C1 vertebra.
1.1 Spinal cord in transverse section. The
gray matter forms an H shape, with two
dorsal horns (a) and two ventral horns (b).
The white matter tracts are divided into
dorsal funiculi, between the dorsal roots (c);
lateral funiculi, between dorsal and ventral
roots (d); and ventral funiculi, between the
ventral roots (e).
Small Animal Spinal Disorders
1.2 Position of the spinal cord segments

in the cervical and cranial thoracic
vertebrae. The cervical intumescence
(C6T2) lies within vertebrae C4T2.
Thus, lesions as far cranial as C4/C5
vertebrae may cause LMN signs in the
thoracic limbs.
4
5
6
1.3 Normal cervical MRI. A: Sagittal, T2-weighted MRI of the cervical spine in a 7-year-old Doberman. It had undergone a ventral
slot at C6/7 2 years ago and is now clinically normal. The other discs show the expected high signal intensity of the nucleus
pulposus (arrow). B: Transverse, T2-weighted MRI of the cervical spine through mid-C5 vertebra; the gray matter is visible clearly
within the spinal cord, which is surrounded by a hyperintense rim of cerebrospinal fluid (arrow). The black crescent on one side
between CSF and epidural fat is a chemical shift artifact.
C5
T2
C6
T1
C8
C7
The rest of the cervical spinal nerves exit the vertebral

canal cranial to the vertebrae of the same annotation,
except C8 nerves, which exit between C7 and T1 vertebrae. The thoracic and lumbar spinal nerves exit
behind the same-named vertebrae.
The nerve roots are partly ensheathed by meninges,
which are continuous with the epineurium.
Relationship of spinal cord segments

to vertebrae
a
d
1.4 Brachial plexus: lateral thoracic (a); ulnar (b); median (c);
radial (d); axillary (e); and musculocutaneous nerves (f). The
subscapular and suprascapular nerves arise just cranial to
the musculocutaneous nerve.
Some spinal cord segments lie in the vertebra of the

same annotation, but others do not (1.2, 1.5). Neurological lesion localization refers to spinal cord segments.
It is therefore important to understand the relationship
between vertebrae and spinal cord segments. Although
regional nerves of the brachial and lumbosacral plexuses
are relatively constant in their distribution they can
vary considerably in their origins. The plexuses arise
from C6 to T1 and from L5 to S3 spinal cord segment
in the majority of animals (1.4, 1.8A). However, the
brachial plexus arises from C5 to T1 in about 20% of
dogs and from C6 to T2 in another 20% (Evans, 1993).
Similarly the lumbosacral plexus arises from L3 to S1
in 20% and from L6 to S3 in 20% (Fletcher, 1970).
This can have some impact on the neurological localization within these regions.
Functional anatomy
1.5 Position of spinal cord segments within the lumbar vertebrae. Segments L1 and L2 lie in their respective vertebrae. The lumbar
intumescence lies within vertebrae L3L5; lesions as far cranial as L3/4 intervertebral disc may cause LMN signs in the pelvic limbs.
The sacral segments S1S3 are within L5 vertebra in most dogs (the 5 in L5 resembles the S for sacral). The spinal cord ends in
L6 in most dogs; L7 in cats. The cauda equina runs from L5 vertebra into the sacrum (1.7, 1.8B).
1.6 Normal thoracolumbar MRI. A: Sagittal, T2-weighted MRI of the thoracolumbar spine of a normal 6-year-old Golden retriever.
Nucleus pulposus in this pulse sequence shows intermediate signal intensity; the thin layer of cerebrospinal fluid (CSF) is of
intermediate signal intensity (arrow); epidural fat is high signal intensity (arrowhead). Note the nutrient vessels arising from the aorta
(arrowheads) (Parker, 1973). B: Transverse, T1-weighted MRI of the thoracic spine at the level of T13. The spinal cord is surrounded
by high signal epidural fat (arrow); CSF is of low signal intensity and does not show clearly. Note the aorta (arrowhead) (1.23).
a
1.7 Nerve roots of the cauda equina. Conus medullaris (a). Nerve roots and rootlets (b). Intervertebral foramen (c). Spinal ganglion
(d). Spinal nerve (e). L6 vertebra (f). L7 vertebra (g). Sacrum (h).
Cauda equina
Meninges
The nerves of the cauda equina have a typical peripheral

nerve structure and are partly ensheathed by the
meninges (1.7). They tolerate deformation better than
spinal cord, and there is also a large epidural space in the
region of the cauda equina (1.8B, 1.9). Thus they are
usually more resistant to injury than spinal cord tissue,
but if severe damage occurs, recovery is unlikely (13.22).
The meninges (1.10) surround the central nervous

system (CNS). The arachnoid mater and pia mater
together are termed the leptomeninges. Between the
pia mater and the arachnoid mater is a space, the
subarachnoid space, which is filled with cerebrospinal
fluid (CSF). The subarachnoid space is traversed by
the arachnoid trabeculae, which suspend the spinal
L5
L6
L5
L7
L6
a
1.8 A: Lumbosacral plexus: femoral (a);

obturator (b); sciatic (c); and pudendal
nerves (d). B: Dorsal, T1-weighted
MRI of a normal 6-year-old German
short-haired pointer to show the nerve
roots of the cauda equina surrounded
by high signal epidural fat (same dog
as 1.9).
L7
d
c
B
1.9 Normal lumbosacral MRI. A: Sagittal,
T2-weighted MRI of the lumbosacral
spine in extension (same dog as 1.8).
There is loss of signal of the L7/S1
nucleus pulposus (arrowhead) but there is
no dorsal displacement of the disc.
B: Transverse, T1-weighted MRI through
the L7/S1 disc space and foramen of the
same dog; the nerve roots (arrowheads)
are surrounded by high-signal epidural fat.
See also 1.19B.
1.11 Section of spinal cord to show the subdural space (between

dura and arachnoid mater, 1.10) filled with Indian ink. White
denticulate ligaments (arrows) anchor the pia and spinal cord to
the dura; they also restrict ventral extension of the ink within the
subdural space (from Penderis et al., 1999).
1.10 There are three layers of meninges. The most superficial
is the dura mater, which is composed of dense connective
tissue (blue). The thin arachnoid mater (red) is inside the dura
mater and lies adjacent to it. These two membranes follow the
larger contours of the spinal cord. The pia mater (green) is a
delicate layer that lies directly on the surface of the spinal cord.
cord within the CSF. The denticulate ligaments also

traverse the subarachnoid space to anchor the spinal cord
(1.11). The meninges caudal to the conus medullaris
form the filum terminale, which extends into the
sacrocaudal vertebrae. The meningeal sac is outlined by

myelography or MRI. The caudal limit to its extension
varies between animals; it can terminate anywhere
between L7 and the caudal vertebrae, but usually ends
in the sacrum (see Chapter 10).
Cerebrospinal fluid
Cerebrospinal fluid is formed in the brain, mainly by
the choroid plexuses, with contributions from the
Functional anatomy
Sensory to brain
1.12 The reflex arc of the LMN is influenced

by the UMN systems.
Upper motor neuron
Spinal ganglion
Sensory
fiber
Lower motor neuron
leptomeninges (pia-arachnoid mater) and the ependymal lining.

Cerebrospinal fluid flows mainly in a caudal direction.
Most leaves the fourth ventricle of the brain through the
lateral apertures into the subarachnoid space, with some
entering the central canal of the spinal cord. It is absorbed
through the arachnoid villi in cerebral venous sinuses as
well as by venules in the subarachnoid space, lymphatics
around the spinal nerves, and the ependymal lining.
The CSF is normally a clear, colorless fluid with a
very low protein and cellular content. It suspends and
protects the brain and spinal cord against shock, allows
some variation in the volume of the CNS without altering pressure, and has some nutritional and metabolic
functions.
The caudal direction of flow of CSF has some clinical
relevance (14.3A). Cerebrospinal fluid collected caudal
to a lesion is also more likely to provide diagnostic
information (see Chapter 4).
Spinal cord white matter tracts

ASCENDING SENSORY TRACTS
Sensory information is gathered from the peripheral
nervous system via sensory axons. The cell bodies of
these axons lie in the spinal ganglia (dorsal root ganglia). Central projections of these axons ascend in the
spinal cord to the brain (1.12).
Proprioception is transmitted in the tracts of the
dorsal and lateral funiculi. Axons project to either the
somesthetic cerebral cortex or to the cerebellum.
Temperature and superficial pain sensation are transmitted by the myelinated fibers of several tracts,
including the lateral spinothalamic in the lateral funiculus. Severe pain sensation (deep pain or nociception) is
carried by non-myelinated fibers, particularly in the

propriospinal and spinoreticular tracts. These tracts lie
close to the junction of gray and white matter and therefore a lesion has to be extensive in order to damage
all deep pain fibers at a given level. Pain fibers cross
and re-cross the midline in a multisynaptic arrangement
throughout the spinal cord, providing a diffuse bilateral
pattern of ascending pain fibers from each limb.
Information on the degree of urinary bladder filling
is transmitted to the brain in the spinothalamic tract.
DESCENDING MOTOR TRACTS

Two systems are responsible for the transmission of
motor functionthe upper motor neuron (UMN) and
LMN systems (1.12, 1.13).
The LMN is the effector neuron of the reflex arc. The
cell bodies are the ventral horn cells, which lie in the
ventral gray matter of the spinal cord. The axons leave
the spinal cord in the ventral roots and pass through the
brachial and lumbosacral plexuses to form the peripheral
nerve trunks of the limbs. The sensory arm of the reflex
arc is the sensory neuron. It arises in the periphery and
enters the spinal cord via the dorsal root. It projects to
the LMN (via an interneuron in some reflex pathways)
and a branch also ascends in the spinal cord.
Function of flexor muscles is facilitated by the corticospinal and rubrospinal tracts. The fibers of the corticospinal tract arise in the cerebral cortex, and most
decussate at the spinomedullary junction and descend
in the lateral corticospinal tracts of the lateral funiculi.
Fibers that do not decussate descend in the ventral
corticospinal tracts, which lie in the ventral funiculi.
The rubrospinal fibers originate in the red nucleus of
the brainstem, cross the midline and descend in the
1.13 The UMN system is the sum effect of the

various individual descending pathways. The
UMN system in general moderates LMN activity,
initiates voluntary movement and maintains
normal muscle tone.
UMN
LMN
LMN
rubrospinal tract of the lateral funiculus. The vestibulospinal tracts and reticulospinal tracts also influence
motor function. The function of extensor muscles is
facilitated by these tracts, which lie in the ventral funiculi. The vestibulospinal fibers arise in the ipsilateral
vestibular nuclei. They facilitate extensors and inhibit
flexors on the ipsilateral side, and have the opposite
effect on the muscles of the contralateral limbs.
Voluntary bladder emptying is mediated through
fibers in the tectospinal and reticulospinal tracts of the
ventral funiculi.
ASCENDING MOTOR TRACT

In dogs, an ascending motor tract originates in the border cells of the dorsolateral gray matter of the L1L7
spinal cord segments. Their axons pass cranially in the
fasciculus proprius of the lateral funiculus to inhibit
the extensor muscles of the thoracic limbs. Interference
with this pathway, as seen in some peracute, severe
thoracic spinal cord lesions, is manifest as the Schiff
Sherrington sign (2.17).
Spinal cord nerve fibers and the

effect of compression
The white matter tracts of the spinal cord are composed
of nerve fibers of different sizes, most of which have a
myelin sheath. The largest fibers are myelinated, which
are the most rapidly conducting; they transmit proprioception. Motor fibers are intermediate-sized myelinated
fibers. Pain perception is transmitted by the smallest
myelinated fibers and by non-myelinated fibers.
Larger diameter fibers are more susceptible to injury
than fibers of lesser diameter; small fibers are the most
resistant. The progression of clinical signs seen with
increasing spinal cord damage is explained largely by
this feature. Mild lesions cause loss of proprioception.
Increasingly severe lesions cause loss of the ability to
bear weight, loss of voluntary movement and, finally,

loss of deep pain sensation.
The position of the spinal cord tracts also contributes
to the progression of signs. The ascending proprioceptive
tracts lie superficially in the spinal cord and, therefore,
are most susceptible to compression. In contrast, the
spinothalamic tracts and ascending propriospinal pathways, which carry pain perception, are more deeply positioned, and the fibers cross the spinal cord at various
levels. Thus a lesion must involve most of the diameter
of the spinal cord for the patient to lose deep pain sensation (6.2) (Olby et al., 2003). This point and the fact
that pain fibers are the most resistant to pressure explain
why loss of deep pain sensation is such a severe clinical
sign (see Assessing the severity of the lesion, page 31).
Animals with lesions severe enough to damage transmission of deep pain along the cervical spinal cord either do
not survive their injury or are at high risk of death from
hypoventilation because of loss of respiratory muscle
function (see Chapter 2).
SKELETON
The vertebral column is composed of a series of vertebrae, most of which are joined by the intervertebral
discs and by synovial joints between the articular
processes.
Vertebrae
The numbers of different types of vertebrae are as
follows: cervical 7; thoracic 13; lumbar 7; sacral 3; and
caudal 20 (approximately). Variations are possible,
particularly in the transition zones between thoracic
to lumbar, and lumbar to sacral vertebrae. The most
common variations are in the number of ribs (8.19,
8.21), and abnormal articulations with the ilium (see
Chapter 10). The importance of this lies in recognizing
Functional anatomy
e
d
1.14 The atlas C1 (a) articulates with the skull (b)

via the atlanto-occipital joints. C1 has prominent
transverse processes (c), which can be palpated
easily. There is no spinous process on the dorsal
arch (d). There are lateral vertebral foramina in
the vertebral arch (e), through which pass the C1
spinal nerves. The axis or C2 has a large spinous
process (f), which extends cranially over the atlas
and is connected to it by the dorsal atlantoaxial
ligament (g). The atlas and axis are connected by
multiple ligaments (1.35) and by synovial joints
between the articular processes (h), which lie
ventral to the vertebral canal.
c
h
1.15 CT scans of a normal dog through

A: the mid-body of the atlas C1: note
the transverse foramina and the dens on
the floor of the vertebral canal (9.1), and
B: the cranial portion of the axis C2.
surgical landmarks (8.198.21). The total number of

thoracic and lumbar vertebrae is generally 20.
The vertebrae have various common features, but
there are differences between the groups. Each vertebra
has a vertebral body, which lies ventral to the spinal cord,
and is joined to its neighbors by intervertebral discs.
In immature animals, the vertebral bodies have cranial
and caudal growth plates, which close by about 11 months
of age in dogs (Hare, 1961). The center of the vertebral
body is composed of cancellous bone, which is red and
relatively soft. The margins of the vertebral body are
made of hard, dense, white cortical bone, which also
forms the vertebral end plates adjacent to the intervertebral discs. The types of bone provide an important guide
to the depth of penetration in surgical procedures (8.32,
8.33, 10.30).
Each vertebra has a vertebral arch, which forms the
dorsal and lateral parts of the vertebral canal enclosing
the spinal cord. The arch is made up of the pedicles
lateral to the vertebral canal and of the lamina dorsally.

The vertebral arch also has cortical and cancellous bone,
although the cancellous bone may be thin in small dogs
and cats. Most vertebrae have transverse processes projecting laterally from the vertebral body, a spinous process
projecting dorsally from the lamina, and cranial and caudal articular processes on the vertebral arch. Mammillary
processes of the thoracic vertebrae are short, knob-like
dorsal projections from the transverse processes. On lumbar vertebrae they are dorsal projections of the cranial
articular processes (Evans, 1993). Other bony processes
vary with the group of vertebrae. Between each pair of
vertebrae there is an intervertebral foramen, through
which pass the spinal nerves and blood vessels.
CERVICAL VERTEBRAE
There are seven cervical vertebrae. The first two are distinct: the atlas (C1) and axis (C2) (1.14, 1.15). The vertebral body of C1 is very small, the bulk of the vertebra
being composed of lateral masses. Caudally on the body

of C1, there are two articular processes, which articulate
with C2. There is no intervertebral disc between C1 and
C2. A prominent ventral tubercle lies on the caudoventral aspect of C1, just cranial to the intervertebral space,
which can be a useful landmark at surgery. There are
transverse foramina in the transverse processes, through
which pass the vertebral arteries (1.36). The dens projects cranially from the vertebral body of C2 into the
atlas, lying on the floor of the vertebral canal (9.1). The
dens originates embryologically as part of C1. It has a
growth plate at its attachment to the body of C2, which
can separate after trauma. The atlas has three ossification centers at birth; the axis has four at birth and
another two or three that develop after birth (Hare,
1961; Cook and Oliver, 1981).
The other cervical vertebrae each have a similar morphology (1.16, 1.17). The large transverse processes of
C6 are particularly large and project ventrally; they are

important surgical landmarks. The C5/6 intervertebral
disc lies between the cranial edges of these transverse
processes (4.6). A vascular channel runs through the
center of these vertebrae and can give rise to hemorrhage and other complications (1.18, 1.19).
THORACIC AND LUMBAR VERTEBRAE

The 13 thoracic vertebrae articulate with the ribs
(1.201.22A). The spinous processes of the cranial thoracic vertebrae (to T10) slant caudally; those of the last
two thoracic vertebrae slant cranially. The site of this
change in direction (which may vary) is termed the anticlinal vertebra (1.20B). The articular processes from T1
to T10 sit at the base of the spinous process. They are
in the same oblique horizontal plane as in the cervical
vertebrae, with the caudal process of the cranial-most
a
d
c
b
1.16 Cervical vertebrae. The spinous processes are small (a),
and the transverse processes (b) project laterally (with the
exception of C6, where they are directed ventrally (4.6, 7.37B).
There are transverse foramina (c), through which pass the
vertebral arteries (1.17, 1.36), except in C7. The articular
processes (d) lie in an oblique dorsal plane.
1.18 Photograph of the floor of the vertebral canal of C4, C5

and C6 with the lamina and pedicles removed. Note the large
vascular foramen in the center of each vertebra (arrows). This
can result in significant hemorrhage when drilling into bone
and can impact screw placement (11.16).
1.17 CT scans through the mid-bodies of A: C5; B: C6; and C: C7 vertebrae. Note the transverse foramen in C5 and C6. The
transverse processes of C7 have no foramina and have a characteristic horizontal orientation.
Functional anatomy
1.19 Nutrient foramen (arrows) in the center of

lumbar vertebrae. A: Transverse CT scan of
L4 vertebra. B: Transverse, T1-weighted MRI of
L7 vertebra after fat suppression (same dog as
1.8). Bleeding from this structure is a common
cause of severe hemorrhage when drilling into
a vertebra; it can be stopped by direct pressure
or with bone wax (1.6A, 1.18, 1.38, 8.2B).
b
c
A
b
a
B
1.20 A: Thoracic vertebrae. The vertebral bodies are small with
large spinous processes (a). The articular processes vary with the
location along the thoracic spine (b). The transverse processes are
short and have a fovea (c) that articulates with a rib. Caudal to the
mid-thoracic area, caudally projecting accessory processes are
present on the pedicle (13.55A). B: Radiograph of the anticlinal
region. Spinous processes of cranial thoracic vertebrae (a); caudal
thoracic and lumbar vertebrae (b); and T11, the anticlinal vertebra
(c). This can be a useful surgical landmark, but the position must
be ascertained in each animal from the radiographs.
vertebra overlying the cranial process of the caudalmost vertebra (1.20A). In the caudal thoracic vertebra,
the articular processes adopt a more vertical orientation, as between the lumbar vertebrae (13.55). The
change in orientation occurs near the anticlinal vertebra.
The location of the vertebral canal relative to external vertebra landmarks at the thoracolumbar junction is
variable, due largely to the orientation of the ribs (1.21,
1.22). It appears fairly constant within each breed of
dog and also in the cat.
The aorta and vena cava are in close proximity to
many of the thoracic and lumbar vertebrae. Damage
could be caused to these vascular structures when drilling or placing implants into vertebral bodies (Garcia
et al., 1994). The relationship of the aorta to T13 vertebra is shown in 1.6B and 1.23A; to L1 in 1.23B and
4.44B; and to L2 in 4.43.
There are seven lumbar vertebrae (1.241.28A). The
L1 transverse processes are usually small, an important
feature in identifying the thoracolumbar junction at
surgery (8.19, 8.20). They are also difficult to palpate
as they are obscured by the last rib. More caudally, the
transverse processes are narrower and longer.
The three sacral vertebrae are fused into one body,
which articulates with the pelvis via the sacroiliac joint
(1.28B). There is a marked notch in the cranial lamina of
the sacrum, such that the lamina is not complete over
the cauda equina at the lumbosacral junction (10.48).
There are two pairs of dorsal and ventral sacral foramina, through which pass spinal nerves and blood vessels.
The sacrum articulates caudally with the first caudal
vertebra (1.26).
Articulations
SYNOVIAL ARTICULATIONS
The articular processes of the vertebral bodies have dorsal articulations (except for C1/2 where the articulations
10
1.21 CT scans through the mid-bodies of A: T11, T12, T13 and L1 vertebrae in a Dachshund (same dog as 8.60, 8.61B); and
B: T12, T13, L1 and L2 vertebrae in a Springer spaniel after myelography. There is accumulation of contrast within the center of the
spinal cord at T12, as well as leakage through a presumed dural tear (arrow) after trauma (13.10). The rib articulates progressively
higher up the pedicle from T11 vertebra to T13.
1.22 CT scans through the
mid-bodies of A: T12 in a cat; and
B: L2 in a cat. T12 vertebra of a
cat is shown in 1.23A. Compare
these images with the dogs shown
in 1.21.
B
1.23 The proximity of the aorta
(arrowheads) is shown relative to
A: T12T13 intervertebral space in
a cat (shown on a CT scan); and
B: L1 vertebra in a dog (shown on a
T1-weighted MRI). The vena cava is also
evident (arrow). The cava lies to the right
and somewhat ventral to the aorta.
Functional anatomy
d
a
c
d
1.24 Lumbar vertebrae. The vertebral bodies are long (particularly

in cats). The spinous processes are short and blunt and slant
cranially (a). The transverse processes project laterally and
cranially (b). The articular processes are vertically oriented (c);
the caudal process of the cranial-most vertebra lies medial to the
cranial process of the caudal-most vertebra. There are accessory
processes on the pedicles (13.55C), which project caudally (d).
1.26 Lumbosacral vertebrae. L7 differs from the other lumbar

vertebrae in that the spinous process is shorter (a). The
intervertebral foramen (b) lies cranial to the lumbosacral disc
(1.27). The sacrum comprises the three fused sacral vertebrae (c),
the spinous processes of which are fused and form a continuous
ridge of bone (d). The lateral wing of the sacrum (e) articulates
laterally with the wing of the ilium.
Sacrum
L7
1.27 Sagittal MRI of a dog to show that the intervertebral

foramen of L7 (arrows) extends much more cranially than it does
in other vertebrae (L6/7 and L7/S1 disc spaces shown by
arrowheads). This foramen is thought to act as a stress riser for
L7 fractures (1.26, 13.14, 13.21).
are ventral, and between the fused sacral vertebrae).

These joints have a capsule, articular cartilage and synovial fluid.
INTERVERTEBRAL DISCS
B
1.25 CT scans through the mid-bodies of A: L6; and B: L7
vertebrae. Note that the transverse process is located a little more
dorsally on L7 relative to the vertebral canal than on L6 (13.56B).
The vertebral bodies are joined by intervertebral discs

(1.3A, 1.6A, 1.9A, 1.29), with the exception of C1/2
along with the fused sacral vertebrae.
The intervertebral discs provide flexibility to the
vertebral column, and act as shock absorbers for the
spine. The capacity to absorb shock is diminished by
age and degenerative changes. Intervertebral discs have
a poor blood supply; nutrients gain access by diffusion.
The anulus fibrosus is supplied with pain fibers, mainly
in the outer laminae (Forsythe and Ghoshal, 1984).
11
12
1.28 The L7 nerve root runs in a lateral recess in the floor of

the vertebral canal of L7 before emerging from the intervertebral
foramen. A: The lateral recesses can be seen clearly (arrows) in
this CT scan taken through the L7 vertebra. B: CT scan taken
through the sacrum and sacroiliac joints.
b
1.29 The divisions of the intervertebral disc are the outer anulus
fibrosus (a) and the inner nucleus pulposus (b). The anulus
fibrosus is made up of concentric fibrous laminae. The laminae
may not form complete rings, but they are interconnected with
the adjacent laminae, thus forming a strong complete structure.
The anulus fibrosus is thicker ventrally and laterally than it is
dorsally and the nucleus pulposus is therefore eccentrically
positioned (1.9A). The anulus fibrosus is attached firmly to the
vertebral end plates by deeply penetrating (Sharpeys) fibers.
The nucleus pulposus is a gelatinous structure in young dogs,
but its characteristics change with age (see below).
Degeneration of the intervertebral discs occurs with

age and may precede disc herniation (Hansen, 1952). In
humans there is a general loss of water content (Jenkins
et al., 1985; Hickey et al., 1986) and of proteoglycans
with age (Taylor et al., 2000). Similar changes have
also been reported in dogs (Gysling, 1985; Cole et al.,
1986; Bray and Burbidge, 1998). The two main types of
disc degeneration in dogs are chondroid and fibroid
metamorphosis:
Chondroid metamorphosis occurs in
chondrodystrophoid breeds during the first 2 years
of life. As the disc degenerates, it dehydrates and at
the same time the nucleus pulposus is invaded by
hyaline cartilage. These two processes interfere
with the shock absorbing capacity of the disc by
reducing the hydrostatic properties of the nucleus
pulposus, and by weakening the fibers of the anulus
fibrosus. In most Dachshunds, by 2 years of age the
majority of discs have undergone chondroid
metamorphosis, and many nuclei have also
mineralized, changing from the former jelly-like
consistency into a dry, gritty substance. Normal
wear and tear often causes severe weakening of the
intervertebral discs, especially at the thoracolumbar
junction. This explains why the peak incidence of
disc disease is between 3 and 6 years of age for
most of the chondrodystrophoid breeds of dog.
Herniation of this type of disc is often termed
Hansen type I or a disc extrusion (1.30, 1.33).
Fibroid metamorphosis occurs in nonchondrodystrophoid breeds after middle age. The
nucleus pulposus dehydrates but is invaded by
fibrocartilage rather than hyaline cartilage. This
process has a much later onset than in chondroid
metamorphosis, and the discs are usually quite
normal while the dog is young and active. The
nucleus pulposus does not undergo mineralization
as frequently as in discs that undergo chondroid
metaplasia. Clinical problems occur primarily in
older dogs and this type of herniation is often
termed Hansen type II disease (1.31). Hansen
type II lesions may or may not be responsible for
clinical signs (see below).
DEFINITION OF DISC LESIONS

Although the Hansen classification system is useful
as a general guide, there is considerable overlap between
type I and type II disc lesions. Each type of lesion can
also occur in both chondrodystrophoid and nonchondrodystrophoid breeds (Cudia and Duval, 1997).
Furthermore, there is a spectrum of chondrodystrophoid breeds ranging from extreme like the Shih Tzu,
Functional anatomy
1.30 Hansen type I disc extrusion is most common in

chondrodystrophoid breeds. Following chondroid metamorphosis,
the nucleus pulposus extrudes into the vertebral canal
through the damaged anulus fibrosus. The nucleus may take a
tortuous route through the anular fibers or may explode through
a large defect (1.33).
et al., 1995). The classification scheme used most

commonly for CT and MRI employs the terms normal,
bulge, protrusion, or extrusion. Extrusion is a fairly
distinct category used to describe disc material that has
clearly escaped the normal boundaries of the disc,
often to cause injury to neural structures.
Disc extrusions are rare in asymptomatic humans,
occurring in only 2 of 98 people in one study (BrantZawadzki et al., 1995). However the terms normal,
bulge or protrusion include discs with identical degrees
of disruption and associated symptoms in humans.
Furthermore, there is only moderate interobserver
agreement when applying these three terms to MRIs of
disc lesions (Brant-Zawadzki et al., 1995; Milette et al.,
1999). A purely morphological description of the displacement of disc material is therefore preferred to use
of the terms normal, bulge and protrusion. An example
would be dorsolateral displacement of the L7/S1 disc
to the right causing impingement on the L7 nerve root
(Gorman and Hodak, 1997; Milette et al., 1999). Some
intervertebral discs may have an internal derangement
without obvious change in the contour of the disc.
These changes can nevertheless be an important cause
of clinical signs in humans and can only be diagnosed by
using either MRI or discography (Milette et al., 1999)
(see Chapter 4).
LIGAMENTS
1.31 Hansen type II disc herniation. This occurs mainly

following fibroid metamorphosis, in the non-chondrodystrophoid
breeds. The anulus fibrosus is damaged and there is
displacement of the intervertebral disc into the vertebral canal.
classical like the Beagle, to less obvious like the Basset

hound. An additional confounding factor is the high
incidence of small- and medium-sized disc lesions in
clinically normal animals. There is also a lack of a standardized nomenclature with which to describe disc
abnormalities and this is a serious hindrance to proper
communication and particularly to the accurate use of
MRI (see Chapters 4 and 10) (Brant-Zawadzki et al.,
1995; Milette et al., 1999). The term herniation is
poorly defined and is therefore misused frequently. It
is probably best reserved as a general term denoting a
non-specific type of disc abnormality and should not be
used to indicate clinical significance (Brant-Zawadzki
The nuchal ligament extends from the dorsal arch of the

axis to the spinous processes of the cranial thoracic vertebrae (11.47, 11.48). This ligament lies deep in the
dorsal cervical musculature. It is a large structure but
can be sectioned at surgery if necessary without impairing head and neck support. The supraspinous ligament
(1.32) continues along the tips of the spinal processes.
The lumbodorsal fascia blends with this structure in the
thoracolumbar region. The interspinous ligament (1.32)
is a fascial sheet that is found between the spinous
processes and is continuous with the lumbodorsal fascia
in the lumbosacral region. The ligaments inside and outside the vertebral canal have a significant role in spinal
stability and mobility (1.321.35). The ligamentum
flavum (10.27) is found in the roof of the vertebral canal
and in the space between adjacent vertebral laminae.
It is continuous with the joint capsules of the articular
processes, and may be significantly thickened in some
disease conditions, particularly lumbosacral disease
and cervical spondylomyelopathy (CSM). The fibers of
the dorsal longitudinal ligament (1.33, 1.34) merge
with those of the anulus fibrosus of the intervertebral
disc; both carry pain fibers (Forsythe and Ghoshal,
1984). The presence of the intercapital ligament (1.34)
13
14
a
b
e
a
1.32 The cut-away portion reveals the weak ventral longitudinal

ligament (a) running along the ventral surface of the vertebral
bodies and the much more substantial dorsal longitudinal
ligament (b) on the floor of the vertebral canal (1.33, 1.34). Also
shown are the supraspinous ligament (c); interspinous ligament
(d); intertransverse ligaments (e); and joint capsule covering the
articular processes (f).
1.34 The dorsal longitudinal ligament (a) lies on the floor of the
vertebral canal. The ligament is compact and narrow over the
vertebral bodies (b). It diverges and is thus thinner over the
intervertebral disc (c). Between the heads of the ribs (except T1,
T12 and T13) there is a reinforcing intercapital ligament (d),
which lies under the dorsal longitudinal ligament.
1.33 Dorsal view of the floor of the vertebral canal with the
spinal cord removed to show a type I disc extrusion. The
calcified nuclear material lies at the level of the intervertebral
foramen, to one side of the dorsal longitudinal ligament. Note
how this ligament fans out over, and then merges with, the
dorsal surface of the anulus fibrosus.
contributes to the low incidence of intervertebral disc

extrusions in the thoracic spine between T2 and T11
(Wilkens et al., 1996).
BLOOD SUPPLY
Vertebral column
The arterial supply to the vertebral column is segmental, with a spinal branch entering the vertebral canal via
the intervertebral foramen, closely associated with the
spinal nerve. The origin of the branches varies between
the regions of the spine (1.361.39) (Forsythe and
Ghoshal, 1984).
Venous drainage is via the internal vertebral venous
plexus, which comprises two valveless veins on the floor
of the vertebral canal (often termed the venous
sinuses). The veins converge at mid-vertebral body (and
1.35 The most significant ligament between C1 and C2 is

the transverse ligament of the atlas (a), which runs between the
sides of the atlas and over the dens, holding it down on to the
floor of the atlas. Less significant are the apical ligament of
the dens (b) (dens to foramen magnum) and the alar ligaments
(c) (dens to occipital bones). The dorsal atlantoaxial ligament
runs between the dorsal arch of C1 and the spinous process of
C2 (1.14). There are two ventral synovial articulations between
C1 and C2.
Functional anatomy
e
k
l
m
j
d
c
f
b
a
1.36 Blood supply to the cervical spine. The arterial supply to the cervical vertebrae is from the paired vertebral arteries (a), which
run cranially from their origin on the subclavian arteries in the thorax. The arteries run through the transverse foramina (b) in the
transverse processes of the vertebrae (except C7). At each segment, there are dorsal (c) and ventral (d) muscular branches.
A significant vessel runs near the caudal edge of the articular processes (e). A spinal branch (f) enters the vertebral canal at each
intervertebral foramen; these supply the spinal cord. The vertebral artery branches at the atlas. The dorsal branch (g) runs over the
transverse process of C1, anastomoses with a branch of the occipital artery (h), and enters the vertebral canal through the lateral
vertebral foramen of C1 (i). The ventral branch runs under the transverse process and also anastomoses with a branch of the
occipital artery ( j). The internal vertebral venous plexus (venous sinus) (k) lies on the floor of the vertebral canal. The veins converge
at midvertebral body (and sometimes join) (l), and then diverge again over the intervertebral disc (m). In the atlas and axis, the veins
of the internal vertebral venous plexus are much more laterally positioned. Thus, in attempting to collect CSF from the
cerebellomedullary cistern, the veins may be perforated if the needle strays from the midline.
1.37 The thoracic spine is supplied by
spinal branches (a) from the intercostal
arteries (b), which enter the vertebral
canal via the intervertebral foramina.
The internal vertebral venous plexus
drains into the major veins of the dorsal
thorax (c), mainly the azygous vein.
a
15
16
d
a
1.38 The lumbar spine is supplied by spinal

branches (a) of the lumbar arteries (b), which
arise from the aorta (c). Each lumbar artery
also gives rise to a nutrient vessel that enters
the vertebral body (Evans, 1993). A dorsal
branch runs caudally behind the articular
processes in the musculature (d). The lumbar
internal vertebral venous plexus drains into
major veins of the abdomen (e), mainly the
azygous vein and the caudal vena cava.
g
e
1.39 The blood supply to the spinal cord arises from the spinal arteries (a), which enter the vertebral canal through the intervertebral
foramina. These branch into dorsal (b) and ventral (c) radicular arteries. These arteries supply an anastomotic network on the surface
of the spinal cord, deep to the dura mater. There are paired dorsolateral spinal arteries (d), which run on the dorsal spinal cord.
These vessels may be tortuous and not recognizable as a distinct entity. There is a ventral spinal artery (e), which runs in the ventral
fissure. There are multiple anastomotic arteries connecting the main vessels. The segmental arteries are inconsistently present, such
that several segments may be supplied by one spinal artery. The distribution is also not symmetrical. The spinal cord substance is
supplied by various arteries that penetrate the surface. The vertical arteries (f) arise from the ventral spinal artery and pass dorsally
through the ventral fissure. They supply most of the gray matter and some white matter. Radial arteries (g) pass centrally from the
arteries on the cord surface, and enter the spinal cord substance. They supply the white matter and the peripheral gray matter.
The areas of spinal cord supplied by each branch have been described (de Lahunta and Alexander, 1976). The venous drainage of
the cord is also in a radial pattern, to a network of surface veins (h). These drain into the internal vertebral venous plexus on the floor
of the vertebral canal (i). These are large, valve-free vessels that have occasional anastomoses in the midline. The plexus drains at
the intervertebral foramina through the intervertebral veins. There are also veins draining the vertebral bodies into the plexus, and
drainage through the vertebral bodies (1.19).
Functional anatomy
sometimes join) and diverge over the intervertebral disc

(7.43). They are thin walled and easily damaged. The
venous plexus drains at the intervertebral foramina via
the intervertebral veins into the vertebral veins. The
intervertebral veins may be single at each foramen, or
may be paired, in which case they surround the spinal
nerve. The intervertebral veins are fragile and can bleed
profusely if damaged.
REFERENCES
Brant-Zawadzki, M.N., Jensen, M.C., Obuchowski, N., Ross, J.S., Modic,
M.T. (1995) Interobserver and intraobserver variability in interpretation
of lumbar disc abnormalities. A comparison of two nomenclatures. Spine
20, 12571263; discussion 1264.
Bray, J.P., Burbidge, H.M. (1998) The canine intervertebral disk. Part Two:
degenerative changesnon-chondrodystrophoid versus chondrodystrophoid disks. Journal of the American Animal Hospital Association 34,
135144.
Cole, T.C., Ghosh, P., Taylor, T.K.F. (1986) Variations of the proteoglycans
of the canine intervertebral disc with ageing. Biochimica et Biophysica
Acta 880, 209219.
Cook, J.R., Oliver, J.E. (1981) Atlantoaxial luxation in the dog. Compendium
on Continuing Education for the Practicing Veterinarian 3, 242250.
Cudia, S.P., Duval, J.M. (1997) Thoracolumbar intervertebral disk disease
in large, nonchondrodystrophic dogs: a retrospective study. Journal of the
American Animal Hospital Association 33, 456460.
de Lahunta, A., Alexander, J.W. (1976) Ischemic myelopathy secondary
to presumed fibrocartilaginous embolism in nine dogs. Journal of the
Americal Animal Hospital Association 12, 3748.
Evans, H.E. (1993) Millers Anatomy of the Dog, 3rd edn. Philadelphia:
WB Saunders.
Fletcher, T.F. (1970) Lumbosacral plexus and pelvic limb myotomes of the
dog. American Journal of Veterinary Research 31, 3541.
Forsythe, W.B., Ghoshal, N.G. (1984) Innervation of the canine thoracolumbar vertebral column. The Anatomical Record 208, 5763.
Garcia, J.N.P., Milthorpe, B.K., Russell, D., Johnson, K.A. (1994)
Biomechanical study of canine spinal fracture fixation using pins or bone
screws with polymethylmethacrylate. Veterinary Surgery 23, 322329.
Gorman, W.F., Hodak, J.A. (1997) Herniated intervertebral disc without
pain. Journal of the Oklahoma State Medical Association 90, 185190.
Gysling, C. (1985) Ageing process of intervertebral disks in the German
Shepherd dog [Abstract of Dissertation, Zurich Univ.]. Schweizer Archiv
fur Tierheilkunde 127, 5354.
Hansen, H.J. (1952) A pathologicanatomical study on disc degeneration
in dogs. Acta Orthopaedica Scandinavia Suppl. 11, 1117.
Hare, W.C.D. (1961) Radiographic anatomy of the cervical region of

the canine vertebral column. Part I: Fully developed vertebrae II:
Developing vertebrae. Journal of the American Veterinary Medical
Association 139, 209220.
Hickey, D.S., Aspden, R.M., Hukins, D.W., Jenkins, J.P., Isherwood, I.
(1986) Analysis of magnetic resonance images from normal and degenerate lumbar intervertebral discs. Spine 11, 702708.
Jenkins, J.P., Hickey, D.S., Zhu, X.P., Machin, M., Isherwood, I. (1985)
MR imaging of the intervertebral disc: a quantitative study. British
Journal of Radiology 58, 705709.
Milette, P.C., Fontaine, S., Lepanto, L., Cardinal, E., Breton, G. (1999)
Differentiating lumbar disc protrusions, disc bulges, and discs with normal contour but abnormal signal intensity. Magnetic resonance imaging
with discographic correlations. Spine 24, 4453.
Olby, N.J., Harris, T., Munana, K.R., Skeen, T.M., Sharp, N.J.H. (2003)
Long-term functional outcome of dogs with severe injuries of the thoracolumbar spinal cord: 87 cases (19962001). Journal of the American
Veterinary Medical Association 222, 762769.
Parker, A.J. (1973) Distribution of spinal branches of the thoracolumbar
segmental arteries in dogs. American Journal of Veterinary Research 34,
13511353.
Penderis, J., Sullivan, M., Schwarz, T., Griffiths, I.R. (1999) Subdural
injection of contrast medium as a complication of myelography. Journal
of Small Animal Practice 40, 173176.
Taylor, T.K., Melrose, J., Burkhardt, D., Ghosh, P., Claes, L.E., Kettler, A.,
Wilke, H.J. (2000) Spinal biomechanics and aging are major determinants of the proteoglycan metabolism of intervertebral disc cells. Spine
25, 30143020.
Wilkens, B.E., Selcer, R., Adams, W.H., Thomas, W.B. (1996) T9T10
intervertebral disc herniation in three dogs. Veterinary and Comparative
Orthopaedics and Traumatology 9, 177178.
FURTHER READING
Boyd, J.S., Patterson, C. (1991) A Colour Atlas of Clinical Anatomy of the
Dog and Cat. London: Wolfe Publishing.
Caulkins, S.E., Purinton, P.T., Oliver, J.E. (1989) Arterial supply to the
spinal cord of dogs and cats. American Journal of Veterinary Research 50,
425430.
de Lahunta, A. (1983) Veterinary Neuroanatomy and Clinical Neurology.
Philadelphia: WB Saunders.
Jenkins, T.W. (1978) Functional Mammalian Neuroanatomy. Philadelphia:
Lea & Febiger.
King, A.S. (1987) Physiological and Clinical Anatomy of the Domestic
Mammals, Vol. 1, Central nervous system. Oxford: Oxford University
Press.
Worthman, R.P. (1956) The longitudinal vertebral venous sinuses of the
dog. American Journal of Veterinary Research 17, 341363.
17
Patient examination
Approach to the patient

History
Chapter
age should be considered, but again this information

must be used with care.
19
19
HISTORY
Physical examination
19
Neurological examination 20
Stage 1: Patient in upright position 20
Stage 2: Patient in lateral recumbency 25
Localization of lesions 28
Assessment of the brachial and lumbosacral
plexus 30
Assessing the severity of the lesion
Determining the etiology
References
31
32
Taking a history and performing a full clinical examination are prerequisites to the neurological examination.
The history often leads to a provisional diagnosis. Of
particular note is evidence of trauma; whether the condition is progressive, static, or episodic; previous episodes
of disease; signs of pain; vaccination status; travel history;
and urinary function. In particular, does the animal let
the owner know that it needs to urinate, can it void a
stream of urine or does it just dribble urine without
being aware of it? Owners sometimes say that their dog
is incontinent because every time they pick it up it urinates, when in fact this just occurs because of pressure
being placed on its abdomen.
32
Further reading
33
The clinical syndromes seen in animals with spinal disease are recognized generally from either the history or
the physical findings. Spinal disease should also be considered in animals with non-specific pain, exercise intolerance or lameness not caused by orthopedic disease
(McDonnell et al., 2001). This chapter discusses the
approach to a patient in which spinal disease is suspected.
PHYSICAL EXAMINATION
A general physical examination must be made in all
patients. If there has been trauma or if anesthesia is contemplated, involvement of other body systems must be
determined (Neer, 1992) (2.1). Also, some patients in
which spinal disease is suspected have disorders of other
APPROACH TO THE PATIENT

The aims of patient examination are as follows:
Determine whether the problem is spinal in origin.
Locate the site of the disorder.
Assess the severity of the neurological deficit.
Identify the disease process.
Determine the most appropriate form of treatment.
Predict the prognosis.
Knowledge of the breed incidence of spinal diseases
is useful in the initial assessment, but it is a mistake to use
such information as the only basis for diagnosis. Similarly,
2.1 Damage to thoracic or abdominal viscera is common

after trauma. Here bile duct rupture has given rise to extensive
peritoneal contamination (see page 282).
20
systems. It is not at all unusual for orthopedic disorders to

mimic spinal conditions; examples are given in Table 2.1.
Careful clinical examination should identify such
problems, and particular note should be made of joint
pain or enlargement, as these signs are present in many
dogs misdiagnosed as having neurological disorders.
The presence of any spinal pain or deformity should
also be noted. The quality of the femoral pulse must
be determined, particularly in acutely paralysed cats
although thromboembolism also occurs in dogs (Boswood
et al., 2000) (see Chapter 14).
It is straightforward to perform a screening neurological examination as part of a general physical examination (Table 2.2). If abnormalities are detected, carry
out a more complete neurological examination.
NEUROLOGICAL EXAMINATION
The neurological examination is carried out with the aims
of determining the precise location of the spinal lesion
and its severity. The neurological examination described
here is performed readily with the animal upright in the
first instance, and later placed in lateral recumbency.
It is useful to have a form on which to write the findings of the neurological examination (2.2). This insures
that no aspect of the examination is missed, provides a
permanent record, prevents errors and permits more
accurate comparisons of any serial examinations. A video
can also be made of the patient for this purpose.
Stage 1: Patient in upright position

ASSESS ATTITUDE, POSTURE AND GAIT
Watch the patient as it relaxes in the examination room.
Let it move to the best of its ability, unless it has an
acute spinal injury, when movement should be restricted.
Note the degree of motor function, the gait (particularly
noting any asymmetry) and the general demeanor. In
cats, this part of the examination is particularly important as later parts may be difficult to perform. It is useful to listen to dogs as they walk on a hard surface; if
proprioceptive deficits are present, the examiner may
hear the claws scuff.
DETERMINE THE LOCOMOTOR STATUS

The animal is encouraged to move, except where an
acute spinal injury has occurred or if there is severe
Table 2.1 Disorders that can mimic spinal disease

Systemic
disorders
Uni- or Bilateral
orthopedic disorders
Generalized orthopedic
disorders
Neuromuscular
disorders
Endocarditis
Cardiac insufficiency
Hypertension
Upper airway disease
Hyperkalemia
Hypokalemia
Hypocalcemia
Hypoglycemia
Hyperthyroidism
Addisons disease
Pheochromocytoma
Osteochondritis dissecans
Cranial cruciate ligament injury
Tibial crest avulsion
Fractures
Coxofemoral osteoarthritis
Patellar luxation
Septic arthritis
Biceps tendonitis
Infraspinatus contracture
Gracilis contracture
Achilles tendon rupture
Psoas muscle injury
Hypertrophic osteodystrophy
Polyarthritis
Panosteitis
Generalized myopathies
Ischemic neuromyopathy
Neuropathies
Radiculopathies
Junctionopathies
Table 2.2 Screening neurological examination
Observation
Mental status
Posture
Gait
Postural
reactions
Cranial
nerves
Spinal
reflexes
Spinal
hyperesthesia
Bladder
function
Paw position
Hopping
Menace
Pupillary light reflex
Oculovestibular
response
Jaw tone; temporal
muscle mass
Facial sensation
Palpebral reflex
Patellar
Withdrawal
Perineal
Cutaneous trunci
reflexes
Cervical
Thoracolumbar
Lumbosacral
Historical
Leakage?
Voiding?
Patient examination
SPINAL REFLEXES
HISTORY
PHYSICAL EXAMINATION
OBSERVATION
Reflex
(Nerve)
(Spinal cord segments)
LEFT
RIGHT
Triceps
(Radial)
(C7-T1)
Biceps
(Musculocutaneous)
(C6-C8)
Withdrawal (thoracic limb)
(Multiple)
(C6-T2)
Patellar
(Femoral)
(L4-L6)
Gastrocnemius
(Tibial, sciatic)
(L6-S1)
Withdrawal (pelvic limb)
(Sciatic)
(L6-S1)
Mental status (e.g. alert, depressed, stupor, coma)

Posture (e.g. normal, paraparesis, hemiparesis, head tilt, tremor)
Gait (e.g. ataxia, circling)
POSTURAL REACTIONS
LEFT
RIGHT
Hopping
Front
Rear
Paw position
Front
Rear
Reflex step
Front
Rear
Tactile placing
Front
Rear
Visual placing
Front
Rear
Perineal
(S1-S2)
URINARY FUNCTION
Hemistanding
Voluntary urination?
Hemiwalking
Wheelbarrowing
Bladder distention?
Extensor postural thrust

Overflow / ease of manual expression
CRANIAL NERVES
SPINAL HYPERAESTHESIA
Test
(Innervation)
LEFT
RIGHT
Menace response
(II & VII)
Vision
(II)
SML
Pupil size
Pupillary light reflex
(II & III)
Stimulate left eye
Stimulate right eye
Strabismus
(III, IV & VI)
Spontaneous nystagmus
(III, IV, VI & VIII)
Positional nystagmus
Oculovestibular response
Facial sensation
(V)
Jaw tone
(V)
Temporal muscle mass
(V)
Corneal reflex
(V, VI & VII)
Facial symmetry
(VIII)
Palpebral reflex
(V & VII)
Hearing (hand clap)
(VIII)
Swallowing or gag reflex
(IX & X)
Tongue
(XII)
EYES
PANNICULUS REFLEX
Level of cut-off (dermatome)
LEFT
SML
RIGHT
DEEP PAIN PERCEPTION

Thoracic limb
Pelvic limb
Tail
LESION LOCALIZATION
BRAIN
Side
Forebrain
Brainstem
Cerebellum
Vestibular - peripheral
Vestibular - central
Multifocal
SPINAL CORD
C1 - C5
C6 - T2
T3 - L3
L4 - S3
MULTIFOCAL CNS
PERIPHERAL NERVE
Local
Generalised
Horners Syndrome
(Sympathetic)
NEUROMUSCULAR
Fundic examination
MUSCULAR
MUSCLE PALPATION
NORMAL
Tone
Atrophy
KEY: ABSENT 0; REDUCED 1; NORMAL 2; INCREASED 3; CLONUS 4
2.2 Form for neurological examination. The DAMNIT scheme can be listed on the reverse side with room to add differential
diagnoses and the diagnostic plan(s).
21
22
A
2.4 Proprioception: paw position response where an animals
body-weight is supported fully and then each paw is turned
over individually to bring the dorsal surface into contact with the
ground. Normal animals return the paw to an upright position
almost immediately; those with neurological disease cranial to
the limb may leave the paw flexed. If proprioception is normal,
spinal cord disease is unlikely to be present.
B
2.3 A: Hopping in the thoracic limb. The patients body is
supported with only one limb bearing weight on the ground,
and then it is moved laterally. The animal will not be able to hop
normally if it has a proprioceptive deficit or impaired motor
function. In a large dog this can be done by just lifting one
thoracic limb and then moving the dog sideways. B: Hopping
in the pelvic limb. Large dogs can be hopped as described for
the thoracic limb.
pain. Patients that appear paraplegic at rest may show

some voluntary movement if supported by a sling
(15.9A) or if held by the base of their tails. Unilateral
weaknesses and sensory deficits may be revealed by
hopping (2.3), hemistanding and hemiwalking tests. As
hopping also requires good motor function and intact
proprioception, it is included in the section on proprioceptive testing below.
Assess muscle strength, if the patient is able to stand,
by pressing down on the shoulders and hips. Following
these tests, the locomotor status can then be classified,
for example paraparetic, hemiparetic and tetraparetic.
ASSESS PROPRIOCEPTION
This is evaluated in the standing animal by the paw
position test and the reflex step (2.4, 2.6). Animals
with deficits of proprioception that can still walk often
wear the dorsum of their digit(s) or claw(s).
2.5 Boxer dog with proprioceptive loss in the left pelvic limb.
Severe deficits such as this are evident even if the animals
weight is not supported properly.
Tactile placing can be tested as in 2.10 but with

the animals eyes covered or, in small patients, with
the animals body curved around the examiners
abdomen so that its head is pointing away from the
table. The thoracic and pelvic limbs on the side nearest
to the table are then touched gently against the table
edge. The animal should immediately lift up the
foot; the contralateral limbs are then repositioned so
that they are adjacent to the table and the process is
repeated.
Patient examination
2.6 A, B: Proprioception can also be

tested by the reflex step, where a piece
of paper is placed under the foot and
pulled laterally. The animal should return
the foot briskly to a normal position; an
abnormal response is to let the foot slide
away from the body. This test probably
assesses proprioception more in the
proximal part of the limb than the test
shown in 2.4.
A
2.7 Wheelbarrowing test. The pelvic limbs are lifted off the
ground as shown and the animal made to walk forward.
This test can reveal thoracic limb paresis and exacerbates
asymmetrical lesions. Elevating the head on this test will
sometimes unmask hypermetria.
2.9 Extensor postural thrust. The patient is held up, as shown,

and lowered to the surface. The normal animal will push away
with the pelvic limbs and step backwards. This test is useful in
revealing pelvic limb deficits.
2.8 Wheelbarrow testing in large dogs is particularly useful to

identify subtle weakness in the thoracic limbs that might indicate
a cervical lesion in a dog that at first sight appears to be only
weak in its rear limbs.
PALPATE THE ABDOMEN

Determine the degree of bladder filling, and the ease
with which urine is expressed by palpating the
abdomen. Urinary incontinence is often a feature of
spinal disorders, and some assessment of urinary function should have been gained from the history.
2.10 Placing test. Here visual placing is being evaluated as the

animal can see the table edge. Visual placing evaluates vision
and motor function. Tactile placing evaluates sensation,
proprioception and motor function.
23
24
If neurogenic urinary incontinence is present, it is

important to determine if it is LMN or UMN in nature
(Table 15.6):
The LMN bladder is typically large, flaccid, and
easily expressed. This is usually associated with
lesions of the sacral spinal cord segments or nerve
roots. Some animals, particularly cats with
sacrocaudal injuries, can nevertheless be difficult to
express due to high sphincter tone, probably from
unopposed sympathetic tone via the uninjured
hypogastric nerve (see page 351).
The UMN bladder is characteristically tense
and difficult to express unless grossly distended,
because urethral sphincter tone is often increased.
This is seen with lesions cranial to the sacral
segments, usually those affecting the T3L3
spinal cord region (see Disorders of micturition,
page 351).
The importance of determining the type of incontinence is two-fold. Appropriate drug therapy can be determined from this information (see Pharmacological
manipulation of micturition, page 351). Also, the prognosis for recovery of urinary function is often worse for
LMN incontinence compared to most UMN lesions.
intact on the contralateral side. Thoracolumbar lesions

may interfere with the afferent part of the cutaneous
trunci reflex, which will be intact only in dermatomes
supplied by segments cranial to a spinal cord lesion and
will be absent in the segments caudal to it. Lesions caudal to L1 spinal cord segment usually have a normal
cutaneous trunci response as the L1 dermatome extends
to the level of the tuber coxae (2.12).
PALPATE THE SPINE

Determine the presence of spinal hyperesthesia by
palpating the vertebral column and evaluating the
patients response (2.13, 2.14). This is an important
step in the examination. The degree of pressure to be
T13
L1
T13
T12
T11
THE CUTANEOUS TRUNCI REFLEXES

This is tested by pinching the skin along the dorsal surface of the trunk with fine forceps and observing the
twitch of the cutaneous trunci muscle on both the ipsilateral and, to a lesser extent, the contralateral side
(2.11, 2.12). There is crossing of the pathway within
the spinal cord, leading to a bilateral response after unilateral stimulation. In thoracic limb paresis, the cutaneous trunci reflex will also be absent on the affected
side if the lesion involves the C8 and T1 spinal cord
segments, roots or nerves (1.4 and page 30), but will be
2.11 The cutaneous trunci reflex is activated by pinching the

skin over the lumbar spine with forceps or by a gentle needle
prick, which leads to a twitch of the cutaneous trunci muscle.
2.12 The cutaneous trunci reflex. Stimulation of the skin of

the back activates the reflex, the efferent arm of which is the
lateral thoracic nerve arising from segments C8 and T1
and leaving the caudal part of the brachial plexus (a). The
dermatomes are positioned some distance caudal to their
respective vertebrae as illustrated here.
2.13 Examining the neck for evidence of pain. Palpating the

neck muscles and moving the neck is usually adequate to reveal
pain, as evidenced by an increase in tension (guarding) and by
muscle fasciculations.
Patient examination
applied when determining the presence of pain varies

between patients. In a dog with neck pain caused by a
cervical disc extrusion, it is often adequate to palpate
the cervical muscles gently. It is not usually necessary
to put the head in extreme positions to reach this conclusion. Some animals do not show pain but guard
against neck movement with an increase in resistance
compared to normal. In some more stoic dogs, greater
force may be required either by downward pressure on
the spinous processes or along the transverse processes.
The transverse processes of C6 (4.6) can also be
manipulated percutaneously in most dogs to assess for
low cervical pain.
Spinal pain can arise as a result of discogenic pain
(Morgan et al., 1993); dorsal longitudinal ligament
damage; nerve root irritation; stretching or inflammation
of the meninges; and bone pain. In humans, back pain
can also arise from the sacroiliac and facet joints; these
are likely to be potential causes of pain in animals as
well but would be much harder to confirm than in
humans (Pang et al., 1998) (see Chapter 14).
In some animals with spinal disease, pain is the only
clinical abnormality, neurological function being normal.
CRANIAL NERVE EXAMINATION

Even though abnormal cranial nerve findings may not be
expected in spinal disorders, some animals with multifocal neurological disease present predominantly with
signs of spinal dysfunction (14.15). It is therefore important to assess the entire nervous system (de Lahunta,
2001; Braund and Sharp, 2003). In addition, particular
note should be made of the presence of any component
of Horners syndrome (ptosis, miosis, enophthalmos and
third eyelid protrusion), which occurs with interference
to the sympathetic supply to the eye. Note that some
animals will only show miosis without the other components. The resultant anisocoria is missed easily unless
tested specifically (2.15). The pupils should be assessed
carefully for anisocoria by looking through an ophthalmoscope held about 60 cm from the animals face. This will
produce a tapetal reflection that allows easy comparison
of pupil sizes even in a well-lit room; it is particularly
important in animals with a dark iris. Isolated miosis may
be a feature of spinal disease if the cervical or cranial thoracic spinal cord segments, or the nerve roots of the
brachial plexus, are involved. An ophthalmoscopic examination should also be carried out to look for fundic
lesions indicative of inflammatory CNS disease.
Stage 2: Patient in lateral recumbency
2.14 Palpating the thoracolumbar spine for pain. For

assessment of lumbosacral hyperesthesia see page 184.
The patient is then placed in lateral recumbency and

each limb evaluated, with the aim of placing it into one
of the following categories:
Normal.
LMN-type abnormality.
UMN-type abnormality.
An understanding of functional anatomy is required
to appreciate the difference between these types of
deficit (1.12, 1.13).
The effect of lesions on the LMN and UMN systems
can be considered in terms of motor function, muscle
atrophy, muscle tone and local reflexes. The clinical
2.15 A: Isolated miosis due to loss of

sympathetic outflow after brachial plexus
injury. B: Iatrogenic Horners syndrome
in a cat that has just had inflammatory
polyps removed from the middle ear.
The ptosis and prolapsed third eyelid are
visible although the pupil is somewhat
obscured. Tumors of the brachial plexus
may demonstrate several components of
Horners syndrome but often only show
miosis in the early stages.
25
26
Table 2.3 Differentiation of LMN vs UMN abnormalities

Lower motor neuron (LMN)
Upper motor neuron (UMN)
Motor function
Paresis or paralysis
Paresis or paralysis
Reflexes
Absent or reduced
Normal or increased
Muscle tone
Reduced
Normal or increased
Muscle atrophy
Severe, earlyneurogenic
Late, milddisuse
2.16 Neurogenic atrophy. It is useful to evaluate muscles

that have a definitive bony border, for example the spinatus
muscles as shown, the cranial tibial muscle and the muscles
over the pelvis.
2.17 Paraplegic dog with stiff, hyperextended thoracic limbs

and digits; sometimes the neck is also extendedthe
SchiffSherrington sign. Thoracic spinal cord lesions may
interfere with inhibitory neurons that have their cell bodies (the
border cells) in the L1L7 spinal cord segments (Braund and
Sharp, 2003). Axons from these cells pass cranially to inhibit
neurons supplying the thoracic limb extensor muscles.
MUSCLE MASS
signs that allow differentiation between UMN and
LMN abnormalities are summarized in Table 2.3.
LMN deficits are characterized by paresis or paralysis; severe (neurogenic) muscle atrophy (2.16);
reduced tone in the affected muscles; and depressed or
absent reflexes. UMN deficits show paresis or paralysis; mild (disuse) muscle atrophy; normal or increased
muscle tone; and normal or hyperactive reflexes. There
is some variation in mild cases, but from the neurological examination, it should be possible to categorize each
limb as being normal, UMN-type abnormality, or
LMN-type abnormality.
Muscle atrophy is assessed by observing and palpating

major muscle groups (2.16).
MUSCLE TONE
Test muscle tone by gently flexing and extending
the joints. Normally there is some resistance to such
manipulation. An incorrect impression of increased
tone may be gained in excitable or fractious animals,
or if the animal has a painful orthopedic condition.
Increased tone in the thoracic limbs is seen in the
SchiffSherrington sign (2.17). This sign (see Ascending
motor tract, page 6) does not necessarily denote a
hopeless prognosis (13.7).
MOTOR FUNCTION
REFLEX TESTING
This has been evaluated previously. With the animal in

lateral recumbency, the examination proceeds as follows.
There are a number of local reflexes available for examination, but it is usual to concentrate on the patellar,
Patient examination
2.20 In the pelvic limb critical assessment should be made of

hock flexion to assess for evidence of LMN weakness (2.27). It
may also be necessary to stimulate the withdrawal reflex with
forceps placed over the nail bed or digit in order to evoke a
behaviural response for deep pain evaluation (2.21).
2.18 The flexor or withdrawal reflex is stimulated by pinching
the toe, which results in flexion of the limb. It is important to
persist with the stimulus until it is clear that all the limb joints are
flexing. Loss of elbow flexion is often the most sensitive
indicator of a weak thoracic limb withdrawal reflex. While the
flexor reflex is being evoked, the contralateral limb should also
be observed for reflex extensionthe crossed extensor reflex.
2.19 The patellar reflex is evoked by tapping the straight

patellar ligament. It can be done on either the upper or the lower
limb, in contrast to the withdrawal reflex that is unreliable in the
lower limb.
flexor (withdrawal) and perineal reflexes (2.182.20).

The perineal reflex can be elicited by squeezing the base
of penis or the perineal region. Anal tone can be assessed
by a rectal examination or using a thermometer.
Other reflexes such as the triceps, biceps, and gastrocnemius may also be tested. However, they are
found inconsistently in normal animals and their main
significance is in finding hyperactive responses in UMN
disorders. Increased patellar reflexes are often seen in
UMN lesions cranial to the L4L6 spinal cord segments.
When this occurs there are sustained contractions known
as clonus, where the limb vibrates briefly after the first

initial kick. It must be remembered, however, that
many animals with UMN lesions have a normal patellar
reflex. A different type of exaggerated patellar reflex
can be seen in animals with LMN lesions involving the
sciatic nerve or its origins. The hamstring muscle group
innervated by the sciatic nerve normally counteracts or
damps the reflex kick of the patellar reflex. Interference
with sciatic nerve function can reduce this damping
effect and lead to an increased reflex, so that the lower
limb will oscillate after an initial brisk kick. This is
termed pseudohyperreflexia or an oscillating patellar
reflex. In tense patients, the patellar reflex may not
respond at all in the upper limb whereas it does in the
lower limb; it is then important to test the reflexes
with the dog on its opposite side. In most animals the
reflex can also be elicited in the standing position if the
limb is relaxed.
DEEP PAIN SENSATION OR NOCICEPTION

On testing the withdrawal reflex, note the patients
behaviural response to the stimulus (2.21). Formerly
called deep pain sensation, this is more correctly
referred to as nociception although the terms are used
interchangeably in this text (de Lahunta, 2001). A turn
of the head or vocalization should be seen in response to
the pinch, indicating that the painful stimulus has been
transmitted up the spinal cord to the brain. If there is no
such response then the stimulus must be increased, usually by using larger instruments such as needle holders,
pliers, or even with an electrical stimulator (13.36B)
across the digit or nail bed. Confirmation of a loss of
deep pain is often provided by the animal showing a line
of analgesia over its flank. The animal responds to
27
28
a
b
SITE OF INJURY
2.21 Testing nociception or deep pain sensation using pliers

(Scott, 1997; Scott and McKee, 1999). The reflex withdrawal of
the limb must not be mistaken for a behaviural response.
Similarly, some animals will react to the change in body position
induced by the withdrawal reflex. Ideally, the withdrawal reflex
should be initiated first, and then further pressure applied to
evaluate nociception (6.2, 13.36B).
pinching skin of dermatomes arising cranial to a lesion

but shows no response in those arising caudally.
Animals with cervical injuries very rarely lose deep
pain. This is because animals that suffer spinal cord
injuries severe enough to interfere with nociception
usually are left with no motor function, which results in
rapid asphyxiation. This emphasizes the importance of
assessing for hypoventilation in any severely tetraparetic
or tetraplegic animal (6.1, 7.11). One example where
there may be a unilateral decrease in deep pain after
cervical injury is following catastrophic failure of a disc
after trauma. Here the nucleus pulposus explodes dorsolaterally to cause devastating, asymmetrical neurological deficits (Griffiths, 1970).
Absence of deep pain sensation therefore indicates
severe spinal cord damage (see Assessing the severity
of the lesion, page 31; Spinal cord nerve fibers and the
effect of compression, page 6; Mechanisms of recovery after spinal cord injury, page 87).
LOCALIZATION OF LESIONS
On the basis of the neurological examination, it is usually possible to identify the location of the spinal cord
lesion (2.22). Functionally, the spinal cord may be
divided into four regions:
A: C1C5.
B: C6T2 (cervical intumescence).
C: T3L3.
D: L4S3 (lumbar intumescence).
Areas A and C, the cervical and thoracolumbar spinal
cord, convey primarily the UMNs. Areas B and D, the
cervicothoracic and lumbosacral spinal cord, provide
THORACIC LIMB
DEFICIT
PELVIC LIMB
DEFICIT
UMN
UMN
LMN
UMN
NORMAL
UMN
NORMAL
LMN
C1C5
C6T2
T3L3
L4S3
2.22 Lesions in specific regions will produce different

combinations of neurological signs. Lesions in the cervical spinal
cord (a) produce UMN signs in all four limbs. Lesions in the
segments of the cervical intumescence (b) produce LMN deficits
in the thoracic limbs and UMN signs in the pelvic limbs. Lesions
in the thoracolumbar cord (c) produce UMN signs in the pelvic
limbs only, with normal thoracic limbs (the SchiffSherrington
sign may be present in peracute lesions). Lesions in the lumbar
intumescence (d) produce LMN signs in the pelvic limbs, tail
and perineum; the thoracic limbs are normal.
innervation to the thoracic and pelvic limbs respectively. Area B also conveys UMNs to the pelvic limbs.
It can be seen in 1.12 that part of the LMN lies
within the spinal cord; lesions in areas B and D of the
spinal cord will therefore produce LMN signs in the
limbs. Variations are possible, for example in some animals with myelopathy affecting the caudal cervical
segments, the thoracic limbs often show a mixture
of UMN and LMN signs. They may show increased
thoracic limb muscle tone, which is an UMN effect on
the elbow and carpal extensor muscles (Seim and
Withrow, 1982). In addition, there is often an associated LMN weakness of the elbow flexors resulting in a
weak withdrawal reflex. Dogs with severe C6T2 signs
also have a short-strided, choppy or disconnected
thoracic limb gait and a long-strided, ataxic pelvic limb
gait (see Chapter 11). In contrast, dogs with C1C5
signs often show a long-strided or floating thoracic
limb gait together with a long-strided, ataxic pelvic limb
gait (Baum et al., 1992).
Patient examination
When considering the neurological localization,

remember that the spinal cord segments are not all
contained in the vertebra of the same number, especially in the cervical and lumbar region (1.2, 1.5).
Examples of potential pitfalls in neurological examination include:
LMN deficits in all limbs generally indicate diffuse
peripheral nerve or neuromuscular disease.
Occasionally, deficits may only be present in either
the thoracic limbs or the pelvic limbs early in the
course of generalized LMN diseases.
If there are any cranial nerve deficits, or signs
such as a change in mentation or behavior, then
it is very unlikely that the animal only has a spinal
cord lesion.
2.23 Eight-year-old Golden retriever with acute onset of left

hemiparesis and inability to stand. Neurological deficits localized
to C1C5 spinal cord segments. No lesions were detected in
the cervical spine; CT scan revealed a mass in the left medulla.
In retrospect, the only clue to a brainstem lesion was a subtle
decrease in mental status.
Rarely an animal will have no cranial nerve deficits

and will appear to localize to the C1C5 spinal
cord segments when it actually has a lesion in its
brainstem (2.23).
Animals with cerebellar disease as well as those with
spinal cord lesions affecting the spinocerebellar
tracts often show hypermetric and dysmetric limb
movements. Animals with primary cerebellar disease
(see Neuroaxonal dystrophy, page 320), can usually
be distinguished as they have additional signs such as
head tremors, dysmetric movements or menace
deficits (Holliday, 1980).
Animals with cervical lesions can sometimes show
vestibular signs. Local nerve blocks or neurectomy
of the first cervical dorsal nerve roots produces
vestibular signs in monkeys and in horses (Biemond
and De Jong, 1969; Mayhew, 1999).
Central cord syndrome can cause unusual signs.
With this condition a high-cervical lesion causes
tetraparesis with much more severe deficits in the
thoracic limbs than in the pelvic limbs; spinal
reflexes are preserved in all four limbs.
An LMN lesion may obscure a concomitant UMN
lesion (2.24).
Animals with some spinal cord lesions, such as
arachnoid cysts or intramedullary tumors, may
show urinary or fecal incontinence while still able
to walk (12.4).
Apparent paraparesis due to a mild cervical lesion
(see Chapters 7 and 11).
A lesion in the sacral spinal cord segments can
mimic nerve root compression at the lumbosacral
junction (2.25).
Animals with a choppy thoracic limb gait and
paraparesis may have a lesion between T1T6
(3.36A).
2.24 Aged Dalmatian with mild, progressive paraparesis, dilated anus and urinary incontinence. Neurological deficits localized to
the L7S2 spinal cord segments. A CT scan of the lumbosacral space was unremarkable (10.2B, 10.10A). A: Myelography
revealed masses (arrowheads) in the subarachnoid space at C1/2 on the right; caudal C2 on the left. B: Cranial C6 on the right;
mid-C7 on the right; and a number of small filling defects over the lumbar cord (not shown). The mass at C2 was a nerve sheath
tumor with subarachnoid metastasis (4.27, 4.28).
29
30
2.25 Seven-year-old Shih Tzu that presented initially with loss of tail tone and then mild paraparesis, anal arreflexia and incontinence
over the next 8 months. Localization was to the cauda equina or L4S3 spinal cord segments. T1-weighted MRIs post contrast
medium. A: sagittal and B: transverse images show that the dog has an enhancing mass (arrows), most likely a nerve sheath tumor,
involving the right L5 nerve root and invading the sacral spinal cord segments. The lumbosacral disc space appears normal.
2.26 A: Dog with a brachial plexus

avulsion injury affecting the caudal portion
of its left brachial plexus. It is unable to fix
the carpal or elbow joint but can still flex
its elbow joint due to preserved cranial
plexal function. It has also begun to
mutilate its foot. B: This dog sustained an
injury to both the cranial and caudal
portions of its right brachial plexus. It is
unable to flex the elbow or to bear weight
on the limb (note dropped elbow and
collapsed carpal joints) and is at high risk
of abrading its foot.
Assessment of the brachial and

lumbosacral plexus
Lesions affecting these areas may be missed if specific
testing is not performed. Animals must be assessed for
brachial plexus lesions after trauma (13.1) or when
they have a chronic, progressive thoracic limb lameness
(12.2). This is done by testing for anisocoria (2.15);
unilateral loss of the cutaneous trunci reflex (2.11,
2.12); reduced thoracic limb flexor response (2.18,
2.26); and reduced nociception (2.21).
Animals with lesions of the brachial plexus may
have deficits referable to the cranial brachial plexus,
the caudal brachial plexus, or both (1.4, 2.26). The
suprascapular, subscapular, musculocutaneous and axillary nerves derive mainly from the cranial portions of
the plexus (C5, C6 and C7). Deficits of the cranial

plexus cause loss of elbow flexion and may result in
subluxation of the shoulder joint. The radial, median
and ulnar nerves derive mainly from more caudal portions of the plexus (C7, C8, T1 and T2). Deficits of
the caudal plexus cause an inability to fix the elbow or
carpus and therefore the animal cannot bear weight on
the limb. In addition they often have ipsilateral (full or
partial) Horners syndrome and cutaneous trunci reflex
deficit (1.4, 2.15). Elbow flexion is preserved when
the cranial plexus is intact (2.26).
Animals must be assessed for lumbosacral plexus
lesions after trauma by testing for sciatic, pudendal or
pelvic nerve deficits. One of the most specific tests for
a sciatic nerve deficit is to assess hock flexion (2.27).
Patient examination
2.28 Sensory deficits secondary to a brachial plexus lesion

may give rise occasionally to trophic ulceration as shown here
(arrowheads). Trophic ulcers occur due to an inability to protect
the foot and also from abnormal forces placed on the foot due
to digital paralysis (Hunt and Chapman, 1991).
2.27 When assessing the withdrawal reflex in the pelvic limb, all
three joints should be assessed. An animal can flex both its hip
and stifle joints by using the femoral and sciatic nerves whereas
it can only flex its hock joint by using the peroneal division of the
sciatic nerve. Weak or absent hock flexion can be appreciated
best by pulling the hock joint gently into extension and then
applying a stimulus to the digits.
Loss of proprioception is a more sensitive indicator

although this does not differentiate an UMN lesion
from a LMN lesion.
Sensory loss in small animals usually reflects spinal
cord or nerve root lesions. Occasionally, it reflects
peripheral nerve injury and specific test sites for the various cutaneous sensory nerves have been defined (Bailey
and Kitchell, 1987). For the thoracic limb the main sites
are:
Musculocutaneous nervejust distal to the medial
epicondyle of the humerus.
Radial nervedorsal aspect of the third or fourth
digit.
Ulnar nervejust distal to the olecranon on the
caudolateral aspect of the antebrachium and on the
lateral surface of the fifth digit.
For the pelvic limb the main sites are:
Saphenous (femoral) nerve4 cm distal to the
medial condyle of the tibia.
Superficial peronealdorsal aspect of the 3rd

digit, proximal end.
Deep peronealskin between the 2nd and 3rd
digits, dorsal aspect.
Tibial nerveskin at the proximal border of the
tarsal pad.
Sensory loss following brachial or lumbosacral plexus
injury occasionally gives rise to trophic ulceration (2.28).
ASSESSING THE SEVERITY OF

THE LESION
Assessing the severity of a lesion plays a major part in
the diagnostic procedure. In certain patients, it has as
much bearing on prognosis as the etiology; if a poor
prognosis is suggested from the neurological examination further investigations may be deemed unnecessary.
In general, patients with spinal diseases that show
LMN deficits have a worse prognosis for a return to
function than those showing UMN deficits. One exception to this is for dogs with thoracolumbar disc disease
(Dhupa et al., 1999) (see Chapter 8).
In UMN injuries, rate of onset, duration, and degree
of spinal cord damage all affect the clinical signs. Rapid
progression of acute signs tends to reflect acute decompensation, which if treated early has a favorable outcome unless the damage is irreversible. Slow progression
31
32
generally reflects a chronic process that is gradually

outstripping secondary compensation mechanisms
(Gilson, 2003). For thoracic and lumbar lesions, the
degree of dysfunction can be classified as grades 15
(Scott, 1997; Scott and McKee, 1999):
1. Pain only.
2. Paraparesiswalking.
3. Paraparesisnot walking.
4. Paraplegia.
5. Paraplegia with loss of deep pain sensation.
The animal usually progresses through these stages as
it deteriorates unless the disease is peracute in onset,
such as trauma or a vascular lesion. The scheme applies
to both UMN and LMN lesions. Control of micturition
is often lost between grade 4 and grade 5 but full
control of continence may be incomplete as early as
grade 2. Recovery of spinal cord function usually progresses steadily in the reverse order, unless the animal
develops either a reflex bladder or spinal walking (6.2,
13.30, 13.34). Recovery after spinal injury is usually
defined as recovery of continence and of the ability to
walk unaided.
The severity of the neurological deficits depends on
two anatomical features: the position of the tracts in the
spinal cord that carry the respective function and the
diameter of the fibers transmitting that function (see
Spinal cord nerve fibers and the effect of compression,
page 6). The prognosis worsens with increasing neurological deficit, as in general this reflects greater spinal
cord damage. The presence of a crossed extensor reflex
(seen in UMN lesions) and the SchiffSherrington sign
indicate severe lesions, but are not in themselves
prognostic indicators. SchiffSherrington syndrome in
particular is often thought to indicate a poor prognosis
but this is largely because it is only seen after severe
spinal cord injury. However, some animals with Schiff
Sherrington syndrome make a good recovery (13.7) and
so prognosis should be inferred from the presence or
absence of nociception. The prognosis for patients without deep pain sensation (grade 5) is guarded for most
disorders, especially if it has been absent for longer than
48 h (see also page 132). Unfortunately, although it is
usually obvious how long an individual animal has been
paraplegic it is seldom clear exactly when that animal lost
deep pain. Loss of deep pain sensation after disc disease
is associated with a 5060% chance of recovery following
surgery (see Chapter 8, page 138 and Table 8.1a). Animals
that have absent deep pain sensation after trauma carry a
much worse prognosis whatever the duration of the clinical signs (Olby et al., 2003) (see Chapter 13, page 302).
In other grades of dysfunction, the prognosis is also
somewhat dependent on the etiology. For example, a
Dachshund with grade 3 signs could possibly have a
thoracolumbar disc extrusion or a spinal tumor. Clearly,

the prognosis for such disorders could differ markedly.
Spinal shock is a phenomenon that affects primarily
humans and higher primates. It is where an UMN lesion
is associated with a temporary loss of spinal reflexes
below the lesion; in humans this can last for a week or
more. If it occurs in dogs and cats it is transient and is
only seen within the first few hours after injury. It could
confound neurological localization and therefore in animals examined immediately after trauma a repeat examination should be performed a few hours later (Walmsley
and Tracey, 1983; Gopal and Jeffery, 2001).
Cervical spinal cord lesions display a similar progression of signs to that described above but urinary incontinence or loss of nociception is very rare. Severe
cervical spinal cord lesions may result in respiratory
failure caused by interference with respiratory control,
although this is uncommon (6.1, 7.11).
DETERMINING THE ETIOLOGY

Once the lesion has been located, a list of differential
diagnoses can be made. Many components go into this
assessment, including breed and age of the patient, history, presenting signs, progression, and physical and
neurological findings. It is important not to depend
entirely on one feature to make a firm diagnosis. It is
reasonable to start from the assumption that the most
common disease is the cause, but this should ideally
either be confirmed or attempts made to exclude less
likely conditions, particularly if the patient is not progressing as expected. Differential diagnosis is discussed
in Chapter 3.
REFERENCES
Bailey, C.S., Kitchell, R.L. (1987) Cutaneous sensory testing in the dog.
Journal of Veterinary Internal Medicine 1, 128135.
Baum, F., Trotter, E.J., de Lahunta, A.D. (1992) Cervical fibrotic stenosis
in a young Rottweiler. Journal of the American Veterinary Medical
Biemond, A., De Jong, J.M. (1969) On cervical nystagmus and related disorders. Brain 92, 437458.
Boswood, A., Lamb, C.R., White, R.N. (2000) Aortic and iliac thrombosis
in six dogs. Journal of Small Animal Practice 41, 109114.
Braund, K.G., Sharp, N.J.H. (2003) Neurological examination and localization. In: D. Slatter (ed.), Textbook of Small Animal Surgery, 3rd edn,
10921107. Philadelphia: Elsevier Science.
de Lahunta, A. (2001) Neurological examination. In: K.G. Braund (ed.),
Clinical Neurology in Small AnimalLocalization, Diagnosis and Treatment.
Ithaca: International Veterinary Information Service. http://www.ivis.org/
special_books/Braund/deLahunta/chapter_frm.asp?LA1
Dhupa, S., Glickman, N.W., Waters, D.J., Dhupa, S. (1999) Functional
outcome in dogs after surgical treatment of caudal lumbar intervertebral
disk herniation. Journal of the American Animal Hospital Association 35,
323331.
Gilson, S.D. (2003) Neuro-oncologic surgery. In: D. Slatter (ed.), Textbook of
Small Animal Surgery, 3rd edn, 12771286. Philadelphia: Elsevier Science.
Patient examination
Gopal, M.S., Jeffery, N.D. (2001) Magnetic resonance imaging in the diagnosis and treatment of a canine spinal cord injury. Journal of Small
Animal Practice 42, 2931.
Griffiths, I. (1970) A syndrome produced by dorso-lateral explosions of
the cervical intervertebral discs. Veterinary Record 87, 737741.
Holliday, T.A. (1980) Clinical signs of acute and chronic experimental lesions
of the cerebellum. Veterinary Science Communications 3, 259278.
Hunt, G.B., Chapman, B.L. (1991) Trophic ulceration of two digital pads.
Australian Veterinary Practitioner 21, 196, 205.
Mayhew, I.G. (1999) The healthy spinal cord. American Association of
Equine Practitioners 45, 5666.
McDonnell, J.J., Platt, S.R., Clayton, L.A. (2001) Neurologic conditions
causing lameness in companion animals. Veterinary Clinics of North
America, Small Animal Practice 31, 1738.
Morgan, P.W., Parent, J., Holmberg, D.L. (1993) Cervical pain secondary
to intervertebral disc disease in dogs; radiographic findings and surgical
implications. Progress in Veterinary Neurology 4, 7680.
Neer, T.M. (1992) A review of disorders of the gallbladder and extrahepatic biliary tract in the dog and cat. Journal of Veterinary Internal
Medicine 6, 186192.
Long-term functional outcome of dogs with severe spinal cord injuries.
Journal of Neurotrauma (in press).
Pang, W.W., Mok, M.S., Lin, M.L., Chang, D.P., Hwang, M.H. (1998)
Application of spinal pain mapping in the diagnosis of low back pain
analysis of 104 cases. Acta Anaesthesiology Singapore 36, 7174.
Scott, H.W. (1997) Hemilaminectomy for the treatment of thoracolumbar

disc disease in the dog: a follow-up study of 40 cases. Journal of Small
Scott, H.W., McKee, W.M. (1999) Laminectomy for 34 dogs with thoracolumbar intervertebral disc disease and loss of deep pain perception.
Journal of Small Animal Practice 40, 417422.
Seim, H.B., Withrow, S.J. (1982) Pathophysiology and diagnosis of caudal
cervical spondylo-myelopathy with emphasis on the Doberman Pinscher.
Journal of the American Animal Hospital Association 18, 241251.
Walmsley, B., Tracey, D.J. (1983) The effect of transection and cold block
of the spinal cord on synaptic transmission between Ia afferents and
motoneurones. Neuroscience 9, 445451.
FURTHER READING
de Lahunta, A. (2001) Neurological examination. In: K.G. Braund (ed.),
Clinical Neurology in Small AnimalsLocalization, Diagnosis and
Treatment. Ithaca: International Veterinary Information Service
(www.ivis.org), B0201.1001. http://www.ivis.org/special_books/Braund/
deLahunta/chapter_frm.asp?LA1
33
Diagnosis and differential

diagnosis
Differential diagnosis
A: C1C5 35
B: C6T2 35
C: T3L3 36
D: L4S3 37
Further reading
35
39
On the basis of the neurological examination the lesion is

localized to one of the following four areas of the spinal
cord (2.22):
A: C1C5.
B: C6T2.
C: T3L3.
D: L4S3.
These areas relate to spinal cord segments. Refer to 1.2
and 1.5 to identify the location of these areas within the
vertebral column.
A differential diagnosis list is then drawn up, which
embraces all the disease conditions that could be present.
It is usual to create this list using the DAMNIT format
(Table 3.1), the elements of which can also be expressed
using the acronym VITAMIN D. Conditions that occur
frequently in each of the four spinal cord regions are
listed in Tables 3.23.5. When the animal has no neurological deficits, consider the conditions listed in Table 2.1
(disorders mimicking spinal disease) and Table 3.6 (diffuse or non-localizable pain). Other differential diagnoses
are listed in Box 7.2 (neck pain), Box 8.1 (thoracolumbar disc disease), Box 9.1 (atlantoaxial disease), Table
10.2 (lumbosacral disease), Table 11.1 (cervical spondylomyelopathy), and Box 13.1 (trauma). Non-spinal causes
of neurological disease and differentials for exercise intolerance are included in Table 3.6.
The list of differential diagnoses for any given patient
can be shortened considerably if some basic information
is taken into account, such as the following:
The age of the patient.
Whether the condition is acute or chronic,
progressive or static.
Presence or absence of spinal pain.
Chapter
The neurological localization will also exclude some

conditions (see Chapter 2).
The breed of the patient may also indicate that certain

diseases are more likely to be present, but this should
not be the only information on which the diagnosis is
based. An additional way to narrow the list is by a process
of elimination. For the dog in 3.1, the signalment and
history were not suggestive of degenerative, anomalous
or metabolic conditions; there was no pain to suggest
trauma; and the pattern of progression made vascular
disease unlikely. The two most probable differential
diagnoses were therefore either neoplasia or infectious
and inflammatory disease.
DIFFERENTIAL DIAGNOSIS
A: C1C5
In adult dogs, intervertebral disc disease (Table 3.2) is
the most common condition affecting this region. In
dogs less than 2 years of age the differential diagnoses are
different. Here, atlantoaxial subluxation, inflammatory
CNS disease, discospondylitis and trauma are the most
likely causes. Spinal tumors can occur in dogs of any
age. The most common tumors in the cervical spine are
meningiomas and nerve sheath tumors; both are more
common in older dogs.
Some dogs with cervical spondylomyelopathy (CSM)
have neurological signs with a C1C5 pattern of dysfunction (see page 28), although the lesion is often
more caudal in the cervical spine. Acute, non-painful,
non-progressive deficits usually result from ischemic
myelopathy due to fibrocartilaginous embolism (FCE).
Signs are often asymmetrical and severe.
In cats, clinical disc disease is rare in the neck.
The likely diagnoses are trauma, neoplasia (usually
lymphoma) and inflammatory diseases, particularly
feline infectious peritonitis (FIP). Atlantoaxial subluxation is rare in cats.
B: C6T2
Similar considerations apply here to those causing C1C5
signs, although atlantoaxial subluxation does not occur.
36
Table 3.1 Differential diagnoses for spinal disease using the DAMNIT scheme *
Category
Diseases (common conditions in bold)
Degenerative
Degenerative myelopathy
Synovial cysts
Leukodystrophies
Lumbosacral disease
Facet joint pain
Cervical spondylomyelopathy
Disc disease
Lysosomal storage disease
Cervical fibrotic stenosis
Sacroiliac joint pain
Anomalous
Syringo(hydro)myelia
Pilonidal (dermoid) sinus
Spinal arachnoid cyst
Epidermoid cyst
Atlantoaxial subluxation
Spina bifida
Tethered cord syndrome
Sacral osteochondritis dissecans
Vertebral malformations
Cartilaginous exostoses
Meningo(myelo)celes
Spinal dysraphism
Hereditary myelopathy
Sacrocaudal dysgenesis
Schmorls node
Atlantooccipital dysplasia
Metabolic
See Table 3.6, Exercise intolerance
Lysosomal storage disease
Neoplastic
Primary or secondary tumor
Epidural lipomatosis
Nutritional
Hypervitaminosis A
Thiamine deficiency
Idiopathic
Tumoral calcinosis
Disseminated idiopathic skeletal hyperostosis
Iatrogenic injury
Peridural scar
Diagnostic, radiation or surgical injury
Inflammatory and
Infectious
Discospondylitis
Spinal epidural empyema
Gelfoam reaction
Meningo(encephalo)myelitis
Cauda equina neuritis
Foreign body migration
Traumatic
Sacrocaudal injury
Brachial plexus avulsion myelopathy
Gunshot injury
Dural tear
Traumatic disc injury
Fracture/luxation
Toxic
See Table 3.6, Exercise intolerance
Vascular
Fibrocartilaginous embolism
Spinal cord hematoma
Vascular malformation
Fat graft necrosis
Traumatic feline ischemic myelopathy

Neurogenic claudication
Ascending myelomalacia
Spinal cord hemorrhage
* Common conditions are in bold.
Intervertebral disc disease is less frequent in the caudal

cervical spine, but does occur, occasionally even at C7/T1.
Cervical spondylomyelopathy is most prevalent in
the caudal cervical spine of Dobermans and Great Danes.
Ischemic myelopathy also occurs with some frequency
in this region (see Table 3.3).
C: T3L3
This region of the spine accounts for most cases of spinal
disease. Disc herniation is the most likely diagnosis in
dogs older than 1 year. In younger dogs, inflammatory
CNS disease, discospondylitis and trauma are common. In cats, disc disease is uncommon but does occur,
Diagnosis and differential diagnosis
Table 3.2 Diseases causing signs that localize to C1C5

(see also Table 3.1)
Painful
Non-painful
Acute
Chronic
Atlantoaxial
subluxation
CSM
Disc disease
Neoplasia
Discospondylitis
Meningomyelitis
Fracture/luxation
Spinal cord
hematoma
Atlantoaxial
subluxation
Atlantooccipital
dysplasia
CSM
Cervical fibrotic stenosis
Synovial cysts
Syringohydromyelia
Hypervitaminosis A
FCE

Tumoral calcinosis
Neoplasia
CSM, cervical spondylomyelopathy.
Painful
Non-painful
Chronic
Painful
Disc disease
Neoplasia
Discospondylitis
Meningomyelitis
Fracture/luxation
Ascending
myelomalacia
Disc disease
Synovial cysts
Tumoral calcinosis
Non-painful
Traumatic feline
ischemic
myelopathy
FCE
Degenerative
myelopathy
Neoplasia
Acute
Chronic
Painful
Neoplasia
Discospondylitis
Meningomyelitis
Fracture/luxation
Disc disease
Ascending
myelomalacia
Psoas muscle injury
Lumbosacral
disease
Tethered cord
syndrome
Sacral OCD
Non-painful
Cauda equina
neuritis
Ischemic
neuromyopathy
FCE
Degenerative
myelopathy
Spina bifida
Sacrocaudal
dysgenesis
Dermoid sinus
Neoplasia
Chronic
CSM
CSM
Disc disease
Synovial cysts
Neoplasia
Discospondylitis
Meningomyelitis
Fracture/luxation
Spinal cord hematoma
Brachial plexus
avulsion
FCE
Acute
Table 3.5 Diseases causing signs that localize to L4S3

Table 3.3 Diseases causing signs that localize to C6T2

Acute
Table 3.4 Diseases causing signs that localize to T3L3

Spinal arachnoid
cyst
Dermoid sinus
Neoplasia
CSM, cervical spondylomyelopathy.
particularly in aged animals. Trauma, neoplasia and

inflammatory diseases are the most likely causes of
feline thoracolumbar spinal disease.
In large-breed, older dogs with chronic signs, the primary differential diagnosis is degenerative myelopathy.
Chronic disc herniation (Hansen type II), synovial
cyst(s), neoplasia and mild L4S3 lesions must be eliminated before this diagnosis is reached (Table 3.4).
D: L4S3
Subacute or chronic presentations with signs localizing
to this region are usually referable to the lumbosacral
junction (Table 10.2). Neoplasia or discospondylitis
OCD, osteochondritis dissecans.
may be present. Lumbosacral disease is common in

large breeds of dog, but is less common in smaller dogs
and is rare in cats. In sacrocaudal injuries, signs referable to the bladder, perineum, and even paraparesis
may be seen.
Degenerative myelopathy may appear to localize to
this region if the patellar reflexes are absent. This is
due to nerve root involvement; the lesion is still largely
T3L3 in nature. Ischemic myelopathy is also seen here.
In young animals, congenital defects of the vertebrae or
spinal cord are likely (Table 3.5). Exterior signs of spina
bifida may be present, but this disease is uncommon.
Taking the signalment and neurological localization into
account can therefore reduce the differential diagnosis
37
38
3.1 Eight-year-old Akita that presented with progressive tail

paralysis, loss of anal reflex, incontinence and pelvic limb
weakness over 3 weeks. The myelogram indicates a discrete
filling defect within the dural sac over L6 vertebra. Final
diagnosis was an intradural fibrosarcoma.
list. It may be reasonable, in some circumstances, to

proceed with treatment without further investigations,
for example in a middle-aged Dachshund with mild
T3L3 signs and where conservative treatment is
planned. However, if the patient does not progress as
expected, or if there is any doubt about the diagnosis,
the provisional diagnosis should be reassessed and more
3.2 German shepherd dog with severe lumbosacral pain.

Rectal examination revealed an asymmetrical prostate gland
fixed to the pelvis. Ultrasound revealed an enlarged prostate
(arrowheads) with mineralization (arrow) and small cystic cavities
suggestive of prostatitis or carcinoma. The dog also had
lumbosacral disease (10.4) and dermatofibrosis with renal
cystadenocarcinoma (4.35). Nevertheless, it tested positive for
Ehrlichia canis and returned to work as a search and rescue
dog after treatment with doxycycline.
definitive diagnostic tests instituted (see Chapter 4).

Occasionally, an animal may present with more than
one disease that could explain its clinical signs (3.2).
Neurological localization should define the relevant
Table 3.6 Differential diagnoses for diffuse pain, non-spinal neurological disease and exercise intolerance
Diffuse pain
Non-spinal neurological disease
Exercise intolerance
Polyarthritis
Polymyositis
Pancreatitis
Renal or ureteral calculi
Gallstones
Gastrointestinal parasites
Kidney worm
Osteoporotic vertebral fracture
Mid-thoracic vertebral lesion
Other abdominal pain
Meningoencephalitis
Thalamic pain syndrome
Prostatic disease
Urethral tumor
Myositis or myopathy
Peripheral neuropathies
Radiculopathies
Myasthenia gravis
Tick paralysis
Subacute organophosphate toxicity
Myasthenia gravis
Mild cervical myelopathy
Addisons disease
Hypoglycemia
Polymyositis
Toxic myopathy
Hypokalemic polymyopathy
Hyperthyroidism
Congenital myopathy
Tick paralysis
Coonhound paralysis
Lumbosacral pain
Idiopathic polyradiculoneuritis
Protozoal myositis and
polyradiculoneuritis
Subacute organophosphate
toxicity
Botulism
Peripheral neuropathy
Intermittent claudication
Cardiac disease
Upper airway disease
Botulism
Localized tetanus
Brain tumor
Psoas muscle hemorrhage
Diagnosis and differential diagnosis
lesion but occasionally each disease may localize to the

same region. This shows that the clinician must strive
to keep an open mind and try to rule out each potential
diagnosis in step-wise fashion. Response to treatment is
one useful means of doing this, in which case the simplest therapy should be tried first. The case discussed
in 3.2 illustrates the importance of considering a full
range of differential diagnoses for each case and not
simply assuming that the most common disease process

is the one responsible for the clinical signs.
FURTHER READING
Braund, K.G. (2003) Clinical Neurology in Small Animals: Localization,
Diagnosis and Treatment. http://www.ivis.org/special_books/Braund/
toc.asp
39
Diagnostic aids
Routine laboratory analysis

Hematology 41
Biochemistry 41
Urinalysis 42
Serology 42
Microbiology 42
Other 42
41
4
Biopsy
61
Key issues for future investigation

References
Cerebrospinal fluid 42
Indications and contraindications for CSF
collection 43
Collection of CSF 43 (64)
Sample handling and laboratory analysis 43
Normal CSF findings 45
Abnormal CSF findings 45
Normal findings in the face of disease 45
Radiography 45
Survey radiographs 45
Radiographic positioning and normal spinal
radiographs 46
Special radiographic procedures
Myelography 48 (70)
Epidurography 50
Discography 50
48
Principles of spinal radiology 50

Myelographic interpretation 50
Complications of myelography
Chapter
62
Procedures 64
Collection of CSF 64
Myelography 70
The exact selection of diagnostic tests to be employed

varies with the circumstances. In the following chapters, recommendations as to the tests most likely to provide a diagnosis in a particular disease are given. Clearly,
individual preference, experience and the availability
of equipment will influence these decisions.
ROUTINE LABORATORY ANALYSIS

A complete blood count, chemistry profile and urinalysis should be checked in all patients undergoing nonemergency procedures. They rarely provide definitive
diagnostic information in neurological diseases, but can
be important for detecting concurrent disorders (Braund
and Sharp, 2003) (Table 3.6). In emergency situations,
analysis may be restricted to packed cell volume, total
protein, serum glucose and urea concentrations.
Hematology
In the majority of animals with spinal disease, the
hemogram is unremarkable, although a stress leukogram (lymphopenia, eosinopenia and leukocytosis) is a
common finding.
Patients with inflammatory diseases of the spinal
cord and meninges usually have normal hemograms.
Dogs with discospondylitis or rickettsial disease and
cats with feline infectious peritonitis (FIP) may have
inflammatory hemograms, but this is not consistent.
51
Other imaging techniques 54

Ultrasonography 54
Computed tomography 55
Magnetic resonance imaging 57
Scintigraphy 59
Clinical electrophysiology 60
Electromyography 60
Spinal cord evoked response 60
F waves and cord dorsum potentials
62
Biochemistry
61
Metabolic diseases may cause generalized weakness,

which could be mistaken for spinal disease. Examples
42
include Addisons disease, Cushings disease, hypothyroidism, hypoglycemia, hypocalcemia, hyponatremia

and hypokalemic polymyopathy.
Raised serum creatine kinase concentrations are usually indicative of muscle injury. This increases the index
of suspicion for protozoal disease, especially when multifocal neurological deficits, pulmonary or hepatic disease
coexist (4.46). Note that serum creatine kinase concentrations may be raised in patients that suffer trauma
or are recumbent for lengthy periods.
Abnormal biochemical findings may be seen in certain
spinal diseases. Hypocalcemia is seen in some patients
with generalized bone disorders, for example in secondary hyperparathyroidism. Hypercalcemia is seen in some
patients with lymphoma or myeloma; this is a paraneoplastic effect. Hypergammaglobulinemia may be seen
in myeloma and in chronic, infectious diseases such as
FIP; serum electrophoresis is useful to differentiate
the two.
Urinalysis
Urinalysis may provide specific diagnostic information
regarding renal and hepatic function or urinary tract
infection.
If bladder function is in doubt in patients with spinal
disease, urinalysis should be performed on a sample
collected by cystocentesis when the patient is first
evaluated. Ultrasound may aid collection of urine if the
bladder is not palpable. Subsequent urinalysis should
be performed periodically until normal urinary function returns (see Chapter 15).
Urinary tract infection (UTI) associated with urine
retention is present frequently in animals with spinal
disease. Urinalysis reveals high white blood cell counts
and elevated total protein, and bacteria may be present
in the sediment. If so, urine culture should be performed
and antibiotic sensitivity determined. Fungal hyphae
may be detected in dogs with systemic aspergillosis,
especially if the urine has been at room temperature
for more than 12 h.
For routine urinalysis, aseptic catheterization or
cystocentesis is suitable. For culture, cystocentesis is
preferred.
BenceJones proteinuria may be a feature of
myeloma.
Serology
Serology is useful in many diseases, and should be considered in any spinal patient where the origin of the disorder is obscure. Specific examples include Neospora
caninum, Cryptococcus neoformans, Coccidioides immitis,
Bartonella sp., Ehrlichia sp., and also Brucella canis in
dogs with discospondylitis. Acute and convalescent
titers are usually necessary for Rickettsia sp. and

Toxoplasma gondii.
Canine distemper virus (CDV) and FIP both cause
neurological lesions, which may present with spinal
involvement. The presence of serum IgG is not confirmatory of central nervous system (CNS) involvement by
these viruses unless it can be shown to be due to intrathecal production rather than caused by leakage from
serum. Ante-mortem confirmation of these diseases
is challenging.
Non-spinal causes of weakness may be detected by
serum tests such as cholinesterase concentrations in
organophosphate toxicity or acetylcholine receptor
antibody titers in animals with myasthenia gravis
(Table 3.6).
Microbiology
Urine culture is indicated in patients with cystitis related
to neurological disease. Sensitivity testing should also be
performed to determine the most appropriate antibiotic
with which to treat infections. However, bladder infections usually only resolve when urinary function returns
to normal.
In patients with discospondylitis, urine and blood
culture may be attempted to isolate the causative
organism along with direct aspiration from the affected
disc space(s).
If wound infections occur following surgery, culture
from the depths of the wound should be performed
after aseptic preparation of the skin surface (see
Wound infection, page 357).
Other
Endocrine testing, analysis of Von Willebrand (VW) factor, and bleeding times may all be deemed necessary
during preoperative evaluation (see Chapters 6 and 11).
Coagulation profiles and testing for warfarin are indicated in animals with coagulopathies. Blood gas analysis should be performed in tetraplegic animals where
hypoventilation is suspected (7.11).
CEREBROSPINAL FLUID
Cerebrospinal fluid (CSF) analysis is an important tool
in the investigation of neurological patients, but there
are many limitations to CSF analysis. Abnormal CSF is
strongly indicative of neurological disease, but it is a
relatively non-specific finding.
Cerebrospinal fluid should be collected routinely
from either the cerebello-medullary cistern (CMC
cisterna magna) or the lumbar region. It is important
to do this prior to performing myelography as injection
of contrast medium induces a sterile meningitis that
Diagnostic aids
will confound subsequent interpretation of CSF analysis

for 1 week or more (Widmer et al., 1992). When radiography and myelography are normal, and in patients
with multifocal signs, CSF should always be analysed.
Even when the diagnosis is obvious, the information
gained may be very useful on a retrospective basis. An
example would be a dog with a thoracolumbar disc
extrusion that has a very high protein level in CSF and
that subsequently develops myelomalacia after surgery.
Collection of CSF is a very low-risk procedure when
performed with care and when taking appropriate precautions. It is nevertheless an invasive test and informed
consent should always be obtained. Owners should also
be appraised of any specific risk factors before the procedure (see below). Collection of CSF should always
be performed with the patient intubated and under
general anesthesia. Ideally, corticosteroids should not be
given prior to CSF collection as they may alter results
and render interpretation inaccurate.
Following spinal trauma unless a myelogram is to

be performed.
Patients that could have atlantoaxial subluxation
should not have CSF collected from the CMC
because of the danger associated with patient
positioning.
If raised intracranial pressure or brain herniation
could be present then the risk of collection is
increased significantly. This may be so in any
patient with a severe intracranial lesion, but
particularly following head trauma, with spaceoccupying lesions, or with severe inflammatory
disease. Signs indicative of raised intracranial
pressure include depression, stupor or coma;
hypoventilation; bradycardia; dilated pupils or
anisocoria; and cranial nerve deficits such as
decreased or absent physiologic nystagmus and
poor or absent pupillary light reflexes. Collection
of CSF from either of the spinal sites in the face of
increased intracranial pressure increases the
likelihood of brain herniation.
Indications and contraindications

for CSF collection
INDICATIONS
Collection of CSF
See Procedures, page 64.
Suspected intracranial disease (see Contraindications, below).

Suspected multifocal disease.
Spinal disease where survey radiographs are
normal.
Suspected polyneuropathy.
Ideally in every patient unless contraindicated.
CONTRAINDICATIONS
Where general anesthesia is contraindicated.
Sample handling and laboratory

analysis
Analysis of CSF should include gross examination, total
and differential white blood cell (WBC) count and total
protein (4.1). A sedimentation technique (4.2) should
be used unless the sample can be sent to a specialized
laboratory that will perform cytocentrifugation (Cellio,
4.1 A: Left to rightnormal CSF; xanthochromia (which is caused by hemoglobin breakdown products and indicates previous
intrathecal hemorrhage); and hemorrhage at collection. B: Degenerate neutrophils containing bacteria from the CSF of a dog with
severe, diffuse pain and fever secondary to myelography performed 24 h earlier using iohexol from a contaminated, multi-use bottle.
A filter should be used with multi-use containers.
43
44
B
4.2 Hans Wolfgang Kolmel Sedimentation Chamber (Free
University of Berlin, with permission). A: Glass microscope
slide is placed on the apparatus, B: filterpaper is placed on the
glass slide, and C: the chamber is fitted. CSF is placed in the
chamber and the filter paper absorbs fluid, while cells remain
on the slide. The slide is then stained for cytology and
differential count.
2001). White cell analysis should be performed ideally

on fresh CSF within 30 min of collection as the cells
deteriorate rapidly (Bienzle et al., 2000). If this is not
possible, the specimen should be split in two aliquots
and then one aliquot is put into a small EDTA tube or a
drop of serum or an equal volume of 4% formalin solution added in order to preserve the cells (Evans, 1989;
Bienzle et al., 2000). Other analyses performed on
CSF include immunology and electrophoresis.
automatic means. Undiluted CSF is used, unless cell

counts are particularly high in which case the CSF will
appear turbid.
Normal CSF is free of red blood cells. Normal white
cell counts vary by laboratory, but are generally less
than 5 cells per l (5 106 per liter). An increase in
WBCs is termed pleocytosis (4.1B).
GROSS EXAMINATION
Mild blood contamination has little effect on WBC

counts in CSF. There are formulae for correcting CSF
cell counts that take the peripheral blood WBC count
into consideration (Rossmeissl et al., 2002). However,
contaminating red blood cells (RBCs) do not greatly alter
WBC counts, especially those already markedly elevated
(Wilson and Stevens, 1977). It is probably adequate to
remember that WBCs and RBCs are in a ratio of approximately 1 : 500 in blood, and to take this into account
when viewing cell counts in blood-contaminated CSF.
Normal CSF is clear and colorless. Color change and

turbidity may be noted in some diseases but these do
not occur unless the abnormalities in cell count or protein are marked. The most frequent color changes are
due to hemorrhage or xanthochromia (4.1A).
CELL COUNTS
These are performed using a hemocytometer; the
cell numbers are too low to be counted accurately by
CELL COUNTS AND BLOOD

CONTAMINATION
Diagnostic aids
DIFFERENTIAL WHITE CELL COUNTS

AND CYTOLOGY
This is an important part of the examination of CSF,
even in the face of a normal total WBC count. The
WBCs must be concentrated by centrifugation or sedimentation (4.2), mounted on a slide and stained (Cellio,
2001). Most WBCs in normal CSF are mononuclear
lymphocytes, monocytes and occasional neutrophils or
macrophages (Duque et al., 2002).
The differential WBC count is most useful in distinguishing acute and chronic inflammation, for example
between non-infectious suppurative meningoencephalitis and granulomatous meningoencephalitis (Thomson
et al., 1989, 1990; Chrisman, 1992). Occasionally, high
white cell counts will be seen in disc disease as well
(Thomson et al., 1989).
PROTEIN CONTENT
Protein content of CSF is estimated by a number of
methods; use of a professional laboratory is best. Normal
CSF collected from the CMC has a protein concentration of less than 30 mg/dl; it may be up to 45 mg/dl at
the lumbar site. Total protein may be increased in many
diseases but is non-specific. Electrophoresis provides
information regarding the composition of the CSF
protein.
MICROBIOLOGY
Examination for the presence of bacteria may be performed by microscopy of stained samples (4.1B) and by
culture. If CSF culture is attempted, enhancement
using blood culture bottles is recommended. Bacterial
diseases of the CNS are uncommon in dogs and cats and
even when present, culture of CSF tends to be unrewarding (Remedios et al., 1996; Lavely et al., 2002;
Radaelli and Platt, 2002) (see below). Care should be
exercised in interpreting positive results in the absence
of pleocytosis, as bacteria may be contaminants.
Normal CSF findings

Normal CSF findings are given in Table 4.1.
Table 4.1 Normal values for commonly evaluated CSF

parameters
Color
Specific gravity
White cell count
Differential white cells
Clear, colorless
1.0041.006
5/l
Lymphocytes and monocytes
Total protein
Cerebello-medullary cistern
Lumbar
1030 mg/dl (0.10.3 g/l)

1045 mg/dl (0.10.45 g/l)
Mononuclear pleocytosis is generally indicative of

chronic inflammation, for example canine distemper or
granulomatous meningoencephalomyelitis.
Neutrophilic pleocytosis is seen mostly with infectious
or non-infectious meningoencephalomyelitis (aseptic
meningitis) but may also be seen with meningiomas and
vascular disease.
Mixed pleocytosis is non-specific, but can be seen in
some tumors, particularly meningiomas.
Identification of neoplastic cells in CSF is unusual. It
is most likely in lymphoma and choroid plexus tumors
(see Chapter 12).
Increased protein in the absence of pleocytosis
(albuminocytological dissociation) is indicative of noninflammatory disorders such as neoplasia or vascular
disease.
Bacteria may be seen in CSF occasionally (4.1B).
These may be pathogenic, but if present in the absence
of a neutrophilic pleocytosis, they are usually the
result of reagent contamination. Intracellular bacteria
should always be viewed as being significant (Munana,
1996).
Normal findings in the face of

disease
Many animals with CNS disease, including spinal diseases, have normal CSF (Thomson et al., 1989, 1990;
Chrisman, 1992; Cellio, 2001). This is particularly so if
the CSF sample is collected cranial to the lesion; that
is, from the CMC in thoracolumbar and lumbosacral
lesions. Thus, finding normal CSF does not rule out the
possibility of a lesion being present.
Abnormal CSF findings

Abnormal CSF is indicative of CNS (or nerve root)
disease (Murray and Cuddon, 2002).
High numbers of RBCs in the CSF are generally
indicative of contamination at collection. Hemorrhage
in the CNS may be inferred from the presence of xanthochromia (4.1A) or erythrophagocytosis (macrophages
containing RBCs).
RADIOGRAPHY
Survey radiographs
Radiography is a valuable diagnostic aid for spinal
patients. Good quality survey radiographs will provide
the diagnosis in some cases. To obtain diagnostic
radiographs of the spine, general anesthesia is usually
45
46
4.3 Radiographic positioning aids.

A: Radiolucent foam wedges covered in
plastic, floppy sandbags and ties.
B: Foam wedges, ties, floppy sandbag
along with 2.5 cm tape and a trough (see
examples of use in 4.44.14).
4.4 Lateral cervical spine. The patient is in lateral recumbency

with the head extended and the thoracic limbs pulled caudally.
A wedge is placed under the sternum to prevent rotation
(visible in 4.10). It is usually best to take two views, one centered
at C2 and one at C6.
necessary to position the patient correctly and to minimize radiation exposure of personnel. In many animals
with spinal disease, the radiological features are subtle
so that accurate positioning is essential. An exception is
in acute spinal injuries, where general anesthesia or
manipulations could exacerbate the neural damage.
The area of interest must be centered properly and the
beam collimated closely. Survey radiographs of large
areas of the spine are rarely useful. Stress radiography
may be helpful but is not without risk (Farrow, 1982)
(9.5, 10.12, 11.10, 13.17).
Radiographic positioning and

normal spinal radiographs
Various positioning aids facilitate spinal radiography
(4.3). Techniques for radiographing the spine are given
in the following section.
CERVICAL SPINE (4.44.9)

Stressed survey views of the cervical spine are not generally of value except when done with care to assess the
atlantoaxial joint (9.3).
4.5 Lateral cervical spine. A foam wedge is placed under the

middle part of the neck to avoid sagging, which could result in
distortion of the intervertebral spaces.
a
b
4.6 Lateral cervical radiograph of a normal dog. Note that the

transverse processes are superimposed (a); the intervertebral
spaces near the middle of the film are clear and the end plates
are parallel to the beam. The transverse processes of C6 are
large and project ventrally (b), (7.37B). See 4.16 for cranial
cervical radiograph. See also 11.13B.
ATLANTOAXIAL JOINT (4.8, 4.9)

Projections are as for the cervical spine with the beam
centered appropriately. For the lateral projection, a mild
degree of flexion may be employed to demonstrate
Diagnostic aids
4.9 Lateral atlantoaxial radiograph of a normal dog. Note the

normal relationship between the dorsal arch of C1 and the
spinous process of C2. (a) Wing of atlas. Rotating the head
slightly makes the dens more visible.
4.7 Ventrodorsal cervical spine. The patient is placed in

dorsal recumbency, with the whole body aligned vertically.
The beam is centered on the area of interest. It is useful to
remove the endotracheal tube for this view, particularly in
myelography.
4.8 Ventrodorsal atlantoaxial radiograph of a normal dog.

Note the dens (arrow) and the atlantoaxial articulations
(arrowheads).
subluxation, but this must not be excessive (9.3, 9.5).

The use of the open-mouth view to radiograph the dens
is not recommended, as flexing the neck can be dangerous. The ventrodorsal projection is adequate and the
cervical extension also alleviates cord compression. The
spinal deviation present in many patients with atlantoaxial subluxation can make positioning difficult.
A common error is to try to evaluate the atlantoaxial
joint in poorly positioned films taken of a conscious
4.10 Lateral thoracolumbar spine. The patient is in lateral

recumbency with the limbs positioned as shown. Foam wedges
are placed under the sternum (arrow) and between the limbs to
prevent rotation and under the lumbar vertebrae to avoid
sagging. The beam is centered on the area of interest and
collimated closely over the spine to reduce soft tissue scatter
and improve radiographic quality.
patient with the area of interest near the edge of the

film. Accurate interpretation is often impossible (see
Chapter 9).
THORACOLUMBAR SPINE (4.104.13)

A common error is to center the beam in the midlumbar region when evaluating the spine for disc disease; most herniations occur in the T11L1 region, and
the beam should be centered accordingly.
47
48
4.11 Ventrodorsal thoracolumbar spine. The patient is in dorsal

recumbency with the limbs extended. It is important that the body
is upright and the beam collimated.
4.14 Lateral lumbosacral spine. Positioning is crucial here as

rotation may induce artefacts. The patient is positioned as for
the thoracolumbar spine, taking particular care to keep both
pelvic limbs parallel to the table top.
a
b
4.12 Lateral thoracolumbar spinal radiograph of a normal dog.

The dog is well positioned, with the ribs (a) and transverse
processes (b) superimposed. The vertebral end plates in the
middle of the image are parallel to the beam, whereas those
near the right edge are not.
4.15 Lateral lumbosacral radiograph of a normal dog. In a

well-positioned radiograph, the wings of the ilium will be
directly superimposed.
LUMBOSACRAL SPINE (4.144.15)
SPECIAL RADIOGRAPHIC
PROCEDURES
Myelography
4.13 Ventrodorsal thoracolumbar spinal radiograph of a

normal dog.
In myelography the spinal cord is outlined by positive

contrast medium injected into the subarachnoid space.
The indications for myelography are:
Where the neurological examination indicates a spinal
lesion, but none is visible on survey radiographs.
To determine the significance of multiple lesions
identified on survey radiographs.
To determine the presence of spinal cord
compression and especially to evaluate dynamic
lesions (10.7, 11.411.6).
To assist in deciding whether or not to perform
surgery as well as the type of procedure to be
performed.
Diagnostic aids
This procedure is contraindicated if general anesthesia

or spinal puncture is unsafe (see Cerebrospinal fluid,
page 42), or where inflammatory disease of the CNS is
present. The presence of mild CSF pleocytosis is not
in itself a contraindication, as it is present in many
compressive spinal diseases, including disc herniation
(Thomson et al., 1989, 1990).
The choice of contrast medium is important, as many
are extremely irritating to neural tissue. Hyperosmolar
or hypertonic solutions must be avoided. A non-ionic,
water-soluble contrast medium must be usedIohexol
(Omnipaque, Nycomed) is the contrast medium of
choice (Wheeler and Davies, 1985; Allen and Wood,
1988). Concentrations in the range of 180350 mg
iodine/ml are usedmost often 240 or 300 mg iodine/
ml. The contrast medium should be warmed to body
temperature before injection (Lamb, 1994).
Performance of a myelogram does not preclude the
need to take high-quality survey radiographs. The practice of using the survey films only to establish radiographic exposure factors is deplored, as significant
information that would be visible on good survey films
may be masked by the myelogram. The region of interest
is dictated by the neurological examination, remembering that occasionally a lower motor neuron (LMN)
lesion can mask an upper motor neuron (UMN) lesion
(2.24).
Myelography is carried out with the patient under
general anesthesia. Injection of contrast medium is
either at the CMC or in the lumbar subarachnoid
space. The techniques for spinal puncture are described above in CSF collection. Contrast is injected
following collection of the CSF sample (4.64, 4.66).
4.16 Normal lateral myelogram of the cranial cervical region.

Note that the contrast columns are wide in C1 and C2. The
ventral column thins and is elevated over the C2/3 intervertebral
space and to a lesser degree over the other spaces.
4.17 Normal lateral myelogram of the caudal cervical region.
CERVICAL MYELOGRAPHY (4.16, 4.17)

Technique is described under Procedures, page 70.
LUMBAR MYELOGRAPHY (4.184.20)
Technique is described under Procedures, page 71.
Lumbar myelography is required in many circumstances, the most common being an acute disc extrusion.
Here, injection of contrast medium requires relatively
high pressures to delineate the lesion, as the spinal cord
is often swollen or under pressure. Some radiologists
prefer lumbar myelography for all patients.
Epidural leakage is often a problem in lumbar
myelography, which makes interpretation of the study
very difficult (4.67). Slow injection of the contrast
medium may reduce the chance of epidural leakage.
Another common complication is poor filling of the
subarachnoid space, especially in animals with spinal
cord swelling (8.1).
4.18 Normal lumbar lateral myelogram. Note the streaked

appearance of the contrast medium where it outlines the nerves
of the cauda equina.
4.19 Lateral lumbosacral myelogram from a normal dog. Note

that the contrast column passes well into the sacrum.
49
50
Normal
Extradural
Intradural
extramedullary
4.20 Ventrodorsal lumbosacral myelogram of a normal dog.

The separate contrast columns can be seen in L5 (arrows).
The columns converge and cross the lumbosacral junction.
Superimposition of the spinous processes and the presence of
colonic contents can make interpretation difficult.
Intramedullary
Epidurography
In epidurography, contrast medium is introduced into
the epidural space. For some radiologists this was the
preferred method for evaluating the lumbosacral region
(Barthez et al., 1994; Ramirez and Thrall, 1998) (10.8B).
CT or MRI are superior in most cases.
Discography
In discography, contrast medium is injected into the
nucleus pulposus of the intervertebral disc (Sisson et al.,
1992; Barthez et al., 1994). Normally, only a very small
amount can be introduced (0.10.2 ml in the lumbosacral disc). If there is damage to the anulus fibrosus,
more contrast can be injected and the leakage will be evident on subsequent radiographs (10.8A). The technique
may be particularly useful if combined with CT
(Ohnmeiss et al., 1997; Milette et al., 1999).
PRINCIPLES OF SPINAL
RADIOLOGY
It is important to take a systematic approach to evaluating spinal radiographs. The clinician has the advantage
of having similar adjacent structures with which to compare the suspected lesion. Accuracy of interpretation is
limited by the radiographs, and it is well worth paying
attention to obtaining good images as discussed above.
Also, knowledge of the normal radiographic anatomy is
required, as discussed in this chapter and in Chapter 1.
One system used to evaluate lateral radiographs is
to assess each structure in turn from cranial to caudal:
start ventrally with the hypaxial muscles and end with
4.21 Spinal lesions are classified according to their location

relative to the spinal cord and the dura mater. Examples of
potential causes are given in Table 4.2.
the spinous processes dorsally. For ventrodorsal radiographs, start on the left side and work cranial to caudal.
This technique should be employed even if there is an
obvious lesion, to avoid overlooking other important
features or additional lesions.
General radiological principles dictate that images
are evaluated for the following features: position, size,
number, contour, architecture and opacity.
Many of these alterations are seen in spinal disease
and most are shown in this book.
Myelographic interpretation
From the myelogram, it is usually possible to gain an
impression of the location of the lesion relative to the
spinal cord (4.244.31, 14.4). Note that lesions in all
locations may give the appearance of an expanded or
swollen cord in an image taken face on to the lesion.
For this reason, it is essential that perpendicular
projections (lateral and ventrodorsal) are made (4.30,
4.31, 7.2).
Interpretation is relatively straightforward if there
are changes in the skeleton at the site of the myelographic abnormality, for example in disc herniation or
vertebral neoplasia (4.47). Generally, it is possible to
make some estimate of the location of the lesion as
shown in 4.21. Causes of these compression patterns
are given in Table 4.2. Extradural lesions are shown in
Diagnostic aids
Table 4.2 Common causes of myelographic abnormalities
Extradural
Intradural
extramedullary
Intramedullary
Degenerative
Disc herniation
Synovial cyst
Congenital
Arachnoid cyst
Syringohydromyelia
Neoplasia
Primary or metastatic
Vertebral or soft tissue
Meningioma
Nerve sheath tumor
Nephroblastoma
Glioma
Ependymoma
Metastatic
Inflammatory
Discospondylitis
Epidural abscess
Myelitis
Trauma
Disc
Bone fragments
Vascular
Hematoma
Ischemic myelopathy (acute stage)

Hematoma
4.23 Extradural lesion. Myelogram of a Doberman with Wobbler

syndrome and a large, extradural lesion due to what appears to
be a simple disc extrusion at C6/7. The ventral contrast column is
split (arrow and arrowhead) over the C6 vertebra (4.24, 4.25).
4.22 Extradural lesion. Diagram of the myelographic pattern

of an extradural lesion.
4.224.25, 7.2, 8.1; intraduralextramedullary lesions

are shown in 4.264.28; and intramedullary lesions are
shown in 4.294.31.
A CT scan often enhances the utility of the
myelogram markedly and also aids surgical planning
(4.25, 11.5).
EXTRADURAL LESIONS
INTRADURALEXTRAMEDULLARY LESIONS
INTRAMEDULLARY LESIONS
Spinal cord swelling may be evident although myelography rarely delineates parenchymal lesions. One
exception is dogs with generalized malacia of the spinal

cord (13.10, 14.18). CT reconstruction can further
enhance lesion detection after myelography (4.41) but
MRI is superior for most parenchymal lesions.
COMPLICATIONS OF
MYELOGRAPHY
The most common complications after cervical injection
are either that the contrast has entered the subdural
space (1.11); that the needle has been displaced from
the vertebral canal; that excessive contrast has entered
the cranial vault (4.65); or either that the contrast has
not reached the lesion or that it has flowed past the
lesion (Lamb, 1994; Penderis et al., 1999)(4.32). The
most common problem after lumbar myelography is
51
52
4.24 Extradural lesion causing column splitting. Diagram of

column splitting, which generally indicates an asymmetrical
extradural lesion (4.25, 4.33).
4.26 Intraduralextramedullary lesion. Diagram of the typical

golf-tee pattern (arrows).
4.27 Intraduralextramedullary lesion. Ventrodorsal myelogram

at L1 of a 10-year-old mixed-breed dog with recurrence of
paraparesis after removal of a nerve sheath tumor at L3/4.
There was evidence of spread along the subarachnoid space
(4.28). Necropsy revealed an anaplastic sarcoma of meningeal
origin. An MRI from this dog is shown in 4.44B.
B
4.25 Same dog as in 4.23. A: Column splitting was due to a
probable left-sided synovial cyst (its location shifted slightly into
C7 with the dog on its back for the CT). A ventral slot was done
at C6/7 but little disc material was removed and the dog was
much worse the next day. B: A postoperative CT myelogram
showed more marked compression, due in part to collapse of
the interspace. Compression in this dog was probably
exacerbated by the synovial cyst (11.8). In retrospect this dog
should have undergone a distraction-stabilization procedure.
4.28 Intraduralextramedullary lesions. Ventrodorsal

myelogram of this dog is shown in 4.27. An MRI of this area is
shown in 4.44B.
Diagnostic aids
4.31 Intramedullary lesion. Lateral myelogram of the dog shown

in 4.30. The myelographic columns diverge in this view as well
as in the ventrodorsal view.
4.29 Intramedullary lesion. Diagram showing spinal cord
swelling.
4.32 Poor contrast filling is common over the caudal cervical

spine. The cranial cervical and thoracic spines can be elevated
in order to cause the heavier contrast agent to pool in the
caudal cervical area (from McKee et al., 2000).
4.30 Intramedullary lesion. Ventrodorsal myelogram at L1/2

of a 7-year-old Doberman with progressive paraparesis of
2 weeks duration. The myelographic columns diverge in this
ventrodorsal view, as they do in the lateral view (4.31). Diagnosis
was a highly-invasive neurofibrosarcoma with extensive spinal
cord invasion but only a very small mass around one dorsal
nerve rootlet.
poor myelographic quality (8.1A); this can be due to

several reasons. Epidural leakage may obscure the subarachnoid contrast medium (4.67). Fortunately, epidural
contrast is usually absorbed more rapidly than the subarachnoid contrast medium and so resolution may
improve after 1020 min. Another potential problem is
inadvertent injection into the subdural space rather
than the subarachnoid space. This is not uncommon

one study found a higher incidence with injections at
the CMC (13%) than after lumbar injection (2.6%).
Subdural contrast is usually restricted to the dorsal
area and this is probably due to a restriction of ventral
flow caused by the denticulate ligaments (Lamb, 1994,
1997; Scrivani et al., 1997; Penderis et al., 1999;
Scrivani, 2000) (1.11).
Contrast in the subarachnoid space may need to be
pooled by manipulating the animals position (4.32).
Positioning the animal in sternal recumbency with its
head elevated slightly can cause contrast to collect in
the caudal cervical area for a dorsoventral view; analogous pooling techniques can be applied to the lumbar
spine. One final potential problem is that the lesion
may only be visible when the animal is lying on one side
but not when it is on the other side (4.33).
Neurological deterioration can occur in some animals
after myelography (Allen and Wood, 1988). Largebreed dogs with significant cervical cord lesions; dogs
with chronic spinal cord compression, meningitis or
extradural tumors; and those with degenerative
myelopathy seem to be affected most often. Fortunately,
this deterioration is usually transient, most patients
returning to their pre-myelogram status within a few
53
54
4.33 Some lesions may be missed if

only one lateral view is taken. A: There is
a generalized loss of contrast over the
L1/2 space in this radiograph taken with
the dog in right lateral recumbency.
B: Same dog in left lateral recumbency.
There is marked splitting of the ventral
contrast column due to an asymmetric
lesion. This indicates that the lesion lies
on the left side of the vertebral canal
(Matteucci et al., 1999).
A
4.34 Injection of contrast into the spinal cord parenchyma. This

can cause severe neurological deficits and focal malacia, which
is related to the volume injected. If injection is performed without
fluoroscopic guidance, a test injection should be used (Parker
et al., 1975; Servo and Laasonen, 1985; Kirberger and Wrigley,
1993).
days. Clearly, if the spinal puncture or injection technique itself is at fault, significant neurological damage
may result (4.34). Injection of contrast into the central
canal can occur in cisternal myelography; the effect on
the patient varies but is generally serious. Cardiovascular
effects are usually seen and neurological deterioration is
likely. Central canal injection in the lumbar spine can
cause temporary deterioration of deficits and the severity is related to the volume injected. The likelihood of
this complication increases with injections at sites cranial
to L5/6 (Kirberger and Wrigley, 1993). This is particularly true in small dogs where the spinal cord terminates
more caudally than in large dogs (Morgan et al., 1987).
Injection of contrast medium can be made between T13
and L2 vertebrae but the incidence of epidural leakage
(4.67), and of central canal injection is then higher compared to more caudal injections (McCartney, 1997).
Nevertheless, provided that a test injection is employed,
this technique can be useful when injection is not successful at more caudal sites. It is also useful in dogs with
lesions that localize to the L4S3 spinal cord segments as
there is then much less risk of injecting into the lesion
itself; use of this site is much less desirable in dogs when
the lesion is located in the thoracolumbar region.
Manipulations involved in obtaining flexion, extension or traction radiographs during myelography can lead
to neurological deterioration and it is wise not to
maintain these positions for excessive periods of time,
particularly when extending the cervical spine (11.9,
11.10).
Seizures occur infrequently on recovery from anesthesia following myelography (Wheeler and Davies, 1985;
Allen and Wood, 1988). The site of injection appears not
to influence the frequency of seizures; dogs with CSM
may be prone to this complication (Lewis and Hosgood,
1992). Seizures are best managed by intravenous
diazepam or barbiturates and so it is prudent to leave an
intravenous catheter in place in any patient recovering
after a myelogram to facilitate medication.
OTHER IMAGING TECHNIQUES

Ultrasonography
This technique has two broad indications. The first and
most important is to evaluate the heart and abdomen.
The abdomen should be examined in animals that
might have concurrent lesions caused by trauma or by
neoplastic, inflammatory or endocrine disease (4.35).
Echocardiography is indicated in dogs with cardiac disease secondary to degenerative, neoplastic, inflammatory or traumatic disease.
The second indication is to visualize the nervous
system. This can be done either through natural portals such as a fontanelle or the atlantoaxial (4.36) and
lumbosacral spaces. It can also be done through thin
areas of the skull in toy breed dogs with hydrocephalus
(a condition that may coexist with syringohydromyelia).
In addition, the examination can be performed at
surgery through a laminectomy defect or ventral slot.
The main value of intraoperative ultrasonography
is that the surgeon may be able to visualize a tumor
or disc material that is obscured by the spinal cord. It
can therefore serve as a real-time aid to localizing and
defining the extent of a lesion. This could make the
difference between a complete or incomplete tumor
Diagnostic aids
4.35 German shepherd dog that presented with mild

neurological deficits and lumbosacral pain. The dog had small
dermal nodules all over its skin suggestive of dermatofibrosis,
an inherited condition that is associated with renal cystadenocarcinoma (Moe and Lium, 1997). Ultrasound revealed
multiple masses within each kidney (arrow). This dog also had
several other disorders (3.2, 10.4); it died 6 months later.
resection or prevent the surgeon leaving behind significant amounts of disc material (Nakayama, 1993;
Finn Bodner et al., 1995; Hudson et al., 1995; Ham
et al., 1995) (7.8, 8.6). Sonography has also proven
useful in determining the extent of arachnoid cysts
(Galloway et al., 1999). In addition, it can show
irregular new bone on vertebrae affected by tumors or
discospondylitis.
Computed tomography
Computed tomography (CT) of the vertebral column
is very useful in certain circumstances, particularly
in patients with mineralized disc extrusions, vertebral
tumors or cervical spondylomyelopathy. Contrast
enhancement by myelography (7.3A), rather than
intravenous contrast administration as performed in
intracranial imaging, outlines the subarachnoid space if
needed. A much lower dose of contrast medium is
required for CT myelography than for conventional
myelograms. If only a CT study is planned, the dose of
contrast medium is reduced to about one quarter of the
4.36 Ultrasonography can be used to examine the spinal cord;

normal C1 region seen through the atlantooccipital space. The
dorsal and ventral dura mater is visible (arrows); the two lines
ventrally probably represent the dura and pia mater. The
hyperechoic area centrally represents the central canal; CSF in
the subarachnoid space is anechoic (Finn Bodner et al., 1995).
4.37 A: Dog in dorsal recumbency,
intubated and connected to a ventilator
prior to undergoing a CT scan; this
minimizes movement artefacts caused
by breathing (for both CT and MRI).
Breathing artefact is less of a problem
for lesions in the cranial cervical or
lumbosacral regions. B: Normal L1/2
intervertebral space of a dog to show
epidural fat within each foramen
(arrowheads). See also 1.23A. This
pattern is lost after disc extrusion (4.40A).
dose for conventional myelography. If the CT study

is performed directly after a conventional myelogram,
the natural dilution and absorption of the contrast
medium is sometimes adequate to reduce the contrast
concentration in some animals or the patient can be positioned to allow contrast to flow away from the lesion
(taking care that it does not run into the head). Animals
are usually positioned in dorsal recumbency (4.37).
Animals in ventral recumbency tend to lean to one side
or the other, especially those with a narrow chest but
some fractures may be more stable in this position
(4.38). In general CT images should be made at right
angles to the vertebral canal and not angled as shown
in 4.39.
For non-mineralized disc extrusions, either CT myelography or intravenous contrast-enhanced CT may be
required to see the lesion (Sharp et al., 1995) (11.52).
Use of contrast medium is usually unnecessary for
chondrodystrophoid breeds of dog (8.2A, 8.3) and will
also tend to obscure some of the more subtle features
of disc disease. The study is therefore non-invasive and
55
56
4.38 Positioning for a CT scan

demands the same precautions as
for radiography; unstable lesions and
dynamic studies warrant particular
care. A: This dog had an L2 fracture
(arrowhead), which was displaced
markedly by putting the dog into
dorsal recumbency (scout CT image).
B: It was repositioned in ventral
recumbency to reduce the
displacement (arrowhead) and made a
good recovery after surgery (13.48).
4.39 A: Cranial and caudal extent of

extruded disc material shown on a
scout image. This also identifies the
exact disc or vertebra(e) involved, which
is particularly useful in surgical planning.
B: Transverse CT scan through the
C2/3 disc to show the mineralized
extrusion (arrow).
B
4.40 A: CT scan through the T12/13
intervertebral disc space from the same
dog as shown in 4.37B. There is loss of
the normal epidural fat pattern at the
level of the right foramen (arrow). Disc
extrusion was confirmed at surgery.
B: Hemorrhage in the epidural or
subdural space (arrow) following a disc
extrusion in the thoracolumbar area.
Often the rim of hyperattenuation is less
distinct than this (Olby et al., 2000).
Contrast has not been administered.
also very fast (510 min depending on the scanner).

Although CT scans are usually more expensive than
myelograms, the reduced anesthesia time means that
the final cost is often similar. The region from caudal T9
vertebra to cranial L4 vertebra is scanned for dogs with
lesions that localize to the T3L3 spinal cord segments.
The region from caudal L3 vertebra to mid-sacrum is
scanned for dogs with lesions that localize to the L4S3
spinal cord segments. Three-millimetre slices are usually adequate although additional 1 mm slices can be
made through the region of interest (Olby et al., 2000).
CT is more useful than myelography for surgical planning and can also be used to depict the cranial and caudal extent of disc material clearly (Olby et al., 1999)
(4.39). Survey radiographs are recommended in addition to the CT scan in order to allow the surgeon to identify any anatomical anomalies (8.198.21).
Interpreting CT scans in dogs with thoracolumbar

disc disease has been reviewed (Olby et al., 2000).
Normal CT images are shown in Chapter 1 (1.17, 1.21,
1.22, 1.23A, 1.25, 1.28). Images should be viewed
initially using soft-tissue parameters (4.40) and then
subsequently using bone parameters (1.17). The normal
spinal cord has intermediate attenuation on transverse
images, equivalent to that of soft tissue such as adjacent kidney. Spinal cord causes more attenuation than
epidural fat, which is usually most evident at the level
of the disc space, especially adjacent to the intervertebral foramen (1.23A, 4.37B). Displacement of
epidural fat at this level (4.40A), or an increase in
the attenuation characteristics of the epidural fat, is
a useful indicator of a disc extrusion. Another very
common feature is the presence of a focal, heterogenous,
hyperattenuating mass of mineralized disc material in an
Diagnostic aids
4.41 The A: lateral and B: ventrodorsal myelogram in this dog revealed slight expansion of the spinal cord over the cranial C5/6
disc space. The myelogram was followed by a CT scan, which was inconclusive (not shown). C: 3D reconstruction of the CT
myelogram identified an expansile lesion within the spinal cord. The dog was euthanized; necropsy was not performed. Had MRI
been available it would have been a more efficient way to demonstrate this lesion.
epidural location (8.3). A less common presentation

occurs when disc material and blood are spread more
diffusely along and around the spinal cord. In such cases
the epidural mass is relatively small in individual transverse images and is only slightly more attenuating than
the spinal cord. It does, however, still cause more attenuation than epidural fat, which is either displaced or
infiltrated with blood and nuclear material (Olby et al.,
2000). In some animals a rim of increased attenuation is
visible around the spinal cord, which almost certainly
represents epidural or subdural hemorrhage (Penderis
et al., 1999; Olby et al., 2000; Tidwell et al., 2002)
(4.40B). The exact appearance of hemorrhage may
depend on how recent it is and if the hemoglobin has
degraded (Tidwell et al., 1994). Adjacent chronic disc
lesions, which are frequent and often incidental findings
in dogs with acute extrusions, are recognized by their
extreme hyperattenuation, which may approach that of
cortical bone. The degree of attenuation can be quantified by measuring the Hounsfield units (CT numbers)
within a defined region of interest.
For dogs with suspected cervical disc extrusions, the
scan extends from the mid-atlas (useful to help rule
out atlantoaxial instability by demonstrating a normal
dens) to cranial T1 vertebra. Attenuation characteristics of cervical disc extrusions in chondrodystrophoid
breeds are similar to those seen in the thoracolumbar
area (4.39B, 7.4A) except that extruded disc material
rarely extends over more than one vertebra and diffuse,
subdural hemorrhage is unusual. Reconstruction techniques can be used to enhance further the diagnostic
value of CT in disc disease (7.4, 7.51) as well as for
other conditions (4.41).
Comparisons between CT and MRI in humans show
that each has its own strengths and weaknesses (see also
Chapter 10, pages 186187). CT is superior for identifying
(cervical) fractures whereas MRI is better for soft tissue
lesions (Cirillo et al., 1988; Kent et al., 1992; Klein

et al., 1999). MRI is more accurate for showing the
extent of medullary bone involvement in osteosarcoma
(OFlanagan et al., 1991); is superior to myelography
and CT myelography in cervical myelopathy (Masaryk
et al., 1986); is at least as accurate as CT myelography for
identifying disc disease and for differentiating scar tissue
from recurrent disc (Modic et al., 1984; Sotiropoulos
et al., 1989; Yousem et al., 1992); and is superior to CT
at identifying epidural abscess (Angtuaco et al., 1987).
CT myelography is superior to MRI in detecting multiple myeloma (Mahnken et al., 2002); and in delineating
meningeal carcinomatosis (Krol et al., 1988) CT myelography is equivalent to MRI for diagnosis of nerve root
avulsion in brachial plexus injury (Doi et al., 2002), and
is a very accurate means of identifying cervical disc herniations (Houser et al., 1995). CT usually gives superior
spatial resolution to MRI, which can be important in
identifying spondylosis and the lateral extent of disc
extrusion (Karnaze et al., 1988; Jones et al., 2000a).
Magnetic resonance imaging

Magnetic resonance imaging (MRI) has an important role in imaging spinal diseases, particularly in lumbosacral disease and parenchymal spinal cord lesions
(deHaan, 1993; Kippenes et al., 1999). Normal anatomy and basic changes in canine disc disease have been
reviewed (Sether et al., 1990a,b; Kippenes et al., 1999)
(see pages 1113). Imaging of spinal lesions is covered
below and in the relevant chapters (7.3B, 8.2B, 9.7,
10.12, 12.3, 12.6B, 12.41A, 13.7, 14.3A, 14.15, 14.17,
14.19). The region of interest to be covered by the scan
is dictated by the neurological localization, bearing in
mind that occasionally a LMN lesion can obscure an
UMN lesion (2.24). The animal is usually positioned in
dorsal recumbency to minimize movement artefact
57
58
4.42 A: T2-weighted, sagittal MRI showing

spinal cord edema (arrow) caused by an
acute disc herniation at C5/6 in a dog with
Wobbler syndrome. There is also a large,
more chronic lesion at C6/7. Both disc
spaces have reduced central signal intensity
in this T2-weighted image. B: T1-weighted,
transverse image at C5/6 to show the spinal
cord (arrowhead) being compressed by a
dorsally displaced, non-mineralized disc
(arrow). Compare to 4.44A, 7.3B and
11.52. Some scanners indicate the level of
the transverse image on each image (inset).
A
B
4.43 A: Transverse, T1-weighted MRI of
the L2/3 disc in an 11-year-old paraparetic
Labrador (same dog as in 8.2B). The
spinal cord is deformed by a large,
somewhat right-sided, low signal mass.
B: Corresponding image after gadoteridol
(Prohance, Bracco Diagnostics,
Mississauga, Ontario) administration.
There is a rim of enhancement to the mass
(arrow) (Vroomen et al., 1998; Saifuddin
et al., 1999). Mini-hemilaminectomy with
corpectomy was used to remove the
chronic disc material.
from breathing (4.37A) and sagittal scans using T1 and

T2 weighting are used to identify the lesion. T2weighted images highlight water and are especially useful for this as many lesions have an associated spinal cord
edema that highlights the lesion clearly (4.42A, 13.7,
14.15, 14.19). Transverse images are then made through
the region of interest (4.42B, 4.43, 4.44A). Intravenous
contrast (gadolidium [Gd-DPTA, Magnavist, Berlex laboratories, Liberty Corner, NJ] or gadoteridol [Prohance,
Bracco Diagnostics, Mississauga, Ontario] both used at
0.10.2 mmol/kg) is administered if needed, followed
by another series of sagittal and transverse T1 scans
(4.43B, 4.44B). Additional pulse sequences may also be
added as required (Kippenes et al., 1999) (9.7B).
The MRI features of cervical disc lesions in dogs
are similar to those described in humans. Loss of signal intensity of the nucleus pulposus on T2-weighted
images is a common finding but is also non-specific. A
more reliable feature is displacement of epidural fat by
extruded disc material as shown in both sagittal and
transverse plane images. Extruded disc material has low
signal intensity and it may alter the shape of the spinal
cord in a transverse plane (Levitski et al., 1999b;
Yamada et al., 2001). These features are also consistent
with the changes seen in the lumbosacral region

(Adams et al., 1995) (see Chapter 10). In acute thoracolumbar disc extrusions there is often an associated
hemorrhage at the site of extrusion secondary to tearing of the internal venous plexus (Sether et al., 1990a;
Olby et al., 2000). Mineralized disc material may show
as a signal void (i.e. hypointense) and the extradural
spinal cord compression is usually caused by a combination of disc material and hematoma. The hematoma is
initially signal hypointense on T1- and T2-weighted
images but between 2 and 7 days after onset the signal
becomes more hyperintense as deoxyhemoglobin is
converted to methemoglobin (Tidwell et al., 2002).
There is often only minimal contrast enhancement of
disc material or hematoma but this can identify which
lesion is more recent (Sether et al., 1990a; Tidwell and
Jones, 1999) (4.43, 8.2B). The terminology used to
describe disc lesions is discussed on page 12:
Herniation is a general term denoting a nonspecific type of disc abnormality and should not be
used to indicate clinical significance.
Extrusion describes disc material that has clearly
escaped the normal boundaries of the disc and is
usually significant clinically.
Diagnostic aids
4.44 A: Transverse image of the

dog shown in 7.3B. There is a large,
hypointense extrusion of mineralized disc
material (arrow) causing marked
unilateral compression of the spinal cord
(arrowhead). The inset shows the level
of this image (C45). B: T1-weighted
MRI of the dog shown in 4.27 and 4.28
after contrast. The enhancing mass
(arrowhead) is causing marked
compression of the spinal cord. It is not
clear from the MRI if the tumor is extraor intra-medullary (Kippenes et al.,
1999).
The terms bulge and protrusion are also nonspecific and a morphological description of the
disc displacement is preferred.
Criteria for differentiating clinically insignificant
lesions from incidental, age-related changes (7.15A)
are discussed below and in Chapter 10.
It must be remembered that some disc herniations are
incidental, age-related findings (4.42A) that are not
responsible for the clinical signs (Jensen et al., 1994;
Milette et al., 1999; Jones and Inzana, 2000b) (see
Chapters 1, 10, page 188 and 11, page 213). MRI of people without back pain has shown that 50% have one disc
herniation and 25% have two (Jensen et al., 1994).
Interpretation must therefore be made using clinical
signs, electromyographic (EMG) changes in local muscles, and imaging characteristics (Ramirez and Thrall,
1998). Indicators of clinical relevance in dogs include
changes in posture or evidence of pain such as spontaneous vocalization or discomfort on palpation of local
paralumbar muscles (Chrisman, 1975; Nardin et al.,
1999). Overall, MRI shows excellent anatomical detail
with reasonably high sensitivity but only low specificity,
resulting in a high rate of false positives. In contrast to
MRI, electromyography has high specificity and is therefore useful to confirm which lesions are physiologically
significant, thereby avoiding unnecessary interventions
(Nardin et al., 1999; Robinson, 1999). Imaging features
of disc lesions that suggest physiological relevance include
edema in bone marrow adjacent to the disc (Sether et al.,
1990a; Morrison et al., 2000); edema within the spinal
cord over the affected disc space (4.42, 13.7); contrast
enhancement, which shows a high correlation with rupture of the anulus (Vroomen et al., 1998; Saifuddin
et al., 1999) (4.43); and changes within the disc itself
(Jenkins et al., 1985; Milette et al., 1999) (8.2B).
MRI signal characteristics within the disc itself are
useful in characterizing internal derangement, which
in humans with low back pain can indicate that the
lesion is symptomatic even in the absence of a change
in disc contour. Decreased central disc signal intensity
on T2-weighted images is highly predictive of anular

tears extending into or beyond the outer anulus and a
majority of these are responsible for symptoms. A
localized area of hypersignal on T2-weighted images is
also a reliable marker of symptomatic, outer anular disruption. However, a disc can also have normal MRI
characteristics and yet still be symptomatic (Milette
et al., 1999). Changes in disc signal intensity on T2weighted images may prove to be useful for dogs and
cats when a lesion does not involve clear extrusion of
disc material, but this remains to be proven.
MRI features of spinal cord tumors (4.44B, 12.6B,
12.41A) have been reviewed (Kippenes et al., 1999).
Localization of the mass is initially made using sagittal T2weighted images. Identification of tumor extent and its
relationship to surrounding tissues is best done with T1weighted images after contrast administration. Extradural
tumors can usually be identified accurately although their
signal characteristics and enhancement patterns
are not consistent. Contrast enhancement of vertebral
tumors may be of less value than for tumors in other locations. Assessment of bone infiltration is facilitated by use
of additional pulse sequences such as fat suppression or
T2-weighted gradient echo, which delineate tumor by its
hyperintense signal characteristics. Determining the
intraduralextramedullary compartment is not always
straightforward and it is not uncommon for this classification to be inaccurate even though most of these tumors
show marked contrast enhancement (4.44B). CSF imaging studies (MR myelography) may improve anatomic
classification (Kippenes et al., 1999).
Other indications for MRI include degenerative
(Levitski et al., 1999a,b; Lipsitz et al., 2001; Webb
et al., 2001), anomalous (14.3A), inflammatory (Kraft
et al., 1998) (14.15), traumatic (13.7), and vascular
(14.17, 14.19) lesions (Gopal and Jeffery, 2001).
Scintigraphy
Bone scintigraphy evaluates certain functional aspects of
bone, particularly related to blood supply and metabolic
59
60
4.45 Nuclear bone scan from a 10year-old German shepherd dog with
paraparesis and back pain of 1 week
duration. A: An osteoproductive lesion is
visible at T5 (arrow); note also the soft
tissue mass ventral to the affected
vertebral body. B: Increased uptake of
technetium is evident on the bone scan.
Final diagnosis was osteosarcoma.
activity. Certain types of lesions will produce a hot

spot on a bone scan (4.45); this may be evident before
radiographic changes are visible. Particular uses of bone
scintigraphy include discospondylitis, vertebral tumors,
and small pathological fractures secondary to osteoporosis (Lamb et al., 1990; Stefanacci and Wheeler, 1991;
Cook et al., 2002).
CLINICAL ELECTROPHYSIOLOGY
Electromyography
Electromyography (EMG) is the method by which the
electrical activity of muscle is studied and analyzed.
This technique can assist the neurosurgeon by helping
to localize some of the more subtle lesions as being
either UMN or LMN in nature. If LMN deficits are
present they will generally be associated with spontaneous electrical activity in those muscles supplied by the
injured spinal cord segment or nerve root (4.46). It is
important to remember that this spontaneous activity
takes between 4 and 7 days to appear, which is the time
taken for the axons to degenerate from the site of injury
to the neuromuscular junctions (Cuddon et al., 2003).
Two broad categories of LMN lesions can be seen.
The first relates to lesions that are restricted to either the
brachial or lumbosacral regions, such as brachial plexus
avulsion injury or damage to the cauda equina. In this
case, the spontaneous activity will be restricted to a
discrete muscle group(s). Use of EMG in radiculopathy
4.46 Spontaneous electrical activity

recorded from the gastrocnemius
muscle of a 7-month-old Sharpei with
protozoal polyradiculoneuritis and
myositis. The dog was paraplegic with
LMN deficits. A: The two main types
of spontaneous activity are visible:
1. Fibrillation potential. 2. Positive
sharp wave. The dog tested positive for
Neospora caninum; creatine kinase (CK)
was elevated. B: Muscle biopsy
identified numerous protozoal organisms.
has high specificity and as such is a very useful complement to MRI. In particular, the combination of these
two tests is one way to overcome the high rate of falsepositive diagnosis that can occur using MRI alone (Nardin
et al., 1999; Robinson, 1999). The second category of
LMN disease is when the animal is suffering from a
generalized peripheral neuropathy, or possibly a myopathy, in which case spontaneous activity will be widespread throughout the body and will not be restricted
to any particular muscle group.
If an UMN lesion is present, electromyography of
the paraspinal muscles can sometimes be useful to help
localize the lesion. Although UMN deficits are evident
mainly through their effects on white matter tracts,
they will also impair neurons in the local gray matter
at the same level. Damage to these neurons often
produces spontaneous activity in epaxial and hypaxial
spinal muscles adjacent to the injured spinal cord segment. This activity seems to occur as a result of both
degenerative and irritative effects, so the latter may be
evident on an acute basis (Chrisman, 1975).
Spinal cord evoked response

In this technique, a recording electrode is placed on the
dorsal lamina of a cervical, thoracic, or lumbar vertebra
in order to detect an electrical response in the spinal
cord. The impulse in the spinal cord is created by stimulating a peripheral sensory nerve distal to the lesion.
The impulse or evoked response needs to undergo signal
Diagnostic aids
averaging to remove background information, much like

a brainstem auditory evoked response. The spinal cord
evoked response has the potential to give prognostic
information about the functional state of the spinal cord
after trauma. The technique is performed routinely in
humans undergoing surgery for scoliosis and can reduce
the incidence of postoperative paraparesis and paraplegia by more than 60% (Dawson et al., 1991).
Various parameters can be recorded from the evoked
waveforms traveling in the spinal cord, but no single
one has proved to be predictive of the eventual outcome
in dogs with acute spinal cord injuries. Mathematical
manipulations of the data may provide useful information
but it can be difficult to identify and interpret the waveforms accurately in dogs with severe spinal cord injuries,
as the waveforms are often small and dispersed
(Sylvestre et al., 1993). It would be a tremendous advantage to have a more objective criterion than nociception
to predict the eventual outcome in an animal with severe
spinal cord injury. The main challenges are the need for
one person to be dedicated solely to the monitoring technique, the variation in size of canine and feline patients,
and the fact that pre-existing neurological deficits make it
hard to get useful potentials (Cuddon et al., 2003).
Further study is required before the spinal cord evoked
response can make a significant contribution to the management of clinical cases (Holliday, 1992; Poncelet et al.,
1993; Cuddon et al., 2003).
B
4.47 A: Radiograph of the lumbar spine of a 4-year-old Golden
retriever with a 2-week history of lumbar pain and asymmetrical
paraparesis. Loss of bone within the L4 vertebral body is seen.
B: Myelogram of the same dog showing ventral extradural
compression. A spinal radiographic survey, bone scan and
thoracic radiographs revealed no other lesions.
F waves and cord dorsum potentials

F waves are probably of more value than peripheral
motor nerve conduction studies for most spinal diseases. They represent delayed action potentials that
arrive a few milliseconds after the compound muscle
action potential. F waves are initiated by antidromic
conduction of a stimulus along a motor axon that elicits an action potential from the ventral horn cell in the
spinal cord. This secondary action potential is then conducted orthodromically down the motor axon in the
ventral nerve root and peripheral nerve to produce a
second muscle-evoked action potential, which is
known as an F wave. These waveforms therefore provide a specific way to assess nerve roots; the longer
path taken by the F wave also means that any abnormalities tend to be magnified compared to standard
nerve conduction studies. They are particularly useful
in assessing nerve root disease in cauda equina syndrome
(Cuddon et al., 2003) (see Chapter 10).
Cord dorsum potentials are a spinal-cord evoked
response that arise from purely sensory input of
peripheral nerves. They are useful both for evaluating
radiculopathies and also for myelopathies that involve
the intumescences (Cuddon et al., 2003).
4.48 Fine needle aspiration was performed under fluoroscopy.

The material collected was diagnostic of myeloma. The dog
was treated by radiation and chemotherapy (4.49).
BIOPSY
(4.474.49)
Tumors are the most likely reason for biopsy and

are discussed in Chapter 12. Most biopsies of spinal
tissue will be performed following surgical exposure of
the lesion.
61
62
4.49 Same dog as shown in

4.474.48, above. A: Sixty five
months after therapy it was
euthanized due to a pathologic
fracture of C7 caused by recurrence
of myeloma. B: The original lesion
at L4 (arrowhead) had not changed
although neoplastic plasma cells
were still present at this location as
well as in the spleen and kidney
(Rusbridge et al., 1999).
Surgical exposure can sometimes be avoided by fine

needle aspiration of tissue from within the vertebral
canal or vertebral body, preferably under fluoroscopic
(4.48) or CT (12.4) guidance (Irving and McMillan,
1990). A core of vertebral bone can also be taken using
a Jamshidi needle (5.35).

Can MRI signal characteristics and contrast enhancement
be used to differentiate disc lesions that are significant
clinically from those that are not significant?
REFERENCES
Adams, W.H., Daniel, G.B., Pardo, A.D., Selcer, R.R. (1995) Magnetic resonance imaging of the caudal lumbar and lumbosacral spine in 13 dogs
(19901993). Veterinary Radiology and Ultrasound 36, 313.
Allen, G.S., Wood, A.K.W. (1988) Iohexol myelography in the dog.
Veterinary Radiology 29, 7882.
Angtuaco, E.J., McConnell, J.R., Chadduck, W.M., Flanigan, S. (1987) MR
imaging of spinal epidural sepsis. American Journal of Roentgenology
149, 12491253.
Barthez, P.Y., Morgan, J.P., Lipsitz, D. (1994) Discography and
epidurography for evaluation of the lumbosacral junction in dogs with
cauda equina syndrome. Veterinary Radiology and Ultrasound 35,
152157.
Bienzle, D., McDonnell, J.J., Stanton, J.B. (2000) Analysis of cerebrospinal
fluid from dogs and cats after 24 and 48 hours of storage. Journal of the
American Veterinary Medical Association 216, 17611764.
Braund, K.G., Sharp, N.J.H. (2003) Neurological examination and localization. In: D. Slatter (ed.), Textbook of Small Animal Surgery, 3rd edn,
Cellio, B.C. (2001) Collecting, processing, and preparing cerebrospinal
fluid in dogs and cats. Compendium on Continuing Education for the
Practicing Veterinarian 23, 786794.
Chrisman, C.L. (1975) Electromyography in the localization of spinal cord
and nerve root neoplasia in dogs and cats. Journal of the American
Chrisman, C.L. (1992) Cerebrospinal fluid analysis. Veterinary Clinics of
North America, Small Animal Practice 22, 781810.
Cirillo, S., Simonetti, L., La Tessa, G., Elefante, R., Smaltino, F. (1988)
[MR vs CT: which is the diagnostic advantage in neuroradiology].
Radiologia Medica (Torino) 76, 390398.
Cook, G.J., Hannaford, E., See, M., Clarke, S.E., Fogelman, I. (2002) The
value of bone scintigraphy in the evaluation of osteoporotic patients with
back pain. Scandinavian Journal of Rheumatology 31, 245248.
Cuddon, P.A., Murray, M., Kraus, K. (2003) Electrodiagnosis. In: D. Slatter
(ed.), Textbook of Small Animal Surgery, 3rd edn, 11081117.
Philadelphia: Elsevier Science.
Dawson, E.G., Sherman, J.E., Kanim, L.E., Nuwer, M.R. (1991) Spinal
cord monitoring. Results of the Scoliosis Research Society and the
European Spinal Deformity Society survey. Spine 16, S361364.
De Haan, J.J., Shelton, S.B., Ackerman, N. (1993) Magnetic resonance
imaging in the diagnosis of degenerative lumbosacral stenosis in four
dogs. Veterinary Surgery 22, 14.
Doi, K., Otsuka, K., Okamoto, Y., Fujii, H., Hattori, Y., Baliarsing, A.S.
(2002) Cervical nerve root avulsion in brachial plexus injuries: magnetic
resonance imaging classification and comparison with myelography and
computerized tomography myelography. Journal of Neurosurgery 96,
277284.
Duque, C., Parent, J., Bienzle, D. (2002) The immunophenotype of blood
and cerebrospinal fluid mononuclear cells in dogs. Journal of Veterinary
Internal Medicine 16, 714719.
Evans, R.E. (1989) Haematology, biochemistry, cerebrospinal fluid analysis
and other clinicopathological investigations. In: S.J. Wheeler (ed.),
Manual of Small Animal Neurology, 4962. Cheltenham: BSAVA
Publications.
Farrow, C.S. (1982) Stress radiography: applications in small animal practice. Journal of the American Veterinary Medical Association 181,
777784.
Finn Bodner, S.T., Hudson, J.A., Coates, J.R., Sorjonen, D.C., Simpson,
S.T., Cox, N.R., Wright, J.C., Garrett, P.D., Steiss, J.E., Vaughn, D.M.,
Miller, S.C., Brown, S.A. (1995) Ultrasonographic anatomy of the
normal canine spinal cord and correlation with histopathology after
induced spinal cord trauma. Veterinary Radiology and Ultrasound 36,
3948.
Galloway, A.M., Curtis, N.C., Sommerlad, S.F., Watt, P.R. (1999)
Correlative imaging findings in seven dogs and one cat with spinal arachnoid cysts. Veterinary Radiology and Ultrasound 40, 445452.
Ham, L.M.v., Bree, H.J.v., Tshamala, M., Thoonen, H. (1995) Use of computed tomography and computed tomographic myelography for assessment of spinal tumoral calcinosis in a dog. Veterinary Radiology and
Ultrasound 36, 115118.
Holliday, T.A. (1992) Electrodiagnostic examination. Somatosensory
evoked potentials and electromyography. Veterinary Clinics of North
Diagnostic aids
Houser, O.W., Onofrio, B.M., Miller, G.M., Folger, W.N., Smith, P.L.
(1995) Cervical disk prolapse. Mayo Clinic Proceedings 70, 939945.
Hudson, J.A., Finn Bodner, S.T., Coates, J.R., Sorjonen, D.C., Simpson, S.T.,
Wright, J.C., Steiss, J.E., Cox, N.R., Vaughn, D.M., Miller, S.C.,
Brown, S.A., Garrett, P.D. (1995) Color doppler imaging and doppler
spectral analysis in the spinal cord of normal dogs. Veterinary Radiology
and Ultrasound 36, 542547.
Irving, G., McMillan, M.C. (1990) Fluoroscopically guided percutaneous
fine-needle aspiration biopsy of thoracolumbar spinal lesions in cats.
Progress in Veterinary Neurology 1, 473475.
Jenkins, J.P., Hickey, D.S., Zhu, X.P., Machin, M., Isherwood, I. (1985)
MR imaging of the intervertebral disc: a quantitative study. British
Jensen, M., Brant, Z.M., Obuchowski, N., Modic, M., Malkasian, D., Ross, J.
(1994) Magnetic resonance imaging of the lumbar spine in people without back pain. New England Journal of Medicine 331, 6973.
Jones, J.C., Banfield, C.M., Ward, D.L. (2000a) Association between postoperative outcome and results of magnetic resonance imaging and computed
tomography in working dogs with degenerative lumbosacral stenosis.
Journal of the American Veterinary Medical Association 216, 17691774.
Jones, J.C., Inzana, K.D. (2000b) Subclinical CT abnormalities in the
lumbosacral spine of older large-breed dogs. Veterinary Radiology and
Karnaze, M.G., Gado, M.H., Sartor, K.J., Hodges, F.J., 3rd (1988)
Comparison of MR and CT myelography in imaging the cervical and
thoracic spine. American Journal of Roentgenology 150, 397403.
Kent, D.L., Haynor, D.R., Larson, E.B., Deyo, R.A. (1992) Diagnosis of lumbar spinal stenosis in adults: a metaanalysis of the accuracy of CT, MR, and
myelography. American Journal of Roentgenology 158, 11351144.
Kippenes, H., Gavin, P.R., Bagley, R.S., Silver, G.M., Tucker, R.L.,
Sande, R.D. (1999) Magnetic resonance imaging features of tumors of
the spine and spinal cord in dogs. Veterinary Radiology and Ultrasound
40, 627633.
Kirberger, R.M., Wrigley, R.H. (1993) Myelography in the dog: review of
patients with contrast medium in the central canal. Veterinary Radiology
Klein, G.R., Vaccaro, A.R., Albert, T.J., Schweitzer, M., Deely, D., Karasick,
D., Cotler, J.M. (1999) Efficacy of magnetic resonance imaging in the
evaluation of posterior cervical spine fractures. Spine 24, 771774.
Kraft, S.L., Mussman, J.M., Smith, T., Biller, D.S., Hoskinson, J.J. (1998)
Magnetic resonance imaging of presumptive lumbosacral discospondylitis in a dog. Veterinary Radiology and Ultrasound 39, 913.
Krol, G., Sze, G., Malkin, M., Walker, R. (1988) MR of cranial and spinal
meningeal carcinomatosis: comparison with CT and myelography.
American Journal of Roentgenology 151, 583588.
Lamb, C., Berg, J., Bengtson, A. (1990) Preoperative measurement of
canine primary bone tumors, using radiography and bone scintigraphy.
Lamb, C.R. (1994) Common difficulties with myelographic diagnosis of
acute intervertebral disc prolapse in the dog. Journal of Small Animal
Practice 35, 549558.
Lamb, C.R. (1997) Subdural collection of contrast medium during
cervical myelography. Veterinary Radiology and Ultrasound 38, 479481.
Lavely, J., Vernau, K.M., LeCouteur, R.A. (2002) Spinal epidural empyema
in seven dogs. Journal of Veterinary Internal Medicine 16, 369.
Levitski, R., Chauvet, A., Lipsitz, D. (1999a) Cervical myelopathy associated with extradural synovial cysts in 4 dogs. Journal of Veterinary
Levitski, R.E., Lipsitz, D., Chauvet, A.E. (1999b) Magnetic resonance
imaging of the cervical spine in 27 dogs. Veterinary Radiology and
Lewis, D.D., Hosgood, G. (1992) Complications associated with the use
of iohexol for myelography of the cervical vertebral column in dogs:
66 cases (19881990). Journal of the American Veterinary Medical
Lipsitz, D., Levitski, R.E., Chauvet, A.E., Berry, W.L. (2001) Magnetic
resonance imaging features of cervical stenotic myelopathy in 21 dogs.
Veterinary Radiology and Ultrasound 42, 2027.
Mahnken, A.H., Wildberger, J.E., Gehbauer, G., Schmitz-Rode, T.,
Blaum, M., Fabry, U., Gunther, R.W. (2002) Multidetector CT of the
spine in multiple myeloma: comparison with MR imaging and radiography. American Journal of Roentgenology 178, 14291436.
Masaryk, T.J., Modic, M.T., Geisinger, M.A., Standefer, J., Hardy, R.W.,
Boumphrey, F., Duchesneau, P.M. (1986) Cervical myelopathy: a comparison of magnetic resonance and myelography. Journal of Computer
Assisted Tomography 10, 184194.
Matteucci, M.L., Ramirez, O., III, Thrall, D.E. (1999) Radiographic diagnosis: effect of right vs. left lateral recumbency on myelographic appearance of a lateralized extradural mass. Veterinary Radiology and
McCartney, W.T. (1997) Lumbar myelography in 79 dogs, using different
puncture sites. Veterinary Record 141, 417419.
McKee, W., Penderis, J., Dennis, R. (2000) Radiology corner: obstruction
of contrast medium flow during cervical myelography. Veterinary
Radiology and Ultrasound 41, 342343.
Modic, M.T., Pavlicek, W., Weinstein, M.A., Boumphrey, F., Ngo, F.,
Hardy, R., Duchesneau, P.M. (1984) Magnetic resonance imaging of intervertebral disk disease. Clinical and pulse sequence considerations.
Radiology 152, 103111.
Moe, L., Lium, B. (1997) Hereditary multifocal renal cystadenocarcinomas
and nodular dermatofibrosis in 51 German shepherd dogs. Journal of
Small Animal Practice 38, 498505.
Morgan, J.P., Atilola, M., Bailey, C.S. (1987) Vertebral canal and spinal
cord mensuration: a comparative study of its effect on lumbosacral
myelography in the Dachshund and German Shepherd dog. Journal of
the American Veterinary Medical Association 191, 951957.
Morrison, J.L., Kaplan, P.A., Dussault, R.G., Anderson, M.W. (2000)
Pedicle marrow signal intensity changes in the lumbar spine: a manifestation of facet degenerative joint disease. Skeletal Radiology 29, 703707.
Munana, K.R. (1996) Encephalitis and meningitis. Veterinary Clinics of
North America, Small Animal Practice 26, 857874.
Murray, M., Cuddon, P.A. (2002) Cerebrospinal fluid analysis in acute
canine polyradiculoneuritis: albumin quotient and immunoglobulin
index determination using polyacrylamide gel electrophoresis. Journal of
Veterinary Internal Medicine 16, 370.
Nakayama, M. (1993) Intraoperative spinal ultrasonography in dogs: normal
findings and case-history reports. Veterinary Radiology and Ultrasound
34, 264268.
Nardin, R.A., Patel, M.R., Gudas, T.F., Rutkove, S.B., Raynor, E.M. (1999)
Electromyography and magnetic resonance imaging in the evaluation of
radiculopathy. Muscle Nerve 22, 151155.
OFlanagan, S.J., Stack, J.P., McGee, H.M., Dervan, P., Hurson, B. (1991)
Imaging of intramedullary tumor spread in osteosarcoma. A comparison
of techniques. The Journal of Bone and Joint SurgeryBritish volume 73,
9981001.
Ohnmeiss, D.D., Vanharanta, H., Ekholm, J. (1997) Degree of disc disruption and lower extremity pain. Spine 22, 16001605.
Olby, N.J., Munana, K.R., Sharp, N.J.H., Flegel, T., Van Camp, S.,
Berry, C.R., Thrall, D.G. (1999) A comparison of computed tomography
and myelography in the diagnosis of acute intervertebral disc disease in
dogs. Journal of Veterinary Internal Medicine 13, 239.
Olby, N.J., Munana, K.R., Sharp, N.J.H., Thrall, D.E. (2000) The computed
tomographic appearance of acute thoracolumbar intervertebral disc herniations in dogs. Veterinary Radiology and Ultrasound 41, 396402.
Parker, A.J., Cusick, P.K., Stowater, J.L., Park, R.D. (1975) Effects of
intramedullary injection of methiodal and lidocaine on spinal cords
of dogs. American Journal of Veterinary Research 36, 15291532.
Penderis, J., Sullivan, M., Schwarz, T., Griffiths, I.R. (1999) Subdural
injection of contrast medium as a complication of myelography. Journal
Poncelet, L., Michaux, C., Balligand, M. (1993) Somatosensory potentials
in dogs with naturally acquired thoracolumbar spinal cord disease.
American Journal of Veterinary Research 54, 19351941.
Radaelli, S.T., Platt, S.R. (2002) Bacterial meningoencephalomyelitis
in dogs: a retrospective study of 23 cases (19901999). Journal of
63
64
Ramirez, O., III, Thrall, D.E.(1998) A review of imaging techniques for

canine cauda equina syndrome. Veterinary Radiology and Ultrasound 39,
283296.
Remedios, A.M., Wagner, R., Caulkett, N.A., Duke, T. (1996)
Epidural abscess and discospondylitis in a dog after administration of a
lumbosacral epidural analgesic. Canadian Veterinary Journal 37, 106107.
Robinson, L.R. (1999) Electromyography, magnetic resonance imaging, and
radiculopathy: its time to focus on specificity. Muscle Nerve 22, 149150.
Rossmeissl, J.H., Troy, G.C., Inzana, K.D., Jortner, B., Boon, G.D. (2002)
Effects of blood contamination on canine cerebrospinal fluid white blood
cell counts and total protein concentrations. Journal of Veterinary
Rusbridge, C., Wheeler, S.J., Lamb, C.R., Page, R.L., Carmichael, S.,
Brearley, M.J., Bjornson, A.P. (1999) Vertebral plasma cell tumors in 8
Saifuddin, A., Mitchell, R., Taylor, B.A. (1999) Extradural inflammation
associated with annular tears: demonstration with gadolinium-enhanced
lumbar spine MRI. European Spine Journal 8, 3439.
Scrivani, P.V. (2000) Myelographic artifacts. Veterinary Clinics of North
America, Small Animal Practice 30, 303314, vi.
Scrivani, P.V., Barthez, P.Y., Leveille, R., Schrader, S.C., Reed, S.M. (1997)
Subdural injection of contrast medium during cervical myelography.
Servo, A., Laasonen, E.M. (1985) Accidental introduction of contrast
medium into the cervical spinal cord. A case report. Neuroradiology 27,
8082.
Sether, L.A., Nguyen, C., Yu, S.N., Haughton, V.M., Ho, K.C., Biller, D.S.,
Strandt, J.A., Eurell, J.C. (1990a) Canine intervertebral disks: correlation of anatomy and MR imaging. Radiology 175, 207211.
Sether, L.A., Yu, S., Haughton, V.M., Fischer, M.E. (1990b) Intervertebral
disk: normal age-related changes in MR signal intensity. Radiology 177,
385388.
Sharp, N.J.H., Cofone, M., Robertson, I.D., DeCarlo, A., Smith, G.K.,
Thrall, D.E. (1995) Computed tomography in the evaluation of caudal
cervical spondylomyelopathy of the Doberman Pinscher. Veterinary
Sisson, A.F., LeCouteur, R.A., Ingram, J.T., Park, R.D., Child, G. (1992)
Diagnosis of cauda equina abnormalities by using electromyography,
discography, and epidurography in dogs. Journal of Veterinary Internal
Medicine 6, 253263.
Sotiropoulos, S., Chafetz, N.I., Lang, P., Winkler, M., Morris, J.M., Weinstein,
P.R., Genant, H.K. (1989) Differentiation between postoperative scar and
recurrent disk herniation: prospective comparison of MR, CT, and contrastenhanced CT. American Journal of Neuroradiology 10, 639643.
Stefanacci, J.D., Wheeler, S.J. (1991) Skeletal scintigraphy in canine
discospondylitis. American College of Veterinary Radiology 66.
Sylvestre, A.M., Cockshutt, J.R., Parent, J.M., Brooke, J.D., Holmberg, D.L.,
Partlow, G.D. (1993) Magnetic motor evoked potentials for assessing
spinal cord integrity in dogs with intervertebral disc diseases. Veterinary
Surgery 22, 510.
Thomson, C.E., Kornegay, J.N., Stevens, J.B. (1989) Canine intervertebral
disc disease: changes in the cerebrospinal fluid. Journal of Small Animal
Thomson, C.E., Kornegay, J.N., Stevens, J.B. (1990) Analysis of cerebrospinal fluid from the cerebellomedullary and lumbar cisterns of dogs
with focal neurologic disease: 145 cases (19851987). Journal of the
Tidwell, A.S., Jones, J.C. (1999) Advanced imaging concepts: a pictorial
glossary of CT and MRI technology. Clinical Techniques in Small Animal
Practice 14, 65111.
Tidwell, A.S., Mahony, O.M., Moore, R.P., Fitzmaurice, S.N. (1994)
Computed tomography of an acute hemorrhagic cerebral infarct in a
dog. Veterinary Radiology and Ultrasound 35, 290296.
Tidwell, A.S., Specht, A., Blaeser, L., Kent, M. (2002) Magnetic resonance
imaging features of extradural hematomas associated with intervertebral
disc herniation in a dog. Veterinary Radiology and Ultrasound 43,
319324.
Vroomen, P.C., Van Hapert, S.J., Van Acker, R.E., Beuls, E.A., Kessels, A.G.,
Wilmink, J.T. (1998) The clinical significance of gadolinium enhancement of lumbar disc herniations and nerve roots on preoperative MRI.
Neuroradiology 40, 800806.
Webb, A.A., Pharr, J.W., Lew, L.J., Tryon, K.A. (2001) MR imaging findings in a dog with lumbar ganglion cysts. Veterinary Radiology and
Ultrasound 42, 913.
Wheeler, S.J., Davies, J.V. (1985) Iohexol myelography in the dog and cat:
a series of one hundred cases, and a comparison with metrizamide and
iopamidol. Journal of Small Animal Practice 26, 247256.
Widmer, W.R., DeNicola, D.B., Blevins, W.E., Cook, J.R., Jr, Cantwell, H.D.,
Teclaw, R.F. (1992) Cerebrospinal fluid changes after iopamidol and
metrizamide myelography in clinically normal dogs. American Journal of
Veterinary Research 53, 396401.
Wilson, J.W., Stevens, J.B. (1977) Effects of blood contamination on cerebrospinal fluid analysis. Journal of the American Veterinary Medical
Yamada, K., Nakagawa, M., Kato, T., Shigeno, S., Hirose, T., Miyahara, K.,
Sato, M. (2001) Application of short-time magnetic resonance examination for intervertebral disc diseases in dogs. Journal of Veterinary
Medical Science 63, 5154.
Yousem, D.M., Atlas, S.W., Hackney, D.B. (1992) Cervical spine disk
herniation: comparison of CT and 3DFT gradient echo MR scans.
Journal of Computer Assisted Tomography 16, 345351.
PROCEDURES
Collection of CSF
General anesthesia is required for CSF collection in dogs and cats; the depth must be suitable so that the
animal will not move during the procedure. Patients should be intubated and ventilator support must be
available. The collection site must be clipped and prepared aseptically. Sterile surgical gloves should be worn. An
assistant is required to hold the patient in the correct position when collecting from the cerebello-medullary
cistern (CMC).
Which collection site to use warrants some consideration? The two sites available are the CMC and the lumbar
region. Collection from the CMC is easier and is less likely to produce a sample contaminated with blood.
However, as CSF flows in a cranial to caudal direction, abnormal CSF is more likely to be present caudal to a
lesion. Thus, lumbar CSF is more likely to be useful diagnostically. However, lumbar collection is more difficult and
blood contamination occurs more often.
If raised intracranial pressure is present, lumbar collection is somewhat safer. This is not because brain
herniation is any less likely, but because the risk of direct damage to already-herniated tissue is not a factor with
lumbar collection. If CSF collection is necessary when intracranial pressure could be elevated (see page 43), then
Diagnostic aids
the animal should be ventilated prophylactically and a capnometer used to maintain end-tidal pCO2 at
3035 mmHg. Mannitol, lasix and dexamethasone should also be on hand for emergency use (see below).
Spinal needles are preferred for CSF collection (4.50). The CSF is collected into sterile vials; plain vials without
anticoagulant are generally used.
COLLECTION FROM THE CEREBELLOMEDULLARY CISTERN (4.504.58)

Because the neck is flexed severely (4.51), a kink-proof endotracheal tube should be used. Alternatively, the cuff
is deflated to allow space around the tube for ventilation should the tube obstruct.
An imaginary line is drawn between the wings of the atlas (4.52, 4.53). The ideal site is in the midline of the
patient, half way between the occipital protuberance and the line joining the wings of the atlas. Just behind the
occipital protuberance, there is a slight depression in the muscles. This is not the site of needle penetration but
is cranial to it; inserting the needle there usually leads to it striking the bone of the skull. The needle is inserted
perpendicular to the skin in the midline (4.54, 4.58). In small dogs and especially in cats, it is much safer to pick
up the skin and first penetrate the skin with the needle pointing away from the spinal cord.
There are two main methods for advancing the needle:
1.
Remove the stilette once the tip of the needle enters muscle and then advance the needle until CSF
appears in the hub of the needle.
2.
Advance the needle with the stilette still in place in small increments, removing the stilette between each
movement to check whether CSF is present in the needle. A slight pop can sometimes be felt when the
needle enters the subarachnoid space but this sensation can also be produced by movements of the
needle if the tip has been blunted on bone.
Method (1) is simpler for most clinicians and is less hazardous.

If CSF does not flow or is blood tinged, a number of possible complications could have arisen:
If the tip of the needle hits bone, it is most common for it to be too far rostral. The point of the needle is
redirected caudally by withdrawing it to a subcutaneous position and moving the hub rostrally.
If CSF flows, but is tinged with a trickle of bright red blood in the base of the hub, a dural vessel has been
penetrated. This may clear after a few seconds, and then the CSF may be collected. Rotating the needle helps
to clear this type of hemorrhage. The initial blood contamination can be evacuated from the hub of the needle
using a separate, small needle and syringe although it is often better to just remove the needle and start again.
Dark venous blood flows from the needle. This indicates that a venous structure has been penetrated,
usually because the needle has strayed from the midline. This causes no harm but it is best to start again
with a fresh needle.
No CSF flows, even though all the features of a successful tap appear to be present and the needle
appears to have penetrated to an adequate depth. The stilette should be replaced in case the needle has
plugged with tissue or blood clot. If CSF still does not flow, this may be because either the needle is far
off the midline; the brain has herniated; a mass occupies part of the cisterna magna; or the spinal cord
has been penetrated. The latter is a significant risk if the needle is advanced with the stilette in place.
The needle should be removed, the patient placed in a normal position and the respiratory pattern
observed. If respiration is normal after several minutes, the procedure may be repeated or a lumbar
puncture performed. In some patients with soft tissue lesions in the cranial cervical vertebral canal, CSF
cannot be collected from the CMC.
The patient moves suddenly; this is usually because the spinal cord has been damaged. The needle must
be removed and the patient ventilated. This is a potentially serious situation leading to marked neurological
deterioration or death. It may be useful to give the patient methylprednisolone sodium succinate (MPSS)
(30 mg/kg IV) in an attempt to limit the spinal cord damage.
The CSF flows forcefully from the needle. It is not always clear why this happens, but it clearly indicates
high CSF pressure at the CMC. Some dogs do not suffer any apparent deleterious effects from this, but
others do deteriorate. The needle should be removed and the precautions mentioned above taken. It may
65
66
also be useful to hyperventilate the patient as discussed below if increased intracranial pressure is
suspected as the cause.
CSF enters the hub but then flow stops abruptly and the animals heart rate drops below 40/min. This
indicates that brain herniation has occurred; the needle must be removed and the animal ventilated to
maintain end-tidal pCO2 at 3035 mmHg. Mannitol (1.0 g/kg) and furosemide should be administered
intravenously. This strategy is often successful in reversing the herniation if instituted immediately; a
capnometer and emergency drugs should therefore always be available and ready for immediate use for
any patient considered to be at increased risk of herniating.
4.50 Spinal needles have a stilette and a shallow

bevel. A notch in the hub indicates the side of
the bevel, which in myelography should be
pointing in the direction in which contrast is
intended to flow.
4.50
4.51 Patient positioned in right lateral recumbency for

collection of CSF by a right-handed operator.
The head is held by an assistant in 90 flexion
with the nose parallel to the tabletop. The neck
is positioned close to the edge of the table. A
foam wedge may be used to support the nose.
4.51
4.52 Diagram to show the site of CMC puncture.

Landmarks are the lateral margin of the wings
of the atlas (a), the occipital protuberance
(b), and the midline (c). The location of the
cisterna magna is outlined by contrast in 4.65.
4.52
4.53 Landmarks for CMC puncture. These are the

lateral margin of the wings of the atlas (a), the
occipital protuberance (b), and the midline.
b
a
4.53
Diagnostic aids
4.54 Needle insertion. The operator is wearing sterile

gloves, and the site has been prepared
aseptically. The left hand is identifying the
landmarks. The thumb and upper fingers are
palpating the wings of the atlas. The middle
finger is on the midline, just behind the intended
site. The 4th or 5th fingers of the right hand
should always be rested on the animals head in
case the table or dog move inadvertently.
4.54
4.55 The needle is in place with the stilette removed; a

drop of CSF is seen. Spinal needles become
plugged only rarely. If plugging is suspected, the
stilette should be inserted then removed.
4.55
4.56 Fluid is collected into a sterile vial.
4.56
4.57 Fluid can also be collected into a sterile syringe,

which is not used to aspirate CSF but just to
catch drops as they fall.
4.57
67
68
4.58 Diagram to show position of needle. The landmarks are as in

4.52 (a) and (b). Technique for injection of contrast at the end of
CSF collection is shown in 4.64.
4.58
COLLECTION FROM THE LUMBAR SPINE (4.594.63)

Collection of CSF from the lumbar site is more likely to fail than from the CMC. Contamination is also more
frequent. Collection is usually from L4/5 or L5/6 in dogs (4.59); in the cat L6/7 can be used.
If CSF does not flow or is blood tinged, a number of possible complications may have arisen:
The needle strikes bone. The space available for penetration of the vertebral canal is relatively small,
particularly in large dogs with significant vertebral new bone formation. Also, the ligamentous tissue may be
mineralized and have the feel of bone. Repeated attempts may be required to position the needle and, in
some patients, the technique may fail. Flexing the spine by drawing the pelvic limbs forward may open up
the interarcuate spaces.
The considerations for hemorrhage are similar to those mentioned above. If the needle appears to be
positioned correctly and yet CSF does not appear, the needle can be rotated slowly or withdrawn slightly.
Provided that intracranial pressure is normal, one or both jugular veins may be occluded to cause CSF to
flow. If these approaches are not successful then a second attempt at collection can be made. Aspiration
of CSF with a syringe is possible, but blood contamination is more likely.
4.59 Lumbar puncture between L5 and L6. The tip of

the L6 spinous process is palpated. The needle
is inserted alongside the caudal edge of the L6
spinous process and directed cranially and
ventrally through the ligamentum flavum into the
vertebral canal (4.67).
4.59
Diagnostic aids
4.60 Here the needle is in position alongside

the spinous process of L6; the patient is
in lateral recumbency. The technique can
also be done with the animal in ventral
recumbency, useful when contrast is to
be used for myelography.
4.60
4.61 The needle is advanced cranioventrally such

that the point reaches the midline at the
vertebral lamina. It may be necessary to step
the needle off the lamina into the ligamentum
flavum. This ligament can be tough and require
considerable force to penetrate it. The stilette is
removed intermittently to check progress. There
is often a twitch of the tail and pelvic limbs as
the needle touches or penetrates the neural
tissues. CSF may flow on penetrating the
subarachnoid space but more slowly than at
the CMC.
4.61
4.62 Needle in situ. Another method of lumbar

puncture is to find the cranial edge of the
spinous process and insert the needle vertically
into the ligamentum flavum.
4.62
4.63 Collection of CSF using a syringe.
4.63
69
70
Myelography
As a guide, a dose of iohexol contrast medium from 0.25 to 0.5 ml/kg body weight is used, although this can vary
with the site of the injection and the expected location of the lesion. Better contrast may be obtained when using
a concentration of 300 mg iodine/ml compared to 240 mg iodine/ml.
CERVICAL MYELOGRAPHY (4.64, 4.65)

If possible the table is tilted by 510 at injection. The bevel of the needle (indicated by a notch on the hub)
should be directed caudally. Injection should be performed preferably by connecting the syringe to the spinal
needle using a flexible tube, which is pre-filled with contrast before injection. Contrast should ideally be filtered if
multi-use vials are used routinely (4.1B). It is vital that the needle is not advanced into the spinal cord substance;
use of an extension tube can help to avoid needle movement during injection and the needle should be held as
described in 4.64.
After injection the needle is removed and the patients head is elevated. The legs and maxilla are tied to the
table edge using rope ties. Initially, the patient is tilted at 30 with the head up and an immediate radiograph is
taken, further radiographs being taken depending on the progress of the contrast column. If the X-ray tube can
be tilted parallel to the table, radiographs may be taken in this position. The angle of tilt may be increased to 45
or more and further radiographs taken until a lesion is demonstrated or the contrast ceases to flow. When the
lesion is delineated, ventrodorsal and oblique views should be taken.
If a tilting table is not available, the patient must be held up by the thoracic limbs, with the head above the body.
This can be difficult in large dogs and use of a tilting table is recommended.
In normal dogs, the column will usually reach the lumbosacral region in a maximum of 10 min on a suitably
tilted table. Occasionally, the column stops in a totally unexpected position, often the low cervical region, and this
should not be considered diagnostic until further time has elapsed and the table tilted to a steeper angle.
4.64 Position of patient and needle. In this patient, an extension tube

has been applied to the spinal needle for the injection of contrast
medium for myelography. The position of the right hand as
shown is dangerous and NOT recommended. It is much safer
to maintain the needle in position by resting the 4th and 5th
finger of the right hand on the dogs head as shown for CSF
collection (4.54).
4.64
Diagnostic aids
4.65 Close-up of a myelogram with the neck flexed to

show the atlanto-occipital space and the
cerebellomedullary cistern distended with
contrast (arrow). The cranial border of the atlas is
shown (arrowhead). The cerebellar folia are
outlined by contrast ( ); cranial flow of contrast
*
should be avoided by pointing the bevel caudally
while injecting and then elevating the head
promptly.
4.65
LUMBAR MYELOGRAPHY (4.664.68)

To expedite imaging and minimize the amount of repositioning, the order for radiographs should be as follows:
ventrodorsal survey radiographs; lateral survey radiographs, then injection of contrast with lateral images;
followed by a final repositioning for ventrodorsal and oblique images.
Ideally, injection should be made at the L5/L6 space as the incidence of complications increases with
injections at sites cranial to this (Kirberger and Wrigley, 1993). Injections at L6/L7 are prone to epidural injection,
although this tends to be true more in large dogs and L6/L7 can be more reliable than L5/L6 in small dogs (4.68).
Epidural leakage may also be more likely when the pelvic limbs are flexed. The bevel of the needle (indicated by
a notch on the hub) should be directed cranially. Injection may be performed directly by applying a syringe to the
spinal needle, or via a flexible tube, which is pre-filled with contrast before injection. A test injection of 0.20.4 ml
of contrast should always be made to insure that it is not being delivered into the spinal cord parenchyma (4.34).
Provided that this is not the case, a lateral radiograph is then taken once most of the contrast dose has been
injected but with the needle still in place. If the spinal cord is swollen then radiographs should ideally be taken as
contrast is being injected in order to best outline the subarachnoid space. In such cases it is useful to take the
dorsoventral radiograph by repositioning the animal with the needle still in situ to permit a further injection of
contrast to be made. If the needle has been inserted with the pelvic limbs flexed, they should be maintained in
this position as the animal is repositioned. Once the spinal needle has been removed, ventrodorsal and oblique
radiographs should be taken rapidly as epidural leakage occurs through the puncture hole, especially when the
spinal cord is markedly swollen. When epidural leakage does occur it tends to clear faster than the contrast in the
subarachnoid space and so serial images should be taken in the hope that resolution improves; if it does not then
the study will need to be repeated.
4.66 Contrast injection using an extension tube.

Injection should be slow to avoid epidural
leakage. The apparatus may need to be held in
place to prevent the extension tube and needle
from disconnecting.
4.66
71
72
4.67 Radiograph to show position of needle for

lumbar puncture. A myelogram has been
performed; contrast can be seen in the
subarachnoid space. Note that there is epidural
contrast (arrow) in the cranial lumbar vertebral
canal. This is a common complication in lumbar
myelograms and when present makes
interpretation much harder.
4.67
4.68 T2-weighted MRI of the lumbar region of a largebreed dog to show the high-signal epidural
(arrowheads) and subarachnoid spaces (arrow).
Note that the epidural space is usually much
wider relative to the subarachnoid space at L6/7
than at L5/6. The L4/5, L5/6 and L6/7 discs
show normal signal intensities but there is loss of
signal of the L7/S1 disc along with extension of
the disc dorsally compressing the nerve roots of
the cauda equina.
4.68
SPECIAL MYELOGRAPHIC CONSIDERATIONS

The simple technique of taking a radiograph with the animal lying on the contralateral side should also be
considered (4.33). Oblique projections may also be very useful in myelography; they are of most value from the
ventrodorsal position, although interpreting these projections can sometimes be challenging. A follow-up CT
scan is preferred after myelography when available, when clarification is required.
In some circumstances, special stressed positions may give more information about a lesion seen in a neutral
position, or they may reveal a lesion not apparent previously.
Cervical spine
In cervical spondylomyelopathy special positions during myelography may be useful:
The traction view, where tension is applied to the cervical spine. It can indicate whether the lesion is
dynamic or static, which can then have a bearing on surgical planning (11.4, 11.5).
The extension view, where the neck is extended dorsally, can reveal other lesions that may become significant
in the future; that is, the domino effect (11.23). This view is not without hazard and must be done with care
(11.9, 11.10).
Flexed and extended views may reveal lesions that cause minor trauma as the animal moves its neck
around (11.6).
Lumbosacral spine
In general a cervical injection is preferred to a lumbar injection to evaluate this region,

even though this necessitates that contrast be made to flow caudally by using gravity. The reason is because
after lumbar puncture there is a risk of either injecting into the lesion (2.25) or of epidural leakage (4.67). An alternative is to use an L1/2 injection site (McCartney,1997). Flexion and extension myelography may reveal lumbosacral lesions not apparent on neutral positions (10.7), but it is possible to get false-positive results with this
technique. Dynamic views may also be used during CT (10.11) and MRI (10.12) although the patient will then
need to be repositioned and a second scan conducted.
Instrumentation
Draping
Chapter
78
References
79
Spinal surgery requires some special instruments and

many of the items are illustrated in this chapter (5.1
5.35). However, the starting point is a well-equipped
general surgical pack, as would be used for most soft
5.1 Operating loupes. Some form of magnification is almost

essential for neurosurgery.
5.2 An operating microscope is extremely useful when

dissecting close to the spinal cord.
5.3 Headlightgood lighting is also essential for neurosurgery

and this is facilitated by a fiberoptic light source (Goring et al.,
1991), ideally one that attaches to the surgeons head and can
be flipped up or down as needed.
5.4 Anatomical specimens such as this canine spine are

invaluable for reference and for surgical planning.
74
5.5 Sew in, waterproof drape. See also 12.32.
5.8 Blunt self-retaining retractors are invaluable in the ventral

approach to the cervical spine. Either Gosset (illustrated) or
pediatric Balfour (9.18).
5.6 Gelpi self-retaining retractors. These are particularly

useful in all dorsal and dorsolateral approaches to the spine
(10.26, 13.60).
5.7 Multi-toothed self-retaining retractors. Illustrated here are

Weitlander and AdsonBaby (top) retractors. West retractors are
similar (9.20). These facilitate ventral cervical approaches.
5.9 Hand-held retractors. The small Hohmann retractor (above)

is useful in ventral repair of atlantoaxial subluxation. The Senn
(below) or Langenbeck retractor is useful in dorsal, dorsolateral
and lateral approaches to the spine (8.158.18).
5.10 A laminectomy spreader is very useful to reduce some

fractures and to distract over-ridden vertebrae (Boudrieau,
1997) (13.64).
Instrumentation
5.14 #7 scalpel handle with #11 blade. This is particularly

useful for disc fenestration and ligament removal (7.34, 7.45).
5.11 Electrosurgical instruments. The monopolar system is

used for coagulation and for incising tissues (8.14). Monopolar
cautery should not be used in close proximity to the spine, as
the current travels through the patient, and this can lead to
spinal cord damage. Close to the spinal cord, bipolar or
microbipolar cautery (5.12) must be used for coagulation
(8.24).
5.15 Instruments for fenestration and disc removal from the

vertebral canal. From top: Rosen mobilizer, Shea curette and
House curette (5.16). The latter is also very useful for removal
of bone during the final stages of laminectomy (8.37) and
ventral spinal decompression (7.51).
5.12 Microbipolar cautery should be used when working

close to neural tissues.
B
5.13 Periosteal elevator and freer, for removal of muscle from
the vertebrae (7.38) and for vertebral distraction (11.33).
5.16 Close up of House curette.
75
76
5.17 Nerve hook (below) for retraction of spinal nerves. Dental

tartar scrapers (blunt and pointed) for fenestration and removal
of disc material from the vertebral canal (5.18).
5.20 Small instruments for use with higher-powered
magnification.
5.21 These delicate rongeurs are used for fine bone removal.
Larger, double action rongeurs are available for heavier bone
removal (10.28).
5.18 Close up of dental instruments (8.35, 8.36, 8.39 and 8.41).
5.19 25 gauge hypodermic needles. These can be held at the

hub by needle holders. They can be used for cutting or the tip
may be twisted off using needle holders and then bent to size
as a blunt retractor or scraper. Blunt instruments made in this
way are particularly useful when working down a ventral slot,
especially in small dogs (11.55B).
5.22 Mini chuck for insertion of small pins, such as those used
for atlantoaxial subluxation (Chapter 9) and to stabilize articular
facets after spinal trauma (13.48A).
Instrumentation
5.23 Pneumatic systems are preferable for bone removal in

neurosurgical procedures. Illustrated here is the Hall
Surgairtome 2 with long bur guard, and round and oval burs
(8.30, 11.53).
5.24 Angled bur guard is useful for the cement plug technique
used in Wobbler surgery (11.38).
5.25 An electrical drill is an alternative to the pneumatic drill.

The Dremel model is an example; it is considerably less
expensive than the pneumatic system and does have
acceptable performance. Sterilization is a problem; ethylene
oxide systems can be used but their availability is restricted.
Methods of wrapping the instrument in sterile drapes are
available, but are less satisfactory, and we do not recommend
them in view of the requirement for asepsis in spinal surgery.
Disadvantages of an electric drill are that it is slower; more
pressure may be needed, which can cause bone necrosis
and possibly inadvertent injury to the spinal cord; and the
burs may clog more rapidly (Shires et al., 1986; Bitetto and
Kapatkin, 1989).
tissue and orthopedic procedures (Cockshutt, 2003).

With experience, surgeons find which instruments they
prefer for spinal surgery, but there is no doubt that certain instruments make the tasks easier and, thus, more
efficient.
5.26 Adson suction tip and bulb syringe for bur irrigation
(8.29, 8.30, 8.398.41).
5.27 Bone wax is used for hemostasis; it is used to plug small

vessels in bone by pressing over the bleeding site (8.31, 13.58).
5.28 Cellulose surgical spears (Ultracell sponges, Ultracell

medical Technologies Inc., North Stonington, CT.) are a very
useful type of sponge that does not shed material like standard
sponges. They are recommended to absorb fluid in contact with
the spinal cord.
5.29 Surgicel. If this product is used, it must always be

removed from the laminectomy site once hemostasis is
achieved as it swells and may cause compression. Surgifoam
hemostat, however, may be left in situ since it does not swell.
77
78
5.30 Gelfoam (Pharmacia, Kalamazoo, MI) absorbable gelatin

sponge is an alternative product for hemostasis that can be left
in place.
5.31 Hemoclips (Pilling Weck Inc., Research Triangle Park, NC)

are often the only way to control hemorrhage from largerdiameter veins in proximity to the intervertebral foramen (11.51).
B
5.33 Methylmethacrylate bone cement is used in several of the
procedures described in this book. It is worth emphasizing the
requirement for sterility when using these products (13.60).
Basic surgical techniques for neurosurgery share

most of the principles of soft tissue and orthopedic
surgery (Cockshutt, 2003; Shmon, 2003).
5.32 Positive profile pins (see Chapter 9 and also 13.48B);
small diameter pins are best inserted using a mini chuck (5.22).
These pins also have a roughened end opposite the threaded
end so that they do not need to be notched like smooth pins.
Small diameter negative profile pins can also be used as even
they have 46 times the pullout strength of equivalent diameter
smooth pins (Degernes et al., 1998; Sandman et al., 2001).
DRAPING
The patient should be completely covered by sterile,
waterproof surgical drapes. Their additional expense
must be weighed against the morbidity and the cost of
Instrumentation
5.34 Bone graft is used in several situations to promote vertebral

fusion. Illustrated is a curette and bowl for graft collection (11.24).
placed on top of the waterproof drapes prevent tissues

from drying out under surgical lights. They also help
self-retaining retractors maintain exposure of the surgical field (10.26, 12.36) and are less likely than standard
sponges to be left in the wound inadvertently. Tissues
should be irrigated regularly with sterile saline to prevent desiccation and to reduce airborne contamination.
After completion of surgery, the site must be inspected
to insure that nothing has been left in the wound.
Retractors are invaluable for providing adequate
exposure to perform the procedure. This can be supplemented by use of fine stay sutures, particularly
those applied to the dura (see 14.514.8).
REFERENCES
5.35 A Jamshidi needle is invaluable for taking a core biopsy

of a vertebral mass.
a surgical wound infection caused by organisms penetrating a porous drape.

Sew-in or clip-in drapes may be used to isolate the
wound further; ideally these should also be waterproof
(5.5). Moistened laparotomy sponges or towel drapes
Bitetto, W.V., Kapatkin, A.S. (1989) Intraoperative problems associated

with intervertebral disc disease. Problems in Veterinary Medicine 1,
434444.
Boudrieau, R.J. (1997) Distraction-stabilization using the ScovilleHaverfield self-retaining laminectomy retractors for repair of 2nd cervical vertebral fractures in 3 dogs. Veterinary and Comparative
Cockshutt, J. (2003) Principles of surgical asepsis. In: D. Slatter (ed.),
Textbook of Small Animal Surgery, 3rd edn, 149154. Philadelphia: WB
Saunders.
Degernes, L.A., Roe, S.C., Abrams, C.F., Jr (1998) Holding power of different pin designs and pin insertion methods in avian cortical bone.
Veterinary Surgery 27, 301306.
Goring, R.L., Beale, B.S., Faulkner, R.F. (1991) The inverted cone decompression technique: a surgical treatment for cervical vertebral instability
Wobbler syndrome in Doberman Pinschers. Part 1. Journal of the
Sandman, K.M., Smith, C.W., Harari, J., Manfra Maretta, S., Pijanowski,
G.J. (2001) Comparison of pull-out resistance of Kirschner wires and
Imex miniature interface fixation pins in polyurethane foam. Veterinary
and Comparative Orthopaedics and Traumatology 15, 1822.
Shires, P.K., Roberts, E.D., Hulse, D.A., Zeman, D.H., Kearney, M.T.
(1986) Hemilaminectomy using an autoclavable electrical drill compared to a pneumatic drill. Journal of the American Animal Hospital
Shmon, C. (2003) Assessment and preparation of the surgical patient and
operating team. In: D. Slatter (ed.), Textbook of Small Animal Surgery,
3rd edn, 162178. Philadelphia: WB Saunders.
79
Preoperative assessment
Pharmacological considerations 83
Antibiotics 83
Non-steroidal anti-inflammatories 83
Corticosteroids 83
Preoperative analgesia 84
Adverse drug reactions and interactions
85
85
References
Skin or periodontal disease
Cushings disease
Orthopedic disease
Hypothyroidism
Cardiac disease
Diabetes mellitus
Hepatic disease
Disorders of hemostasis
Renal disease
Urinary tract infection
Neoplasia
Metastatic disease
87
Prostatic disease
Pyometra
(Box 6.1). This is particularly true for older patients,

following trauma (Box 13.2) and dogs with cervical
spondylomyelopathy (see Chapter 11, page 213).
Concurrent disease
88
Undetected primary
Surgical considerations 86
Laminectomy healing 86
Durotomy 87
Myelotomy 87
Mechanisms of recovery after spinal cord injury
Client communication
Box 6.1 Conditions that may complicate the

management of a neurological patient
Clinical assessment 81
Concurrent disease 81
Hemostasis 82
Hypoventilation 82
Anesthetic considerations
Premedication 85
Induction 85
Maintenance 85
Recovery 85
Complications 86
Chapter
88
88
Accurate preoperative assessment of the neurosurgical

patient is extremely important, particularly in view of the
time, expense and complexity of the procedures performed. Anticipation of the most likely complications
may lessen their impact or even allow them to be avoided
altogether.
CLINICAL ASSESSMENT
It is easy to overlook concurrent clinical or subclinical
disorders that may have an important bearing on the case
The physical examination must be thorough. Skin disorders may necessitate delaying elective surgery because
pyoderma, clipper rash, or povidoneiodine reactions
may predispose to wound infection. Fleas should be
eliminated to prevent them from entering the surgical
field. Osteoarthritic joints may cause major interference
with rehabilitation from neurological deficits, especially
for large or obese dogs.
Laboratory evaluation should include a hematological
and biochemical profile, along with a urinalysis (see
Chapter 4). In cats, the feline leukemia virus (FeLV)
and feline immunodeficiency virus (FIV) status should
be determined. When positive, the likelihood of lymphoma as a cause for neurological deficits increases
markedly. If there is suspicion of cardiac dysfunction in
large- or giant-breed dogs, suitable investigations should
82
be performed, such as an electrocardiogram (ECG) and

echocardiogram (Calvert and Wall, 2001). Chest radiographs are essential in these patients:
For older animals (7 years of age).
When neoplasia is a differential diagnosis.
For any recumbent, tetraparetic patient.
Following trauma.
Similarly, an ultrasound or CT scan of the abdomen
is useful to identify intercurrent disease in the first two
categories, above. Endocrine disorders may predispose a
patient to complications, for example, urinary tract infections (UTI) and delayed wound healing in Cushings disease and diabetes mellitus or pathological fractures in
Cushings disease (Gehlbach et al., 2000; Hosgood, 2003).
Neuropathies or myopathies associated with hypothyroidism may complicate the patients neurological
status. The assessment and, if possible, stabilization of
endocrine disorders is always recommended prior to any
neurosurgical procedure.
Hemostasis
Hemostasis is divided into primary and secondary events
(Kerwin and Maudlin, 2003). Primary hemostasis
depends on platelet aggregation and adhesion, and usually causes most concern to the neurosurgeon. It may be
disturbed in the following conditions:
von Willebrand disease (see Chapter 11, page 217).
Severe thrombocytopenia
(platelet count 20 000/l).
Azotemia (serum creatinine 5.6 mg/dl).
Non-steroidal anti-inflammatory drug (NSAID)induced platelet dysfunction, including aspirin at
all therapeutic doses and especially above 25 mg/kg.
Anticoagulant intoxication.
Disorders of primary hemostasis should be identified
by a platelet count and bleeding time evaluation (11.11).
Perturbed hemostasis can cause decreased visualization
during surgery, postoperative bruising (15.2), hematoma
formation and other more serious complications (15.40).
The dog shown in 6.1, represented 4 months after surgery with a packed cell volume (PCV) of 12% from a
bleeding gastric ulcer but survived after a transfusion.
The owner had treated the dog with aspirin at 40 mg/kg
for 2 weeks.
Hypoventilation
This is a serious potential complication that can occur in
any animal with a serious cervical spinal cord injury.
Patients that are severely tetraparetic on initial presentation should be evaluated for hypoventilation using blood
gas analysis or a capnometer. Tetraplegic animals are
at even higher risk. Occasionally an animal may need
urgent ventilatory assistance prior to surgery in order to

survive long enough to benefit from definitive treatment
(Boudrieau, 1997). Any severely tetraparetic animal can
also develop this complication in the postoperative
period and the owner must be appraised of this risk prior
to surgery (6.1, 7.11).
Hypoventilation after spinal cord injury can occur
due to three main mechanisms:
Hemorrhage or edema affecting the respiratory
centers in the medulla and C1 spinal cord, which
lead to decreased respiratory drive.
Cervical myelopathy severe enough to interrupt
conduction along all motor fibers in the spinal cord,
which causes near-paralysis of the respiratory
muscles (see pages 28, 216).
Diaphragmatic paralysis due to a lesion of the
phrenic LMNs (C5 spinal cord segment or nerve
roots); this is the least important mechanism unless
combined with an injury to the white matter at that
level (Smith and Walter, 1985; Blass et al., 1988;
Beal et al., 2001).
The end result of these injuries is either that there is
decreased respiratory drive or there is insufficient communication between the respiratory center and the respiratory muscles (Beal et al., 2001). Apnea can also
occur following the injection of subarachnoid contrast
at the cerebello-medullary cistern (CMC) (Widmer
et al., 1992a), or after manipulation during imaging (Seim
and Prata, 1982) (11.10).
6.1 Myelogram from a 6-month-old Great Dane with sudden

onset tetraplegia and dyspnea. It had neck pain with deficits
localized to the C1C5 spinal cord. Blood gas analysis showed
a paCO2 of 45 mmHg (N3545). Ventral slot at C3/4 retrieved
a large amount of nucleus pulposus from the vertebral canal.
On extubation the dog did not breathe spontaneously; its
paCO2 was 91 mmHg. It was ventilated overnight but attempts
to wean it from the ventilator were unsuccessful until 4 days
later. It was able to walk 2 weeks after surgery and made a
good recovery (see Hemostasis).
PHARMACOLOGICAL
CONSIDERATIONS
Antibiotics
(Reed, 2002), or given together with corticosteroids, as

potentially severe gastrointestinal erosions or ulcerations
are likely.
Most neurosurgical procedures can be classified as clean

and uncontaminated. However, prophylactic antibiotics
are indicated if sterility is broken or if one or more of
the factors in Box 6.2 apply.
In such cases, perioperative cefazolin (20 mg/kg IV) is
recommended for its good tissue penetration and broad
spectrum of activity against staphylococci and other
Gram-positive organisms (Dunning, 2003). A short, decisive window for prophylactic antibiotic use extends from
the start of surgery to a maximum of 3 h afterwards. It is
at this time that bacterial contamination of tissue can be
suppressed by antimicrobial therapy. There is no advantage to using IV antibiotics before the start of surgery, or
for continuing them beyond its completion, except in
concomitant diseases such as pyoderma or UTI (Rosin,
1988). A penicillin-derivative or a cephalosporin is usually a suitable initial choice for retention cystitis. Final
antibiotic selection for UTI should be based on urine
culture and sensitivity whenever possible, especially
when the animal has been hospitalized and a nosocomial
infection is much more likely (Dunning, 2003).
Corticosteroids
Box 6.2 Some factors that can predispose to

wound infection
Dermatitis
Periodontal disease
Urinary tract infection
Obesity
Shock or sepsis
Use of a surgical implant
Cushings disease
Corticosteroid or non-steroidal anti-inflammatory drug

(NSAID) use
Diabetes mellitus
Surgical time over 90 min
Excessive use of electrocautery
Non-steroidal anti-inflammatories
These drugs are valuable analgesics although they are
often less effective than corticosteroids as specific
anti-inflammatory agents for neurological disease. They
can have a number of important adverse effects. Such
effects are most marked in drugs that inhibit mainly
cyclooxygenase-1 (COX-1) as opposed to COX-2
(Kay-Mugford et al., 2000) (see Chapter 15, page 341).
NSAIDs must not be used at more than the recommended dose, used in combination with other NSAIDs
Corticosteroids are used widely in all types of spinal cord

conditions, mainly because they are so effective at
reducing the associated inflammation and at relieving
pain. However, there is a widespread misunderstanding
that these beneficial effects also apply to the neural
injury itself. While corticosteroids may reduce inflammation associated with, for example, extruded disc
material, they can also prejudice the survival of any
injured neurons by interfering with their glucose metabolism (Sapolsky, 1994; Smith-Swintosky et al., 1996).
This may be of little consequence with less severe spinal
cord lesions but it could be critical in severe injuries. In
such cases the animal might feel better but may have a
decreased chance for neurological recovery. In addition,
there is almost certainly no neuroprotective benefit
provided by any corticosteroid with the exception of
methylprednisolone sodium succinate (MPSS) (Heary
et al., 1997; Olby, 1999; Hurlbert, 2000; Bracken and
Holford, 2002). Corticosteroids can also precipitate gastrointestinal bleeding in as many as 15% of neurosurgical
patients, with mortality rates of up to 2%. Dexamethasone is most likely to cause problems and has no role in
the management of spinal trauma (Moore and Withrow,
1982). Standard gastrointestinal protectant agents may
not be effective in preventing corticosteroid-induced
side-effects (Hanson et al., 1997). Duodenal or colonic
perforation are the most serious potential complications
(Toombs et al., 1986; Hinton et al., 2002). Routine glucocorticoid therapy in spinal patients is strongly discouraged unless these drugs are used for short periods at
anti-inflammatory doses in animals with mild neurological deficits (LeCouteur and Sturgess, 2003).
METHYLPREDNISOLONE SODIUM
SUCCINATE (MPSS)
MPSS has received wide interest in human, and, to a
lesser extent, veterinary medicine in the past 10 years
because of proposed benefits in spinal cord injury. After
acute injury, the blood supply to the spinal cord is progressively reduced. When the injured tissue is reperfused,
massive amounts of highly reactive chemicals called free
radicals are liberated. These free radicals are especially
damaging to the plasma membrane of cells via a process
called lipid peroxidation. Free radical-induced lipid peroxidation is now recognized as a key pathophysiological
mechanism for irreversible tissue loss following spinal
cord trauma and ischemia (Brown and Hall, 1992).
83
84
The neuroprotective effect of MPSS is exerted by its

actions as a free-radical scavenger. These benefits are not
due to its glucocorticoid activity and only occur in doses
far exceeding those that saturate all glucocorticoid
receptors. The optimal neuroprotective dose of MPSS
has been determined to be 30 mg/kg, whereas doses of
60 mg/kg were detrimental and doses of 15 mg/kg had
no effect. Benefit has only been observed in humans
with spinal cord injury who received treatment within
8 h of injury (Bracken et al., 1990, 1992).
In animals, the suggested dosage regime is an initial
IV bolus of 30 mg/kg MPSS, followed by 15 mg/kg IV
2 and 6 h later, then 2.5 mg/kg IV per hour for a further
24 h (Brown and Hall, 1992). The bolus doses should
be given slowly to avoid vomiting and hypotension. The
continuous IV dose is not widely used in small animal
patients; humans are given an initial bolus of 30 mg/kg,
followed by an infusion of 5.4 mg/kg/h for 24 h
(Bracken et al., 1990).
A similar regime might also provide some benefit when
given prior to spinal cord decompression for lesions that
have been present for longer than 8 h. However, this
hypothesis has not yet been tested and might even have
an adverse effect (Olby, 1999; Bracken, 2000a,b, 2001).
Vitamin E also has protective effects on the spinal cord
when given at 10002000 IU per animal per day for
5 days and is an alternative prior to elective surgery
(Fehlings et al., 1989; Olby, 1999).
No serious side-effects were reported in one study
using high dose MPSS therapy in 86 dogs with thoracolumbar disc disease (Siemering and Vroman, 1992). In
another study, 35 of 108 dogs developed complications
such as diarrhea or melena but none were considered
serious and they usually resolved without therapy
(Culbert et al., 1998). However, endoscopy revealed
severe, subclinical gastric hemorrhage in 90% of dogs
after MPSS treatment (Rohrer et al., 1999a,b); and 90%
of dogs treated with MPSS prior to spinal surgery get
occult gastrointestinal bleeding, although it is rarely
severe enough to warrant intervention (Hanson et al.,
1997; Rohrer et al., 1999a). Caution is advisable if considering MPSS therapy after treatment with other types
of corticosteroid or with NSAIDs. Gastrointestinal barrier
disruption and bloody diarrhea can lead to bacteremia,
which is undesirable in a surgical patient (Epstein et al.,
1992). Complications of MPSS reported in humans
include pneumonia, sepsis, immunosuppression and pancreatitis (Levy et al., 1996; Bracken et al., 1997; Gerndt
et al., 1997; Matsumoto et al., 2001).
A number of studies have now raised serious questions
about the value of MPSS in humans (George et al.,
1995; Gerhart et al., 1995; Levy et al., 1996; Heary et al.,
1997; Nesathurai, 1998; Hurlbert, 2000; LeCouteur and
Sturgess, 2003). The benefits reported originally were

mainly to upper body function and not a regaining of the
ability to walk (Bracken et al., 1990, 1992, 1997). These
relatively small changes are crucial in people but are of
much less significance in animals. They may also represent an improvement in gray matter function whereas
white matter survival is much more important in
animals (Jeffery and Blakemore, 1999a). Furthermore,
the statistical basis of the original human clinical trials has
now been questioned (Hurlbert, 2000). Although highdose MPSS therapy may be an advance in the management of acute spinal cord injury, it is not a panacea.
Administration after the 8-h therapeutic window worsens the outcome in humans while the length of this window in dogs and cats has not been determined (Bracken,
2000a,b, 2001; Hurlbert, 2000). The deleterious effect
of delayed administration of MPSS is probably due to
interference with neuronal glucose metabolism (Sapolsky,
1994; Smith-Swintosky et al., 1996; LeCouteur and
Sturgess, 2003). Therapy with MPSS must therefore be
looked upon at best as a way to complement, but not
replace, current veterinary neurosurgical techniques
(LeCouteur and Sturgess, 2003).
Preoperative analgesia
Fentanyl patches provide a useful way to deliver preoperative analgesia and they can also be combined with
one of the analgesics discussed under postoperative care
(see Chapter 15), however the rate of absorption of fentanyl can vary considerably in an individual animal over
time and can also vary at different times in the same animal
(Kyles et al., 1996; Egger et al., 1998). The dose range is
25 g/kg/h. Increasing the patch size did not increase
the plasma fentanyl concentration in one study but did
in a second (Egger et al., 1998; Welch et al., 2002a). The
analgesic effect provided by fentanyl in dogs is equivalent
to that provided by intramuscular oxymorphone and is
often superior to epidural morphine; in cats it is superior
to butorphanol (Kyles et al., 1998; Robinson et al., 1999;
Franks et al., 2000). The onset of action is faster in cats
(26 h) than in dogs (2436 h) (Scherk Nixon, 1996;
Robinson et al., 1999). The patch should therefore be
placed 24 h prior to surgery in dogs and supplemental
analgesia is often required during this period (Egger
et al., 1998). Concentrations fall rapidly after patch
removal and are often below therapeutic concentrations
within 1 h (Egger et al., 1998).
Respiratory depression can be a serious side-effect in
humans but was not seen in dogs at high doses
(5 g/kg/h) following thoracotomy (Welch et al., 2002a).
Caution should be used in dogs that are hypoventilating
due to head or spinal cord injury (see Chapter 15).
Other potential side-effects in dogs include bradycardia,
dysphoria and vomiting; acepromazine and glycopyrrolate may alleviate these effects. Skin reactions may also
occur (15.1). The patch must not be placed on a heating
pad as this will increase the rate of absorption from the
patch (Egger et al., 1998). Fentanyl is a useful way to
provide non-invasive, inexpensive, long-lasting analgesia
that is tolerated well (Scherk Nixon, 1996; Kyles, 1998).
However, careful monitoring is needed to insure adequate analgesia and minimize adverse effects (Scherk
Nixon, 1996; Kyles, 1998) (see also Postoperative analgesia, page 339).
Adverse drug reactions and

interactions
Adverse effects of drugs are reported only rarely in animals but are likely to become a more serious problem,
especially with the increasing awareness of such events
in humans (Stillman, 1989; Weinblatt, 1989). One
potential adverse effect is a drug interaction, for example renal failure as has been reported in people using
NSAIDs and angiotensin-converting enzyme (ACE)
inhibitors together (Seelig et al., 1990). Cyclosporin can
cause adverse interactions in dogs when given together
with either ketoconazole or with ivermectin (Myre et al.,
1991; Roulet et al., 2003). A second type of adverse
effect is an unexpected reaction to an individual drug,
for example enrofloxacin as a (dose-related) cause of
blindness in cats or metronidazole as a cause of central
vestibular disease (Dow et al., 1989; Gelatt et al., 2001).
Hepatotoxicosis, keratoconjunctivitis and aplastic
anemia have been reported in dogs given trimethoprimsulfonamide combinations (Diehl and Roberts, 1991;
Rowland et al., 1992; Fox et al., 1993). Anaphylactictype reactions can occur in dogs given cefalosporins
during surgery; to avoid this complication occurring
intraoperatively it is preferable to give the first dose
along with the premedication where possible
(Anderson and Adkinson, 1987; Grouhi et al., 1999).
Adverse drug events have been reviewed extensively in
humans and in animals (Boothe, 1990; Scott and Miller,
1998, 1999; Ament et al., 2000; Maddison et al., 2000;
Papadogiannakis, 2000).
ANESTHETIC CONSIDERATIONS
Many neurosurgical patients will be dehydrated at presentation, usually when pain or weakness reduce fluid
intake. They must be rehydrated adequately prior to
anesthetic induction, because of the detrimental effect
of hypotension on spinal cord perfusion (Tator and
Fehlings, 1991; Nuwer, 1999). Another way to reduce
dehydration is to increase preoperative fluid intake; fasting recommendations for humans now permit water up
until 2 h prior to anesthesia for this reason (Crenshaw

and Winslow, 2002).
Premedication
Premedication should have a calming effect and relieve
pain. Hypotension and loss of protective muscle tone
should be avoided in an animal with severe neurological
deficits or an unstable vertebral column. Glycopyrrolate
is the preferred anticholinergic as it is less likely to
induce tachycardia; atropine may still be required to treat
severe bradyarrhythmias (Stauffer et al., 1988).
Induction
A laryngoscope should be used for intubation with as little movement of the spine as possible, especially in dogs
with unstable lesions or lesions in the cervical area. An
armoured endotracheal tube is recommended if cerebrospinal fluid (CSF) is to be taken from the cerebellomedullary cistern (CMC), if stress radiographs are to be
taken, or if a ventral approach to the neck is to be used.
Maintenance
Isofluorane is the usual inhalation agent of choice,
because it is both less depressant to the cardiovascular
system and less arrhythmogenic than halothane. Methoxyfluorane provides good muscle relaxation and better
postoperative analgesia than isofluorane, but is contraindicated in animals receiving opioids or NSAIDs, or in
those with renal disease. Sevofluorane is a good inhalational agent in dogs and may be superior to isofluorane
(Branson et al., 2001). Autoregulation in the brain is
preserved better with sevofluorane than with isofluorane
(Summors et al., 1999; Endoh et al., 2001); both agents
maintain good spinal cord blood flow (Hoffman et al.,
1991; Crawford et al., 1992).
Maintenance of anesthesia should be at a depth sufficient to prevent movement of the patient, which could
be dangerous during surgery. Mechanical ventilation is
recommended in large or debilitated animals, those that
are positioned in dorsal recumbency and those with
neurological deficits cranial to the thoracolumbar junction. Barotrauma must be avoided, especially in small
animals (Manning and Brunson, 1994; Parent et al.,
1996). Nitrous oxide is not recommended as it may
result in a rapid onset of hypoxia if the patient hypoventilates during CSF collection or if it will be disconnected
temporarily from the gas supply during radiography.
Recovery
The recovery from anesthesia should be smooth, and
this may be helped by the use of narcotics, diazepam, or
both. If the animal has undergone myelography, its head
85
86
must be kept elevated during the entire recovery period.

Seizures should be treated promptly with diazepam
(0.4 mg/kg IV), repeated as necessary. The use of iohexol
for myelography means that the risk of seizures is low,
although Doberman pinschers and other large-breed dogs
with cervical spondylomyelopathy may be at increased
risk and warrant careful monitoring (Lewis and Hosgood,
1992). It is prudent to leave an IV catheter in place until
the patient is fully recovered.
Complications
Anesthesia must optimize both cardiovascular and
respiratory function in order to minimize ischemia of the
spinal cord. Spinal cord blood flow is autoregulated in a
manner analogous to cerebral blood flow; volatile anesthetic agents may depress this autoregulation and trauma
to the spinal cord can abolish it altogether (Tator, 1991;
Nuwer, 1999; Olby and Jeffery, 2003). Any cardiac
arrhythmia or systemic hypotension will compound this by
reducing spinal cord blood flow further and so should be
avoided if at all possible.
Iohexol is a safe agent for myelography (Wheeler and
Davies, 1985), although the incidence of seizures can be
around 7% in large-breed dogs with Wobbler syndome
(Lewis and Hosgood, 1992; Widmer et al., 1992b).
Seizures can cause the animal to extend its neck, which
might have an adverse effect on a dynamic cervical lesion
(11.10). It is recommended that the patient be at an adequate depth of anesthesia before the injection is made,
that the contrast be used at body temperature, and that
it is injected at a steady rate (Lamb, 1994). In one study,
six out of 66 dogs developed bradycardia during iohexol
injection and one died; two also developed bradycardia
during recovery (Lewis and Hosgood, 1992). The ECG
must be monitored closely during the study and any problems treated promptly. Hypothermia is another potential
problem, particularly in small animals or if the anesthetic
period is prolonged.
Surgical techniques can also affect the patient during
anesthesia. Cervical spinal surgery is associated with a
much higher risk of arrhythmias and ventricular premature contractions than thoracolumbar surgery (Stauffer
et al., 1988). This may result from manipulation of either
the spinal cord itself, or of nerves in the ventral neck. Both
positioning the spine in extension and manipulation of the
vagosympathetic trunk should be minimized. A paramedian approach to the neck provides better protection for
vital structures and may therefore reduce the incidence of
intraoperative arrhythmias (11.2511.27). Magnification
is recommended when performing surgery on the cervical
spinal cord in order to reduce unnecessary manipulation
(see Chapters 7 and 11). Severe cervical myelopathies can
result in sympathetic blockade (11.10), which can be fatal
(Rosenbluth and Meirowsky, 1953; Seim and Prata, 1982;

Clark, 1986). Surgery in the cranial and mid-thoracic
spine can also impair autonomic control of cardiovascular
function and so the patient must again be monitored
closely.
SURGICAL CONSIDERATIONS
Laminectomy healing
A laminectomy heals as the hematoma forms a fibrous
callus, which then undergoes metaplasia to cartilage and
bone (Trotter et al., 1988). Adhesions can involve the
dura and nerve root(s) if no attempt is made to protect
them after laminectomy (Cook et al., 1994). Adhesions
around local nerve roots can cause considerable postoperative morbidity in humans, probably by causing a
tethering effect (Songer et al., 1990; Geisler, 1999).
This has not been identified specifically in animals
although those with postoperative back pain and evidence of probable peridural fibrosis on CT scan have
been reported (Olby et al., 2000). The preferred way to
assess epidural scarring is by using MRI (Ross et al.,
1999). The more scar tissue that is evident on MRI, the
more pain is reported in humans after lumbar discectomy
(Ross et al., 1996; Maroon et al., 1999). It is advisable to
keep the exposed dura mater separate from the damaged
epaxial muscles during healing in order to minimize the
development of adhesions, which can even cause subsequent spinal cord compression (Trotter et al., 1988; Cook
et al., 1994).
Several implants have been used to minimize adhesions but none is ideal (Cook et al., 1994). One study
using Gelfoam (Pharmacia, Kalamazoo, MI) showed no
reduction of peridural scarring in dogs; two more studies
showed that scarring was actually increased (Gill et al.,
1979; Songer et al., 1990; Robertson et al., 1993) and a
fourth study showed reduction in scar using Gelfoam
(LaRocca and Macnab, 1974).
Better overall results seem to be obtained in dogs using
autogenous fat (Gill et al., 1979; Cook et al., 1994).
Histopathological evaluation of free fat grafts after
laminectomy for disc disease in 21 dogs showed that
5090% of the graft was made up of fat with no scar. In
a further eight dogs, myelography was used to verify the
lack of scar formation in the subarachnoid space over
the surgical site. Overall follow-up times ranged from
1 month to 5 years (Biggart, 1988). CT scans can also
be used to verify the presence and viability of fat graft at
a previous laminectomy site (Biggart, 1988; Olby
et al., 2000). However, two other studies found either
no benefit from free fat grafts (Songer et al., 1995), or
that grafts do not prevent dural adhesions (Trevor et al.,
1991). Free fat grafts must revascularize and so should be
no more than 35 mm thick in order to minimize the risk

of aseptic necrosis (8.8, 8.54, 12.10). An initial inflammatory phase resolves as fibrosis increases over 816
weeks. By this time the graft has contracted to roughly
half its original size (Trevor et al., 1991; Cook et al.,
1994). One study has shown that pedicle grafts in dogs
have no advantage over free fat grafts; a second study
showed pedicle grafts to be superior (Gill et al., 1979;
Trevor et al., 1991). This discrepancy could simply
reflect the small number of dogs in the studies. A commercially available alternative is ADCON-L (Gliatech
Inc., Cleveland, OH), a novel, porcine-derived, polyglycan
implant that blocks in-growth of fibroblasts. It was shown
to be effective in dogs (Einhaus et al., 1997; BenDebba
et al., 1999; Geisler, 1999); it can even be used to deliver
morphine locally to the surgical site (Mastronardi et al.,
2002). There is still some controversy about its value and
price may be an issue for veterinary use (Richter et al.,
2001). A resorbable, polymer barrier film (Hydrosorb
TS, Macropore Biosurgery Inc., San Diego, CA) has
proven to be superior to ADCON-L in rats and was also
moderately effective in dogs compared to controls
(Welch et al., 2002b).
Durotomy
Durotomy, or incision of the dura mater, results in mild,
temporary deficits in normal dogs (Parker and Smith,
1972). As the dural incision heals, the edges rejoin but
they often adhere to the spinal cord as well (Trevor et al.,
1991). Durotomy is necessary to evaluate intradural
and intramedullary lesions (Jeffery and Phillips, 1995)
(12.40). It is sometimes performed at laminectomy in
an attempt to decompress the spinal cord further (8.49,
8.50). Although this may improve decompression, the
relative risks and benefits are unclear (Perkins and Deane,
1988). In experimental situations, durotomy may have
value when performed immediately after trauma, but
was found to have no benefit 2 h after the injury (Parker
and Smith, 1974, 1975). However, studies of intracranial pressure show that craniotomy and durotomy
lower pressure by 15 and 65%, respectively (Bagley
et al., 1996). This suggests that durotomy might also produce a significant decompressive benefit for spinal cord.
Durotomy is not recommended in order to provide prognostic information after spinal cord injury and euthanasia
should not be based solely on the gross appearance of the
spinal cord (Salisbury and Cook, 1988) (8.50).
Myelotomy
Myelotomy involves incising the spinal cord on the dorsal
midline to the level of the central canal. It has been performed in normal dogs at the T3L3 region with minimal
residual neurological sequelae (Teague and Brasmer,

1978). It has also been used with good results to aid
removal of a spinal cord hematoma and a tumor
(Martin et al., 1986; Jeffery and Phillips, 1995). No
therapeutic value has been demonstrated in acute spinal
trauma and the technique is difficult to perform in cats
without causing additional injury (Hoerlein et al., 1985).
Mechanisms of recovery after

spinal cord injury
Recovery occurs mainly by the re-establishment of a
normal spinal cord microenvironment and forming new
patterns of central nervous system (CNS) circuitry
(Jeffery and Blakemore, 1999b). Nociception is normally regained first then motor function and continence
followed by proprioception (see Chapter 2). Interestingly,
some animals that fail to regain deep pain and continence
still recover the ability to walk (13.30, 13.34). Although
often termed spinal reflex walking, this type of recovery
probably requires some input from higher centers
mediated through a few intact axons surviving across
the lesion. Survival of small groups of axons can occur
because many severe spinal cord injuries are centered on
the gray matter and there is relative sparing of peripheral
white matter (Griffiths, 1978; Olby et al., 2003) (6.2).
Evidence of higher input in dogs that walk without
recovering deep pain is inferred from the fact that many
regain a voluntary tail wag (Olby et al., 2003). However,
these dogs nearly always remain incontinent and the
recovery of walking ability takes at least 4 months and
usually much longer (Olby et al., 2003).
6.2 Thoracic spinal cord from a dog that did not regain deep
pain sensation after a spinal fracture 6 weeks previously. There
is a large area of central necrosis with small groups of surviving
axons in the periphery (arrows). These axons may mediate late
recovery of motor function in dogs that never regain deep pain
sensation and that remain incontinent (Olby et al., 2002).
87
88
CLIENT COMMUNICATION
The client should always be offered the textbook
approach for managing a patient in terms of the diagnostic
evaluation and treatment. This should be offered regardless of the clinicians perception of whether the client will
pursue treatment, and even when it involves referral.
In this manner, the client is made aware that the very best
is available for their animal, and any compromises arrived
at will be seen in the proper perspective.
Whenever possible, the clinical problems and the diagnostic and therapeutic options should be explained to the
client in non-scientific terms. Anatomical specimens, websites, drawings and the animals own laboratory data and
images should be used to illustrate the situation. On the
other hand, care must be taken not to overload the client
with excessive information, and they should be given time
and the opportunity to think privately about their decision. It is also important that the client be given as accurate
a prognosis as possible, both for the degree of recovery and
the time required. Giving the client prior knowledge of
the probable prognosis and of the most likely complications is invaluable should the patient suffer either a protracted recovery or setbacks in its postoperative course.
The client should never be pressured into consenting
to a course with which they are not comfortable, even if
the clinician feels that it is in the animals best interests.
If complications do arise the client may then hold the
clinician responsible, not only for the decision but also
for the complications.
Finally, the client should be given, and asked to sign, a
consent form and a written estimate of the likely cost
for the entire procedure. It is vital that the client be
made aware that there is some risk even in a routine general anesthetic and scan. This estimate should include a
price range to take into account any foreseeable variations in the postoperative course. Should the bill begin to
approach the upper end of the estimate, or if unforeseen
complications develop, the client must be notified
immediately (Thacher, 1989).
1. Does MPSS provide significant neuroprotection in dogs
after spinal cord trauma?
2. Does MPSS provide benefit when given prophylactically
prior to surgery?
3. What are the relative merits of MPSS and Vitamin E?
REFERENCES
Ament, P.W., Bertolino, J.G., Liszewski, J.L. (2000) Clinically significant
drug interactions. American Family Physician 61, 17451754.
Anderson, J.A., Adkinson, N.F., Jr (1987) Allergic reactions to drugs and

biologic agents. Journal of the American Medical Association 258,
28912899.
Bagley, R.S., Harrington, M.L., Pluhar, G.E., Keegan, R.D., Greene, S.A.,
Moore, M.P., Gavin, P.R. (1996) Effect of craniectomy/durotomy alone
and in combination with hyperventilation, diuretics, and corticosteroids
on intracranial pressure in clinically normal dogs. American Journal of
Beal, M.W., Paglia, D.T., Griffin, G.M., Hughes, D., King, L.G. (2001)
Ventilatory failure, ventilator management, and outcome in dogs with
cervical spinal disorders: 14 cases (19911999). Journal of the American
BenDebba, M., Augustus van Alphen, H., Long, D.M. (1999) Association
between peridural scar and activity-related pain after lumbar discectomy. Neurological Research 21 Suppl 1, S3742.
Biggart, J.F., III (1988) Prevention of laminectomy membrane by free fat
grafts after laminectomy in dogs with disk herniations. Veterinary
Surgery 17, 2829.
Blass, C.E., Waldron, D.R., Van Ee, R.T. (1988) Cervical stabilization in
three dogs using Steinmann pins and methylmethacrylate. Journal of the
Boothe, D.M. (1990) Drug therapy in cats: mechanisms and avoidance
of adverse drug reactions. Journal of the American Veterinary Medical
Boudrieau, R.J. (1997) Distractionstabilization using the ScovilleHaverfield self-retaining laminectomy retractors for repair of 2nd cervical
vertebral fractures in 3 dogs. Veterinary and Comparative Orthopaedics
and Traumatology 10, 71.
Bracken, M.B. (2000a) Methylprednisolone and spinal cord injury. Journal
of Neurosurgery 93 Suppl 1, 175179.
Bracken, M.B. (2000b) The use of methylprednisolone. Journal of
Neurosurgery 93 Suppl 2, 340341.
Bracken, M.B. (2001) High dose methylprednisolone must be given for 24
or 48 hours after acute spinal cord injury. British Medical Journal 322,
862863.
Bracken, M.B., Holford, T.R. (2002) Neurological and functional status
1 year after acute spinal cord injury: estimates of functional recovery in
National Acute Spinal Cord Injury Study II from results modeled in
National Acute Spinal Cord Injury Study III. Journal of Neurosurgery 96,
259266.
Bracken, M.B., Shepard, M.J., Collins, W.F., Holford, T.R., Young, W., Baskin,
D.S., Eisenberg, H.M., Flamm, E., Leo-Summers, L., Maroon, J., et al.
(1990) A randomized, controlled trial of methylprednisolone or naloxone
in the treatment of acute spinal-cord injury. Results of the Second National
Acute Spinal Cord Injury Study. New England Journal of Medicine 322,
14051411.
Bracken, M.B., Shepard, M.J., Collins, W.F., Jr, Holford, T.R., Baskin, D.S.,
Eisenberg, H.M., Flamm, E., Leo-Summers, L., Maroon, J.C., Marshall,
L.F., et al. (1992) Methylprednisolone or naloxone treatment after acute
spinal cord injury: 1-year follow-up data. Results of the Second National
Acute Spinal Cord Injury Study. Journal of Neurosurgery 76, 2331.
Bracken, M.B., Shepard, M.J., Holford, T.R., Leo-Summers, L., Aldrich, E.F.,
Fazl, M., Fehlings, M., Herr, D.L., Hitchon, P.W., Marshall, L.F., et al. (1997)
Administration of methylprednisolone for 24 or 48 hours or tirilazad mesylate for 48 hours in the treatment of acute spinal cord injury. Results of the
Third National Acute Spinal Cord Injury Randomized Controlled Trial.
National Acute Spinal Cord Injury Study. Journal of the American Medical
Branson, K.R., Quandt, J.E., Martinez, E.A., Carroll, G.L., Trim, C.M.,
Dodam, J.R., Hartsfield, S.M., Matthews, N.S., Mackenthun, A., Beleau,
M.H. (2001) A multisite case report on the clinical use of sevoflurane
in dogs. Journal of the American Animal Hospital Association 37,
420432.
Brown, S.A., Hall, E.D. (1992) Role of oxygen-derived free radicals in the
pathogenesis of shock and trauma, with focus on central nervous system
injuries. Journal of the American Veterinary Medical Association 200,
18491859.
Calvert, C.A., Wall, M. (2001) Results of ambulatory electrocardiography
in overtly healthy Doberman Pinschers with equivocal echocardiographic evidence of dilated cardiomyopathy. Journal of the American
Clark, D.M. (1986) An analysis of intraoperative and early postoperative

mortality associated with cervical spinal decompressive surgery in the
dog. Journal of the American Animal Hospital Association 22, 739744.
Cook, S., Prewett, A., Dalton, J., Whitecloud, T.R. (1994) Reduction in
perineural scar formation after laminectomy with Polyactive membrane
sheets. Spine 19, 18151825.
Crawford, M.W., Lerman, J., Pilato, M., Orrego, H., Saldivia, V.,
Carmichael, F.J. (1992) Haemodynamic and organ blood flow responses
to sevoflurane during spontaneous ventilation in the rat: a doseresponse
study. Canadian Journal of Anaesthesia 39, 270276.
Crenshaw, J.T., Winslow, E.H. (2002) Original Research: Preoperative
Fasting: Old Habits Die Hard: Research and published guidelines no
longer support the routine use of NPO after midnight, but the practice
persists. American Journal of Nursing 102, 3644.
Culbert, L.A., Marino, D.J., Baule, R.M., Van-Knox, I.-W. (1998) Complications associated with high-dose prednisolone sodium succinate therapy
in dogs with neurological injury. Journal of the American Animal Hospital
Diehl, K.J., Roberts, S.M. (1991) Keratoconjunctivitis sicca in dogs associated with sulfonamide therapy: 16 cases (19801990). Progress in
Veterinary and Comparative Ophthalmology 1, 276282.
Dow, S.W., LeCouteur, R.A., Poss, M.L., Beadleston, D. (1989) Central
nervous system toxicosis associated with metronidazole treatment of
dogs: Five cases (19841987). Journal of the American Veterinary
Medical Association 195, 365368.
Dunning, D. (2003) Surgical wound infection and the use of antimicrobials. In: D. Slatter (ed.), Textbook of Small Animal Surgery, 3rd edn,
Egger, C.M., Duke, T., Archer, J., Cribb, P.H. (1998) Comparison of
plasma fentanyl concentrations by using three transdermal fentanyl
patch sizes in dogs. Veterinary Surgery 27, 159166.
Einhaus, S.L., Robertson, J.T., Dohan, F.C., Jr, Wujek, J.R., Ahmad, S.
(1997) Reduction of peridural fibrosis after lumbar laminotomy and discectomy in dogs by a resorbable gel (ADCON-L). Spine 22, 14401446;
discussion 14461447.
Endoh, H., Honda, T., Ohashi, S., Hida, S. (2001) [Cerebral autoregulation during sevoflurane or isoflurane anesthesia: evaluation with transient hyperemic response]. Masui 50, 13161321.
Epstein, M.D., Banducci, D.R., Manders, E.K. (1992) The role of the
gastrointestinal tract in the development of burn sepsis. Plastic
Reconstructive Surgery 90, 524531.
Fehlings, M.G., Tator, C.H., Linden, R.D. (1989) The effect of nimodipine
and dextran on axonal function and blood flow following experimental
spinal cord injury. Journal of Neurosurgery 71, 403416.
Fox, L.E., Ford, S., Alleman, A.R., Homer, B.L., Harvey, J.W. (1993)
Aplastic anemia associated with prolonged high-dose trimethoprim
sulfadiazine administration in two dogs. Veterinary Clinical Pathology
22, 8992.
Franks, J.N., Boothe, H.W., Taylor, L., Geller, S., Carroll, G.L., Cracas, V.,
Boothe, D.M. (2000) Evaluation of transdermal fentanyl patches for
analgesia in cats undergoing onychectomy. Journal of the American
Gehlbach, S.H., Bigelow, C., Heimisdottir, M., May, S., Walker, M.,
Kirkwood, J.R. (2000) Recognition of vertebral fracture in a clinical setting. Osteoporosis International 11, 577582.
Geisler, F.H. (1999) Prevention of peridural fibrosis: current methodologies. Neurological Research 21 Suppl 1, S922.
Gelatt, K.N., van der Woerdt, A., Ketring, K.L., Andrew, S.E.,
Brooks, D.E., Biros, D.J., Denis, H.M., Cutler, T.J. (2001) Enrofloxacinassociated retinal degeneration in cats. Veterinary Ophthalmology
4, 99106.
George, E.R., Scholten, D.J., Buechler, C.M., Jordan-Tibbs, J., Mattice, C.,
Albrecht, R.M. (1995) Failure of methylprednisolone to improve the
outcome of spinal cord injuries. American Surgeon 61, 659663.
Gerhart, K.A., Johnson, R.L., Menconi, J., Hoffman, R.E., Lammertse, D.P.
(1995) Utilization and effectiveness of methylprednisolone in a populationbased sample of spinal cord injured persons. Paraplegia 33, 316321.
Gerndt, S.J., Rodriguez, J.L., Pawlik, J.W., Taheri, P.A., Wahl, W.L.,
Micheals, A.J., Papadopoulos, S.M. (1997) Consequences of high-dose
steroid therapy for acute spinal cord injury. Journal of Trauma 42,
279284.
Gill, G., Sakovich, L., Thompson, E. (1979) Pedicle fat grafts for the prevention of scar formation after laminectomy. An experimental study in
dogs. Spine 4, 176186.
Griffiths, I.R. (1978) Spinal cord injuries: a pathological study of naturally
occurring lesions in the dog and cat. Journal of Comparative Pathology
88, 303315.
Grouhi, M., Hummel, D., Roifman, C.M. (1999) Anaphylactic reaction to
oral cefaclor in a child. Pediatrics 103, e50.
Hanson, S.M., Bostwick, D.R., Twedt, D.C., Smith, M.O. (1997) Clinical
evaluation of cimetidine, sucralfate, and misoprostol for prevention of
gastrointestinal tract bleeding in dogs undergoing spinal surgery.
Heary, R.F., Vaccaro, A.R., Mesa, J.J., Northrup, B.E., Albert, T.J.,
Balderston, R.A., Cotler, J.M. (1997) Steroids and gunshot wounds to
the spine. Neurosurgery 41, 576583.
Hinton, L.E., McLoughlin, M.A., Johnson, S.E., Weisbrode, S.E. (2002)
Spontaneous gastroduodenal perforation in 16 dogs and seven cats
(19821999). Journal of the American Animal Hospital Association 38,
176187.
Hoerlein, B.F., Redding, R.W., Hoff, E.J., McGuire, J.A. (1985) Evaluation
of naloxone, crocetin, thyrotropin releasing hormone, methylprednisolone, partial myelotomy, and hemilaminectomy in the treatment of
acute spinal cord trauma. Journal of the American Animal Hospital
Hoffman, W.E., Edelman, G., Kochs, E., Werner, C., Segil, L., Albrecht,
R.F. (1991) Cerebral autoregulation in awake versus isofluraneanesthetized rats. Anesthesia and Analgesia 73, 753757.
Hosgood, G. (2003) Wound repair and specific tissue response to injury.
In: D. Slatter (ed.), Textbook of Small Animal Surgery, 3rd edn, 6686.
Hurlbert, R.J. (2000) Methylprednisolone for acute spinal cord injury: an
inappropriate standard of care. Journal of Neurosurgery 93 (1 Suppl), 17.
Jeffery, N.D., Blakemore, W.F. (1999a) Spinal cord injury in small animals
2. Current and future options for therapy. Veterinary Record 145,
183190.
Jeffery, N.D., Blakemore, W.F. (1999b) Spinal cord injury in small animals.
1. Mechanisms of spontaneous recovery. Veterinary Record 144,
407413.
Jeffery, N.D., Phillips, S.M. (1995) Surgical treatment of intramedullary
spinal cord neoplasia in two dogs. Journal of Small Animal Practice
36, 553557.
Kay-Mugford, P., Benn, S.J., LaMarre, J., Conlon, P. (2000) In vitro effects
of nonsteroidal anti-inflammatory drugs on cyclooxygenase activity in
dogs. American Journal of Veterinary Research 61, 802810.
Kerwin, S.C., Maudlin, G.E. (2003) Hemostasis, surgical bleeding, and
transfusion. In: D. Slatter (ed.), Textbook of Small Animal Surgery, 3rd
edn, 4465. Philadelphia: Elsevier Science.
Kyles, A.E. (1998) Transdermal fentanyl. Compendium on Continuing
Education for the Practicing Veterinarian 20, 721722, 724726.
Kyles, A.E., Hardie, E.M., Hansen, B.D., Papich, M.G. (1998) Comparison of
transdermal fentanyl and intramuscular oxymorphone on post-operative
behaviour after ovariohysterectomy in dogs. Research in Veterinary Science
65, 245251.
Kyles, A.E., Papich, M., Hardie, E.M. (1996) Disposition of transdermally
administered fentanyl in dogs. American Journal of Veterinary Research
57, 715719.
Lamb, C.R. (1994) Common difficulties with myelographic diagnosis of
acute intervertebral disc prolapse in the dog. Journal of Small Animal
LaRocca, H., Macnab, I. (1974) The laminectomy membrane. Studies in
its evolution, characteristics, effects and prophylaxis in dogs. The Journal
of Bone and Joint SurgeryBritish volume 56B, 545550.
LeCouteur, R.A., Sturgess, B. (2003) Spinal fractures and luxations. In:
D. Slatter (ed.), Textbook of Small Animal Surgery, 3rd edn, 12441260.
Levy, M.L., Gans, W., Wijesinghe, H.S., SooHoo, W.E., Adkins, R.H.,
Stillerman, C.B. (1996) Use of methylprednisolone as an adjunct in the
management of patients with penetrating spinal cord injury: outcome
analysis. Neurosurgery 39, 11411148; discussion 11481149.
Lewis, D.D., Hosgood, G. (1992) Complications associated with the
use of iohexol for myelography of the cervical vertebral column in
89
90
dogs: 66 cases (19881990). Journal of the American Veterinary Medical

Maddison, J., Ettinger, S.J., Feldman, E.C. (2000) Adverse drug reactions.
In: S.J. Ettinger and E.C. Feldman (eds), Textbook of Veterinary Internal
Medicine: Diseases of the Dog and Cat, Vols 1 and 2, 5th edn, 320325.
Manning, M.M., Brunson, D.B. (1994) Barotrauma in a cat. Journal of the
Maroon, J.C., Abla, A., Bost, J. (1999) Association between peridural scar
and persistent low back pain after lumbar discectomy. Neurological
Research 21 Suppl 1, S4346.
Martin, R.A., Shell, L., Dodds, W.J. (1986) Focal intramedullary spinal
cord hematoma in a dog. Journal of the American Animal Hospital
Mastronardi, L., Pappagallo, M., Puzzilli, F., Tatta, C. (2002) Efficacy of
the morphine-Adcon-L compound in the management of postoperative
pain after lumbar microdiscectomy. Neurosurgery 50, 518524; discussion 524525.
Matsumoto, T., Tamaki, T., Kawakami, M., Yoshida, M., Ando, M., Yamada, H.
(2001) Early complications of high-dose methylprednisolone sodium
succinate treatment in the follow-up of acute cervical spinal cord injury.
Spine 26, 426430.
Moore, R.W., Withrow, S.J. (1982) Gastrointestinal hemorrhage and pancreatitis associated with intervertebral disk disease in the dog. Journal of
Myre, S.A., Schoeder, T.J., Grund, V.R., Wandstrat, T.L., Nicely, P.G.,
Pesce, A.J., First, M.R. (1991) Critical ketoconazole dosage range for
ciclosporin clearance inhibition in the dog. Pharmacology 43, 233241.
Nesathurai, S. (1998) Steroids and spinal cord injury: revisiting the
NASCIS 2 and NASCIS 3 trials. Journal of Trauma 45, 10881093.
Nuwer, M.R. (1999) Spinal cord monitoring. Muscle and Nerve 22,
16201630.
Olby, N. (1999) Current concepts in the management of acute spinal cord
injury. Journal of Veterinary Internal Medicine 13, 399407.
Olby, N.J., Jeffery, N.D. (2003) Pathogenesis of diseases of the central
nervous system. In: D. Slatter (ed.), Textbook of Small Animal Surgery,
3rd edn, 11321147. Philadelphia: Elsevier Science.
Long-term functional outcome of dogs with severe injuries of the
thoracolumbar spinal cord. Journal of the American Veterinary Medical
Papadogiannakis, E.I. (2000) Cutaneous adverse drug reactions in the dog
and cat. European Journal of Companion Animal Practice 10, 7177.
Parent, C., King, L.G., Van Winkle, T.J., Walker, L.M. (1996) Respiratory
function and treatment in dogs with acute respiratory distress syndrome:
Parker, A.J., Smith, C.W. (1972) Functional recovery following incision of
spinal meninges in dogs. Research in Veterinary Science 13, 418421.
Parker, A.J., Smith, C.W. (1974) Functional recovery from spinal cord
trauma following incision of spinal meninges in dogs. Research in
Veterinary Science 16, 276279.
Parker, A.J., Smith, C.W. (1975) Functional recovery from spinal cord
trauma following delayed incision of spinal meninges in dogs. Research in
Veterinary Science 18, 110112.
Perkins, P.G., Deane, R.H. (1988) Long-term follow-up of six patients
with acute spinal injury following dural decompression. Injury 19,
397401.
Reed, S. (2002) Nonsteroidal anti-inflammatory drug-induced duodenal
ulceration and perforation in a mature rottweiler. Canadian Veterinary
Journal 43, 971972.
Richter, H.P., Kast, E., Tomczak, R., Besenfelder, W., Gaus, W. (2001)
Results of applying ADCON-L gel after lumbar discectomy: the German
ADCON-L study. Journal of Neurosurgery 95, 179189.
Robertson, J., Meric, A., Dohan, F.J., Schweitzer, J., Wujek, J., Ahmad, S.
(1993) The reduction of postlaminectomy peridural fibrosis in rabbits
by a carbohydrate polymer. Journal of Neurosurgery 79, 8995.
Robinson, T.M., Kruse Elliott, K.T., Markel, M.D., Pluhar, G.E., Massa, K.,
Bjorling, D.E. (1999) A comparison of transdermal fentanyl versus
epidural morphine for analgesia in dogs undergoing major orthopedic
surgery. Journal of the American Animal Hospital Association
35, 95100.
Rohrer, C.R., Hill, R.C., Fischer, A., Fox, L.E., Schaer, M., Ginn, P.E.,
Casanova, J.M., Burrows, C.F. (1999a) Gastric hemorrhage in dogs given
high doses of methylprednisolone sodium succinate. American Journal of
Rohrer, C.R., Hill, R.C., Fischer, A., Fox, L.E., Schaer, M., Ginn, P.E.,
Preast, V.A., Burrows, C.F. (1999b) Efficacy of misoprostol in prevention of gastric hemorrhage in dogs treated with high doses of methylprednisolone sodium succinate. American Journal of Veterinary Research
60, 982985.
Rosenbluth, P.R., Meirowsky, A.M. (1953) Sympathetic blockade, an acute
cervical cord syndrome. Journal of Neurosurgery 10, 107112.
Rosin, E. (1988) The timing of antibiotic administration for antimicrobial
prophylaxis in surgery. Veterinary Surgery 17, 181.
Ross, J., Obuchowski, N., Modic, M. (1999) MR evaluation of epidural
fibrosis: proposed grading system with intra- and inter-observer variability. Neurological Research 21 Suppl 1, S2326.
Ross, J.S., Robertson, J.T., Frederickson, R.C., Petrie, J.L., Obuchowski, N.,
Modic, M.T., deTribolet, N. (1996) Association between peridural scar
and recurrent radicular pain after lumbar discectomy: magnetic resonance evaluation. ADCON-L European Study Group. Neurosurgery 38,
855861; discussion 861863.
Roulet, A., Puel, O., Gesta, S., Lepage, J.F., Drag, M., Soll, M., Alvinerie, M.,
Pineau, T. (2003) MDR1-deficient genotype in Collie dogs hypersensitive to the P-glycoprotein substrate ivermectin. European Journal of
Pharmacology 460, 8591.
Rowland, P.H., Center, S.A., Dougherty, S.A. (1992) Presumptive
trimethoprimsulfadiazine-related hepatotoxicosis in a dog. Journal of
Salisbury, S.K., Cook, J.R., Jr (1988) Recovery of neurological function
following focal myelomalacia in a cat. Journal of the American Animal
Hospital Association 24, 227230.
Sapolsky, R.M. (1994) Glucocorticoids, stress and exacerbation of excitotoxic neuron death. Seminars in Neuroscience 6, 323331.
Scherk Nixon, M. (1996) A study of the use of a transdermal fentanyl
patch in cats. Journal of the American Animal Hospital Association 32,
1924.
Scott, D.W., Miller, W.H., Jr (1998) Idiosyncratic cutaneous adverse
drug reactions in the cat: literature review and report of 14 cases
(19901996). Feline Practice 26, 1015.
Scott, D.W., Miller, W.H., Jr (1999) Idiosyncratic cutaneous adverse
drug reactions in the dog: literature review and report of 101 cases
(19901996). Canine Practice 24, 1622.
Seelig, C.B., Maloley, P.A., Campbell, J.R. (1990) Nephrotoxicity associated with concomitant ACE inhibitor and NSAID therapy. Southern
Medical Journal 83, 11441148.
Seim, H.B., Prata, R.G. (1982) Ventral decompression for the treatment
of cervical disk disease in the dog: a review of 54 cases. Journal of the
Siemering, G.B., Vroman, M.L. (1992) High dose methylprednisolone
sodium succinate: an adjunct to surgery for canine intervertebral disc
herniation. Veterinary Surgery 21, 406.
Smith, G.K., Walter, M.C. (1985) Fractures and luxations of the spine.
In: C.D. Newton and D. Nunamaker (eds), Textbook of Small Animal
Orthopedics, 307322. Philadelphia: JB Lippincott.
Smith-Swintosky, V.L., Pettigrew, L.C., Sapolsky, R.M., Phares, C.,
Craddock, S.D., Brooke, S.M., Mattson, M.P. (1996) Metapyrone, an
inhibitor of glucocorticoid production, reduces brain injury induced by
focal and global ischemia and seizures. Journal of Cerebral Blood Flow
and Metabolism 16, 585598.
Songer, M., Ghosh, L., Spencer, D. (1990) Effects of sodium hyaluronate
on peridural fibrosis after lumbar laminotomy and discectomy. Spine
15, 550554.
Songer, M.N., Rauschning, W., Carson, E.W., Pandit, S.M. (1995) Analysis
of peridural scar formation and its prevention after lumbar laminotomy
and discectomy in dogs. Spine 20, 571580; discussion 579580.
Stauffer, J.L., Gleed, R.D., Short, C.E., Erb, H.N., Schukken, Y.H. (1988)
Cardiac dysrhythmias during anesthesia for cervical decompression in
the dog. American Journal of Veterinary Research 49, 11431146.
Stillman, M.T. (1989) Interaction and selection of therapeutic agents in
the elderly: NSAIDs and the ageing kidney. Scandinavian Journal of
Rheumatology Supplement 82, 3338.
Summors, A.C., Gupta, A.K., Matta, B.F. (1999) Dynamic cerebral
autoregulation during sevoflurane anesthesia: a comparison with isoflurane. Anesthesia and Analgesia 88, 341345.
Tator, C.H. (1991) Review of experimental spinal cord injury with emphasis
on the local and systemic circulatory effects. Neurochirurgie 37, 291302.
Tator, C.H., Fehlings, M.G. (1991) Review of the secondary injury theory
of acute spinal cord trauma with emphasis on vascular mechanisms.
Journal of Neurosurgery 75, 1526.
Teague, H.D., Brasmer, T.H. (1978) Midline myelotomy of the clinically
normal canine spinal cord. American Journal of Veterinary Research
39, 15841590.
Thacher, C. (1989) Problems with owner communications. Problems in
Veterinary Medicine 1, 478490.
Toombs, J.P., Collins, L.G., Graves, G.M., Crowe, D.T., Caywood, D.D.
(1986) Colonic perforation in corticosteroid-treated dogs. Journal of the
Trevor, P.B., Martin, R.A., Saunders, G.K., Trotter, E.J. (1991) Healing
characteristics of free and pedicle fat grafts after dorsal laminectomy and
durotomy in dogs. Veterinary Surgery 20, 282290.
Trotter, E.J., Crissman, J., Robson, D., Babish, J. (1988) Influence of

nonbiologic implants on laminectomy membrane formation in dogs.
Weinblatt, M.E. (1989) Drug interactions with non steroidal anti-inflammatory
drugs (NSAIDs). Scandinavian Journal of Rheumatology Supplement
83, 710.
Welch, J.A., Wohl, J.S., Wright, J.C. (2002a) Evaluation of postoperative
respiratory function by serial blood gas analysis in dogs treated with
transdermal fentanyl. Journal of Veterinary Emergency and Critical Care
Medicine 12, 8187.
Welch, W.C., Cornwall, G.B., Toth, J.M., Turner, A.S., Thomas, K.A.,
Gerszten, P.C., Nemoto, E.M. (2002b) Use of polylactide resorbable
film as an adhesion barrier. Orthopedics 25 (10 Suppl), S11211130.
Wheeler, S.J., Davies, J.V. (1985) Iohexol myelography in the dog and cat:
a series of one hundred cases, and a comparison with metrizamide and
iopamidol. Journal of Small Animal Practice 26, 247256.
Widmer, W.R., DeNicola, D.B., Blevins, W.E., Cook, J.R., Jr, Cantwell,
H.D., Teclaw, R.F. (1992a) Cerebrospinal fluid changes after iopamidol
and metrizamide myelography in clinically normal dogs. American
Journal of Veterinary Research 53, 396401.
Widmer, W.R., Blevins, W.E., Jakovljevic, S., Teclaw, R.F., Han, C.M.,
Hurd, C.D. (1992b) Iohexol and iopamidol myelography in the dog:
a clinical trial comparing adverse effects and myelographic quality.
91
Cervical disc disease
Clinical signs
94
Treatment options 96
Non-surgical treatment 96
Surgical treatment 96
Fenestration 96
Ventral decompression 96
Ventral decompression and fixation
Dorsal decompression 98
CLINICAL SIGNS
97
Complications 98
Intraoperative 98
Early postoperative 99
Postoperative care
Prognosis
102
102
Cervical disc disease in cats
103

References
et al., 1992). Dachshunds, Beagles, Poodles, Spaniels,

Shih Tzus, Pekingese and Chihuahuas are affected
most often. Large-breed dogs also suffer from cervical
disc disease, usually as part of the syndrome of cervical
spondylomyelopathy (see Chapter 11). Most small dogs
suffer signs after 2 years of age, with a mean onset at
6 years (Goggin et al., 1970; Gage, 1975).
93
Diagnosis 94
Survey radiography
CSF analysis 94
Myelography 95
CT and MRI 95
Chapter
The predominant clinical sign is severe neck pain (7.1

and Box 7.1), which may be acute or chronic (Fry
et al., 1991; Smith et al., 1997). In making a diagnosis,
it is usually adequate to palpate the spine and muscles
of the neck rather than flex and extend the neck to
confirm pain (2.13). This is one of the few conditions
that cause dogs to scream spontaneously. Often the
pain is unremitting and not responsive to medication
although dogs with more marked neurological deficits
often show less pain. Affected dogs may be reluctant to
eat unless the food is raised off the floor.
Nerve root signature (pain on traction of the limb)
is another frequent finding (Seim and Prata, 1982;
Fry et al., 1991; Morgan et al., 1993). It may present as
103
104
Procedures 106
Approach to the ventral neck 106
Fenestration 109
Ventral decompression with
distraction-stabilization 117
Dorsal hemilaminectomy 119
Dorsal laminectomy 120
Cervical disc disease is a frequent disorder of dogs.
Small dogs are affected commonly, particularly those
with chondrodystrophoid characteristics although the
condition can occur in any breed (Gage, 1975; Dallman
7.1 Dachshund with neck pain caused by a cervical disc

extrusion. Note low head position and thoracic limb held up due
to root signature. The hunched posture can be mistaken for
thoracolumbar pain; history and physical examination
differentiates the two.
94
an apparent orthopedic lameness but nerve root pain

can usually be elicited by neck palpation or limb traction. This sign usually indicates that more caudal discs
are affected but can also be seen with C2/3 discs
(Morgan et al., 1993). Neurological deficits may be
restricted to one thoracic limb or the dog may show
Box 7.1 Clinical signs of cervical disc disease
Neck pain
Spontaneous screaming
Low head carriage
Thoracic limb lameness or paresis
Thoracic limb nerve root signature
Hemiparesis
Tetraparesis
hemiparesis, tetraparesis or even tetraplegia with

hypoventilation (6.1, 7.11). Neurological deficits are
more common with lesions at C4/5 to C6/7 inclusive,
while neck pain without deficits is more common with
lesions at C2/3 and C3/4 (Lemarie et al., 2000). This
may reflect the greater degree of space in the cranial
vertebral canal compared to more caudally.
Most dogs have Hansen type I extrusions. Hansen
type II herniations do occur, generally in larger breed
dogs (7.56). The C2/3 disc is involved most frequently,
with the incidence decreasing caudally (Seim and Prata,
1982; Fry et al., 1991). Lesions at C6/7 are the least
common, with the exception of large-breed dogs as part
of cervical spondylomyelopathy and also in Pekingese
(Fry et al., 1991). The C7/T1 disc herniates occasionally
(7.56). Intracranial lesions can cause neck pain on rare
occasions and should be considered under differential
diagnosis (Coates and Dewey, 1998) (Box 7.2).
Box 7.2 Differential diagnoses for neck pain
DIAGNOSIS
Survey radiography
Schmorls node
Facet joint pain
Synovial cyst
Bicipital bursitis
Temporomandibular joint (TMJ) lesion or oropharyngeal pain
Otitis media
Syringohydromyelia
Osteoporotic pathological fracture
Soft tissue tumor in the neck or bone tumor
Intracranial lesion
Mid-thoracic lesion, e.g. T5/6 discospondylitis
CSF analysis
Thoracic lesion, e.g. pleuritic pain
Meningomyelitis
Polyarthritis
Analysis of CSF is useful to eliminate inflammatory

CNS disease. Results of CSF analysis may be abnormal
in disc disease, but elevations of protein and cells are
usually mild (Thomson et al., 1989). Conditions other
than disc disease are more common in dogs less than
2 years old or in aged animals (Box 7.2).
Polymyositis
Spinal cord hematoma or hemorrhage
Diagnosis is based on the clinical signs described above.

Survey radiographic features of chronic disc disease are
common incidental findings in older dogs. Narrowing
of the intervertebral space or dorsal displacement of
mineralized disc material is suggestive of disc extrusion
(Morgan et al., 1993) (7.2). Discography may be useful
for non-mineralized, lateral extrusions (Felts and Prata,
1983; Wrigley and Reuter, 1984). Myelography or
advanced imaging is necessary for definitive diagnosis
(Somerville et al., 2001).
7.2 A: Lateral myelogram from a

10-year-old Shih Tzu that presented with
neck pain and no neurological deficits.
There is a narrow intervertebral space
at C5/6 with extruded, mineralized disc
material and splitting of the ventral
contrast column at this level (4.234.25).
B: Dorsoventral myelogram. Further
images from the same dog are shown
in 7.3, 7.9 and 7.37.
Myelography
CT and MRI
Myelography or advanced imaging is necessary to

confirm the diagnosis and because multiple discs are
sometimes involved. In some lateral or intraforaminal
extrusions the myelogram is normal but oblique views
may reveal the offending disc (Felts and Prata, 1983).
Oblique views are also useful in determining the side
on which an asymmetrical extrusion lies, although CT
or MRI are better.
Depending on the scanner, CT is usually much faster

to perform than myelography and is also more accurate
for surgical planning (Olby et al., 1999). When available,
CT is the imaging modality of choice for chondrodystrophoid dogs with suspected cervical disc disease
(4.39B, 7.4, 7.5). The cross-sectional image helps the
surgeon to plan the ventral slot and is also useful to
identify lateralized or foraminal lesions (Hara et al., 1994;
C3
C4
C5
C6
7.3 A: CT scan through C5/6 made

after myelography (same dog as shown
in 7.2). The connection between the disc
extrusion and the rest of the nucleus
pulposus is seen clearly. B: T2-weighted,
sagittal MRI to show a disc extrusion at
C4/5 (transverse image, 4.44A). The C3/4
and C5/6 discs are of relatively normal
signal intensity but C2/3 and especially
C4/5 and C6/7 discs are degenerate with
low signal intensity (Sether et al., 1990;
Levitski et al., 1999).
B
7.4 A: Transverse CT and B: 3D
reconstruction of CT scan from a
5-year-old Lhasa Apso that had neck
pain and was unable to walk. The scan
revealed a large mineralized extrusion at
the C4/5 disc space. The postoperative
scan is shown in 7.5.
7.5 A: Transverse and B: 3D

reconstruction of the post-surgical CT
scan from the same dog as shown
in 7.4. A ventral slot was performed,
which permitted the surgeon to remove
the majority of the extruded material.
Note the small amount of residual
material present to one side of the slot
(arrowhead and arrow). The width of this
slot is 39% of the vertebral body width.
95
96
7.6 Dog with severe neck pain and root

signature but no neurological deficits.
A: Transverse CT scan and B: 3D
reconstruction reveals disc material
within the intervertebral foramen at
C5/6 (arrowheads). The disc material
was removed using a blunt probe after
a standard ventral approach with
elevation of the longus colli muscles
from the ventral and lateral aspects of
the anulus fibrosus. Fenestration was
also performed. The dog was pain free
within 2 days.
A
Bagley et al., 1996) (7.6). The scanner can also be used to

show the exact location and extent of disc material on a
scout image (4.39A). A heterogenous, hyperattenuating,
extradural mass with loss of epidural fat is a characteristic
feature of a mineralized disc extrusion (Olby et al., 2000).
If CT is not diagnostic it can be followed by a myelogram
or a CT myelogram (Hara et al., 1994). For more information on performing and interpreting CT images in disc
disease, see Chapter 4, page 5557. MRI may be better
than CT, particularly when the disc material is not mineralized, but scan times are usually longer and the cost is
higher (Levitski et al., 1999) (4.42, 4.44A, 7.3B).
TREATMENT OPTIONS
Treatment may be non-surgical or surgical.
Non-surgical treatment
This entails cage rest and use of anti-inflammatory medications. It can be tried in any patient unless marked
neurological deficits are present. Either non-steroidal
anti-inflammatory drugs (NSAIDs), low-dose prednisolone, or narcotics may be used, sometimes combined with diazepam or methocarbamol (Tables 15.1,
15.2). Acupuncture may also be of benefit (Janssens,
1985). The catastrophic worsening of neurological status that can occur with medical treatment of thoracolumbar discs is rare with cervical disc disease. However,
the neck pain in cervical disc disease seems to be less
responsive to non-surgical treatment than does the pain
from thoracolumbar disc disease. Progression of signs or
lack of response in 12 weeks indicates treatment failure. A dog that is responding well to non-surgical treatment should be kept rested for at least 6 weeks
after clinical signs have resolved. Recurrence of clinical
signs after non-surgical treatment has been reported
in over 30% of patients (Russell and Griffiths, 1968;
Janssens, 1985). The role of chemonucleolysis is unclear
and cervical discs must first be exposed surgically if this
procedure is planned (Atilola et al., 1993).
Surgical treatment
Indications for surgical treatment include:
Failure of non-surgical treatment.
Unremitting pain.
Severe or progressive neurological deficits.
Ventral decompression is the preferred procedure. It
may need to be combined with stabilization for caudal
disc extrusions. Dorsal or dorsolateral decompression
may be required in extrusions that cannot be reached
via ventral slot or if there is doubt about the diagnosis.
Fenestration (7.307.36)
Although the value and desirability of fenestration has
been questioned (Fingeroth, 1989), it should prevent
further extrusion of disc material into the vertebral canal
and so reduce the recurrence rate (Russell and Griffiths,
1968). It also appears effective for dogs with discogenic
pain (Morgan et al., 1993 (Algorithm 7.1)). It is usual to
fenestrate the discs from C2/3 to C5/6 inclusive; C6/7
is fenestrated if there is evidence of disease. Advantages
and disadvantages of fenestration compared to ventral
slot decompression are shown in Table 7.1. Fenestration
is only recommended as a primary procedure for dogs
with discogenic pain and as a prophylactic procedure in
combination with ventral decompression. In small breed
dogs that develop signs suggestive of Wobbler syndrome, fenestration is probably contraindicated as it
may exacerbate bulging of the dorsal anulus (7.15A).
Ventral decompression (7.377.52)
General indications for decompression include:
Presence of neurological deficits.
Spinal cord compression on neuroimaging.
Failure of fenestration.
Removal of disc material by ventral slot decompression provides the most rapid resolution of clinical signs
and it is therefore the treatment of choice. Accurate
Neuroimaging
No spinal cord
compression
Traction
non-responsive
Reassess
diagnosis
Traction
responsive
(small dogs*)
Single
lesions
Fenestrate
for
discogenic
pain
Cranial discs
C2/3 to C3/4
Multiple
lesions
Caudal discs
C4/5 to C6/7
Dorsal
laminectomy
C7/T1 or very
lateral extrusion
Ventral
slot
Dorsolateral
hemilaminectomy
or dorsal
laminectomy
Ventral slot and

stabilize if width
close to 50%
Distraction
(7.55)
*less than 10 kg body weight
Algorithm 7.1 Surgical decision-making in cervical disc disease.
Table 7.1 Comparison of ventral fenestration and slot

Fenestration
Slot
Technically
Easy
Difficult
Accurate identification
of disc involved
Not required
Required
Special equipment
required
No
Yes
Potential for iatrogenic

damage
Unlikely
Possible
Removal of disc material

from vertebral canal
No
Yes
Resolution of neck pain
Slow in
many dogs
Usually within
a few days
identification of the disc involved is an obvious prerequisite. The width of the slot should be about one third the
width of the vertebral body and certainly no more than
50% (7.10, 7.16). These parameters are especially
important for the C4/5 to C6/7 discs inclusive (Fitch
et al., 2000; Lemarie et al., 2000). The inverted cone
technique can be used to improve access while minimizing any potential instability (Goring et al., 1991) (7.50).
Postoperative fusion occurs in a proportion of dogs

following ventral decompression but how well may
depend on the width of the slot (11.12); narrow slots
are less likely to fuse than wide slots (Gilpin, 1976;
Seim and Prata, 1982). If osseous fusion does occur
across the slot it can then predispose even small breed
dogs to domino lesions, although probably to a lesser
degree than it does in large-breed dogs (Prata and Stoll,
1973; Bagley et al., 1993) (7.14).
Ventral decompression and

fixation (7.537.55)
Postoperative instability or subluxation are important
potential complications that can be prevented by
fixation of the interspace at the time of ventral decompression (Fitch et al., 2000; Lemarie et al., 2000).
Fixation is not usually necessary for cranial disc lesions
(C2/3 and C3/4) but should be considered for caudal
lesions (C4/5C6/7) when the slot width is nearing
50% of the vertebral body width (Lemarie et al., 2000)
(7.10, 7.16). Fixation is probably not necessary for
caudal lesions if the slot dimensions are less than 50%
of the vertebral body width. When subluxation does
occur following ventral decompression the rescue
technique of choice is distraction-stabilization. Several
methods can be used for fixation after ventral slot
(see also pages 118, 220). Cement plugs are very difficult to use in small dogs (7.13, 7.54). Bone autografts
97
98
from the ilium are effective but necessitate two surgical approaches (Prata and Stoll, 1973). A more recent
study preferred a bone allograft block placed in the
interspace without use of additional implants (Lemarie
et al., 2000). Autografts give better radiographic fusion
than allografts but other factors such as graft site
morbidity also affect final outcome in humans (Floyd
and Ohnmeiss, 2000). Whatever method is used, bony
fusion should be encouraged so that long-term fixation
does not depend solely on the implant(s). The prognosis is good for dogs that subluxate after a ventral
slot provided that the site can be stabilized (Lemarie
et al., 2000).
In general, adherence to the recommendation to
limit the width of the slot to near 33% is preferable to
managing potential or subsequent complications.
Dorsal decompression (7.567.59)

Dorsal laminectomy is much easier technically in small
dogs than large dogs and short-term morbidity is less of
a problem (Gill et al., 1996; De Risio et al., 2002).
Dorsal laminectomy has been proposed as an alternative to ventral decompression for small dogs but ventral
slot has the advantage of permitting removal of disc
material (Gill et al., 1996; Fitch et al., 2000). Dorsal
laminectomy is therefore best reserved for lesions at
C7/T1, where there is doubt regarding the diagnosis, or
for dogs with multiple lesions (7.15, 7.59).
Hemilaminectomy can be performed in the cervical
region using a lateral approach but a dorsolateral
approach is easier (Lipsitz and Bailey, 1995) (7.567.58,
12.1512.29). Hemilaminectomy is indicated where
disc material is situated too laterally to access via ventral
slot, for some lesions at C7/T1 (7.56), or if there
is doubt about the diagnosis (Seim and Prata, 1982).
Intraforaminal extrusions approached in this way or
laterally (Lipsity & Bailey, 1995) may occasionally be
accessible without entering the vertebral canal (Prata
and Stoll, 1973; Felts and Prata, 1983) (7.6).
occur if the intervertebral space is explored recklessly.

Other technical errors can occur during fenestration,
ventral decompression (7.7, 7.8), or implant placement
(Box 7.3).
Hemorrhage can be problematical at various stages.
Dorsal laminectomy can damage vessels in the epidural
space, especially at the level of a foramen (Hurov,
1979) (7.59, 11.51, 11.52). During a ventral approach
the longus colli muscles tend to bleed when removed
C4
C5
7.7 This dog presented with neck pain and no neurological

deficits. Myelography revealed a mineralized extrusion at the
C4/5 space. A ventral slot was performed with fenestration
between C2/3 and C6/7 inclusive. There was considerable
hemorrhage from the vertebral venous plexus and no disc
material was retrieved from the vertebral canal. It was assumed
that the response would be at least equivalent to a fenestration;
re-examination was scheduled in 2 weeks.
C4
C5
COMPLICATIONS
Intraoperative
The ventral surgical approach itself has few complications unless the surgeon does not identify the midline
correctly and damages the vertebral artery. It is also possible to damage vital structures such as the recurrent
laryngeal nerve (7.25, 7.54); the esophagus can be perforated if mistaken for the longus colli muscle and
pleura or other vital structures may be damaged at the
level of the thoracic inlet (Funkquist and Svalastoga,
1979). Spinal cord damage during fenestration can
7.8 Two weeks after surgery the dog was tetraparetic and
unable to stand. The ventral slot had been performed at the
correct interspace; the cranial margin is just visible in the
caudal portion of C4 (arrowhead). However, the degree of
spinal cord compression is now worse than before. The slot
was re-explored; it was noted to have been directed to one
side. The slot was redirected to the midline and a large
amount of disc material was removed; hemorrhage was
minimal. The dog made a good recovery and was able to walk
2 weeks later.
from their insertion on the ventral process. Muscle

dissection should be restricted to the midline to avoid
bleeding from the vertebral artery or its branches. The
vertebral arteries may also be damaged during removal
of lateral or foraminal disc material from either a ventral or a lateral approach (Felts and Prata, 1983) (7.6).
Bleeding from the bone during drilling can be controlled with bone wax. The vertebral venous plexus
(7.43, 7.44, 11.29B) is damaged easily and hemorrhage can be so severe that it is impossible to continue
to explore the vertebral canal (7.7); it may occasionally
even be fatal (Clark, 1986). Severe hemorrhage may
occur in as many as one quarter of dogs undergoing
ventral slot, even without pre-existing coagulopathy
(Smith et al., 1997). Concurrent use of aspirin or the
presence of any coagulopathy increases the risk (15.2).
Steps recommended in case of severe venous plexus
hemorrhage are outlined in Table 7.2. Venous plexus
hemorrhage can lead on rare occasions to complications
in the early postoperative period. Two dogs that had
venous plexus hemorrhage during ventral decompression developed acute, progressive tetraparesis within
12 h. Both had large blood clots compressing the spinal
cord but recovered well after evacuation combined
with hemostasis using thrombin (Seim and Prata,
1982) (Table 7.2).
Box 7.3 Intraoperative complications
Early postoperative
These relate mainly to continued pain; neurological
deterioration (7.8); or respiratory complications (Box
7.4). After fenestration neck pain can persist in a significant number of dogs and may take 12 months to
resolve completely (Denny, 1978) (see Prognosis,
page 102). Neurological deterioration after fenestration can also occur, probably when incorrect fenestration technique leads to disc material being forced into
the vertebral canal (Tomlinson, 1985).
After ventral slot decompression neck pain should
improve within a few days. Moderate or severe pain
has been reported 3 days after surgery in up to 65% of
dogs although this could have been due to excessive
slot width in some of these dogs (Fitch et al., 2000).
Persistence of severe pain beyond 72 h, or progressive
neurological deficits, warrant repeat imaging (Seim and
Box 7.4 Early postoperative complications
Residual or increased compression at the site (7.8)
Residual material within a foramen
Subluxation or instability at the surgical site (7.10, 7.16, 7.52)
Extradural hematoma
Implant complications (7.13, 7.54)
New extrusion at another site
Hyperflexion injury
Dyspnea (7.12)
Infection or sepsis
Vascular decompensation of spinal cord
Iatrogenic injury
Cardiopulmonary arrest
Hypoventilation (6.1, 7.11)
Implant complications
Edema (7.12)
Hemorrhage (7.7)
Seroma or hematoma of wound
Disc extrusion due to fenestration
Horners syndrome
Ventral slot too wide (7.16)
Megaesophagus
Surgery performed at the wrong site
Laryngeal paralysis
Redistribution of disc material
Aspiration
Inadequate decompression (7.8)
Pneumonia
Table 7.2 Hemorrhage from vertebral venous plexus

Preoperative
Intraoperative
Postoperative
Evaluate for coagulopathy

Cross-match for high-risk patients
Place finger over the slot if severe

Relieve retraction pressure on jugular veins
Plug slot with macerated muscle tissue
Pack slot with Gelfoam
Occlude plexus with direct pressure (11.29B)
Avoid long-term continuous suction of hemorrhage
Increase fluid rate, consider plasma or blood product
Check for neurological deterioration
99
100
A
R
7.10 Ventrodorsal radiograph of a dog where subluxation had

occurred following a very wide ventral slot (59%). The dog had
severe neck pain for a week after surgery, which prompted
repeat radiography. Treatment by distraction-stabilization was
attempted but the dog died (7.16, 7.52).
B
7.9 Residual disc material does not always cause persistent
pain. A: Preoperative 3D reconstruction of the CT myelogram
shown in 7.3B. There is a large mass of mineralized material
on the left side of the floor of the vertebral canal. B: 3D
reconstruction made immediately following surgery. A ventral
slot has been performed but removal of disc material was
suboptimal. Despite this, the dog was almost pain free the day
after surgery and was normal within a few days. The slot width
in this dog is approximately 38%.
Prata, 1982; Smith et al., 1997). This may mean that

some dogs are imaged unnecessarily (7.9) but will also
identify quickly any dogs that need further surgery.
Potential causes of severe postoperative pain are
shown in Box 7.4. One common cause is when some of
the disc material has simply been moved to a new location during attempts at retrieval, which sets up a new
area of irritation. A poorly executed slot is actually no
better than a fenestration and can be worse (Chambers
et al., 1986) (4.25, 7.8, 11.21).
Vertebral instability is another important cause of
persistent pain after surgery and can occur if the slot
is made too wide (Seim and Prata, 1982). Range of
motion increases in cadaver spines by 3040% after
fenestration and by 66% after a ventral slot, even if
the slot is only one third of the vertebral width (Macy
et al., 1999; Wolf and Roe, personal communication).

Subluxation leads to severe pain or marked deterioration in neurological status due to nerve root, meningeal
or spinal cord compression (7.10, 7.16, 7.52). Signs
associated with this usually occur within a week of surgery but can be delayed for up to 3 months (Lemarie
et al., 2000). If subluxation does occur, the lesion should
be managed by distraction-stabilization, as for a traumatic fracture or subluxation. It is likely that instability
is under-recognized as a cause of continued pain after
surgery (Macy et al., 1999; Fitch et al., 2000, Lemarie
et al., 2000).
Infection at the surgical site may cause either incisional drainage or systemic signs (Chambers et al., 1982;
Fry et al., 1991; Lipsitz and Bailey, 1995). It can occur
following dorsal or ventral decompression, especially
if the dog becomes bacteremic for any reason (see page
84). Discospondylitis (14.11) or even epidural abscessation (14.14) may also develop.
Respiratory arrest during or after surgery has been
reported in several studies (Clark, 1986; Waters, 1989;
Smith et al., 1997; Beal et al., 2001). This can be due to
either cardiopulmonary or neurological abnormalities
(see Chapter 6, page 82). Some dogs develop more
severe neurological deficits following surgery so that
C7
A
7.11 Some dogs only need to be ventilated for a day or two

but others require ventilation for 12 weeks before they can be
weaned off mechanical support (Beal et al., 2001) (6.1).
C7
7.12 This dog shows profound swelling 2 days after a ventral

slot; the base of the tongue was also affected. It caused this
dog no problems but can cause dysphagia or dyspnea in some
animals. Hot packing and anti-inflammatory medications are
indicated (Jerram et al., 1997) (see page 357). The possibility of
infection should be investigated.
they are unable to breathe spontaneously. In these animals the only option is to put them on a mechanical
ventilator. Ventilation should be considered once the
paCO2 exceeds 50 mmHg and the dog should either be
ventilated or euthanized on humane grounds if the
paCO2 is greater than 70 mmHg (7.11).
Swelling or edema in the ventral neck can also cause
problems after surgery (Fry et al., 1991; Smith et al.,
1997) (7.12). Care with hemostasis and wound closure
helps to avoid this. Seroma is a particular risk after dorsal decompression (15.34).
Technical problems may occur when trying to place
implants in small dogs (7.13, 7.54). Other potential
7.13 Nine-year-old pug with chronic tetraparesis localizing to

C6T2 spinal cord. A: There is severe spinal cord compression
at C5/6 and C6/7, which was reduced partially by traction.
Cement plugs were placed at C5/6 and C6/7 using anchor
holes in the end plates and the dog was put in a splint
(Dixon et al., 1996) (11.39). B: It developed severe neck pain
5 days later; neurological status was unchanged. A myelogram
showed the lesion at C6/7 had improved but C5/6 was worse.
Disc material could have extruded spontaneously, been forced
into the canal during plug placement, or the distraction had
failed. There was no evidence of cement in the canal. The dog
was euthanized; no necropsy was performed.
complications include trauma during recovery (11.22);

megaesophagus (15.40) and Horners syndrome
(Boydell, 1995); seizures after myelography; urinary
tract infection; diarrhea and sepsis (Fry et al., 1991).
Any dog that has undergone ventral decompression in the past, or that has congenitally fused vertebrae,
could go on to develop a domino lesion. The incidence
is much lower than in Wobbler syndrome but it does
occur (Prata and Stoll, 1973; Bagley et al., 1996) (7.14).
The overall risks of serious complications or mortality with cervical surgery are higher than in thoracolumbar surgery. The highest mortality rate is for dogs
that are unable to walk prior to surgery, especially those
with pre-existing disease. Nine of 37 such dogs (25%)
in one study died or were euthanized. Causes of death
101
102
T1
T1
included cardiac arrest, pulmonary thromboembolism

and respiratory dysfunction (Waters, 1989). Recumbent
dogs are also at high risk of developing pneumonia, as
are those that develop laryngeal paralysis (Prata and
Stoll, 1973; Lemarie et al., 2000) (see Chapter 15).
Three of 52 dogs in another study died after ventral
decompression for cervical disc herniation: one from
uncontrollable venous plexus hemorrhage and two
from acute bradycardia and hypotension (Clark, 1986).
Two of these dogs presented with only neck pain.
Dysrhythmias are common during cervical disc surgery.
They were reported in 15 of 48 dogs (31%), two of
which died, emphasizing the importance of good anesthetic monitoring (Stauffer et al., 1988). Hypotension
is also common during ventral decompression and may
influence outcome (Griffiths, 1973; Cybulski, 1988;
Smith et al., 1997) (see page 86).
In summary, a high index of suspicion for complications must be maintained if the surgery does not go as
planned, if the response to surgery is poor and certainly
if the animals neurological status worsens after surgery
(7.7). In such cases, repeat imaging is strongly recommended (7.77.9).
POSTOPERATIVE CARE
7.14 A: Eight-year-old Silky terrier

(5 kg) with a disc extrusion at C5/6.
A ventral slot was performed and
the dog made a good recovery.
B: Eighteen months later the dog
showed recurrence of neck pain, root
signature, mild tetraparesis and ataxia.
A repeat myelogram revealed no
residual compression at C5/6 but
extrusion of disc material at C4/5 and
C6/7. Note the fusion at C5/6 (7.53,
7.54).
(see Chapter 15)
Most patients should be much less painful within a day

or two of ventral slot surgery. Two weeks of restricted
exercise should be enforced, even if improvement has
been marked. Leash exercise for urination and defecation is allowed, but a harness should be used rather
than a collar. If discomfort persists, anti-inflammatory
drugs may be needed (Table 15.2) along with
diazepam, methocarbamol, ultrasound (15.13B), laser
(15.14A), or acupuncture. Severe pain beyond 48 h
warrants repeat imaging (see Complications, page 99).
Dogs with preoperative neurological deficits benefit
from physiotherapy once the pain has subsided. Largebreed dogs are prone to problems related to recumbency,
and strict attention to bedding and cleanliness is

required (see Chapter 15).
PROGNOSIS
Resolution of clinical signs may be slow after fenestration. Of 12 dogs with neck pain, six recovered rapidly
but five continued to have intermittent pain for 14
weeks. Of a further 16 dogs with thoracic limb paresis,
12 (75%) recovered in an average of 38 weeks. Of
nine with severe deficits (tetraparesis or tetraplegia),
five (56%) recovered in 16 weeks (Denny, 1978).
Fenestration has been compared to ventral slot decompression in 111 dogs that were able to walk prior to
surgery. Intraoperative and postoperative complications,
other than simple incisional swelling, were more common after ventral decompression (30% vs 12%), resulting
in longer hospital stays (7.12). However, neurological
recovery was slower following fenestration. Fenestrated
dogs were three times more likely to have static signs
at discharge and twice as likely to be unchanged at their
last hospital visit (Fry et al., 1991). Ventral slot decompression gives much better results overall (Table 7.3).
This is probably because even dogs with no neurological deficits usually have substantial amounts of
disc material in the vertebral canal.
The prognosis for dogs with pain or moderate
neurological deficits is usually good after a ventral slot
(Table 7.3). The prognosis for dogs that cannot walk is
more guarded. Complete recoveries are reported in
5560% of such dogs, a further 1520% recover with
residual deficits and 2025% die or are euthanized.
Dogs that do not walk within 2 weeks are likely to
continue to have residual deficits (Waters, 1989; Smith
et al., 1997). Others have reported better recovery
rates for severely affected dogs (Seim and Prata, 1982).
The outcome after ventral slot decompression in
small-breed dogs appears to be better for dogs with
cranial cervical lesions (C2/3 or C3/4) than for those
Table 7.3 Results of ventral slot decompression for cervical disc

Neurological
status
No. of
dogs
No. recovered*
by 2 days (%)
No. recovered*
by 7 days (%)
No. recovered*
by 28 days (%)
No. recovered*
by 365 days (%)
No recovery
Pain / RS1
33
16 (48)
24 (73)
30 (91)
33 (100)
Can walk1
14
6 (43)
10 (71)
12 (86)
14 (100)
Can not walk1,2
18
5 (28)
10 (56)
14 (78)
15 (83)
3 (17)
1
Seim and Prata (1982); 2 Waters (1989).
* Recovered defined as no neck pain; ability to walk for dogs unable to walk before surgery.
RS, Root signature.
with caudal cervical lesions (C4/5C6/7 inclusive)

(Waters, 1989; Fitch et al., 2000). Long-term resolution of signs after ventral slot decompression was seen
in 31 of 47 dogs with cranial lesions (66%) compared to
only 10 of 48 with caudal lesions (21%). The poor
outcome for dogs with caudal lesions undergoing a
ventral slot contrasts with the long-term resolution
seen in 12 of 15 dogs (80%) that were also stabilized
and distracted intraoperatively at the ventral slot site
(Fitch et al., 2000).
One of the main reasons for failure to improve, or for
recurrence of signs after ventral slot decompression, is
postoperative instability or subluxation. These problems have only been reported in dogs with caudal
lesions (Seim and Prata, 1982; Wheeler and Sharp,
1994; Smith et al., 1997; Fitch et al., 2000; Lemarie
et al., 2000). One factor that increases the risk considerably is the width of the slot. In dogs that suffer subluxation the slot is usually too wide; reported as
3980% of the vertebral width and a median of 50%
(Fitch et al., 2000; Lemarie et al., 2000) (7.10, 7.16).

Subluxation does appear to be a significant risk when a
wide slot is performed for caudal lesions. It is not clear
if this is also true for small dogs when the slot width
is kept to about one third of vertebral body width
(Lemarie et al., 2000). If ventral slot decompression is
to be performed in dogs with caudal lesions, the slot
proportions should be close to 33% or an inverted cone
should be considered (Goring et al., 1991). If the width
of a caudal slot is nearing 50% then the site should also
be stabilized (Lemarie et al., 2000).
CERVICAL DISC DISEASE IN CATS

Disc herniation is not uncommon in cats, particularly in
the cervical region, but clinical signs related to these
lesions are rare (Heavner, 1971; Littlewood et al., 1984;
Kathmann et al., 2000; Knipe et al., 2001; Muana et al.,
2001). The principles of diagnosis and treatment are
discussed above.

1. How to manage dogs with multiple extrusions of mineralized disc material (7.15A)? Options include dorsal laminectomy (Gill
et al., 1996), or several ventral slot(s) with bone graft distraction. For the latter procedure it would help to be able to decide
which lesion(s) is most significant (4.43).
2. How to deal with multiple, dynamic lesions in small dogs (7.15B)? Options include dorsal laminectomy, fusion using cement
plugs (but see 7.13), or possibly several ventral slots, each with bone graft or cement wedges (page 118).
3. What is the maximum recommended width for a slot and what are the risks of instability (7.16)?
4. Is there a need for routine fixation to prevent luxation in dogs with caudal lesions or is this only necessary when the slot width
is 50% (Lemarie et al., 2000)?
103
104
7.15 A: Multiple herniations in a Pomeranian with neck pain, mild tetraparesis and a disconnected gait (see page 28). Extradural
compression is evident from C2/3 to C7/T1 inclusive; several lesions were of mineralized material. Fenestration of these disc spaces
led to recurrence of signs within a year. B: Miniature pinscher with neck pain and thoracic limb root signature. Traction-responsive,
ventral extradural compression is present from C2/3 to C6/7 inclusive.
7.16 A: Ventrodorsal survey radiograph from a 6-year-old Yorkshire terrier that suffered a vertebral subluxation at the site of a
previous ventral slot at C4/5 (arrow). The width of the slot is 55% (same dog as in 7.52). B: Seven-year-old Dachshund with
neck pain and root signature that was unable to walk after ventral slot at C3/4; slot width is 66%. An external splint was applied;
the dog could walk within 2 weeks.
5. Are cranial lesions more resistant to subluxation (Lemarie et al., 2000) (7.16B)?
6. What is the best way to fuse interspaces in small-breed dogs? Options include metal and bone cement (page 118) or bone
allografts combined with ventral slot (Fitch et al., 2000). Non-biological materials that permit bone in-growth may become
available in the future (Cook et al., 1994; Emery et al., 1996).
REFERENCES
Atilola, M.A.O., Cockshutt, J.R., McLaughlin, R., Cochrane, S.M.,
Pennock, P.W. (1993) Collagenase chemonucleolysisa long term radiographic study in normal dogs. Veterinary Radiology and Ultrasound 34,
321324.
Bagley, R.S., Forrest, L.J., Cauzinille, L., Hopkins, A.L., Kornegay, J.N.
(1993) Cervical vertebral fusion and concurrent intervertebral disc extrusion in four dogs. Veterinary Radiology and Ultrasound 34, 336339.
Bagley, R.S., Tucker, R., Harrington, M.L. (1996) Lateral and foraminal
disk extrusion in dogs. Compendium on Continuing Education for the
Practicing Veterinarian 18, 795804.
Boydell, P. (1995) Horners syndrome following cervical spinal surgery in
the dog. Journal of Small Animal Practice 36, 510512.
Chambers, J.N., Oliver, J.E., Jr, Kornegay, J.N., Malnati, G.A. (1982)
Ventral decompression for caudal cervical disk herniation in large- and
giant-breed dogs. Journal of the American Veterinary Medical Association
180, 410414.
Chambers, J.N., Oliver, J.E., Jr, Bjorling, D.E. (1986) Update on ventral
decompression for caudal cervical disk herniation in Doberman Pinschers.
Clark, D.M. (1986) An analysis of intraoperative and early postoperative
mortality associated with cervical spinal decompressive surgery in the
Coates, J.R., Dewey, C.W. (1998) Cervical spinal hyperesthesia as a clinical sign of intracranial disease. Compendium on Continuing Education for
the Practicing Veterinarian 20, 10251037.
Cook, S.D., Dalton, J.E., Tan, E.H., Tejeiro, W.V., Young, M.J.,
Whitecloud, T.S. 3rd (1994) In vivo evaluation of anterior cervical
fusions with hydroxylapatite graft material. Spine 19, 18561866.
Cybulski, G., DAngelo, C.M. (1988) Neurological deterioration after
laminectomy for spondylotic cervical myeloradiculopathy: the putative
role of spinal cord ischaemia. Journal of Neurology, Neurosurgery and

Psychiatry 51, 717718.
Dallman, M.J., Palettas, P., Bojrab, M.J. (1992) Characteristics of dogs
admitted for treatment of cervical intervertebral disk disease: 105 cases
(19721982). Journal of the American Veterinary Medical Association
200, 20092011.
De Risio, L., Muana, K.R., Murray, M., Olby, N.J., Sharp, N.J.H.,
Cuddon, P. (2002) Dorsal laminectomy for caudal cervical
spondylomyelopathy: postoperative recovery and long-term follow-up in
20 dogs. Veterinary Surgery 31, 418427.
Denny, H.R. (1978) The surgical treatment of cervical disc protrusions in
the dog: a review of 40 cases. Journal of Small Animal Practice 19,
251257.
Dixon, B.C., Tomlinson, J.L., Kraus, K.H. (1996) Modified distractionstabilization technique using an interbody polymethyl methacrylate
plug in dogs with caudal cervical spondylomyelopathy. Journal of the
Emery, S.E., Fuller, D.A., Stevenson, S. (1996) Ceramic anterior spinal
fusion. Biologic and biomechanical comparison in a canine model. Spine
21, 27132719.
Felts, J.F., Prata, R.G. (1983) Cervical disk disease in the dog: intraforaminal and lateral extrusions. Journal of the American Animal Hospital
Fingeroth, J.M. (1989) Fenestration. Pros and cons. Problems in Veterinary
Medicine 1, 445466.
Fitch, R.B., Kerwin, S.C., Hosgood, G. (2000) Caudal cervical intervertebral disk disease in the small dog: role of distraction and stabilization in
ventral slot decompression. Journal of the American Animal Hospital
Floyd, T., Ohnmeiss, D. (2000) A meta-analysis of autograft versus allograft in anterior cervical fusion. European Spine Journal 9, 398403.
Fry, T.R., Johnson, A.L., Hungerford, L., Toombs, J. (1991) Surgical
treatment of cervical disc herniations in ambulatory dogs. Ventral
decompression vs. fenestration, 111 cases (19801988). Progress in
Veterinary Neurology 2, 165173.
Funkquist, B., Svalastoga, E. (1979) A simplified surgical approach to the last
two cervical discs of the dog. Journal of Small Animal Practice 20, 593601.
Gage, E.D. (1975) Incidence of clinical disc disease in the dog. Journal of
the American Animal Hospital Association 11, 167174.
Gill, P.J., Lippincott, C.L., Anderson, S.M. (1996) Dorsal laminectomy
in the treatment of cervical intervertebral disk disease in small dogs: a
retrospective study of 30 cases. Journal of the American Animal
Gilpin, G.N. (1976) Evaluation of three techniques of ventral decompression of the cervical spinal cord in the dog. Journal of the American
Goggin, J.E., Li, A., Franti, C.E. (1970) Canine intervertebral disk disease:
Characterization by age, breed, sex and anatomic site of involvment.
Griffiths, I.R. (1973) Spinal cord blood flow in dogs: the effect of blood
pressure. Journal of Neurology, Neurosurgery and Psychiatry 36, 914.
Hara, Y., Tagawa, M., Ejima, H., Orima, H., Fujita, M. (1994) Usefulness
of computed tomography after myelography for surgery on dogs with
cervical intervertebral disc protrusion. Journal of Veterinary Medical
Science 56, 791794.
Heavner, J.E. (1971) Intervertebral disc syndrome in the cat. Journal of the
Hurov, L. (1979) Dorsal decompressive cervical laminectomy in the dog:
surgical considerations and clinical cases. Journal of the American Animal
Janssens, L.A.A. (1985) The treatment of canine cervical disc disease by
acupuncture: a review of 32 cases. Journal of Small Animal Practice 26,
203212.
Jerram, R.M., Hart, R.C., Schulz, K.S. (1997) Postoperative management
of the canine spinal surgery patientpart I. Compendium on Continuing
Education for the Practicing Veterinarian 19, 147161.
Kathmann, I., Cizinauskas, S., Rytz, U., Lang, J., Jaggy, A. (2000) Case
report spontaneous lumbar intervertebral disc protrusion in cats: literature review and case presentations. Journal of Feline Medicine and
Surgery 2, 207212.
Knipe, M.F., Vernau, K.M., Hornof, W.J., LeCouteur, R.A. (2001)
Intervertebral disc extrusion in six cats. Journal of Feline Medicine and
Surgery 3, 161168.
Lemarie, R.J., Kerwin, S.C., Partington, B.P., Hosgood, G. (2000) Vertebral
subluxation following ventral cervical decompression in the dog. Journal of
Levitski, R.E., Lipsitz, D., Chauvet, A.E. (1999) Magnetic resonance imaging of the cervical spine in 27 dogs. Veterinary Radiology and Ultrasound
40, 332341.
Lipsitz, D., Bailey, C.S. (1995) Clinical use of the lateral cervical approach
for cervical spinal cord and nerve root disease: eight cases. Progress in
Littlewood, J.D., Herrtage, M.E., Palmer, A.C. (1984) Intervertebral disc
protrusion in a cat. Journal of Small Animal Practice 25, 119127.
Macy, N.B., Les, C.M., Stover, S.M., Kass, P.H. (1999) Effect of disk
fenestration on sagittal kinematics of the canine C5C6 intervertebral
space. Veterinary Surgery 28, 171179.
Morgan, P.W., Parent, J., Holmberg, D.L. (1993) Cervical pain secondary
to intervertebral disc disease in dogs; radiographic findings and surgical
implications. Progress in Veterinary Neurology 4, 7680.
Muana, K.R., Olby, N.J., Sharp, N.J.H., Skeen, T.M. (2001) Intervertebral
disk disease in 10 cats. Journal of the American Animal Hospital Association
37, 384389.
Olby, N.J., Munana, K.R., Sharp, N.J.H., Flegel, T., Van Camp, S., Berry,
C.R., Thrall, D.G. (1999) A comparison of computed tomography and
myelography in the diagnosis of acute intervertebral disc disease in dogs.
Olby, N.J., Munana, K.R., Sharp, N.J.H., Thrall, D.E. (2000) The
computed tomographic appearance of acute thoracolumbar intervertebral
disc herniations in dogs. Veterinary Radiology and Ultrasound 41,
396402.
Prata, R.G., Stoll, S.G. (1973) Ventral decompression and fusion for the
treatment of cervical disc disease in the dog. Journal of the American
Animal Hospital Association 9, 462472.
Russell, S.W., Griffiths, R.C. (1968) Recurrence of cervical disc syndrome
in surgically and conservatively treated dogs. Journal of the American
Sether, L.A., Yu, S., Haughton, V.M., Fischer, M.E. (1990) Intervertebral
disk: normal age-related changes in MR signal intensity. Radiology 177,
385388.
Smith, B.A., Hosgood, G., Kerwin, S.C. (1997) Ventral slot decompression for cervical intervertebral disc disease in 112 dogs. Australian
Veterinary Practitioner 27, 5864.
Somerville, M.E., Anderson, S.M., Gill, P.J. (2001) Accuracy of localization of cervical intervertebral disk extrusion or protrusion using survey
radiography in dogs. Journal of the American Animal Hospital
Stauffer, J.L., Gleed, R.D., Short, C.E., Erb, H.N., Schukken, Y.H. (1988)
Cardiac dysrhythmias during anesthesia for cervical decompression in
the dog. American Journal of Veterinary Research 49, 11431146.
Tomlinson, J. (1985) Tetraparesis following cervical disk fenestration in
two dogs. Journal of the American Veterinary Medical Association 187,
7677.
Waters, D.J. (1989) Nonambulatory tetraparesis secondary to cervical disk
disease in the dog. Journal of the American Animal Hospital Association
25, 647653.
Wheeler, S.J., Sharp, N.J.H. (1994) Small Animal Spinal Disorders:
Diagnosis and Surgery, 1st edn. St Louis: Mosby, page 82.
Wrigley, R.H., Reuter, R.E. (1984) Canine cervical discography. Veterinary
Radiology 25, 274279.
105
106
PROCEDURES
Approach to the ventral neck (7.177.29)
The cervical spine is extended over a sandbag; the thoracic limbs are pulled in a caudal direction, and tape is used to
immobilize the head and thorax (7.17). It is important to have the dog straight to insure that the neck is aligned correctly.
Note that extension of the caudal portion of the neck as shown here will tend to close the dorsal intervertebral space of
C5/6 and C6/7 spaces. Ventral slot decompression at these sites is easier to perform if the neck is put in a more neutral
position. If access needs to be improved to C6/7, it is preferable to put the neck in gentle traction using a weight on the
maxilla (11.24), or consider a dorsal (11.4411.55, 7.59) or dorsolateral laminectomy (7.567.58, 12.1512.29).
Deep anatomy includes the carotid sheaths, recurrent laryngeal nerves and esophagus (7.257.27). Careful dissection
of the loose fascial layers in a longitudinal direction will expose these structures. Finger dissection here is safe and
effective. Further dissection of the deep fascia reveals the longus colli muscles, which lie ventral to the cervical vertebrae
(7.28). If preferred, the paramedian approach to the ventral neck can be used instead of the approach described here
in order to provide greater protection for blood vessels, nerves, trachea and esophagus (11.2511.27).
Blunt-jawed self-retaining retractors (5.8) are used initially to maintain exposure (7.29). The retractors must not interfere
with gas flow in the airway. Palpate the end of the endotracheal tube and insure that it is positioned distal to the retractors.
Moist towels or sponges are used to protect the tissues. The longus colli muscles join in the midline and insert on the ventral
process of the vertebra, thus overlying the intervertebral discs (7.29). Accurate identification of the intervertebral discs at this
stage is crucial. The ventral process of C5 vertebral body lies in the midline between the cranial borders of the large C6 transverse processes (4.6, 7.37B); the C5/6 intervertebral disc lies immediately behind this ventral process. The ventral process
of C1 is particularly prominent and sharp (7.19); this also can or may be palpated. There is no intervertebral disc at C1/2.
7.17
7.17 Positioning of dog for ventral approach to the

neck. Excessive extension will close the dorsal
disc space; mild traction may be preferable
(11.24A).
7.18 Landmarks are (a) larynx; (b) wing of atlas; and

(c) manubrium of sternum. Incision site is depicted
on the midline.
c
a
7.18
7.19 Landmarks can be palpated; in this illustration

the surgeon is palpating the wings of the atlas
(left hand) and the prominent transverse
process of C6 (4.6, 7.37B).
7.19
7.20 Superficial anatomysee

also 7.21. The skin and
superficial fascia have
been divided to reveal the
sternocephalicus muscles
(a), and sternohyoid
muscles (b). The cervical
vertebrae C1 through C7
are seen in the
background.
b
a
7.20
7.21 The skin and superficial fascia have been

divided. This reveals the sternocephalicus
muscles (a), and the sternohyoid muscles (b).
The sternocephalicus muscles should be divided
to the manubrium, particularly if access to the
caudal cervical vertebrae is required.
7.21
7.22 The sternohyoid muscles are divided in the

midline. If the median raphe is not apparent, apply
finger pressure to the muscles over the trachea
and the raphe will become visible.
7.22
7.23 Making a small incision in the fascia and then

dissecting the fascia bluntly usually prevents
bleeding. The caudal thyroid vein lies in the
connective tissue with small branches on each
side. This vein should be preserved if possible;
the branches may be cauterized. The trachea is
immediately under the fascia.
7.23
107
108
7.24 Once the sternohyoid muscles have been

divided the trachea is visible. Here the separation
is made with fingers.
7.24
7.25 Close up view of the trachea (a); recurrent

laryngeal nerve (b); and carotid sheath (c). The
approach is between the recurrent laryngeal
nerve (which remains associated with the
trachea) and the carotid sheath at the level
indicated by the arrow. At this stage the
esophagus must be identified or the surgeon
could damage it inadvertently (7.26).
a
c
7.25
7.26 There are several layers of loose fascia that must

be penetrated before reaching the longus colli
muscles. The esophagus is identified by its pink
color compared to the darker red of the longus
colli muscles; differentiation is aided by having an
esophageal stethoscope in place (see 7.27 for
labels). The recurrent laryngeal nerve has been
separated from the trachea. This is for illustration
only and is not done normally.
7.27 Note the recurrent laryngeal nerve (a),

and the carotid sheath containing the
carotid artery along with the more
prominent, white vagosympathetic
trunk (b). The trachea and esophagus
have been retracted away from the
surgeon; they should be kept together
on the same side of the incision.
b
7.26
b
7.27
7.28 Retraction of the vital structures exposes the

longus colli muscles. The transverse processes of
C6 are large and are directed ventrally; they are
palpated readily as shown here (4.6, 7.37B).
7.28
7.29 Deep anatomy. Note the pattern of

longus colli muscle bellies running
cranially. Once landmarks have been
identified and any fenestrations
completed, pressure on the jugular
veins should be relieved by
substituting smaller Gelpi retractors
over the space of interest if a ventral
slot is to be performed.
7.29
Fenestration (7.307.36)
During fenestration it is important not to confuse the ventral process of C2C5 vertebrae with the transverse processes
(see under Approach, page 106). The longus colli muscles run cranially to insert on the ventral process in the midline.
The transverse processes can be palpated on each side of the ventral process. The three rows of structures (ventral
processes in the midline, transverse processes laterally on each side) should be identified before proceeding.
7.30 To gain access to the disc, the longus colli

muscles are divided in the midline, immediately
caudal to the ventral process. They may be cut
or division may be achieved with a pair of small,
curved hemostats.
7.30
109
110
7.31 Method of fenestration. The anulus fibrosus is cut

in the form of a window to allow access to the
nucleus pulposus (7.34).
7.31
7.32
7.32 The nuclear material is removed (7.35). It is

important not to push the instrument deeply
through the dorsal anulus fibrosus into the
vertebral canal. Radiographs provide some guide
to the depth of the intervertebral space as long as
magnification is taken into account. The window in
the anulus must be larger than the instrument to
avoid creating a piston effect, which could force
disc material dorsally into the vertebral canal.
7.33 If greater exposure of the disc is required, the

muscle separation is maintained by self-retaining
retractors. The longus colli tendons may also be
cut. The ventral surface of the vertebra (a), and
the ventral anulus fibrosus (b), can be seen. The
surgeons finger is on the ventral process of the
vertebra cranial to the disc.
7.33
7.34 Fenestration is done by cutting a hole in the

anulus fibrosus; this should be made as large as
possible to allow complete evacuation of the
nucleus pulposus. A fresh #11 scalpel blade is
used to make the opening, either by making four
cuts in a rectangular pattern in the anulus
fibrosus, or by using a sawing motion in an oval
pattern until the piece of anulus is free. The
surgeons finger is on the ventral process in this
picture.
7.34
7.35 Disc fenestrated. Nucleus pulposus is oozing

from the intervertebral space. The removed piece
of anulus fibrosus lies cranial to the disc space
(arrow). A small curette or blunt instrument is
used to remove the nucleus pulposus by entering
the space, dragging towards the surgeon and
7.35
then lifting the material out. This maneuver
should be repeated until there is a reasonable
certainty that all the nuclear material has been removed. Care must be taken to avoid penetrating the
vertebral canal dorsally (7.32).
7.36
7.36 Sagittal diagram shows that fenestration only

allows access to disc material in the
intervertebral space. Disc material lying in the
vertebral canal cannot be removed by
fenestration.
Ventral decompression (7.377.52)

Sometimes the dorsal longitudinal ligament remains after removal of the anulus fibrosus. The ventral surface (the side
nearest the surgeon) of the ligament is slightly roughened and has longitudinal fibers; the dorsal surface is smooth and
shiny. The identity of any tissue exposed in the vertebral canal must be established before further manipulations or incisions are made. It is often necessary to complete bone removal before tissue can be identified clearly.
If hemorrhage from the venous plexus does occur, it may be possible to control it with Gelfoam, direct pressure
(11.29), or a small piece of muscle (Table 7.2). Pressure on the jugular veins should be relieved. Work should then continue at the other end of the slot. Suction may be maintained while disc material is being removed, but careful note must
be taken of the amount of blood aspirated. If hemorrhage is severe, the slot is packed and time allowed for it to stop
before re-exploration.
111
112
7.37 A: View of cervical vertebrae showing ventral slot

decompression. The procedure allows access to the
vertebral canal through the vertebral body. Correct
identification of the intended space is crucial (see
Approach, page 106). Where concomitant fenestration
is to be performed this should be done first in order to
verify landmarks. B: 3D reconstruction of a ventral slot
(same dog as shown in 7.9) to illustrate the appearance
from a ventral aspect at C5/6. The overall width of this
slot is 38%. Note the prominent wings of C6
(arrowheads).
7.37
A
7.38 The site for the ventral slot is prepared by

removing the musculature from the vertebrae on
both sides of the midline. The longus colli
muscles are separated from the ventral process.
Hemostasis at this stage is important, as
excessive bleeding will obscure the site of the
slot. Muscle separation is maintained with selfretaining retractors and tissues are protected
with moist sponges.
7.38
7.39
7.39 Sagittal section shows that the slot should be

started more in the vertebra cranial to the disc
than the one caudal, because of the angulation
of the disc space relative to the vertebral canal.
This will also allow room for screws or pins to
be placed should the slot collapse or subluxate
(7.10, 7.16, 7.52, 11.21).
7.40
7.40 The slot is commenced in the position described

in 7.39. Initially drilling is only into the cranial
vertebra. The cortical bone has been penetrated
to reveal purple cancellous bone; the
intervertebral disc is caudal to this. The aim is to
create a slot approximately one third of the width
of the vertebra and one third of its length, with
the disc space at its center once the vertebral
canal is reached (7.39, 7.41). This size of slot
should not result in instability and if kept on midline should avoid the internal vertebral venous plexus
(7.8, 7.43, 7.44).
7.41
7.41 Partially completed slot. Somewhat more

advanced than in the previous illustration.
Hemorrhage from the bone is controlled with
bone wax. It is important to keep the bur
irrigated and the site free of debris. It is useful to
stop drilling periodically to clean the site and to
assess progress.
7.42 The floor of the slot is shown. It is important to

judge the depth of the slot accurately. Cortical
bone is white and hard (a), in contrast to the dark
cancellous bone (b). The cancellous bone is
removed over the whole slot area, using a small
bur. The remaining cortical bone is then thinned
to allow easy removal; it is best to thin the whole
area before entering the vertebral canal (8.36,
10.34). The dorsal anulus fibrosus is visible (c).
The shape of the opening into the vertebral canal
must take into account the location of the venous
plexus (7.43, 7.44).
b
a
7.42
113
114
7.43 Latex-infused preparation to demonstrate the

location of vertebral venous plexus on the floor of
the vertebral canal. The dorsal lamina has been
removed for this view from above the spine. The
discs are represented in orange. Note how the
venous plexus converges in the mid-portion of
the vertebra and diverges over the disc
(1.36, 7.44).
7.43
7.44 This series of CT scans was made after intravenous contrast to

highlight the vertebral venous plexus (arrows). A: Middle of C6
vertebra. B: At the C6/7 disc space. C: Middle of C7. Note how
the vessels diverge at the level of the disc space but converge
again over the vertebral bodies.
7.44
A
7.45 The dorsal anulus fibrosus is incised on each

side of the slot using a fresh #11 blade. One
incision can be seen uppermost in the picture
(arrow). The near-side incision is being made.
The incision is then extended to where the
cortical bone has been removed.
7.45
7.46 The anulus fibrosus is lifted with a suitable

instrument (in this case, a pointed tartar scraper)
and dissected free. Removal can be facilitated by
preserving a small knub of anulus to grasp with
fine rongeurs (11.29A).
7.46
7.47 The dorsal anulus has been removed, allowing a

view into the vertebral canal. Here the glossy
dura mater is visible. When extruded disc
material is present, it is wise to remove it from
the midline first to avoid damaging the venous
plexus early in the procedure.
7.47
7.48 The slot is almost complete. The floor of the slot is

excised with fine rongeurs or a small curette after
bone has been removed and ligamentous tissue
cut with a scalpel. Great patience is required
when removing herniated disc material;
magnification and additional lighting are preferred
(5.15.3). Once all material has been removed,
the spinal cord will lie adjacent to the edges of the
slot as shown. If the disc is asymmetrically
positioned, it is possible to shape the base of the
slot to allow more exposure on one side although
this may increase the risk of bleeding (7.50B).
7.48
115
116
7.49 Removal of mineralized disc material is generally

straightforward although patience is required. As
in the thoracolumbar region, additional disc
material usually remains in the vertebral canal
unless the dura returns to lie level with the bone.
7.49
7.50
7.50 In small dogs or dogs

with very large
extrusions the inverted
cone technique can
improve access to the
vertebral canal A: This
5-year-old Dachshund
A
B
had severe neck pain
and mild tetraparesis. The 3D reconstruction of the preoperative CT shows a huge extrusion at C2/3,
viewed looking caudally. B: 3D reconstruction made from the immediate postoperative CT scan looking
cranially to C1. The slot has been extended to one side (arrowhead) as an inverted cone in order to give
greater access without causing instability (Goring et al., 1991). Nearly all of the material has been
removed. The width is approximately 33%.
7.51
7.51 Access to disc material in the vertebral canal via

a ventral slot. A: Sagittal diagram. Note the
starting position of the slot relative to the disc.
B: Inverted preoperative (left) and postoperative
(right) 3D reconstruction of the dog shown in
7.50. The extrusion at C2/3 is indicated in the
left-hand panel (*). The margins of the slot are
outlined in the right-hand panel by arrowheads.
7.52 A: Postoperative radiograph of a

dog that had a subluxation after
ventral slot at C4/5. The width of
the slot was 55% of the vertebral
body (7.16). B: Stabilization was
performed using pins and cement.
The dog suffered respiratory arrest
and did not recover from surgery.
C6
C6
7.52
A
Ventral decompression with distraction-stabilization (7.537.55)

(see also Chapter 11)
Preferred methods of distraction-stabilization use a corticocancellous autograft, allograft (Veterinary Transplant Services,
Seattle, WA), or metal and bone cement (Fitch et al., 2000); cement plugs are difficult to use in small dogs (7.13, 7.54).
Initial distraction can be accomplished as shown in 11.3111.34, by K-wires (page 235), or by a rope around the dogs
upper canine teeth and tied to the surgical table (7.55, 11.24). After distraction a bone allograft (Veterinary Transplant
Services, Seattle, WA), cement (page 118), or an autograft from the wing of C6, sternum or ilium, can be wedged into
the slot to maintain distraction (Fitch et al., 2000). Metal and cement implants may also be used (7.55).
7.53 A: This 8-kg Silky terrier

had undergone a ventral
slot at C5/6 and then
suffered domino lesions at
C4/5 and C6/7 some18
months later (same dog as
7.14). B: Both lesions were
traction-responsive.
T1
T1
7.53
A
7.54 A: Cement plugs at C4/5 and

C6/7 using anchor holes in the
end plates (11.38) with
cancellous grafting from C4 to
T1
T1
C7 (Dixon et al., 1996). An
external splint was not used. B:
Six-week follow-up shows end7.54
plate fracture at C4/5 and
A
B
ventral displacement of the
implant at C6/7 (7.13). The dog was doing well clinically but had developed iatrogenic laryngeal paralysis.
117
118
7.55
7.55 Miniature pinscher with severe neck pain and left thoracic limb root
signature. A: Preoperative myelogram shows collapse and
mineralized material at C6/7. B: CT myelogram shows mineralized
extrusion or osteophytes within the left foramen also (arrowhead)
(Prata and Stoll, 1973). C: Postoperative radiograph of stabilization
using 2.7-mm screws and bone cement following traction on the
maxilla as the cement hardened (compare disc space to A). The
caudal screw is positioned poorly but the dog has shown no
clinical signs for 6 years.
A
L
Distraction can be maintained with a small amount of bone cement wedged into a partial ventral slot. The edges of
the slot are undercut to prevent slippage of the cement. This method can also be used to rescue a wide or
subluxated ventral slot, as the main goal here is to prevent excessive collapse or subluxation as seen in 7.10.
Gelfoam (Pharmacia, Kalamazoo, MI) should be used to insulate the spinal cord (page 235 and 239). The cement
wedge should be kept to the ventral half of the interspace.
Dorsal hemilaminectomy (7.567.58) (see also page 266)
7.56
7.56 A: Transverse and B: 3D

reconstruction of a CT
scan from a 12-year-old
Labrador retriever with
paraparesis, worse on the
right side. Neurological
examination revealed a
disconnected thoracic
limb gait (see page 28)
and guarding of the neck. A
B
These findings suggested
a caudal cervical lesion; a C7/T1 ventral, extradural compression was confirmed by myelography. CT
scan revealed the mass to be mainly right-sided. Differential diagnoses were a disc, nerve root tumor or
meningioma.
7.57 A: Preoperative
myelogram and B: 3D
reconstruction of the
postoperative CT
scan to show the
hemilaminectomy
(arrowheads) used to
access the mass. The gap
between C6 and C7
(arrow) is an artefact of
reconstruction. Head is to
the right in these images.
7.57
7.58
7.58 Postoperative A:
Transverse and B: 3D
reconstruction of a CT
scan from the same dog as
shown in 7.567.57. The
mass was confirmed to be
a large, fibrous disc
herniation; about 75% of it
was removed using a
B
A
combination of scalpel and
rongeurs. A pocket of gas is visible at the surgical site in A. The extent of the hemilaminectomy is shown
clearly in B. The dog made a rapid recovery.
119
120
Dorsal laminectomy (7.59)

The surgical approach is described in Chapter 11 (11.4411.55). Large veins may be encountered in the dorsal epidural
space, especially close to the intervertebral foramen (Hurov, 1979) (11.51, 11.52).
R
7.59
7.59 A: Ventrodorsal myelogram from a dog with hemiparesis and neck

pain shows left-sided, extradural compression over C7 vertebra.
B: CT myelogram shows mineralized material to the left of the dural
tube over mid-C7 vertebral body. C: Intraoperative photograph shows
mineralized disc material to one side of the dural tube (arrowheads).
The dura was retracted using stay sutures to improve access.
Fenestration was performed using a ventral approach after the
laminectomy. The dogs neurological status was improved the next day
and it was almost normal 1 month later.
Thoracolumbar disc disease
Clinical signs
122
CLINICAL SIGNS
Treatment 123
Non-surgical 123
Decompression (hemilaminectomy,
mini-hemilaminectomy or pediculectomy)
Fenestration without decompression 127
126
Complications 127
Intraoperative complications 127
Early postoperative complications 127
Late postoperative complications 130
Postoperative care
Prognosis
132
132
Thoracolumbar disc disease in cats

References
frequently, usually after middle age. However, disc disease

should not be overlooked in large dogs as they can have
acute, Hansen Type I extrusions as well as Type II lesions
(see pages 12 and 5658). L1/2 is affected most often in
large dogs (Cudia and Duval, 1997).
121
Diagnosis 122
Radiography 122
Cerebrospinal fluid analysis
Myelography 122
CT and MR imaging 123
Chapter
133
133
134
Procedures 136
Dorsolateral hemilaminectomy 136
Pediculectomy and mini-hemilaminectomy
Lateral fenestration 154
151
Thoracolumbar disc disease is a common disorder in

dogs that affects mainly chondrodystrophoid breeds.
Peak incidence in these breeds is between 3 and 6 years
of age. Over 50% of lesions occur at the T12/13 and
T13/L1 discs and more than 85% occur between
T11/12 and L2/3 inclusive; disc extrusion occurs occasionally as far cranial as T9/10 (Wilkens et al., 1996).
Non-chondrodystrophoid breeds are affected less
Back pain and neurological deficits in the pelvic limbs

are features of thoracolumbar disc disease; urinary
dysfunction may occur with more severe lesions (see
page 31). The back pain is usually less dramatic than
that associated with cervical disc disease. The dog may
show kyphosis and a reluctance to run or jump; discomfort can usually be elicited by palpation over the
thoracolumbar region. Pain without neurological deficits
can be misinterpreted as being cervical, orthopedic or
abdominal in origin (Table 2.1, Table 3.6, Box 7.2).
The pain in thoracolumbar disc disease arises because
of combinations of anulus fibrosus and dorsal longitudinal ligament damage with associated meningeal and
nerve root irritation. Neurological deficits range from
mild ataxia and paraparesis to paraplegia, which may be
accompanied by depressed or absent nociception caudal to the lesion (Olby et al., 2001) (see Assessing the
severity of the lesion, page 31).
Neurological deficits usually become more severe
with increasing spinal cord compression (see Spinal cord
nerve fibers and the effect of compression, page 6).
In addition to the mass effect of the disc material, the
rate at which spinal cord compression occurs is also very
important. In extreme cases there may be a combination
of mass effect and impact injury to the spinal cord,
resulting from explosive disc rupture. In the majority of
dogs the deficits are upper motor neuron (UMN) in
nature and there may be an associated cutaneous trunci
reflex cut-off (see page 24). Approximately 1015% of
dogs show lower motor neuron (LMN) deficits because
of lesions between L3/4 and L6/7 discs inclusive (see
1.5). The most common LMN sign is loss of the patellar
reflex indicating L5 (range L4 to L6) spinal cord segment
involvement, usually due to L3/4 disc herniation.
Differential diagnoses are listed in Box 8.1.
122
Box 8.1 Differential diagnoses for thoracolumbar

disc disease and for back pain
Lumbar synovial cyst
Lumbar facet joint pain
Congenital, e.g. arachnoid cyst
Neoplasia
Tumoral calcinosis
Discospondylitis
Inflammatory CNS disease
Epidural abscess
Polyarthritis
Polymyositis
Trauma
Dural tear
Ischemic myelopathy
Psoas muscle injury
Abdominal and pelvic pain due to:
Pancreatitis
Renal pain
Ureteral calculi
Gallstones
Gastrointestinal parasites
Kidney worm
Prostatic disease
Urethral tumor
DIAGNOSIS
Radiography
Survey radiographs may indicate if disc disease is present but are only 6070% accurate in identifying the
exact location (Kirberger et al., 1992; Olby et al.,
1994). Survey radiographs must not be used as the sole
means of confirming the diagnosis if decompressive
surgery is planned. The main roles of radiography are to
help rule out differential diagnoses (Box 8.1) and to
confirm anatomical landmarks (8.198.21).
B
8.1 A: Lateral myelogram from a dog with a disc extrusion at
T12/13. There is poor filling of the subarachnoid space over T12
vertebral body because of a ventral, extradural mass. Filling can
often be improved by injecting contrast as the radiograph is
being taken. B: The ventrodorsal view of the same dog reveals
poor contrast filling of the subarachnoid space over T12 and
T13 vertebral bodies. The mass is mainly left-sided over the
T12/13 space (arrow) (from Olby et al., 2000).
do show mild abnormalities (Thomson et al., 1989).

Routine collection and analysis from the CMC insures
that the clinician is able to detect meningitis. Analysis
of CSF may show that imaging is unnecessary or that
myelography is contraindicated and it also provides a
sample for further diagnostic testing should the myelogram prove to be non-diagnostic.
Cerebrospinal fluid analysis

Ideally, CSF should be collected routinely from the
cerebello-medullary cistern (CMC) unless fluid can be
obtained from the lumbar region prior to contrast injection. CSF taken after myelography is almost impossible
to interpret as most contrast agents induce a short-term,
sterile meningitis (Widmer et al., 1992) (see page 43).
Analysis of CSF can assist in refining the differential
diagnosis (Box 8.1) although some dogs with disc disease
Myelography
Either myelography or advanced imaging should be
performed for definitive diagnosis. A lumbar injection
is preferred for myelography because there is often
considerable spinal cord swelling, which tends to cause
cervical myelograms to stop cranial to the lesion (see
Chapter 4, page 71). Lateral and ventrodorsal images
should be taken (8.1A, B). If it is not clear on which
8.2 A: CT scan from the dog shown

in 8.1. This image, performed prior to
myelography, shows a centrally located,
mineralized mass occupying much of
the vertebral canal (arrows) (from Olby
et al., 2000). B: Sagittal T2-weighted
MRI through the lumbar region of an
11-year-old paraparetic Labrador with
disc herniations at T12/13L3/4; L2/3
and L3/4 have decreased signal intensity
(see 4.43). Note the nutrient arteries
entering L4 (arrow) (Parker, 1973).
B
8.3 A: Transverse CT image and B: 3D
reconstruction from different dogs, each
demonstrating disc material occupying
both sides of the vertebral canal
(arrows). These images show why it can
sometimes be hard to judge on which
side to perform surgery (Schulz et al.,
1998; Grevel and Schwartau, 1997).
side the disc material is located, oblique views or a CT

myelogram should be used.
some disc herniations that are only incidental findings

and that are not responsible for causing clinical signs
(Milette et al., 1999; Olby et al., 2000) (see Chapter 4).
CT and MR imaging
CT is more accurate and usually much faster than myelography, especially in chondrodystrophoid breeds (Olby
et al., 1999). CT is non-invasive as mineralized disc
material shows clearly without the need for contrast, even
when it is not visible on survey radiographs. It is usually
much easier to decide what side(s) the disc material is
on from a CT than from a myelogram (see Chapter 4,
page 55; 8.2A, 8.3). A heterogenous, hyperattenuating,
extradural mass with loss of epidural fat are characteristic features of mineralized disc extrusions (Olby et al.,
2000) (4.37B, 4.40A, 8.3A). Contrast medium may be
necessary if the extruded material is not mineralized
(4.40A, 11.52). CT has become increasingly popular
with neurologists for diagnosing disc disease in chondrodystrophoid breeds and is the modality of choice for
many. It does not show the extent of spinal cord swelling
as well as myelography but, conversely, swelling does not
degrade the CT image. If a CT is not diagnostic it can
always be followed by a myelogram. Scanners can also
be used to show the exact location and extent of disc
material on a scout image (4.39, 4.44A).
MRI also provides transverse imaging and is superior to
CT when the disc material is not mineralized (4.42, 4.43,
8.2B). The increased sensitivity of CT and MRI reveal
TREATMENT
Many dogs will recover from moderate neurological
deficits following either non-surgical or surgical treatment. Certain generalizations can be made regarding
the advantages of each type of therapy. An algorithm
for surgical decision-making is shown in Algorithm 8.1.
Patients with marked deficits (see Assessing the
severity of spinal cord injury, page 31) seen within 8 h
of spinal cord injury may benefit from concomitant
methylprednisolone sodium succinate (MPSS) therapy
(see page 83).
Non-surgical
Strict cage rest is the overriding principle here although
judicious use of analgesics or anti-inflammatory drugs
may also be needed. These drugs should be withheld
when feasible in order to encourage the animal to rest.
The animal must rest quietly in a confined space (traveling cage size) for at least 4 weeks, during which time
it should only be removed to urinate and defecate.
A satisfactory response should be followed by a further
2 weeks rest and a gradual increase in activity between
123
124
Neurological examination
Grades 1 or 2
First episode
Grades 3 or 4
Grade 5 or
rapid
progression
Persistent or
recurrent
Neuroimaging
Rest and monitor
Elective
neuroimaging
URGENT
ELECTIVE
Fenestration alone
Decompression
Large dog
Acute
Pediculectomy
mini-hemilaminectomy, or
hemilaminectomy and stabilize
DO NOT FENESTRATE
AFFECTED DISC
Small dog
Chronic
Pediculectomy
mini-hemilaminectomy, or
hemilaminectomy and stabilize
DO NOT FENESTRATE
AFFECTED DISC?
Acute
Decompress in
the least
invasive way
FENESTRATE
Algorithm 8.1 Surgical decision-making in thoracolumbar disc disease.
Table 8.1 Results of treatment for thoracolumbar disc disease (See also Table 8.1a)
Treatment as % success (no. of dogs)
Neurological grade
Conservative4
Decompression1,2,5,6,7,8
Fenestration3
1no deficits
100 (8/8)
97 (29/30)
92 (1/12)
2paresis walking
84 (32/38)
95 (36/38)
93 (40/43)
3paresis not walking
84 (32/38)
93 (43/46)
85 (22/26)
4paraplegia
81 (13/16)
95 (37/39)
88 (1/8)
64 (86/135)*
33 (2/6)
5no deep pain

1
7 (1/14)
Anderson et al., 1991; 2 Black, 1988; 3 Butterworth and Denny, 1991; 4 Davies and Sharp, 1983; 5 Olby et al., 2003;
Scott and McKee, 1999; 7 Sukhiani et al., 1996; 8 Yovich et al.,1994.
*See Table 8.1a.
the 6th and 8th weeks. This is the minimum time needed
for an avascular structure like the anulus fibrosus to
repair. Activities like jumping should be avoided for
46 months.
Animals that will not rest or are confined inadequately
may fail to respond or get worse. The patient must be
evaluated regularly for any deterioration, which indicates
treatment failure, as does a lack of improvement within
2 weeks.
Advantages of non-surgical treatment are minimal
expense and equipment-needs. Treatment can be continued at home, ideally after an initial few days of direct
observation. Overall recovery rates are good for dogs
with grade 1 to 3 deficits (Table 8.1). About 50% of
paraplegic animals that are also incontinent will recover,
Table 8.1a Results of hemilaminectomy for dogs with no

deep pain (grade 5)percentage (nos)
Deep pain checked in all dogs at
follow-up*
57 (38/67)2,3
Deep pain not checked in all

dogs at follow-up
71 (48/68)1,4,5
Total grade 5 dogs recovering
64 (86/135)1,2,3,4,5
Anderson et al., 1991; 2 Olby et al., 2003; 3 Scott, 1997; 4 Scott and
McKee, 1999; 5 Yovich et al., 1994.
*An additional 8/67 dogs (12%) regained voluntary motor function
without regaining deep pain (Olby et al., 2003).
but non-surgical treatment is ineffective for dogs with

grade 5 lesions (Davies and Sharp, 1983).
Although a useful initial option for some dogs with
grade 1 or 2 lesions, non-surgical therapy is rarely the
treatment of choice for dogs with grade 3 lesions or
worse, especially if there are no financial constraints. The
major long-term problem is that over one third of dogs
will suffer recurrence (Table 8.2). Another disadvantage
is that the dog can deteriorate during treatment, possibly
as far as grade 5, often due to poor owner compliance.
In addition, there is a natural tendency to minimize the
diagnostic evaluations for a dog to be treated by cage
rest and consequently other causes for the neurological
deficits can be overlooked. Physical therapy must also
be delayed until the latter part of the treatment period
and recovery of the neurological deficits may be slow
or incomplete. Dogs that suffer recurrences after nonsurgical treatment may also have more severe deficits
compared to dogs suffering recurrences after surgical
therapy (Davies and Sharp, 1983; Dhupa et al., 1999a).
A short course of corticosteroids without cage rest does
not constitute effective non-surgical treatment. A high
proportion of dogs referred for emergency decompression have been treated in the preceding days or weeks
using corticosteroids without cage confinement (8.4).
Humans experience a euphoric effect when on steroids
and a similar phenomenon may also make dogs more
active (Swinburn et al., 1988; Assimes and Lessard,
1999). Unrestrained activity renders the dog susceptible
to further herniation of disc material and severe neurological deficits.
Table 8.2 Recovery times and recurrence rates after treatment for thoracolumbar disc disease
Mean recovery time in weeks
Neurological grade
Conservative4
Decompression7,8,9,10
Fenestration2
1no deficits
2paresis walking
3paresis not walking
2 (25, 673)
68
4paraplegia
912
14
68
5no deep pain
N/A
510 (59, 77, 363)
10
Recurrence of signs (%) (see Prognosis)
344, 406
193,10, 166, 138, 67, 41*
04, 22, 156
Brisson et al., 2002; 2 Butterworth and Denny, 1991; 3 Cudia and Duval, 1997; 4 Davies and Sharp, 1983; 5 Davis and Brown, 2001; 6 Levine and
Caywood, 1984; 7 Olby et al., 2003; 8 Scott, 1997; 9 Scott and McKee, 1999; 10 Yovich et al., 1994.
*After hemilaminectomy with fenestration of a variable number of disc spaces.
N/A, not available.
125
126
8.4 Paraparetic Dachshund given large doses of corticosteroids

without confinement. The dog had diarrhea, its packed cell
volume (PCV) was 21% and its serum albumin 2.4 g/dl
(normal 3.0). PCV was 15% the next day and the albumin
1.7 g/dl, presumably secondary to corticosteroid-induced
gastrointestinal bleeding. Melena was only evident after 4 days.
The dog required 7 days intensive care prior to laminectomy at
T11/12. One year previously it had been decompressed at
T13/L1 and fenestrated from T12/13 to L3/4 inclusive.
Decompression (hemilaminectomy,
mini-hemilaminectomy or pediculectomy)
Although decompression is the treatment of choice for
disc disease it does require good-quality imaging. Imaging
usually identifies the affected interspace but determining which side to decompress can be more problematic.
Myelography is more reliable than neurological signs,
but CT or MR imaging are superior (Smith et al., 1997;
Schulz et al., 1998; Olby et al., 1999). This is because
spinal cord swelling can confound myelographic interpretation regarding the side of the lesion and because
the disc material may actually be on both sides (Schulz
et al., 1998) (8.3). Good surgical technique is needed
to retrieve as much disc material as possible once the
spinal cord has been exposed. Any material on the contralateral side must be removed by probing over or under
the spinal cord; alternatively bilateral decompression
may be needed (8.3). Bilateral hemilaminectomy does
not appear to prejudice the outcome in small dogs
despite causing an increase in range of motion or even
overt instability (Anderson et al., 1991; Shires et al.,
1991; Grevel and Schwartau, 1997; Corse et al., 2002;
Viguier et al., 2002). Instability will be worsened by
concomitant fenestration (Shires et al., 1991). Preservation of facet joints is therefore recommended on at
least one side (Yovich et al., 1994).
Decompression is the treatment of choice for dogs with
spinal cord compression causing persistent or recurrent
grade 1 signs and for most dogs with neurological
deficits. The rate of recovery is faster after decompression
than after either non-surgical treatment or fenestration

(Table 8.2) and there is less likelihood of residual neurological deficits (see Prognosis, page 132). Corticosteroids provide no overall benefit when used with
decompressive surgery, except possibly MPSS for dogs
that present within the first 8 h of injury (Olby, 1999)
(see page 83). Disadvantages of decompression relate
mainly to the need for advanced imaging and for special
equipment and expertise (8.118.62). Specific recommendations for decompression include:
Progressive, persistent, or recurrent clinical signs.
Evidence of severe spinal cord compression.
Grade 5 lesionsthese should be decompressed
as soon as possible (see page 132).
Concomitant fenestration should be performed at
the time of the decompression (8.53). Probable exceptions include chronic disc herniations or extrusions in
large-breed dogs because of the destabilizing effect
(Shires et al., 1991) (8.51, 8.52). As many discs as possible should be fenestrated in chondrodystrophoid
breeds, especially the discs between T11/12 and L2/3
inclusive. Decompression without fenestration can
result in recurrence rates as high as 32% (Levine and
Caywood, 1984; McKee, 1992). Up to 10% of dogs
where adjacent discs were not fenestrated routinely at
the time of decompression subsequently require another
decompression due to a second herniation (Smith et al.,
1997; Dhupa et al., 1999a). The re-operation rate falls
to 4% when prophylactic fenestration of adjacent discs
is performed (Brisson et al., 2002).
Dorsal laminectomy is not recommended as it has
no advantages over hemilaminectomy and causes
considerably more biomechanical instability (Smith
and Walter, 1988). It may also increase intradiscal
pressure, which could affect recurrence rates
adversely (Lin et al., 1978; Shires et al., 1991).
Hemilaminectomy gives better access to extruded
disc material than dorsal laminectomy and also
makes fenestration easier (McKee, 1992; Muir et al.,
1995). Compared to standard hemilaminectomy,
pediculectomy and mini-hemilaminectomy have
the advantage of preserving the articular processes
(8.57), which is likely to retain stability especially
when the disc is also fenestrated (Shires et al.,
1991; Hill et al., 2000; Viguier et al., 2002). The
decreased bone removal should also reduce surgery
time and overall morbidity (Jeffery, 1988; Lubbe
et al., 1994; McCartney, 1997).
Mini-hemilaminectomy removes less bone than a
standard hemilaminectomy but only gives good
access to the ventral portion of the vertebral canal
(8.578.59). The greater tendency for hemorrhage
when working around the foramen can also
exacerbate the decreased exposure with

mini-hemilaminectomy.
Pediculectomy also preserves the facet joint and
removes less bone than a standard hemilaminectomy
(8.608.61). In contrast to a mini-hemilaminectomy,
pediculectomy avoids the region of the foramen and
its vessels. However, it also gives restricted access,
especially as most disc material tends to be
concentrated over the foramen (8.61B). To improve
access, one or two pediculectomies can be merged
with a mini-hemilaminectomy (Biggart, 1988; Lubbe
et al., 1994; McCartney, 1997) (8.62). This combination can then readily be expanded to a standard
hemilaminectomy if additional exposure is necessary.
Fenestration without decompression

Some surgeons believe that fenestration alone has no
merit as the sole surgical procedure. Decompression is
the preferred treatment for extradural compression
caused by disc disease because even dogs that present
with back pain alone usually have significant spinal cord
compression (Sukhiani et al., 1996). There are, however,
a few situations in which to consider fenestration alone:
Presumed discogenic pain.
Recurrent back pain with minimal spinal cord
compression.
Recurrent, mild ataxia and paresis associated with
multiple small herniations.
When fenestration is not combined with decompression it is done most easily in small dogs using a lateral
approach. Although relatively straightforward in principle, fenestration requires a thorough understanding of
anatomy and, like decompression, should be undertaken only after careful preparation and practice on
cadavers. It is usual to fenestrate the discs from T11/12
to L3/4 via the lateral approach (8.638.78). The more
caudal lumbar discs should be fenestrated in dogs with
LMN deficits, taking particular care not to damage the
large ventral branches of the spinal nerves at this level.
The main advantage of fenestrating the majority
of high-risk discs is that it reduces recurrence rates
(Table 8.2). There is also no need for special instrumentation. In comparison to non-surgical therapy, fenestration has the added advantage that physiotherapy
can be instituted immediately.
The main disadvantages of fenestration are that
recovery rates and times are prolonged for dogs with
grade 3 and 4 deficits compared to decompression
(Butterworth and Denny, 1991; Davies and Brown,
2001) (Tables 8.1, 8.2). Results are much worse than
decompression for dogs with grade 5 deficits (Table 8.1).
Residual neurological deficits are also more common
than after decompression. Therefore fenestration alone
cannot be regarded as the treatment of choice for dogs

with marked spinal cord compression. Other potential
problems are that improper technique can occasionally
force more disc material into the vertebral canal (8.77)
and fenestration of chronic discs or in large dogs may
decrease stability (Shires et al., 1991) (8.52).
COMPLICATIONS
Complications common to both surgical and non-surgical
management include gastrointestinal (GI) ulceration,
iatrogenic hyperadrenocorticism, pancreatitis, pulmonary thromboembolism (PTE), deep vein thrombosis
(DVT), decubitus, urine scald, urine retention and
urinary tract infection (UTI) (see Chapters 6 and 15).
Intraoperative complications
These are listed in Box 8.2. The main problems are
finding the correct disc space and removing all of the
disc material. The problems are discussed below.
Box 8.2 Intraoperative complications
Improper identification of the surgical site (8.6, 8.21)
Inability to find disc material (8.6, 8.46, 8.61)
Inadequate removal of disc material (8.42, 8.61B)
Diffuse disc material (8.5, 8.55)
Adhesion of disc material to the dura (8.51)
Excessive hemorrhage (8.43)
Hemilaminectomy done on wrong side (8.6, 8.46, 8.61B)
Severe spinal cord swelling (8.49, 8.50, 8.55)
Early postoperative complications

Several factors can account for neurological deterioration in the early postoperative period (Box 8.3).
Box 8.3 Early postoperative complications
Myelomalacia (8.5)
Self-mutilation
Inadequate decompression (8.6)
Scoliosis
Second disc extrusion (8.7)
Body wall flaccidity
Fat graft (or Gelfoam) reaction (8.8)
Pneumothorax
Instability
Femoral nerve paralysis
Wound infection
Cutaneous fistula
127
128
Progressive myelomalacia (8.5) affects about 10% of

dogs that present without nociception (Scott and
McKee, 1999; Olby et al., 2003). Most affected dogs
develop paralysis over less than 12 h; occasionally
dogs present with grade 4 deficits but then go on to
develop myelomalacia (Griffiths, 1972). The
condition usually develops within 5 days of initial
paralysis, with a range of 110 days; signs may
therefore only become evident in the postoperative
period. The condition then progresses over 37 days
(Funkquist, 1962; Olby et al., 2003). Warning signs
include depression, anorexia, vomiting, hypotension,
toxemia, profound hyperesthesia, a cutaneous
trunci cut-off that moves cranially, progression from
UMN to LMN deficits and tetraparesis with
abdominal breathing. Myelography reveals diffuse
contrast medium infiltration into the cord
parenchyma (Lu et al., 2002) (14.18). CSF usually
shows very high protein levels even when taken from
the CMC. Widespread malacia with epidural and
subarachnoid hemorrhages develop, although these
are not always evident at the time of surgery
(8.5, 8.50). In most dogs with progressive
8.5 Dorsal laminectomy performed in a dog with clinical

signs of progressive myelomalacia shows a swollen and grossly
hemorrhagic spinal cord (arrow). It is very important to
distinguish this situation from the more common one of focal
malacia shown in 8.50 (see Prognosis, page 132).
myelomalacia the disc material spreads extensively

along the epidural space, often encircling the dura
mater but causing no direct spinal cord compression.
It is likely that release of catecholamines and other
substances causes severe, progressive vasospasm
(Griffiths, 1972). As soon as clinical signs of this
condition are recognized, euthanasia should be
performed on humane grounds as patients usually
progress from hypoventilation to asphyxiation.
Occasionally the condition stops progressing before it
kills the dog. Differential diagnoses for progressive
myelomalacia include any coagulopathy that could
cause intradural hemorrhage.
If little or no disc material is retrieved and the
animal shows no improvement, then repeat imaging
should be considered promptly. It is possible that
the wrong site was decompressed (8.6); more disc
material was hidden on the opposite side of the
spinal cord (Schulz et al., 1998) (8.6); disc material
was missed at surgery (Dhupa et al., 1999a) (8.40,
8.61B); or that the dog is developing progressive
myelomalacia (8.5, 8.55).
Occasionally a dog will herniate a second disc in
the early postoperative period. This can occur as a
complication of improper fenestration (8.7) or
from manipulation and loss of muscle tone under
anesthesia.
Aseptic necrosis of an excessively thick fat graft
can cause deterioration within a few days of
surgery (8.8). Adverse reactions have also been
reported after Gelfoam (Pharmacia, Kalamazoo,
MI) (Muir et al., 1995; Songer et al., 1990).
Some dogs will deteriorate after surgery yet none of
the above factors are present. Iatrogenic damage to
the spinal cord is rare but can occur either during
myelography (4.34) or at surgery. Postoperative
deterioration seems to be a particular problem in
dogs with chronic compression (8.51, 8.52). This
apparent decompensation may be related to poor
8.6 This dog deteriorated after hemilaminectomy the previous day. Myelography
had identified an L2/3 disc; the surgeon
believed that decompression was from L2
to L4 on the right but no disc material was
detected. 3D reconstruction of a CT scan
made 24 h after hemilaminectomy revealed
decompression was actually from L1 to L3.
A: Coronal reconstruction shows material
situated over L2/3 (arrow) at the caudal end
of the decompression (arrowheads). The
line shows the level of the transverse image
in B. B: Transverse view looking forward to
show the residual compression. A left-sided
mini-hemilaminectomy was performed over
L2/3; the dog made an excellent recovery.
8.7 Series of four CT scans from a dog

with acute paraparesis. Images in A and
C were made at presentation; A shows a
disc extrusion at T12/13. Images in B
and D were made when the dog lost
deep pain 36 h after surgery at T12/13;
D shows a second disc extrusion at
T11/12. A: Right-sided, extradural disc
material (arrows) at T12/13. Material
was retrieved by right-sided
hemilaminectomy; fenestration was from
T12/13 to L3/4 inclusive. Deep pain was
present 12 h after surgery but was
absent after 36 h. B: T12/13 at 36 h after
surgery showing the hemilaminectomy
defect with no residual disc material at
this site (the increased opacity was a
hematoma).C: T11/12 shown prior to
surgery, made at the same time as the
image in A. D: T11/12 shown 36 h after
surgery, at the same time as the image
in B. Mineralized disc material is now
visible at T11/12 (arrows); this was not
present prior to surgery. A second
hemilaminectomy revealed new material
from a herniation of the T11/12 disc (not
fenestrated previously).
8.8 This dog had good deep pain 24 h after hemilaminectomy at L1/2 but had lost it after 48 h. Myelogram at 48 h reveals
extradural compression over the hemilaminectomy defect. A: Marked ventral deviation of the dorsal contrast column (arrowheads).
B: Thinning of right contrast column between arrowheads. The fat graft was swollen and edematous (12.10). It was replaced by
Gelfoam (Pharmacia, Kalamazoo, MI) and the dog made a good recovery.
129
130
Table 8.3 Outcome for dogs according to size following decompression for thoracolumbar disc disease
Neurological
grade
As % success (total
number)ALL dogs*
As % success (total
number)LARGE dogs
2paresis, walking
95 (38) 2,5,7
92 (13) 3
3, 4not walking, paraplegia
94 (85) 2,5,7
90 (31) 3
5no deep pain
64 (135) 1,4,5,6,7
25 (4) 3
Anderson et al., 1991; 2 Black, 1988; 3 Cudia and Duval, 1997; 4 Olby et al., 2003; 5 Scott, 1997; 6 Scott and McKee,
1999; 7 Yovich et al., 1994.
*Mostly small-breed dogs.
perfusion in injured spinal cord segments, a

reperfusion injury, or because dogs with chronic
compression may have only just enough surviving
axons to walk and so cannot afford to lose any more
(see page 293, 132) (Blight and Decrescito, 1986;
Cybulski and DAngelo, 1988; Basso et al., 1996;
Jeffery and Blakemore, 1999; Olby et al., 2003).
Another potential cause for either early
deterioration or long-term morbidity is low-grade
instability after facet joint removal. Facet removal
has minimal effects on lateral bending but significant
effects during rotation, especially when it is
combined with fenestration (Shires et al., 1991;
Schulz et al., 1996; Viguier et al., 2002). Unilateral
facetectomy has no effect on strength in studies of
flexion and extension but does cause a significant
increase (11%) in range of motion, which could have
adverse effects on a spinal cord that is already
severely injured (Smith and Walter, 1988). The
biomechanical studies performed to date may not
address all types of forces acting on the spine and
have only been done in normal cadavers (Smith
and Walter, 1988); the effects of severe disc
degeneration on stability are unknown. In addition,
animals with more severe injuries can probably
generate only minimal protection from their
paraspinal muscles. These various factors may be
even more important in large-breed dogs, and may
help to explain their apparent lower recovery rates,
more frequent residual deficits and longer recovery
times following grade 5 lesions compared to smaller
dogs (Cudia and Duval, 1997; Olby et al., 2003)
(Tables 8.2, 8.3). Facet preservation or concomitant
vertebral stabilization are recommended in active,
large-breed dogs (McKee, 2000) (see Prognosis,
page 132). In humans, the more extensive the
removal of bone the greater the subsequent
morbidity (Eule et al., 1999; Papagelopoulos et al.,
1997); therefore surgeries that preserve the facet
joints are a logical development for dogs and are in

keeping with the trend towards microdiscectomy in
humans (Hermantin et al., 1999) (8.6, 8.568.62).
Fenestration can decrease stability, which may be
of clinical relevance in large-breed dogs or at sites
of chronic compression (Shires et al., 1991). It
could also cause collapse of an interspace thereby
increasing compression (McKee, 2000).
Other potential problems during the early
postoperative period include postoperative pain;
wound discharge (14 of 264 dogs or 5%) (Hosgood,
1992); UTI (9 of 36 dogs or 26% recovering after
decompression for grade 5 deficits) (Olby et al.,
2003); or self-mutilation of the feet or penis (Olby
et al., 2003) (see page 359). Scoliosis or flaccidity
of the body wall is reported in up to 10% of dogs
undergoing fenestration (Bartels et al., 1983; Black,
1988) (8.9). Pneumothorax (6 of 127 dogs or 5%)
and femoral nerve paresis (4 of 127 dogs or 3%) may
also occur (Bartels et al., 1983; Sukhiani et al., 1996).
These problems are usually transient (8.9), but
scoliosis can be permanent (Yovich et al., 1994).
Late postoperative complications

The main problems in the late postoperative period are
listed in Box 8.4.
Restrictive, peridural fibrosis or laminectomy scar is

recognized only rarely in dogs (Applewhite et al.,
1999)(see Chapter 6). It is most likely in a dog
that has had a previous, wide, dorsal laminectomy;
if hemilaminectomy was combined with
excessive removal of bone dorsally; following
hemilaminectomy over several interspaces (8.56); or
if there has been a chronic reaction to the material
placed at the surgery site (Muir et al., 1995).
Infection of an intervertebral space may occur,
which can be iatrogenic (Funkquist, 1978) (see
Discospondylitis, page 326, 13.34, 14.14).
8.9 Mild kyphosis and body wall flaccidity (arrowheads) in a

Cocker spaniel the day after a right-sided hemilaminectomy
and fenestration for intervertebral disc disease. Signs resolved
gradually over a few days and were assumed to have been due
to neurapraxic injury to nerve root(s) or spinal nerve(s).
Box 8.4 Late postoperative complications
Peridural fibrosis
Residual neurological deficits
Infection
Fecal incontinence
Disc extrusion at a new site
Recurrent UTI
There may be recurrence of signs, which usually

occurs between 1 month and 2 years after surgery
due to late disc herniation at a new space (Dhupa
et al., 1999a) (8.4). Although still controversial, it
appears that fenestration reduces the recurrence
rate and that the more discs fenestrated, the greater
this reduction (Funkquist, 1978; Fingeroth, 1989;
Yovich et al., 1994; Olby et al., 2003 (Table 8.2).
Twelve of 265 dogs suffered recurrences (4%) in
one study; in 10 out of 12 of these dogs (83%)
recurrence was at an interspace that had not been
fenestrated at the original surgery (Brisson et al.,
2002).
Residual neurological deficits, usually mild
paraparesis or pelvic limb ataxia, affect about
2025% of dogs presenting with severe deficits
(McCartney, 1997; Scott, 1997). This rate appears
to be higher (39%) for large-breed dogs (Cudia and
Duval, 1997) (see Prognosis, page 132). One factor

that may contribute to the residual neurological
deficits in some dogs is inadequate spinal cord
decompression. Postoperative vertebral canal stenosis
can cause development of syringohydromyelia in
humans and may be an overlooked cause of
residual neurological deficits in animals (PerrouinVerbe et al., 1998; Fischbein et al., 1999; Bains
et al., 2001).
Fecal incontinence has been reported in 5 to 39% of
dogs recovering from surgery (Anderson et al., 1991;
Cudia and Duval, 1997; Dhupa et al., 1999a; Olby
et al., 2003). This problem is usually only an
intermittent one. Of 36 dogs with no deep pain that
recovered nociception after surgery, 14 (39%) had
fecal incontinence but the owners only perceived
this to be a problem in three (8%) (Olby et al.,
2003). Mild urinary incontinence also occurred in
most of the dogs that suffered from postoperative
fecal incontinence (31%) (Olby et al., 2003).
Recurrent UTI occurs in some dogs recovering from
grade 5 deficits (Olby et al., 2003). This may be due
to an underlying problem, such as pyelonephritis or
cystic calculi. However, a more important cause of
recurrent UTI is when a dog recovers the ability to
walk without recovering continence or nociception
(also termed spinal reflex walking, see Chapter 6,
page 87). This problem is almost certainly underrecognized. It was generally believed that useful
motor function only recovers once deep pain had
already returned and so most dogs are not checked
for deep pain once they begin to walk (Wheeler and
Sharp, 1994; Oliver et al., 1997). However, 8 of 19
dogs (42%) that never regained deep pain sensation
after surgery still recovered the ability to walk. These
dogs took between 4 and 18 months (mean 9
months) to walk and yet none had deep pain
sensation on re-examination and all suffered from
incontinence and recurrent UTI. Interestingly,
despite these problems the owner of each dog was
happy with the outcome. If all dogs that regained
motor function in this study are considered to have
had a successful outcome, then 44 of the 64 dogs
(69%) presenting with grade 5 lesions recovered
overall (Olby et al., 2003). The small population of
dogs that walk without regaining deep pain could
explain the apparent discrepancy in outcome
between studies that checked all dogs specifically
for recovery of nociception at long-term follow-up
and studies that used telephone follow-up for
some dogs (Table 8.1a). Dogs with recurrent UTI
should therefore be assessed critically for deep
pain sensation.
131
132
POSTOPERATIVE CARE
(see Chapter 15)
Extended cage rest is not usually necessary although it

may be recommended for large-breed dogs, after bilateral hemilaminectomy or after decompression of
chronic disc lesions. Postoperative physical therapy
together with restricted exercise on a leash are indicated (see Chapter 15). Corticosteroids are not recommended other than possibly low-dose prednisone for
analgesia. By far the most important aspect of postoperative care is to insure that the bladder is emptied regularly of all urine. Urinary retention is the most common
postoperative problem in dogs paralyzed due to thoracolumbar disc disease. Such dogs must have regular urinalysis and may require pharmacological intervention
(see Control of urinary function, page 350; Table 15.7).
PROGNOSIS
The prognosis is excellent for dogs with grade 1, 2 and
3 deficits, especially following decompression. Dogs
with grade 3 and 4 deficits have a better outcome after
surgery and the response is better after decompression
than after fenestration alone. Decompression is clearly
the treatment of choice for dogs with grade 5 lesions,
with between 60 and 70% making a functional recovery
in most studies (Tables 8.1, 8.1a). When deep pain was
assessed using bone clamps or pliers the rate was 62%
(23/37) (Scott and McKee, 1999). Specific evaluation
of all dogs for recovery of deep pain, rather than the
ability to walk, also brings the recovery rate close to
60% (Olby et al., 2003) (Table 8.1a). Most dogs that
make a functional recovery have regained deep pain
within 2 weeks. However, 19% (8/42) recovered deep
pain in the 3rd or 4th weeks and so a final assessment
should not be made prior to 1 month post-surgery
(Scott and McKee, 1999; Olby et al., 2003). Recovery
of deep pain may take even longer for large-breed dogs
(Olby et al., 2003).
Studies show conflicting results for both the prognostic
value of the speed of onset of clinical signs and of spinal
cord swelling at myelography (Duval et al., 1996; Scott
and McKee, 1999; Olby et al., 2003). The prognosis
for dogs with LMN signs was no worse than those with
UMN signs (Dhupa et al., 1999b; Olby et al., 2003).
Furthermore, prognosis did not correlate with the duration of signs prior to surgery. Good results were still
obtained if surgery was performed within 48 h, or even
more than 72 h after onset of signs (Anderson et al.,
1991; Scott and McKee, 1999; Olby et al., 2003), which
is in contrast to the recommendations in many standard
textbooks. Taken together, these data suggest that a delay
in decompression of a few hours seems to make little difference to outcome, with the possible exception of dogs
with rapidly progressive signs (Anderson et al., 1991;

Scott and McKee, 1999; Olby et al., 2003). However,
logic would still dictate that early decompression be a
high priority when a dog presents with severe deficits.
Prognosis has been inferred from the presence of
spinal cord liquifaction following a durotomy. However,
euthanasia based on a malacic, toothpaste-like spinal cord
at durotomy almost certainly results in the death of some
dogs that would otherwise have recovered (Olby et al.,
2003) (8.50). In a study where durotomy was routine
and dogs with malacic cords were euthanized, only 20 of
46 dogs (43%) with grade 5 lesions recovered. If the
16 dogs that were euthanized at surgery are excluded,
the recovery rate for the other 30 dogs becomes 66%
(Duval et al., 1996). This is more in line with the overall
recovery rate for dogs with grade 5 lesions shown in
Table 8.1. As few as 510% of axons surviving within a
lesion appear to allow functional recovery and these
would be impossible to identify at durotomy (Blight
and Decrescito, 1986; Basso et al., 1996; Jeffery and
Blakemore, 1999; Olby et al., 2003). Therefore the significance of focal malacia, as well as the difference
between it and the syndrome of progressive myelomalacia, should not be determined at surgery. Dogs or cats with
extensive malacia should only be euthanized based on
either a lack of recovery within 4 weeks or if they develop
clinical signs of progressive myelomalacia (Salisbury and
Cook, 1988; Muir et al., 1995; Olby et al., 2003).
Recurrence rates of clinical signs once a dog has recovered from surgery ranged from 69% (Table 8.2).
Confirmed recurrences at a new disc space (8.4, 8.7)
occur in 58% of dogs but these are probably underestimates of the overall figure (Muir et al., 1995; Smith et al.,
1997; Dhupa et al., 1999a). When a recurrence does
occur it is often at a site that has not been fenestrated
(Levine and Caywood, 1984; Muir et al., 1995; Smith
et al., 1997) (8.4, 8.7). Fenestration is a low-risk procedure that, when combined with decompression, appears
to lower the recurrence rate. However, this assertion
remains to be proven by randomized, prospective studies;
the role of fenestration in chronic disc disease also needs
to be defined (Levine and Caywood, 1984; McKee,
1992; Lubbe et al., 1994; Yovich et al., 1994; Brisson
et al., 2002; Olby et al., 2003) (8.51, 8.52). Fortunately,
if dogs do require a second decompression for recurrent
extrusion the prognosis appears to be no worse than after
the first surgery (Dhupa et al., 1999a).
Although the recovery rate for large-breed dogs is
reported as 91%, this only applies to dogs that present
with good deep pain (grades 24) (Cudia and Duval,
1997). Results for dogs with grade 5 lesions seem to be
much worse. Only 1 of 4 large dogs (25%) with grade 5
signs recovered compared to 86 of 135 small dogs (64%,
8.10 Lateral myelogram of a cat with a

disc extrusion causing extradural
compression at T13/L1. The CT image
taken at this interspace following
myelography shows a mineralized,
left-sided extradural mass (arrows).
Table 8.3). Recovery times after grade 5 lesions also

appear to be prolonged for large-breed dogs, with a mean
of 36 weeks compared to 57 weeks for small breeds
(Cudia and Duval, 1997; Scott and McKee, 1999; Olby
et al., 2003) (Table 8.2). Furthermore, of the 41 largebreed dogs that recovered following decompression,
16 (39%) had residual neurological deficits compared
to 2025% of small-breed dogs (Cudia and Duval, 1997;
McCartney, 1997; Scott, 1997) (Table 8.2). In addition,
8 of the 41 large dogs (19%) that recovered then
suffered a recurrence of signs, of which 5 became paralyzed (Cudia and Duval, 1997). This recurrence rate is
higher than the rates from most other studies (Table 8.2).
If this trend towards disappointing results is borne out
in subsequent studies of large-breed dogs then results
might be improved through facet joint preservation
using pediculectomy or mini-hemilaminectomy, provided that all disc material can be removed. An alternative is for some form of stabilization to be applied
following a standard hemilaminectomy. In addition, fenestration should probably not be performed at the
affected site as it can compromise stability (Shires et al.,
1991; McKee, 2000).
The most important points relating to prognosis are
that:
The clinician is able to assess the neurological
status of the dog accurately.
Any dog presenting with, or subsequently developing,
a grade 5 neurological status should undergo
decompression as soon after presentation as possible.
If deep pain does not return within 4 weeks the
prognosis for full recovery is poor (Olby et al., 2003).
MPSS produces only minor improvements in
human spinal cord injury and is unlikely to cause
significant improvements in outcome for dogs or
cats (Hurlbert, 2000) (see page 83). It may help
some animals with grade 5 lesions but it must be
combined with surgery.
Other corticosteroids are contraindicated as they
may worsen neurological outcome and increase the
complication rate (see page 83). They relieve pain

and inflammation but do not lessen spinal cord
injury. Recovery from spinal cord injury takes time;
this should be combined with decompression
where indicated.
THORACOLUMBAR DISC
DISEASE IN CATS
Neurological deficits caused by disc disease in cats are
more common than was thought previously. Affected
cats tend to be older but often have acute, type I
extrusions (8.10). Cats with neurological deficits
may not show back pain. Lesions are visible on CT
scan although subarachnoid contrast may be helpful.
Differential diagnoses to be considered include trauma,
neoplasia (lymphoma), inflammatory CNS disease
(feline infectious peritonitisFIP) and ischemic neuromyopathy (see Chapter 14). A diagnosis of feline disc
disease should only be made after thorough patient evaluation and neuroimaging (Knipe et al., 2001; Munana
et al., 2001). The response to surgery appears to be
equivalent to that reported in dogs.

1. Is there any clinical advantage for either small- or
large-breed dogs in preserving the articular facets?
2. Does durotomy provide any additional decompression for
spinal cord swelling?
3. Does the long-term prognosis depend on the extent of
bone removal after extensive hemilaminectomy?
4. How effectively do we remove 100% of extruded disc
material and does this affect outcome?
5. Should a chronic disc be fenestrated; how do we define
chronic disc?
133
134
REFERENCES
Anderson, S.M., Lippincott, C.L., Gill, P.J. (1991) Hemilaminectomy in
dogs without deep pain perception. A retrospective study of 32 cases.
California Veterinarian 45, 2428.
Applewhite, A.A., Wilkens, B.E., McDonald, D.E., Radasch, R.M.,
Barstad, R.D. (1999) Potential central nervous system complications of
von Willebrands disease. Journal of the American Animal Hospital
Assimes, T.L., Lessard, M.L. (1999) The use of perioperative corticosteroids in craniomaxillofacial surgery. Plastic Reconstructive Surgery
103, 313321; quiz 322.
Bagley, R.S., Harrington, M.L., Pluhar, G.E., Keegan, R.D., Greene, S.A.,
Moore, M.P., Gavin, P.R. (1996) Effect of craniectomy/durotomy alone
and in combination with hyperventilation, diuretics, and corticosteroids
on intracranial pressure in clinically normal dogs. American Journal of
Bains, R.S., Althausen, P.L., Gitlin, G.N., Gupta, M.C., Benson, D.R.
(2001) The role of acute decompression and restoration of spinal alignment
in the prevention of post-traumatic syringomyelia: case report and
review of recent literature. Spine 26, E399402.
Bartels, K.E., Creed, J.E., Yturraspe, D.J. (1983) Complications associated
with the dorsolateral muscle-separating approach for thoracolumbar disk
fenestration in the dog. Journal of the American Veterinary Medical
Basso, D.M., Beattie, M.S., Bresnahan, J.C. (1996) Graded histological and
locomotor outcomes after spinal cord contusion using the NYU weightdrop device versus transection. Experimental Neurology 139, 244256.
Biggart, J.F., III (1988) Prevention of laminectomy membrane by free fat
grafts after laminectomy in dogs with disk herniations. Veterinary Surgery
17, 2829.
Bitetto, W.V., Thacher, C. (1987) A modified lateral decompressive technique for treatment of canine intervertebral disk disease. Journal of the
Black, A.P. (1988) Lateral spinal decompression in the dog; a review of
39 cases. Journal of Small Animal Practice 29, 581588.
Blight, A.R., Decrescito, V. (1986) Morphometric analysis of experimental
spinal cord injury in the cat: the relation of injury intensity to survival of
myelinated axons. Neuroscience 19, 321341.
Braund, K.G., Taylor, T.K.F., Ghosh, P., Sherwood, A.A. (1976) Lateral spinal
decompression in the dog. Journal of Small Animal Practice 17, 583592.
Brisson, B.A., Moffat, S.L., Parent, J.M. (2002) Canine intervertebral disc
disease; retrospective evaluation of 265 cases: 19951999. Veterinary
Surgery 31, 500.
Butterworth, S.J., Denny, H.R. (1991) Follow-up study of 100 cases with
thoracolumbar disc protrusions treated by lateral fenestration. Journal of
Corse, M.R., Renberg, W.C., Friis, E.A. (2002) Biomechanical effects of
multiple hemilaminectomy in the canine lumbar spine. Veterinary Sugery
31, 500.
Cudia, S.P., Duval, J.M. (1997) Thoracolumbar intervertebral disk disease
in large, nonchondrodystrophic dogs: a retrospective study. Journal of the
Cybulski, G.R., DAngelo, C.M. (1988) Neurological deterioration after
role of spinal cord ischemia. Journal of Neurology 51, 717718.
Davies, J.V., Sharp, N.J.H. (1983) A comparison of conservative treatment
and fenestration for thoracolumbar intervertebral disc disease in the dog.
Davis, G.J., Brown, D.C. (2001) Prognostic indicators for time to ambulation in non-ambulatory dogs with acute thoracolumbar disk extrusions:
115 cases. Veterinary Surgery 30, 491.
Dhupa, S., Glickman, N., Waters, D.J., Sarit, D. (1999a) Reoperative neurosurgery in dogs with thoracolumbar disc disease. Veterinary Surgery
28, 421428.
Dhupa, S., Glickman, N.W., Waters, D.J., Dhupa, S. (1999b) Functional
outcome in dogs after surgical treatment of caudal lumbar intervertebral
disk herniation. Journal of the American Animal Hospital Association 35,
323331.
Duval, J., Dewey, C., Roberts, R., Aron, D. (1996) Spinal cord swelling as
a myelographic indicator of prognosis: a retrospective study in dogs with
intervertebral disc disease and loss of deep pain perception. Veterinary

Surgery 25, 612.
Eule, J., Breeze, R., Kindt, G. (1999) Bilateral partial laminectomy: a
treatment for lumbar spinal stenosis and midline disc herniation.
Surgical Neurology 52, 329337; discussion 337338.
Fingeroth, J.M. (1989) Fenestration. Pros and cons. Problems in Veterinary
Medicine 1, 445466.
Fischbein, N.J., Dillon, W.P., Cobbs, C., Weinstein, P.R. (1999) The presyrinx state: a reversible myelopathic condition that may precede
syringomyelia. American Journal of Neuroradiology 20, 720.
Funkquist, B. (1962) Thoraco-lumbar disk protrusion with severe cord
compression in the dog. I: Clinical and pathoanatomical observations
with special reference to the role of development of the symptoms of
motor loss. Acta Veterinaria Scandinavia 3, 256274.
Funkquist, B. (1978) Investigations of the therapeutic and prophylactic
effects of disc evacuation in cases of thoraco-lumbar herniated discs in
dogs. Acta Veterinaria Scandinavica 19, 441457.
Grevel, V., Schwartau, K. (1997) Hemilaminectomy in thoracolumbar disc
disease in the dog. Part 2: Intraoperative findings and surgical results.
Kleintierpraxis 42, 173196.
Griffiths, I.R. (1972) The extensive myelopathy of intervertebral disc protrusions in dogs (the ascending syndrome). Journal of Small Animal
Henry, A., Tunkel, R., Arbit, E., Ku, A., Lachmann, E. (1997) Tethered
thoracic cord resulting from spinal cord herniation. The Archives of
Physical Medicine and Rehabilitation 78, 530533.
Hermantin, F., Peters, T., Quartararo, L., Kambin, P. (1999) A prospective,
randomized study comparing the results of open discectomy with those
of video-assisted arthroscopic microdiscectomy. The Journal of Bone and
Joint SurgeryAmerican volume 81, 958965.
Hill, T.P., Lubbe, A.M., Guthrie, A.J. (2000) Lumbar spine stability following hemilaminectomy, pediculectomy, and fenestration. Veterinary
Holmberg, D.L., Palmer, N.C., Vanpelt, D., Willan, A.R. (1990) A comparison of manual and power-assisted thoracolumbar disc fenestration in
Hosgood, G. (1992) Wound complications following thoracolumbar
laminectomy in the dog: a retrospective study of 264 procedures.
inappropriate standard of care. Journal of Neurosurgery 93 (1 Suppl), 17.
Jeffery, N.D. (1988) Treatment of acute and chronic thoracolumbar disc
disease by mini hemilaminectomy. Journal of Small Animal Practice
29, 611616.
Jeffery, N.D., Blakemore, W.F. (1999) Spinal cord injury in small animals 2.
Current and future options for therapy. Veterinary Record 145, 183190.
Kirberger, R.M., Roos, C.J., Lubbe, A.M. (1992) The radiological diagnosis
of thoracolumbar disc disease in the Dachshund. Veterinary Radiology
Knipe, M.F., Vernau, K.M., Hornof, W.J., LeCouteur, R.A. (2001)
Intervertebral disc extrusion in six cats. Journal of Feline Medicine and
Surgery 3, 161168.
Levine, S.H., Caywood, D.D. (1984) Recurrence of neurological deficits in
dogs treated for thoracolumbar disk disease. Journal of the American
Lin, H.S., Liu, Y.K., Adams, K.H. (1978) Mechanical response of the lumbar intervertebral joint under physiological (complex) loading. The
Journal of Bone and Joint SurgeryAmerican volume 60, 4155.
Lu, D., Lamb, C.R., Targett, M.P. (2002) Results of myelography in seven dogs
with myelomalacia. Veterinary Radiology and Ultrasound 43, 326330.
Lubbe, A.M., Kirberger, R.M., Verstraete, F.J.M. (1994) Pediculectomy
for thoracolumbar spinal decompression in the Dachshund. Journal of
Maroon, J.C., Abla, A., Bost, J. (1999) Association between peridural scar
and persistent low back pain after lumbar discectomy. Neurological
Research 21, Suppl 1, S4346.
McCartney, W. (1997) Partial pediculectomy for the treatment of thoracolumbar disc disease. Veterinary and Comparative Orthopaedics and
Traumatology 10, 117121.
McKee, M. (2000) Intervertebral disc disease in the dog: 2. Management
options. In Practice 22, 458471.
McKee, W.M. (1992) A comparison of hemilaminectomy (with concomitant disc fenestration) and dorsal laminectomy for the treatment of
thoracolumbar disc protrusion in dogs. Veterinary Record 130, 296300.
Moissonnier, P., Carozzo, C., Meheust, P. (2002) Lateral corpectomy
as a treatment of chronic disc herniation in 15 dogs. Veterinary Surgery
31, 296.
Muir, P., Johnson, K.A., Manley, P.A., Dueland, R.T. (1995) Comparison of
hemilaminectomy and dorsal laminectomy for thoracolumbar intervertebral
disc extrusion in Dachshunds. Journal of Small Animal Practice 36,
360367.
Munana, K.R., Olby, N.J., Sharp, N.J., Skeen, T.M. (2001) Intervertebral
disk disease in 10 cats. Journal of the American Animal Hospital
Olby, N.J., Dyce, J., Houlton, J.E.F. (1994) Correlation of plain radiographic
and lumbar myelographic findings with surgical findings in
thoracolumbar disc disease. Journal of Small Animal Practice 35, 345350.
Olby, N.J., Munana, K.R., Sharp, N.J.H., Flegel, T., Van Camp, S.,
Berry, C.R., Thrall, D.G. (1999) A comparison of computed tomography
and myelography in the diagnosis of acute intervertebral disc disease in
Olby, N.J., de Risio, L., Munana, K.R., Wosar, M.A., Skeen, T.M.,
Sharp, N.J.H., Keene, B.W. (2001) Development of a functional scoring
system in dogs with acute spinal cord injuries. American Journal of
Long-term functional outcome of dogs with severe spinal cord injuries.
Oliver, J.E., Lorenz, M.D., Kornegay, J.N. (1997) Handbook of Veterinary
Neurology, 3rd edn. Philadelphia: WB Saunders.
Osterholm, J. (1974) The pathophysiological response to spinal cord injury.
The current status of related research. Journal of Neurosurgery 40, 533.
Papagelopoulos, P., Peterson, H., Ebersold, M., Emmanuel, P., Choudhury, S.,
Quast, L. (1997) Spinal column deformity and instability after lumbar
or thoracolumbar laminectomy for intraspinal tumors in children and
young adults. Spine 22, 442451.
13511353.
Perrouin-Verbe, B., Lenne-Aurier, K., Robert, R., Auffray-Calvier, E., Richard,
I., Mauduyt de la Greve, I., Mathe, J.F. (1998) Post-traumatic
syringomyelia and post-traumatic spinal canal stenosis: a direct relationship:
review of 75 patients with a spinal cord injury. Spinal Cord 36, 137143.
Ross, J.S., Robertson, J.T., Frederickson, R.C., Petrie, J.L., Obuchowski, N.,
Modic, M.T., deTribolet, N. (1996) Association between peridural
scar and recurrent radicular pain after lumbar discectomy: magnetic resonance evaluation. ADCON-L European Study Group. Neurosurgery
38, 855861; discussion 861863.
Salisbury, S.K., Cook, J.R., Jr (1988) Recovery of neurological function
following focal myelomalacia in a cat. Journal of the American Animal
Schulz, K.S., Waldron, D.R., Grant, J.W., Shell, L., Smith, G., Shires, P.K.
(1996) Biomechanics of the thoracolumbar vertebral column of dogs
during lateral bending. American Journal of Veterinary Research 57,
12281232.
Schulz, K.S., Walker, M., Moon, M., Waldron, D., Slater, M., McDonald, D.E.
(1998) Correlation of clinical, radiographic, and surgical localization of
intervertebral disc extrusion in small-breed dogs: a prospective study of
50 cases. Veterinary Surgery 27, 105111.
Scott, H.W. (1997) Hemilaminectomy for the treatment of thoracolumbar
disc disease in the dog: a follow-up study of 40 cases. Journal of Small
Scott, H.W., McKee, W.M. (1999) Laminectomy for 34 dogs with thoracolumbar intervertebral disc disease and loss of deep pain perception.
Shires, P.K., Waldron, D.R., Hedlund, C.S., Blass, C.E., Massoudi, L.
(1991) A biomechanical study of rotational instability in unaltered and
surgically altered canine thoracolumbar vertebral motion units. Progress
in Veterinary Neurology 2, 614.
Smith, G.K., Walter, M.C. (1988) Spinal decompressive procedures and
dorsal compartment injuries: comparative biomechanical study in canine
cadavers. American Journal of Veterinary Research 49, 266273.
Smith, J.D., Newell, S.M., Budsberg, S.C., Bennett, R.A. (1997)
Incidence of contralateral versus ipsilateral neurological signs associated
with lateralised Hansen type I disc extrusion. Journal of Small Animal
Songer, M., Ghosh, L., Spencer, D. (1990) Effects of sodium hyaluronate
on peridural fibrosis after lumbar laminotomy and discectomy. Spine 15,
550554.
Sukhiani, H.R., Parent, J.M., Atilola, M.A.O., Holmberg, D.L. (1996)
Intervertebral disk disease in dogs with signs of back pain alone: 25 cases
209, 12751279.
Swinburn, C.R., Wakefield, J.M., Newman, S.P., Jones, P.W. (1988)
Evidence of prednisolone induced mood change (steroid euphoria) in
patients with chronic obstructive airways disease. British Journal of
Clinical Pharmacology 26, 709713.
Trevor, P.B., Martin, R.A., Saunders, G.K., Trotter, E.J. (1991) Healing
characteristics of free and pedicle fat grafts after dorsal laminectomy and
durotomy in dogs. Veterinary Surgery 20, 282290.
Viguier, E., Petit-Etienne, G., Magnier, J., Diop, A., Lavaste, F. (2002)
Mobility of T13-L1 after spinal cord decompression procedures in dogs
(an in vitro study). Veterinary Surgery 31, 297.
Wheeler, S.J., Sharp, N.J.H. (1994) Small Animal Spinal Disorders:
Diagnosis and Surgery, 1st edn. St Louis: Mosby.
Widmer, W.R., DeNicola, D.B., Blevins, W.E., Cook, J.R., Jr, Cantwell, H.D.,
Teclaw, R.F. (1992) Cerebrospinal fluid changes after iopamidol and
metrizamide myelography in clinically normal dogs. American Journal of
Wilkens, B.E., Selcer, R., Adams, W.H., Thomas, W.B. (1996) T9T10
intervertebral disc herniation in three dogs. Veterinary and Comparative
Yovich, J.C., Read, R., Eger, C. (1994) Modified lateral spinal decompression in 61 dogs with thoracolumbar disc protrusion. Journal of Small
135
136
PROCEDURES
Dorsolateral hemilaminectomy (8.118.56)
8.11 Completed hemilaminectomy with extruded disc

material in situ.
Ideal prerequisites to perform this procedure are:
Identification of the affected interspace(s).
Knowledge of how disc material is distributed threedimensionally across the vertebral canal at the
affected interspace(s).
Adequate exposure of the interspace(s) to be

decompressed.
Removal of a minimum of bone in order to access disc material and decompress the spinal cord. In
retrospect, the surgeon could have removed this disc material by mini-hemilaminectomy.
Removal of extruded disc material without traumatizing the spinal cord.
8.11
8.12 Incision site for caudal

thoracic and cranial lumbar
intervertebral spaces. The
incision can be modified
according to the exact
site of compression.
Mini-hemilaminectomy is
made easier if the dog is
rotated somewhat more
to one side than for a
standard hemilaminectomy.
8.12
8.13
Skin and superficial tissues are incised 1 cm

from midline so that the incision does not rest
over the spinous processes on closure.
Subcutaneous fat is reflected for 1 cm on either
side of midline to facilitate closure. This reveals
the lumbodorsal fascia, which is incised on the
near side of each spinous process over five
vertebrae. A periosteal elevator is then used to
lever muscle away from the near side of each
spinous process. Muscular insertions on the
cranial and caudal ends of each process are
cut. One process is exposed already (arrow).
8.13
8.14 Here a second spinous process has been

exposed (arrow). Electrocautery to cut muscular
insertions on the spinous processes reduces
minor hemorrhage during these superficial stages
of vertebral exposure.
8.14
8.15 A Langenbeck or Senn retractor (5.9) is placed

adjacent to the spinous process and pulled
laterally, and slightly cranially, to facilitate
exposure of the dorsal surface of the articular
facet joint.
8.15
8.16 Continued retraction has exposed the articular

processes (arrow) with muscles still attached.
While maintaining traction, the muscular
attachments onto the articular facets are cut
as close as possible to the joint capsule (white
arrows). Bipolar electrocautery is helpful to retard
hemorrhage.
8.16
8.17 The muscular attachments have been severed,

and the isolated articular facets are visible clearly
(arrow). This process is now repeated for one to
two facet joints on each side of the site to be
decompressed. Gelpi retractors are placed to
facilitate exposure.
8.17
137
138
8.18 Landmarks must now be

evaluated carefully to
insure that the final
dissection is done at
the correct location.
Landmarks include the
spinous process of the
anticlinal vertebra (1.20B);
the difference between the
transverse process of the
first lumbar vertebra
8.18
and the head of the last
rib (8.19, 8.20); and the
relationship of the spinous process of the sixth lumbar vertebra to the wings of the ilium (4.15).
8.19 3D reconstruction of a CT scan to show that the

caudally directed, proximal portion of the
thirteenth rib is distinct from the short, somewhat
cranially directed, first lumbar transverse
process. It is crucial that landmarks are
identified accurately prior to further dissection
(8.6). Vertebral anomalies occur in 1015%
of Dachshunds (Jeffery, 1988; Kirberger et al.,
1992) (8.21).
8.19
8.20 The proximal portion of the thirteenth rib is much

longer (arrow) than the short transverse process
of the first lumbar vertebra. The tip of the
transverse process can also be palpated with
an elevator, in distinction to the ribs. The angle
of the last rib and the size of the first transverse
process do vary considerably between animals.
The transverse process is also deeper than the
rib. This dog has a neurovascular bundle running
over the rib.
8.20
8.21 This CT reconstruction is from a dog with an

anomalous rib; the surgeon was misled into
performing the hemilaminectomy at the wrong
interspace. On finding no disc material, the
surgeon rechecked the scout image of the
CT scan and noticed that there was a vestigial
thirteenth rib. This had been misidentified as the
first lumbar transverse process.
8.21
8.22 Once landmarks are ascertained, a Gelpi

retractor is placed dorsal to the articular facet
on either side of the interspace(s) to be
decompressed. Retractors are positioned between
an interspinous space and the epaxial muscles; a
moistened laparotomy sponge is placed over the
muscle for protection prior to distraction.
8.22
8.23 If the anatomy around the chosen intervertebral

space is not clear, a periosteal elevator and dry
gauze sponge can be used to clean away tags of
muscle. This should not be done excessively as
a large area of periosteal irritation could increase
postoperative pain. Here the sponge has clarified
an accessory process.
8.23
139
140
8.24 There is a long, tendinous attachment of the

longissimus muscle (arrowhead) to the accessory
process (arrow) of the vertebra just in front of each
disc. There is frequently a small artery located
dorsal to this process that requires bipolar cautery.
(*) Articular process.
8.24
8.25 Cutting the longissimus tendon (arrowhead)

close to its insertion on the accessory process
(asterisk) by cutting away from the underlying
spinal artery and nerve. The tendon is exposed
by putting pressure on it with an elevator (arrow).
Tendons are cut over one to three
interspace(s).
8.25
8.26 The tendinous attachment to the accessory

process has now been cut and the main
branches of the spinal artery, vein and nerve
appear as a common neurovascular bundle
(arrow). (*) Articular process. Saline-soaked
laparotomy sponges are repositioned in such
a way as not to interfere with drilling.
8.27 Diagram to show the

relationship of the
accessory process to the
intervertebral foramen,
through which the
neurovascular bundle
passes.
8.27
8.26
8.28 Rongeurs are used to remove the articular

processes at the site of entry into the vertebral
canal. Rongeurs can be used to perform the
entire hemilaminectomy if desired; adjacent
spinous processes are grasped with bone
clamps, which are then used to lever the two
vertebrae apart. The rongeurs can then be
introduced carefully into the intervertebral
foramen to start bone removal.
8.29 Relationship of the suction tip (arrow) in the

gutter formed between the vertebral body and
the epaxial muscle mass, the tip of the irrigator
over the hemilaminectomy site, and the drill itself.
Note how the surgeon is holding the drill,
steadying the bur guard with the other hand to
resist excessive downward force. One hand
should also rest gently on the dog, taking care
not to compromise ventilation. The articular
facets have been removed.
8.28
8.29
8.30 Bone debris can be irrigated away while drilling

by using a jet of water as shown (arrow).
Alternatively, a slow drip of saline can be used
to cool the bur with periodic flushing of debris.
Here the surgeon has repositioned a finger
lower down the bur guard to improve fine motor
control.
8.30
8.31 Bleeding from cancellous bone is controlled

using bone wax. Note that cancellous bone
is visible over both pedicles but not centrally
over the inner articular facet joint (arrowheads).
The surgeon will not encounter any cancellous
bone at the level of the joint prior to entering the
vertebral canal.
8.31
141
142
8.32 Bone has been removed more extensively,

revealing the gray/pink cancellous bone (arrows)
of the caudal vertebral body. Drilling is advanced
further over the cranial vertebral body, where
cancellous bone has been removed to reveal the
inner cortical bone plate (asterisk). Note the seam
of cortical bone at the facet joint (arrowheads).
8.32
8.33 The inner cortical bone plate of each vertebral

body has been exposed in this dog (arrows).
The next stage is to drill to eggshell thickness
over the entire hemilaminectomy defect.
8.33
In order to remove the inner cortical bone, the surgeon makes fine movements of the bur as shown in 8.34.
Holding the drill close to the remaining plate of bone the surgeon makes fine movements, approximately 1 mm in
excursion, until the bur is felt to bite and cut away fragments of bone. The softness of the inner plate should be
tested often using the bur as a probe but with the power off. This method of drilling is continued until the inner
plate yields to gentle pressure, at which point it can be broken down with fine instruments. The surgeon should
again note that there is a seam of cortical bone at the level of the articular facet joint that unites the outer and
inner plates with no intervening cancellous bone.
8.34 In this diagram, the surgeon has started drilling for a

mini-hemilaminectomy. The aim should be to perform the least
invasive surgery possible for the distribution of disc material (see
page 126). When feasible, start with a pediculectomy or
mini-hemilaminectomy and either combine the two or extend
subsequently to a standard hemilaminectomy if necessary
(8.57, 8.62).
8.34
8.35 Once through the inner cortical plate, the surgeon may want to
palpate the floor of the canal with instruments to clarify exactly how
the defect needs to be enlarged.
8.35
8.36 Here the thinned, inner cortical plate has been

broken. All bone should be removed from
within the defect before starting to remove disc
material. It is easier to drill all remaining bone
away when disc material still covers the spinal
cord rather than once the cord has moved back
to the edges of the defect.
8.36
8.37 Extruded disc is often

mixed with old
hemorrhage; it is still
covered here by
endosteum (arrows).
The surgeon is using a
8.37
House curette (5.16) to
clear bone from around
B
A
the entire circumference
of the bone window. All force applied to this instrument is directed away from the spinal cord.
143
144
8.38 A: After all bone has been

removed, a thin
endosteum (arrow) often
remains over the extruded
disc material and dura
8.38
(8.37). B: The endosteum
must be broken down
A
B
before disc material can be
accessed. Dura (arrow) is usually much whiter than endosteum. Endosteum is thinner than dura and
tears easily. 8.44 shows the same dog after decompression.
8.39 Instruments are then used

to tease the extruded disc
material (arrows) away
from the dura mater. The
tough, medial portion of
joint capsule (arrowhead)
8.39
may tear or, in bigger dogs,
A
it may need to be cut.
8.40 A space is visible (vertical line) between dura and

bone when significant disc material remains.
The gray mass under the spinal cord is disc
material (arrowhead); it is separating the dura
from the floor of the vertebral canal. 8.62A is
from the same dog and shows how the dura
returns to the floor of the canal once all disc
material has been evacuated.
8.40
8.41 Gentle exploration beneath the spinal cord with a

blunt probe shown in three successive images. This
permits ventrally located material to be retrieved even
from the opposite side of the vertebral canal, as
shown. The suction tip should not touch the dura; a
soft, red rubber catheter tip should be attached to the
metal tip when in close proximity to spinal cord.
8.41
A
8.42 CT reconstruction of a hemilaminectomy defect to show

the idealized motion for retrieving material from
underneath the spinal cord. A long, circular scooping
motion retrieves material better than more linear
movements that often just serve to push material farther
away. The venous plexus (venous sinus) may be
lacerated accidentally at this stage.
8.42
145
146
8.43 Control of venous hemorrhage, often arising from

the junction of the spinal vein and the venous
plexus, using a small piece of muscle tissue
(arrow) cut from the adjacent epaxial
musculature. The muscle is macerated then
pressed firmly over the vein; it should be removed
once decompression is finished. Alternatives
include Gelfoam (Pharmacia, Kalamazoo, MI),
direct pressure on either side of the tear, or a
small Hemoclip (Pilling Weck Inc., Research
Triangle Park, NC) for a tear in a spinal vein.
8.44 When all disc material has been removed the

dura returns to lie adjacent to the floor and
pedicle of the vertebral canal; the spinal cord
may remain indented as here (same dog as in
8.38). Other clues to the adequacy of
decompression are reappearance of epidural fat
at each end of the bony defect along with
pulsation of the dura (caused by CSF pulsing in
synchrony with venous flow).
8.43
8.44
When no disc material can be found, it may be that:
The surgeon has approached at the wrong site (8.6).
The surgeon has approached on the wrong side.
There has been a small volume, high-energy disc herniation causing an impact injury.
The dog could have progressive myelomalacia.
The original diagnosis could be wrong.
If landmarks confirm that the site is correct (8.20), the defect must be extended in whichever direction the spinal
cord appears most compressed or swollen (8.56, 8.61A). If the spinal cord still appears compressed after removal
of all visible disc material then it is possible that more material exists on the far side of the spinal cord (8.3, 8.6). The
defect may then have to be extended either to one side (8.45), dorsally (8.46) or bilaterally (Schulz et al., 1998).
8.45 To extend the hemilaminectomy, the surgeon

should start drilling a short distance away from
the existing opening. The two should then be
merged to form one larger opening by breaking
the intervening bone bridge (arrow). This is safer
than trying to drill outwards from the edge of the
original hemilaminectomy defect as the drill bit
tends to slip off the edge and could injure the
spinal cord.
8.45
8.46 Drilling can continue dorsally if disc material lies

on the opposite side of the vertebral canal.
Excessive bone removal could cause delayed
injury from restrictive fibrosis; to avoid this
the cancellous and inner cortical bone can
sometimes be removed while sparing the outer
plate. If the disc material is still not visible, it is
safer to start a mini-hemilaminectomy on the
contralateral side.
8.46
8.47 A: This hemilaminectomy has been extended

ventrally to expose the spinal nerve and ganglion
(arrow). Bleeding from the spinal artery may
require bipolar cautery. The venous plexus is just
visible on the floor of the vertebral canal
(arrowhead); it is damaged easily, especially
where the spinal vein joins the plexus (8.43).
B: Lateral corpectomy (arrowheads) of one or
two vertebral bodies improves ventral access
further, especially to chronic disc herniations
(Moissonnier et al., 2002). Drilling should
proceed under the cortical bone plate that forms
the floor of the canal before collapsing the
cortical plate ventrally along with the disc.
The area of the corpectomy is occupied by air;
a large air pocket also lies adjacent to the fat
graft placed over the mini-hemilaminectomy site.
8.47
A
B
8.48
8.48 Access to the vertebral canal

varies with the location.
Hemilaminectomy at A:
lumbar vertebrae (L2) often
gives better access than at
B: thoracic vertebrae (T10/11)
due mainly to the rib heads.
A
147
148
8.49 A durotomy has been performed using a

25-gauge hypodermic needle to reveal the spinal
cord, small pial blood vessels and dorsal nerve
rootlets (arrow). A guarded scalpel or fine
scissors may also be used to incise the dura.
Collapse of the subarachnoid space has caused
the dura mater to become translucent (8.5,
8.50). The venous plexus is visible clearly on the
floor of the vertebral canal (arrowhead).
8.49
Although durotomy may provide additional spinal cord decompression, the relative risks and benefits are not
clear (see page 87):
Local necrosis of gray matter over one or two spinal cord segments is common after severe injury and it is
not unusual for this material to ooze out after durotomy (8.50).
This does not necessarily indicate a poor prognosis, such as with complete spinal cord transection, or that
there is progressive myelomalacia (Salisbury and Cook, 1988).
It is only necessary for 510% of the white matter to survive in order for an animal to recover; this would
be very hard to quantify after durotomy (Blight and Decrescito, 1986; Basso et al., 1996; Jeffery and
Blakemore, 1999; Olby et al., 2003) (see Prognosis, page 132).
Durotomy should therefore not be used to infer prognosis (Muir et al., 1995); its use purely as an additional
decompressive procedure may warrant further study (Anderson et al., 1991; Bagley et al., 1996).
Routine use of durotomy is not yet recommended in dogs with grade 5 deficits, particularly as more than
half of these dogs recover without one (Muir et al., 1995) (see Prognosis, page 132).
There is a small risk of spinal cord herniation through the durotomy defect, which can cause severe
distortion and compression (Osterholm, 1974; Henry et al., 1997).
8.50 Minihemilaminectomy and durotomy

(arrowheads) performed in a 25 kg dog
presenting with no deep pain sensation. There
is extensive bruising with focal malacia (white
arrow) of the spinal cord. This dog regained
nociception three days after surgery and made
a good recovery within six weeks of surgery.
The facet joint is shown by an asterisk.
8.50
8.51 Nerve roots can cause significant tethering of spinal

cord over the disc. This can be relieved by rhizotomy
(arrow) except at the lumbosacral plexus. Great care
must be taken when removing a chronic disc like this
(see Early postoperative complications, page 127). If
a plane of dissection can be developed between dura
and adhesions, the extrusion can then be cut away by
scalpel. Corpectomy facilitates this (8.47B). Otherwise
the spinal cord is just decompressed as shown without prolonged attempts at removal.
8.51
Rhizotomy improved the degree of decompression although it is not known if cutting the associated segmental
blood supply has any adverse effect on perfusion at that level.
8.52 Diagram to show the relationship of

the spinal nerve, the completed
hemilaminectomy defect and the
intervertebral disc space. These
relationships should be understood
clearly prior to fenestration (8.53).
The surgeon must also take care
not to injure the exposed spinal
cord inadvertently.
8.52
8.53 Fenestration is usually performed after

hemilaminectomy. If the surgeon is not familiar
with fenestration, then it is best to fenestrate the
hemilaminectomy site first as shown here (arrow),
in order to understand the relative positions of the
disc and spinal cord. In practice the spinal nerve is
seen only rarely. Here a scalpel blade has been
used to make a deep cut into the anulus (8.70,
8.78). It is not clear if a chronic disc should be
fenestrated (page 126).
8.53
8.54 Prior to closure, the hemilaminectomy site is

covered by a 35 mm thick layer of subcutaneous
fat demonstrated here lying over a cloth drape.
The graft shrinks to 50% of its original size
and so should be considerably larger than the
laminectomy defect (Trevor et al., 1991). It is
held in place by the epaxial musculature
after retractors are removed; alternatively
saline-soaked Gelfoam (Pharmacia, Kalamazoo,
MI) can be used. The fat graft must not be too
thick (8.8, 12.10).
8.54
149
150
8.55 3D CT reconstruction showing the postoperative

appearance of a standard hemilaminectomy. This
image, along with several other reconstructions
(8.48, 8.56, 8.608.62), shows postoperative
laminectomies from several dogs. This scan was
performed in a dog that lost deep pain 2 days
after a hemilaminectomy at L1/2; no residual disc
material was identified and the dog had
progressive myelomalacia at necropsy.
8.55
Decompression may be necessary over three or four vertebrae if:
The extrusion cannot be found (8.61).
Disc material is spread over a wide area.
The spinal cord is very swollen.
Extensive hemilaminectomy does not appear to affect recovery rates but movement at the surgical site is
increased (Corse et al., 2002); this may cause delayed morbidity in some dogs (Anderson et al., 1991; Grevel and
Schwartau, 1997; Applewhite et al., 1999; Scott and McKee, 1999).
8.56
8.56 This dog had CT for recurrent back pain 4 years

after hemilaminectomy from T13 to L3. No cause
was identified; the pain may have been related to
laminectomy scar (Applewhite et al., 1999). There
is a direct correlation in humans between
persistent low back pain and extensive peridural
scarring (Ross et al., 1996; Maroon et al., 1999).
Where possible, less extensive decompression
should be considered (8.578.62).
Pediculectomy and mini-hemilaminectomy (8.578.62)

Approaches include lateral (8.588.59) or standard dorsolateral (8.118.27) muscle separation. The lateral
approach only works well for lean, small dogs. With a standard dorsolateral approach the dog should be rotated
away from the surgeon in order to perform these surgeries as this makes it much easier to gain access under the
facet joint (Bitetto and Thacher, 1987).
8.57 Difference between A: hemilaminectomy,

B: mini-hemilaminectomy, and C: pediculectomy.
Hemilaminectomy sacrifices a facet joint; the other procedures
do not. Mini-hemilaminectomy enlarges the region around the
foramen. Pediculectomy spares the foramen but removes bone
between adjacent facet joints. As access is restricted with
mini-hemilaminectomy (8.59B) and pediculectomy (8.61), it is
often better to combine them (8.62).
8.57
A
151
152
8.58 Transverse section through the lumbar region

to show the lateral approach for
mini-hemilaminectomy or pediculectomy.
The transverse processes are palpated and
then the epaxial muscles are elevated dorsally
to reveal the lateral aspect of the vertebrae
(Braund et al., 1976) (8.59, 8.638.65). A curette
is shown retrieving extruded nucleus pulposus
following a mini-hemilaminectomy.
8.58
8.59 A: ArmyNavy retractors elevate the

epaxial muscles and expose the
vertebrae laterally (articular processes
are obscured by the muscle and
retractors). B: Mini-hemilaminectomy
provides good access to the
mid-ventral vertebral canal. The floor
of the vertebral canal is obscured here
by a large, chronic disc (arrow)
extruding dorsolaterally.
8.59
A
8.60 Pediculectomy via a standard dorsolateral approach

(8.118.27). A: Intraoperative photograph, and
B: 3D reconstruction of a postoperative CT scan from the
same dog. Pediculectomy gives good access over the
mid-vertebral body and preserves the articular processes.
Access to the intervertebral space is poor compared to
mini- and standard hemilaminectomy (8.61B).
8.60
A
8.61 A: When imaging is inconclusive, multiple pediculectomies

(arrows) can be used to search for disc, hemorrhagic fat, or
spinal cord swelling; the defect can be enlarged once the
correct location is found. Although not ideal, this is often
quicker than extensive laminectomy (8.56).
B: A disadvantage of pediculectomy is that material can be
missed easily. Mid-sagittal 3D reconstruction through the
vertebral canal of a dog where disc material (arrow) was
missed after pediculectomy (same dog as 1.21A, 8.60).
8.61
A
153
154
8.62 A mini-hemilaminectomy has been combined with two

pediculectomies to give much better access than either
procedure alone (Biggart, 1988; Lubbe et al., 1994;
McCartney, 1997). A: Intraoperative photograph, and B: a
3D reconstruction of a postoperative CT scan from the same
dog as shown in 8.40. The dorsolateral approach used here
makes it easier to extend the bony defect and is therefore
preferred over the lateral approach shown in 8.58 and 8.59.
8.62
A
Lateral fenestration (8.638.78)

The lateral approach for fenestration can be made from either side but is normally undertaken from the left side.
The depth of the epaxial muscle makes the approach progressively more challenging in dogs greater than 10 kg
bodyweight.
8.63 Positioning and skin incision

for lateral fenestration.
Thoracic limbs are tied
forwards with the pelvic limbs
extended backwards. A thin
sandbag is placed under the
dog at the level of the
thoracolumbar junction to
open up the disc spaces of
8.63
interest; this may be moved
as required to approach the
lumbar discs. The skin incision is made at the level of the lumbar transverse processes and extends from
approximately T8 to L5.
8.64 The thick, shiny, lumbodorsal fascia lies under

the subcutaneous fat and fascia. The
subcutaneous fat is reflected for 1 cm on either
side of the proposed incision. The lumbodorsal
fascia is then incised as shown. Reflection of fat
creates additional dead space but a wide
exposure makes it easier to repair the fascia
later.
8.64
8.65 The deep layer of fat, which can be substantial

even in lean dogs, has been incised to reveal the
iliocostalis lumborum and the thirteenth rib
(arrow). The longissimus dorsi muscle is covered
by a fascial sheath, which is just visible under the
layer of fat (arrowhead). Exposure dorsally is
more than is usually needed but has been made
to show the longissimus muscle.
8.65
8.66 Diagram to show the deep

anatomy and muscle
separation through the
iliocostalis lumborum muscle,
in this case over the L1/2
intervertebral space. The
longissimus muscle is in the
dorsocaudal part of the
surgical field (arrow). The
iliocostalis muscle fibers are
8.66
seen running obliquely
(arrowhead) to insert on the ribs.
8.67 Close up to show separation of the iliocostalis

lumborum muscle over the T13/L1 disc by
opening a pair of Metzenbaum scissors in the
same direction as the muscle fibers. The
thirteenth rib (with periosteum reflected for clarity)
is shown clearly in the lower left-hand side of the
picture.
8.67
8.68 Retraction of the iliocostalis muscle. This reveals

the body (*) and transverse process (arrowhead)
of the L1 vertebra along with the T13/L1 disc
(arrow). Note the fibers of the anulus fibrosus
(over which lies a fine layer of connective tissue).
This is best removed by using a periosteal
elevator covered with a surgical swab (8.23),
pushing the tissue in a craniodorsal direction.
8.68
155
156
8.69 Close up diagram of the site

approached in 8.68. Muscle
separation has revealed the
lateral aspect of the anulus
fibrosus, which lies just cranial
to the transverse process of
the lumbar vertebra. Note the
small vein that lies over the
craniodorsal anulus; this is
retracted as described in
8.68. The incision in the
anulus is shown by the
dotted line. Note also the
accessory process, which
marks the dorsal margin of the
intervertebral foramen (arrow).
The exact dorsal limit for the
fenestration varies slightly
depending on the disc space
and also from dog to dog. A
skeleton is a useful reference
(5.4). A blunt periosteal probe
can be used to palpate the
curving lateral surface of each
disc before fenestrating it. A
fresh approach is made for
each disc space.
8.69
8.70 Incising the disc with a #11 scalpel

blade reveals the window of anulus and
the jelly-like nucleus pulposus (arrow);
angle same as in 8.68.
8.70
To fenestrate a disc:
The window is made by four separate stab incisions, which are joined at the corners.
A sawing motion is used to cut instead of direct pressure.
The hole in the anulus must be made larger than any instrument used for removal of nucleus pulposus.
Disc removal is by small curette, a Rosen mobilizer or, failing that, by twisting the #11 blade.
It is important to remove as much nuclear material as possible, as any left will remain in the intervertebral
space and could cause a clinical problem later.
Power fenestration with a small drill bit may allow more complete evacuation (Holmberg et al., 1990). Care
must be taken to prevent the drill bit from slipping toward the foramen.
8.71 The approach to thoracic discs is different.

They may be approached by separating the
iliocostalis lumborum muscle as for lumbar discs.
Alternatively, the muscle can be cut close to its
insertions on the 12th and 13th ribs.
8.71
8.72 Retraction of the iliocostalis lumborum muscle

dorsally reveals the levator costarum muscle
(arrow). This will be separated from the rib and
retracted in a cranial direction.
8.72
8.73 The periosteum of the rib has been incised at the

caudal border of the levator costarum muscle
and elevated using a thin periosteal elevator.
Periosteum is elevated adjacent to the neck of
the rib, a small retractor is being used to retract
the levator costarum muscle cranially. The plane
of dissection is medial and dorsal to the pleural
reflection. Dissection of the deep fascia that
attaches to the cranial margin of the rib (arrow) is
then continued proximally, ventral to the head of
the rib.
8.73
157
158

features described in
8.73. The levator
costarum muscle is being
elevated from the rib with
a periosteal elevator. A
hand-held retractor is
positioned to keep the
iliocostalis muscles
retracted dorsally. Gelpi
retractors can be used
here but they must not
tear the pleura. A change
in respiratory character or
volume indicates possible
pneumothorax.
8.74

lateral aspect of the disc
exposed, with the levator
costarum muscle
retracted cranioventrally,
and the epaxial muscles
retracted dorsally. The
site of the fenestration is
shown.
8.75
8.76 Blunt dissection has been completed and the

disc is now visible clearly (*). Handheld retractors
are being used to elevate the epaxial muscle
away from the last two ribs.
8.76
8.77 A #11 scalpel blade cutting deeply into the

anulus. The fenestration is then completed as
described in 8.70. The surgeon should insure
that the window in the anulus is slightly larger
than the instrument to be used for curettage or
material might be forced inadvertently into the
vertebral canal.
8.77
8.78 Diagram of a transverse section through the

lumbar region of a dog to show the lateral
approach for fenestration. This approach allows
access to the lateral anulus fibrosus and to disc
material in the intervertebral space. However,
there is no access to disc material in the
vertebral canal unless a pediculectomy or
mini-hemilaminectomy is also performed
(8.578.62).
8.78
159
Clinical signs
161
Diagnosis 163
Examination 163
Radiography 163
163
Treatment 164
Surgery 165
Complications 167
Ventral fusion 167
Dorsal fixation 168
Postoperative care
Prognosis
164
167
169
169
Atlantoaxial subluxation in cats
169

References
170
170
Procedures 171
Ventral transarticular fixation 171
Multiple ventral implants and bone cement
Dorsal wire fixation 178
Dorsal cross-pin fixation 180
177
The atlantoaxial joint allows rotation of the head; C1

pivots around the dens of C2 but the joint permits little
flexion. There is no intervertebral disc between C1 and
C2 (4.8, 4.9) and the relationship between these vertebrae is maintained largely by ligaments (1.14, 1.35).
Clinical signs in congenital atlantoaxial subluxation
are usually seen in immature patients although signs can
develop at any age (Thomas et al., 1991; McCarthy
et al., 1995; Beaver et al., 2000) (9.1, 9.32). The disorder
is encountered most often in small breeds of dog,
Chapter
particularly Yorkshire terriers, Chihuahuas and Miniature

poodles (Denny et al., 1988; Thomas et al., 1991;
McCarthy et al., 1995; Beaver et al., 2000). Rare cases
occur in cats and in large-breed dogs (9.4). Atlantoaxial
subluxation can cause clinical signs in breeds such as the
Rottweiler (Rochat and Shores, 1999; Wheeler, 1992),
Doberman (Huibregtse et al., 1992; LeCouteur and
Child, 1995), Basset hound (Hurov, 1979), Standard
poodle (Knipe et al., 2002), Weimaraner, and German
shepherd dog (Read et al., 1987).
A number of pathological processes can lead to
atlantoaxial subluxation:
Absence of the dens (9.2B).
Fracture or separation of the dens (9.2C).
Failure of the ligaments due to either malformation
or rupture (9.2D).
Most dogs with congenital lesions have either
absence or hypoplasia of the dens (46%; 9.4), 30%
have a malformed dens (9.1B), and 24% have a normal
dens (Beaver et al., 2000). In dogs with a normal or
malformed dens, abnormalities of the transverse
ligament of the atlas can lead to subluxation (Watson
et al., 1989). This is serious as the dens tends to protrude dorsally into the spinal cord (9.1B, 9.2D,
9.19B).Occipito-atlantoaxial malformation can also
occur (Read et al., 1987).
Most patients have an underlying congenital abnormality but trauma can cause failure of normal elements
in this region (9.2C) (see Chapter 13). Minor trauma
may also precipitate a crisis in a dog that has a congenital abnormality but has shown no clinical signs previously (Thomas et al., 1991; Beaver et al., 2000).
CLINICAL SIGNS
Neck pain is seen in most dogs following traumatic
lesions and in 3060% of dogs with congenital lesions
(Thomas et al., 1991; Beaver et al., 2000). Neurological
signs reflect cervical spinal cord compression. In mild
cases only proprioceptive deficits are seen. Tetraparesis
indicates more significant spinal cord compression.
Asymmetry of signs, or preferential involvement of
162
either thoracic or pelvic limbs, may occur. Tetraplegia

is rarely encountered but if present the dog must be
checked for respiratory failure (Beaver et al., 2000) (see
Chapter 2). Signs referable to brain involvement have
also been reported and should be evaluated carefully.

Hydrocephalus has been reported in dogs with
atlantoaxial subluxation and could cause the animal to
show forebrain signs (Chambers et al., 1977; Denny
9.1 3D reconstruction of CT scans; the dens is indicated by arrows. A: A normal toy-breed dog. B: A dog with atlantoaxial subluxation
and a malformed dens, which occupies much of the vertebral canal. The spinous process of C2 is also angled away from the arch of
the axis (arrowhead). Two-dimensional images of these dogs are shown in 9.6; sagittal 3D reconstructions in 9.19; postoperative
radiographs for dog B in 9.32.
9.2 A: Diagram to show the normal relationship between C1 and C2 (1.14, 1.35, 9.19). B: Congenital absence or hypoplasia of
the dens is the most common abnormality. C: There is an ossification center between the dens and the body of C2, which
predisposes to fracture or separation. D: Rupture or malformation of the ligaments.
et al., 1988). Hydrocephalus may also be accompanied

by syringohydromyelia (9.12). Another potential explanation for signs of forebrain disease in dogs with
atlantoaxial subluxation is hepatic encephalopathy, which
is over-represented in toy-breed dogs and coexisted in
two of six dogs undergoing surgery for atlantoaxial subluxation (Schulz et al., 1997). Forebrain signs of disorientation and behavior change, along with vestibular
deficits, have also been associated with basilar artery
compression caused by the dens. The clinical signs
resolved completely after surgery (Jaggy et al., 1991).
Torticollis has been described with atlantoaxial lesions
and could be due to syringohydromyelia or a vestibular
sign secondary to a high cervical lesion (Johnson and Hulse,
1989; Gibson et al., 1995; Mayhew, 1999) (see page 29).
DIAGNOSIS
Examination
Atlantoaxial subluxation should be considered in any
young, small-breed dog with the clinical signs described.
Neurological examination indicates a lesion between
C1 and C5 and palpation of the neck often localizes
the origin of the pain to the C1C2 region. It is unwise
to flex the neck forcibly in a patient where atlantoaxial
Box 9.1 Primary differential diagnoses for

atlantoaxial subluxation (see also Box 7.2)
Cervical disc extrusion (older than 1 year)
Syringohydromyelia
Neoplasia
Inflammatory CNS disease
Intracranial disease
Discospondylitis
Polyarthritis
Polymyositis
Trauma
subluxation is suspected, as this may worsen the situation

considerably.
The differential diagnosis for a small-breed dog with
cranial cervical signs is given in Box 9.1. Inflammatory
CNS disease is the most likely consideration in immature
dogs. Cervical disc disease is more likely in older dogs but
is rare in dogs less than 2 years of age. Discospondylitis or
fractures could be present at any age. Atlantoaxial subluxation is encountered in cats rarely (see page 169).
Radiography
SURVEY RADIOGRAPHY
Survey radiographs provide the diagnosis in most cases
(9.3). General anesthesia is usually required, although
great care must be taken when intubating the patient. If
non-surgical management is to be used following trauma
then it may be worth trying to obtain diagnostic images
without anesthesia (see page 283). However, accurate
positioning is essential to evaluate the cranial cervical
region and this may be impossible in the conscious patient,
particularly if severe neck pain is present (see page 46).
It is a common error to diagnose congenital atlantoaxial
subluxation on radiographs of conscious dogs where the
positioning is inadequate and the region of interest is
far from the center of the film. Fluoroscopic observation
while flexing the neck gently in a conscious animal can
provide rapid and accurate diagnosis by revealing the
dynamic nature of the lesion while allowing the animal to
maintain some protective muscle tone.
The lateral projection is the most useful. Mild flexion
of the cranial cervical region may be required to demonstrate misalignment but this must not be excessive. A
ventrodorsal view will highlight the dens; it is safe to
position the dog in dorsal recumbency with the neck
extended for the ventrodorsal projection, which is
preferable to the open-mouth view. Oblique radiographs
can also provide an excellent image of the dens (9.3B,
9.4B) (Cook and Oliver, 1981).
9.3 Survey radiographs from a

1-year-old Toy poodle showing
a marked increase in the gap
between the dorsal arch of C1 and
the spinous process of C2 when
the neck is flexed (B). Postoperative
radiographs are shown in 9.8.
163
164
9.4 Eight year-old Bull mastiff with acute pelvic limb ataxia. Survey radiographs revealed severe hypoplasia of the dens; a lumbar
myelogram revealed no other abnormality. A: Flexed view and B: extended view of C1 and C2. Note the narrowing of the dorsal
subarachnoid space and the hypoplastic dens (arrowhead). Neurological deficits worsened after the myelogram but the dog made
a good recovery and had no neurological deficits 4 months later. Subclinical atlantoaxial malformation has been reported in humans
(McKeever, 1968).
9.5 Myelogram in a 4-year-old Toy

poodle demonstrates that there may
be little spinal cord compression if
the dens is absent or hypoplastic.
Postoperative radiographs of this dog
are shown in 9.29.
B
9.6 A: CT scan through the atlas of a
normal toy-breed dog. The dens is of
normal size and occupies the floor of the
vertebral canal. B: Dog with a malformed
dens. The same dogs are also shown
in 9.1 and 9.19.
MYELOGRAPHY
Myelography should not be necessary for diagnosis and
any post-myelographic seizures could be disastrous.
Cerebello-medullary cistern puncture either for myelography or CSF sampling should not be performed in
dogs that could have atlantoaxial subluxation (9.5);
lumbar puncture is preferable.
CT AND MRI
Although much simpler techniques are available to
assess atlantoaxial stability, advanced imaging can add
important preoperative information (Johnson and
Hulse, 1989) (9.6, 9.12). CT provides excellent bone
imaging, which assists surgical planning (Johnson and

Hulse, 1989) (9.33). It will also reveal abnormal dens
conformation, which is seen in over 70% of dogs (Denny
et al., 1988; Beaver et al., 2000) (9.6). An MRI (9.7)
might prove to be prognostic if there is extensive spinal
cord malacia and it will also reveal syringohydromyelia
(Sanders et al., 2000) (9.12).
TREATMENT
Non-surgical treatment entails cage rest, application of
a neck brace (13.18) and use of analgesics (Tables 15.1,
9.7 Sagittal MRIs of a dog with atlantoaxial subluxation. A: T2-weighted and B: short tau inversion recovery (STIR) images.
Note the high signal intensity (edema) and severe spinal cord compression at C1C2 (arrowheads).
Algorithm 9.1 Surgical decision-making

in atlantoaxial subluxation.
Survey radiographs
Malformed dens
Normal or
hypoplastic dens
Consider odontoidectomy
Fusion with multiple

implants and cement
Failure
Success
CT scan
Dorsal cross
pinning
External
splint
Replace implant
15.2). This approach can produce surprisingly good

results. Six toy-breed dogs were managed non-surgically,
four of which had been unable to walk. After 14 weeks
all six dogs could walk without neurological deficits
(Hawthorne et al., 1998). Four out of another six dogs
also did well with non-surgical management (Lorinson
et al., 1998). Despite these excellent results after nonsurgical management, the concern is that any improvement might be lost after brace removal and return to
normal activity (9.32). Nevertheless, it can provide a
very useful alternative for very young dogs, for dogs that
cannot walk, or for other high-risk patients. Surgical
management should be recommended for congenital
lesions once the neurological deficits improve or the dog
is large enough, at least until the long-term outcome of

non-surgical management is known. Non-surgical treatment does give excellent long-term outcomes for animals that fracture a normal atlantoaxial articulation;
results are often superior to surgery (see page 295).
Surgery
Surgical treatment is indicated in most patients with
congenital lesions. Even dogs with profound neurological
deficits are likely to benefit from stabilization (Thomas
et al., 1991; Beaver et al., 2000). The two main options
are either ventral fusion or dorsal stabilization. Ventral
fusion is the treatment of choice. An algorithm for surgical decision-making is shown in Algorithm 9.1.
165
166
9.8 Reduction after transarticular screw

fixation is good. Note the hypoplastic
dens. Same dog as in 9.3. Angles are
mediolateral: left 41 and right 38;
ventrodorsal 33 (9.25).
Table 9.1 Results of ventral fixation for atlantoaxial subluxation

Transarticular fixation
Multiple implant fixation
(1)
(2)
(6)
Total
transarticular
(5)
(4)
(3)
Total
multiple
Success rate after

1st surgery (%)
31/40 (78)
9/10 (90)
8/18 (44)
48/68 (71)
5/6 (83)
5/6 (83)
12/13 (92)
22/25 (88)
Median follow-up
months
12 (mean)
N/A
21
N/A
36
Residual ataxia (%)
6/31 (19)
N/A
N/A
6/31 (19)
0/5
N/A
4/13 (31)
4/18 (22)
Residual neck pain (%)
3/31 (10)
N/A
N/A
3/31 (10)
0/5
N/A
0/13
0/18
Second surgery (%)
4/35 (11)
0/9
3/18 (17)*
7/62 (11)
0/5
1/5** (20) 0/13
1/23 (4)
Mortality (%)
5/40 (13)
1/10 (10)
7/18 (39)
13/68 (19)
1/6 (17)
0/6
2/25 (8)
1/13 (8)
(1) Beaver et al., 2000; (2) Denny et al., 1988; (3) Knipe et al., 2002; (4) Sanders et al., 2000; (5) Schulz et al., 1997; (6) Thomas et al., 1991.
* Two of these recovered.
** Signs recurred but did not actually undergo second surgery.
N/A, not available.
VENTRAL FUSION
Fusion of the atlantoaxial joints can be performed using
either transarticular fixation (9.8) or using multiple
implants and bone cement (9.329.34). Good results
can be obtained using transarticular fixation; threaded
pins or screws give superior results to smooth pins.
However, there is little margin for error with only
two implants and overall failure rates approach 30%
with transarticular fixation (Table 9.1). Furthermore,
radiographic evidence suggests that fusion is often
delayed or incomplete (Thomson and Read, 1996;
Beaver et al., 2000). Multiple implant techniques
should reduce the failure rate; they add little technical
difficulty and almost certainly provide more rigid
fixation (Schulz et al., 1997; Sanders et al., 2000;
Knipe et al., 2002) (9.329.34). Use of multiple
implants is therefore the technique of choice for ventral

fixation. Following ventral fixation, odontoidectomy
may rarely be indicated via a ventral slot in C1 (9.6,
9.27, 9.32).
DORSAL STABILIZATION
Dorsal wiring has a high failure rate (Table 9.2), such
that many animals need repeat surgery (9.11). Reinforcement with bone cement may be useful but any thermal
necrosis could weaken the bone further (Renegar and
Stoll, 1979; Cook and Oliver, 1981; Martinez et al.,
1997). It does provide an option should ventral fusion
fail (Beaver et al., 2000), but a second attempt at fusion
using multiple implants or by dorsal cross-pinning are
more logical approaches (Thomas et al., 1991; Jeffery,
1996; Schulz et al., 1997; Sanders et al., 2000).
Table 9.2 Results of dorsal suture fixation for atlantoaxial subluxation

Dorsal wire or suture fixation
(1)
(2)
(3)
(4)
Total dorsal
Success rate after 1st surgery (%)
9/12 (75)
6/6 (100)
7/13 (54)
2/7 (29)
24/38 (63)
Median follow-up months
23
N/A
N/A
Residual ataxia (%)
4/9 (44)
5/6 (83)
N/A
N/A
9/15 (60)
Residual neck pain (%)
1/9 (11)
0/6
N/A
N/A
1/15 (6)
Second surgery (%)
2/11 (18)
2/6 (33)
N/A
3/7 (43)*
7/24 (29)
Mortality (%)
1/12 (8)
0/6
5/13 (38)
0/7
6/38 (16)
(1) Beaver et al., 2000; (2) Chambers et al., 1977; (3) Denny et al., 1988; (4) Thomas et al., 1991.
* One dog recovered.
COMPLICATIONS
The most common non-surgical and surgical complications are listed in Table 9.3.
The splinted animal must initially be assessed daily for
problems (Table 9.3), including the tendency for moisture to seep into the splint around the mouth. Dyspnea
can occur if the splint is too restrictive in the pharyngeal
or thoracic regions but it must also be tight enough to stay
in place (Schulz et al., 1997). Aspiration is a particular
problem if the neck is extended excessively as it is almost
impossible for the dog to swallow in this position.
Nasogastric or gastrostomy feeding may reduce the risk
of aspiration and also helps to keep the splint clean
(Shelton et al., 1991).
Ventral fusion
The ventral approach to the neck is straightforward but
care must be taken to avoid damage to vital structures,
particularly the recurrent laryngeal nerve and the vascular supply to the thyroid gland. Particular care must be
taken to prevent excessive traction or compression on
the trachea and esophagus (Thomas et al., 1991).
The main complications are death, or implant failure
that necessitates a second surgery (Table 9.3). The
primary causes of death are cardiac or respiratory arrest
and pulmonary edema (Thomas et al., 1991; Schulz
et al., 1997; Beaver et al., 2000). One potential cause of
pulmonary edema is barotrauma; care must be taken
when ventilating toy-breed dogs (see page 85).
Implant failure is more likely after ventral transarticular fixation than after use of multiple implants. The main
Table 9.3 Complications of non-surgical and surgical

treatment
Non-surgical
complications
Surgical
complications
Dyspnea
Aspiration
Abrasions
Decubitus
Otitis externa
Dermatitis
Recurrence
Cardiac arrest
Respiratory arrest
Pulmonary edema
Implant failure (9.11, 9.31)
Soft tissue injury
Tracheal collapse
causes of implant failure are K-wire migration or loss of

reduction; these tend to occur in the first 3 weeks
(Johnson and Hulse, 1989; Thomas et al., 1991; Schulz
et al., 1997; Beaver et al., 2000). Implant failure in some
dogs may not cause clinical signs if preceded by sufficient
fusion (Wheeler, 1992). Threaded pins are preferred
over Kirschner wires as they are much less likely to
migrate and have greater pullout strength (Johnson and
Hulse, 1989; Sandman et al., 2001). In one dog a K wire
migrated into the oral cavity although it caused no clinical signs other than a cough (Schulz et al., 1997) (9.9).
The disadvantage of using only two implants in
transarticular fashion is the relatively high failure rate
(Table 9.1). Failure rates should be lower when screws
or threaded pins are used instead of K wires but failure
can still occur even when implant position is excellent
(9.10). In ventral fusion the implants are on the compression side of the vertebral column and so are subject
to greater stresses than implants on the dorsal (tension)
side (Jeffery, 1996; Rochat and Shores, 1999). Even
167
168
9.9 A: Postoperative radiographs from a 1-year-old Toy poodle show K-wires with bone cement over the exposed ends. Angles are
mediolateral: left 53 and right 55; ventrodorsal angle is 2530. B: Failure occurred as one implant is anchored only poorly in C1.
An external splint was applied but both pins had failed by 5 weeks. The splint was maintained for 12 weeks in all; the dog was
pain-free with no deficits 4 years later apart from a lump on its neck from a migrated pin. A metallic skin staple is visible.
9.10 A: Transarticular fixation using lag

screws and cancellous bone graft;
postoperative alignment is excellent.
Angles are mediolateral: left 22 and right
45. B: The dog was also placed in a
splint but despite this one screw was
loose at the 6-week recheck and there
was no bridging callus. After a further
6 weeks in the splint good fusion was
evident and the dog made an excellent
recovery.
when failure does not occur, fusion is often delayed

(Sorjonen and Shires, 1981; Thomson and Read, 1996;
Beaver et al., 2000). Nine of 12 immature dogs showed
no bony fusion 6 weeks after transarticular fixation
despite use of a bone graft. Nevertheless, the fixations
were stable in 10 of the 12 dogs, usually due to fibrocartilaginous tissue (Sorjonen and Shires, 1981).
If implant failure does occur, either a splint can be used
or a second surgery may be performed. Good results can
often be obtained using non-surgical management when
the failure is only partial (9.9, 9.10). A second surgery is
recommended for complete failure of fixation (Thomas
et al., 1991; Beaver et al., 2000). After failure of either
a dorsal or ventral surgery, rescue is probably best
attempted using either multiple implants and bone
cement or dorsal cross-pinning (Thomas et al., 1991;
Jeffery, 1996; Schulz et al., 1997; Sanders et al., 2000;
(9.329.34; 9.43). Mini plates are another alternative
but there is little margin for error when placing the
screws in C2 (Stead et al., 1993).
Other postoperative complications include gagging
or coughing; laryngeal paralysis (Beaver et al., 2000;
Sanders et al., 2000); tracheal collapse (Thomas et al.,

1991); tracheal necrosis (Thomas et al., 1991); pneumonia (Rochat and Shores, 1999; Beaver et al., 2000);
dyspnea and sudden death (Denny et al., 1988; Thomas
et al., 1991). Tracheostomy may be necessary to relieve
iatrogenic laryngeal paresis or paralysis (Sanders et al.,
2000). Respiratory dysfunction can also result from
spinal cord injury impairing respiratory drive or motor
function (Thomas et al., 1991; Schulz et al., 1997)
(7.11). Non-respiratory complications include torticollis
(Beaver et al., 2000); delayed fracture of C1 (Johnson
and Hulse, 1989; Beaver et al., 2000); Horners syndrome (Jaggy et al., 1991); and esophageal stricture
secondary to gastric reflux (Schulz et al., 1997).
Dorsal fixation
The main complications are death or failure of fixation.
Death is due to cardiac or respiratory arrest, which
may occur during placement or tightening of the wires
(Denny et al., 1988). The wire suture can also break or
cheese-wire though C2 spinous process (Beaver et al.,
2000; Thomas et al., 1991)(9.11). Almost 30% of dogs
9.11 Failure of dorsal fixation. 2/0 Teflon-coated, polypropylene

suture was used instead of wire in this immature dog to avoid
cheese wiring through the spinous process of C2 but it broke
after 2 days. The radiograph shows loss of reduction; the hole
drilled in C2 spinous process is visible (arrowheads). The first
suture was replaced with 0.6-cm Dacron tape; it broke at the
cranial edge of C1 after a further 7 days. Number 2 Mersiline
was used for the third surgery but almost tore through the
spinous process of C2. The spinous process was then reinforced
with bone cement and the dog placed in a splint for 4 weeks,
which caused severe otitis externa. The dog was walking
normally 4 months later.
require a second surgery after standard dorsal fixation (Table 9.2). Other failures reported after dorsal
wiring include wire suture breakage and fracture of the
spinous process of C2 (Chambers et al., 1977; Thomas
et al., 1991; Beaver et al., 2000).
POSTOPERATIVE CARE
(see Chapter 15)
Pain is common postoperatively and adequate analgesia

should be provided, avoiding respiratory depressant
drugs in tetraplegic or severely tetraparetic animals
(Tables 15.1, 15.2). Strict rest is enforced for 612
weeks following surgery. Osseous fusion, or at least
stability of implants for a minimum of 6 weeks, is
the goal. External support is useful after surgery,
especially with transarticular or dorsal wire fixation.
The splint should be removed periodically for a thorough inspection.
PROGNOSIS
The prognosis for dogs with congenital lesions is good
if the animal survives the perioperative period (Beaver
et al., 2000). The best predictor of a successful outcome is when the onset of signs is prior to 2 years of
age (Beaver et al., 2000). The final outcome also tends
to be better if signs have been present for less than
10 months, if the dog can still walk, and if the reduction

after surgery is good (Beaver et al., 2000). Despite the
guarded prognosis for dogs with severe neurological
deficits, 9 of 13 dogs (69%) that were unable to walk
before surgery had a good outcome (Thomas et al.,
1991; Knipe et al., 2002). Neck pain persists in about
10% of dogs and residual ataxia in about 20% after
transarticular pin fixation (Beaver et al., 2000). The
rate of residual ataxia is two to three times as high after
dorsal wire fixation, presumably as the fibrosis at the
surgical site provides less stability than osseous or
fibrocartilaginous fusion ventrally (Beaver et al., 2000)
(Tables 9.1, 9.2).
Overall failure rates are as follows:
After dorsal wire fixation techniques the failure
rate is 37%; the mortality rate is 16% and 29%
of dogs need a second surgery (Table 9.2).
A literature summary reported a similar failure
rate in 20/52 dogs (38%) (McCarthy et al., 1995).
Outcome after dorsal wiring is no different to that
after ventral transarticular fixation using smooth
pins (McCarthy et al., 1995; Beaver et al., 2000).
After transarticular fixation the overall failure rate
is 29%; the mortality rate is 19% and 11% of dogs
need a second surgery. However, when results for
lag screw transarticular fixation are considered
specifically the results are much better than for
smooth pins (Denny et al., 1988; McCarthy et al.,
1995). Threaded pins should also give much better
results than smooth pins due to their higher pullout
strengths (Sandman et al., 2001).
After multiple ventral implants and bone cement
the failure rate is 12%; the mortality rate is 8% and
4% of dogs need a second surgery (Table 9.1).
Reported outcomes are therefore best for techniques
that employ multiple ventral implants and bone
cement. Dogs requiring a second surgery have no worse
outcomes than dogs that undergo only one (Beaver
et al., 2000).
ATLANTOAXIAL SUBLUXATION
IN CATS
Clinical signs of congenital luxation are similar to those
seen in dogs. All three cats described had good outcomes after transarticular pin fixation (Jaggy et al.,
1991; Shelton et al., 1991; Thomson and Read, 1996).
Differential diagnoses for cats include disc disease (see
Chapter 8), mucopolysaccharidosis, hypervitaminosis
A (see Chapter 14), lymphoma or other spinal cord
tumors (see Chapter 12), trauma (see Chapter 13),
meningoencephalitis, discospondylitis and possibly
fibrocartilaginous embolism (FCE) (see Chapter 14).
169
170

1. What is the long-term outcome after non-surgical
management for atlantoaxial subluxation?
2. Is a splint necessary to provide additional stability after
surgery?
3. Is there any advantage to performing routine
odontoidectomy for a malformed dens?
4. What is the prevalence and significance of
syringohydromyelia in atlantoaxial subluxation (
9.12 Syringohydromyelia has been reported as an incidental

finding on MRI of dogs with atlantoaxial subluxation (Sanders
et al., 2000). It could explain residual signs (9.32) or failure to
improve after stabilization. This dog shows high signal within
the spinal cord over the first and second cervical vertebrae
(arrow). There is also severe compression due to instability at
the atlantoaxial joint (arrowhead).
REFERENCES
Beaver, D.P., Ellison, G.W., Lewis, D.D., Goring, R.L., Kubilis, P.S.,
Barchard, C. (2000) Risk factors affecting the outcome of surgery for
atlantoaxial subluxation in dogs: 46 cases (19781998). Journal of the
Chambers, J.N., Betts, C.W., Oliver, J.E. (1977) The use of nonmetallic
suture material for stabilization of atlantoaxial subluxation. Journal of the
Cook, J.R., Oliver, J.E. (1981) Atlantoaxial luxation in the dog. Compendium
on Continuing Education for the Practicing Veterinarian 3, 242250.
Denny, H.R., Gibbs, C., Waterman, A. (1988) Atlantoaxial subluxation in
the dog: a review of thirty cases and an evaluation of treatment by lag
screw fixation. Journal of Small Animal Practice 26, 3747.
Gibson, K.L., Ihle, S.L., Hogan, P.M. (1995) Severe spinal cord compression
caused by a dorsally angulated dens. Progress in Veterinary Neurology 6,
5557.
Hawthorne, J.C., Cornell, K.K., Blevins, W.E., Waters, D.J. (1998) Nonsurgical treatment of atlantoaxial instability: A retrospective study.
Huibregtse, B.A., Smith, C.W., Fagin, B.D. (1992) Atlantoaxial luxation in
a Doberman Pinscher. Canine Practice 17, 710.
Hurov, L. (1979) Congenital atlantoaxial malformation and acute subluxation in a mature Basset Houndsurgical treatment by wire stabilization.
Jaggy, A., Hutto, V.L., Roberts, R.E., Oliver, J.E. (1991) Occipitoatlantoaxial malformation with atlantoaxial subluxation in a cat. Journal of Small
Jeffery, N.D. (1996) Dorsal cross pinning of the atlantoaxial joint: new
surgical technique for atlantoaxial subluxation. Journal of Small Animal
Practice 37, 2629.
Johnson, S.G., Hulse, D.A. (1989) Odontoid dysplasia with atlantoaxial
instability in a dog. Journal of the American Animal Hospital Association
25, 400408.
Knipe, M.F., Sturges, B.K., Vernau, K.M., Berry, W.L., Dickinson, P.J.,
Anor, S., LeCouteur, R.A. (2002) Atlantoaxial instability in 17 dogs.
LeCouteur, R.A., Child, G. (1995) Diseases of the spinal cord. In:
S.J. Ettinger (ed.), Textbook of Veterinary Internal Medicine, 629695.
Lorinson, D., Bright, R.M., Thomas, W.B., Selcer, R.R., Wilkens, B.A. (1998)
Atlantoaxial subluxation in dogs: the results of conservative and surgical
therapy. Canine Practice 23, 1618.
Martinez, S.A., Arnoczky, S.P., Flo, G.L., Brinker, W.O. (1997) Dissipation
of heat during polymerization of acrylics used for external skeletal fixator connecting bars. Veterinary Surgery 26, 290294.
Mayhew, I.G. (1999) The healthy spinal cord. American Association of
Equine Practitioners; 5666.
McCarthy, R.J., Lewis, D.D., Hosgood, G. (1995) Atlantoaxial subluxation in
dogs. Compendium on Continuing Education for the Practicing Veterinarian
17, 215226.
McKeever, F.M. (1968) Atlantoaxial instability. Surgical Clinics of North
America 48, 13751390.
Read, R., Brett, S., Cahill, J. (1987) Surgical treatment of occipitoatlantoaxial malformation in the dog. Australian Veterinary Practitioner
17, 184189.
Renegar, W.R., Stoll, S.G. (1979) The use of methylmethacrylate bone
cement in the repair of atlantoaxial subluxation stabilization failures.
Case report and discussion. Journal of the American Animal Hospital
Rochat, M.C., Shores, A. (1999) Fixation of an atlantoaxial subluxation by
use of cannulated screws. Veterinary and Comparative Orthopaedics and
Sanders, S.G., Bagley, R.S., Silver, G.M. (2000) Complications associated
with ventral screws, pins and polymethylmethacrylate for the treatment
of atlantoaxial instability in 8 dogs. Journal of Veterinary Internal
Medicine 14, 339.
Sandman, K.M., Smith, C.W., Harari, J., Manfra Maretta, S., Pijanowski, G.J.
(2001) Comparison of pull-out resistance of Kirschner wires and Imex
miniature interface fixation pins in polyurethane foam. Veterinary and
Comparative Orthopaedics and Traumatology 15, 1822.
Schulz, K.S., Waldron, D.R., Fahie, M. (1997) Application of ventral pins
and polymethylmethacrylate for the management of atlantoaxial instability: results in nine dogs. Veterinary Surgery 26, 317325.
Shelton, S.B., Bellah, J., Chrisman, C., McMullen, D. (1991) Hypoplasia
of the odontoid process and secondary atlantoaxial luxation in a Siamese
cat. Progress in Veterinary Neurology 2, 209211.
Sorjonen, D.C., Shires, P.K. (1981) Atlantoaxial instability: a ventral surgical
technique for decompression, fixation, and fusion. Veterinary Surgery
10, 2229.
Stead, A.C., Anderson, A.A., Coughlan, A. (1993) Bone plating to stabilise
atlantoaxial subluxation in four dogs. Journal of Small Animal Practice
34, 462465.
Swaim, S.F., Greene, C.E. (1975) Odontoidectomy in a dog. Journal of the
Thomas, W.B., Sorjonen, D.C., Simpson, S.T. (1991) Surgical management of atlantoaxial subluxation in 23 dogs. Veterinary Surgery 20,
409412.
Thomson, M.J., Read, R.A. (1996) Surgical stabilisation of the atlantoaxial
joint in a cat. Veterinary and Comparative Orthopaedics and Traumatology
9, 3639.
Watson, A.G., de Lahunta, A. (1989) Atlantoaxial subluxation and absence
of transverse ligament of the atlas in a dog. Journal of the American
Wheeler, S.J. (1992) Atlantoaxial subluxation with absence of the dens
in a Rottweiler. Journal of Small Animal Practice 33, 9093.
PROCEDURES
The patient is placed with the neck in extension; this helps to reduce the subluxation. Although positioning the
neck like this helps, reduction of C1 and C2 is usually incomplete. The mid-body of C2 must be grasped with
bone holding forceps and pulled in a caudoventral direction (9.19). Breaking down the joint capsule may also be
necessary to complete reduction. Occasionally malformation or secondary changes make complete reduction
impossible (9.32). The approach can either be the standard ventral approach (see page 106) or the paramedian
approach to the neck (see page 232). Preserving the thyroid artery (9.18) is much easier when using the paramedian approach as the carotid artery and its thyroid branch are reflected along with the trachea and larynx. The
dogs head is to the left in all illustrations.
Ventral transarticular fixation (9.139.31)
9.13 Positioning for surgery. The area is prepared and

draped, including the proximal humerus to allow
bone graft collection (11.24).
9.13
9.14 Site of incision. Note that the incision extends

cranial to the larynx. At the cranial edge of the
skin incision is the hyoid venous arch, which
should be divided.
9.14
9.15 Incision through the superficial fascia reveals

the sternohyoid muscles (7.21, 7.22). The larynx
(a) is at the cranial end of the incision, and the
trachea is visible (b).
a
b
9.15
171
172
9.16 The sternothyroid muscle inserts on the thyroid

cartilage. The muscle is sectioned along the
dotted line. The vascular bundle shown in 9.18
can be seen under the muscle insertion.
9.16
9.17 The sternothyroid muscle is mobilized and

divided close to the larynx. Leave an adequate
portion near the thyroid cartilage for repair. The
thyroid gland is visible (*).
9.17
9.18 A vascular bundle supplies the thyroid gland;

only one artery supplies each gland and so it
must be preserved. Adequate padding with
moist towels or sponges must be provided for
the trachea, esophagus and thyroid gland.
Insufficient protection has been provided here to
the trachea, which could lead to necrosis
(Thomas et al., 1991).
9.18
9.19
9.19 These sagittal 3D

reconstructions of CT
scans show how
subluxation can make
the initial approach
challenging. A: Normal
dog to show congruity of
B
A
the ventral surfaces of C1
and C2. B: In most dogs with atlantoaxial subluxation, the dens, atlantoaxial joints and the cranial portion
of C2 are obscured under the body of C1. Reduction is in the direction of the arrow. Images from the
same dogs are shown in 9.1 and 9.6.
9.20 Deep fascia dissected with C1 and C2 reduced

to show the tendons of the longus colli muscles
inserting on the ventral process of C1 (arrow).
9.20
9.21 Diagram of deep anatomy as shown in 9.20.

Note the relationship of the soft tissues to the
underlying skeletal structures.
9.21
9.22 The tendons of the longus colli muscles are

elevated from the ventral process of C1. The
muscles are elevated caudolaterally from the
body of C2 (exposed here). Dissection of the
fascia reveals the joint capsule of the C1/C2
joints. Here the capsule has been incised and
removed partially on the dogs right side (arrow).
9.22
9.23 The joint spaces can be seen clearly. The

articular cartilage is removed with a curette, #11
blade or a small bur. If a bur is used, care must
be taken not to weaken the cortical bone on
either side of the joint.
9.23
173
174
9.24
9.24 The joint space can be opened using a dental

tartar scraper or small Hohmann retractor. This is
a useful maneuver in order to support C2 while
drilling. Here the cranial articular cartilage has
been removed revealing the subchondral
bone (arrow).
9.25 Position of screws. Ideally, the screws are angled

away from the midline at approximately 30,
towards the medial angle of the alar notch, and
downward (i.e. dorsally) at approximately 20 from
the horizontal, which in practice means using as
flat a trajectory as possible (Sorjonen and Shires,
1981).
9.25
9.26 Placement of screws. In most miniature dogs,

1.5-mm cortical bone screws are used.
A 1.1-mm hole is drilled in C2 and across the
articulation. The body of C2 has a tendency to
move down, away from the surgeon; use of the
stabilization technique shown in 9.19 and 9.24
can help to prevent this. Once the hole has been
drilled in one side of C2, a tartar scraper may be
inserted in the hole to stabilize the vertebra while
the other hole in C2 is drilled.
9.26
The following order for drilling, tapping and screw placement is preferred:
1. Drill both 1.1-mm holes in C2.
2. Drill 1.1-mm hole through one of these holes into C1 on one side only. Insure that far cortex is penetrated.
3. Drill 1.5-mm glide hole on the same side in C2.
4. Measure depth of the hole through C2 into C1.
5. Tap this hole with 1.5-mm tap.
6. Place screw through C2 into C1. Do not tighten fully.
7. Repeat steps 2 to 6 on the other side.
Performing the operations in the order shown above avoids the problem of drilling holes and not being able to
locate one after the other screw has been tightened. A similar sequence can be used with threaded pins. These
still require a pilot hole and should be inserted using a low power setting or ideally with a small hand-held chuck
(5.22). Pins must also be encased completely within bone cement (Johnson and Hulse, 1989). Cannulated
screws provide another alternative as they permit repositioning of the guide wires if necessary before placing
screws. These screws are self-tapping and come in 4.0- and 3.0-mm diameters (Rochat and Shores, 1999).
The average mediolateral pin angle in dogs undergoing a successful stabilization was between 22 and 27,
with a theoretical ideal of 29 and a practical range of 10 and 45. If implants are directed too laterally across the
atlantoaxial joints, they may damage the vertebral artery; if directed too medially they may damage spinal cord
(Rochat and Shores, 1999). The average ventrodorsal angle in the same dogs was between 28.5 and 34.5 with
a theoretical ideal of 22 and a practical range of 15 and 45 (9.8, 9.29, 9.31). The theoretical ideal ventrodorsal
angle is hard to approach and in practice the angle should be as near horizontal as possible. Starting the implants
as far caudally as possible in C2 may facilitate this (Sorjonen and Shires, 1981; Jaggy et al., 1991). Malformation
of C1 or C2 can make assessment of the correct angles very difficult (Thomas et al., 1991). Use of a goniometer
during surgery may help improve accuracy (Thomson and Read, 1996). Vertebral positioning must also be perfectly symmetrical.
Odontoidectomy is usually described prior to reduction but may be easier and safer once the joint is reduced
and stabilized (Johnson and Hulse, 1989; Jaggy et al., 1991; Gibson et al., 1995) (9.27). Access to the dens is
through a short slot in C1. Odontoidectomy is recommended if the dens has a marked dorsal angulation (Swaim
and Greene, 1975; Johnson and Hulse, 1989; Jaggy et al., 1991; Gibson et al., 1995) but is not necessary in
most dogs (Schulz et al., 1997).
9.27 Here one screw is in position, and the other is

being inserted prior to final tightening.
9.27
175
176
9.28 Cancellous bone is harvested from the proximal

humerus and placed in and around the joint
space before tightening the screws (11.24).
The longus colli is then apposed with absorbable
sutures and the sternothyroid muscle is repaired.
9.28
9.29 A: The ventrodorsal screw

angle is nearly perfect (25).
B: The mediolateral angles
are a little high but still
acceptable (left 43 and
right 40). This dog, also
shown in 9.5, was normal at
3-month follow-up.
9.29
A
9.30 If the ventrodorsal angle is

suboptimal then failure
occurs due to insufficient
bone purchase (Rochat
and Shores, 1999). Despite
implant failure, this 1-year9.30
old Toy poodle was
managed in a splint and was A
normal 3 years later.
9.31 The screw on the left is not angled adequately away from the midline
and was probably not crossing the joint. It snapped soon after
surgery and the dog had a partial recurrence of signs. The dog was
managed by addition of an external support and made a good
recovery. Angles are mediolateral: left 14 and right 41.
9.31
Multiple ventral implants and bone cement (9.329.34)

Screws can be directed laterally into the thick bone just rostral to each caudal articular surface of C1 (1.15,
9.33A) and also into the bone caudal to the cranial articular surface of C2 (1.15B). The screw heads are then
united with bone cement (9.32). Variations on this technique include placement of two implants in C1, two in C2,
and two more in C3 (Sanders et al., 2000); and the pattern shown in 9.34 that incorporates transarticular
implants (Schulz et al., 1997). Implants of the largest possible size should be chosen and they should penetrate
two cortices for greatest pullout strength (Sandman et al., 2001). All implants should be covered completely by
cement. Threaded pins are superior to smooth pins and should ideally be put in by hand using a mini-chuck
(5.22). Whatever pattern of multiple implant is used, the length of implant and the volume of cement must be
considered in relation to the amount of room existing at this location (9.34). Care must be taken to protect soft
tissues from the heat of polymerization and to avoid rough edges that might abrade tissues.
9.32 This Miniature poodle was

tetraparetic at 6 weeks of age
but improved dramatically in a
splint. The dog presented
with neck pain, tetraparesis
and fecal incontinence at 3.5
years. Fixation was with
9.32
multiple implants and bone
B
cement. A and B: Despite
A
excellent reduction dorsally
there is a marked step on the floor of the vertebral canal; the malformed dens probably prevented complete
reduction. Odontoidectomy may have helped but the risk of iatrogenic spinal cord trauma was considered too
high (9.6). The dog was pain-free but still incontinent at 1-year follow-up.
9.33 CT scans to show A: site of

implant placement in the
thick bone just rostral to the
caudal articular surfaces of
C1 (arrows) and B: the
method for cross-pinning
into the caudal vertebral
body of C2 (arrows).
9.33
Implants should penetrate
two cortices for maximum
A
B
holding power (Sandman et al.,
2001). When available, intraoperative radiography is recommended to confirm implant placement (Rochat
and Shores, 1999). See also 1.14, 1.15.
177
178
9.34 Postoperative A: lateral and B:

ventrodorsal radiographs to show
fixation in a 1-year-old Yorkshire
terrier using six Kirschner wires,
cancellous bone graft and bone
cement. Two wires were placed into
each of the pedicles of C1 (9.33A).
Two wires were placed across the
9.34
atlantoaxial joints as described in
B
9.249.27. Two wires were placed into A
the caudal body of C2 at approximately
30 to the transverse plane (9.33B); ideally they should not cross C2/3 disc space (from Schulz et al., 1997).
Dorsal wire fixation (9.359.42)
9.35 Incision site relative to

the skeleton. The finger
is on the occipital
protuberance; the
incision is made just off
the midline. Note that
the neck must be flexed,
which is not ideal
(page 171).
9.35
9.36 Exposure of the spinous process of C2.

Manipulation and movement of the vertebrae
must be kept to a minimum; it is therefore
preferable to use sharp dissection.
9.36
9.37 Diagram to show the relationship

between skeletal, vascular and
nervous structures. (a) Dorsal
notch of the foramen magnum;
(b) dorsal arch of C1; (c) spinous
process of C2; (d) C2 nerve roots
and vessels. Note also the
vertebral artery and its branches
(1.36).
9.37
9.38 Two holes are drilled in the spinous process of C2.
9.38
9.39 The dorsal atlantoaxial ligament between C1

arch and C2 spinous process is disrupted in
atlantoaxial subluxation. The periosteum and soft
tissues are removed to allow access to the
vertebral canal between C1 and C2. A double
loop of wire is passed under the arch of C1 in a
cranial direction. Pressure on the spinal cord
must be avoided. The internal periosteum of the
vertebral canal may be continuous with the dura
mater at this stage, and gentle dissection is
required to allow passage of the wire.
9.39
9.40 The loop of wire is retrieved from the atlantooccipital space. This space may need to be
enlarged in order to grasp the wire. This is best
achieved by removing bone from the occiput,
thus preserving the dorsal arch of C1 required
for the fixation.
9.40
179
180
9.41 Arrangement of wires after tightening and

cutting the ends. If the spinous process is too
small to drill holes into, the suture can be
passed under the C2 spinous process
(Chambers et al., 1977).
9.41
9.42 Postoperative radiograph showing reduction of

the subluxation. Note that the wire under the
lamina of C1 is causing some spinal cord
compression.
9.42
Dorsal cross-pin fixation (9.43)

This should be stronger than dorsal wiring; it does not carry the risk of spinal cord trauma and provides a useful
rescue technique for any fixation failure. A disadvantage is that it still relies on the spinous process of C2. Care
must be taken during exposure of the wings of the atlas to avoid the C1 nerve root and also the vertebral artery as
it exits the transverse foramen (1.36). With the vertebrae reduced, a K-wire is driven in a ventrolateral direction
across each side of the spinous process of C2 to engage and penetrate the caudal half of the wing of C1.
Predrilling a small initial hole and then using a bone gouge is necessary to stop the pin from slipping caudally off
the sloped surface of the wing. Each K-wire is then removed and replaced by a positive profile pin. The pin usually
has to curve slightly between C1 and C2 in order to make good bone contact; the curve is best induced using the
bone gouge. Bone cement is then applied to the exposed aspects of the pins but direct contact should be avoided
between the cement and the spinous process of C2 in order to reduce the risk of thermal necrosis weakening
the bone of toy-breed dogs. Bone graft is also applied between the arch of C1 and the spinous process of C2 to
promote a fibro-osseous union (Jeffery, 1996).
9.43 A,B: Dorsal cross-pin fixation.
9.43
A
Lumbosacral disease
Clinical signs
Chapter
10
183
Diagnosis 183
Examination 183
Differential diagnosis 184
Electrophysiology 184
Radiography 185
Treatment 188
Surgery 189
Foraminal decompression or facetectomy
Dorsal fusionfixation 190
190
Complications 191
Intraoperative 191
Early postoperative 191
Late postoperative 191
10.1 Dog in a typical posture of low back pain. It also had

lameness in one pelvic limb and marked pain on palpation over
the lumbosacral junction.
Postoperative care
cauda equina. The spinal cord ends within L6 vertebra

in most dogs or in L7 in cats and some small dogs
(Fletcher and Kitchell, 1966) (1.5, 1.8B, 1.9). The L7
nerve roots run in unique troughs called the lateral
recesses within L7 vertebra (1.28B).
Motion in the normal lumbar spine is greatest at the
lumbosacral joint (Burger and Lang, 1993). In some dogs,
abnormal motion probably leads to degenerative changes
such as spondylosis deformans, osteophyte proliferation and soft tissue overgrowth of the joint capsules.
These degenerative changes appear to reduce the overall range of motion and clinical problems may arise due
to subsequent compression of neural structures in the
vertebral canal and intervertebral foramina (Mattoon
and Koblik, 1993; Schmid and Lang, 1993). This
sequence of events resembles those thought to occur in
hip dysplasia (Chambers et al., 1988).
A number of abnormalities may combine to cause
compression of the cauda equina or L7 nerve roots
(10.210.4). These include:
Prognosis
192
192

References
193
193
Procedures 195
Foraminal decompression and facetectomy
Dorsal fixationfusion 204
202
The clinical signs seen with lumbosacral lesions differ

from those seen at other locations of the spine (10.1),
mainly because of the unique anatomical structure of the
region. The vertebral canal in this region contains only
182
10.2 A: Sagittal section to show the normal relationship between L7 and the sacrum. B: 3D reconstruction from a dog with a
normal lumbosacral spine (same dog as in 2.24). The dog was free of signs for lumbosacral disease and a transverse CT study
through this region was normal (10.10A).
10.3 A: Hansen type II disc

herniation compressing the cauda
equina (10.10B). B: 3D
reconstruction of a CT scan from a
7-year-old Mastiff with chronic, lower
back pain, reduced hock flexion and
low tail carriage. A large, disc
herniation was excised via dorsal
laminectomy. The dog then developed
urinary retention due to increased
sphincter tone; this resolved after
2 weeks of catheterization (OBrien,
1988; Coates, 1999).
A
10.4 A: Ventral subluxation of the sacrum relative to L7. This may occur in normal dogs, although a step of more than 4 mm
suggests an abnormal lumbosacral junction (Wright, 1980; Denny et al., 1982; Schmid and Lang, 1993). B: Six-year-old working
German shepherd dog presenting with mild paraparesis, weak hock flexion and lumbosacral pain. The dog was seropositive for
Ehrlichia canis and returned to work after antibiotics; the subluxation did not seem to cause any clinical signs (3.2, 4.35).
Lumbosacral disease
Stenosis (multilevel) of the vertebral canal (Meij,

1993; Jones et al., 1995a,b, 1996b).
Hansen type II disc herniation at the L7/S1
intervertebral space (Jones et al., 1996b).
Subluxation, osteophytosis or thickening of the
articular processes (Jones et al., 1994).
Epidural fibrosis (Oliver et al., 1978; Sisson et al.,
1992; Jones et al., 1996b, 1999).
Soft tissue proliferation, usually of the joint capsule
or ligamentous structures (Jones et al., 1999).
Vascular compromise of the spinal nerves (Tarvin and
Prata, 1980; Jones et al., 1996a, 1999; Porter, 1996).
Osteochondrosis of the sacrum (Lang et al., 1992;
Hanna, 2001).
Instability and misalignment between L7 and S1
(Schmid and Lang, 1993).
The exact role of lumbosacral instability is unclear
and it is difficult to quantify regardless of the imaging
technique (Schmid and Lang, 1993; Fox et al., 1996).
This can pose a problem for surgical decision-making
(Algorithm 10.1).
Larger-breed dogs, particularly German shepherd
dogs, are affected most frequently although signs are
also reported in small dogs (Tarvin and Prata, 1980;
Lang et al., 1992; Jones et al., 1995b). Young, working
dogs that have been trained heavily are particularly
prone to this disorder (Wheeler, 1992; Danielsson and
Sjostrom, 1999; Jones et al., 2000). The condition is
rare in cats (Hurov, 1985; Stoll, 1996). Although
degenerative and congenital conditions are the focus of
this chapter, other disorders can affect the lumbosacral
spine. Tumors (see Chapter 12) and fractures (see
Chapter 13) should be considered in the differential
diagnosis. Discospondylitis also occurs at the lumbosacral junction (14.12); its management is discussed in
Chapter 14 (page 327). Articular (synovial) cysts have
also been described as a cause of lumbosacral pain in a
German shepherd dog (Webb et al., 2001).
CLINICAL SIGNS
Lumbosacral lesions can cause pelvic limb gait abnormalities, lameness, or lower motor neuron (LMN) neurological deficits. Pain is common but additional signs vary
depending on the nature and severity of the neurological
impairment (McKee, 1993) (Table 10.1). There may be
pain with no deficits; mild paresis with proprioceptive
deficits; or paraparesis, tail paralysis and incontinence
(Oliver et al., 1978; de Risio et al., 2001). Limb signs
include lameness or sciatic deficits affecting the caudal
thigh muscles and those distal to the stifle (Tarvin and
Prata, 1980; Meij et al., 1993). Lameness in performance
animals is often exacerbated by work (Jones et al., 2000).
Table 10.1 Clinical signs of lumbosacral disease

Mild
Severe
Low back pain palpation,

tail elevation, per rectum
Difficulty sitting
Difficulty jumping
Difficulty climbing
Pelvic limb lameness,
worsening with exercise
Hyperesthesia, pruritus
Mild pelvic limb paresis

Pelvic limb muscle atrophy
Tail paresis
Fecal incontinence
Anal hyporeflexia
Urinary incontinence
Self-mutilation
Incontinence results from pelvic and pudendal nerve

dysfunction. Urinary incontinence is usually LMN in
nature with dribbling of urine and a bladder that is
easily expressed by manual pressure (10.3) (see Control
of urinary function, page 350). Rarely, urinary dysfunction may be the only clinical sign (Tarvin and Prata,
1980). Fecal incontinence seems to be related mainly
to poor anal tone, which may be present even when the
anal reflex is intact.
Dogs with chronic degenerative lumbosacral lesions
may present with non-specific clinical signs, but low back
pain is quite different from that seen in thoracolumbar
lesions. Diagnosis of lumbosacral disease depends on
recognizing the historical features and clinical signs and
on a careful physical examination, which should pinpoint the source of pain.
DIAGNOSIS
Examination
Clinical signs may be vague in some dogs, which makes
accurate diagnosis difficult. A thorough physical,
orthopedic and neurological examination is essential,
along with a rectal examination. In view of the difficulty in interpreting some of the ancillary diagnostic
tests the clinical signs may provide the main basis for
reaching a diagnosis (Table 10.1).
The anal reflex, sphincter tone and tail tone should
be evaluated. Reduced or absent hock flexion during
the withdrawal reflex (2.27) is a sensitive indicator of
motor dysfunction in the sciatic nerve (but mild lesions
will only cause loss of proprioception). The patellar
reflex may also appear exaggerated if sciatic nerve
function is depressed. This phenomenon of pseudohyperreflexia must be differentiated from the increased
reflex that occurs with upper motor neuron (UMN)
deficits seen with lesions cranial to the L4 segment.
Pseudohyperreflexia results from decreased tone in the
muscles innervated by the sciatic nerve, which normally
183
184
counteract the extension of the stifle induced by the

patellar reflex (see page 27).
Hyperesthesia is a frequent finding and the hindquarters should be palpated carefully to locate the focus
of pain. Manipulation of the limbs and spine may provoke a pain response, but it may be difficult to distinguish pain associated with spinal disease from that
caused by orthopedic problems. Direct pressure over the
lumbosacral joint, either with the examiners knee
between the dogs rear limbs to support its pelvis and lift
its feet off the ground slightly or with the dog lying on its
side (or per rectum), may help to differentiate lumbosacral pain from orthopedic disease. Pain may also be
elicited by elevation of or traction on the tail and when
rotating the lumbosacral joint by swinging both rear
limbs from one side to the other (Sjostrum, 2003). A
final test is to have the dog stand up on its rear legs, with
its front feet resting on a chair and its spine extended,
and then apply pressure over the lumbosacral space.
If lameness or weakness worsen with exercise and
resolve to some degree with rest then they may be due
to neurogenic claudication, a condition well recognized
in humans. This is due to failure of arterial vasodilation in
affected nerve roots during exercise. As well as lameness, paresthesia and sphincter disturbance may also
occur (Markwalder, 1993; Porter, 1996). Neurogenic
claudication has also been recognized in dogs, one of
which had concomitant stranguria (Tarvin and Prata,
1980).
Differentials vary depending on whether the dog has
non-specific signs or obvious neurological deficits (Table
3.5 and 10.2). Dogs with the orthopedic diseases listed
in Table 10.2 have a normal neurological examination.
Dogs with pain and no neurological deficits could have
synovial cysts (Webb et al., 2001); tethering of the dural
sac (Huttmann et al., 2001; Shamir et al., 2001); inflammation of disc, meninges or nerve root; or possibly either
facet or sacroiliac joint pain (Pang et al., 1998) (see
Chapter 14). Dogs with degenerative myelopathy are
pain free and the withdrawal reflex is normal, but the
patellar reflex may be depressed because of dorsal nerve
root involvement (see Chapter 14). It may not be possible to differentiate degenerative lumbosacral disease,
discospondylitis and neoplasia on physical examination.
In some dogs, disorders such as degenerative myelopathy
(10.53) or a thoracolumbar disc(s) may coexist with
lumbosacral disease and this can confound the neurological localization (2.24). The deficits of lumbosacral disease are referable to the L4S3 nerve roots and so may
be difficult to differentiate from L4S3 lesions that
are situated within the dural sac (2.25). In such dogs,
imaging should extend as far forward as the L4/5 disc

space to include the sacral spinal cord.
Electrophysiology
Clinical electrophysiological studies are useful to confirm
LMN disease and nerve root dysfunction (Sisson et al.,
1992; Meij, 1993; de Risio et al., 2001; Cuddon, 2002;
Cuddon et al., 2003) (see page 60). Electromyography
of the limbs, tail and perineum may reveal spontaneous
activity, which is consistent with LMN involvement.
However, a normal electromyogram does not eliminate
the possibility of lumbosacral disease (de Risio et al.,
2001). Electromyography is particularly useful to reduce
the number of false-positive diagnoses associated with
MRI evidence of nerve root disorders (Nardin et al.,
1999) (see page 60). Nerve conduction studies and Fwave latencies may also be useful (Cuddon, 2002;
Cuddon et al., 2003) (see page 61). Abnormal findings
confirm LMN disease but do not specify the etiology.
Table 10.2 Differential diagnosis of lumbosacral disease
Mild clinical signs

Neurological disorders
Synovial cyst
Schmorls node
Discogenic pain
Facet joint pain
Congenital anomalyspina
bifida, dermoid sinus
Tethering of the terminal
spinal cord
Peripheral neuropathy
Neoplasiaspinal cord or
nerve root
Discospondylitis
Occult discospondylitis
Meningomyelitis
Polyradiculoneuritis
Orthopedic disorders
Coxofemoral arthritis
Cruciate rupture
Gracilis contracture
Psoas muscle injury
Pelvic limb lameness, other
Other
Prostatic disease
Urethral neoplasia
Anal sac adenocarcinoma
Severe neurological
deficits
Ischemic myelopathy
Neoplasia
Discospondylitis or epidural
abscess
Polyradiculoneuritis
Myelitis
Trauma
Lumbosacral disease
Radiography
SURVEY RADIOGRAPHY
It is advisable that the patient be anesthetized or
sedated heavily when radiographing the lumbosacral
joint. Rotation of the spine and pelvis must be avoided
(4.14, 4.15). The value of flexed and extended positional survey radiographs appears limited (Mattoon and
Koblik, 1993; Schmid and Lang, 1993). The main role
of survey radiographs is to rule out neoplasia (12.11A),
trauma (13.21) and discospondylitis (14.12).
Many clinically normal dogs have radiographic abnormalities of the lumbosacral junction (10.5). Conversely,
occasional dogs with lumbosacral disease will have
normal survey radiographs (Ness, 1994). Identification

of sacral osteochondrosis or transitional vertebrae
increases substantially the likelihood that the dogs signs
are due to cauda equina compression (Lang et al., 1992;
Morgan et al., 1993; Morgan, 1999) (10.6).
Neurological localization of a patient may indicate a
lesion of L4S3 spinal cord or nerve roots. It may be possible to define the location more accurately, for example
by the anal reflex. Consideration of the diagnostic imaging in many of these patients frequently focuses on the
lumbosacral joint alone. This policy runs the risk of missing lesions elsewhere in the vertebral column that are
involving the L4S3 cord segments (2.25) or ascribing
significance to incidental lesions (10.4).
MYELOGRAPHY
10.5 Lumbosacral spondylosis deformans and narrowing of the

intervertebral space. This type of change is seen frequently in
older, large-breed dogs but in itself is not diagnostic of disease
(Morgan et al., 1989; Ramirez and Thrall, 1998).
Myelography (or MRI) is particularly useful in dogs that

might have a spinal cord lesion (2.25). The subarachnoid
space extends beyond the lumbosacral junction in about
80% of dogs (Lang, 1988; Ness, 1994; de Risio et al.,
2001). Myelography can therefore be useful to assess
the low lumbar region as well as the rest of the spinal
cord (Ramirez and Thrall, 1998; Sjostrum, 2003).
Cervical injection is preferred, as it avoids the potential
for epidural contrast leakage in the area of interest.
Filling of the dural sac is improved by flexing the spine
(Lang, 1988). Lesions may cause dorsal elevation or
attenuation of the column; flexionextension studies
may also be useful (10.7), but false-positive results occur
(Lang, 1988; Watt, 1991; Danielsson and Sjostrom,
1999). Furthermore, a normal study does not preclude
cauda equina involvement (Ramirez and Thrall, 1998).
10.6 A, B: Five-year-old German shepherd dog with lumbosacral pain. Sacral OCD is present; the fragment moves with flexion and
extension (Lang et al., 1992). Subluxation of the articular facets (arrowhead) with foraminal narrowing is also evident (10.11, 10.46).
C: Four-year-old dog with transitional vertebra and left nerve root signature. 3D reconstruction of the CT scan shows foraminal
narrowing on the left due to a large articular facet (10.9). There is only one transverse process (arrow); the other is fused to the ilium.
185
186
Sjostrom, 1999) (10.8). Discography is abnormal if

more than 0.3 ml of contrast medium can be injected
(Barthez et al., 1994). When both are to be done the
discogram should precede the epidurogram. Positional
views may be employed with epidurography (Selcer
et al., 1988; Ramirez and Thrall, 1998). If both myelography and epidurography are planned, the myelogram
should be performed first, as the presence of epidural
contrast complicates myelographic interpretation. The
utility of discography may be enhanced if it is followed
by CT and discography in humans may be more sensitive in detecting anular tears than MRI (Ohnmeiss
et al., 1997; Milette et al., 1999).
DISCOGRAPHY AND EPIDUROGRAPHY

These two methods can be very useful, especially when
used in combination (Selcer et al., 1988; Barthez et al.,
1994; Ramirez and Thrall, 1998; Danielsson and
COMPUTED TOMOGRAPHY
CT is particularly useful as it shows clearly the vertebral
canal (10.10), lateral recesses (1.28), intervertebral
foramina (1.26, 1.27) and articular processes (10.9) in
cross-sectional images (Jones et al., 1995a, 1996b).
Reformatting can be used to create dorsal and sagittal
images (10.11) and facet joint subluxation may be visible using bone window images or 3D reconstructions
(Jones et al., 1994; Meij et al., 1996) (10.6, 10.48). CT
can also be used for dynamic studies (10.11) and is helpful when performed after surgery (Meij et al., 1996;
Jones et al., 2000) (10.54). The primary abnormalities
visible on CT are a replacement of epidural fat with soft
tissue density (10.10), which often represents epidural
fibrosis, and multilevel compression (Jones et al., 1996b)
(10.17). Use of intravenous contrast may improve surgical decision-making by identifying compressed tissues,
which tend to enhance (Jones et al., 1999) (10.40).
Some CT findings, such as loss of epidural fat, may not
be of clinical significance in older animals (Jones and
Inzana, 2000). In contrast to MRI, CT offers lower cost
B
10.7 Seven-year-old German shepherd dog with mild paraparesis
and lumbosacral pain. There is a change in the degree of
compression at the lumbosacral space between flexion and
extension. A: Good filling of the dural sac when the spine is
flexed; mild disc herniation is present at L6/7. B: There is dorsal
displacement of the dural sac in extension and marked
attenuation of contrast filling from L6/7 caudally. Fixation fusion
was done using two screws and a bone graft but fixation was
suboptimal (10.16).
10.8 A: Discogram demonstrating a large disc herniation. B: The needle was then withdrawn slightly to perform the
epidurogram, which also outlines the bulging disc. Attenuation of more than 50% of the canal is considered abnormal (Ramirez
and Thrall, 1998).
Lumbosacral disease
and thinner slice thickness along with better discrimination of bone spurs, articular process disease, soft tissue
calcification, and soft tissue gas opacities (Jones et al.,
2000). Scans should include at least the L6 vertebral
body and, ideally, additional images should also be made
through the L4/5 and L5/6 disc spaces (10.17).
component to the compression (10.12). One potential

disadvantage of MRI is over-diagnosis. Even expert,
human neuroradiologists tend to over-interpret the significance of lesions seen on MRI (Deyo, 1994; Jensen
MRI
This provides better soft tissue resolution than CT as
well as an ability to acquire images in multiple planes
without image degradation; earlier detection of disc
degeneration (1.9, 4.68) and evaluation of the entire
lumbar spine in a single sagittal examination (Ramirez
and Thrall, 1998; Jones et al., 2000) (1.9, 2.25, 10.12).
Slice thickness is often greater than for CT, however,
which increases volume averaging artefacts (Jones et al.,
2000). Transverse images provide the best visualization
of disc or foraminal anatomy (Adams et al., 1995) (1.9).
They also reveal lesions in a foramen that cannot be
detected by myelography or epidurography (Chambers
et al., 1997). Sagittal views made with the spine in
flexion and then extension help to identify a dynamic
B
10.10 A: Transverse CT scan at the level of the L7/S1 disc in a
dog with a normal lumbosacral spine. The nerve roots and
dural sac are surrounded by epidural fat. Same dog as in 10.2.
B: In this dog (also shown in 10.3) there is almost complete loss
of epidural fat and a general increase in soft tissue opacity.
B
10.9 Marked foraminal compression is evident on A: transverse
CT scan, and B: 3D reconstruction. Same dog as in 10.6C.
10.11 Sagittal reformatted CT scan made with the dog in an

A: extended position and B: flexed position. There is stenosis of
the vertebral canal and foramina with a bulging L7/S1 disc. An
OCD lesion is visible on the sacrum and the degree of compression
increases with extension. Same dog as in 10.6 A, B and 10.46.
187
188
et al., 1994). Therefore, the clinician should not rely

exclusively on imaging to confirm the diagnosis (BrantZawadzki et al., 1995; Gorman and Hodak, 1997;
Milette et al., 1999). Part of the problem is a confusion
in terminology. Classification of lesions into a disc bulge,
protrusion or extrusion is not very helpful in humans
(Milette et al., 1999) (see pages 13 and 58). In particular,
bulges or protrusions are common incidental findings in
people without back pain (Deyo, 1994; Jensen et al.,
1994). Criteria in humans for disruption of disc anatomy
include a loss of disc height; decrease in central disc signal on T2-weighting; or localized, peripheral hyperintensity on T2-weighting. Most outer anulus disruptions in
humans are symptomatic (Milette et al., 1999).
The choice of diagnostic tests is largely dependent
on availability and clinician preference. Physical and
neurological examination, survey radiography, CSF sampling at both the cerebello-medullary cistern (CMC)
and lumbar sites, and clinical electrophysiology are
recommended. Imaging by CT or MR is done when
available as they provide cross-sectional images (Ramirez
and Thrall, 1998; de Risio et al., 2001). CT and MRI can
also be used to compare the degree of compression with
the spine in flexed and extended positions (Bagley,
2003) (10.6A,B, 10.11, 10.12, 10.46). Either myelography or MRI should be performed if the dog has signs of
spinal cord disease (2.25). MRI is the best, single imaging modality overall for humans (Kent et al., 1992). The
position of the dog in the scanner can be very important;
a neutral or flexed position reduces compression compared to an extended position (Adams et al., 1995; Jones
et al., 2000). If CT or MRI is not available, myelography
is performed followed by discography and epidurography if necessary (Sisson et al., 1992; Sjostrum, 2003).
Some lesions detected by CT or MRI do not cause
clinical signs and this must be considered prior to instituting treatment (Deyo, 1994; Jensen et al., 1994; Jones
and Inzana, 2000). A herniated disc is detected in a high
proportion of people who have never had spinal pain and
so imaging alone, both in dogs and in humans, should not
serve as confirmation of a clinical diagnosis (Gorman and
Hodak, 1997). Electrophysiological testing provides an
additional means of independent confirmation of lesions
(Nardin et al., 1999).
TREATMENT
Most dogs are treated initially with rest and antiinflammatory medication (Table 15.2). This may be successful if pain is the main clinical sign; 34 months of
exercise restriction produced improvement in 8/16 dogs
(50%) (Ness, 1994). Such a course is seldom effective in
working dogs and signs often recur when normal activities are resumed (Danielsson and Sjostrom, 1999;
Janssens et al., 2000; Sjostrum, 2003).
Surgical treatment
Surgical treatment is indicated when non-surgical treatment has failed; in working dogs; and in those with pain
or neurological deficits (Sjostrum, 2003). Further indications for surgery include CT or MR findings of increased
soft tissue suggestive of epidural fibrosis, especially if it
enhances with contrast. The choice of surgical procedure
is then between dorsal laminectomy, distraction and
fusion, or a combination of the two. Definitive criteria for
these procedures are lacking. Laminectomy alone is not
effective for dogs with chronic incontinence (de Risio
et al., 2001). Fusion alone may be insufficient for dogs with
neurological deficits or severe pain (Slocum and Devine,
10.12 T2-weighted MRIs of a 7-year-old Labrador with lumbosacral pain. A: Image made with the spine in flexion. There is loss of
signal in the L7/S1 disc but little compression of the cauda equina (arrowhead). B: Image from the same dog with the spine in
extension. There is severe compression at L7/S1 caused by both disc and ligamentum flavum (arrowhead). A large disc herniation
was confirmed at surgery.
Lumbosacral disease
excised once the cauda equina is retracted laterally when

there is marked bulging of the disc (10.3, 10.39). Routine
fenestration may also be warranted (Danielsson and
Sjostrom, 1999). Any redundant joint capsule should also
be resected. Laminectomy often provides rapid relief of
pain with improvement of mild gait abnormalities and
minor neurological deficits (Danielsson and Sjostrom,
1999; Risio et al., 2001). It does not address instability if
this is a contributing factor. Decompression by laminectomy can be combined with fixation and fusion (Slocum
and Devine, 1998; Bagley, 2003) (10.4610.54). As dogs
with incontinence of more than 6 weeks duration
respond poorly to laminectomy alone they should also
undergo fixation and fusion (de Risio et al., 2001).
Results of dorsal laminectomy are shown in Table 10.3.
1998). These presentations warrant combined dorsal

laminectomy and fixationfusion. Additional indications
for distraction and fusion include a marked change in the
degree of compression between flexion and extension or
a telescoping of the facets, which suggest instability
(10.6A,B, 10.4610.48). An algorithm for surgical
decision-making is shown in Algorithm 10.1.
SURGERY
Dorsal laminectomy (10.1810.39)
Decompression of the cauda equina and spinal nerves can
be achieved by dorsal laminectomy (10.13), which can
be combined with foraminal decompression or even facetectomy (10.4010.45). The anulus fibrosus should be
Pain, mild
Neurological
deficits or
severe pain
without
incontinence
Dorsal
laminectomy
+/ foraminal
decompression
Fixation
fusion
Incontinence
Severe
radiculopathy

in lumbosacral disease.
Severe
instability
on imaging
Dorsal
laminectomy, +/
foraminal
decompression
AND fixation
fusion
Dorsal laminectomy,
foraminal
decompression
or facetectomy
+/ fusion
10.13 Postoperative appearance 1 year after dorsal laminectomy; this dog remained painful until it underwent a fixation fusion
(10.53). An asterisk indicates the L7/S1 disc space. A: Mid-sagittal 3D reconstruction after CT scan. B: Dorsal view of the 3D
reconstruction shown in A. Same dog as shown in 10.6A,B, 10.11 and 10.46.
189
190
Epidural fibrosis is a common finding; this is

probably a response to chronic compression (10.17B).
Magnification is recommended if attempts are made to
break down adhesions around nerve roots (Oliver et al.,
1978; Fingeroth et al., 1989; Sisson et al., 1992;
Markwalder, 1993; Jones et al., 1996b, 1999).
Foraminal decompression or
facetectomy (10.4010.45)
An increase in soft tissue within the foramen, loss of
epidural fat or facet osteophytosis are indicators of probable radiculopathy (Adams et al., 1995; Jones et al.,
1996b). Contrast enhancement after CT or MRI further
increases preoperative suspicion of foraminal disease
(Jones et al., 1999, 2000) (10.40). Together with radicular pain and delayed late waves (see Chapter 4), these
features are indicators for some form of foraminal
decompression (Oliver et al., 1978; Chambers et al.,
1988). The laminectomy can be simply undercut or
widened over the foramen and any osteophytes can also
be removed by undercutting (10.4110.43). However,
the foramen can be difficult to examine fully until the
facet is removed (Watt, 1991). The risks and benefits of
facetectomy are unclear as is the need for subsequent
stabilization (Slocum and Devine, 1998) (10.45).
Lumbosacral stiffness is decreased markedly after bilateral facetectomy (Smith and Bebchuk, 2002).
Dorsal fusionfixation (10.4610.54)

There are a number of similarities between lumbosacral
disease and caudal cervical spondylomyelopathy (see
Chapter 11). A rationale can therefore be made for

lumbosacral distraction and stabilization (Oliver et al.,
1978). The principle is the same, namely to open the
vertebral canal and intervertebral foramen, relieve
pressure on neural tissues, and prevent abnormal motion.
Distraction of the dorsal aspect of L7 and S1 is achieved
using screws placed through the articular processes of
L7/S1 and into the body of the sacrum (10.14). Fusion is
promoted by removing the articular cartilage from the
facet joints and by a cancellous bone graft (10.52). This
can be combined with dorsal laminectomy, especially
for dogs with marked neurological deficits, severe pain,
10.14 Middle-aged German shepherd dog that presented with

mild pelvic limb weakness, lumbosacral pain, anal hyporeflexia
and dribbling urine. CT scan revealed marked telescoping of the
sacrum relative to L7. Screw fixation was performed using
4.0 mm cancellous screws. The screws had loosened slightly at
6-week follow-up but the dog had an excellent outcome with
resolution of all signs for 3 years.
Table10.3 Results after dorsal laminectomy ( or foraminal decompression)

Oliver et al.,
1978
Chambers et al.,
1988
Danielsson and
Sjostrom, 1999
Risio et al.,
2001
Dogs (n)
10
26
131
69
Good outcome* n (%)
9 (90)
19 (73)
122 (93)
54 (78)
Mean follow-up months (range)
13 (148)
21 (255)
26 /17 (573)
38 /22 (696)
Median follow-up months
18
N/A
36
Incontinence resolved
3/4
1/8
N/A
5/11
*Substantial improvement and resolution of any incontinence.

N/A, not available.
Lumbosacral disease
or a large disc herniation (Slocum and Devine, 1998;

Bagley, 2003).
distal cortex in order to avoid damage to the lumbosacral

trunk of the sciatic nerve (1.8, 10.15).
COMPLICATIONS
Early postoperative
These have been reviewed extensively after lumbar

disc surgery in humans but have not been reviewed for
dogs (Stolke et al., 1989; Fritsch et al., 1996; Gibson
et al., 2002). Intraoperative, early postoperative and
late complications are listed in Table 10.4.
Seroma formation is particularly likely in the lumbosacral region if there is any dead space (Oliver et al.,
1978; Slocum and Devine, 1986; Chambers et al.,
1988; Watt, 1991; Ness, 1994). Infection can occur,
especially as some dogs have established urinary tract
infection (UTI) (Oliver et al., 1978). Occasionally, a dog
may be unable to void after surgery because of increased
sympathetic tone (10.3), analogous to that seen after
sacrocaudal injury (see page 351).
If pain persists after surgery this suggests implant failure, residual compression or possibly instability (10.53).
Intraoperative
The surgeon must be vigilant while using a scalpel in the
vertebral canal to fenestrate the L7/S1 disc, and also
while using a bur to undercut a facet, to avoid inadvertent
damage to the cauda equina. In some cases an AO drill
guard can be used to protect neural tissues from the bur.
The base of the L7 articular process must not be thinned
excessively (10.41, 10.42), especially when foraminal
decompression is performed. In addition, over-tightening
must be avoided when placing a screw across the L7/S1
facet or the facet could fracture. When placing screws,
the drill bit, tap and screw must not extend beyond the
Table 10.4 Intraoperative, early postoperative and late

postoperative complications
Intraoperative
Iatrogenic nerve
root injury
Fracture of L7
articular facet
Poor implant
position (10.16)
Spinal nerve injury
Inadequate
decompression
Early
postoperative
Late
postoperative
Implant failure
Instability (10.16)
Dorsal seroma
Infection
Increased sphincter
tone
Implant failure
Recurrence of
signs (10.53)
Dural herniation
Scar formation
Infection (10.45)
Late postoperative
Implant loosening or failure is always a risk after instrumented fusion (Fox et al., 1996; Gibson et al., 2002).
It is usually due to poor implant selection or suboptimal technique (10.16). Sometimes implant failure does
not cause clinical problems, especially if physiological
fusion progresses fast enough (Slocum and Devine,
1986; Bagley, 2003) (10.14, 10.53).
Postoperative imaging for recurrent lumbosacral pain
may reveal residual disc material or fibrous scar, especially when intravenous contrast is used (Jones et al.,
1999, 2000). Postoperative scarring has been reported
in several studies (Adams et al., 1995; Danielsson and
Sjostrom, 1999; Risio et al., 2001). Herniation of the
dural sac through the laminectomy defect has also been
reported; this may be secondary to tearing of the dura
(Lang, 1988; Markwalder, 1993; Fox et al., 1996). Patient
age and operative time are predictors of complications
in humans undergoing lumbar disc surgery (Stolke
et al., 1989). Additional reasons for failure to improve
include surgery done at the wrong site, insufficient
10.15 Nine-year-old Pug that underwent dorsal laminectomy and fixation

fusion using 2.7 mm screws and a
bone graft. These screws are actually
still contained within bone because of
the contour of the ventral sacrum
(1.28B, 10.46, 10.50, 10.54). However,
if they were any longer there would be
risk of damaging the proximal sciatic
nerve, which lies in very close proximity
to the ventral surface of the sacroiliac
joint. Same dog as in 10.47.
A
191
192
10.16 A: Eight week follow-up radiograph reveals implant failure due to poor
technique. The screws were put in at too
steep an angle in the sagittal plane.
Immediate postoperative radiographs
and fluoroscopy showed the screws to
be in bone and not the disc space; the
dog was confined in the hope that fusion
would occur before implant failure. One
screw has now backed out (10.14,
10.15, 10.53). B: 3D reconstruction
shows the screws also do not diverge
sufficiently from each other in a transverse plane (10.50).
A
decompression or removal of disc material, a second disc

lesion, extravertebral compression or nerve root trauma
(Fritsch et al., 1996; Janssens et al., 2000).
Facet fracture can cause persistent signs after dorsal
laminectomy. It can occur after excessive thinning of
the base of the L7 facet at laminectomy or from overtightening of a screw (Adams et al., 1995). Repeat
imaging is recommended when postoperative problems
develop as very good outcomes have been reported
after a second surgery (Danielsson and Sjostrom, 1999;
Jones et al., 2000; Risio et al., 2001).
POSTOPERATIVE CARE
(see Chapter 15)
Strict rest is enforced for 3 months following surgery,

followed by a gradual return to fitness over a further 2
months. Dogs that are allowed to move freely too soon
after surgery are at risk of making a poor recovery
(de Risio et al., 2001). For working and athletic dogs,
an additional month of gradual transition to full work is
recommended (Sjostrum, 2003). After fixationfusion
the dog is confined for at least 68 weeks or until there
is radiographic evidence of fusion. Physical therapy, leash
walking and swimming are used with a gradual return to
normal activity over a further 23 months (Bagley, 2003).
Long-term non-steroidal anti-inflammatory drugs
(NSAIDs) have been suggested to reduce postoperative
scarring and improve outcome but this is unsubstantiated (Janssens et al., 2000).
PROGNOSIS
Laminectomy provides rapid relief of pain in most dogs.
Similarly, lameness and mild neurological deficits usually
improve rapidly. More severe deficits probably carry a
less favorable prognosis (Denny et al., 1982; Chambers
et al., 1988; Chambers, 1989). Few studies document
long-term follow-up after lumbosacral surgery. Studies
with mean follow-up periods beyond 1 year show overall
success rates from 73 to 93% (Table 10.3). The presence
of an osteochondritis dissecans (OCD) lesion does not

appear to affect prognosis after surgery (Hanna, 2001).
The prognosis for working dogs is similar with 67 of 88
(76%) returning to normal duties after dorsal decompression (Danielsson and Sjostrom, 1999; Jones et al., 2000).
Some of these working dogs had also undergone foraminal decompression or facetectomy (Jones et al., 2000).
The presence of fecal incontinence, and both the presence and the duration of urinary incontinence prior to
surgery, are negative prognostic factors in dogs with lumbosacral disease. Incontinence for longer than 6 weeks
carries a guarded prognosis (Risio et al., 2001). Success
rates after surgery are known to decline over time in
humans and this also seems to be true in dogs (Fox et al.,
1996; Javid and Hadar, 1998; Janssens et al., 2000).
Few specific results are available for either foraminal
decompression or facetectomy. Foraminal decompression
in 12 dogs did not improve outcome compared to 15 dogs
in which laminectomy alone was performed (de Risio
et al., 2001). Facetectomy gave good long-term results in
15/15 dogs but many were of small breeds (Tarvin and
Prata, 1980). Nine of 11 larger-breed dogs improved
after facet removal but were often left with residual signs
(Denny et al., 1982; Watt, 1991; Ness, 1994).
Repeat surgery is indicated for dogs with a poor initial
outcome or that suffer a recurrence of signs following
dorsal laminectomy. Good results were reported for
four of six dogs undergoing a second surgery; with scar
tissue being the most common finding on re-operation
(Danielsson and Sjostrom, 1999; de Risio et al., 2001).
There are only two limited studies with long-term follow-up after fixationfusion. All eight dogs in one study
and five of five in another had good outcomes (Slocum
and Devine, 1986; Meheust, 2000). Although one study
in humans reported better results after fusion than after
decompression alone, a large meta-analysis of all available data in people found that fusion did not improve
outcome and was associated with a higher complication
rate (Fox et al., 1996; Gibson et al., 2002).
Lumbosacral disease

1. What is the significance of instability and how should it be assessed (Fox et al., 1996; Gibson et al., 2002) (10.53)?
2. What is the role of facetectomy and should it be followed by stabilization (10.45)?
3. Is routine dorsal anulectomy of benefit or does it increase collapse of the lumbosacral space (Danielsson and Sjostrom,
1999; Janssens et al., 2000; de Risio et al., 2001)?
4. Does grafting at the disc space promote fusion (page 328) (Auger et al., 2000)?
5. Why is epidural fibrosis so common and is it a consequence of multilevel compression (Jayson, 1992; Porter, 1996)?
6. How should lesions at more than one location be dealt with; that is, should decompression be extended over L6/7 for the
dog in 10.17A?
7. What are the long-term outcomes after fixationfusion?
8. When is fusion indicated and when is decompression indicated?
10.17 A: This dog has a lesion at
L7/S1 and another at L6/7. B: This
dog has an isolated L7/S1 lesion.
Several studies have emphasized the
role of multilevel compression (Jayson,
1992; Markwalder, 1993; Jones et al.,
1996b; Porter, 1996).
REFERENCES
Adams, W.H., Daniel, G.B., Pardo, A.D., Selcer, R.R. (1995) Magnetic resonance imaging of the caudal lumbar and lumbosacral spine in 13 dogs
(19901993). Veterinary Radiology and Ultrasound 36, 313.
Auger, J., Dupuis, J., Quesnel, A., Beauregard, G. (2000) Surgical treatment of lumbosacral instability caused by discospondylitis in four dogs.
Bagley, R.S. (2003) Surgical stabilization of the lumbosacral joint. In:
Barthez, P.Y., Morgan, J.P., Lipsitz, D. (1994) Discography and epidurography for evaluation of the lumbosacral junction in dogs with cauda equina
syndrome. Veterinary Radiology and Ultrasound 35, 152157.
Brant-Zawadzki, M.N., Jensen, M.C., Obuchowski, N., Ross, J.S., Modic,
M.T. (1995) Interobserver and intraobserver variability in interpretation
of lumbar disc abnormalities. A comparison of two nomenclatures. Spine
20, 12571263; discussion 1264.
Burger, R., Lang, J. (1993) Kinematic study of the lumbar and lumbosacral
spine in the German Shepherd Dog. Part 2. Results of experiments.
Schweizer Archiv fur Tierheilkunde 135, 3543.
Chambers, J.N. (1989) Lumbosacral stenosis. Confusion, surgery, prognosis. Veterinary Medicine Report 1, 228229.
Chambers, J.N., Barbera, A., Selcer, B.A. (1988) Results of treatment of
degenerative lumbosacral stenosis in dogs by exploration and excision.
Veterinary Comparative Orthopaedics and Traumatology 3, 130133.
Chambers, J.N., Selcer, B.A., Sullivan, S.A., Coates, J.R. (1997) Diagnosis
of lateralized lumbosacral disk herniation with magnetic resonance
imaging. Journal of the American Animal Hospital Association 33, 296299.
Coates, J.R. (1999) Urethral dyssynergia in lumbosacral syndrome.
American College of Veterinary Internal Medicine 1999; 299301.
Cuddon, P.A. (2002) Electrophysiology in neuromuscular disease.
Veterinary Clinics of North America, Small Animal Practice 32, 3162.
Cuddon, P.A., Murray, M., Kraus, K. (2003) Electrodiagnosis. In: D. Slatter
(ed.) Textbook of Small Animal Surgery, 3rd edn, 11081117.
Danielsson, F., Sjostrom, L. (1999) Surgical treatment of degenerative
lumbosacral stenosis in dogs. Veterinary Surgery 28, 9198.
Denny, H.R., Gibbs, C., Holt, P.E. (1982) The diagnosis and treatment of
cauda equina lesions in the dog. Journal of Small Animal Practice 23,
425443.
de Risio, L., Sharp, N.J.H., Olby, N.J., Munana, K.R., Thomas, W.B. (2001)
Predictors of outcome after dorsal decompressive laminectomy for degenerative lumbosacral stenosis in dogs: 69 cases (19871997). Journal of the
Deyo, R. (1994) Magnetic resonance imaging of the lumbar spine. Terrific
test or tar baby? New England Journal of Medicine 331, 115116.
Ebraheim, N.A., Haman, S.P., Xu, R., Stanescu, S., Yeasting, R.A. (2000)
The lumbosacral nerves in relation to dorsal S1 screw placement and
their locations on plain radiographs. Orthopedics 23, 245247.
Fingeroth, J.M., Johnson, G.C., Burt, J.K., Fenner, W.R., Cain, L.S. (1989)
Neuroradiographic diagnosis and surgical repair of tethered cord syndrome
193
194
in an English Bulldog with spina bifida and myeloschisis. Journal of the

Fletcher, T.F., Kitchell, R.L. (1966) Anatomical studies on the spinal cord segments of the dog. American Journal of Veterinary Research 27, 17591767.
Fox, M., Onofrio, B., Hanssen, A. (1996) Clinical outcomes and radiological
instability following decompressive lumbar laminectomy for degenerative
spinal stenosis: a comparison of patients undergoing concomitant arthrodesis versus decompression alone. Journal of Neurosurgery 85, 793802.
Fritsch, E.W., Heisel, J., Rupp, S. (1996) The failed back surgery syndrome: reasons, intraoperative findings, and long-term results: a report
of 182 operative treatments. Spine 21, 626633.
Gibson, J.N.A., Waddell, G., Grant, I.C. (2002) Surgery for degenerative
lumbar spondylosis (Cochrane Review). Issue 1 ed: The Cochrane
Library. http:/www.update-software.com/cochrane/
Hanna, F.Y. (2001) Lumbosacral osteochondrosis: radiological features and
surgical management in 34 dogs. Journal of Small Animal Practice 42,
272278.
Hurov, L. (1985) Laminectomy for treatment of cauda equina syndrome in a
cat. Journal of the American Veterinary Medical Association 186, 504505.
Huttmann, S., Krauss, J., Collmann, H., Sorensen, N., Roosen, K. (2001)
Surgical management of tethered spinal cord in adults: report of 54
cases. Journal of Neurosurgery 95 (2 Suppl), 173178.
Janssens, L.A.A., Moens, Y., Coppens, P., Peremans, K., Vinck, H. (2000)
Lumbosacral degenerative stenosis in the dog: the results of dorsal
decompression with dorsal anulectomy and nuclectomy. Veterinary and
Javid, M., Hadar, E. (1998) Long-term follow-up review of patients who
underwent laminectomy for lumbar stenosis: a prospective study.
Jayson, M.I. (1992) The role of vascular damage and fibrosis in the pathogenesis of nerve root damage. Clinical Orthopaedics and Related
Research 279, 4048.
Jensen, M., Brant,Z.M., Obuchowski, N., Modic, M., Malkasian, D., Ross,
J. (1994) Magnetic resonance imaging of the lumbar spine in people without back pain. New England Journal of Medicine 331, 6973.
Jones, J.C., Inzana, K.D. (2000) Subclinical CT abnormalities in the lumbosacral spine of older large-breed dogs. Veterinary Radiology and
Jones, J.C., Wilson, M.E., Bartels, J.E. (1994) A review of high resolution
computed tomography and a proposed technique for regional examination of the canine lumbosacral spine. Veterinary Radiology and
Jones, J.C., Cartee, R.E., Bartels, J.E. (1995a) Computed tomographic
anatomy of the canine lumbosacral spine. Veterinary Radiology and
Jones, J.C., Wright, J.C., Bartels, J.E. (1995b) Computed tomographic
morphometry of the lumbosacral spine of dogs. American Journal of
Jones, J.C., Hudson, J.A., Sorjonen, D.C., Hoffman, C.E., Braund, K.G.,
Wright, J.C., Garrett, P.D., Bartels, J.E. (1996a) Effects of experimental
nerve root compression on arterial blood flow velocity in the seventh lumbar spinal ganglion of the dog: measurement using intraoperative Doppler
ultrasonography. Veterinary Radiology and Ultrasound 37, 133140.
Jones, J.C., Sorjonen, D.C., Simpson, S.T., Coates, J.R., Lenz, S.D.,
Hathcock, J.T., Agee, M.W., Bartels, J.E. (1996b) Comparison between
computed tomographic and surgical findings in nine large-breed dogs with
lumbosacral stenosis. Veterinary Radiology and Ultrasound 37, 247256.
Jones, J.C., Shires, P.K., Inzana, K.D., Sponenberg, D.P., Massicotte, C.,
Renberg, W., Giroux, A. (1999) Evaluation of canine lumbosacral stenosis using intravenous contrast-enhanced computed tomography. Veterinary
Jones, J.C., Banfield, C.M., Ward, D.L. (2000) Association between postoperative outcome and results of magnetic resonance imaging and computed
tomography in working dogs with degenerative lumbosacral stenosis.
Jones, J.C., Shires, P.K., Inzana, K.D., Mosby, A.D., Sponenberg, D.P.,
Lanz, O.I. (2002) Use of computed tomographic densitometry to quantify
contrast enhancement of compressive soft tissues in the canine lumbosacral
vertebral canal. American Journal of Veterinary Research 63, 733737.
Kent, D.L., Haynor, D.R., Larson, E.B., Deyo, R.A. (1992) Diagnosis of lumbar spinal stenosis in adults: a metaanalysis of the accuracy of CT, MR, and
myelography. American Journal of Roentgenology 158, 11351144.
Lang, J. (1988) Flexionextension myelography of the cauda equina.
Lang, J., Hani, H., Schawalder, P. (1992) A sacral lesion resembling osteochondrosis in the German Shepherd Dog. Veterinary Radiology and
Markwalder, T.M. (1993) Surgical management of neurogenic claudication
in 100 patients with lumbar spinal stenosis due to degenerative spondylolisthesis. Acta Neurochirurgica (Wein) 120, 136142.
Mattoon, J.S., Koblik, P.D. (1993) Quantitative survey radiographic evaluation of the lumbosacral spine of normal dogs and dogs with degenerative
lumbosacral stenosis. Veterinary Radiology and Ultrasound 34, 194206.
McKee, W.M. (1993) Differential diagnosis of cauda equina syndrome.
In Practice 15, 243244.
McKee, W.M., Mitten, R.W., Labuc, R.H. (1990) Surgical treatment of
lumbosacral discospondylitis by a distractionfusion technique. Journal
Meheust, P. (2000) A new surgical technique for lumbosacral stabilization:
arthrodesis using the pedicle screw fixation. A clinical study of 5 cases.
Pratique Medicale and Chirurgicale de lAnimal de Compagnie 35, 201207.
Meheust, P., Mallet, C., Marouze, C. (2000) A new surgical technique for
lumbosacral stabilization: arthrodesis using the pedicle screw fixation.
Anatomical aspects. Pratique Medicale and Chirurgicale de lAnimal de
Compagnie 35, 193199.
Meij, B.P. (1993) Pelvic limb lameness associated with nerve root compression: diagnosis, neurosurgery and follow-up in two dogs. Veterinary
Surgery 22, 249.
Meij, B.P., Voorhout, G., Wolvekamp, W.T.C. (1996) Epidural lipomatosis
in a six-year-old dachshund. Veterinary Record 138, 492495.
Morgan, J.P. (1999) Transitional lumbosacral vertebral anomaly in the dog:
a radiographic study. Journal of Small Animal Practice 40, 167172.
Morgan, J.P., Hansson, K., Miyabayashi, T. (1989) Spondylosis deformans
in the female Beagle dog: a radiographic study. Journal of Small Animal
Morgan, J.P., Bahr, A., Franti, C.E., Bailey, C.S. (1993) Lumbosacral transitional vertebrae as a predisposing cause of cauda equina syndrome in
German Shepherd Dogs; 161 cases (19871990). Journal of the
Nardin, R.A., Patel, M.R., Gudas, T.F., Rutkove, S.B., Raynor, E.M. (1999)
Electromyography and magnetic resonance imaging in the evaluation of
radiculopathy. Muscle and Nerve 22, 151155.
Ness, M.G. (1994) Degenerative lumbosacral stenosis in the dog: a review
of 30 cases. Journal of Small Animal Practice 35, 185190.
O Brien, D. (1988) Neurogenic disorders of micturition. Veterinary
Clinics of North America, Small Animal Practice 18, 529544.
Ohnmeiss, D.D., Vanharanta, H., Ekholm, J. (1997) Degree of disc disruption and lower extremity pain. Spine 22, 16001605.
Oliver, J.E., Jr., Selcer, R.R., Simpson, S. (1978) Cauda equina compression from lumbosacral malarticulation and malformation in the
dog. Journal of the American Veterinary Medical Association 173,
207214.
Palmer, R.H., Chambers, J.N. (1991) Canine lumbosacral diseases. Part II.
Definitive diagnosis, treatment, and prognosis. Compendium on
Continuing Education for the Practicing Veterinarian 13, 213222.
analysis of 104 cases. Acta Anaesthesiological Sinica 36, 7174.
Porter, R.W. (1996) Spinal stenosis and neurogenic claudication. Spine 21,
20462052.
Ramirez, O., III, Thrall, D.E. (1998) A review of imaging techniques for
canine cauda equina syndrome. Veterinary Radiology and Ultrasound 39,
283296.
Schmid, V., Lang, J. (1993) Measurements on the lumbosacral junction in
normal dogs and those with cauda equina compression. Journal of Small
Lumbosacral disease
Selcer, B.A., Chambers, J.N., Schwensen, K. (1988) Epidurography as a diagnostic aid in canine lumbosacral compressive disease: 47 cases (19811986).
Veterinary Comparative Orthopaedics and Traumatology 2, 97103.
Shamir, M., Rochkind, S., Johnston, D. (2001) Surgical treatment of tethered spinal cord syndrome in a dog with myelomeningocele. Veterinary
Record 148, 755756.
Sisson, A.F., LeCouteur, R.A., Ingram, J.T., Park, R.D., Child, G. (1992)
Diagnosis of cauda equina abnormalities by using electromyography,
discography, and epidurography in dogs. Journal of Veterinary Internal
Medicine 6, 253263.
Sjostrum, L. (2003) Decompression of lumbosacral disease. In: D. Slatter
Slocum, B., Devine, T. (1986) L7S1 fixationfusion for treatment of
cauda equina compression in the dog. Journal of the American Veterinary
Slocum, B., Devine, T. (1998) L7S1 fixationfusion technique for cauda
equina syndrome. In: M.J. Bojrab (ed.), Current Techniques in Small
Animal Surgery, 5th edn. 861864. Philadelphia: Lea & Febiger.
Smith, M.E.H., Bebchuk, T. (2002) An in vitro biomechanical study of the

effects of surgical modification upon the canine lumbosacral spine.
Stolke, D., Sollmann, W.P., Seifert, V. (1989) Intra- and postoperative
complications in lumbar disc surgery. Spine 14, 5659.
Stoll, S.G. (1996) Alternative method of traction stabilization of L7S1.
San Francisco: American College of Veterinary Surgeons Symposium,
116117.
Tarvin, G., Prata, R.G. (1980) Lumbosacral stenosis in dogs. Journal of the
Watt, P.R. (1991) Degenerative lumbosacral stenosis in 18 dogs. Journal of
Ultrasound 42, 913.
Wheeler, S.J. (1992) Lumbosacral disease. Veterinary Clinics of North
Wright, J.A. (1980) Spondylosis deformans of the lumbosacral joint in
dogs. Journal of Small Animal Practice 21, 4555.
PROCEDURES
Discectomy may or may not provide additional benefit to laminectomy although it is logical for dogs in which the
disc is causing significant compression (Danielsson and Sjostrom, 1999; Janssens et al., 2000; de Risio et al.,
2001). A cancellous bone graft can be placed into the disc space following curettage and this may enhance
fusion (Auger et al., 2000) (10.52, 14.12).
10.18 Positioning of dog for

lumbosacral surgery.
Note that the pelvic limbs
are drawn forward. The
interarcuate space and
dorsal anulus are
widened by either supporting the pelvis over a
10.18
sandbag or by pulling the
hocks further forward
with the hips flexed (Oliver et al., 1978; Watt, 1991; Slocum and Devine, 1998; Bagley, 2003; Sjostrum,
2003). The dogs head is to the left in all illustrations.
10.19 The important landmarks are being palpated.

The surgeons right hand is on the cranial dorsal iliac spine. The left index finger is on the
spinous process of L6. The spinous process of
L7 is shorter and often cannot be palpated
(10.2B, 10.24, 10.26).
10.19
195
196
10.20 Here the upper instrument is over the lumbosacral space, and the lower instrument on the
wing of the left ilium.
10.20
10.21 The skin incision is made just lateral to

the midline. It extends from the spinous
process of L5 to the caudal end of the
fused spinous processes of the sacrum.
Here the skin and superficial fascia (a)
have been incised revealing the deep
lumbodorsal fascia (b). The fascia,
which may be quite thick, is undermined
slightly to facilitate closure (8.13) prior
to incision.
10.22 The lumbodorsal fascia is incised around

and between the spinous processes
(a). This reveals the epaxial musculature.
Here the fascia has been incised and
retracted on the lower side of the
illustration but is still intact on the upper
side (b). The lumbodorsal fascia merges
with the interspinous ligament that lies
between the spinous processes.
(c) Transverse process L6. (d) Transverse
process L7. (e) Wing of ilium.
10.21
b
a
10.22
10.23 Here the fascia has been retracted on

both sides, revealing the epaxial muscles.
The spinous processes are seen in the
midline (a). The interspinous fascia
between the spinous processes is thick.
10.23
Lumbosacral disease
10.24 The muscles are elevated from the spinous processes and retracted. The L6
(a), L7 (b), and sacral (c) spinous
processes are visible. Note that the spinous process of L6 is taller than L7.
c
b
10.24
10.25 A periosteal elevator is used to remove the

muscles from the spinous processes on
both sides of the vertebral column. It is
useful to insert self-retaining retractors to
maintain the exposure (see 10.24 for key).
10.25
10.26 The muscles have been retracted further.

In the illustration, L6 (a), L7 (b) and the
sacrum (c) have been exposed.
c
b
10.26
10.27 Closer view of 10.26. The curette is on the

ligamentum flavum. (a) L7 vertebrae.
(b) Sacrum. In some dogs with
lumbosacral disease the sacrum
telescopes cranioventrally and the dorsal
lamina of the sacrum lies adjacent to, or
even under, the lamina of L7 (10.46).
There is then no interarcuate space.
a
10.27
197
198
10.28 The spinous processes of L7 (shown here)

and S1 are removed with rongeurs or
bone cutters.
10.28
10.29 The laminectomy is made with a power

bur. In this illustration the laminectomy
has been commenced in L7; note the
dark red cancellous bone (a). At least half
of L7 is removed and most of the sacrum
(Danielsson and Sjostrom, 1999). The
remnant of sacral spinous process is
visible (b). The laminectomy should be
restricted to the lamina at this stage,
preserving the articular processes.
b
a
10.29
10.30 The laminectomy has been continued into

both L7 and the sacrum down to the
inner cortical bone. The margin between
dark cancellous bone and white cortical
bone is seen clearly (arrows).
10.30
Lumbosacral disease
10.31 The ligamentum flavum is removed by

sharp dissection. Great care must be
taken not to penetrate too deeply, risking
damage to the cauda equina
(Markwalder, 1993; Fox et al., 1996), as
there is no bony protection overlying the
neural structures at this level. In dogs with
marked compression and dorsal elevation
of the cauda equina it is safer to enter
the vertebral canal over mid-L7 or
mid-sacrum and then peel away the
ligamentum flavum from either side.
10.31
10.32 Removal of the ligamentum flavum

exposes the epidural fat (a) and the cauda
equina (b). Epidural fat should be removed
only when necessary.
10.32
10.33
10.33 The final shelf of bone is thinned with

the bur.
199
200
10.34 It is best to bur the bone to eggshell thickness

to allow easy removal. The tip of the probe is
visible through the thinned bone (arrow).
10.34
10.35 The inner shelf of bone is removed carefully with

rongeurs. The width of this laminectomy is
greater than is ideal if screw fixation were to be
performed subsequently (compare to 10.41,
10.51 and 10.52)
10.35
10.36 The laminectomy is completed with rongeurs or

a curette. The cauda equina is then inspected.
The S1 nerve roots (a) are usually seen adjacent
to the dural tube of the cauda equina (b). The
L7 nerve root lies further laterally, in the lateral
recess and under the articular process (1.27,
1.28). A herniated disc may either be visible
directly (10.44), or palpable by running a probe
along the floor of the vertebral canal (which is
flat in normal dogs) (10.38A).
10.36
10.37 Laminectomy has been completed. Part of L7

nerve root (a) is seen in the lateral recess,
before exiting the intervertebral foramen. (b) S1
nerve root. (c) S2 nerve root. (d) Remainder of
cauda equina.
10.37
c d
Lumbosacral disease
10.38 Ten-year-old Rottweiler that had been dribbling

urine for 2 months and had lumbosacral pain,
pelvic limb ataxia, and marked reduction of
right-sided hock flexion with a lick granuloma over
the right tarsus. A CT myelogram revealed that
the terminal dural sac was displaced to the left by
a large soft tissue mass. A: The mass was
palpable just in front of the lumbosacral joint.
B: Retraction of the right S1 nerve root revealed a
huge disc extrusion (*). C: Continued retraction
reveals the extent of the mass compressing the
right L7 nerve root. D: After fenestration with a
scalpel the mass was removed using rongeurs to
leave a large crater (arrowhead), which was then
covered with a thin fat graft. The clinical signs
resolved slowly after surgery. Two years later the
dog was pain free, continent and its hock lesion
had healed.
10.38
A
201
202
10.39 Here the cauda equina is being retracted medially

to allow a bulging disc to be fenestrated on one
side (arrowheads). Half of the anulus fibrosus is
incised carefully with a #11 blade and removed
with fine rongeurs. The outermost incision is just
medial to the venous plexus (Danielsson and
Sjostrom, 1999). The cauda equina is then
retracted to the other side and the second half of
the disc is fenestrated. The nucleus pulposus is
then removed with a curette.
10.39
Foraminal decompression and facetectomy (10.4010.45)

Enhancement of soft tissues within a foramen after intravenous contrast is an indication for foraminal decompression (see page 186, 190; 10.40). Densitometry can be used to assess the degree of contrast enhancement
more accurately although it does not predict the nature of the enhancing tissue (Jones et al., 2002).
Facetectomy can be followed by stabilization using screws and bone graft (Slocum and Devine, 1998), screws
and bone cement (10.45), or screws with connecting bars (Meheust, 2000; Meheust et al., 2000).
Wound closure is routine. A free fat graft, pedicle fat graft or Gelfoam may be placed over the laminectomy site.
Closure must avoid leaving any dead space.
10.40 A: Pre-contrast, and

B: post-contrast
transverse CT images
obtained at the level of
the L7/S1 intervertebral
foramina in a dog with
10.40
degenerative
lumbosacral stenosis.
A
B
Contrast enhancing
tissue fills the left foramen. A small area of enhancement can also be seen on the right side of the
ventral canal (Jones et al., 1999).
10.41 Reconstruction following a dorsal laminectomy.

If imaging indicates nerve root compression, the
laminectomy can be widened still further, initially
only to the medial limit of the facet joint
capsule (Danielson and Sjostrom, 1999). The
intervertebral foramen can be explored with a
probe; the normal L7 nerve root should move
a few millimetres without undue tension.
10.41
Lumbosacral disease
10.42 If a foraminotomy is to be performed, the

laminectomy is extended laterally on that side
dorsal to the foramen. The relationship between
the lumbosacral disc and the foramen is shown
in 1.26 and 1.27; the foramen is cranial to the
disc (10.44). The articular processes should be
preserved and care taken not to weaken the
base of the L7 facet when decompressing the
foramen. Extending the laminectomy all the way
into the foramen will isolate L7 articular process.
10.43 Undercutting of a foramen to remove osteophytes

is illustrated on this transverse CT image. The level
of the laminectomy is shown by white arrowheads.
Additional undercutting of the L7 or S1 facet can
be performed with a curette or bur, as shown by
the area outlined with black arrowheads (Palmer
and Chambers, 1991). Great care is taken to
protect the nerve roots while drilling. The facet
should not be weakened excessively, particularly if
screw fixation is also to be performed.
10.42
10.43
10.44 Facetectomy may be considered if foraminal

decompression does not expose enough of the
L7 nerve root. Here the dural tube is displaced
medially and S1 spinal nerve laterally by an
extruded L7/S1 disc (arrow). The articular
process of L7 has been removed to expose
cartilage of the sacral facet (asterisk). The dorsal
root ganglion of L7 nerve root can be seen
10.44
within the foramen (arrowhead) although it is
covered partly by remnants of joint capsule.
More bone may need to be removed cranially to decompress the nerve root within the lateral recess as
the foramen is located some distance cranial to the disc space (1.27).
203
204
10.45 This dog had shown

intermittent lameness for
one year with no
neurological deficits.
A: After facetectomy,
screws were placed into the
L7 pedicle; this is easier
if a dorsal laminectomy
10.45
is done first so that both
lateral and medial walls of
A
B
the pedicle can be palpated.
B: The cement bridge is visible clearly between L7 and S1; a single screw has also been placed across
the other L7/S1 facet joint (10.5010.53). Six weeks after surgery it was much improved; at 11 months
it developed a draining tract from the wound but was otherwise normal.
Dorsal fixationfusion (10.4610.54)

Indications for fixationfusion include telescoping of the sacrum under L7 and marked instability (10.46). Facet subluxation must be reduced prior to fixation by either positioning the dog as in 10.18 or by using a laminectomy spreader
(10.49, 13.64). Initial reduction can also be obtained using bone holding forceps or towel clamps (10.47).
10.46 This dog initially underwent dorsal laminectomy

for lumbosacral pain (10.6A,B, 10.13).
Preoperative 3D reconstruction of CT scans
performed with the dog in A: flexion, and B:
extension revealed marked compression of the
cauda equina (10.6A,B, 10.11). The L7 articular
facets and ventral osteophytes on L7 are seen
in a constant position but the sacral lamina has
telescoped under the roof of L7 (arrowhead).
The disc is also bulging upwards and there is
severe foraminal narrowing (10.6). Postoperative
images are shown in 10.53.
10.46
A
Lumbosacral disease
10.47 A: Collapse of the dorsal lumbosacral articulation

with telescoping of the sacrum under the L7
lamina. B: The telescoping is reduced and the
distance between the towel clamps is increased
markedly as the L/S articulation is distracted.
Same dog as in 10.15.
10.47
A
205
206
10.48 The degree of

telescoping of the
sacrum relative to L7
increases from A to D
as shown in these 3D
reconstructions from
four different dogs (dog
A is normal). The black
line is drawn at the
cranial edge of each
sacral lamina. The
space between the
lamina of L7 and S1
decreases and the
facet overlap increases.
The best guide to the
degree of subluxation is
provided by opening
the joint capsule
(10.49).
10.48
A
Lumbosacral disease
10.49 A: Subluxation of facet joints is evident by the

amount of articular cartilage exposed on the
sacral facet (arrow). B: Reduction with
a laminectomy spreader anchored at the
laminectomy edges has brought the L7 facet
back to cover the cartilage of the sacral facet
(arrow). The cartilage on both sides of this joint
is removed with a small curette prior to screw
fixation. Bone debris or graft is then packed into
and around the joint (Bagley, 2003).
10.49
A
10.50 Transverse CT image of the LS articulation

from a dog after screw fixation to show the
approximate angles for screw insertion (13.57).
The ideal screw angle is between 30 and 45
from the sagittal plane (Bagley, 2003).
Compare these angles to the suboptimal
angles shown in 10.16. Ideally, screws should
not compromise the sacroiliac joint or cause
injury to the lumbosacral trunk of the sciatic
nerve (Ebraheim et al., 2000).
10.50
10.51 Screw entry points are shown (*) in a 3D CT

reconstruction from a dog that had undergone
previous dorsal laminectomy (10.13). The entry
point should be equal in distance from each
side and from the tip of the facet. The tip of the
10.51
L7 spinous process can also be used as a
guide to the angle of insertion (McKee et al.,
1990). Adequate reduction of the facets is assessed by observing joint congruity (10.49).
207
208
10.52 Screw placement. A: One screw is in position; the

pilot hole is being drilled for the second. A dorsal
laminectomy has already been performed. Ideally
the facet nearest the surgeon (L7) is over-drilled. If
the facet is too narrow, this step can be skipped
rather than risk facet fracture. B: The hole is then
countersunk. Washers can also be used (10.14,
10.15). C: The hole is measured so that the screw
will only just exit the sacrum (10.15); the hole is
then tapped. The tip of the L7 spinous process is
shown (arrowhead). D: Cancellous bone from the
wing of the ilium is packed over the screw; the L7
spinous process can also be used as a source of
corticocancellous graft (Stoll, 1996). Normally, the
laminectomy would be covered first with a fat graft
but it has been left exposed here for orientation.
10.52
A
Lumbosacral disease
10.53 This dog had a poor

response to dorsal
laminectomy done 1 year
previously (same dog as in
10.6A, B, 10.11, 10.13,
10.46). It had episodes of
severe pain; fluoroscopy
10.53
suggested marked lumB
bosacral instability. A: After A
fusion the dog was free of
pain for 4 months but then
developed degenerative myelopathy. B: Necropsy confirmed degenerative myelopathy; it also revealed
that the site had fused and was immobile despite lysis around one screw (arrow) and bending of the
other (arrowhead).
10.54 3D reconstruction of a CT
scan made after fusion using
bone screws and graft
(same dog as 10.53). Areas
of new bone production are
shown (white arrows). The
black arrowhead indicates
the position of the sacral
spinous process; above it is
the bent spinous process of
L7 (white arrowhead).
10.54
209
Clinical signs
Chapter
11
Ligamentous hypertrophy.
Joint capsule proliferation or cyst formation.
Osteophyte production.
Disc herniation.
Vertebral malformation and canal stenosis may be present at birth in the Doberman, suggesting either a congenital or inherited disorder (Burbidge, 2001; Drost
et al., 2002) (11.1).
An early onset of clinical signs is most common in
giant-breed dogs. Most other breeds show clinical signs
from middle age onwards. These usually develop due to
212
Diagnosis 212
Radiography 213
Presurgical evaluation 216
Hypothyroidism 217
Bleeding disorders 217
Treatment 217
Complications 224
Postoperative care 228
Prognosis
229

References
229
229
Procedures 232
Paramedian approach to the ventral neck 232
Vertebral distraction 235
Cement plug 237
Metal implant and bone cement method 239
Dorsal decompression 241
Laminoplasty 245
Cervical spondylomyelopathy (CSM or Wobbler syndrome) is predominantly a syndrome of large- and

giant-breed dogs, particularly Doberman pinschers and
Great Danes. The cause of this disorder is multifactorial. Important contributing factors are:
Vertebral canal malformation, stenosis, or both.
Vertebral instability.
11.1 CT images of C6 and C7 vertebrae from neonatal puppies

A: Doberman B: other large breed. There are significant
differences in vertebral dimensions between the two breed
groups: the cranial vertebral canal is relatively narrow in the
Doberman group and they have vertebral body asymmetry.
Changes were most marked for C7 but were also present at
both C5 and C6 (Burbidge, 2001).
212
acquired soft tissue or osseous lesions, which are probably a consequence of low-grade instability. Compression
is seen mainly at C5/6 and C6/7 in the Doberman but
lesions in other breeds often affect more cranial disc
spaces.
Both C5/6 and C6/7 sites are at high risk of causing
spinal cord compression; twenty per cent of dogs present with both C5/6 and C6/7 lesions at the time of
initial diagnosis. In addition, if either one of these
intervertebral spaces fuses then the other one seems
even more likely to cause compression (Bruecker et al.,
1989a). Any surgical procedure should therefore
address all high-risk spaces in an individual animal
(Dixon et al., 1996; Hilibrand et al., 1999).
CLINICAL SIGNS
The most common presentation is a gait disturbance,
which is most severe in the pelvic limbs and ranges from
mild ataxia to marked paresis and dysmetria. Cervical

hyperesthesia, guarding of the neck, pain on manipulation of the prominent transverse process of C6 (4.6), or
a low carriage of the head may also be seen (11.2).
Lameness and shoulder muscle atrophy in one or both
thoracic limbs, or pain when traction is applied to the
limb (root signature), suggest nerve root compression
(see Chapter 7, page 93). Neurological deficits localize
to either the C1C5 region or to the cranial portion of
C6T2. Dogs with C1C5 signs often show a floating
thoracic limb gait in addition to tetraparesis and dysmetria (Baum et al., 1992). Dogs with C6T2 signs have
more of a short-stepping thoracic limb gait. This causes
the thoracic and pelvic limbs to be advanced at different
rates and produces a characteristic disconnected gait.
The short thoracic limb steps may be due to either
increased thoracic limb muscle tone, which is an upper
motor neuron (UMN) effect on the elbow and carpal
extensor muscles, lower motor neuron (LMN) weakness, or both (Seim and Withrow, 1982). There is often
an associated LMN weakness of the elbow flexors resulting in a weak withdrawal reflex. Pelvic limb muscle tone
and reflexes are normal or exaggerated. Tetraplegia is
uncommon but when present (15.7) the dog must be
assessed for hypoventilation (6.1, 7.11, 11.10).
DIAGNOSIS
11.2 Doberman with CSM. Note the broad-based pelvic limb

stance and lowered head position.
Although the gait abnormalities are often suggestive of

CSM, careful evaluation of the history and thorough
physical and neurological examinations are important
to help rule out the differential diagnoses listed
in Table 11.1 (see also Box 7.2 and Chapter 14).
The clinician should also consider the possibility of a
Table 11.1 Differential diagnoses for cervical spondylomyelopathy (CSM)

Degenerative
disorders
Anomalous/developmental
disorders
Neoplastic
disorders
Inflammatory/infectious
disorders
Ischemic
disorders
Degenerative
myelopathy
Leukodystrophies
Synovial cyst(s)
Orthopedic
disease
Disc extrusion
Cervical
fibrotic stenosis
Congenital vertebral
malformations
Multiple cartilaginous exostoses
Tumoral calcinosis (calcinosis
circumscripta)
Meningocele/
meningomyelocele
Spinal dysraphism
Hydromyelia, syringomyelia,
or both
Pilonidal sinus (dermoid sinus)
Epidermoid cyst
Meningioma,
Nerve sheath
tumor
Other tumors
Discospondylitis
Epidural abscess
Meningomyelitis
Fibrocartilaginous
embolic
myelopathy (FCE)
Spinal cord
hematoma
Ischemic
neuromyopathy
Consider also thoracolumbar lesion; brainstem lesion.
thoracolumbar lesion in a dog with signs restricted to the

pelvic limbs (2.8, 4.30, 4.31). Brainstem lesions can also
mimic cervical spinal cord lesions on rare occasions
(2.23). A further diagnostic challenge is the potential for
over-diagnosis using an MRI without any reference to
the clinical signs. In humans a disc herniation and even
spinal cord compression appear in many asymptomatic
people older than 30 years (Teresi et al., 1987; Hayashi
et al., 1988; Gorman and Hodak, 1997). By the age of
60 years, some 12% of asymptomatic individuals show
compression of the spinal cord on MRI (Gorman and
Hodak, 1997). Incidental, mild disc lesions also occur in
the cervical region of large-breed dogs.
Prior to anesthesia, several potential complicating
conditions should be ruled out:
Chronic active hepatitis of Doberman pinschers
is usually evident on serum biochemistry.
Cardiomyopathy is seen in many large- and
giant-breed dogs, and is often fatal within 6 months
of diagnosis, particularly in male dogs. Even a
subtle arrhythmia should not be discounted (Calvert
et al., 1996). An electrocardiogram (ECG) and
echocardiogram should be performed, realizing that
this still cannot rule out occult cardiomyopathy.
A 24-h ECG may also be of value as a screening
test (Calvert and Wall, 2001).
Hypothyroidism.
Bleeding disorders (page 217).
Radiography
SURVEY RADIOGRAPHY
Survey radiographs are useful to rule out potential
differential diagnoses but are not definitive for CSM.
Severe articular facet changes or vertebral body malformation do raise the index of suspicion for CSM,
especially in giant-breed dogs (11.3).
flexion or extension. Lesions that did not improve were

termed static and those that did improve were termed
dynamic (Seim and Withrow, 1982). Dynamic lesions
can be subdivided further; first by whether or not they
respond to traction (traction-responsive or traction
non-responsive) and then by whether or not they
change during flexion and extension (positional). This
subdivision of lesion types helps the surgeon to decide
on the best procedure to perform. Although there may
be some overlap, lesions fall primarily into one of three
basic types (11.411.6).
Compressive lesions
that improve with traction are termed tractionresponsive (McKee and Sharp, 2003) (11.4). Traction
usually decreases spinal cord compression caused by
anulus fibrosus or ligamentous tissue and therefore
increases dural tube diameter (Rusbridge et al., 1998).
Such traction-responsive lesions would be expected to
benefit from a distraction-stabilization surgery. Such
lesions have also been called dynamic (Seim and
Withrow, 1982; Wheeler and Sharp, 1994).
Traction-responsive lesions
Some compressive lesions do not improve with traction (11.5).

Such lesions are usually caused by new bone formation
or extrusion of nucleus pulposus. They are most likely
to respond either to a ventral slot or dorsal decompression. Such lesions have also been called static (Seim
and Withrow, 1982; Wheeler and Sharp, 1994). They
are less common than traction-responsive lesions (Seim
and Withrow, 1982).
Differentiation between lesions that do or do not
respond to traction can be subjective at times, but this
is still the most logical means of deciding on the best
Traction non-responsive lesions
MYELOGRAPHY
This is the standard means of confirming a diagnosis
of CSM and has the advantage that the lesion can
be observed readily in different positions of the spine.
Contrast should be first concentrated at the lesion by
positioning the site of interest at the low point of the
spine for several minutes (McKee et al., 2000) (4.32).
Lateral, ventrodorsal (for cranial cervical vertebrae),
dorsoventral (for caudal cervical vertebrae), flexion
and traction views should then be taken (Rendano and
Smith, 1981; Lamb, 1995).
The lesion may alter appearance as the relative positions of adjacent vertebrae are changed. Formerly,
lesions were categorized based on whether or not compression changed in the stressed positions of traction,
11.3 Survey radiographs must be taken under general

anesthesia to be of diagnostic quality. This 6-year-old
tetraparetic Great Dane has multiple sites of periarticular new
bone around the facet joints, especially at C5/6 and C6/7. The
CT scan of this dog is shown in 11.7.
213
214
11.4 Traction-responsive type of

dynamic lesion. A: Ventral spinal cord
compression at C6/7 due to redundant
anulus fibrosus in a middle-aged
Doberman. B: Same dog but with
traction applied to the vertebral
column. There is marked relief of spinal
cord compression and widening of the
dural tube dorsally.
11.5 Traction non-responsive type of static lesion in a middle-aged Doberman (same dog as 11.30B). A: Ventral spinal cord
compression at C6/7 due to disc herniation. B: Same dog but with traction applied to the vertebral column. C: CT myelogram of the
same dog to show extruded, mineralized disc material at C6/7 (arrow).
type of surgical procedure (Rusbridge et al., 1998).

Some lesions will also have a positional component
and then a judgement must be made about which is
most significant.
Positional lesions Some dogs have minimal spinal

cord compression in a neutral neck position and therefore show little change with traction. Nevertheless
they may show significant compression in different
neck positions (11.6). In such dogs the degree of compression changes specifically as the neck is moved
between flexed, neutral and gently extended positions
(see below) (Dueland et al., 1973; McKee and Sharp,
2003). These types of dynamic lesion are best termed
positional as they are worsened by positions that
reflect normal neck motion. In dogs with positional
lesions the spinal cord is probably suffering repeated,
minor trauma during everyday life. Such dogs may
therefore benefit from a stabilization procedure to prevent this repeated, low-grade injury.
Positional studies are not without risk. The extension
view can cause severe exacerbation of spinal cord compression and should be undertaken with extreme care
and preferably only under fluoroscopy (11.9, 11.10).

If this is not available, the dog should be positioned in
mild (not extreme) extension for as short a time as possible. If the dog is not being ventilated, its ability to
breathe spontaneously must be checked carefully as
respiratory arrest can occur in this position (Seim and
Withrow, 1982). Use of the extension view should be
restricted to dogs with suspected positional lesions or
possibly to identify subclinical lesions. Advanced imaging techniques are preferable to diagnose subclinical
lesions when conventional myelography does not provide sufficient information (4.42, 11.711.10).
CT-MYELOGRAPHY
Ideally all conventional myelograms for CSM should be
followed by a CT scan (Sharp et al., 1992). The CT provides excellent bone imaging (11.7) and when used with
contrast it also gives a good transverse image of the spinal
cord (11.8, 11.10). This information can improve surgical planning and may provide prognostic information by
detecting spinal cord atrophy (Chambers and Betts,
1977; Sharp et al., 1995). Sagittal or three-dimensional
reconstructions may also prove useful (11.8B). Finally,
11.6 Dynamic, positional lesions in a

9-month-old Doberman. A: Neutral
view showing minimal spinal cord
compression. B: There is no change in
appearance with traction. C: In flexion
the craniodorsal aspects of both C4
and C5 vertebral bodies cause mild
attenuation of the ventral subarachnoid
space. D: There is moderate
attenuation of the dorsal subarachnoid
space over C3/4 and C4/5
intervertebral spaces in extension
(arrows) (Dueland et al., 1973).
11.7 CT scan at C5/6 from the dog

shown in 11.3. The conventional
myelogram was of poor quality.
A: Transverse CT image after
myelography reveals dramatic
overgrowth of the right articular facet
joint (arrowhead). B: 3D reconstruction
demonstrates the overgrown right facets
and also shows marked impingement of
both right and left facet joints into the
vertebral canal (arrows) (Massicotte
et al., 1999).
11.8 A: Transverse CT image at C5/6

in a 2-year-old Mastiff reveals marked,
asymmetrical, extradural compression
caused by a soft tissue mass adjacent to
the left articular facet joint (same dog as
in 11.56, 11.57). B: 3D reconstruction
suggests that this is cystic in nature
(arrow). This was suspected to be a
synovial cyst (Levitski et al., 1999;
Dickinson et al., 2001; Lipsitz et al.,
2001). See page 320.
215
216
11.9 Myelogram of a 5-year-old

Rottweiler with tetraparesis. The dog
had deteriorated after a ventral slot at
C6/7 1 week previously. A: The lateral
view shows no compression. B: The
ventrodorsal view shows moderate
extradural compression at C6/7 on
the right side. The radiopaque object
over the spinal cord is a pet
identification chip. CT was performed
with the dog first in a neutral and then
an extended position (11.10).
11.10 A: Same dog as in 11.9.

CT image with the neck in a neutral
position showing moderate attenuation
of the subarachnoid space over C6/7.
The central defect in the vertebral
body of C6 is due to the ventral slot.
B: CT image from the same dog with
the neck in an extended position.
Spinal cord compression is much
worse; this positional effect also led to
an unanticipated complication (see
below).
postoperative imaging can be used to gauge the effectiveness of a surgical procedure (4.25, 7.5, 7.8, 7.9,
11.10, 11.12, 11.56, 11.57).
CT can also be used to study positional lesions.
However, keeping the neck in an extended position for
the duration of the scan can cause severe injury (11.10).
The dog shown in 11.9 and 11.10 suffered seizures
during recovery, which were very difficult to control.
He was put on a ventilator but became profoundly
hypotensive and died 12 h after the study. The cause
of death was not determined but was probably complicated by sympathetic blockade secondary to
severe cervical spinal cord compression (see page 82)
(Rosenbluth and Meirowsky, 1953; Seim and Withrow,
1982; Clark, 1986). It is likely that maintaining the
neck in an extended position for the time that it took
to complete the second CT scan caused severe spinal
cord injury; this was probably exacerbated by the
seizures.
disadvantages are the lack of general availability and,

depending on the machine, the time required to complete a study. Dynamic studies are possible but do require
that the patient be repositioned and then re-imaged
(10.12, 11.10). However, dynamic studies are less
important if the cement plug distraction-stabilization
technique is to be used because this surgery can be
used for nearly all types of lesion (Algorithms 11.1A,B).
It is then only important for MRI to differentiate a disc
extrusion from other lesion types (4.42, 4.43).
A temporary worsening of neurological deficits may
occur after imaging dogs with CSM, especially after
using subarachnoid contrast. Imaging ideally should
be undertaken 48 h before elective surgery in order to
allow the patient to recover. This period also gives
ample opportunity to decide on treatment options, to
discuss these with the owner and to perform a thorough presurgical evaluation.
MR IMAGING
PRESURGICAL EVALUATION
This is the technique of choice for imaging humans

with degenerative diseases of the cervical spine. It is
non-invasive and also provides superior soft tissue resolution to CT-myelography (Lipsitz et al., 2001). The
During the recovery period following imaging, medical

problems that may complicate the situation should be
assessed. Post-myelographic seizures may be more
common in Doberman pinschers with CSM than in
other breeds (Lewis and Hosgood, 1992). This highlights the need for constant monitoring during recovery,
with diazepam on hand to control any seizures as they
could also exacerbate the spinal cord injury (Lipsitz
et al., 2001) (11.10).
Care should be taken to maintain adequate systemic
blood pressure and spinal cord perfusion throughout the
procedure (see page 86). Dogs with chronic spinal cord
lesions are particularly susceptible to postoperative
deterioration, especially following dorsal laminectomy
or ventral slot (Kohno et al., 1997; Rusbridge et al.,
1998; de Risio et al., 2002). Preoperative methylprednisolone sodium succinate (MPSS) may be useful but the
risks and benefits of this strategy remain unproven
(Olby, 1999; Pietila et al., 2000). Preoperative vitamin E
may be a useful alternative for elective surgeries (see
page 85).
Hypothyroidism
Doberman pinschers and Great Danes are predisposed
to hypothyroidism although this disorder is probably
over-diagnosed. Lethargy, muscle weakness and peripheral neuropathy may occur, all of which are undesirable
in a surgical candidate. The best diagnostic test is
thyroid-stimulating hormone (TSH) stimulation but
this is not widely available. In its absence a high endogenous TSH level with a low total T4 is highly suggestive
of hypothyroidism if the history and clinical signs are
compatible. Glucocorticoids, phenobarbitone and some
non-steroidal anti-inflammatory drugs (NSAIDs) may
interfere with thyroid function testing (Gieger et al.,
2000). Hypothyroid dogs should probably receive supplementation for at least 48 h prior to surgery although
hypothyroidism does not induce VW disease as was
once thought, nor does it cause defects in primary
hemostasis (Panciera and Johnson, 1996).
Bleeding disorders
It has been estimated that 16% of Doberman pinschers
in the USA have a bleeding tendency related to VW
disease (Dodds, 1989). Bleeding from the internal vertebral venous plexus is a potential problem during ventral decompression and can be almost impossible to
arrest unless the dog has normal hemostatic abilities.
After a ventral approach to the neck, the inability to
close dead space under the strap muscles can lead to
hematoma formation several days after surgery
(15.40). The easiest way to test an animals VW status is to perform a standardized bleeding time test
(11.11). A stopwatch is started just as the incisions
are made. Blood is blotted at 5-s intervals using filter
paper, taking care not to touch the incision itself. The
mean buccal mucosal bleeding time for normal dogs is
11.11 Illustration of a buccal mucosal bleeding time being

performed in a dog. The gauze strip has been used to keep the
lip turned over and cause slight venous engorgement. A twoblade, spring-loaded device (Simplate II, Organon Teknika,
General Diagnostics, Cambridge, UK) has been used to make
two 6-mm long by 1-mm deep incisions in the upper lip
mucosa. Any obvious blood vessels should be avoided.
2.62 min, with a range of 1.74.2 min (Jergens et al.,

1987). The cuticle bleeding time is harder to standardize than the buccal mucosa bleeding time, but should
identify severe bleeding disorders. A cut is made using
a guillotine nail clipper at the apex of the cuticle. In
normal dogs, bleeding stops within 8 min, but occasional normal animals will bleed for up to 12 min (Giles
et al., 1982; Stokol and Parry, 1998).
Dobermans with prolonged bleeding times, or those
with known VW disease, can be given desmopressin
(DDAVPRhone-Poulenc Rorer) (1.0 g/kg SQ)
immediately prior to surgery (Callan and Giger, 2002;
Kraus et al., 1989). Cryoprecipitate is the optimal
therapy, although fresh or frozen plasma (10 ml/kg of
plasma, taken 3060 min after the donor has been
given desmopressin) may also be useful (Kraus et al.,
1989; Ching et al., 1994; Stokol and Parry, 1998). As a
precaution, dogs at known risk can be cross-matched if
they are to undergo ventral decompression. There does
not appear to be any benefit in giving thyroid hormone
supplementation to euthyroid dogs with prolonged
bleeding times as this does not increase plasma VW factor antigen (Panciera and Johnson, 1994).
TREATMENT
The decision on the best way to treat each patient is
based on the presenting history, neurological status,
results of imaging, and on the owners expectations and
their ability to undertake any necessary aftercare. Most
dogs that show neurological deficits are surgical candidates, but consideration will also be given here to nonsurgical treatment.
217
218
Table 11.2 Reported results for surgical treatment of cervical spondylomyelopathy (CSM)
Ventral slot
Total
Metal and
cement
Cement
plug
Dorsal laminectomy
Total
Dogs (n)
182
273
147
59
411
225
188
146
214
53
Successes
(as %)
14
(78)
18
(67)
13
(91)
45
(76)
37
(90)
18
(82)
14
(78)
11
(79)
20
(93)
45
(85)
Follow-up (months)
(Mean)
1060
(29)
648
N/A
640
N/A
350
(20)
360
(21)
1.553
(17)
N/A
N/A
7108
(38)
No. of repeat episodes

(as %)
4/14
(28)
N/A
N/A
5/13
(38)
8/37
(22)
2/18
(11)
0/14
(0)
N/A
N/A
4/21
(19)
Months to repeat episode

(Mean)
1260
(32)
N/A
N/A
1633
(23)
542
N/A
833
N/A
N/A
N/A
N/A
N/A
448
(23)
9/27
(33)
4/35
(11)
Bruecker et al., 1989a; 2 Bruecker et al., 1989b; 3 Chambers et al., 1986; 4 de Risio et al., 2002; 5 Dixon et al., 1996; 6 Lipsitz et al., 2001;
Rusbridge et al., 1998; 8 Trotter et al., 1976.
N/A, not available.
7
Non-surgical treatment is warranted in two situations.
The first situation is when a normal dog develops neurological deficits following minor trauma; these may resolve
completely within a few weeks unless severe injury is sustained. The second is when a dog develops CSM before it
is skeletally mature and so may benefit from correction of
nutritional imbalances together with severe caloric reduction. This is analogous to the strategy used for management of young horses with Wobbler syndrome but it is
unproven in dogs (Donawick et al., 1993). For most other
dogs, however, surgery is the treatment of choice as the
majority probably show progressive deterioration without
treatment (Jeffery and McKee, 2001).
For most dogs with CSM the surgery is elective. A
24 week trial period of severe exercise reduction and
use of a chest harness is often justified (15.11); if this
fails it often emphasizes the need for surgery to the
owner. Anti-inflammatory doses of prednisolone may
also be used, but for short periods only and preferably
on an alternate day basis (VanGundy, 1988). One additional role for corticosteroids is to use the resultant
response, or lack of response, as a crude indicator of
the reversibility of any neurological deficit prior to performing surgery.
Surgical treatment
A large number of different surgical techniques have
been proposed for CSM, with many of the authors
claiming between 70 and 90% success rates (Table 11.2).
The most logical way to obtain the best overall results is

to consider three basic types of surgery and to perform
these for certain, relatively well-defined indications. The
three types of surgery are:
Ventral decompression.
Vertebral distraction-stabilization.
Dorsal decompression.
The basic indications for each procedure are summarized in Table 11.3. The main factor governing the
choice of surgical procedure is the appearance of the
spinal cord on imaging, particularly the traction and
flexion views after myelography. Some lesions show a
combination of different types of compression and then
a judgement must be made as to which is the major
component (see Myelography and MRI, pages 213
and 216). Algorithms 11.1A and B are algorithms for
surgical decision-making based on whether the dog has
single or multiple lesions. In general:
Dogs with single, ventral lesions that do
not respond to traction should undergo a
ventral slot.
Single and multiple lesions that respond well to
traction should undergo cement plug distractionstabilization. A ventral slot can be as good for single
lesions if performed well.
Single and multiple lesions that are positional
should undergo cement plug distractionstabilization. Some surgeons may prefer dorsal
decompression, particularly for dorsal positional
lesions or synovial cysts (page 320) (Jeffery, 1995;
Dickinson et al., 2001; de Risio et al., 2002).
Table 11.3 General indications for surgical procedures in cervical spondylomyelopathy (CSM)
Procedure
Indication(s) (see also Algorithms 11.1A,B)
Lesion(s) addressed
Ventral decompression
Ventral slot
Single, ventral, traction non-responsive (static) lesion*
Disc extrusion
Can be used for anulus
Ventral slot with metal and

bone cement
To prevent (or to treat) disc space collapse after

ventral slot decompression*
Disc extrusion and anulus
Distraction-stabilization
Cement plug
Single lesions of all types except ventral, traction

non-responsive lesion*
Multiple lesions of all types
Metal and bone cement
Single traction-responsive (dynamic) lesion*

Rescue after failed ventral slot decompression
Synthes locking plate
Single lesionsas for cement plug

Rescue after failed ventral slot decompression
Multiple lesionsneeds further evaluation
Dorsal decompression
Dorsal laminectomy
Laminoplasty
(needs further evaluation)
Single or multiple, dorsal, traction non-responsive

(static) lesion(s)
Single or multilpe, dorsal positional lesion(s)
Single traction non-responsive (static) lesion
Single dorsal, positional lesion
}
}
Osseous or soft tissue compression

Anulus fibrosus
Ligamentum flavum
Osseous vertebral canal stenosis, articular

facet osteophytosis, synovial cyst(s)
*In addition to the operated site, any additional subclinical lesion, in particular at C5/6 and C6/7 interspaces, should be addressed routinely in an
attempt to reduce the incidence of domino lesions (see 11.30 and page 221).
11.12 A: Preoperative CT myelogram at

C5/6 in a tetraparetic Doberman. There
is a soft tissue mass on the right, ventral
to the spinal cord. Radiopaque material
within the right foramen represents
mineralized disc or new bone. A ventral
slot retrieved nucleus pulposus and
anulus fibrosus; the slot was filled with fat
and then cancellous bone. B: Eight
weeks later the spinal cord and ventral
subarachnoid space have expanded.
New bone occupies the ventral portion of
the slot. Contrast material within the
spinal cord may represent a syrinx (Faiss
et al., 1990), (from Sharp et al., 1995).
In all dogs that show compression at only one

interspace, consideration should also be given to
a strategy that will prevent compression from
developing at adjacent high-risk spaces in the
future (see page 221).
VENTRAL DECOMPRESSION (11.2811.30)

(see Chapter 7, page 96)
Ventral decompressive surgery can be very challenging
in CSM because of various combinations of ventral
osteophytosis, misshapen vertebrae, and intraoperative

hemorrhage (Table 7.2). Short-term deterioration is
common even in dogs that have good long-term results
(Rusbridge et al., 1998). This technique is best reserved
for the relief of single-level spinal cord compression
caused by midline, dorsal extrusion of disc material. In
such lesions, a ventral slot should permit retrieval of a
large amount of nucleus pulposus or torn anulus fibrosus from the vertebral canal (11.5, 11.12).
Retrieval of extruded nucleus pulposus is not possible
in some dogs because the compressive lesion consists
219
220
instead solely of anulus fibrosus. In this situation it is

essential that a large portion of this anulus be excised so
that on completion, the dura is visible clearly across the
entire width of the slot (7.48, 11.29B). This may prove
to be very difficult, as the fibers of the anulus tend to
fragment before they can be removed and are often
attached to the internal venous plexus. Removal is facilitated by leaving a plug of anulus to serve as a handle
during removal (11.29A) and by using the inverted cone
technique (Goring et al., 1991) (7.50).
If the surgeon is in doubt about the adequacy of
decompression after a ventral slot then it should be
converted to a distraction-stabilization technique. This
is recommended in order to prevent intervertebral collapse with its attendant risk of anular buckling and
increased spinal cord compression (Chambers et al.,
1986) (4.25, 11.21). The worsening noted after surgery even in dogs that do well long term may suggest
that some degree of collapse is common after ventral
slot (Rusbridge et al., 1998). It is likely that the signs
then improve as the site stabilizes subsequently due to
the development of a fibrous or osseous union (Sharp
et al., 1995). Postoperative imaging can be used to
assess the decompressive effect of a ventral slot and is
certainly indicated if the neurological deficits are
markedly worse after surgery (11.10, 11.21). If a
distraction-stabilization procedure is needed following
a ventral slot, then the metal and bone cement technique
(11.13A, 11.4011.43) or a Synthes locking plate
(11.13B,C) are recommended (Wilson et al., 1994).
A full cement plug should not be used to distract the
site after a ventral slot as there is a high risk of cement
entering the vertebral canal through the defect created
in the dorsal anulus. A further potential problem is that
the cement also tends to collapse into the cancellous

bone with subsequent loss of distraction (Dixon et al.,
1996). A block of corticocancellous autograft (see page
292), small cement wedge (page 118), cancellous bone
allograft (Veterinary Transplant Services, Seattle, WA),
or a rope (11.31), can be used to distract a ventral slot
temporarily prior to fixation (page 117).
The major long-term disadvantage in using a ventral
slot for CSM is that about 30% of dogs undergoing single level decompression suffer a second episode of neurological signs within 23 years (Bruecker et al., 1989b;
Rusbridge et al., 1998) (Table 11.2). In most cases this
is presumed to be due to recurrence at the original site
or a domino lesion at an adjacent space. There are
probably several reasons for this:
Even disc spaces that appear normal often show
subclinical or histological abnormalities in dogs
with CSM, especially the high incidence spaces
C5/6 and C6/7 (1.3) (Seim and Withrow, 1982;
Rusbridge et al., 1998).
Such intervertebral discs may not respond to
fusion of an adjacent interspace in the same way
that a disc would in a normal dog (Cole et al.,
1987; Bruecker et al., 1989a; VanGundy, 1989;
Hilibrand et al., 1999). Signal changes on
MRI in adjacent discs have been detected
in humans within 12 months of fusion (Iseda
et al., 2001).
Fusion does occur at many disc spaces after a
single site ventral slot (11.12). This occurs mainly
after wide slots but the mobility at the slotted
space probably changes even for more narrow
slots (Gilpin, 1976; Chambers et al., 1982;
VanGundy, 1989).
11.13 A: A ventral slot has been performed at C5/6 intervertebral space. Bone screws have been placed in C5 and C6 ready for
bone cement to be applied once the space has been distracted. Distraction using a Gelpi retractor in adjacent disc spaces is no
longer recommended (Wilson et al., 1994) (see Vertebral distraction, page 235). B, C: Seven-month postoperative radiographs
of a Doberman treated using Syncage-C intervertebral implant and Cervical Spine Locking Plate (AO) at C6/7. Outcome at 1 year
was excellent (Matis, 2001).
Taken together, this would suggest that high incidence

disc spaces adjacent to the ventral slot site should
undergo some sort of prophylactic procedure in order
to reduce the incidence of domino lesions (see also
Vertebral decompression, page 234).
The choices of prophylactic procedure for high incidence disc spaces adjacent to a ventral slot are:
1. Place a cancellous bone graft at the adjacent
space(s), combined with forage of ventral
cortical bone (11.30), in order to promote
fusion (Dixon et al., 1996). The aim is to
improve stability and so reduce the likelihood
of this space producing a domino lesion in the
future.
2. Place a cement plug at the adjacent space(s).
However, this approach may increase collapse
of the slotted space. Even if this collapse does not
compress the spinal cord it may cause foraminal
narrowing and radiculopathy.
3. Do nothing at the adjacent space(s) but then
the risk of domino lesions will remain high
(7.14, 11.18, 11.23A).
Grafting and forage are recommended as they carry the
lowest risk of the three options, although at present
there are no data to support their efficacy (11.30).
The threat of a domino lesion is also present after a
single site ventral slot combined with metal and bone
cement fixation. The same three prophylactic options
also apply for this situation, but inducing osseous
fusion at the adjacent space will be more difficult due
to the mass of cement ventral to the slot (see Metal
and bone cement, page 239; 11.18, 11.43).
Fenestration is not a suitable treatment for any dog
with CSM (Lincoln and Pettit, 1985; Jeffery and McKee,
2001). It hastens intervertebral collapse and anular buckling, which cause spinal cord compression. Likewise, fenestration of discs to position a distraction instrument
is not recommended (Wilson et al., 1994) (11.13A).
Others suggest that concomitant forage and grafting may
be sufficient to stabilize an interspace that has been fenestrated for distraction purposes (M. McKee, personal
communication).
VERTEBRAL DISTRACTION-STABILIZATION
(11.3111.43)
The primary indications for distraction and stabilization are the presence of a traction-responsive lesion and
to relieve nerve root compression. Lesions may be
either single or multiple and cause either dorsal or ventral spinal cord compression (Algorithms 11.1A,B).
Distraction-stabilization has been attempted in the
past using a number of different techniques, but the
ones now recommended are the cement plug, Synthes
locking plate (11.13B), and metal implant and bone

cement techniques. The latter technique is limited
to single space distraction. The locking plate may be
suitable for multiple lesions although this needs to be
studied further in dogs. Cement plugs can be used
for nearly all types of lesion. An advantage of all distraction techniques is that they often provide rapid
relief of cervical hyperesthesia related to nerve root
decompression. Unlike the ventral slot, these distraction techniques do not involve entry into the vertebral
canal but this advantage is partly offset by the risk of
implant failure or other implant-associated complications. Future fixation devices will almost certainly be
made of absorbable materials (Vaccaro et al., 2002).
Metal implant and bone cement method

(11.4011.43) Metal implant and bone cement distraction is a well-tested technique with good long-term
follow up results (Table 11.2). It is also the standard
rescue technique for a failed ventral slot. The metal
implants can either be Steinmann pins (11.43B),
threaded pins (McKee and Sharp, 2003), or bone screws
(11.13A, 11.19, 11.4011.43). Attempts to bridge more
than one interspace usually result in implant failure
(Ellison et al., 1988; VanGundy, 1988) (11.17B). If two
spaces must be bridged with cement, implants should
be placed in all three vertebrae and the cement reinforced with a thick Steinmann pin (13.25). The main
disadvantage of single-level metal and bone cement is
the high rate of domino lesions. Subclinical lesions at
adjacent interspaces should therefore be addressed for
the reasons described under ventral decompression. This
is difficult because of the mass of cement (11.19), but
may be best-accomplished using forage (11.30), a
cement plug, or possibly a locking plate (11.13B).
Cement plug (11.3511.39) Cement plugs can be

applied to many types of lesion but the most logical indication is for traction-responsive lesions (page 213). These
implants retain the advantages of the earlier metal washer
technique but have overcome many of the disadvantages
(Wheeler and Sharp, 1994; Dixon et al., 1996; Rusbridge
et al., 1998; McKee et al., 1999). Catastrophic collapse
from end-plate fracture is unusual with the cement plug
(11.20A), presumably as forces are distributed more
evenly over the end plates than they were using washers
(McKee et al., 1999). The main advantage of cement
plugs is that they can be applied easily to more than one
intervertebral space. The surgeon can therefore be more
aggressive in dealing with a dog that has more than one
lesion, any one of which could go on to cause a domino
problem in the future if not treated (Dixon et al., 1996;
Rusbridge et al., 1998).
221
222
Cement plugs should be placed routinely at the highest risk disc spaces (usually C5/6 and C6/7), and
ideally to any other space that shows subclinical
lesions. An alternate approach is to implant cement plugs
at sites of spinal cord compression and fuse other sites
that are considered to be at risk by forage and grafting
(11.30). Obviously not every cervical intervertebral
space can be fused and the maximum is probably three
interspaces (Bolesta et al., 2000). It may also be difficult to decide exactly which sites to fuse in any given
dog. For the Great Dane and other giant breeds, C4/5
or even C3/4 may be involved along with C5/6 and
C6/7 (Olsson et al., 1982; Lewis, 1989; Lipsitz et al.,
2001). For the Doberman usually only C6/7 and C5/6
discs are at high risk although in some dogs C4/5
should probably also be addressed (11.14).
The aim of fusing multiple spaces routinely is to
reduce the incidence of domino lesions (Dixon et al.,
1996; Jeffery and McKee, 2001) (Table 11.2). Support
for this more aggressive approach also comes from
work in humans with cervical spondylosis. In humans,
the risk of new disease at an adjacent level was found to
be significantly lower following a multilevel arthrodesis
than it was following a single-level arthrodesis.
Therefore, it has been proposed in humans that all
degenerated segments causing radiculopathy or

myelopathy should be included in an anterior cervical
arthrodesis (Hilibrand et al., 1999). This is also consistent with reports suggesting that normal canine discs
can adjust to the fusion of adjacent segments, whereas
abnormal discs can not (Cole et al., 1987; Bruecker et al.,
1989a).
Cement plugs are also indicated for dogs with positional lesion(s) (page 214, Algorithms 11.1A,B). The
aim of stabilization in these dogs is to reduce or abolish
movement at the affected interspace. Stabilization may
even benefit dorsally located soft tissue or osseous
lesions by limiting movement locally. Fusion has been
shown to cause the regression of ligamentum flavum
lesions in dogs and of bony articular lesions in horses
(Grant et al., 1985; Seim, 1986; McKee et al., 1990).
Therefore cement plugs should be considered for dogs
with certain types of traction non-responsive lesions
such as those caused by osseous compression of the
dorsal vertebral canal (11.15). The advantage of using
cement plugs over dorsal laminectomy in these situations is that distraction-stabilization should avoid the
short-term morbidity associated with laminectomy and
will also allow the surgeon to address adjacent, subclinical lesions (de Risio et al., 2002).
11.14 A: Cement plugs have been

inserted in this Doberman at the C4/5,
C5/6 and C6/7 intervertebral spaces.
The degree of cement filling is good for
C4/5 and C5/6 but suboptimal for C6/7.
The dog wore an external splint for 8
weeks (11.39). B: Two-month follow-up
radiograph reveals good osseous fusion
at all three spaces. The presenting
complaint of neck pain and exercise
intolerance had resolved.
A
11.15 Young Great Dane with severe,

osseous stenosis from C3/4 to C6/7.
A: Preoperative CT myelogram from site
of maximum compression at C5/6.
Cement plugs were applied at C4/5 and
C5/6; C6/7 was only foraged and
grafted. A restraint device was applied
(13.18B). B: Two-year follow-up, again
from site of maximum compression at
C5/6 (arrow cement plug); there is
bony remodeling and less spinal cord
compression (arrowheads). C6/7 had not
fused; compression here was worse, as
was the dogs neurological status.
A
The main disadvantage of the cement-plug technique

is that it is unclear if an external splint must be used in
the postoperative period. Splints are not tolerated by all
dogs, they may require frequent modification or
replacement and can cause pressure sores (VanGundy,
1988) (11.39). An alternative is to use a Halti (13.18B) or
fixation screw or pin to prevent the cement from falling
out (McKee and Sharp, 2003), (11.16, 11.3611.38),
or to use a locking plate instead (11.13B).
Preliminary results with this technique are encouraging (11.13B,C). A swivel ring in the
plate hole means that screws may be inserted at any
angle within a range of 20 and the screws lock in the
plate via a unique locking mechanism. The Syncage is
designed to maintain distraction; it lies in the intervertebral space and is packed with cancellous bone. These
implants are extremely strong and can even bridge more
than one interspace (McLaughlin et al., 1997; Matis,
2001). Failure rates are lower in humans than for
implants that lack the screw-locking feature (Lowery
and McDonough, 1998). However, their utility for multiple lesions in dogs is not yet clear.
Locking plate
DORSAL DECOMPRESSION
This technique is an alternative for dogs with single dorsal lesions
that do not respond to traction as well as those with
multiple dorsal lesions (Dickinson et al., 2001; Lipsitz
et al., 2001) (11.8, 11.56, 11.57). This technique also
provides an option for dogs with ventral lesions at multiple intervertebral spaces (Lyman, 1991) (7.15).
Traction non-responsive
Ventral
lesion
Ventral slot*
Dorsal
lesion
Positional
Ventral
lesion
Cement plug(s)*,
locking plate,
laminoplasty or
dorsal
laminectomy
Long-term results of dorsal laminectomy appear to be

very good (Lipsitz et al., 2001; de Risio et al., 2002).
The major disadvantage is that there is significant,
short-term morbidity with deterioration in neurological
status, which can cause considerable nursing problems
in giant-breed dogs (Trotter et al., 1976; VanGundy,
1988; de Risio et al., 2002). An extensive soft-tissue
approach is also needed and ventrally located disc
material cannot be removed (VanGundy, 1988). Although
dorsal laminectomy should not cause domino lesions,
recurrence of clinical signs occurs in about 10% of dogs
and is reported to be due to restrictive fibrosis (Trotter
et al., 1976; de Risio et al., 2002) (Table 11.2). Because
of the disadvantages of laminectomy, and because
cement plugs also allow the surgeon to address multiple
lesions, including subclinical ones, the cement-plug
technique is preferred for most types of lesion
(Algorithms 11.1A,B). Dorsal laminectomy has given
good results for dogs with synovial cysts, with four of
six dogs being followed for more than 1 year (Dickinson
et al., 2001). It is not clear if these results will be maintained long term in a larger series of dogs (Dickinson
et al., 2001; Lipsitz et al., 2001).
When it is desirable to provide additional stability
following dorsal decompression, fusion can be encouraged by screwing and then bone grafting the facet joints
(11.56, 11.57). Potential risks include fracture of a
facet and trauma to nerves at the level of the foramen.
Stability of this fixation technique was reported to be
good when spines were assessed grossly at post mortem
2 weeks after surgery (Swaim, 1975). Long-term
results of this approach have not been reported.
Traction responsive
Ventral
lesion
Cement
plug(s)* or
locking
plate
Dorsal
lesion
Cement
plug(s)* or
locking
plate
Cement plug(s)*
or locking plate
*Consider a strategy to address all high-risk disc spaces
Algorithm 11.1A
a single lesion.
Surgical decision-making in dogs with
223
224
Traction non-responsive
Ventral
lesions
Dorsal
lesions
Positional
Ventral
lesions
Dorsal
lesions
Ventral
slot*
Traction responsive
Ventral
lesions
Dorsal
lesions
Cement
plugs
Cement
plugs
Algorithm 11.1B Surgical decision-making

in dogs with multiple lesions.
Cement plugs
or dorsal laminectomy
Laminoplasty (11.58)
This technique has been

reported just once and was then only used as a singlelevel procedure. A 5-year-old Great Dane with marked
ataxia and tetraparesis improved markedly 2 weeks
after surgery. It was normal neurologically 6 months
later but then died of a gastric torsion (McKee, 1988).
Laminoplasty is used widely in humans with severe
vertebral canal stenosis for both single- and multi-level
compression (Kohno et al., 1997; Shaffrey et al.,
1999). Its main advantage over laminectomy is that it
provides protection to the spinal cord from the recurrent compression that can arise from re-growth of a
fibrous or osseous lamina. This is often termed restrictive
fibrosis (Trotter et al., 1976; de Risio et al., 2002).
Laminoplasty may also induce less change in adjacent
discs compared to fusion (Iseda et al., 2001) and warrants further evaluation.
COMPLICATIONS
Some risks are either inherent to surgery on the cervical
spine or to a particular breed and as such should be
explained clearly to the owner. Many of the respiratory
and cardiovascular problems discussed in Chapter 7 also
occur in CSM (Boxes 7.3, 7.4; Table 11.4). These problems emphasize further the need for a thorough presurgical evaluation (Calvert et al., 1996). Complications
can be divided into three main categories (Table 11.4).
These are mainly due to either iatrogenic injury or
technical errors (Table 11.5). Overzealous retraction of
Table 11.4 Main sources of complications
Intraoperative
Iatrogenic injury
Technical errors
Early
postoperative
Late
postoperative
Implant failure
Postoperative
morbidity
Infection
Adjacent segment
disease
Adjacent segment
disease
Recurrence of signs
soft tissues during a ventral approach to the neck can

damage important nerves in the cervical region. This
can induce intraoperative arrhythmias or postoperative laryngeal paralysis (11.20B), Horners syndrome
(Boydell, 1995), or megaesophagus (15.40). Bleeding
from the venous plexus can be a major problem during
ventral decompression and may require a blood transfusion. Excessive retraction or improper patient
positioning can exacerbate the bleeding by compressing
the jugular veins (11.24). Hemorrhage can usually be
arrested by relieving pressure on the jugular veins in
combination with use of a piece of muscle to promote
coagulation (Table 7.2; 8.43). Direct pressure can also
be used (11.29B). Hematoma formation is most likely
to occur in the Doberman and can cause delayed damage to the vagus nerve or its branches (15.40) or even
spinal cord compression (Seim and Prata, 1982) (see
page 100).
Technical errors during surgery are the other main
source of complications in CSM. Examples include
improper selection of the implant (McKee et al., 1990)
(11.16) and poor implant positioning (11.17A, 11.37).
Table 11.5 Specific complications
Intraoperative
Early postoperative
Late postoperative
Iatrogenic neural injury (11.20B)

Wrong surgical site
Hemorrhage
Poor implant selection (11.16)
Poor implant position
(11.17A, 11.37)
Pneumomediastinum
Prolonged extension (11.18)
Vertebral fracture
Implant failure (11.19, 11.20A)

Morbidity after ventral slot (11.21)
Morbidity after dorsal laminectomy
End-plate failure (11.20A)
Self-induced trauma (11.22)
Ventral or epidural hematoma
(15.40)
Dorsal seroma
Ischemic injury
Adjacent segment disease
Discospondylitis
Epidural abscess (14.14)
Recumbency complications
Adjacent segment disease (11.23A)

Restrictive fibrosis
Recurrence of signs (11.8, 11.23B)
Discospondylitis (14.11)
Delayed compression from implant
11.16 A, B: A single midline screw has been used to keep each cement plug in place (11.3511.38). Unfortunately each screw
passed through a vascular foramen (1.18) to enter the vertebral canal and cause dramatic compression of the spinal cord, especially
over C5. C: The screws were removed and replaced by shorter ones. The dog walked without assistance the next day and had an
excellent outcome; the screw probably caused only mild, transient spinal cord compression.
11.17 A, B: These screws are all inserted too caudally in their respective vertebral bodies. Two have penetrated an end plate and
all four have the potential to damage the nerve roots or spinal nerves (Swaim, 1975). C: An excess of bone cement can compress
structures within the thoracic inlet to cause dysphagia or even esophageal perforation (Halligan and Hubschmann, 1993; Dixon
et al., 1996). Note that in this dog, the implants span two intervertebral spaces. If this is done, more implants and cement
reinforcement are needed (page 221) (13.25).
225
226
The surgeon may also cause iatrogenic spinal cord

damage (Read et al., 1983; Lipsitz et al., 2001), vertebral
fracture (McKee et al., 1989), induce pneumomediastinum (Marchevsky and Richardson, 1999), perform
surgery at the wrong site, or simply fail to understand
certain limitations of the technique itself (11.17C).
Prolonged or excessive extension of the neck during
surgery is undesirable as it causes spinal cord compression (11.10). In addition, it could lead to fusing
the vertebrae in extension if the position is not
corrected prior to cement hardening (Bruecker et al.,
1989b) (11.18).

The main complications at this stage are postoperative deterioration and implant failure (Table 11.5).
Refractory neck pain or a marked deterioration in neurological status are indications for repeat imaging and
possibly for a second surgery (Rusbridge et al., 1998;
McKee et al., 1999).
Implant failure can occur such as loss of distraction
after a variety of techniques (Bruecker et al., 1989a;
Wilson et al., 1994; Marchevsky and Richardson, 1999;
McKee et al., 1999; (11.1911.20A). Implant failure may
also be subclinical at times (11.19B, 11.20B, 11.37B).
11.18 A: There is a lesion at C5/6 and a
probable subclinical lesion at C4/5. B: The
C5/6 lesion was distracted manually and
stabilized using screws and bone cement.
Only minimal distraction is visible on the
postoperative myelogram; the caudal spine
was also fixed inadvertently in extension.
The dog did well after surgery and had no
deficits or neck pain 7 months later.
The thread for the cranial screw in C6
vertebra was stripped and it was replaced
with a 6.5-mm screw. This dog is at a
particularly high risk for a domino lesion at
C4/5 or C6/7.
11.19 A: Distraction at C6/7

intervertebral space using two small 2.7mm distraction screws (11.40). Each of
the four 4.5-mm screws only penetrates
one cortex (11.43). B: Loss of distraction
at 7 months; one of the distraction
screws has broken. In addition the 4.5mm screws have changed orientation
probably due to bone remodeling.
The cement has not failed and the dog
was doing well clinically although it had
an episode of neck pain 10 months
post-surgery.
11.20 A: Loss of distraction after a cement plug was used at C5/6 (11.38B). The dog showed sudden deterioration 4 weeks after
surgery. A small fracture is visible at the ventral portion of the plug (arrow) and there is failure of the cranial end plate of C6 (arrowhead)
with recurrent spinal cord compression. B: This dog did well until it fell down the stairs 6 weeks after distraction at C6/7. Failure of a
screw in C7 did not cause clinical signs (McKee et al., 1990). The dog had also shown a change in bark since surgery.
End-plate failure can occur after the cement-plug

technique (11.20A); potential factors include excessive motion, thermal necrosis and possibly weakening
from too large an anchor hole (Boker et al., 1989;
Martinez et al., 1997; Williams et al., 1997).
Inadequate removal of disc material during a ventral
slot decompression can increase spinal cord compression as the intervertebral space collapses (Chambers
et al., 1986; Ellison et al., 1988) (11.21). This complication was seen in up to 20% of dogs that failed to
respond after ventral slot (Chambers et al., 1986).
Even if it does not compress the spinal cord, collapse
can compress the nerve roots in the intervertebral
foramen, thereby increasing cervical hyperesthesia.
Collapse at an interspace can also be exacerbated by
the presence of an unsuspected synovial cyst (4.25).
Even when surgery goes well there can be a deterioration in neurological status after either dorsal laminectomy or ventral slot (Rusbridge et al., 1998; de Risio
et al., 2002). This deterioration increases a dogs
susceptibility to self-induced trauma, particularly as
it recovers from anesthesia or tries to stand up
(Bruecker et al., 1989b) (11.22).
Even without injury, there may be deterioration secondary to decreased stability at the operated site(s)
(Queen et al., 1998; Rusbridge et al., 1998; Macy et al.,

1999). Range of motion increases in cadaver spines by
3040% after fenestration and by 66% after a ventral
slot, even if the slot is only one third of the vertebral
width (Macy et al., 1999; Wolf and Roe, personal communication). Ventral slots with dimensions of greater
than 50% of the vertebral width produce instability in
small dogs and are likely to do the same in larger dogs
with CSM (Seim and Withrow, 1983; Fitch et al.,
2000; Lemarie et al., 2000). Dorsal laminectomy in the
lumbar region is known to reduce stability significantly
(see Chapter 8) but data are not available for the neck.
Postoperative deterioration could also be due to an
ischemic or reperfusion injury of the spinal cord
(Cybulski and DAngelo, 1988). Vascular injuries can
occur either during surgery or in the immediate postoperative period (see pages 86 and 130).
Early development of a second lesion at a new disc
space has been described at 2, 4 and 12 weeks after
surgery (Chambers et al., 1982; Wilson et al., 1994;
Rusbridge et al., 1998). One of these lesions developed
at a space that had been fenestrated in order to place
a Gelpi for distraction and in another the disc was
affected subclinically at the time of the first surgery
(Wilson et al., 1994; Rusbridge et al., 1998).
11.21 A: The preoperative myelogram

reveals ventral, extradural compression at
C6/7 that did not improve with traction.
B: The dog was worse the day after
ventral slot at C6/7; the margins of which
are visible clearly in C7 (arrowhead) and
to a lesser extent in C6. Compression
was now more severe than before
surgery, probably because of inadequate
removal of disc material and intervertebral
collapse (Chambers et al., 1986).
A
11.22 A: This dog made a good

recovery after ventral slot at C5/6 but
became severely tetraparetic 36 h later.
A small fracture is visible at the ventral
aspect of C5 (arrow), presumably due
to hyperflexion injury (Bruecker et al.,
1989b; McKee et al., 1989).
B: Distraction-stabilization were
performed using screws and bone
cement. The dog made a slow
recovery and was able to walk
4 months after surgery (15.11).
227
228
Discospondylitis can also occur on either an early

or delayed basis after surgery (Chambers et al., 1982).
It can sometimes be associated with an epidural
abscess (14.14).

The most important of these is recurrence of clinical
signs (Table 11.5; 11.8, 11.23B). Domino lesions or
adjacent segment disease (11.23A) are a particular
problem and result at least in part from abnormal
stresses imposed on one intervertebral space by fixation of an interspace adjacent to it (Fox et al., 1996;
Hilibrand et al., 1999). These stresses can exacerbate
any pre-existing subclinical instability and so produce
either disc extrusion or hypertrophy of anular or ligamentous structures (Bruecker et al., 1989a).
Recurrence of paraparesis or tetraparesis occurs in up

to one third of dogs after either ventral decompression
or metal implant and bone cement fixation (Table 11.2).
Recurrence can be caused by compression at the original
site or by a domino lesion at an adjacent site (Jeffery and
McKee, 2001). It usually occurs between 6 months
and 4 years after the original surgery, with a mean of
around 2 years (Seim, 1986; Bruecker et al., 1989a,b;
Rusbridge et al., 1998; McKee et al., 1999). Recurrence
seems to be less of a problem for the cement-plug technique, which is probably because all high-risk interspaces
can be stabilized at the same time (Dixon et al., 1996;
Hilibrand et al., 1999).
Recurrence after dorsal decompression is often due
to constrictive fibrosis at the surgical site, also termed
the laminectomy membrane (LaRocca and Macnab,
1974) (page 86). When the roof of the vertebral canal
is removed at laminectomy, the spinal cord is often displaced upwards when ventral extradural compression is
present (Lyman, 1991). However, the scar that forms
as the laminectomy heals often does so in the position
of the original lamina, which can then cause a recurrence of compression (Trotter et al., 1975, 1976;
Lyman, 1991; de Risio et al., 2002). A second surgery
may be successful when signs recur following a dorsal
laminectomy (de Risio et al., 2002) (11.23B). Overall
recurrence rates appear to be similar between dorsal
and ventral decompressive techniques (Jeffery and
McKee, 2001).
Spinal cord compression can also be caused by a relative change in the position of an implant secondary to
bony remodeling and collapse of a distracted interspace
(McKee et al., 1990). In addition, late-onset discospondylitis has been recorded 6 and 26 months after
a surgery and it can also occur as a complication of
an unrelated surgical procedure (Ellison et al., 1988;
Dixon et al., 1996) (14.11).
Postoperative care
B
11.23 A: This dog underwent a ventral slot decompression at
C6/7 1 year previously. It had recurrent tetraparesis due to a new
lesion at C5/6. This is the domino effect. B: CT myelogram
made at C5/6 in a Great Dane that underwent a dorsal
laminectomy 1 year previously. The dog improved but then
deteriorated 10 months later. Proliferation of new bone has almost
reformed the lamina of the vertebral canal resulting in severe spinal
cord compression with atrophy (A and B from Sharp et al., 1992).
Activity must be restricted severely in the immediate

postoperative period. Ataxic dogs should be confined
to a small kennel for 12 weeks to prevent them from
stumbling and hyperflexing their neck (11.22).
Activity can be increased gradually after this period. As
a minimum precaution, neck collar restraint and vigorous exercise should be avoided for 4 months. The first
68 weeks after surgery are the most crucial until significant osseous or fibrous healing has occurred.
Protective neck braces can be used for this period but
may be tolerated poorly (11.39, 13.18).
Dogs that remain recumbent after surgery require a
high level of nursing care (see Chapter 15). Pneumonia
is a particular risk for these dogs (Seim, 1986; Bruecker

et al., 1989b). Recumbent dogs are also prone to muscle
atrophy, decubitus, gastric volvulus (Trotter et al., 1976;
Mason, 1979; Seim, 1986; McKee, 1988) seroma formation at the surgical site (15.34), urinary tract infection
(UTI) (Rusbridge et al., 1998), and urine scald (15.35)
(VanGundy, 1989). Some of these factors can be complicated further by hypothyroidism (VanGundy, 1989;
de Risio et al., 2002). Additional postoperative problems
in CSM include hematoma formation due to VW disease
(Rusbridge et al., 1998) (15.40), diarrhea and gastrointestinal hemorrhage, especially after excessive corticosteroid use (VanGundy, 1988), pancreatitis (Read et al.,
1983; VanGundy, 1988); excoriation of digits (Trotter
et al., 1976) (15.7, 15.10), discospondylitis or epidural
abscess (Chambers et al., 1982; Dixon et al., 1996) (14.11,
14.14), sepsis (Black, 1986) (14.14), and multiple
abscessation (Seim, 1986).
PROGNOSIS
The seriousness of this condition is best illustrated by
the fact that a quarter of dogs with CSM in one series
were euthanized within 6 weeks of surgery as a result
of neurological problems (Seim, 1986). Overall longterm mortality rates for CSM vary from 19 to 43%
(Seim et al., 1986; Dixon et al., 1996; Rusbridge et al.,
1998; McKee et al., 1999; de Risio et al., 2002). Dogs
with more than one lesion generally have a worse prognosis than dogs with single lesions, and dogs with
chronic tetraparesis have a guarded prognosis (Seim,
1986). Most severely tetraparetic dogs that are going to
recover will do so within 6 weeks. In some dogs surgery
will only halt the progression of disease (VanGundy,
1988). This is presumably because there is often significant loss of neural tissue at the site of the lesion (Read
et al., 1983; Sharp et al., 1995). It is therefore likely
that the outcome will be better if surgery is done earlier
in the disease process (Chambers et al., 1986; VanGundy,
1988; McKee et al., 1990).
The general estimate provided by Seim is still useful
to predict the likely outcome after surgery for
CSM. For dogs with single lesions, about 80% of those
that are walking prior to surgery will have a favorable
outcome (Seim, 1986). Success rates quoted in the
literature vary from 70 to 90% but these figures do
not include dogs that undergo euthanasia later for
reasons related to CSM, usually due to recurrence of
signs (Chambers et al., 1986; de Risio et al., 2002)
(Table 11.2). Any apparent differences in success rates
between the various techniques are not statistically significant (Jeffery and McKee, 2001). There are conflicting results from the limited studies conducted to date
on whether dogs that are unable to walk before surgery

have a poor prognosis; the opposite may even be true
(Seim, 1986; Ellison et al., 1988; Bruecker et al.,
1989a; Dixon et al., 1996; Jeffery and McKee, 2001;
de Risio et al., 2002). Clearly there is a need to reduce
the overall mortality rate for dogs with this condition
and in particular to reduce the high risk of recurrent
spinal cord compression.
1. Does grafting an interspace that is judged to have
subclinical disease increase its stability and thereby result
in a decreased incidence of domino lesions (11.30B)?
2. Does forage of the ventral cortex of adjacent vertebrae
enhance the degree of fusion at a site grafted with
cancellous bone (11.30A)?
3. Does the routine fusion of C5/6 and C6/7 interspaces
decrease the incidence of domino lesions or will this
approach just serve to transfer the problem along to C4/5
or C7/T1? If this approach works, how does the surgeon
decide how many spaces to fuse (Bolesta et al., 2000)
(11.14)?
4. Is there any objective means to decide whether an
interspace is at risk of a domino lesion (Mitchell et al.,
2001)? Will MRI be able to identify such an interspace
and does this technique have any advantage over use of
an extension view under myelography?
5. If the aim is to induce fusion between adjacent vertebrae,
at what stage should this be considered successful
(Fox et al., 1996; Gibson et al., 2002)? Does the
presence of a radiolucent line in the bridging spondyle
mean that there is insufficient stability?
REFERENCES
Anderson, G.I. (1988) Polymethylmethacrylate: a review of the implications and complications of its use in orthopaedic surgery. Veterinary and
Bagley, R.S., Silver, G.M., Connors, R.L., Harrington, M.L., Cambridge,
A.J., Moore, M.P. (2000) Exogenous spinal trauma: surgical therapy and
aftercare. Compendium on Continuing Education for the Practicing
Veterinarian 22, 218230.
Baum, F., de Lahunta, A., Trotter, E.J. (1992) Cervical fibrotic stenosis in a
young Rottweiler. Journal of the American Veterinary Medical Association
201, 12221224.
Black, A.P. (1986) Results of arthrodesis as a treatment for caudal cervical
spondylo-myelopathy in nine dogs. Australian Veterinary Practitioner
16, 147148.
Boker, D., Schultheiss, R., Probst, E. (1989) Radiologic long-term results
after cervical vertebral interbody fusion with polymethyl methacrylat
(PMMA). Neurosurgical Review 12, 217221.
Bolesta, M.J., Rechtine, G.R., 2nd, Chrin, A.M. (2000) Three- and fourlevel anterior cervical discectomy and fusion with plate fixation:
a prospective study. Spine 25, 20402044; discussion 20452046.
Boydell, P. (1995) Horners syndrome following cervical spinal surgery in
the dog. Journal of Small Animal Practice 36, 510512.
229
230
Bruecker, K.A., Seim, H.B., Blass, C.E. (1989a) Caudal cervical spondylomyelopathy: decompression by linear traction and stabilization with
Steinmann pins and polymethyl methacrylate. Journal of the American
Bruecker, K.A., Seim, H.B., Withrow, S.J. (1989b) Clinical evaluation of
three surgical methods for treatment of caudal cervical spondylomyelopathy of dogs. Veterinary Surgery 18, 197203.
Burbidge, H.M. (2001) Caudal Cervical Vertebral Malformation in
the Dobermann Pinscher, 121135. PhD thesis. Massey University,
New Zealand.
Callan, M.B., Giger, U. (2002) Effect of desmopressin acetate administration on primary hemostasis in Doberman Pinschers with type-1 von
Willebrand disease as assessed by a point-of-care instrument. American
Calvert, C.A., Wall, M. (2001) Results of ambulatory electrocardiography
in overtly healthy Doberman Pinschers with equivocal echocardiographic evidence of dilated cardiomyopathy. Journal of the American
Calvert, C.A., Jacobs, G.J., Pickus, C.W. (1996) Unfavorable influence of
anesthesia and surgery on Doberman Pinschers with occult cardiomyopathy. Journal of the American Animal Hospital Association 32, 5762.
Cechner, P.E. (1980) Ventral cervical disc fenestration in the dog: a modified technique. Journal of the American Animal Hospital Association
16, 647650.
Chambers, J.N., Betts, C.W. (1977) Caudal cervical spondylopathy in the
dog: a review of 20 clinical cases and the literature. Journal of the
Chambers, J.N., Oliver, J.E., Jr, Kornegay, J.N., Malnati, G.A. (1982)
Ventral decompression for caudal cervical disk herniation in large- and
giant-breed dogs. Journal of the American Veterinary Medical Association
180, 410414.
Chambers, J.N., Oliver, J.E., Jr, Bjorling, D.E. (1986) Update on ventral
decompression for caudal cervical disk herniation in Doberman Pinschers.
Ching, Y., Meyers, K.M., Brassard, J.A., Wardrop, K.J. (1994) Effect of
cryoprecipitate and plasma on von Willebrand factor multimeters and
bleeding time in Doberman Pinchers with type-1 von Willebrands disease. American Journal of Veterinary Research 55, 102110.
Clark, D.M. (1986) An analysis of intraoperative and early postoperative mortality associated with cervical spinal decompressive surgery in the dog. Journal of the American Animal Hospital Association 22,
739744.
Cole, T., Ghosh, P., Hannan, N., Taylor, T., Bellenger, C. (1987) The
response of the canine intervertebral disc to immobilization produced by
spinal arthrodesis is dependent on constitutional factors. Journal of
Orthopedic Research 5, 337347.
Cybulski, G.R., DAngelo, C.M. (1988) Neurological deterioration after
role of spinal cord ischaemia. Journal of Neurology, Neurosurgery and
De Risio, L., Munana, K., Murray, M., Olby, N., Sharp, N.J., Cuddon, P.
(2002) Dorsal laminectomy for caudal cervical spondylomyelopathy:
postoperative recovery and long-term follow-up in 20 dogs. Veterinary
Surgery 31, 418427.
Dickinson, P.J., Sturges, B.K., Berry, W.L., Vernau, K.M., Koblik, P.D.,
LeCouteur, R.A. (2001) Extradural spinal synovial cysts in nine dogs.
Dixon, B.C., Tomlinson, J.L., Kraus, K.H. (1996) Modified distractionstabilization technique using an interbody polymethyl methacrylate plug
in dogs with caudal cervical spondylomyelopathy. Journal of the American
Dodds, W.J. (1989) Acquired von Willibrands disease in dogs. American
Animal Hospital Association 1989; 614619.
Donawick, W.J., Mayhew, I.G., Galligan, D.T., Green, S.L., Stanley, E.K.,
Osborne, J. (1993) Results of a low-protein, low-energy diet and confinement on young horses with wobbles. Proceedings of the Thirty Ninth
Annual Convention of the American Association of Equine Practitioners,
Texas, USA, December 58, 1993, 125127.
Drost, W.T., Lehenbauer, T.W., Reeves, J. (2002) Mensuration of cervical
vertebral ratios in Doberman pinschers and Great Danes. Veterinary
Dueland, R., Furneaux, R.W., Kaye, M.M. (1973) Spinal fusion and dorsal
laminectomy for midcervical spondylolishthesis in a dog. Journal of the
Ellison, G.W., Seim, H.B., Clemmons, R.M. (1988) Distracted cervical
spinal fusion for management of caudal cervical spondylomyelopathy in
large-breed dogs. Journal of the American Veterinary Medical Association
193, 447453.
Faiss, J.H., Schroth, G., Grodd, W., Koenig, E., Will, B., Thron, A. (1990)
Central spinal cord lesions in stenosis of the cervical canal. Neuroradiology
32, 117123.
Fitch, R.B., Kerwin, S.C., Hosgood, G. (2000) Caudal cervical intervertebral disk disease in the small dog: role of distraction and stabilization in
ventral slot decompression. Journal of the American Animal Hospital
Fox, M., Onofrio, B., Hanssen, A. (1996) Clinical outcomes and radiological instability following decompressive lumbar laminectomy for
degenerative spinal stenosis: a comparison of patients undergoing concomitant arthrodesis versus decompression alone. Journal of Neurosurgery
85, 793802.
Gibson, J.N.A., Waddell, G., Grant, I.C. (2002) Surgery for degenerative
lumbar spondylosis (Cochrane Review). Issue 1 ed: The Cochrane
Library. http://www.update-software.com/cochrane/
Gieger, T.L., Hosgood, G., Taboada, J., Wolfsheimer, K.J., Mueller, P.B.
(2000) Thyroid function and serum hepatic enzyme activity in dogs after
phenobarbital administration. Journal of Veterinary Internal Medicine
14, 277281.
Giles, A.R., Tinlin, S., Greenwood, R. (1982) A canine model of hemophilic (factor VIII:C deficiency) bleeding. Blood 60, 727730.
Gilpin, G.N. (1976) Evaluation of three techniques of ventral decompression of the cervical spinal cord in the dog. Journal of the American
Grant, B., Hoskinson, J.J., Barbee, D.D., Gavin, P.R., Sande, R.D.,
Bayly, W.M. (1985) Ventral stabilization for decompression of caudal
cervical spinal cord compression in the horse. Proceedings of 31st Annual
Convention of the American Association of Equine Practitioners 1985;
75103.
Halligan, M., Hubschmann, O. (1993) Short-term and long-term failures
of anterior polymethylmethacrylate construct with esophageal perforation. Spine 18, 759761.
Hayashi, H., Okada, K., Hashimoto, J., Tada, K., Ueno, R. (1988) Cervical
spondylotic myelopathy in the aged patient. A radiographic evaluation of
the aging changes in the cervical spine and etiologic factors of myelopathy. Spine 13, 618625.
Hilibrand, A., Carlson, G., Palumbo, M., Jones, P., Bohlman, H. (1999)
Radiculopathy and myelopathy at segments adjacent to the site of a previous anterior cervical arthrodesis. The Journal of Bone and Joint
SurgeryAmerican volume 81, 519528.
Iseda, T., Goya, T., Nakano, S., Kodama, T., Moriyama, T., Wakisaka, S.
(2001) Serial changes in signal intensities of the adjacent discs on T2weighted sagittal images after surgical treatment of cervical spondylosis:
anterior interbody fusion versus expansive laminoplasty. Acta
Neurochirurgica (Wein) 143, 707710.
Jeffery, N.D. (1995) The wobbler syndrome. In: N.D. Jeffery (ed.),
Handbook of Small Animal Spinal Surgery, 169186. London: WB
Saunders.
Jeffery, N.D., McKee, W.M. (2001) Surgery for disc-associated wobbler
syndrome in the dogan examination of the controversy. Journal of
Jergens, A.E., Turrentine, M.A., Kraus, K.H., Johnson, G.S. (1987) Buccal
mucosa bleeding times of healthy dogs and of dogs in various pathologic
states, including thrombocytopenia, uremia, and von Willebrands disease. American Journal of Veterinary Research 48, 13371342.
Kohno, K., Kumon, Y., Oka, Y., Matsui, S., Ohue, S., Sakaki, S. (1997)
Evaluation of prognostic factors following expansive laminoplasty for
cervical spinal stenotic myelopathy. Surgical Neurology 48, 237245.
Krag, M.H. (1991) Biomechanics of thoracolumbar spinal fixation. A
review. Spine 16 (3 Suppl), S8499.
Kraus, K.H., Turrentine, M.A., Jergens, A.E., Johnson, G.S. (1989) Effect
of desmopressin acetate on bleeding times and plasma von Willebrand
factor in Doberman Pinscher dogs with von Willebrands disease.
Lamb, C.R. (1995) The dorsoventral cervical myelogram. Veterinary
LaRocca, H., Macnab, I. (1974) The laminectomy membrane. Studies in
its evolution, characteristics, effects and prophylaxis in dogs. The Journal
of Bone and Joint SurgeryBritish volume 56B, 545550.
Lemarie, R.J., Kerwin, S.C., Partington, B.P., Hosgood, G. (2000)
Vertebral subluxation following ventral cervical decompression in the
Levitski, R., Chauvet, A., Lipsitz, D. (1999) Cervical myelopathy associated with extradural synovial cysts in 4 dogs. Journal of Veterinary
Lewis, D.D., Hosgood, G. (1992) Complications associated with the use
of iohexol for myelography of the cervical vertebral column in dogs: 66
cases (19881990). Journal of the American Veterinary Medical Association 200, 13811384.
Lewis, D.G. (1989) Cervical spondylomyelopathy (wobbler syndrome)
in the dog: a study based on 224 cases. Journal of Small Animal Practice
30, 657665.
Lincoln, J.D., Pettit, G.D. (1985) Evaluation of fenestration for treatment
of degenerative disc disease in the caudal cervical region of large dogs.
Lipsitz, D., Levitski, R.E., Chauvet, A.E., Berry, W.L. (2001) Magnetic resonance imaging features of cervical stenotic myelopathy in 21 dogs.
Lowery, G.L., McDonough, R.F. (1998) The significance of hardware failure in anterior cervical plate fixation. Patients with 2- to 7-year followup. Spine 23, 181186; discussion 186187.
Lyman, R. (1991) Wobbler syndrome. Continuous dorsal laminectomy is
the procedure of choice. Progress in Veterinary Neurology 2, 143146.
Macy, N.B., Les, C.M., Stover, S.M., Kass, P.H. (1999) Effect of disk fenestration on sagittal kinematics of the canine C5C6 intervertebral
space. Veterinary Surgery 28, 171179.
Marchevsky, A.M., Richardson, J.L. (1999) Disc extrusion in a Rottweiler
dog with caudal cervical spondylomyelopathy after failure of intervertebral distraction/stabilisation. Australian Veterinary Journal 77,
295297.
Martinez, S.A., Arnoczky, S.P., Flo, G.L., Brinker, W.O. (1997) Dissipation
of heat during polymerization of acrylics used for external skeletal fixator connecting bars. Veterinary Surgery 26, 290294.
Mason, T.A. (1979) Cervical vertebral instability (wobbler syndrome) in
the dog. Veterinary Record 104, 142145.
Massicotte, C., Jones, J., Newman, S., Moon, M. (1999) Wobbler syndrome
due to cervical stenosis in a Great Dane puppy. Canine Practice 24, 1821.
Matis, U. (2001) AO spinal implants for canine wobbler syndrome.
McKee, M., Sharp, N. (2003) Cervical spondylopathy. In: D. Slatter (ed.),
Small Animal Surgery, 3rd edn, 11801192. Philadelphia: WB Saunders.
McKee, W., Penderis, J., Dennis, R. (2000) Radiology corner: obstruction
of contrast medium flow during cervical myelography. Veterinary
McKee, W.M. (1988) Dorsal laminar elevation as a treatment for cervical
vertebral canal stenosis in the dog. Journal of Small Animal Practice 29,
95103.
McKee, W.M., Lavelle, R.B., Mason, T.A. (1989) Vertebral stabilisation for
cervical spondylopathy using a screw and washer technique. Journal of
McKee, W.M., Lavelle, R.B., Richardson, J.L., Mason, T.A. (1990)
Vertebral distraction-fusion for cervical spondylopathy using a screw
and double washer technique. Journal of Small Animal Practice 31,
2126.
McKee, W.M., Butterworth, S.J., Scott, H.W. (1999) Management of
cervical spondylopathy-associated intervertebral disc protrusions using
metal washers in 78 dogs. Journal of Small Animal Practice 40, 465472.
McLaughlin, M.R., Purighalla, V., Pizzi, F.J. (1997) Cost advantages of twolevel anterior cervical fusion with rigid internal fixation for radiculopathy
and degenerative disease. Surgical Neurology 48, 560565.
Mitchell, R.A.S., Innes, J.F., McNally, D. (2001) Pressure profilometry of
the lumbosacral disk in dogs. American Journal of Veterinary Research
62, 17341739.
Olsson, S.E., Stavenborn, M., Hoppe, F. (1982) Dynamic compression of
the cervical spinal cord. A myelographic and pathologic investigation in
Great Dane dogs. Acta Veterinaria Scandinavica 23, 6578.
Panciera, D.L., Johnson, G.S. (1994) Plasma von Willebrand factor antigen
concentration in dogs with hypothyroidism. Journal of the American
Panciera, D.L., Johnson, G.S. (1996) Plasma von Willebrand factor antigen
concentration and buccal mucosal bleeding time in dogs with experimental hypothyroidism. Journal of Veterinary Internal Medicine 10, 6064.
Pietila, T.A., Stendel, R., Schilling, A., Krznaric, I., Brock, M. (2000)
Surgical treatment of spinal hemangioblastomas. Acta Neurochirurgica
(Wein) 142, 879886.
Queen, J.P., Coughlan, A.R., May, C., Bennett, D., Penderis, J. (1998)
Management of disc-associated wobbler syndrome with a partial slot
fenestration and position screw technique. Journal of Small Animal
Read, R.A., Robins, G.M., Carlisle, C.H. (1983) Caudal cervical spondylomyelopathy (wobbler syndrome) in the dog: a review of thirty cases.
Rendano, V.T., Jr, Smith, L.L. (1981) Cervical vertebral malformationmalarticulation (wobbler syndrome)the value of the ventrodorsal view
in defining lateral spinal cord compression in the dog. Journal of the
Roosen, K., Grote, W., Liesegang, J., Linke, V. (1978) Epidural temperature changes during anterior cervical interbody fusion with polymethylmethacrylate. Advances in Neurosurgery 5, 373375.
Rosenbluth, P.R., Meirowsky, A.M. (1953) Sympathetic blockade, an acute
cervical cord syndrome. Journal of Neurosurgery 10, 107112.
Rusbridge, C., Wheeler, S.J., Torrington, A.M., Pead, M.J., Carmichael, S.
(1998) Comparison of two surgical techniques for the management of
cervical spondylomyelopathy in Dobermanns. Journal of Small Animal
Seim, H.B. (1986) Caudocervical spondylomyelopathy. 14th Annual
Veterinary Surgical Forum 1986; 7278.
Seim, H.B., Withrow, S.J. (1982) Pathophysiology and diagnosis of caudal
cervical spondylo-myelopathy with emphasis on the Doberman Pinscher.
Seim, H.B., Withrow, S.J. (1983) Ventral decompression for the treatment
of herniated cervical intervertebral disk in the dog. In: M.J. Bojrab (ed.),
Current Techniques in Small Animal Surgery, 2nd edn, 544548.
Shaffrey, C., Wiggins, G., Piccirilli, C., Young, J., Lovell, L. (1999)
Modified open-door laminoplasty for treatment of neurological deficits
in younger patients with congenital spinal stenosis: analysis of clinical
and radiographic data. Journal of Neurosurgery 90 (4 Suppl), 170177.
Sharp, N.J.H., Wheeler, S.J., Cofone, M. (1992) Radiological evaluation of
wobbler syndromecaudal cervical spondylomyelopathy. Journal of
Sharp, N.J.H., Cofone, M., Robertson, I.D., DeCarlo, A., Smith, G.K.,
Thrall, D.E. (1995) Computed tomography in the evaluation of caudal
cervical spondylomyelopathy of the Doberman Pinscher. Veterinary
Stokol, T., Parry, B.W. (1998) Efficacy of fresh-frozen plasma and cryoprecipitate in dogs with von Willebrands disease or hemophilia A. Journal
of Veterinary Internal Medicine 12, 8492.
Swaim, S.F. (1975) Evaluation of four techniques of cervical spinal fixation
in dogs. Journal of the American Veterinary Medical Association 166,
10801086.
Teresi, L.M., Lufkin, R.B., Reicher, M.A., Moffit, B.J., Vinuela, F.V.,
Wilson, G.M., Bentson, J.R., Hanafee, W.N. (1987) Asymptomatic
231
232
degenerative disk disease and spondylosis of the cervical spine: MR

imaging. Radiology 164, 8388.
Trotter, E.J., Brasmer, T.H., de Lahunta, A. (1975) Modified deep dorsal
laminectomy in the dog. Cornell Veterinarian 65, 402427.
Trotter, E.J., de Lahunta, A., Geary, J.C., Brasmer, T.H. (1976) Caudal cervical vertebral malformation-malarticulation in Great Danes and
Doberman Pinschers. Journal of the American Veterinary Medical
Vaccaro, A.R., Venger, B.H., Kelleher, P.M., Singh, K., Carrino, J.A.,
Albert, T., Hilibrand, A. (2002) Use of a bioabsorbable anterior cervical
plate in the treatment of cervical degenerative and traumatic disk disruption. Orthopedics 25 (10 Suppl), s11911199; discussion s1199.
VanGundy, T. (1989) Canine wobbler syndrome. Part II. Treatment.
Compendium on Continuing Education for the Practicing Veterinarian 11,
269284.
VanGundy, T.E. (1988) Disc-associated wobbler syndrome in the
Doberman Pinscher. Veterinary Clinics of North America, Small Animal
Wheeler, S.J., Sharp, N.J.H. (1994) Caudal cervical spondylomyelopathy.

In: S.J. Wheeler and N.J.H. Sharp (eds), Small Animal Spinal
DisordersDiagnosis and Surgery, 135155. London: Mosby-Wolfe.
Williams, N., Tomlinson, J.L., Hahn, A.W., Constantinescu, G.M., WagnerMann, C. (1997) Heat conduction of fixator pins with polymethylmethacrylate external fixation. Veterinary and Comparative Orthopaedics
Wilson, E.R., Aron, D.N., Roberts, R.E. (1994) Observation of a secondary
compressive lesion after treatment of caudal cervical spondylomyelopathy in a dog. Journal of the American Veterinary Medical Association
205, 12971299.
Zindrick, M.R., Wiltse, L.L., Widell, E.H., Thomas, J.C., Holland, W.R.,
Field, B.T., Spencer, C.W. (1986) A biomechanical study of intrapeduncular screw fixation in the lumbosacral spine. Clinical Orthopaedics and
Related Research 203, 99112.
PROCEDURES
Paramedian approach to the ventral neck
Excessive extension should be avoided during positioning as it tends to close the dorsal part of the disc space and
also increases spinal cord compression. Rather than position the neck in extension to improve access to the caudal vertebrae, the dog is positioned with the neck in a relatively neutral position and with a weight tied to the rostral
maxilla. The mild traction also tends to open the disc space, keep the anulus under tension and reduce spinal cord
compression (Goring et al., 1991) (11.24). Exposure to the caudal vertebrae may also be improved using the lateral
muscle separating approach and the strap muscles then serve to protect the trachea, recurrent laryngeal nerve,
carotid sheath and jugular vein (11.2511.27). Regardless of the approach used it is important to take great care
during dissection near the thoracic inlet and to protect all exposed tissues with damp sponges (Cechner, 1980).
11.24
11.24 A: Patient positioning for a ventral

decompression or distraction-stabilization
using traction. The sandbag shown under the
dogs head can be removed to produce mild
cervical extension if needed. B: Harvesting of
cancellous bone from a proximal humerus
(inset). Both shoulders should be included in
the operative field. Bone graft can be used to
promote fusion following a ventral slot,
distraction-stabilization or after vertebral forage
at an adjacent site (11.30).
A
11.25 Initially the standard, midline, ventral approach

to the neck is followed as described in
Chapter 7 (7.177.21). Once the strap muscles
have been exposed the surgeon can use a more
lateral approach. The right sternothyroideus
muscle is first separated from the right
sternocephalicus muscle (Cechner, 1980).
11.25
11.26 An Army Navy retractor is used to retract all

structures ventral to the longus colli muscles
away from the surgeon except for the right
sternocephalicus, which is pulled towards the
surgeon (Cechner, 1980).
11.26
11.27 The dissection is continued through the loose

cervical fascia until the longus colli muscles are
exposed. Balfour self-retaining retractors are
used to keep structures away from the midline
while the dissection is continued to expose the
vertebrae of interest. The esophagus is visible
at the caudal end of the dissection (arrow).
11.27
Ventral decompression
The ventral slot technique can be challenging in CSM because of ventral osteophytosis, malformed vertebral
bodies, intraoperative hemorrhage, friable anulus fibrosus, or a combination of these (11.28, 11.29). Surgery at
C6/7 may be especially difficult but access can be improved by taking particular care with patient positioning, by
use of traction (11.24) and possibly by using a paramedian approach (11.2511.27). Decompression should be
performed as described under 7.377.51. If the slot is too narrow the decompression will be inadequate and the
outcome will be more like a fenestration (Chambers et al., 1986; Ellison et al., 1988). However, a slot that
approaches 50% of the vertebral body carries a high risk of subluxation (Fitch et al., 2000; Lemarie et al., 2000).
In general the slot should be about one third of the vertebral width unless it is shaped like an inverted cone
(Goring et al., 1991; Lemarie et al., 2000) (7.50).
233
234
Osseous fusion is common following a ventral slot (Swaim, 1975; Gilpin, 1976; Sharp et al., 1995) (11.12). This
may be beneficial in preventing motion at the operated site but probably causes increased stress at adjacent spaces
(Fox et al., 1996; Hilibrand et al., 1999). Normal disc spaces can tolerate this stress but subclinically affected spaces
suffer a high incidence of domino lesions (Bruecker et al., 1989a; Hilibrand et al., 1999) (Table 11.2). A technique is
therefore needed to prevent domino lesions after single site ventral slot surgery. One technique that might accomplish this is to try to promote fusion at subclinically affected spaces by forage and grafting (11.30).
11.28 3D reconstruction of a CT scan made the day

after a ventral slot was performed at C5/6
intervertebral space (same dog as in 4.25).
Suction and excellent lighting, such as a
fiberoptic source (5.3), are essential; access
may be improved by modifying the slot into an
inverted cone (7.50) (Goring et al., 1991).
11.28
11.29 A: Removal of anulus can be especially

difficult in CSM. It can be aided by preserving
a small knub of anulus during drilling (arrow).
This can then be grasped like a handle to
facilitate its removal after a window has been
cut into the vertebral canal (Jeffery, 1995)
(7.457.49). B: Hemorrhage from the
vertebral venous plexus can be hard to
control (7.44; Table 7.2); occlusion of the
vertebral venous plexus is by pressure or a
small piece of muscle tissue (8.43). The most
likely puncture sites for the venous plexus are
at each corner. The width of the slot in this
illustration is much greater than generally
recommended unless some type of
stabilization is to be applied afterwards
(11.4111.43).
11.29
A
11.30 A: After a ventral slot at C6/7

(not shown) cortical bone was
removed over the ventral
aspect of C5 and C6 vertebrae
(arrowheads) without damaging
the C5/6 disc (arrow).
Cancellous bone was then
11.30
applied over both disc spaces
A
B
(Dixon et al., 1996; Hilibrand
et al., 1999). B: A ventral slot at C6/7 with forage of the ventral cortical bone on either side of C5/6 (different
dog to A, see 11.5). Both sites show good ventral bridging with new bone at 6 weeks; this progressed to
complete fusion by 4 months. The dogs neurological deficits had resolved completely by 6 months.
Vertebral distraction (11.3111.34)

The temporary distraction device can be custom-made as shown in 11.33A but a Freer or Synthes periosteal elevator is just as effective. Use of this initial distraction technique allows one side of the space to be curetted; the distractor is then repositioned to curette the other side. It is especially important to remove as much material as
possible from the craniodorsal portion of the disc space, although this process is harder to perform properly at C6/7
than at other spaces (11.14). The dorsal anulus must be left intact to prevent cement from entering the vertebral
canal. If the anulus is torn, it is possible to protect the dura prior to introducing cement using Gelfoam (Pharmacia,
Kalamazoo, MI) as this has good thermo-insulating properties (Roosen et al., 1978; Boker et al., 1989).
Distraction can be done using Gelpi retractors in the vertebral bodies as shown in 11.38A (Dixon et al., 1996)
but ideally not by placing a retractor in the adjacent disc spaces (Bruecker et al., 1989a; Dixon et al., 1996; Wilson
et al., 1994) (11.13A). Distraction can also be accomplished by an assistant pulling on a rope around the dogs
upper canine teeth (7.55) but is unreliable and often produces inadequate distraction (11.18). Finally, small K-wires
may be used to maintain distraction like the screws in 11.40, but are then removed once the cement plug hardens.
a
11.31 A: Sagittal section through the cervical spine

to show compression ventrally from anulus (a)
and dorsally from ligamentum flavum (b). B:
Distraction relieves compression caused by
redundant tissues. Gelpi retractors maintain
distraction; one tip is against the caudal end
plate of the cranial vertebra and the other in a
small hole drilled in the body of the caudal
vertebra (11.34). A rope pulling on the dogs
maxilla (11.24A), and tied to a hook on the
surgical table, with or without the Gelpi
distraction shown here, are the most reliable
techniques to permit cement hardening.
11.31
Ventral
Dorsal
A
Ventral
Dorsal
B
235
236
11.32
11.32 A: Distraction has been started by making a wide

fenestration at the C6/7 disc. The window in the
ventral anulus fibrosus (a) and the nucleus pulposus
(b) are visible. B: The C6/7 disc is curetted
thoroughly following fenestration. This is easier with
the space distracted as shown in 11.33B.
11.33 A: A distraction device is introduced into the

intervertebral space to aid curettage. B: The device
is then turned through 90 to distract the two
vertebrae and widen the intervertebral space.
When all nucleus and end-plate cartilage have
been removed (Dixon et al., 1996), a rope or Gelpi
retractor is used to maintain distraction as shown in
11.31B and 11.34.
11.33
A
11.34 A: A channel is drilled with a fine bur just under

the cortical bone on the ventral surface of the
vertebra. This hole starts about two thirds the
distance along the vertebra caudal to the disc
space to be distracted. One end of the Gelpi is
hooked into this hole; the other is set against
the caudal end plate of the vertebra in front of
the disc space to be distracted (11.31B). The
space is distracted as shown in 11.33B and the
Gelpi is locked. B: The Gelpi retractor is locked.
The initial distraction device is now removed
and the space checked again to insure that no
nuclear material remains. The space is now
ready to be filled with cement. Care must be
taken not to dislodge the Gelpi once it is in
place; sudden collapse of the distracted space
could injure the spinal cord (Dixon et al., 1996).
11.34
A
Cement plug (11.3511.39)
11.35 Cement is mixed until the powder is dissolved

and then it is transferred quickly to a 60-ml
catheter-tipped syringe. Attached to the end of
the syringe is an 8 French red rubber tube with
the end cut short so that its tip just fits into the
disc space. Cement is injected into the space
taking care to fill the dorsal portion first without
trapping air under the cement. The Gelpi is then
left in place for 1012 min.
11.35
The handle of a small instrument is used to compress the cement on either side of the Gelpi tip; this tip must be
kept clear of cement to facilitate its removal after cement hardens.
Ideally the two or three high-risk spaces are distracted using this procedure as shown in 11.16 and 11.37.
Alternatively, any subclinically affected spaces can be foraged and then grafted as shown in 11.30.
If no hole has been drilled into the end plate to anchor the cement plug, then a single 4.5-mm screw can be
used to prevent the plug from falling out. A single screw is inserted by drilling and tapping the first cortex only; the
screw then taps its own thread (11.36) as tapping may weaken fixation in cancellous bone (Krag, 1991; Bagley et
al., 2000). The second (inner) cortex should prevent the screw from entering the vertebral canal (see 11.16 for
potential complication). The advantages of using this technique to anchor the plug is that the vertebral end plates
should be less likely to collapse as they are not weakened (11.20A, 11.38) and there is also no need for the dog
to wear a splint after surgery. Some collapse of the space does still occur, presumably from thermal damage to the
end plates (Anderson, 1988; Boker et al., 1989), and to pressure-induced resorption. This technique has worked
237
238
well in the small number of dogs studied provided that the screw is positioned so the head sits over the cement
plug (11.36, 11.37). A similar approach uses threaded pins to anchor the plug instead (McKee and Sharp, 2003).
The original method for anchoring a cement plug is to drill holes in each end plate and to fill these with cement
(Dixon et al., 1996) (11.38). Having an angled bur guard for the drill (5.24) makes it easier to drill the anchor hole
into the cranial end plate.
It is not clear if an external splint provides additional benefit following the standard cement-plug technique.
Although dogs that were not splinted had outcomes as good as those that were, most surgeons who use this
technique recommend use of a splint (Dixon et al., 1996). The dog should be assessed closely for the appearance of rub sores, discomfort, odor or any type of malaise that might signify infection under the splint. If there is
any doubt it should be replaced. A Halti and chest harness are an easier way to restrict movement (13.18B).
11.36 A: Cement has hardened

(arrow). A screw, then
cancellous bone is
applied over the distracted
space. B: Good new bone
production is evident at
C6/7 5 weeks after surgery
11.36
with nearly complete
bridging callus. Note the
B
A
radiolucent area ventral to
the cement (arrow), which also occurs with the standard technique (Dixon et al., 1996). The implant
survived this male St Bernard jumping a fence to reach a bitch in heat and falling on the far side 2 weeks
after surgery. The dog was normal at 4 months.
11.37 A: These screw heads do

not cover the cement plug
adequately (11.16, 11.36).
B: Follow-up radiograph
12 weeks after surgery
reveals that the plugs have
11.37
fallen out (Marchevsky and
Richardson, 1999). Despite A
this the dog did well clinically
and its neurological deficits had resolved by 4 months.
11.38 A: Distraction can be

accomplished using a Gelpi
retractor positioned within
each vertebral body as
shown here (Dixon et al.,
1996). B: Excellent filling of
the C5/6 disc space;
cement can be seen clearly
11.38
where it has entered the
caudal anchor hole
A
B
(arrow). Despite this the
plug failed due to end-plate collapse 4 weeks after surgery (11.20A) and the dog was euthanized.
It had not worn an external splint although this is not a prerequisite for a successful outcome (Dixon
et al., 1996).
11.39 A splint can be made in one piece from

fiberglass; it is easiest to apply under general
anesthesia but can also usually be done under
heavy sedation. Whatever material is used has
to be well padded and the entire splint should
be replaced every 2 weeks, or sooner should
problems arise. A Halti headcollar tied to a
chest harness is a simple and effective way to
limit neck movement (13.18B).
11.39
Metal implant and bone cement method (11.4011.43)

Initial vertebral distraction for the metal and bone cement technique cannot be performed using the method
shown in 11.31 and 11.34 because of the final position of the cement. The disc space is first distracted as shown
in 11.33B or 11.40. Then a bone allograft (Veterinary Transplant Services, Seattle, WA) or an autograft from either
the wing of C6, the sternum or the ilium can be wedged into the slot to maintain distraction prior to cement application (Fitch et al., 2000). Use of such a graft has the additional advantage that it will promote an osseous union.
Alternatively, distraction can be maintained by a rope around the maxillary teeth (11.24) that is then tied to the
surgical table.
11.40 A: Another distraction

option is to use two,
2.7-mm distraction screws.
With the space distracted
(11.33) these are directed
from the ventral vertebral
body to cross the disc
space and rest against the
end plate of the caudal
vertebra (Queen et al.,
11.40
1998). B: The distraction is
asymmetrical as one of the
B
A
distraction screws was
aimed too laterally; this is exacerbated by the curved shape of the end plate (11.19).
The anchoring pins or screws should enter the bone close to the midline (11.41) and are then driven towards
the contralateral articular facets, away from the vertebral canal. Two implants are placed in the vertebrae on each
side of the affected space. The vertebrae are then distracted and this position is maintained by molding bone
cement carefully around each one of the screws.
Even when a ventral slot is not needed to produce decompression, a shallow slot is useful in order to pack with
bone graft and so promote fusion across the interspace. Gelfoam (Pharmacia, Kalamazoo, MI) can be used to
insulate the graft as the cement hardens (Roosen et al., 1978; Boker et al., 1989). If pins or screws are to be combined with a ventral slot it is important to position the screws precisely as there is only a limited amount of bone
available. There is more room for implants in the cranial vertebra if they are inserted a little more parallel to the end
plates than shown in 11.43 (Jeffery and McKee, 2001). The screws have better purchase if they penetrate two
cortices but there is less risk of penetrating the vertebral canal or damaging nerve roots if they do not (Swaim,
1975; Zindrick et al., 1986) (11.40, 11.43). If only one cortex is penetrated then the implant will be strengthened
considerably by use of three screws in each vertebra (Garcia et al., 1994).
239
240
11.41 A: Placement of pins or

screws in cervical vertebra
for distraction-stabilization
(13.56A). B: The diverging
path of the screw in each
vertebra is shown clearly in
this dorsoventral radiograph.
This dog had also
undergone a ventral slot
decompression; the cranial
portion of the slot is visible
(arrowhead).
11.41
B
11.42 Sagittal section shows how the ventral slot should

be started almost entirely in the vertebra cranial to
the disc. Owing to the angle of the end plates, and
the position of the slot, there is often less room for
screws in the cranial vertebra than in the caudal
vertebra (11.1711.19, 11.22, 11.4011.43).
11.43 A: Postoperative radiograph of screws and bone

cement in position after a ventral slot.
A hole was only drilled and tapped for the first
(ventral) cortex. Screw position is good but still
there was only room for one 4.5-mm and one
3.5-mm screw in C6 due to the angle used and
the proximity of the cranial edge of the slot
(arrow). B: Pins can also be used as the metal
implant. Positive profile pins are preferred to the
smooth pins shown here. Fully threaded screws
are also preferred to the partially threaded ones
shown in 11.22B as the thread junction can act
as a stress riser.
11.42
11.43
A
Dorsal decompression (11.4411.55)

Dorsal decompression may be limited to two vertebrae or it can extend over multiple vertebrae (Lyman, 1991). It
is advisable to preserve the facet joints in dogs with CSM and so the limit of bone removal is usually the medial
aspect of the facet joint(s) (Trotter et al., 1976; VanGundy, 1988). After the initial approach (11.4411.47), the nuchal
ligament is divided in the midline (11.48). The approach can be alongside the ligament rather than dividing it and it can
even be sectioned if additional exposure is required. It can be repaired later but this is not necessary.
There is a layer of thick fascia and ligamentum flavum between adjacent vertebrae. It can be removed using sharp
dissection but this must be done with great care as the spinal cord is not protected by bone between the vertebral
arches. It is probably safer to remove this layer of fascia after entering the vertebral canal on either side.
Screws are preferred over wire for providing additional stability after dorsal laminectomy (Swaim, 1975; Trotter
et al., 1976) (11.56, 11.57). Cartilage is debrided from the joint surfaces and bone debris, or ideally cancellous
bone graft, is packed around the joints. The main risks with the screw fixation technique are facet fracture or nerve
root damage. The latter could have severe consequences if major nerves to the brachial plexus are injured.
11.44 The dog is positioned with its

limbs pulled cranially and its
caudal neck supported by a
sandbag. The site of incision
when approaching the caudal
cervical vertebrae is from
C3 to T2.
11.44
11.45 Skin and superficial fascia incised and retracted

to reveal the aponeurosis of the rhomboideus
and trapezius muscles.
11.45
11.46 Diagram to show the position of

muscles relative to underlying
skeletal structures. The muscles
are divided in the midline. Bleeding
from the large neurovascular
bundles that penetrate the fascia
must be controlled to prevent
postoperative hematoma (15.34).
The nuchal ligament can be
palpated at this stage.
11.46
241
242
11.47 The fascia overlying the nuchal ligament is

divided and cleared. The ligament is freed back
to the prominent spinous process of T1. Here
the nuchal ligament has been mobilized,
revealing the paired spinalis muscles (arrows).
11.47
11.48 A: The nuchal ligament

attaches to the spinous
process of T1 and then
continues caudally with
the interspinous
ligament. The spinous
process of C7 can
usually be palpated at
this stage. The relative
size of C7 and T1
11.48
should be ascertained
from lateral radiographs
A
as they may vary between
animals. B: Nuchal ligament being divided.
11.49 Closer view showing spinalis muscles attached

to spinous process of C7 (arrow).
11.49
11.50 The muscles are elevated from the spinous

process (a) and lamina of the vertebra.
The spinous process is then removed with
bone cutters.
a
11.50
11.51 The spinal cord is now visible (a). Large veins may
be present in the epidural space (11.52). Bleeding
may respond to bipolar cautery but a Hemoclip
(Pilling Weck Inc., Research Triangle Park, NC) or
fine suture is needed if a vein is torn.
a
11.51
11.52 CT scans of a tetraplegic

Doberman after IV contrast.
A: Image at the unaffected
C6/7 disc space shows how
the vertebral venous plexus
communicates with
intervertebral veins at the
level of the disc space
11.52
(arrows). These veins may
extend into the dorsal epidural A
B
space; the exact pattern is
variable. B: Image through C5/6 disc space to show the spinal cord compressed by a lateralized,
non-mineralized disc extrusion (arrowhead) that was confirmed at surgery (from Sharp et al., 1995).
11.53 The laminectomy is performed with a bur. Close

attention is paid to the layers of bone as they are
removed (8.32, 10.29, 10.30).
11.53
243
244
11.54 A: Spinal cord exposed (see also 7.59). The

laminectomy can be continued laterally to
remove proliferated articular processes and
synovial tissue (11.57). B: The laminectomy can
be continued to one side to expose the nerve
root more fully. However, this almost certainly
decreases stability.
11.54
A
11.55 A: Myelogram of a 6-year-old Rottweiler with marked

tetraparesis of 12 months duration. There is a
dilation of the ventral and, to an even greater extent,
dorsal subarachnoid space. B: Intraoperative
photograph after dorsal laminectomy to show the
exposed spinal cord and dilation of the subarachnoid
space (arrow). Marsupialization of the arachnoid cyst
is in progress (14.514.8).
11.55
A
11.56 Two-year-old mastiff with

severe osseous compression
at C2/3 (11.57A) that had a
dorsal laminectomy and
4.5-mm lag screws inserted
across the facet joints. The
11.56
dog recovered well but
developed mild left-sided
A
hemiparesis 3 months later.
This follow-up myelogram shows good decompression at C2/3
(11.57B) compared to the preoperative myelogram (not shown) but a
left-sided synovial cyst at C5/6 (11.8). The owner declined further
surgery; the dog improved and was doing well 3 years later.
11.57 3D reconstructions of the

(A) preoperative, and (B)
3-month postoperative CT
myelograms at C2/3; same
dog as shown in 11.8 and
11.56. Undercutting was
used to remove compressive
periarticular bone on the
medial aspects of the facet
11.57
joints; this has widened the
A
vertebral canal significantly
(10.43) (Trotter et al., 1976).
Laminoplasty
Laminoplasty can be performed using a variety of techniques (McKee, 1988; Kohno et al., 1997; Shaffrey et al.,
1999). In the method illustrated in 11.58, a fine-tipped bur is used to make a cut three quarters of the way along
the midline of the C6 dorsal lamina. Great care is needed to avoid the spinal cord, especially cranially. The cut is
extended into a T-shape caudally. Grooves are cut down to the inner cortical bone at the junction of the lamina
with each pedicle. The laminar flap is elevated sufficiently to make a small bur hole in its medial edge for 0
polypropylene suture. The inner cortical bone is weakened further by drilling until each flap can be elevated
almost vertically. The interarcuate ligament and any dural adhesions must be separated with care. 3.5-mm
screws are placed in lag fashion across the C5/6 articular facets in order to attach and tighten the sutures. A cancellous bone graft is packed around the joints and a fat graft then placed over the spinal cord (McKee, 1988).
245
246
11.58 Laminoplasty of the cervical vertebrae as

performed in a 5-year-old Great Dane with
stenosis at C5/6 (McKee, 1988). Many different
types of laminoplasty are used in humans
(Shaffery et al., 1999) (from McKee, 1988).
11.58
Neoplasia
Clinical signs
Chapter
12
247
References
Tumor biology 248

Extradural tumors 248
Intradural/extramedullary
Intramedullary 250
Procedures 266
Hemilaminectomy 266
Vertebrectomy 273
Nerve sheath tumor resection
249
Diagnosis 250
Radiography 250
Ultrasound 250
Electrophysiology 250
Myelography 251
Computed tomography 251
Magnetic resonance imaging 251
Biopsy 251
Staging
263
275
Although tumors are uncommon causes of spinal disease, they are significant once the more common problems such as disc disease (in dogs) and trauma are
eliminated. Older animals are usually affected, although
certain tumor types do occur in young individuals.
252
CLINICAL SIGNS
Treatment 254
Medical treatment 254
Radiation treatment 254
Complications
258
Postoperative care
259
Prognosis 259
Extradural tumors 260
Intradural/extramedullary tumors
Intramedullary tumors 261
Feline spinal tumors
Diagnosis 262
Treatment 263
Prognosis 263
260
261
263
Animals with vertebral column tumors present with

a fairly typical pattern of initial non-specific discomfort,
followed by development of progressive neurological
deficits and more definitive evidence of spinal pain.
Severe pain and significant neurological dysfunction
develop eventually if the condition is allowed to progress.
Marked weight loss or localized muscle atrophy is often
seen. One exception to this pattern is a precipitous
decline following initial mild or even imperceptible signs.
This can occur with vertebral body tumors where
pathological fracture develops (12.12), or in soft tissue
tumors where the animal tolerates a slowly progressing
degree of spinal cord compression until it decompensates suddenly. Vascular interference by the tumor may
also be responsible for acute deterioration.
Most tumors of the spinal cord cause little or no
discomfort. Although incontinence usually develops
only after an animal has lost all motor function (see
Chapter 2), some animals with intramedullary tumors
develop urinary or fecal incontinence while they are
still able to walk. This can occur even for lesions affecting
248
T3L3 segments (Prata, 1977; Jeffery and Phillips,

1995) (12.4).
Tumors involving the brachial or lumbosacral plexus
may present initially as a slowly progressive unilateral
lameness (McDonnell et al., 2001). This can present a
diagnostic challenge and may lead to extended efforts
to find a non-neurological abnormality; even including
surgery for a suspected orthopedic lesion (McCarthy
et al., 1993). Rectal examination may identify a mass
within the pelvic canal (such as an enlarged lumbosacral
trunk, the intrapelvic portion of the sciatic nerve, 1.8A)
(Niles et al., 2001). Animals with brachial plexus
tumors may have an axillary mass or show pain on
palpation of the area (Carmichael and Griffiths, 1981;
Brehm et al., 1995). The miosis of partial Horners
syndrome (Bradley et al., 1982) (2.15A), unilateral
loss of the cutaneous trunci reflex with a consensual
response (Carmichael and Griffiths, 1981) (2.11, 2.12),
or diaphragmatic hemiparalysis (13.12) may be other
important clues to brachial plexus neoplasia (Bradley
et al., 1982; Wheeler et al., 1986; Bailey, 1990). Signs
tend to be insidious in onset, with spinal cord dysfunction occurring only if the vertebral canal is invaded.
Even then half of all dogs with nerve sheath tumors
and abnormal myelograms show no neurological
deficits in the pelvic limbs (Bradley et al., 1982).
The lower motor neuron (LMN) deficits make electromyography a useful test, first for confirming that the
animal has neurological disease and then in identifying
which nerve roots are involved (see Electromyography,
page 250).
Involvement of other body systems with diffuse neoplasia or metastasis may cause clinical signs unrelated
to the nervous system. This is most frequent in animals
with lymphoma. Paraneoplastic effects may also be
seen, particularly lymphoma-related hypercalcemia.
TUMOR BIOLOGY
There are a large number of tumor types that affect
the spine. Many tumors are only reported in small
numbers and these have been reviewed elsewhere
(LeCouteur and Child, 1995; LeCouteur, 2001; Oakley
and Paterson, 2003). Common tumor types are listed
in Table 12.1.
Spinal tumors may be classified by histopathology
(Table 12.1) or according to their anatomic location in
the spine (4.21):
Extradural.
Intradural/extramedullary.
Intramedullary.
Extradural tumors are the most prevalent type in
dogs, accounting for approximately half of all cases.
Intradural/extramedullary tumors make up another
third and intramedullary tumors the remainder
(LeCouteur, 2001). Tumors of bone and nerve roots are
most common in dogs; extradural lymphoma is most
common in cats (Oakley and Patterson, 2003). The
biology and expected behavior of a tumor must be
understood if it is to be diagnosed, staged, and treated
properly.
Extradural tumors
As the name implies these tumors lie outside the dura
mater. They typically cause pain with rapid neurological deterioration and include vertebral body tumors
and those that occupy the epidural space. Most vertebral body tumors are osteosarcomas; fibrosarcomas are
the next most common type (Dernell et al., 2000).
Chondrosarcoma, hemangiosarcoma and myeloma
are less frequent (Dernell et al., 2000; LeCouteur,
2001; Withrow and MacEwan, 2001). Meningiomas
and nerve sheath tumors can be extradural in location
Table 12.1 Classification of major spinal tumor types according to location within the spine
Extradural
Intramedullary
Primary
Osteosarcoma
Fibrosarcoma
Chondrosarcoma
Lymphoma
Hemangiosarcoma
Meningioma
Nerve sheath tumours
(schwannoma, neurofibroma,
neurofibrosarcoma, lymphoma)
Nephroblastoma
Sarcoma
Lymphoma
Ependymoma
Glioma
Lymphoma
Hemangiosarcoma
Reticulosis (focal
granulomatous
meningoencephalomyelitis (GME))
Metastatic
Metastatic
Metastatic
Carcinoma
Sarcoma
Melanoma
Lymphoma
Myeloma
Neoplasia
but are discussed under intradural tumors. Extradural tumors may also be metastatic in origin (osteosarcomas or carcinomas). Discovery of an extradural
tumor indicates the possibility of a primary mass elsewhere, particularly the mammary or thyroid glands,
kidney or urinary bladder.
Osteosarcoma and fibrosarcoma of the spine show
similar biological behavior. The usual cause of treatment failure is local recurrence (Dernell et al., 2000).
Spinal osteosarcomas are moderately aggressive with a
metastatic rate reported as 17% (Heyman et al., 1992).
A new grading system has been developed for osteosarcoma; most tumors are high grade (Kirpensteijn et al.,
2002). Grading for fibrosarcoma is less clear cut
(Ciekot et al., 1994; Hung et al., 2000).
Vertebral plasma cell tumors may exist as a solitary plasmacytoma or as multiple myeloma. Solitary
plasmacytoma is not associated with paraneoplastic
syndromes and usually does not cause monoclonal gammopathy. It has a better prognosis than multiple
myeloma but in humans about 50% go on to develop
disseminated disease (Rusbridge et al., 1999) (12.12).
This may be because a third of humans thought to have
solitary disease actually have occult multiple myeloma
on MRI (Moulopoulos et al., 1993). Multiple myeloma
does cause monoclonal gammopathy and additional
clinical signs can occur from hyperviscosity or paraneoplastic syndromes (Rusbridge et al., 1999).
Lymphoma has the propensity to involve the dura
mater and the spinal cord substance itself but is included
here as the usual site is extradural (Britt et al., 1984).
Canine lymphoma is usually a manifestation of multicentric disease. Lymphoma can affect nerve roots. Spinal
cord compression is often extradural and the mass can
sometimes be mistaken for epidural fat (Summers et al.,
1994). Some dogs show meningeal infiltration and most
have neoplastic lymphocytes in the CSF (Rosin, 1982;
Britt et al., 1984; Couto et al., 1984). Grading techniques for multicentric lymphoma can predict outcome
and these may be applicable to spinal lymphoma (Kiupel
et al., 1999; Dobson et al., 2001). Malignant histiocytosis and mastocytosis also affect the spinal cord occasionally (Moore and Rosin, 1986; Tyrrell and Davis, 2001;
Uchida et al., 2001; Moore et al., 2002). Feline lymphoma is discussed on page 261.
These tumors lie within the dura mater but outside
the spinal cord parenchyma. The most common tumors
in this category are meningiomas and nerve sheath
tumors (neurofibroma, neurofibrosarcoma, schwannoma, lymphoma). Both types can also be in an extradural location.
Meningiomas usually cause pain or chronic discomfort with slowly progressive neurological deficits. Some
meningiomas show an intramedullary pattern on myelography due to associated edema, even though they are
located outside the spinal cord (Fingeroth et al., 1987).
They tend to be discrete, firm to rubbery extramedullary masses. Meningiomas usually compress rather
than infiltrate spinal cord but some show invasion along
the perivascular space (Summers et al., 1994). Four of
13 spinal cord meningiomas in one report were invasive
(Fingeroth et al., 1987). Distant metastasis has been
reported for intracranial but not for spinal meningioma.
Meningiomas show a remarkable histopathological diversity, probably because the meninges arise from both
neural crest and mesodermal cells. They originate from
arachnoid granulations and cap cells, and often have receptors for progesterone (Summers et al., 1994; Theon
et al., 2000). Histopathological grading is not standardized although characteristics of benign versus malignant
meningiomas have been reviewed (Ribas et al., 1991;
Summers et al., 1994; Theon et al., 2000).
Nerve sheath tumors usually cause a chronic, progressive lameness. The two most common sites of origin are
from the plexus or from a nerve root within the
vertebral canal (Brehm et al., 1995). Brachial plexus
tumors are fusiform thickenings of one or more nerve
trunks, which sometimes coalesce into a common mass.
Spinal cord compression is usually a late event for
brachial plexus tumors but is often a presenting sign for
tumors that arise from within a nerve root (Brehm
et al., 1995). Malignant peripheral nerve sheath tumor is
the term preferred to malignant schwannoma or neurofibrosarcoma as it is usually not possible to discriminate
the cell of origin. These tumors are usually malignant
both cytologically and biologically (Summers et al.,
1994). Nerve sheath tumors often invade extensively
along the nerve (Oliver et al., 1965; Bradley et al., 1982;
McCarthy et al., 1993; Brehm et al., 1995; Jones et al.,
1995). Invasion into the spinal cord can occur but local
or distant metastasis is unusual (Brehm et al., 1995;
Sharp, 1988; Summers et al., 1994). They also metastasize occasionally along the subarachnoid space (2.24,
4.27, 4.28, 4.44B). Histopathologically, nerve sheath
tumors are often anaplastic with abundant mitosis and
necrosis. Specific grading systems are not yet standardized (Summers et al., 1994; Kuntz et al., 1997).
Nephroblastoma is another type of intradural/
extramedullary tumor that occurs between T10 and L2
vertebrae in young dogs (Pearson et al., 1997; Summers
et al., 1988; Terrell et al., 2000) (12.40). It was
referred to formerly as neuroepithelioma but this
tumor is really an extrarenal nephroblastoma equivalent to Wilms tumor in humans (Pearson et al., 1997).
249
250
Spinal cord compression is often severe but the mass

tends to be well-circumscribed and the prognosis has
been very good after surgical excision in some dogs
(Ferretti et al., 1993; Siegel et al., 1996). Metastasis is
rare but some of these tumors are invasive locally
(Baumgartner et al., 1988; Terrell et al., 2000).
Intramedullary
Intramedullary tumors occur within the spinal cord
substance and are the least common type. They typically cause neurological deficits that progress rapidly.
Tumors are either gliomas (12.4) or, very occasionally, metastatic tumors (Waters and Hayden, 1990;
LeCouteur, 2001). Staging relies mainly on neuroimaging
and MRI gives the best definition for intramedullary
lesions. However, MRI may not always differentiate
intramedullary tumors from those in the intradural/
extramedullary compartment. This is due partly to
the limitations of slice thickness and partly because
some tumors are only classified accurately using histopathology (Kippenes et al., 1999). A rare, non-neoplastic
differential diagnosis for an intramedullary spaceoccupying lesion is epidermoid cyst (Tomlinson et al.,
1988).
Radiography
Spinal tumors are usually relatively easy to diagnose
using standard methods. Following a physical and neurological examination, routine blood work and urinalysis
are performed. Radiography is often the technique with
the highest diagnostic yield. Survey radiographs of the
spine may not be diagnostic of soft tissue tumors, part
of the value of negative findings lies in the elimination
of obvious, destructive lesions and of differential diagnoses such as discospondylitis (14.11). The axial and
appendicular skeleton should be surveyed in suspected
plasma cell tumors in order to stage the tumor as solitary or disseminated (Rusbridge et al., 1999). Bony
changes will be seen mainly in vertebral tumors (12.1)
but other tumors, such as meningiomas and nerve sheath
tumors can cause mineralization or bone remodeling.
Spinal tumors do not metastasize frequently to the thorax but chest radiographs (three projectionsboth right
and left laterals with ventrodorsal or dorsoventral
views) should always be taken. Chest radiography is followed by abdominal ultrasound; in some animals the
spinal tumor itself may be a metastatic lesion (Bentley
et al., 1990; Jeffery, 1991).
Ultrasound
DIAGNOSIS
Diagnosis of a tumor is usually straightforward but
degenerative conditions like disc disease (see Chapters
7 and 8) and degenerative myelopathy, congenital
anomaly with late onset such as an arachnoid cyst, infectious inflammatory disease including discospondylitis and
focal granulomatous meningoencephalomyelitis (GME),
or vascular diseases such as ischemic neuromyopathy in
both dogs and cats, warrant consideration (see Chapter
14). Bone tumors may cross a disc space but this is rare;
this helps to differentiate them from discospondylitis
(Moore et al., 2000).
A general examination of the abdomen is indicated to

identify other sites of neoplastic involvement and evaluate
intercurrent disease (Platt et al., 1998). Ultrasonography
can also be very useful to identify some nerve sheath
tumors (Platt et al., 1999; Niles et al., 2001) and
may even be superior to MRI (Donner et al., 1994;
Simonovsky, 1997; Platt et al., 1999).
Electrophysiology
Evidence of denervation on electromyography (EMG)
is highly suggestive of nerve sheath tumor in a dog
with suspicious clinical signs (Wheeler et al., 1986;
Steinberg, 1988; Brehm et al., 1995). Nerve conduction
12.1 A: Four-year-old German shepherd

dog presenting with back pain and
paraparesis. Survey radiographs show a
destructive process affecting the T3
spinous process (arrows). Myelography
revealed extradural compression of the
spinal cord at this site (12.6A). B: Tenyear-old paraparetic Cocker spaniel with
an osteoproductive process within
and ventral to the T5 vertebral body
(arrow-heads) (12.6B).
Neoplasia
studies, late wave assessment and cord dorsum potentials may also provide supportive evidence (Cuddon
et al., 2003).
Myelography
This provides useful information about the location of
the tumor and its position within the vertebral canal
relative to the dura mater and spinal cord (4.274.31).
It is important to take two views (lateral and ventrodorsal) for correct localization (4.30, 4.31). Although
it will remain an important imaging modality, myelography is invasive and image quality can vary due to
technical problems such as contrast leakage or poor
filling of the subarachnoid space (see Chapter 4).
False-negative results can also occur, especially with
nerve sheath tumors (Carmichael and Griffiths, 1981;
Bradley et al., 1982; McCarthy et al., 1993) and myelography may miss as many as 42% of intramedullary
tumors (4.41) (Grem et al., 1985; Waters and Hayden,
1990). Transverse imaging by CT or MRI in general
allows earlier detection and better surgical planning
than myelography (Waters and Hayden, 1990; Brehm
et al., 1995).
Computed tomography
Computed tomography excels in imaging tumors that
destroy cortical bone (Dernell et al., 2000; Moore et al.,
2000) (12.6A, 12.7). Administration of intravenous
contrast may improve detection further, especially for
some nerve sheath tumors (McCarthy et al., 1993;
Niles et al., 2001). Beam-hardening artefact from the
humerus can interfere with resolution of brachial plexus
lesions (Platt et al., 1999). False-positive results can
occur in some animals that show thickening of nerve
roots on CT myelography suggestive of tumor (12.2).
Thickening can also be due to edema and not to neoplastic invasion (Simpson et al., 1999). MRI is probably
better than CT myelography at differentiating tumor
from edema (Simpson et al., 1999).
Magnetic resonance imaging

This is now considered the gold standard imaging
modality for human oncosurgery (Gilson, 2003) (see
page 57). It is superior to CT for imaging metastases
and occult myeloma within medullary bone (OFlanagan
et al., 1991; Rusbridge et al., 1999; Moore et al., 2000).
Sagittal images with T2-weighting provide excellent
lesion localization by identifying peritumoral edema.
Nerve sheath tumors are an exception as they do not
always show abnormalities on T2-weighting although
they usually enhance well with contrast (Kippenes et al.,
1999) (2.25, 12.3). As nerve sheath tumors are often
located away from the midline within the axilla the
study must be centerd to include soft tissues on the
affected side (Kippenes et al., 1999; Levitski et al.,
1999; Platt et al., 1999). For most tumor types, T1weighted images post-contrast are preferred for defining tumor volume and the relationship between the
tumor and surrounding tissues. Assessment of bone
marrow infiltration requires additional pulse sequences
such as fat suppression. The main limitation of MRI is
that some lower field-strength magnets are unable to
image thin enough tissue slices to permit fine anatomic
resolution (Kippenes et al., 1999).
Biopsy
This may be done as a needle aspirate or as an incisional
or excisional procedure. The choice is important and
biopsy tracts must be within the final field of surgical or
radiation treatment. Percutaneous fine needle aspiration
12.2 CT myelogram from a dog with chronic lameness, LMN deficits, Horners syndrome, and a painful mass in the right axilla.
Myelography was unremarkable. A: A CT scan showed subtle, right-sided, extradural compression at the cranial edge of T1
vertebra (arrow). B: Filling defects caused by C8 nerve roots as they cross the subarachnoid space (arrows). The right ventral
nerve root is much thicker than the left side. In this dog the ventral C8 and T1 nerve roots were invaded by tumor (12.49).
251
252
12.3 MRI of a dog with a 5-month history of unilateral pelvic limb lameness. It was unable to extend its stifle due to severe atrophy
and paralysis of the quadriceps muscle, the patellar reflex was absent but the withdrawal reflex was intact. A: Post-contrast,
T1-weighted image shows a high signal area in the dorsal portion of one psoas muscle (arrowhead). B: T2-weighted image with fat
suppression reveals a high signal mass at the same location as shown in A (arrowhead). Diagnosis was a nerve sheath tumor.
12.4 Five-year-old dog with a 1-week history of mild paraparesis and dribbling urine. Myelography revealed lack of subarachnoid
contrast from T10 to T13. A: CT-guided needle biopsy (arrowhead) at T11/12 was non-diagnostic. B: Hemilaminectomy revealed
discolored spinal cord and a needle aspirate was done over T11. C: Cytological diagnosis was ependymoma or oligodendroglioma.
Euthanasia was performed and histopathological diagnosis was poorly differentiated glioma (Fernandez et al., 1997). In humans the
histological grade of ependymoma does not necessarily correlate with prognosis (Mork and Loken, 1977; Ritter et al., 1998).
of the spinal cord was shown to produce no neurological

deficits when conducted at L1/2 or L2/3 in seven
normal dogs (Schulz et al., 1994). It has also been
used safely in five cats: one with an intramedullary
lesion at L2/3 and four with epidural lesions (Irving
and McMillan, 1990). Overall, percutaneous needle
biopsy in humans is rapid and minimally invasive; accuracy rates range from 80 to 95% with image-guidance
(Ayala and Zornosa, 1983; Schiff et al., 1997; Ashizawa
et al., 1999; Gilson, 2003)(12.4).
Incisional biopsy provides a higher diagnostic yield
than percutaneous techniques and larger tissue samples
make histological diagnosis more accurate. However,
incisional biopsy necessitates two-stage procedures unless
combined with frozen sections and it can also disseminate
tumor cells if done poorly (Gilson and Stone, 1990).
Excisional biopsy is better for spinal cord tumors as
most are well localized and excision is also the treatment
of choice except for lymphoma or plasma cell tumors.

Needle biopsy is preferable when lymphoma or plasma
cell tumor is suspected (Rusbridge et al., 1999; Gilson,
2003).
STAGING
General staging includes neurolocalization, screening for
any concurrent or paraneoplastic disorders, radiographic
survey of the entire vertebral column, abdominal ultrasound, and screening for pulmonary metastases using
thoracic radiographs or spiral CT. Aspiration of lymph
nodes and bone marrow may also be indicated (Withrow
and MacEwan, 2001). Screening for bone metastases in
humans is done best with a bone scan as this is more
accurate than survey radiography although false-positive
results have been reported in dogs (Bentley et al., 1990;
Berg and Lamb, 1990; Rusbridge et al., 1999; Dernell
Neoplasia
12.5 Diagram of an anatomic

classification system for the extent of
tumor involvement within and adjacent
to a vertebral body (modified from Gilson,
2003). See below for an explanation of
zones IIV.
12.6 A: CT scan of the dog shown in 12.1A reveals an osteolytic, expansile mass within the spinous process of T3 (zone I,
arrowhead). B: Transverse, T1-weighted MRI of the dog shown in 12.1B reveals a large mass ventral to the vertebra (zone III, white
arrowheads); it has also extended dorsally around the rib (*) into zone II and into the vertebral canal (white arrows). The spinal cord
(black arrow) is displaced by tumor within the vertebral canal (zone IV) (12.41).
et al., 2000) (see page 59). Bone scan may also fail to
identify plasma cell tumors (Rusbridge et al., 1999).
A CSF sample should be collected as neoplastic cells will
be identified occasionally. Neoplastic cells were found in
CSF from 10 of 12 dogs with lymphoma, which is more
often than for feline lymphoma (one third of cats, see
Feline spinal tumors, page 262) (Rosin, 1982; Couto
et al., 1984). Local and regional tumor imaging is done by
myelography, CT or MRI. CT is very useful for looking
at cortical bone but MRI is now the gold standard modality for both neural and vertebral tumors (Gilson, 2003)
(12.3, 12.6B).
Anatomical classification of the extent of vertebral

tumors is used to determine surgical approach and
resection technique and may also assist in determining
the subsequent fixation to be employed. Classification
is based on the region of spine involved and on the portion(s) of vertebral body affected. A guide for anatomic
classification of vertebral body involvement is shown in
12.5 (Gilson, 2003).
For most tumors of zone I and for smaller tumors of
zone II, complete en bloc removal is recommended via
a dorsal or dorsolateral approach (Klopp and Quinn,
2001; Gilson, 2003) (12.6A).
253
254
12.7 A: Survey radiograph of the lumbar

spine in a 9-year-old Rottweiler with
lumbar pain and severe paraparesis.
There is loss of bony detail in the
vertebral arch (arrow). B: CT scan shows
destruction of the pedicle and part of the
vertebral body (zones II and IV). The
extent of bony involvement is clearer on
this than on either the survey radiograph
or the myelogram (not shown). An
intraoperative view is shown in 12.9
(see page 256); the necropsy
appearance in 12.8.
consideration of results in humans, there is a great deal

of room for improvement in our management of animals
with spinal tumors. It is probable that improvements will
be incremental based on gradual advances in diagnostic
and surgical techniques combined with earlier diagnosis
and better adjuvant therapies (Gilson, 2003).
Medical treatment
12.8 Section through the lumbar vertebra at necropsy

performed 4 weeks after imaging and laminectomy revealed
the extent of the tumor re-growth, particularly involving the
vertebral body and the lamina (zones II, III and IV). The CT image
in 12.7 is very similar to this picture.
Resections in zone II at certain sites may be challenging because of the vertebral artery, ribs or wings of
the ilium. Depending on their size, many tumors in
zones II, III and IV will either require piecemeal intracapsular resection, vertebrectomy or both (Chauvet
et al., 1999; Dernell et al., 2000) (12.612.8). Tumors
in zone III must be assessed critically for adherence to
cervical soft tissues or to the thoracoabdominal great
vessels or organs (12.41) (Gilson, 2003).
TREATMENT
Treatment is not possible in every animal with a spinal
tumor, but an aggressive approach will prove rewarding
in some patients. With the exception of myeloma, vertebral body tumors are difficult to treat, but the fundamental techniques are well established. Based on
Medical treatment of spinal tumors is unlikely to provide curative therapy in most patients. The exception
is for animals with multiple myeloma where good
responses are obtained with chemotherapy alone or in
combination with radiotherapy; melphalan and prednisone are the standard chemotherapeutic agents
(Rusbridge et al., 1999). In lymphoma, chemotherapeutic regimens are employed ideally without surgical
intervention. Agents such as cytosine arabinoside may
be added due to the difficulty of getting drugs across
the blood brain barrier (Couto et al., 1984; Withrow
and McEwan, 2001). Lomustine is of value for CNS
tumors (Jeffery and Brearley, 1993; Moore et al., 1999;
Fan and Kitchell, 2000) and oral hydroxyurea may
prove useful for meningioma (Mason et al., 2002).
Doxorubicin or cisplatin (in either systemic or slowrelease formulations) have been used to treat various
other tumor types (Jeffery, 1991; Dernell et al., 2000).
Palliative therapy (analgesics, anti-inflammatory agents)
may also be used to alleviate pain and peritumoral
edema, although when used alone they only relieve
clinical signs for relatively short periods.
Radiation treatment
Ideally, radiation treatment is used as an adjunct to surgery, unless the tumor is small, well localized and particularly sensitive to radiation (e.g. lymphoma or plasma
cell tumor) in which case radiation may be used as the
only therapy (LeCouteur, 2001; Rusbridge et al., 1999;
Withrow and MacEwan, 2001). An important consideration is that previous surgery decreases the radiation
tolerance of brain (and probably spinal cord) by causing
injury to local blood vessels. If surgery reduces tumor
Neoplasia
burden to microscopic levels, then this is not crucial. If

incomplete resection leaves a larger tumor burden, this
factor may prevent the aggressive, postoperative radiation therapy necessary to control macroscopic disease
(Smith and LaRue, 2000). Radiation can also cause
injury to the spinal cord (Tiller-Borcich et al., 1987;
Powers et al., 1992; Rusbridge et al., 1999):
Early delayed effects occur from 2 weeks to
3 months after therapy. They are probably due to
transient demyelination; signs are often temporary
and may respond to systemic corticosteroids.
Late effects occur 6 months or more after
radiotherapy and are caused by vascular endothelial
and glial cell injury. Signs are usually progressive;
this is the main factor limiting radiation dose.
More accurate radiation planning allows higher treatment doses to be delivered to the tumor while sparing
normal tissues (Smith and LaRue, 2000). Several
tumors affecting the spinal cord appear to respond to
radiation (Siegel et al., 1996; Oakley and Patterson,
2003) (Tables 12.412.6). These include meningioma,
ependymoma, nephroblastoma and some but not all
nerve sheath tumors (12.49, 12.50).
Surgical treatment
Owing to the inherent anatomical limitations of dealing
with the CNS, aggressive surgical margins are rarely possible for vertebral or spinal cord tumors. Nevertheless,
removal of vertebral tumors using wide margins for
most of a mass, with a more limited margin for the portion nearest the spinal cord, may not affect adversely
the success of local tumor control (Gilson, 2003).
Effective biological barriers to neoplastic tissue should be
used to advantage; these include collagen rich/vascular
poor structures such as intervertebral disc, cortical bone,
dorsal longitudinal ligament, dura mater, interspinous
and intertransverse ligaments, and fascial sheaths of
paraspinal muscles (1.29, 1.321.35, 8.65). General
principles of surgical oncology must be adhered to in
order to insure a successful outcome (Gilson, 2003).
Normal tissues must be protected by impenetrable barriers and copious lavage with suction used to clear the
surgery field of tumor cells. Contaminated instruments
should be discarded after tumor excision and new materials and instruments are used for closure. The most
proximal portion of the specimen is identified and the
tumor margins are assessed histologically (Gilson,
2003). The goal of surgery is to achieve complete resection or, at minimum, reduce tumor burden to residual
microscopic disease (Algorithm 12.1).
The best results are obtained in humans when the
surgical approach and type of resection is targeted to
the exact location of the mass. Only 4% of patients

with anterior (i.e. ventral) tumors worsened after an
anterior approach whereas 26% worsened after posterior (i.e. dorsal) laminectomy. Similarly, over 30% of
patients with posterior tumors could still walk after
a posterior approach compared to less than 16% of
patients with anterior tumors approached posteriorly
(Gilson, 2003).
Vertebral tumors Dogs may show good initial

improvement after debulking surgery (Moore et al.,
2000), but then often worsen markedly (Bentley et al.,
1990; Heyman et al., 1992; Chauvet et al., 1999)
(12.9). Of 20 dogs undergoing debulking surgery, seven
were unchanged (35%), eight were better (40%), and
five (25%) worse after surgery. Improvement is usually
temporary unless combined with effective adjunctive
therapy (Dernell et al., 2000).
Stabilization is required after most partial and all complete vertebral body resections (Chauvet et al., 1999;
Dernell et al., 2000; Gilson, 2003); temporary stabilization is required prior to removing an entire vertebra
(Chauvet et al., 1999; Gilson, 2003) (12.41). Principles
of spinal stabilization after tumor resection are similar to
those used for spinal trauma (Sundaresan et al., 1990)
(see Chapter 13). The most versatile technique is metal
implants and bone cement. An external fixator has some
advantages but is better suited to techniques that do not
use a cement spacer (Walker et al., 2002). Some studies
report a high incidence of cement failure in humans over
time (McAfee et al., 1986; Sundaresan et al., 1990;
Gilson, 2003). This does not appear to be the case in
animals following spinal trauma, probably because of the
relative difference in life span, although this issue has
not been studied in spinal neoplasia (see Chapters 11
and 13). Whether or not postoperative radiation will
affect implants is also unclear (D.E. Thrall, personal
communication). Titanium implants may be affected
less than other metals and have the advantage that they
also permit postoperative MRI.
Following vertebrectomy the surgeon has three main
options:
1. Replace excised bone using grafts such as freezedried femoral allograft (Veterinary Transplant
Services, Seattle, WA) packed with autogenous,
corticocancellous bone from a vertebral spinous
process. The graft replaces the vertebral body and
is anchored by plates (12.42) or with metal
implants and bone cement. Autologous bone can
also be harvested from ilium, rib, fibula, ulna or
possibly a coccygeal vertebra (Yeh and Hou,
1994). Additional pieces of cortical graft or an
adjacent rib are split longitudinally to make
255
256
Algorithm 12.1 Surgical decision-making in

spinal tumours.
Neurological
localization
Vertebral
tumour
Survey radiographs of
spine and thorax, plus
abdominal ultrasound
Spinal cord
tumour
Biopsy
Neuroimaging - CT or
MRI or (myelography)
Laminectomy
Lymphoma, myeloma,
nerve sheath tumor,
see text
Resection under
magnification
Anatomical
classification
Surgical
approach
and margins
Zone I, Zone II (small)
Dorsal or dorsolateral
approach en bloc
resection
Zone III
Ventral or
lateral approach
intracapsular
or vertebrectomy
Zone IV or
crossing
Zones II or III
Combined approach
vertebrectomy
12.9 Intraoperative photographs from the dog in 12.7 and 12.8. A: Exposure of the vertebral column revealed a mass of abnormal
tissue on one vertebra (arrow). B: Hemilaminectomy was performed; much of the vertebral arch was absent. Debulking of the soft
tissues revealed a large defect in the vertebral body (a). The spinal cord (b), spinal nerve (c) and spinal ganglion (d) are seen. The
dog recovered well and was able to walk within 2 days. Histopathological diagnosis was anaplastic carcinoma. Attempts to find the
primary tumor failed, and the dog was paraparetic again within 4 weeks, when it was euthanized (12.8).
protective strut grafts placed dorsal to the spinal

cord (Chauvet et al., 1999; Gilson, 2003) (12.42).
For further information see bone grafts (see
page 292).
2. Fashion a spacer from bone cement to replace

the vertebral body. This is anchored using metal
implants and additional bone cement. The
advantage of the spacer is that it provides
Neoplasia
immediate stability and, because there is no graft

to revascularize, this technique is not affected by
adjunctive therapies such as radiation (Brasmer and
Lumb, 1972; Sundaresan et al., 1985; Matsui
et al., 1994; Gilson, 2003). Ideally, some stabilization
must also be provided to dorsal spinal elements.
This could be fashioned from cement although
thermal injury to the spinal cord must be avoided,
such as by using Gelfoam (Roosen et al., 1978;
Boker et al., 1989). An alternative is to add
modified segmental fixation (12.41).
3. Simply close the gap and fuse the adjacent
vertebrae to leave a shortened vertebral column. In
one study the L2 vertebra was removed from eight
normal dogs; five of these were clinically normal
11 months later (Yturraspe and Lumb, 1973). The
advantage is that there is no spacer to loosen or
revascularize and there is good protection dorsally
for the spinal cord. Some histological changes were
reported at the surgical site in normal dogs and it is
not known how well a compromised spinal cord
would accept this technique (Yturraspe et al.,
1975). This is probably the best approach of the
three although the ribs may complicate its use in
thoracic areas.
Nerve sheath tumors present a difficult surgical challenge but are also potentially
curable if three criteria can be satisfied:
1. Several centimeters of normal nerve can be excised
with complete histological margins proximal and
distal to each branch of the mass. Simple visual
inspection of margins usually underestimates
the extent of invasion markedly (Bradley et al.,
1982).
2. There is no invasion into the sympathetic nerves
and ganglia or the spinal cord (Bradley et al.,
1982; Bradney and Forsyth, 1986) (12.49,
12.50).
3. There is no local (McCarthy et al., 1993), or
distant metastasis (Oliver et al., 1965; Bradley
et al., 1982; Brehm et al., 1995). These tumors can
sometimes invade through the epineurium and
even into adjacent bone (Jones et al., 1995).
It is usually necessary to amputate the forequarter or
hemipelvis in order to satisfy these criteria (Bailey,
1990; Targett et al., 1993; Simpson et al., 1999) (Table
12.6). This is also necessary due to the diffuse nature
of the tumor within nerves, the high recurrence rate,
and the critical nature of much of the plexus for adequate limb function. For the thoracic limb, neoplastic
infiltration of C8 and T1 spinal nerves, of more than
one spinal nerve or a major portion of the plexus,
Nerve sheath tumors
necessitates that the limb be amputated as the resultant deficit to the radial, median and ulnar nerves will
be too disabling. Good preoperative imaging is especially helpful in surgical planning, such as in determining the exact nerve roots involved and where to
perform the laminectomy (Sakai et al., 1996; Platt
et al., 1999).
Spinal cord tumors For spinal cord tumors,

resection at the pseudocapsule between the mass and
normal neural tissue can be curative for many tumor
types. There is much less risk of postoperative instability
for tumors of the spinal cord than for vertebral tumors.
Extensive dorsal laminectomy warrants some form of
stabilization, especially after pedicle or articular process
removal (Siegel et al., 1996) (12.39). Ideally, surgery for
spinal cord tumors should combine good preoperative
imaging, preferably using MRI, intraoperative imaging
with ultrasound and magnification, together with notouch techniques employing lasers and ultrasonic aspiration (Gilson, 2003). As a minimum, good preoperative
imaging with intraoperative magnification is preferred
for intradural/extramedullary tumors (Ferretti et al.,
1993). Magnification is essential for resection of
intramedullary tumors in order to differentiate tumor,
pseudocapsule and normal tissue; to minimize spinal
cord manipulation; and to allow recognition and ligation
of blood vessels that supply the tumor and not the normal spinal cord (Jeffery and Phillips, 1995; Gilson,
2003). Mass removal may lead to reperfusion injury of
the spinal cord, and spinal cord manipulation can also
cause damage. Use of preoperative vitamin E or perioperative methylprednisolone sodium succinate (MPSS)
may be beneficial (Pietila et al., 2000) (see
page 83).
As microscopic tumor burden often remains after
resection of many tumors, adjunct therapy using
chemotherapy or radiation should be considered for
follow-up treatment (Jeffery, 1991; Siegel et al., 1996).
One possible exception is a well-circumscribed nerve
sheath tumor resected with wide margins of normal
nerve on either side or in dogs with more extensive
nerve sheath tumors that undergo forequarter amputation and for which histological margins show no evidence of residual tumor (12.50).
TUMOR DISSECTION
Gentle tissue handling is vital for tumors in or near the
spinal cord. Manipulation must be kept to a minimum to
avoid iatrogenic damage; it is preferable to avoid touching
the spinal cord at all. If manipulation is necessary then
dural sutures are used (Yturraspe and Lumb, 1973)
(12.40). The spinal nerve and vessels can be sacrificed in
257
258
order to improve access if a tumor is difficult to remove

(8.51), but this should be avoided if possible in the
brachial and lumbosacral plexuses unless the limb can be
sacrificed (Fingeroth et al., 1987). The tumor site is
isolated with cottonoid pledgets (5.28) to both prevent
dissemination of tumor cells and retard migration of
hemorrhage into the subarachnoid space. The goal of
surgery is to remove the tumor without causing damage
to the spinal cord rather than to preserve a beautiful
specimen (Fingeroth et al., 1987). If the tumor involves
dura mater, sharp dissection is used to decrease spinal
cord manipulation. Large dural defects seem to be of no
consequence in animals and can be left open (14.8)
(Fingeroth et al., 1987; Ferretti et al., 1993). Soft tumors
can be removed by gentle suction but it is important to
retain a piece of the mass for histopathology. For tumors
that involve nerve it is crucial to mark the proximal
margins of the mass accurately (12.50) (Targett et al.,
1993; Niles et al., 2001).
Intramedullary tumors are best approached via durotomy, piaotomy and dorsal myelotomy using magnification. An ultrasonic aspirator, microsurgical bipolar
cautery and a laser are particularly helpful at this stage
but good results may be obtained using bipolar cautery,
loupes and copious irrigation with suction to preserve
visibility (Jeffery and Phillips, 1995). Tumor removal is
either by blunt dissection to create a plane between normal tissue and the mass, or the mass is debulked with
the aim that the pseudocapsule then falls in on itself and
a plane can then be developed between normal tissue
and mass (Jeffery and Phillips, 1995; Gilson, 2003).
Although considered highly desirable in people, intraoperative monitoring of spinal cord evoked potentials
requires dedicated personnel and expertise that are
rarely available for animals (Cuddon et al., 2003).
COMPLICATIONS
The main sources of complications are listed in Table
12.2. Intraoperative complications include iatrogenic
spinal cord injury, which can occur from manipulation,
instability or vascular compromise (Brasmer and Lumb,
1972; Fingeroth et al., 1987) (Table 12.3). The surgeon may also be unable to find the tumor, especially if
it is intramedullary. Intraoperative ultrasound is particularly useful in this situation (Nakayama, 1993). The
tumor may be invasive and therefore difficult to resect,
or access to the tumor can be difficult if it extends ventral to the spinal cord. Access may be improved by
performing a partial corpectomy of the vertebral
body (Moissonnier et al., 2002b) (8.47B). If nerve roots
are involved at an intumescence it may be impossible to
Table 12.2 Main sources of complications
Intraoperative
Technical problems
Technical errors
Early
postoperative
Late
postoperative
Infection
Lack of stability
Non-neurological
problems
Infection
Implant failure
Table 12.3 Complications
Intraoperative
Iatrogenic spinal
cord injury
Difficulty locating
tumor
Invasive tumor
Poor access to
tumor
Incomplete
resection
Early
postoperative
Late
postoperative
Instability
Pathological
fracture (12.12,
12.12)
Implant failure
Wound infection
Sepsis
Fat graft necrosis
(12.10)
Diagnostic error
Implant failure
Tumor recurrence
(12.11, 12.50)
Peridural scar
achieve complete resection without amputating the

limb (Fingeroth et al., 1987).
Technical errors during surgery may result in instability
in the early postoperative period (Table 12.3). Dorsal
laminectomy, especially if combined with removal of
the articular processes, causes a significant reduction
in spinal stability (see page 286) and in humans up to
20% of patients experience neurological deterioration
(Findlay, 1984, 1987; Gilson, 2003). Preoperative vertebral body collapse in humans also has a negative
effect on outcome after laminectomy (Findlay, 1987;
Santen et al., 1991; Chauvet et al., 1999) (12.11).
Laminotomy has been used in children instead of
laminectomy to avoid late deformity due to instability
and may be warranted in certain situations for animals
(Fingeroth and Smeak, 1989; Cristante and Herrmann,
1994; Gilson, 2003).
Implant failure may occur due to inadequate fixation
(Brasmer and Lumb, 1972; Yturraspe and Lumb, 1973;
Onimus et al., 1996; Miller et al., 2000). Failure of
bone cement tends to occur in humans over time
although this is less likely in animals, probably due in
part to differences in life span (McAfee et al., 1986;
Jonsson et al., 1994; Yerby et al., 1998; Lu et al., 2001;
Jang et al., 2002) (see page 291). However, the incidence
of metal implant and bone cement failure could prove
to be higher following tumor irradiation than it is in
spinal trauma (D.E. Thrall, personal communication).
Neoplasia
12.10 Aseptic necrosis of fat graft after dorsal laminectomy.

The dogs neurological status deteriorated several days
after surgery. Myelography revealed extradural compression.
The graft was removed and the dog made a good recovery
(8.8).
12.12 This dog had been diagnosed 5 years previously with a

solitary plasma cell tumor in L4 when it was treated with
melphalan, prednisone and irradiation (4.474.49); back pain
recurred after 63 months. A lytic lesion on the spinous process
of T7 was removed but the sample was non-diagnostic. Two
months later the dog had sudden onset of tetraparesis due to a
pathological fracture of C7. Plasma cells were found at necropsy
in C7, T10, L4 and the spleen and kidney (Rusbridge et al., 1999).
a 10% rate of histopathological misdiagnosis has been

reported and the rate may be up to 15% in humans
(Levy et al., 1997a).
Some problems only become apparent later in the
postoperative period (Table 12.3). The most important
problem is tumor recurrence. This can also cause instability but unless vertebral column failure is catastrophic (12.11, 12.12), it may respond to repeat
surgery or adjunctive therapy (Bell et al., 1992;
Schueler et al., 1993).
B
12.11 A: Lateral radiograph of the lumbar spine of a
paraparetic 3-year-old mixed-breed dog with a destructive
lesion in the L5 vertebral body (arrow). It underwent a dorsal
laminectomy and biopsy. B: The dog became paraplegic with
reduced deep pain sensation 2 days later. There is marked
collapse of the L5 vertebral body compared to the preoperative
radiographs. Histological diagnosis was fibrosarcoma (12.42)
(from Chauvet et al., 1999).
Fat grafts used to cover dorsal laminectomies must be

thin enough to revascularize (8.8, 12.12; see also
Laminectomy healing, page 86).
Further potential complications include sepsis, which
is a particular risk for debilitated animals with cancer
(Chauvet et al., 1999), and diagnostic errors. For example,
POSTOPERATIVE CARE
Routine nursing considerations are discussed in
Chapter 15. Any postoperative deterioration in neurological status will necessitate a high standard of nursing
care (Chauvet et al., 1999). If radiation treatment is
planned, it is usual to delay this for 10 days or until the
wound has healed (Siegel et al., 1996).
PROGNOSIS
In some circumstances where a tumor is not amenable
to treatment because of its location or because a pathological fracture has occurred, euthanasia is indicated.
Alternatively, palliative radiation therapy may be considered along with analgesia as discussed in Chapter 15.
259
260
Table 12.4 Outcome of dogs with extradural tumors

Myxoma/
Myxosarcoma3
(n 3)
Osteosarcoma3
Osteosarcoma2
Fibrosarcoma2
Fibrosarcoma1
Adjuvant
treatment
None
None
Variable
Variable
Vaccine
Chemotherapy
radiation
Survival
(months)
11; 19; 35
1; 10
4.4
3.7
24
17, 26, 65*
Final
outcome
N/A
N/A
Variable
Variable
N/A
Recurrence
Tumor
type
Plasma cell
tumors4
1
Chauvet et al., 1999; 2 Dernell et al., 2000; 3 Levy et al., 1997a; 4 Rusbridge et al., 1999.
* One dog euthanized at 4 months due to radiation myelopathy.
N/A, not available.
Treatment may only lead to short-term remission in

some patients, but the quality of life during that period
can be acceptable if pain is controlled adequately.
Remission may be achieved for CNS lymphoma in
some dogs. Survival times, however, range from days to
a few months (Couto et al., 1984; Turner et al., 1992).
Results may be improved by use of radiation therapy
(LeCouteur, 2001).
Surgery without curative intent for soft tissue

sarcomas may produce useful remission if
combined with other treatment modalities
(Dernell et al., 2000). However, the prognosis is
poor if the animal has marked neurological deficits
(Dernell et al., 2000; Gilson, 2003) (12.712.9),
and is also influenced negatively by pathological
fracture (Findlay, 1987; Chauvet et al., 1999)
(12.11).
Extradural tumors
Extradural tumors, with the exception of plasma cell
tumors and lymphomas, carry a guarded prognosis
(Dernell et al., 2000; Gilson, 2003) (Table 12.4).
Preoperative neurological status is one important
determinant of final outcome in both humans and
dogs (Dernell et al., 2000; Gilson, 2003). In
humans, between 60 and 95% of patients who could
walk well before surgery still did so after surgery but
only 3565% of paretic patients and less than 25%
of paraplegic patients walked again (Gilson, 2003).
Another factor is anatomic location; four dogs with
tumors in the dorsal spinal compartment were alive
from 7 to 30 months after excisional surgery; a
fifth died from complications after 2 months
(Klopp and Quinn, 2001).
Local disease control is also an important prognostic
factor for overall survival (Kuntz et al., 1997;
Dernell et al., 2000). Dogs with soft tissue
sarcomas that are resected with incomplete margins
are 10 times more likely to have local recurrence
than dogs with complete margins (Kuntz et al.,
1997). One dog with a fibrosarcoma treated by
vertebrectomy survived for over 2 years (Chauvet
et al., 1999) (12.42). Another dog that underwent a
tail amputation for a caudal osteosarcoma survived
2.5 years (Heyman et al., 1992).
Intradural/extramedullary tumors
Intradural/extramedullary tumors have a variable prognosis depending largely on tumor type:
Meningiomas treated by surgical excision alone in
nine dogs gave good long-term outcomes in six
(Fingeroth et al., 1987). Survival is probably
improved further by use of adjunctive therapies
(Bell et al., 1992; Siegel et al., 1996). Even
incomplete resection of meningioma can give a
good outcome when followed by radiation therapy;
three such dogs had 8-, 15- and 25-month survival
times (Siegel et al., 1996) (12.28). A mean survival
of 19.5 months was reported for another 10 dogs
with meningioma (Moissonnier et al., 2002a).
Survival data for dogs with meningioma are shown
in Table 12.5.
Nerve sheath tumors on the other hand have
given very disappointing results overall. This is
particularly true for dogs with tumors that arise
within the vertebral canal and for those undergoing
incomplete resections (Bradley et al., 1982; Brehm
et al., 1995; Jones et al., 1995; Kuntz et al., 1997;
McCarthy et al., 1993; Schueler et al., 1993;
Targett et al., 1993). Only 6 of 51 dogs (12%)
where margins were not assessed remained disease
free after 1 year (Brehm et al., 1995). Ideally all
Neoplasia
Table 12.5 Outcome of dogs with intradural/extramedullary lesions (excluding nerve sheath tumors (12.6))
Tumor
type
Hemangio-4
sarcoma
Nephroblastoma5
Nephroblastoma2
Nephroblastoma8
Meningioma8
Meningioma7
Meningioma1
Meningioma3
Meningioma6
No. of dogs
10
Adjuvant
treatment
Doxorubicin
Radiation
None
Radiation
Radiation
N/A
Radiation
None
None
Survival
(months)
11
36
6.5
825
median
13.5
46 & 47
19
5 live
6 months;
1 alive at
36 months
Mean
19.5
Final
outcome
Metastasis
N/A
N/A
Recurrence
See
citation
N/A
Recurrence
See
citation
N/A
Bell et al., 1992; 2 Ferretti et al., 1993; 3 Fingeroth et al., 1987; 4 Jeffery, 1991; 5 Jeffery and Phillips, 1995; 6 Moissonier et al., 2002a; 7 Levy et al., 1997a;
Siegel et al., 1996.
N/A, not available.
8
Table 12.6 Outcome for nerve sheath tumors following complete excision
Targett et al.,
1993
Bailey, 1990
Platt et al.,
1999
Niles et al.,
2001
Siegel et al.,
1996
Unpublished
(NCSU)
Number of dogs
Survival (months)
9*, 18*
42*
9*
16*
25*
24*, 36*, 8**
Adjuvant
None
None
None
None
Radiation
None
* Without recurrence.
** Probable recurrence, no necropsy; histological margins had not been evaluated.
NCSU, North Carolina State University.
involved branches are resected along with a

generous margin of normal nerve and histological
margins are evaluated. For tumors involving the
plexus this nearly always necessitates concomitant
amputation (Targett et al., 1993; Simpson et al.,
1999) (12.49, 12.50). Usually nerve sheath tumors
are contained within the nerve by an effective
tissue barrier of epineurium. Therefore
all neoplastic tissue can often be excised with
complete margins, in which case these tumors
should be curable (Targett et al., 1993; Ganju et al.,
2001). Results for dogs where apparent complete
excision was performed are shown in Table 12.6.
Nerve sheath tumors show an unreliable
response to radiation (Siegel et al., 1996)
(12.49, 12.50). Malignant nerve sheath tumors
in humans also have poor outcomes and the best
survival times follow amputation (Ganju et al.,
2001).
Intramedullary tumors
Intramedullary tumor resection has only been reported
in three dogs. Two had ependymomas and both underwent surgery and radiation therapy (Jeffery and
Phillips, 1995; Siegel et al., 1996). One dog made a full
recovery and survived for 70 months (Siegel et al.,
1996). The second had a poorly differentiated tumor;
it could walk well 3 months after surgery but was euthanized because it remained incontinent (Jeffery and
Phillips, 1995) (see also 12.4). A third dog had a paraganglioma that was removed and the dog was still alive
10 months later (Poncelet et al., 1994).
FELINE SPINAL TUMORS

Spinal tumors account for over half of all cases of feline
spinal disease not associated with trauma (Wheeler,
1989). Several types of spinal tumors have been
reported, with lymphoma being the most common.
261
262
Neurological signs are seen in 512% of cats with systemic lymphoma (Spodnick et al., 1992; Lane et al.,
1994). Most lymphomas are solitary and extradural but
in some cats the tumor may extend over multiple
vertebrae or there may be distant meningeal infiltration
(12.13). Nerve root infiltration is also reported
(Spodnick et al., 1992; Lane et al., 1994). About two
thirds of cats with lymphoma are feline leukemia virus
(FeLV) positive (35/54) and have involvement of other
organs (32/50), mostly in the kidney (Spodnick et al.,
1992; Lane et al., 1994; Noonan et al., 1997).
A less common cause of spinal lymphoma is feline
immunodeficiency virus (FIV). In contrast to FeLVassociated lymphoma, which is a disease of T cells,
most FIV-associated lymphomas are B cell in origin (Callanan, 1996) (12.14). Other tumor types
12.13 Myelography showing extradural spinal cord

compression due to FeLV-associated lymphosarcoma in a
1-year-old cat that presented with a sudden onset of lower
motor neuron signs. Treatment was not attempted.
include: meningioma (most common), osteosarcomas,

chondrosarcoma, myeloma, lipoma, glioma and nerve
sheath tumor (Mills et al., 1982; Shell et al., 1987;
Wheeler, 1989; Levy et al., 1997b).
Diagnosis
Some extradural tumors may affect the vertebral
bodies, causing bony changes on survey radiographs,
but this is not typical. Myelography or transverse imaging is required to confirm the diagnosis in most cats
(Asperio et al., 1999). Compression is usually extradural
in lymphoma (12.13, 12.14) although occasionally other
patterns are seen, particularly if nerve roots are affected.
Other findings that may indicate possible neoplasia
include soft tissue opacities in the cranial thorax with
mediastinal lymphosarcoma, organomegaly or pulmonary metastases.
Other diagnostic tests are indicated to confirm the
presence of a tumor and to define its nature. The bone
marrow is involved in approximately 75% of cats
(14/19) with extradural lymphoma and bone marrow
aspiration is indicated for cytological analysis (Spodnick
et al., 1992; Noonan et al., 1997). CSF contains malignant lymphocytes in only 11 of 31 of cats with CNS
lymphoma (Spodnick et al., 1992; Lane et al., 1994;
Noonan et al., 1997). In some instances, fine-needle
aspiration (page 252) or biopsy at exploratory surgery
may be the only means of making a definitive diagnosis
12.14 Myelogram from a cat with acute ataxia, paraparesis, generalized lymphadenopathy and herpes keratitis; it was FeLV negative
but FIV positive. Neurological deficits localized to T3L3 spinal cord; the thoracic limbs were normal. Neoplastic lymphocytes were
detected in the CSF; flow cytology identified these as B-cells. A: Myelogram reveals extradural compression at T12, which was
unexpected given the normal thoracic limbs. B: CT myelography confirms the mass as extradural (arrow). Response to chemotherapy
was excellent and remission lasted 1 year. Radiation produced a second remission; lomustine produced a third, brief remission but the
cat was euthanized 18 months after presentation.
Neoplasia
(Irving and McMillan, 1990; Spodnick et al., 1992). In

an FeLV-positive cat with spinal disease related to an
extradural soft tissue mass demonstrated by neuroimaging, it may be reasonable to make a diagnosis of
lymphoma and to treat accordingly.
Treatment
Chemotherapy often leads to a rapid improvement in
neurological function but remission times may be brief.
Chemotherapeutic protocols are described elsewhere
(Withrow and MacEwan, 2001). Clearly, surgical treatment alone in lymphoma has only a limited role because
of the multifocal nature of the disease (Spodnick et al.,
1992). Radiotherapy may be of benefit in lymphoma,
but should probably be combined with chemotherapy
for the same reason (Lane et al., 1994). If lymphoma is
not suspected strongly, biopsy is required to diagnose
tumor type and can be combined with surgical excision
of the mass (Levy et al., 1997b).
Prognosis
Although initial improvement may be seen after therapy
for lymphoma, long-term prognosis is unfavorable
because of the propensity for multifocal involvement,
local recurrence or the development of systemic FeLVrelated disease. Three of six cats treated with chemotherapy went into complete remission but this usually lasted
less than 6 months (mean 3.5 months). Remission for
more than 1 year was seen in only 2 of 14 cats, one of
which underwent surgery and chemotherapy (Spodnick
et al., 1992; Lane et al., 1994). Cats usually succumb to
systemic effects of the disease (Spodnick et al., 1992;
Lane et al., 1994).

1. How are the sympathetic nerves that arise from the
brachial plexus best assessed for neoplastic involvement
and what is the best way to dissect them out?
2. Should dogs with nerve sheath tumors that are resected
with complete histological margins undergo postoperative
radiation therapy?
3. Is a bone allograft preferable to a cement spacer following
vertebrectomy?
4. Will radiation therapy adversely affect metal implant and
bone cement stabilization after vertebrectomy?
5. Is spondylectomy without vertebral replacement
preferable to replacement of the resected vertebral body
with an allograft or cement spacer?
One cat with meningioma survived almost 4 years;

median survival for four other cats was 6 months (Levy
et al., 1997b). One cat with myeloma survived 3 months
on prednisone therapy (Mills et al., 1982). Complete
excision of a nerve sheath tumor resulted in a 6-year
survival in one cat. Vertebral osteosarcoma may be less
aggressive in cats than in dogs and a survival time of 4.5
years was reported in one cat even after incomplete
excision. Non-lymphoid tumors in cats may therefore
have a better prognosis than vertebral canal lymphoma
(Levy et al., 1997b).
REFERENCES
Ashizawa, R., Ohtsuka, K., Kamimura, M., Ebara, S., Takaoka, K. (1999)
Percutaneous transpedicular biopsy of thoracic and lumbar vertebrae
method and diagnostic validity. Surgical Neurology 52, 545551.
Asperio, R.M., Marzola, P., Zibellini, E., Villa, W., Sbarbati, A., Osculati, F.,
Addis, F. (1999) Use of magnetic resonance imaging for diagnosis of a
spinal tumor in a cat. Veterinary Radiology and Ultrasound 40, 267270.
Ayala, A.G., Zornosa, J. (1983) Primary bone tumors: percutaneous needle
biopsy. Radiologicpathologic study of 222 biopsies. Radiology 149,
675679.
Bailey, C.S. (1990) Long-term survival after surgical excision of a schwannoma of the sixth cervical spinal nerve in a dog. Journal of the American
Baumgartner, W., De Vargas Peixoto, P.F. (1988) Spinal nephroblastoma
(neuroepithelioma) in the dog. Kleintierpraxis 33, 7576, 78.
Bell, F.W., Feeney, D.A., Caywood, D.D., Hultgren, B.D. (1992) External
beam radiation therapy for recurrent intraspinal meningioma in a dog.
Bentley, J.F., Simpson, S.T., Hathcock, J.T., Brawner, W.R., Powers, R.D.
(1990) Metastatic thyroid solid-follicular carcinoma in the cervical portion of the spine of a dog. Journal of the American Veterinary Medical
Berg, J., Lamb, C. (1990) Bone scintigraphy in the initial evaluation of dogs
with primary bone tumors. Journal of the American Veterinary Medical
after cervical vertebral interbody fusion with polymethyl methacrylat
Bradley, R.L., Withrow, S.J., Snyder, S.P. (1982) Nerve sheath tumors in the
Bradney, I.W., Forsyth, W.M. (1986) A Schwannoma causing cervical spinal
cord compression in a dog. Australian Veterinary Journal 63, 374375.
Brasmer, T.H., Lumb, W.V. (1972) Lumbar vertebral prosthesis in the dog.
Brehm, D.M., Vite, C.H., Steinberg, H.S., Haviland, J., Van Winkle, T.
(1995) A retrospective evaluation of 51 cases of peripheral nerve sheath
tumors in the dog. Journal of the American Animal Hospital Association
31, 349359.
Britt, J.O., Jr, Simpson, J.G., Howard, E.B. (1984) Malignant lymphoma of
the meninges in two dogs. Journal of Comparative Pathology 94, 4553.
Callanan, J.J., Jones, B.A., Irvine, J., Willett, B.J., McCandlish, I.A.,
Jarrett, O. (1996) Histologic classification and immunophenotype of
lymphosarcomas in cats with naturally and experimentally acquired feline
immunodeficiency virus infections. Veterinary Pathology 33, 264272.
Carmichael, S., Griffiths, I.R. (1981) Tumours involving the brachial
plexus in seven dogs. Veterinary Record 108, 435437.
Chauvet, A., Hogge, G., Sandin, J., Lipsitz, D. (1999) Vertebrectomy,
bone allograft fusion, and antitumor vaccination for the treatment of vertebral fibrosarcoma in a dog. Veterinary Surgery 28, 480488.
Ciekot, P.A., Powers, B.E., Withrow, S.J., Straw, R.C., Ogilvie, G.K., LaRue,
S.M. (1994) Histologically low-grade, yet biologically high-grade, fibrosarcomas of the mandible and maxilla in dogs: 25 cases (19821991). Journal
of the American Veterinary Medical Association 204, 610615.
263
264
Couto, C.G., Cullen, J., Pedroia, V., Turrel, J.M. (1984) Central nervous
system lymphosarcoma in the dog. Journal of the American Veterinary
Cristante, L., Herrmann, H.D. (1994) Surgical management of
intramedullary spinal cord tumors: functional outcome and sources of
morbidity. Neurosurgery 35, 6974; discussion 7476.
Cuddon, P.A., Murray, M., Kraus, K. (2003) Electrodiagnosis In: D. Slatter
Dernell, W., Van, V.B., Straw, R., LaRue, S., Powers, B., Withrow, S. (2000)
Outcome following treatment of vertebral tumors in 20 dogs (19861995).
Dobson, J.M., Blackwood, L.B., McInnes, E.F., Bostock, D.E., Nicholls, P.,
Hoather, T.M., Tom, B.D.M. (2001) Prognostic variables in canine multicentric lymphosarcoma. Journal of Small Animal Practice 42, 377384.
Donner, T.R., Voorhies, R.M., Kline, D.G. (1994) Neural sheath tumors of
major nerves. Journal of Neurosurgery 81, 362373.
Fan, T.M., Kitchell, B.E. (2000) Lomustine. Compendium on Continuing
Fernandez, F.R., Grindem, C.B., Brown, T.T., Jr, Sharp, N.J.H., Saulnier, M.
(1997) Cytologic and histologic features of a poorly differentiated glioma
in a dog. Veterinary Clinical Pathology 26, 182186.
Ferretti, A., Scanziani, E., Colombo, S. (1993) Surgical treatment of a
spinal cord tumor resembling nephroblastoma in a young dog. Progress in
Findlay, G.F. (1984) Adverse effects of the management of malignant
spinal cord compression. Journal of Neurology, Neurosurgery and
Findlay, G.F. (1987) The role of vertebral body collapse in the management of malignant spinal cord compression. Journal of Neurology,
Neurosurgery and Psychiatry 50, 151154.
Fingeroth, J.M., Smeak, D.D. (1989) Laminotomy of the axis for surgical
access to the cervical spinal cord. Veterinary Surgery 18, 123129.
Fingeroth, J.M., Prata, R.G., Patnaik, A.K. (1987) Spinal meningiomas in
dogs: 13 cases (19721987). Journal of the American Veterinary Medical
Ganju, A., Roosen, N., Kline, D.G., Tiel, R.L. (2001) Outcomes in a consecutive series of 111 surgically treated plexal tumors: a review of the
experience at the Louisiana State University Health Sciences Center.
Gilson, S.D. (2003) Neuro-oncologic surgery In: D. Slatter (ed.), Textbook of
Small Animal Surgery, 3rd edn, 12771286. Philadelphia: Elsevier Science.
Gilson, S.D., Stone, E.A. (1990) Surgically induced tumour seeding in
eight dogs and two cats. Journal of the American Veterinary Medical
Grem, J.L., Burgess, J., Trump, D.L. (1985) Clinical features and natural
history of intramedullary spinal cord metastasis. Cancer 56, 23052314.
Heyman, S.J., Diefenderfer, D.L., Goldsschmidt, M.H., Newton, C.D.
(1992) Canine axial skeletal osteosarcoma: a retrospective study of 116
cases (1986 to 1989). Veterinary Surgery 21, 304310.
Hill, T.P., Lubbe, A.M., Guthrie, A.J. (2000) Lumbar spine stability following hemilaminectomy, pediculectomy, and fenestration. Veterinary
Hung, L.C., Pong, V.F., Cheng, C.R., Wong, F.I., Chu, R.M. (2000) An
improved system for quantifying AgNOR and PCNA in canine tumors.
Anticancer Research 20, 32733280.
Irving, G., McMillan, M.C. (1990) Fluoroscopically guided percutaneous
fine-needle aspiration biopsy of thoracolumbar spinal lesions in cats.
Progress in Veterinary Neurology 1, 473475.
Jang, J.S., Lee, S.H., Rhee, C.H. (2002) Polymethylmethacrylateaugmented screw fixation for stabilization in metastatic spinal tumors.
Technical note. Journal of Neurosurgery 96, 131134.
Jeffery, N., Brearley, M.J. (1993) Brain tumours in the dog: treatment of
10 cases and review of recent literature. Journal of Small Animal
Jeffery, N.D. (1991) Treatment of epidural haemangiosarcoma in a dog.
Jeffery, N.D., Phillips, S.M. (1995) Surgical treatment of intramedullary spinal
cord neoplasia in two dogs. Journal of Small Animal Practice 36, 553557.
Jones, B.R., Alley, M.R., Johnstone, A.C., Jones, J.M., Cahill, J.I.,
McPherson, C. (1995) Nerve sheath tumours in the dog and cat. New
Zealand Veterinary Journal 43, 190196.
Jonsson, B., Jonsson, H., Jr., Karlstrom, G., Sjostrom, L. (1994) Surgery of
cervical spine metastases: a retrospective study. European Spine Journal
3, 7683.
Kippenes, H., Gavin, P.R., Bagley, R.S., Silver, G.M., Tucker, R.L., Sande,
R.D. (1999) Magnetic resonance imaging features of tumors of the spine
and spinal cord in dogs. Veterinary Radiology and Ultrasound 40, 627633.
Kirpensteijn, J., Kik, M., Rutteman, G.R., Teske, E. (2002) Prognostic significance of a new histologic grading system for canine osteosarcoma.
Veterinary Pathology 39, 240246.
Kiupel, M., Teske, E., Bostock, D. (1999) Prognostic factors for treated
canine malignant lymphoma. Veterinary Pathology 36, 292300.
Klopp, L.S., Quinn, M.K. (2001) En-Bloc excision and comparative
surgical staging of dorsal compartment vertebral tumors. Veterinary
Surgery 30, 497.
Knecht, C.D., Greene, J.A. (1977) Surgical approach to the brachial plexus
in small animals. Journal of the American Animal Hospital Association
13, 592594.
Kuntz, C.A., Dernell, W.S., Powers, B.E., Devitt, C., Straw, R.C.,
Withrow, S.J. (1997) Prognostic factors for surgical treatment of softtissue sarcomas in dogs: 75 cases (19861996). Journal of the American
Lane, S.F., Kornegay, J.N., Duncan, J.R., Oliver, J.E. (1994) Feline spinal
lymphosarcoma: A retrospective evaluation of 23 cats. Journal of Veterinary Internal Medicine 8, 99104.
LeCouteur, R.A. (2001) Tumors of the nervous system. In: S.J. Withrow
and E.G. MacEwan (eds) Small Animal Clinical Oncology, 3rd edn,
514531. Philadelphia: WB Saunders.
LeCouteur, R.A., Child, G. (1995) Diseases of the spinal cord. In:
S.J. Ettinger (ed.) Textbook of Veterinary Internal Medicine, 629695.
Levitski, R., Lipsitz, D., Chauvet, A. (1999) Magnetic resonance imaging of the
cervical spine in 27 dogs. Veterinary Radiology and Ultrasound 40, 332341.
Levy, M.S., Kapatkin, A.S., Patnaik, A.K., Mauldin, G.N., Mauldin, G.E.
(1997a) Spinal tumors in 37 dogs: clinical outcome and long-term survival (19871994). Journal of the American Animal Hospital Association
33, 307312.
Levy, M.S., Mauldin, G., Kapatkin, A.S., Patnaik, A.K. (1997b) Nonlymphoid vertebral canal tumors in cats: 11 cases (19871995). Journal
Lipsitz, D., Bailey, C.S. (1995) Clinical use of the lateral cervical approach
for cervical spinal cord and nerve root disease: eight cases. Progress in
Lu, W.W., Cheung, K.M., Li, Y.W., Luk, K.D., Holmes, A.D., Zhu, Q.A.,
Leong, J.C. (2001) Bioactive bone cement as a principal fixture for
spinal burst fracture: an in vitro biomechanical and morphologic study.
Spine 26, 26842690; discussion 26902691.
Mason, W.P., Gentili, F., Macdonald, D.R., Hariharan, S., Cruz, C.R.,
Abrey, L.E. (2002) Stabilization of disease progression by hydroxyurea in
patients with recurrent or unresectable meningioma. Journal of
Neurosurgery 97, 341346.
Matsui, H., Tatezaki, S., Tsuji, H. (1994) Ceramic vertebral body replacement for metastatic spine tumors. Journal of Spinal Disorders 7, 248254.
McAfee, P.C., Bohlman, H.H., Ducker, T., Eismont, F.J. (1986) Failure of
stabilization of the spine with methylmethacrylate. A retrospective
analysis of twenty-four cases. The Journal of Bone and Joint Surgery
American volume 68, 11451157.
McCarthy, R.J., Feeney, D.A., Lipowitz, A.J. (1993) Preoperative diagnosis
of tumors of the brachial plexus by use of computed tomography in
three dogs. Journal of the American Veterinary Medical Association 202,
291294.
McDonnell, J.J., Platt, S.R., Clayton, L.A. (2001) Neurologic conditions
causing lameness in companion animals. Veterinary Clinics of North
Miller, D.J., Lang, F.F., Walsh, G.L., Abi-Said, D., Wildrick, D.M.,
Gokaslan, Z.L. (2000) Coaxial double-lumen methylmethacrylate
reconstruction in the anterior cervical and upper thoracic spine after
tumor resection. Journal of Neurosurgery 92 (2 Suppl), 181190.
Mills, J.N., Eger, C.E., Robinson, W.F., Penhale, W.J., McKenna, R.P.
(1982) A case of multiple myeloma in a cat. Journal of the American
Moissonier, P., Crigel, M.H., Timmerman, P. (2002a) Spinal tumors in
dogs: Retrospective study of 18 cases. Veterinary Surgery 31, 301.
Neoplasia
Moissonnier, P., Carozzo, C., Meheust, P. (2002b) Lateral corpectomy as a

treatment of chronic disc herniation in 15 dogs. Veterinary Surgery 31, 296.
Moore, A.S., London, C.A., Wood, C.A., Williams, L.E., Cotter, S.M.,
DA, L.H., Frimberger, A.E. (1999) Lomustine (CCNU) for the treatment
of resistant lymphoma in dogs. Journal of Veterinary Internal Medicine
13, 395398.
Moore, G., Mathey, W., Eggers, J., Estep, J. (2000) Osteosarcoma in adjacent
lumbar vertebrae in a dog. Journal of the American Veterinary Medical
Moore, L.E., Garrett, L.D., Debey, B., Biller, D.S. (2002) Spinal mast cell
tumor in a dog. Journal of the American Animal Hospital Association 38,
6770.
Moore, P.F., Rosin, A. (1986) Malignant histiocytosis of Bernese mountain
dogs. Veterinary Pathology 23, 110.
Mork, S.J., Loken, A.C. (1977) Ependymoma: a follow-up study of 101
cases. Cancer 40, 907915.
Moulopoulos, L.A., Dimopoulos, M.A., Weber, D., Fuller, L., Libshitz, H.I.,
Alexanian, R. (1993) Magnetic resonance imaging in the staging of solitary plasmacytoma of bone. Journal of Clinical Oncology 11, 13111315.
Nakayama, M. (1993) Intraoperative spinal ultrasonography in dogs: normal
findings and case-history reports. Veterinary Radiology and Ultrasound
34, 264268.
Niles, J.D., Dyce, J., Mattoon, J.S. (2001) Computed tomography for the
diagnosis of a lumbosacral nerve sheath tumour and management by
hemipelvectomy. Journal of Small Animal Practice 42, 248252.
Noonan, M., Kline, K., Meleo, K. (1997) Lymphoma of the central nervous
system: a retrospective study of 18 cats. Compendium on Continuing
Oakley, R.E., Patterson, J.S. (2003) Tumors of the central and peripheral
nervous system In: D. Slatter (ed.) Textbook of Small Animal Surgery,
3rd edn, 24052424. Philadelphia: Elsevier Science.
OFlanagan, S.J., Stack, J.P., McGee, H.M., Dervan, P., Hurson, B. (1991)
Imaging of intramedullary tumour spread in osteosarcoma. A comparison of techniques. The Journal of Bone and Joint SurgeryBritish volume
73, 9981001.
Oliver, J.E., Eubank, N.H., Geary, J.C. (1965) Neurofibrosarcoma in a dog.
Onimus, M., Papin, P., Gangloff, S. (1996) Results of surgical treatment
of spinal thoracic and lumbar metastases. European Spine Journal
5, 407411.
13511353.
Pearson, G.R., Gregory, S.P., Charles, A.K. (1997) Immunohistochemical
demonstration of Wilms tumour gene product WT1 in a canine neuroepithelioma providing evidence for its classification as an extrarenal
nephroblastoma. Journal of Comparative Pathology 116, 321327.
Pietila, T.A., Stendel, R., Schilling, A., Krznaric, I., Brock, M. (2000)
Surgical treatment of spinal hemangioblastomas. Acta Neurochirurgica
( Wein) 142, 879886.
Platt, S.R., Sheppard, B.J., Graham, J., Uhl, E.W., Meeks, J., Clemmons, R.M.
(1998) Pheochromocytoma in the vertebral canal of two dogs. Journal of
Platt, S.R., Graham, J., Chrisman, C.L., Collins, K., Sundeep, C.,
Sirninger, J., Newell, S.M., Chandra, S. (1999) Magnetic resonance
imaging and ultrasonography in the diagnosis of a malignant peripheral
nerve sheath tumor in a dog. Veterinary Radiology and Ultrasound
40, 367371.
Poncelet, L., Snaps, F., Jacobovitz, D., Pippendop, B., Balligand, M. (1994)
Successful removal of an intramedullary spinal paraganglioma from a
Powers, B.E., Beck, E.R., Gillette, E.L., Gould, D.H., LeCouter, R.A. (1992)
Pathology of radiation injury to the canine spinal cord. International
Journal of Radiation Oncology Biology Physics 23, 539549.
Prata, R.G. (1977) Diagnosis of spinal cord tumors in the dog. Veterinary
Ribas, J.L., Carpenter, J.L., Mena, H., Schulman, F.Y., Lee, K. (1991)
Central nervous system meningioma in the dog: a review of 50 cases.
Journal of Neuropathology and Experimental Neurology 50, 373.
Ritter, A.M., Hess, K.R., McLendon, R.E., Langford, L.A. (1998)
Ependymomas: MIB-1 proliferation index and survival. Journal of
Neurooncology 40, 5157.
Rosin, A. (1982) Neurologic disease associated with lymphosarcoma in ten
dogs. Journal of the American Veterinary Medical Association 181, 5053.
Rusbridge, C., Wheeler, S.J., Lamb, C.R., Page, R.L., Carmichael, S.,
Brearley, M.J., Bjornson, A.P. (1999) Vertebral plasma cell tumors in 8
Sakai, F., Sone, S., Kiyono, K., Maruyama, Y., Oguchi, K., Imai, N., Li, F.,
Matsubara, M., Ueda, H., Haniuda, M., Kubo, K., Honda, T., Ishii, K.
(1996) Magnetic resonance imaging of neurogenic tumors of the thoracic
inlet: determination of the parent nerve. Journal of Thoracic Imaging 11,
272278.
Santen, D.R., Payne, J.T., Pace, L.W., Kroll, R.A., Johnson, G.C. (1991)
Thoracolumbar vertebral osteochondroma in a young dog. Journal of the
Schiff, D., ONeill, B.P., Suman, V.J. (1997) Spinal epidural metastasis as
the initial manifestation of malignancy: clinical features and diagnostic
approach. Neurology 49, 452456.
Schueler, R.O., Roush, J.K., Oyster, R.A. (1993) Spinal ganglioneuroma in a
dog. Journal of the American Veterinary Medical Association 203, 539541.
Schulz, K.S., Chickering, W.R., Shell, L., Doherty, M.A., Moon, N., Shires, P.
(1994) Fluoroscopicguided and direct spinal cord aspiration in the normal
dog. Progress in Veterinary Neurology 5, 4648.
Sharp, N.J.H. (1988) Craniolateral approach to the canine brachial plexus.
Shell, L., Dallman, M.J., Sponenberg, P. (1987) Chondrosarcoma in a cat
presenting with forelimb monoparalysis. Compendium on Continuing
Siegel, S., Kornegay, J.N., Thrall, D.E. (1996) Postoperative irradiation of spinal
cord tumors in 9 dogs. Veterinary Radiology and Ultrasound 37, 150153.
Simonovsky, V. (1997) Peripheral nerve schwannoma preoperatively diagnosed by sonography: report of three cases and discussion. European
Simpson, D.J., Beck, J.A., Allan, G.S., Culvenor, J.A. (1999) Diagnosis
and excision of a brachial plexus nerve sheath tumour in a dog.
Australian Veterinary Journal 77, 222223.
Smith, M.O., LaRue, S.M. (2000) Lesssons learned from 10 years of irradiating brain tumors. American College of Veterinary Internal Medicine
Annual Conference, Seattle, WA, 2000; 305307.
Spodnick, G.J., Berg, J., Moore, F.M., Cotter, S.M. (1992) Spinal lymphoma
in cats: 21 cases (19761989). Journal of the American Veterinary Medical
Steinberg, H.S. (1988) Brachial plexus injuries and dysfunctions. Veterinary
Summers, B.A., de Lahunta, A., McEntee, M., Kuhajda, F.P. (1988) A novel
intradural extramedullary spinal cord tumor in young dogs. Acta
Neuropathologica 75, 402410.
Summers, B.A., Cummings, J.F., de Lahunta, A. (1994) Veterinary
Neuropathology. St Louis: Mosby.
Sundaresan, N., Galicich, J.H., Lane, J.M., Bains, M.S., McCormack, P.
(1985) Treatment of neoplastic epidural cord compression by vertebral
body resection and stabilization. Journal of Neurosurgery 63, 676684.
Sundaresan, N., Schmidek, H.H., Schiller, A.L. (1990) Tumors of the spine:
Diagnosis and Clinical Management. Philadelphia: WB Saunders.
Targett, M.P., Dyce, J., Houlton, J.E.F. (1993) Tumours involving the nerve
sheaths of the forelimb in dogs. Journal of Small Animal Practice 34,
221225.
Terrell, S.P., Platt, S.R., Chrisman, C.L., Homer, B.L., de Lahunta, A.,
Summers, B.A. (2000) Possible intraspinal metastasis of a canine spinal
cord nephroblastoma. Veterinary Pathology 37, 9497.
Theon, A.P., LeCouteur, R.A., Carr, E.A., Griffey, S.M. (2000) Influence
of tumor cell proliferation and sex-hormone receptors on effectivenessof
radiation therapy for dogs with incompletely resected menin-giomas.
265
266
Tiller-Borcich, J.K., Fike, J.R., Phillips, T.L., Davis, R.L. (1987) Pathology
of delayed radiation brain damage: an experimental canine model.
Radiation Research 110, 161172.
Tomlinson, J., Higgins, R.J., LeCouteur, R.A., Knapp, D. (1988)
Intraspinal epidermoid cyst in a dog. Journal of the American Veterinary
Turner, J.L., Luttgen, P.J., VanGundy, T.E., Roenigk, W.J., Hightower, D.,
Frelier, P.F. (1992) Multicentric osseous lymphoma with spinal
extradural involvement in a dog. Journal of the American Veterinary
Tyrrell, D., Davis, R.M. (2001) Progressive neurological signs associated with
systemic mastocytosis in a dog. Australian Veterinary Journal 79, 106108.
Uchida, K., Morozumi, M., Yamaguchi, R., Tateyama, S. (2001) Diffuse
leptomeningeal malignant histiocytosis in the brain and spinal cord of a
Tibetan Terrier. Veterinary Pathology 38, 219222.
Viguier, E., Petit-Etienne, G., Magnier, J., Diop, A., Lavaste, F. (2002)
Mobility of T13-L1 after spinal cord decompression procedures in dogs
(an in vitro study). Veterinary Surgery 31, 297.
Walker, T.M., Pierce, W.A., Welch, R.D. (2002) External fixation of the
lumbar spine in a canine model. Veterinary Surgery 31, 181188.
Waters, D., Hayden, D. (1990) Intramedullary spinal cord metastasis in

the dog. Journal of Veterinary Internal Medicine 4, 207215.
Wheeler, S.J. (1989) Spinal tumors in cats. Veterinary Annual 29, 270277.
Wheeler, S.J., Clayton-Jones, D.G., Wright, J.A. (1986) The diagnosis of
brachial plexus disorders in dogs: a review of twenty-two cases. Journal
Withrow, S.J., MacEwan, E.G. (2001) Small Animal Clinical Oncology,
3rd edn. Philadelphia: WB Saunders.
Yeh, L.S., Hou, S.M. (1994) Repair of a mandibular defect with a free vascularized coccygeal vertebra transfer in a dog. Veterinary Surgery 23,
281285.
Yerby, S.A., Toh, E., McLain, R.F. (1998) Revision of failed pedicle screws
using hydroxyapatite cement. A biomechanical analysis. Spine 23,
16571661.
Yturraspe, D., Lumb, W., Young, S., Gorman, H. (1975) Neurological and
pathological effects of second lumbar spondylectomy and spinal column
shortening in the dog. Journal of Neurosurgery 42, 4758.
Yturraspe, D.J., Lumb, W.V. (1973) Second lumbar spondylectomy and
shortening of the spinal column of the dog. American Journal of
PROCEDURES
Hemilaminectomy (12.1512.29)
Cervical hemilaminectomy is described below. The ventral slot approach is rarely useful for spinal tumor exploration as
access to the spinal cord and nerve roots is severely limited. Thoracolumbar hemilaminectomy is described in 8.118.37.
12.15
12.15 Patient positioning for

cervical hemilaminectomy
via a dorsolateral approach.
12.16 To approach the midcervical vertebrae, the

incision can extend from the
occipital protuberance to
the spinous process of T2.
The skin incision is to one
side of the midline, which
reduces any tendency for
wound breakdown by
avoiding tension over the
12.16
spinous processes. Good
wound healing is important
in case postoperative radiation therapy is planned. Good hemostasis is also vital; any seroma or
hematoma formation should be attended to carefully (see 15.34).
Neoplasia
12.17 The skin and superficial fascia are incised to

reveal the superficial muscles. Note the dorsal
branches of the cervical nerves emerging in the
midline and diverging laterally. Each is usually
combined with an artery and vein, which must
be ligated or cauterized as necessary. Further
details of anatomy are covered in Chapters 7, 9
and 11.
12.17
12.18 The incision is continued in the midline through

the muscular aponeurosis. The nuchal ligament
is exposed. This may be retracted away from
the surgeon, transected or divided in the
midline.
12.18
12.19 The nuchal ligament has been divided in the

midline as have the spinalis muscles. This
reveals the muscular attachments to the spinous
process of the cervical vertebra (arrow).
12.19
12.20 The spinalis muscles have

been elevated from the
spinous processes (a) and
vertebral laminae (b) on the
side of the spine to be
approached.
12.20
267
268
12.21 The paraspinal muscles are elevated first from

the articular process (a) and then from further
ventrally. Branches of the vertebral artery
emerge here and must be ligated if
encountered. Bleeding from spinal veins is best
occluded with a Hemoclip (Pilling Weck Inc.,
Research Triangle Park, NC) (5.31, 11.51). The
spinous process (b) and lamina (c) are seen
here. (1.16).
b
c
a
12.21
12.22 The articular process is removed with

rongeurs. The laminectomy is commenced
with a bur. Here the cartilage of the articular
surfaces can be seen (arrow).
12.22
12.23 Site of hemilaminectomy.

Here the bone has been
removed in the cranialmost vertebra. The site of
bone removal is shown in
the caudal-most vertebra.
12.23
Neoplasia
12.24 Bone is removed until the vertebral canal is

entered. Here the inner cortical bone is visible, and
ligamentum flavum is exposed in the craniodorsal
corner (arrow). The bone will be burred to eggshell
thickness over the entire defect before the vertebral
canal is entered (8.348.36, 10.34). The inner
portion of the joint capsule is visible in the center of
the defect (a).
12.24
12.25 The laminectomy is continued to reveal the spinal

cord, nerve root (a) and spinal ganglion (b). Large
branches of the vertebral artery and vein are
present at this level; any large tear in a vein is best
closed using a Hemoclip (Pilling Weck Inc.,
Research Triangle Park, NC) (5.31). Hemorrhage
may also occur from the vertebral plexus. This is
best controlled with Gelfoam (Pharmacia,
Kalamazoo, MI), direct pressure, or, as in this case,
a piece of macerated muscle (c).
a
c
b
12.25
12.26 The laminectomy is continued

into the intervertebral foramen
to reveal the nerve root and
spinal ganglion. The
laminectomy may be extended
as appropriate for removal of
neoplasms (7.567.59, 11.54).
12.26
12.27 Myelogram of a dog that

presented with a 3-month
history of neck pain and
right-sided weakness.
An extradural mass is
compressing the spinal
cord at the level of C1
(arrowheads).
12.27
A
269
270
12.28 CT scan without

subarachnoid contrast,
same dog as in 12.27.
A: Mineralized mass to one
side of the vertebral canal
prior to surgery (arrowhead).
B: After surgery the mass is
12.28
still visible along with the
laminectomy defect in C1.
A
B
Complete resection was not
possible as the mass had a consistency similar to bone. Histological diagnosis was meningioma; tissue
consisted of cartilage, mature bone with bone marrow and a few scattered fusiform cells.
12.29 3D reconstruction of the dorsolateral

hemilaminectomy (arrows) performed over the
right side of C1 and the cranial aspect of C2.
The residual mass of the meningioma is evident
within the vertebral canal (arrowhead).
12.29

Cervical dorsal laminectomy is described in 11.4411.55. Thoracolumbar dorsal laminectomy is described below.
Dorsal laminectomy is the approach of choice for most spinal tumors, although hemilaminectomy may be suitable in
some (12.29, 12.40). Good access is often essential for complete tumor resection and to permit the extensive removal
of surrounding dura necessary with meningioma. Access to ventrally located masses may be somewhat restricted and
use of dural stay sutures (12.40, 14.514.8) or rhizotomy (8.51) may be necessary (Fingeroth et al., 1987). Partial
corpectomy of the vertebral body adjacent to the mass (with occlusion of the vertebral venous plexus using bone wax)
may improve access further (8.47B) (Moissonnier et al., 2002b). Dorsal laminectomy with removal of an articular
process does cause significant instability (Smith and Walter, 1988; Shires et al., 1991) (12.39).
12.30
12.30 Positioning of dog for

thoracolumbar dorsal
laminectomy.
Neoplasia
12.31 The skin incision is made

1 cm to one side of the
midline. The precise site of
the incision is governed by
the location of the lesion.
12.31
12.32 The skin is incised just off the midline, at least

two vertebrae on each side of the vertebra to
be approached. The superficial fascia is
mobilized and retracted to reveal the
lumbodorsal fascia. Here the fascia has been
incised bilaterally to reveal the spinous
processes.
12.33 Exposure of the vertebral bodies is done

bilaterally using the methods described under
thoracolumbar hemilaminectomy (8.128.27).
The multifidus muscles have been elevated from
the spinous processes (a) and the vertebral
arch. This reveals the articular processes (b),
and the longissimus tendon attaching to the
accessory process (c). The muscle retraction is
maintained with Gelpi retractors. The dissection
is completed on one side before starting the
other side. The spinous processes are then
removed with rongeurs or bone cutters.
12.32
a
b
c
12.33
12.34 The spinous processes have been removed.

The articular processes are still present (a)
and the ligamentum flavum is visible between
the vertebrae (b).
12.34
271
272
12.35 The laminectomy is commenced in the midline

with a bur, preserving the articular processes at
this stage. Here the cortical bone has been
removed over two vertebrae, revealing dark
cancellous bone. See Chapter 8 for details on
technique for using the bur (8.34).
12.35
12.36 The cancellous bone has been removed with the

bur to reveal the white, inner cortical bone.
This is then thinned to eggshell thickness over
the entire defect (8.338.36).
12.36
12.37 The final layer of bone is removed to reveal the

spinal cord (arrow). This is achieved with fine
rongeurs and a curette. This degree of exposure
may be adequate for some circumstances.
12.37
12.38 The approach can be modified depending on

requirements. Wide exposure of the tumor and
spinal cord is desirable. If the spinal cord is
swollen, care should be taken to remove
enough bone to prevent injury by impingement
of the bone edges (Fingeroth et al., 1987). Here
the pedicle is removed between the articular
processes to expose the lateral aspect of the
spinal cord (arrow). The articular processes are
preserved if at all possible.
12.38
Neoplasia
12.39 Here the left articular processes have been

removed as have the pedicles of two vertebrae.
This gives good access to the nerve root (a).
Note the vertebral plexus lying on the floor of
the vertebral canal (arrows); this should be
avoided. Constrictive fibrosis is a potential risk
with this degree of bone removal. Subsequent
stabilization is advisable (Hill et al., 2000; Viguier
et al., 2002; Gilson, 2003).
12.40 Nephroblastoma in a young dog with acute

paraparesis. A: The dura has been opened and
the edges retracted using silk stay sutures. The
spinal cord, covered by nerve rootlets, has
herniated through the durotomy. It is important
to make a long enough laminectomy and
durotomy to prevent distortion of spinal cord
under pressure. B: Gentle dissection has teased
out a large mass (arrows). One rootlet was
enveloped by the mass but there was no
attachment to the spinal cord. The dog
underwent radiation therapy. It improved
substantially but was then lost to follow-up.
273
Trauma
Chapter
13
Initial assessment 281

Other injuries 281
Cardiothoracic 282
Urinary tract 282
Other 282
Box 13.1 Differential diagnosis of spinal trauma
Neurological examination
Dural tear
Ischemic neuromyopathy (cats and dogs, see Chapter 14)
Traumatic feline ischemic myelopathy (see Chapter 14)
Fibrocartilaginous embolus
Psoas muscle injury
Bilateral pelvic/long bone fracture or cruciate rupture
282
Radiography 283
Survey radiography 283
Additional imaging 283
Pathophysiology 286
Spinal fracture biomechanics 286
Biomechanics of fixation devices 289
Treatment 293
Choice of therapy 293
Anatomical location of the injury
Postoperative care
295
297
Complications 297
Feline spinal injuries
Prognosis
301
301

References
303
Procedures 305
Non-surgical 305
Metal and bone cement 309
Modified segmental fixation 315
External fixation 316
302
Congenital atlantoaxial instability
Degenerative disc disease
Pathological fracture, neoplasia
Trauma can damage the vertebra(e), disc(s), meninges,

spinal cord, or any combination of these. Diagnosis is by
history and physical findings in most dogs, but may be
less obvious in cats (Box 13.1).
After initial assessment of a spinal trauma patient,
consider methylprednisolone sodium succinate (MPSS)
if within 8 h of the injury (see page 83). General principles of treatment can be summarized as follows:
Establish a prognosis and stabilize other injuries.
Provide analgesia and consider whether or not to
use MPSS (see page 83).
Reduce vertebral misalignment to relieve spinal
cord compression.
Provide stability by cage confinement, an external
splint, or an internal or external fixator.
Decompression by laminectomy is not usually
necessary when vertebral alignment is good. It is
recommended for marked extradural compression
from hematoma, disc material or bone fragments.
INITIAL ASSESSMENT
Other injuries
The first priority is to treat shock and other lifethreatening disorders (13.1). A thorough and meticulous
282
reduce spinal cord perfusion (page 86), complicate

anesthesia and may also be fatal (Murtaugh and Ross,
1988; Snyder et al., 2001). Dyspnea is also common
after trauma and some animals may require ventilation
(Campbell and King, 2000; Beal et al., 2001).
Urinary tract
Bladder rupture usually causes discomfort on abdominal
palpation, vomiting, depression and hematuria within
2448 h. Damage to the more proximal urinary tract
may take longer to manifest itself (Weisse et al., 2002).
Pubic or perineal swelling indicates potential urethral
trauma.
13.1 This dog sustained spinal trauma 2 h previously. It has been
strapped to a rigid board for evaluation and transport. The head
can also be taped down for cervical injuries. Early immobilization
is very important as up to 50% of animals continue to deteriorate
after injury, particularly when referral is delayed (McKee, 1990;
Hawthorne et al., 1999). A muzzle would be a useful precaution.
Aggressive intravenous fluid therapy is in progress.
Box 13.2 Other injuries associated with spinal

trauma
Shock
Pulmonary or pleural lesions
Diaphragmatic hernia
Traumatic cardiomyopathy
Damaged urinary tract
Ruptured bile duct
Injury to other abdominal organs
Long bone or pelvic fractures
Multiple spinal injuries
Brachial or lumbosacral plexus injury
Soft tissue injuries
Head injuries
Other
Bile duct rupture, although rare, can result in patient
deterioration up to 15 days after injury (Neer, 1992)
(2.1). As some injuries are not apparent immediately,
the owner should be forewarned of the potential need
for further diagnostic evaluation and treatment (Lanz
et al., 2000).
After initial assessment (including nociceptive evaluationsee below) the patient is given analgesics.
Narcotic agonists are preferred unless there is concomitant hypoventilation; they should also be used with care
after head injury (Table 15.1). Analgesia must be combined with some form of stabilization, so that the animal does not injure itself further. If presented within
8 h of injury high-dose MPSS should be considered
although the available evidence for its use is not compelling (Hurlbert, 2000) (see page 83). Sedatives should
not be used routinely because they may decrease
protective muscle tone; adverse effects on the cardiovascular system may also exacerbate shock and compromise spinal cord blood flow (see page 86).
NEUROLOGICAL EXAMINATION
physical examination is essential at this stage as many
animals have non-neural injuries, which could be overlooked (Box 13.2). In one study, 20% of dogs with
injuries to lumbar vertebrae had concomitant pelvic
fractures; 33% had cardiopulmonary lesions (Turner,
1987). Brachial plexus injury must also be ruled out
(page 30, 2.15, 2.26, 13.1).
Cardiothoracic
Post-traumatic cardiac arrhythmias are common and
are easily overlooked as they can be delayed in onset by
up to 24 h. Continuous electrocardiogram (ECG) monitoring is recommended for 24 h or until any arrhythmia
resolves. Arrhythmias can decrease cardiac output,
The neurological examination is essential to localize the

deficit, to identify multiple lesions, and to determine the
prognosis. The single, most important prognostic factor
following spinal trauma is the presence or absence of
nociception or deep pain sensation (2.21, 13.36B). If
nociception is absent caudal to a traumatic lesion the
prognosis for return of neurological function is poor
(page 302) (Olby et al., 2003). Analgesics should be
given only after this parameter has been assessed as they
could alter the findings. Animals with cervical injuries
very rarely lose nociception (see page 28) but if severely
tetraparetic they must be assessed for hypoventilation
(Griffiths, 1970; Boudrieau, 1997) (6.1). The Schiff
Sherrington response has no prognostic value (see pages
26, 32, 13.7) although severe vertebral displacement
Trauma
does (Bagley, 2000; Papadopoulos et al., 2002) (13.4).

The animal must always be assessed for injury to the
head and brachial plexus (Griffiths et al., 1974; AANS,
2000; Platt et al., 2001) (see pages 29, 30). Screening for
brachial plexus avulsion injury is done by assessing
pupil diameter, cutaneous trunci reflex, thoracic limb
withdrawal and sensation to the digits (Griffiths et al.,
1974; Faissler et al., 2002) (2.20, 2.26).
RADIOGRAPHY
Survey radiography
Once the patient has been stabilized and a neuroanatomical diagnosis has been reached, survey radiographs are
taken. The clinician must remember that radiographs
are no substitute for the neurological examination; it is
not possible to estimate the neurological status from
the radiographs alone unless there is 100% vertebral displacement (McKee, 1990; Bagley, 2000) (13.6, 13.35).
Lateral views are taken first. Oblique views to assess the
articular facets may be taken with the animal in lateral
recumbency and the beam angled (13.2, 13.3),
although CT is more accurate (13.12, 13.16).
Ventrodorsal radiographs are best taken using a horizontal beam. Failing that, extreme care must be taken
when positioning the animal for ventrodorsal or
dorsoventral radiographs (4.38). Temporary stability
should be provided by a wood or Perspex splint.
Anesthesia or heavy sedation is undesirable prior to this
as it will reduce the stabilizing effect of paravertebral
muscle spasm (Blass et al., 1988).
Multiple fractures are not common except when
adjacent vertebrae are involved or with injuries caudal
to the lumbosacral junction (Feeney and Oliver, 1980;
Turner, 1987; McKee, 1990; Selcer et al., 1991;
Hawthorne et al., 1999). The neurological localization
and assessment for focal hyperesthesia should help to
rule out multiple injuries. As lower motor neuron (LMN)
lesions can mask upper motor neuron (UMN) deficits,
the T3L3 region should be surveyed when L4S3
deficits are present, likewise the C15 region with
C6T2 deficits (see page 29).
Additional imaging
Specific indications for spinal cord imaging by myelography, CT or MRI include:
Identification of additional lesions, such as when
survey radiographs do not correlate with the
neurological localization.
Identification of lesions that require removal such
as extradural bone fragments, blood clots or disc
material.
13.2 Diagram to illustrate assessment of the articular facets

with the animal in lateral recumbency and with the X-ray beam
angled. The facet joint closest to the film is highlighted with the
beam angled as in a; the other by the beam angle b.
13.3 This oblique radiograph reveals a fracture of the articular

facets (13.12).
These techniques require general anesthesia. Great

care must be taken when intubating animals with cervical
fractures and when positioning any animal with a fracture
that might be unstable (4.38). Myelography and MRI
often add significant time to the overall procedure. MRI
and CT have the advantage of being non-invasive. Of the
two, CT is faster but MRI provides better images of
extradural and intradural injuries (Fehlings et al., 1999)
(13.7). Additional imaging is not generally indicated in
an animal with good deep pain sensation that is to be
managed non-surgically.
MYELOGRAPHY
Advantages of myelography are that:
The entire spine can be evaluated easily.
It may rule out differential diagnoses (Box 13.1).
Dural tears (Hay and Muir, 2000) (13.10)
and spinal cord transection (13.4) can be
283
284
identified. The spinal cord may be transected

even if radiographs show little evidence of
subluxation as there could have been a transient
yet complete luxation at the time of the original
trauma (13.4).
Routine myelography is not recommended for surgical candidates that have only mild or moderate neurological deficits. These animals should make a very good
recovery provided that they do not deteriorate prior to
definitive stabilization. The benefit of identifying a significant extradural lesion in such animals is outweighed
by the risk of neurological deterioration secondary to
13.4 Myelographic appearance of a dog that has suffered an

anatomical spinal cord transection following vertebral luxation.
Note that there is now only minimal vertebral displacement.
the added manipulation and anesthetic time. In more

severely affected animals the benefits of invasive imaging
may outweigh the risks. However, care must be taken
not to manipulate the spine excessively during injection
of contrast. One important goal of managing animals
with spinal trauma is to prevent deterioration, which is
why non-invasive imaging by CT or MRI is preferred
(Bagley, 2000).
COMPUTED TOMOGRAPHY
Computed tomography should detect bone fragments
within the vertebral canal (13.5, 13.12, 13.16). It is
better than survey radiography at detecting articular
facet fractures and it also allows fracture stability to be
assessed using a three-compartment or column model
(Denis, 1984; Shires et al., 1991). In addition, images
can be used to make 2D and 3D reconstructions of the
lesion (13.5, 13.6).
After localizing the lesion, scans should be made
through the region of interest (see page 55). A CT can
aid the choice of entry points and trajectories if an
implant is to be placed in the vertebral bodies (13.49
13.57). The scan should identify bone fragments, mineralized disc and acute hemorrhage (Tidwell et al., 1994;
Lanz et al., 2000). Use of Hounsfield units (i.e. CT numbers, a measure of tissue attenuation properties) might
also distinguish spinal cord edema, which is potentially
13.5 A CT scan was used in this dog to look for bone fragments in the vertebral canal. A: Survey radiograph. Note the instability
evident between this and image B. B: Sagittal view of 3D reconstruction. A small fragment is visible on the floor of the canal but it is
not causing compression (arrowhead). C: Transverse view of 3D reconstruction. Fracture of the transverse process is visible. The
postoperative radiograph for this dog is shown in 13.24. A myelogram was not performed.
Trauma
reversible, from intramedullary hemorrhage, which carries a poor prognosis (Ramon et al., 1997). CT may
miss traumatic disc herniations although these are usually impact injuries with little residual mass effect. A
disadvantage of CT is that it may underestimate significantly the extent of spinal cord compression and so
should ideally be combined with an MRI (Tator et al.,
1999). Both CT and MRI still carry the risk of fracture
displacement during positioning. Animals with failure
of the disc or end plate (13.5) may be prone to hyperextension injury (4.38). A temporary (non-magnetic
for MRI, below) splint could be used to reduce the risk.
MR IMAGING
A mid-sagittal, T2-weighted MRI allows a large area of
the spinal cord to be surveyed for injury and can also
provide a quantitative assessment of compression. MR
may provide prognostic information by distinguishing

edema from hematoma. Hematoma often carries a
poor prognosis whereas edema has the potential to be
reversible (Ramon et al., 1997; Gopal and Jeffery, 2001)
(13.7). The following are prognostic indicators in humans
(Ramon et al., 1997; Selden et al., 1999):
Presence and extent of any intra-axial hematoma.
The extent of spinal cord edema.
The extent of spinal cord compression by
extra-axial hematoma.
Spinal cord swelling does not interfere with MRI as
it can for myelography (Gopal and Jeffery, 2001) (see
pages 57, 123, 8.1). The main problems with MRI are
availability and cost; greater scan times and slice thickness may also be factors. The ideal way to image human
spinal cord injury is using both MRI and a CT scan
(Fehlings et al., 1999; Tator et al., 1999).
13.6 A: Survey radiograph of a 5-year-old
Labrador that was hit by a car to show
100% displacement of L2 vertebra relative
to L3 in the vertical plane. B: 3D
reconstruction of a CT scan in the same
dog, which illustrates 100% displacement
in the lateral plane as well. The dog was
euthanized as there was no hope of a
functional recovery.
HL
13.7 This dog was paraplegic and incontinent with SchiffSherrington syndrome and UMN deficits 2 days after vehicular trauma.
A: 3-mm, sagittal, T2-weighted image. T13L1, L1L2 and L2L3 discs show subtle loss of signal and extend dorsally; there is
high signal within the spinal cord adjacent to these discs consistent with edema. B: 3-mm, axial, T2-weighted image at L1L2,
same dog. Spinal cord is displaced to the left but not compressed. A low signal, extra-dural mass thought to be a hematoma is
ventrolateral to the spinal cord (arrow). Diagnosis was traumatic disc herniation with focal spinal cord edema, malacia, or both (see
also Gopal and Jeffery, 2001). Treatment was non-surgical; 8 days after trauma the dog could stand and it was normal after 1 month.
285
286
PATHOPHYSIOLOGY
Huge forces are necessary to disrupt the vertebral column and so the spinal cord often sustains a much more
severe injury than it does after a simple disc extrusion
(Tator et al., 1999). Although secondary injury mechanisms are the same, the prognosis is much worse after
trauma than it is after a disc extrusion for animals that
have severe injuries (see page 302). In some animals the
bony elements are spared but catastrophic failure of
the disc can still cause serious spinal cord injury (traumatic disc herniation). The sudden trauma may tear the
anulus causing a previously healthy nucleus pulposus to
rupture explosively into the vertebral canal. This type
of extrusion does not produce the usual mass effect
associated with the mineralized material of degenerative
disc disease. In the cervical region the explosion of
nucleus pulposus is directed dorsolaterally by the dorsal longitudinal ligament and can cause devastating,
asymmetrical neurological deficits. Hemiplegia results
together with loss of sympathetic function and poor
nociception on the affected side (Griffiths, 1970).
Catastrophic disc failure in the thoracolumbar region
can also cause severe spinal cord injury and the sudden
change in subdural pressure may tear the dura mater
(1.21, 13.10). Severe spinal cord injuries with dural
tears can even occur after vigorous running or struggling without any vertebral injury (Hay and Muir,
2000; Yarrow and Jeffery, 2000).
Spinal fracture biomechanics

Forces that act on the vertebral column include bending
(dorsoventral and lateral), rotation, shear, and axial
loading. These can interact to cause a variety of injuries,
but regardless of the exact mechanism one of the
most important clinical questions is How stable is the
fracture/luxation site? Estimates of stability can be
made from knowledge of the elements in the spine that
are compromised. The simplest method is to classify
injuries into three types based on the anatomical compartment that has been compromised. Injuries may
damage the dorsal compartment, the ventral compartment or both (Smith and Walter, 1985) (13.8). Another
method to assess stability is to divide the spine into three
compartments; however, the exact elements of each
compartment vary depending on whether or not the soft
tissues are included (Denis, 1984; Shires et al., 1991).
A modification of these various schemes assesses the
ability of the vertebral column to resist forces applied to
it (13.913.16). This scheme assesses specifically the
integrity of the vertebral body, which acts as a buttress
to resist bending and axial loading, and of the articular
facets, which resist all forces (Smith and Walter, 1985;
Patterson and Smith, 1992). Articular facet integrity is
Dorsal
Ventral
13.8 Diagram to illustrate the dorsal and ventral compartments

of the vertebral column, as referred to in 13.913.16. The major
drawback of this method of classification is that the majority of
injuries damage both dorsal and ventral compartments (Smith
and Walter, 1985).
judged using lateral and oblique radiographs of the spine

(13.2, 13.3) or preferably by CT or MRI (13.12, 13.16).
The vertebral body is assessed readily by survey radiographs (13.14). The disc must also be assessed as it is the
single most important stabilizing factor against rotation
and lateral bending (Shires et al., 1991; Schulz et al.,
1996). The main advantage of this scheme is that it gives
some indication of stability and can serve as a guide to
the fixation procedure best suited to each injury (see
Biomechanics of fixation devices, page 289). The main
categories of injury are listed in Table 13.1:
Table 13.1 Main categories of injury

Failed component(s)
Intact component(s)
I Intervertebral disc
Vertebral buttress and articular

facets
Vertebral buttress and
intervertebral disc
Articular facets and
intervertebral disc
Vertebral buttress
II Articular facet
III Vertebral buttress
IV Articular facets and
intervertebral disc
IV Vertebral buttress and
intervertebral disc
IV Vertebral buttress and
articular facets
IV Vertebral buttress,
articular facets and
intervertebral disc
Articular facets
Intervertebral disc
None
Trauma
I: The disc has failed; the buttress and articular

facets are intact. This injury is surprisingly
unstable due to the importance of the disc for
stability against rotation, lateral bending and
13.9 I: Vertebral buttress and articular facets are intact but the
anulus has failed (arrow) (13.10).
extension (Smith and Walter, 1988; Shires et al.,

1991; Schulz et al., 1996) (13.6, 13.10). It is
sometimes combined with minor fracture(s)
of the vertebral body, but in such cases the
spine can usually still resist axial compression
and some degree of dorsoventral bending
(13.5, 13.17, 13.24). An external splint can
provide adequate stability for many fractures in
this category (Patterson and Smith, 1992).
Metal and bone cement or an external fixator
provide the best means of stabilization (Walker
et al., 2002).
II: A facet has failed; the buttress and disc are
intact. Articular facet fractures that preserve the
ventral buttress are relatively stable provided that
the disc is intact (Shires et al., 1991; Schulz et al.,
1996) (13.12). Facet joints are important for
rotational stability; the metal and bone cement
technique best resists this force (Shires et al.,
1991). Failure of one facet alone causes minimal
instability although cicatrix formation has been
recorded as a late complication of articular facet
fracture (Waters et al., 1994).
III: The buttress has failed; the facets and disc are
intact. Fractures of the ventral buttress are often
unstable and are particularly susceptible to bending
and collapse in axial compression (Walter et al.,
1986) (13.14, 13.32A). Both external and internal
fixators can give excellent results for this type of
injury (Walter et al., 1986; Bagley, 2000; Walker
et al., 2002; LeCouteur and Sturgess, 2003)
(13.21, 13.26, 13.65). An external splint is not
ideal for this type of injury.
B
13.10 This dog presented with paraplegia and a lack of deep
pain sensation in one limb. A: Myelography revealed extensive
spinal cord malacia with leakage of contrast over T12 (Lu et al.,
2002) (13.52A, 14.18). There is an isolated ventral compartment
injury; the ventral buttress is intact but the disc has failed. B: 3D
reconstruction of a CT scan from the same dog. There is
leakage of contrast (arrowhead) secondary to a presumed dural
tear (1.21, 13.52). See 13.20 and 13.30 for follow-up.
13.11 II: The vertebral buttress and disc are intact; an articular
facet is fractured (13.12).
287
288
13.12 II: This cat was tetraparetic and

paralyzed on its left thoracic and pelvic
limbs along with the ipsilateral diaphragm
after being bitten in the neck by a dog.
A: A triangular bone fragment is
overlaying the vertebral canal at C5/6
(arrow). B: Transverse CT image at C5/6
shows an impacted left articular process.
This fracture was considered stable.
The facet was removed; no fixation was
applied. The cat recovered well but
had severe, residual LMN deficits in
its left thoracic limb.
IV: Two or more components fail. An injury is

usually very unstable when two components fail
(13.15, 13.19, 13.32). Failure of two components
is often associated with failure of the third (13.16).
Any fixation technique must now withstand almost
all the forces acting on the vertebral column. For
most dogs the choice is between metal implants
and bone cement or an external fixator. An
external splint does not resist axial compression
and is not ideal when there is major failure of the
buttress (Patterson and Smith, 1992). Cage rest is
also suboptimal for this injury.
The scheme in 13.913.16 can only serve as a
guide to stability . An additional way to assess vertebral
stability is by the use of stress radiography. In its
simplest form, this is done by just comparing vertebral
positions in subsequent radiographs (13.17). The risks
of more deliberate attempts to move the fracture
are obvious and this is best done with caution under
fluoroscopy.
Regardless of the method used, our understanding
of fracture stability remains incomplete (Walter et al.,
1986; Schulz et al., 1996). Shear force and axial
13.13 III: The vertebral buttress is fractured; the articular facets

and disc are intact (13.14).
13.14 This dog jumped out of a truck at 20 mph. It could walk and was continent but had decreased anal tone. There is a fracture
of the L7 buttress; the facets are intact. A: Survey radiograph. B: Mid-sagittal 3D reconstruction of the same dog. C: Dorsal view of
the 3D reconstruction. The facets are intact but overridden (10.48). Postoperative radiographs are shown in 13.21.
Trauma
13.15 IV: Vertebral buttress and articular facet fracture; the disc
may also fail.
loading have not been tested; most models have not

looked at the effect of vertebral body failure; the
definition of instability has been questioned (Fox et al.,
1996); and the strength that a fixator requires to
counter each force is not known (Waldron et al., 1991;
Schulz et al., 1996). Most internal fixation methods
tested to date are far from ideal (Walter et al., 1986).
The relative contribution of the four main forces
acting on the vertebral column is also unknown, and
almost certainly varies between patients. For example,
dorsoventral bending occurs when a paraplegic animal
is lifted or as it attempts to raise its hindquarters.
Rotation may occur as an animal rises from lateral
recumbency or as it moves with its pelvic limbs

dragging to one side. Axial loading occurs from paravertebral muscle spasm. Only the effects of isolated
dorsoventral, lateral bending and rotational forces
have been studied experimentally and these studies
may not reflect the clinical situation accurately. It is
very hard to predict stability for most injuries except
for rare isolated facet lesions. Furthermore, good outcomes may be obtained sometimes using non-surgical
management or with less rigid internal fixation techniques even in unstable fractures (Selcer et al., 1991)
(13.63).
Despite these shortcomings, the least stable fractures must be where there has been catastrophic failure
of two or three components (Schulz et al., 1996; Shires
et al., 1991) (13.16, 13.32). Although in vitro results
for metal and bone cement implants have been conflicting, it is the internal fixation technique with the
best overall success in clinical use (Garcia et al., 1994;
Sharp et al., 1998; Bagley, 2000; Walker et al., 2002;
LeCouteur and Sturgess, 2003). An external fixator
can also be very strong and a splint can be successful in
certain situations (Patterson and Smith, 1992; Walker
et al., 2002).
Biomechanics of fixation devices

NON-SURGICAL MANAGEMENT
Animals with injuries in the cervical or lumbosacral
regions often respond well to non-surgical management
(Smith and Walter, 1985; Hawthorne et al., 1999).
Analgesia is not usually required beyond 96 h. Intractable
pain or neurological deterioration warrants a reassessment of therapy.
13.16 Images of a 1.6 kg dog with articular facet fractures; the vertebral buttress has also failed through the disc. The dog had
decreased anal tone but could walk and was continent despite the marked displacement. A: Survey radiograph. B: Mid-sagittal
3D reconstruction of a CT scan through the fracture site. C: Transverse CT scan to show articular facet fractures. Spinal stapling
was successful although not at all ideal for this fracture (13.63).
289
290
13.17 A, B: Severe instability of a spinal fracture as demonstrated by a change in position between two radiographs (McKee, 1990).
Although there is only a small fracture, loss of the disc removes 3040% of the spines overall stability (Shires et al., 1991).
It is inexpensive, there are no complications related to implants and the spinal cord is not
subject to myelography, anesthetic hypotension or
manipulation. Hospitalization times may be less than
for internal fixation (Selcer et al., 1991).
Advantages
Significant instability cannot be

addressed, reduction is difficult, some animals suffer
prolonged discomfort, and the neurological deficits
worsen in some patients or even become irreversible.
The main problem with medical management of thoracic or lumbar fractures is the longer recovery period
compared to surgery and the much longer time
required to reach optimal neurological status (Selcer
et al., 1991). Cage rest alone is only recommended for
the most stable fractures in the thoracolumbar region
and certainly not when there is severe compression or
when two or three components have failed (13.16,
13.32). Non-operative management in humans is not
really comparable as it involves reduction by traction
and a degree of immobilization that is not possible in
most animals (Fehlings and Tator, 1999; Tator et al.,
1999). An external splint is the nearest equivalent in
animals and this may also help to reduce some displaced fractures (Patterson and Smith, 1992).
Disadvantages
EXTERNAL SPLINT (13.3713.47)

An external splint has been shown to prevent dorsoventral angulation of thoracolumbar spines subjected to
large bending forces. The forces used were in excess of
those causing failure of five internal fixation techniques
in a spinal fracture model and were in the range of
those experienced by a large paralyzed dog undergoing
routine nursing care. External splinting works best with
injuries that preserve the ventral buttress; it can sometimes work with severe compromise of the ventral buttress although it is not the ideal technique for such
injuries (Patterson and Smith, 1992).
The advantages are similar to conservative management but this technique provides much
greater stability (13.46). It can be used as the sole means
of fixation or as a supplement to internal fixation. It is
most applicable to mid-thoracic or lumbar fractures but
can also be used for cervical or lumbosacral injuries.
Advantages
Disadvantages The disadvantages are: the nursing

care needed; variable individual tolerance for splints;
the possibility of neurological deterioration (13.24);
and the development of complications such as urine
scald or decubitus (13.36A, 13.47). When employed as
an adjunct to internal fixation it is difficult to observe
the surgical wound unless the splint is attached by
Velcro straps (13.41).
Good candidates for external fixation include those
with normal nociception, an intact ventral buttress
(13.9, 13.11), and no pelvic, thoracic or major soft tissue injuries (Patterson and Smith, 1992).
Marginal candidates are those with a compromised
ventral buttress (13.13, 13.15) and those with pelvic,
thoracic or soft tissue injuries (Patterson and Smith,
1992).
EXTERNAL FIXATOR (13.6413.66)

External constructs have been compared to both intact
spines and to Steinmann pin and bone cement implants.
Type II external constructs and internal fixation using
eight Steinmann pins with bone cement were stronger
Trauma
than type I external constructs and internal fixation using

four pins with cement. All implants tested were as strong
as an intact spine in extension and rotation and were
stronger in flexion; the type II external construct using a
parabolic arch was the strongest single device. Although
there was no overall difference between the arch construct and internal fixation using eight pins with cement,
the arch was the strongest in flexion. Flexion is considered the main mode of failure for most implants (Lanz
et al., 2000; Walker et al., 2002).
The advantages are that these fixators
can be made to be very strong (Walker et al., 2002); they
work well if the ventral buttress is compromised (13.65);
dissection can be minimized and pins can even be placed
fluoroscopically (Wheeler et al., 2002); and the implant
can be removed easily after healing (Ullman and
Boudrieau, 1993; Lanz et al., 2000) (13.65). It can also
be a very useful rescue technique if other methods fail
(13.65).
The advantages are that it is economical, most clinics have the necessary orthopedic implants
and it can work well when the vertebral buttress is
intact.
Advantages
The disadvantages are that it immobilizes a long segment of the vertebral column with the
weak link being the orthopedic wire, not the Steinmann
pin. In addition, the wire can cause fracture of the articular facets. It is not recommended for stabilizing a compromised vertebral buttress (13.13, 13.15).
Disadvantages
Advantages
Disadvantages include the potential to introduce infection and place excessive tension
on the skin or fascia (Shores et al., 1989; Lanz et al.,
2000); the pin tracts need regular cleaning; additional
postoperative management can be labor intensive
(Lanz et al., 2000); limited pin lengths for large dogs
using some systems (14.13); and the risk of traumatic
removal (13.65). Removal of the device prior to optimal fracture healing could result in catastrophic failure.
Disadvantages
INTERNAL FIXATION
The procedures described are usually performed via a
standard dorsal approach to the thoracolumbar spine
(12.3012.33). The dorsal spinal plate, vertebral body
plate, transilial pin techniques (Lewis et al., 1989), and
spinal stapling are no longer recommended. Bilateral
plating has similar biomechanical properties to pins and
bone cement but is harder to perform over the thoracic
and caudal lumbar regions (Viguier et al., 2002). AO
locking plates may be useful in some situations
(11.13B); these can now be made from bioabsorbable
materials that avoid stress protection by transferring
load gradually to the spine (Yerby et al., 1998; Vaccaro
and Madigan, 2002; Vaccaro et al., 2002) (13.33).
MODIFIED SEGMENTAL FIXATION (13.6113.63)

Although the original spinal stapling technique is inadequate and no longer recommended for all but the
smallest dogs, it is improved by incorporation of the
articular facets into the fixation and by use of several
pins in parallel (McAnulty et al., 1986).
METAL AND BONE CEMENT

This technique was originally described using Steinmann
pins placed in the vertebral bodies (Rouse and Miller,
1975; Blass and Seim, 1984). Threaded metal implants
and bone cement are preferred for most types of spinal
fracture/luxation and have been used in animals of all
sizes with very good results, including those with significant compromise of the ventral buttress (13.32). A
study using negative threaded, 4-mm Steinmann pins
showed that four-pin and eight-pin patterns were at
least as stiff as intact spines. Eight-pin constructs were
significantly stronger than four-pin constructs (Walker
et al., 2002). The optimum configuration of pins or
screws has yet to be determined by biomechanical testing
and a variety of configurations work well in a clinical
setting (13.25, 13.31, 13.48). Failures in clinical cases are
unusual despite mixed results in biomechanical testing
(Garcia et al., 1994; Sharp et al., 1998; Bagley et al.,
2000; Magnier and Lavaste, 2002; Walker et al., 2002;
LeCouteur and Sturgess, 2003). Implant failure occurred
in only one of 19 dogs in one series (Sharp et al., 1998)
(cement fracture shown in 13.28A).
Advantages The advantages are the immobilization

of only a short segment of the vertebral column and its
utility in thoracic and caudal lumbar areas because the
ribs and spinal nerves can be avoided easily. The implants
provide excellent strength and stability against rotation,
flexion and extension (Waldron et al., 1991; Walker et al.,
2002).
The disadvantages include the low
resistance of some configurations to dorsoventral bending (Walter et al., 1986; Walker et al., 2002); migration
of smooth pins (13.29), difficulty in closing soft tissues
over cement if placed in a doughnut pattern, difficulty of
removal; potential for implant failure (13.27, 13.28) or
stress protection (13.33), and the risk of infection
(10.45, 13.28, 13.34).
Failure of Steinmann pins and bone cement occurs
mainly because of pin pullout from the bone (Walter
et al., 1986; Waldron et al., 1991). The cement bar must
Disadvantages
291
292
be thick enough (13.59) and the implants must be strong

enough for each patient (Willer et al., 1991) (13.27).
Although one study showed poor results with screw and
cement implants, it compared 3.2-mm Steinmann pins
to 3.5-mm screws but these screws have only a 2.4-mm
core diameter (Garcia et al., 1994). The strength of a
metal implant is proportional to its radius to the power
of four, so small increases in diameter dramatically
increase implant strength (Beaver et al., 1996). When
pins of the same diameter are used to produce external
fixation and internal fixation of the lumbar spine in a
canine model the techniques have similar mechanical
properties (Walker et al., 2002). The actual strength
required in any given animal is not known but will
obviously depend on the fracture, and the weight and
temperament of the animal. In general, the largest possible implant should be used. Use of threaded in place
of smooth implants should increase the strength of the
fixator because:
Threaded implants have greater resistance to
pullout than equivalent-sized smooth pins.
Threaded implants are less likely to migrate than
smooth pins.
Threaded implants do not need to be notched to
provide anchorage for the cement.
Additional advantages of screws over threaded pins
are that:
Screws do not require cutting, which could cause
rocking and loosening of a pin.
Screws are interchangeable, so the optimum
implant length can be selected.
Screw heads provide additional purchase for the
cement.
In summary:
Positive profile threaded pins or screws should be
used instead of Steinmann pins.
The implant pattern shown in 13.54A (see also
13.20, 13.25, 13.65) is preferred over the
pattern shown in 13.49A. This is because three
implants per fragment are 66% stronger than two
per fragment (Garcia et al., 1994). Implants can be
placed in two vertebrae on each side of the lesion
if necessary (Bagley, 2000; Bagley et al., 2000),
but this immobilizes four or five vertebrae
compared to just the two or three shown in
13.20 and 13.48.
The implant should penetrate two cortices for
greater holding power (Zindrick et al., 1986). The
risk of damage to the aorta or vena cava is low but
this must still be borne in mind when drilling and
when tapping the thread (see page 9, 1.23). The risks
are lowered by using a relatively flat trajectory
(13.52, 13.54B).
The implant diameter should be as large as possible.

4.5-mm cortical screws should be used for dogs
weighing over 30 kg (Beaver et al., 1996) (13.27).
BONE GRAFTS
These have a wide range of potential uses, which include
to enhance the healing of fractures and luxations. Grafts
are either cortical or cancellous in origin; they can be
derived from the host (autograft) or from a donor (allograft) (Millis and Martinez, 2003). Autogenous cancellous graft is the most effective material to improve bone
healing. It does not provide useful mechanical strength
but up to 10% of transplanted cells survive to produce
bone (osteogenesis). It also induces adjacent cells to
make new bone (osteoinduction) and it forms a
weak scaffold for bone ingrowth (osteoconduction).
Harvesting autogenous cancellous bone requires a separate surgical approach; the proximal humerus is the
most accessible site during ventral surgical approaches
(11.24) and the iliac crest during dorsal approaches.
Cortical autografts can be harvested from the
ilium, rib, fibula or ulna.
Caudal vertebra can be used as a free vascularized
graft (Yeh and Hou, 1994).
Corticocancellous autograft can be obtained from
the ilium, sternum, wing of C6 vertebra, or from
adjacent spinous processes (Chauvet et al., 1999).
Cancellous allograft (Veterinary Transplant Services,
Seattle, WA) can be used like fresh cancellous graft;
it can also be combined with a small amount of
autogenous graft. Cancellous allografts contain no
live cells; they are not osteogenic but are both
osteoconductive and osteoinductive. Small blocks of
allograft can also be used to temporarily distract a
ventral slot during fixation.
Cortical allografts (Veterinary Transplant Services,
Seattle, WA) are used to provide some mechanical
support and a strong scaffold for new host bone
ingrowth. They may be packed with cancellous
bone to enhance healing further.
Demineralized bone matrix can also be used to
enhance bone healing. It is available as a powder
(Veterinary Transplant Services, Seattle, WA) made
from cortical bone after acid-extraction of the
mineral to expose osteoinductive proteins. These
proteins then cause adjacent mesenchymal cells to
differentiate into osteogenic cells.
Potential indications for bone grafts in neurosurgery
are to improve healing potential such as in a diabetic or
geriatric animal; for arthrodesis (9.28, 10.52D, 11.30A),
to replace bone loss such as after vertebrectomy
(12.42) and to manage infection (13.28, 14.12).
Trauma
TREATMENT
Choice of therapy
Non-surgical management is indicated for most cervical

injuries and for some lumbosacral injuries (see Anatomical location of the injury, page 295). Surgical management is best for most thoracic and lumbar injuries.
Severe shock and hypotension can exacerbate the neurological deficit and so it may be worth reassessing a
patient that shows a lack of deep pain sensation after a
24-h period of circulatory support. Definitive treatment in such animals should only be undertaken once
the owner has been made fully aware of the prognosis
and the likely time course for recovery (see Prognosis,
page 301).
PATIENTS WITH NO DEEP

PAIN SENSATION
The clinician has several options for animals with no deep
pain due to thoracic or lumbar injuries (Algorithm 13.1):
Perform euthanasia because of the poor prognosis
(Olby et al., 2003).
Presentation
Reassess deep pain after 24 h of circulatory support

combined with an external splint.
Image by myelogram or MRI in case these provide
clear evidence of spinal cord transection, which
carries a hopeless prognosis for recovery (13.4).
Severe compression or parenchymal hemorrhage on
MRI are also likely to be poor prognostic signs in
animals much as they are in humans (Ramon
et al., 1997; Fehlings et al., 1999).
Perform a durotomy as this might identify that the
spinal cord is not in continuity (13.4). However,
a severely compromised spinal cord that nevertheless
remains physically intact may be even less likely to
survive when subjected to additional manipulation.
If there is any tissue in continuity, the animal
should be given the benefit of the doubt. Only
about 5% of axons need to survive across the injury
site in order for a dog or cat to regain the ability
to walk (Blight and Decrescito, 1986; Basso et al.,
1996; Jeffery and Blakemore, 1999; Olby et al.,
2003). The presence of severe, localized malacia is
Euthanasia
Reassess after 24 hours

support with a splint
(+/ MPSS)
Deep pain
present
No deep
pain
See Algorithm 13.2
Examine for
spinal cord
transection by
durotomy
Euthanasia
Imaging
Cord intact
Cord
transected
Euthanasia
No compression
Severe
compression
Stabilize
Decompress
External
fixation
Internal
fixation
Cord
transected
Euthanasia

when deep pain is absent.
293
294
to be expected after any serious injury and on its

own does not warrant euthanasia (Salisbury and
Cook, 1988) (8.50 and see page 132).
If imaging or direct inspection rules out spinal cord
transection, the surgeon can:
Stabilize the spine using internal or external fixation
and hope that the dog recovers or can live in a cart.
Apply an external splint and reassess over 7 days.
If deep pain sensation has not returned by then,
and certainly if it does not return within 1 month,
then it probably never will. The only hope is that
the animal will have a delayed recovery of motor
function but without recovery of continence
or deep pain (Olby et al., 2003) (see pages 87,
131, 302).
Either approach is acceptable in this situation provided that the owner is aware the chances for a full
recovery are probably less than 5%.
PATIENTS WITH INTACT DEEP

PAIN SENSATION
The decision between surgical and non-surgical management for thoracic or lumbar injuries should be made
on an individual basis. A suggested approach is shown

in Algorithm13.2. If the animal satisfies criteria for a
splint (see page 290) then non-surgical therapy can be
considered. With surgery, the risks of implant failure
or infection, overall hospitalization time and cost are
factors to be considered. Surgical patients may stay in
hospital longer than medically treated patients and can
therefore incur more than twice the expense. Final neurological status may not differ between animals treated
medically and those treated by surgery (Selcer et al.,
1991). However, in general:
Surgery is the preferred treatment for thoracic or
lumbar injuries due to the low overall risk of
complications and the difficulties in instituting
non-surgical therapy in many animals. Reduction
of misalignment, additional decompression if
necessary and rigid fixation are the goals of surgery.
For most animals a simple re-alignment provides
sufficient decompression (Bagley, 2000; Bagley
et al., 2000). Immediate decompression has the
most relevance in patients with severe deficits
(Tator et al., 1999; Fehlings et al., 2001;
Papadopoulos et al., 2002).

when deep pain is present.
Consider
anatomical
location
Surgical
Non-surgical
Advanced
imaging
External
fixation
(Splint)
Compression
No
compression
Decompress by
realignment
Improves
Deteriorates
Continue
Surgical
Bone, disc
hematoma
Decompress by
(mini) hemilaminectomy
Fixation
Internal fixation
External fixator
Trauma
Adding a laminectomy will increase surgical

time and decrease stability further. It is only
indicated when there is a large extradural mass
of bone fragment, disc or blood clot (Lanz
et al., 2000).
Hemilaminectomy should be used rather than dorsal

laminectomy (Smith and Walter, 1988). Where
possible, less invasive surgeries that preserve the
facet are preferred over standard hemilaminectomy
because facet removal is very destabilizing,
particularly when the disc is also damaged (Shires
et al., 1991; Schulz et al., 1996) (see page 126).
The vertebral column must always be stabilized
after dorsal laminectomy.
A final factor that may modify the decision regarding
surgery is the anatomical location of the injury.

CERVICAL SPINE
B
13.18 A: Dog with a fracture/luxation of C2 treated with an
external splint. The splint should ideally extend from near the
lateral canthus of the eyes to the mid-thoracic region. B: An
alternative means of stabilizing the upper cervical area using
a chest harness and Halti collar; the Halti with harness can also
be used in cervical spondylomyelopathy (CSM).
This region has the largest ratio of vertebral canal to

spinal cord diameter. Cage rest, with external support
in unstable or displaced fractures (13.18), is the treatment of choice for most cervical fractures unless the
animal is deteriorating neurologically (Hawthorne et al.,
1999). Mortality rates may be as high as 3540%
with surgery (Stone et al., 1979; Hawthorne et al.,
1999). Severe intraoperative hemorrhage can also
occur with C2 fractures and reduction can be challenging (Boudrieau, 1997; Schulz et al., 1997) (13.64).
This is not the easiest location to apply a splint but one
can be made from various materials (13.18).
Surgery may reduce recovery times but is best
reserved for animals that:
Are tetraplegic or have poor ventilatory function.
Show neurological deterioration despite adequate
confinement or external fixation.
Remain very painful beyond an initial 4872 h.
The most useful surgical technique for cervical injuries
is ventral placement of pins or screws and bone cement
(Rouse, 1979; Schulz et al., 1997) (11.22, 13.19 and
see page 221). The main disadvantage of this technique
is that it may fail if used to span more than one intervertebral space, which can be a problem if a vertebral
13.19 This dog was unable to walk after
running into a tree. The C5 fracture was
stabilized with bone screws and cement
(11.22, 13.56A); the dog could walk well
within 2 weeks. Note that the screws in
C7 are placed too far caudally.
295
296
13.20 Dog shown in 13.10 after

stabilization with six 3.5-mm screws
and bone cement. This dog suffered
an unusual complication (shown in
13.30).
13.21 Postoperative radiographs from

the dog shown in 13.14. A: L7 fracture
stabilized with screws and bone cement.
B: Two cancellous screws have been
inserted into the wing of each ilium.
Better purchase would have been
obtained by angling these screws
vertically as shown in 13.65 and 14.12.
The dog was clinically normal 6 weeks
after surgery.
body is shattered (see 11.17C). In this case at least

three implants should be on either side of the fracture
and Steinmann pins used to reinforce the cement
(13.25). If dorsal stability is required, such as after facet
luxation, screws may be placed across the articular
facets (Swaim, 1975; Basinger et al., 1986).
THORACIC AND LUMBAR SPINE

These are the most common regions of the vertebral
column to be injured in animals, and are also at most
risk for devastating neurological consequences because
of the relatively small diameter of the vertebral canal
compared to other areas. Metal implants and bone
cement (13.20) or an external fixator (13.65) are preferred unless the dog is a good candidate for an external splint (Patterson and Smith, 1992) (see page 305).
L6 AND L7 VERTEBRAE AND

LUMBOSACRAL JUNCTION
The vertebral canal is relatively spacious in the caudal
lumbar area because the spinal cord ends around the
6th lumbar vertebra, and only nerve roots occupy
the caudal part of the vertebral canal (1.5, 1.9 and
see page 1). For these reasons, injuries causing severe
displacement may result in only mild neurological

deficits (13.16, 13.26). Fractures in this region can heal
satisfactorily without surgical intervention (Smith and
Walter, 1985; Patterson and Smith, 1992). Candidates
for internal fixation include animals with persistent pain,
marked deficits or deteriorating neurological status.
Screws or threaded pins and bone cement are the preferred technique, with implants in the wing of the ilium
if necessary (13.21, 13.26).
An external fixator also works well in this area (13.65,
13.66, 14.13). A splint or modified segmental fixation
can be effective but ideally the ventral buttress should
then be intact (McAnulty et al., 1986; Patterson and
Smith, 1992) (13.63).
SACRAL, SACROCAUDAL AND

TAIL INJURIES
Fractures of the sacrum have a high incidence of
neurological deficits if they traverse the vertebral canal
or involve the sacral foramina (Kuntz et al., 1995;
Anderson and Coughlan, 1997; Kuntz and Bonagura,
2000) (13.22). Surgery may improve outcome although
the risk of iatrogenic injury is high (Kuntz et al., 1995).
Therefore surgery should be restricted to animals
with severe or progressive deficits. Realignment of the
Trauma
13.22 A: Cat with a sacrocaudal injury and transverse sacral fracture (arrow). Six weeks post trauma the cat still had a paralysed
tail, no anal reflex and was incontinent. Dorsal laminectomy revealed fibrosis; neurolysis produced no improvement and the cat was
euthanized at 18 weeks. Histopathology revealed lesions in the S2, S1 and L7 nerve roots caused by traction injury to the cauda
equina (Smeak and Olmstead, 1985; Kuntz and Bonagura, 2000). B: Necropsy specimen from a dog with a transverse sacral
fracture that has severed all nerve roots at this level (arrowheads).
L7S1 articular facets facilitates reduction of a sacral

fracture (Pare et al., 2001). Sacrocaudal injuries occur
most commonly in cats (Feeney and Oliver, 1980).
The role of tethering in this type of injury is unclear
(Taylor, 1981; Smeak and Olmstead, 1985; LeCouteur
and Sturgess, 2003). Cats that do not recover continence by 4 weeks post-injury have a poor prognosis
(Smeak and Olmstead, 1985). Fractures of the caudal
vertebra are usually treated non-surgically or by tail
amputation.
POSTOPERATIVE CARE
(see also
Chapter 15)
The surgical wound must be examined at least daily for
any sign of infection. A urinary tract infection (UTI)
should be anticipated in any animal whose deficits are
severe enough that it cannot walk. Urine obtained by
cystocentesis should be analysed every 2 days or if the
urine appears cloudy; culture should be done if there is
any sign of infection. Bacteremia from the urinary and
gastrointestinal tracts or from skin lesions increases the
risk of infection at the surgical site (see page 355). Any
sudden increase in pain (13.27) or deterioration in neurological status (13.30) suggests infection or implant
failure and requires repeat radiographs. A deterioration
in patient status can also occur from non-neurological
complications of the original trauma (Box 13.2).
COMPLICATIONS
These are discussed for three major time points
intraoperative, early and late postoperative (Table 13.2).
Table 13.2 Intraoperative complications
Intraoperative
Poor reduction
(13.2413.26)
Poor implant
selection (13.27)
Poor implant
placement
(13.23)
Inadequate
cement
(13.27, 13.28A)
Pneumothorax
Fat embolism
Cardiac
arrythmias
Early
postoperative
Late
postoperative
Complications
of trauma
Implant failure
(13.28, 13.29)
Wound infection
(13.28)
Gastrointestinal
ulceration
Urinary tract
infection
Decubitus
Urine scald
Pneumonia
Vascular
complications
Sepsis
Implant failure
Infection (13.34)
Callus
encroachment
Syringomyelia
Arachnoid cyst
(13.32)
Late deformity
Adjacent segment
disease (13.28)
These fall into two main groups: cardiopulmonary complications and technical errors (Table 13.2). Cardiovascular complications are most likely in animals with
cervical and thoracic injuries (Hawthorne et al., 1999)
(see page 86). Pneumothorax is common and may be
iatrogenic (Swaim, 1971; Blass and Seim, 1984; Selcer
et al., 1991). Fat embolism may occur during fracture
repair (Schwarz et al., 2001). Technical errors in fixation
include improper placement of implants (Blass and Seim,
1984) (13.19, 13.23), poor postoperative alignment
(13.2413.26) that can even lead to a delayed onset
297
298
13.23 One Steinmann pin has been inserted too far and could
damage vital structures in the retroperitoneal space or abdomen
(see 1.23 for relationship of vertebrae to the aorta and vena
cava). This dog was normal at 5-year follow-up.
13.25 A: Mild deformity in extension, and B: significant lateral

misalignment. Use of K wires through the articular facets helps
to prevent this (Berry et al., 1999). The cement bars have been
reinforced with Steinmann pins (13.59). Six 4.5-mm screws
have been used in this 30-kg dog. Despite the poor alignment
the dog recovered from paraparesis to walking within 48 h and
was normal at 6 weeks. See also 13.51.
B
13.24 This dog was progressively paraparetic 7 days after a lowimpact injury. An external splint was applied but the neurological
status deteriorated. A: Preoperative radiograph shows an injury at
L4/5 (same dog as 13.5). B: Although facet joint alignment was
good, the surgeon exerted excessive ventral pressure causing a
misalignment in extension (Matthieson, 1983). Use of a towel or
sandbag under the abdomen helps to prevent this as does gentle
traction on the spinous processes as the cement hardens (Blass
and Seim, 1984; Lewis et al., 1989; Berry et al., 1999). Fortunately
the dog recovered quickly and was normal within 2 months.
syringomyelia in humans (Perrouin-Verbe et al., 1998;

Fischbein et al., 1999; Bains et al., 2001), poor choice
of implant (13.27), and too little cement or an uneven
cement bar (13.27, 13.28A).
13.26 This 11 kg dog was paraparetic with no anal tone 24 h

after trauma. Reduction using a laminectomy spreader, curved
hemostat or Senn retractor could have overcome the
misalignment shown here (Harrington and Bagley, 1998) (13.64).
Correct alignment is not quite so crucial in the caudal lumbar or
cervical regions because there is more space at these sites.
Nerve roots of the cauda equina also tolerate deformity better
than does spinal cord. The dog was normal within 8 weeks
and remained so 2 years later.
Trauma
13.27 Implant failure 4 days after

surgery; the cement did not cover the
cranial screws adequately (13.57). The
3.5-mm screws were also not strong
enough for this 44 kg dog (Beaver et al.,
1996). They were chosen in order to fit
down the pedicle into the base of the
articular facet of L7. This is a useful way
to add an additional implant but should
not be used as the sole fixation because
the pedicle limits the implant diameter.
The rescue procedure is shown in
13.6413.66.
13.28 Implant failure 5 days after surgery. A: The cement has broken; either the bar was not thick enough (13.59) or a bubble
(13.33), fold or thin area was present at this site (McAfee et al., 1986) (13.27). Wound culture yielded Streptococcus fecalis.
B: Wound discharge resolved 2 weeks after the implant was replaced with screws and cement plus modified segmental fixation
(13.6113.63). The dog did well for 2 years when pain and discharge returned. The implant was removed and discharge
resolved again. The dog did well for 3 more years but then developed a marked kyphotic deformity.

A high index of suspicion must be maintained during this
period for complications relating to other organ systems
such as leakage from the urinary tract or bile duct
(Matthieson, 1983). Some complications, especially vascular disorders like pulmonary thromboembolism and
deep vein thrombosis, will benefit from early mobilization and physical therapy (Selcer et al., 1991). If the animal becomes very painful or deteriorates neurologically
then implant failure or infection must be considered
(13.2813.30; see also Table 13.2). External fixation
(13.6413.66) is an excellent rescue technique in such
situations. It would have been a much better rescue
technique for the dog shown in 13.28B; bone grafting
may also help to improve healing at infected sites (Auger
et al., 2000).
The most likely problems at this stage again relate to

either the implant or to delayed infection (Table 13.2).
Of 19 dogs managed by metal and bone cement, five had
infections (13.34). All were treated successfully although
one suffered a late recrudescence (13.28). Three of the
five did not undergo implant removal (Sharp et al., 1998).
Infection can also be a problem underneath external
splints and may even be fatal if detected too late (McAfee
et al., 1986) (13.36A). Callus has been reported to
encroach into the vertebral canal but this is a very unusual
problem (Carberry et al., 1989). Internal fixators do not
generally require removal; late removal could lead to
kyphotic deformity or even pathological fracture (13.28,
13.33). Other potential problems include adjacent segment disease (13.31), late-onset syringomyelia (PerrouinVerbe et al., 1998), or arachnoid cyst formation (13.32).
299
300
13.29 This dog died 1 week after surgery of unknown causes.

The pin was loose in the abdomen but did not appear to be the
cause of death. Implant migration is most likely with smooth pins
(Blass and Seim, 1984; Blass et al., 1988; Wong and Emms,
1992). Every effort should be made to incorporate all implants into
the cement, including any K wires (compare 13.48 and 13.63).
13.31 Three-year follow-up of the dog shown in 13.32.

Although domino lesions have not been reported after fracture
repair there is evidence that additional strain is imposed on
segments adjacent to the fixation (arrows) (Fox et al., 1996;
Wheeler et al., 2002). This could help to explain the failure
shown in 13.30 (Goffin et al., 1995).
13.30 This dog lost deep pain sensation 8 days after surgery at
T12/13 (13.20). This unusual failure may be a type of adjacent
segment disease (MacMillan and Stauffer, 1991; Fox et al.,
1996; Lanz et al., 2000; Wheeler et al., 2002) (11.23, 13.31).
A: Radiography revealed luxation at the adjacent T13/L1
interspace. The implant was intact. B: Marked rotational
deformity was also present. The preoperative myelogram and
CT scan were re-examined but no injury was found at T13/L1.
3D CT scan reconstructions were also examined to view the
articular facets but no abnormality was detected (13.10B). The
dog recovered motor function at 3 months; it could walk by
6 months but did not recover continence or deep pain
(Olby et al., 2003) (see pages 32, 131, 302).
B
13.32 A: Preoperative myelogram of the dog shown in 13.31
and 13.33. The T12 vertebral buttress has failed and the facet
joints have luxated. Despite the severity of the injury the dog
made an excellent recovery and did very well for 3 years when
it developed progressive paraparesis. B: A myelogram then
revealed an arachnoid cyst (arrowhead) at T12/13 (see
pages 321, 323). The dog remained stable for a further 21
months without intervention (Skeen et al., 2003).
Trauma
13.35 Example of the devastating type of spinal injury

commonly encountered in cats (13.6, 13.12).
13.33 A CT scan through the arachnoid cyst shown in 13.32
confirmed a focal dilation of the subarachnoid space (arrowhead).
Note also the large bubble in the cement (arrow). This can
weaken the implant and could predispose it to failure (McAfee
et al., 1986; Anderson, 1988; Beaver et al., 1996) (13.28). There
is also significant loss of cortical bone in the vertebral body when
compared to the ribs. This could be due to stress protection,
which could complicate late implant removal (Craven et al., 1994;
Vaccaro and Madigan, 2002) (13.28).
FELINE SPINAL INJURIES

Although non-surgical therapy is useful it usually is
limited to cage confinement as cats tolerate splints poorly
(Carberry et al., 1989). In theory, any internal fixation
technique can be applied to cats although feline spinal
injuries are often severe (13.35).
PROGNOSIS
13.34 Infection is a potential problem when using bone

cement. A: Reduction of the luxation was poor. B: The dog
developed pain after 6 weeks due to discospondylitis at T13/L1.
The entire implant was removed and the pain resolved after
2 months of antibiotics (Blass and Seim, 1984; Wong and Emms,
1992) (13.28). At 2-year follow-up the dog was walking but had
no deep pain sensation (Olby et al., 2003).
This varies somewhat with the location of injury. Prognosis is very good after cervical injury if the animal does not
die acutely from respiratory arrest (see pages 28, 82). In
the largest series to date, 32 of 39 dogs (82%) with neck
injuries made a functional recovery; results were much
better after non-surgical treatment because the mortality
rate was only one third of that in the surgically treated
dogs. Twenty-five of 28 dogs recovered after non-surgical
management compared with seven of 11 after surgery
(89% vs 64%). Results were good after non-surgical management even for dogs that were unable to walk. Delay in
instituting definitive treatment was a predictor of poor
outcome (Hawthorne et al., 1999). Immediate decompression by vertebral realignment is likely to benefit animals with tetraplegia or hypoventilation secondary to
cervical injury (Papadopoulos et al., 2002).
The prognosis following thoracic and lumbar injury is
similar to that for disc disease provided that the animal
has good deep pain at presentation. However, studies
with good follow-up are limited. Surgical management
of 24 such animals gave good results in 21 (88%),
although 10 had residual deficits and one was euthanized due to implant migration after 22 months (Blass
and Seim, 1984; McKee, 1990). In another study, 19
of 23 dogs (83%) presenting with intact deep pain
made good recoveries (Sharp et al., 1998). Non-surgical
management in a further 25 animals, comprising either
cage confinement or splinting, gave good results in 24
(96%). Three had residual deficits (Carberry et al.,
301
302
1989; McKee, 1990). Non-surgical management using

a splint in 16 dogs gave good results in all animals
although six had residual deficits. Six dogs also developed serious complications and three deteriorated but
these problems were temporary and resolved in all
cases (Patterson and Smith, 1992).
Animals presenting with thoracic or lumbar injuries
and no deep pain rarely recover after trauma. Of nine
such dogs, none regained deep pain over follow-up
periods of 6 weeks to 2 years (Olby et al., 2003). Two
did recover motor function without recovering deep
pain (13.30, 13.34). This unusual phenomenon, which
can also occur after thoracolumbar disc disease (see
pages 32, 131), presumably is mediated by a peripheral
rim of surviving axons at the edge of an extensive central lesion (Griffiths, 1978; Olby et al., 2003) (6.2).
Recovery of motor function without deep pain usually
takes 34 months; it rarely produces a normal gait and
animals are usually incontinent with recurrent urinary
tract infections. Based on extrapolation of data from
disc disease, a full recovery is unlikely unless deep pain
returns within 4 weeks of injury (Olby et al., 2003).
Although MPSS may have some benefit in severe

injuries, its contribution is likely to be marginal (see
page 83). Adequate reduction and fixation by internal
or external means is far more important.
The prognosis after sacral fracture is good with 26 of
32 dogs (81%) recovering (Kuntz et al., 1995). Cats
with sacrocaudal injuries usually recover urinary continence if they retain good anal tone and perineal sensation on initial examination. Cats that do not become
continent within 1 month generally fail to regain urinary
function (Moise and Flanders, 1983; Smeak and
Olmstead, 1985). Animals that remain incontinent
may be candidates for permanent cytostomy tubes
(see 15.32).
The prognosis following brachial plexus injury is
guarded to poor. The best prognostic indicator is the
presence or absence of deep pain sensation in the distal
limb. Five of seven dogs with normal sensation made
complete recoveries whereas only six of 21 with
reduced sensation recovered or remained in stable condition. Abnormal motor nerve conduction is also associated with a negative outcome (Faissler et al., 2002).

1. What is the best way to manage cranial thoracic lesions (13.36A)?
2. Is there any advantage to using more severe stimuli to assess deep pain (13.36B)?
3. Will MRI prove to be prognostic in animals as it may be in humans by detecting spinal cord hemorrhage (Ramon et al., 1997;
Fehlings et al., 1999; Selden et al., 1999)?
4. What can electrophysiology contribute to determining prognosis?
13.36 A: This dog showed progressive paraparesis after trauma. It had subluxation at T2/3, which was managed with an
external splint. This is a difficult area to stabilize so the sides of the splint were extended down each flank to just above the elbow
but this made it impossible to check beneath the splint without removing it. The dog would not eat 5 days after splinting and was
found in extremis on day 6. A huge, phlegmon-like, subcutaneous abscess was present associated with a small decubital lesion.
Bone cement and threaded implants would have been preferable in this dog. B: The prognosis for dogs with traumatic injuries
and without nociception on presentation is very poor (Olby et al., 2003). It would be helpful to be able to differentiate animals
without deep pain that are capable of functional recovery from those with irreversible injuries. Here an electrical stimulator from
an electromyography (EMG) machine is being used but there was no response in this dog. Alternative methods include strong
pliers or a cattle prod (see page 132).
Trauma
REFERENCES
AANS (2000) Guideline to Improve the Outcome of Brain Trauma Patients:
Management and Prognosis of Severe Traumatic Brain Injury, 2nd edn.
Brain Trauma Foundation: American Association of Neurological Surgeons.
http://www.braintrauma.org/index.nsf/Pages/Guidelines-main.
Anderson, A., Coughlan, A.R. (1997) Sacral fractures in dogs and cats: a
classification scheme and review of 51 cases. Journal of Small Animal
Anderson, G.I. (1988) Polymethylmethacrylate: a review of the implications and complications of its use in orthopaedic surgery. Veterinary and
Bagley, R.S. (2000) Spinal fracture or luxation. In: W.B. Thomas (ed.),
Veterinary Clinics of North America Small Animal Practice, 133154.
Bagley, R.S., Silver, G.M., Connors, R.L., Harrington, M.L., Cambridge, A.J.,
Moore, M.P. (2000) Exogenous spinal trauma: surgical therapy and aftercare. Compendium on Continuing Education for the Practicing Veterinarian
22, 218230.
Bains, R.S., Althausen, P.L., Gitlin, G.N., Gupta, M.C., Benson, D.R.
(2001) The role of acute decompression and restoration of spinal alignment in the prevention of post-traumatic syringomyelia: case report and
review of recent literature. Spine 26, E399402.
Basinger, R.R., Bjorling, D.E., Chambers, J.N. (1986) Cervical spinal luxation in two dogs with entrapment of the cranial articular process of C6 over
the caudal articular process of C5. Journal of the American Veterinary
Basso, D.M., Beattie, M.S., Bresnahan, J.C. (1996) Graded histological
and locomotor outcomes after spinal cord contusion using the NYU
weight-drop device versus transection. Experimental Neurology 139,
244256.
Beaver, D.P., MacPherson, G.C., Muir, P., Johnson, K.A. (1996) Methylmethacrylate and bone screw repair of seventh lumbar vertebral
fracture-luxations in dogs. Journal of Small Animal Practice 37,
381386.
Berry, W.L., LeCouteur, R.A., Vernau, K.M., Sturges, B.K. (1999)
Trans-articular process pinning for the initial stabilization of vertebral
fracture/luxations in 18 dogs. Journal of Veterinary Internal Medicine
13, 240.
Blass, C.E., Seim, H.B. (1984) Spinal fixation in dogs using Steinmann pins
and methylmethacrylate. Veterinary Surgery 13, 203210.
Blass, C.E., Waldron, D.R., Van Ee, R.T. (1988) Cervical stabilization in
three dogs using Steinmann pins and methylmethacrylate. Journal of the
Blight, A.R., Decrescito, V. (1986) Morphometric analysis of experimental
spinal cord injury in the cat: the relation of injury intensity to survival of
myelinated axons. Neuroscience 19, 321341.
after cervical vertebral interbody fusion with polymethyl methacrylate
Boudrieau, R.J. (1997) Distraction-stabilization using the ScovilleHaverfield
self-retaining laminectomy retractors for repair of 2nd cervical vertebral
fractures in 3 dogs. Veterinary and Comparative Orthopaedics and
Campbell, V.L., King, L.G. (2000) Pulmonary function, ventilator management, and outcome of dogs with thoracic trauma and pulmonary contusions: 10 cases (19941998). Journal of the American Veterinary
Carberry, C.A., Flanders, J.A., Dietze, A.E., Gilmore, D.R., Trotter, E.J.
(1989) Nonsurgical management of thoracic and lumbar spinal fractures
and fracture luxations in the dog and cat: a review of 17 cases. Journal of
Chauvet, A., Hogge, G., Sandin, J., Lipsitz, D. (1999) Vertebrectomy,
bone allograft fusion, and antitumor vaccination for the treatment of vertebral fibrosarcoma in a dog. Veterinary Surgery 28, 480488.
Craven, T.G., Carson, W.L., Asher, M.A., Robinson, R.G. (1994) The
effects of implant stiffness on the bypassed bone mineral density and
facet fusion stiffness of the canine spine. Spine 19, 16641673.
Denis, F. (1984) Spinal instability as defined by the three-column spine
concept in acute spinal trauma. Clinical Orthopaedics and Related
Research 189, 6576.
Egger, E.L., Histand, M.B., Blass, C.E., Powers, B.E. (1986) Effect of fixation
pin insertion on the bone-pin interface. Veterinary Surgery 15, 246252.
Ethell, M.T., Bennett, R.A., Brown, M.P., Merritt, K., Davidson, J.S., Tran, T.
(2000) In vitro elution of gentamicin, amikacin, and ceftiofur from polymethylmethacrylate and hydroxyapatite cement. Veterinary Surgery 29,
375382.
Faissler, D., Cizinauskas, S., Jaggy, A. (2002) Prognostic factors for functional recovery in dogs with suspected brachial plexus avulsion. Journal
Feeney, D.A., Oliver, J.E. (1980) Blunt spinal trauma in the dog and cat:
neurologic, radiologic, and therapeutic correlations. Journal of the
Fehlings, M., Tator, C. (1999) An evidence-based review of decompressive
surgery in acute spinal cord injury: rationale, indications, and timing based
on experimental and clinical studies. Journal of Neurosurgery 91, 111.
Fehlings, M.G., Rao, S.C., Tator, C.H. et al. (1999) The optimal radiologic
method for assessing spinal canal compromise and cord compression in
patients with cervical spinal cord injury. Part II: Results of a multicenter
study. Spine 24, 605613.
Fehlings, M.G., Sekhon, L.H., Tator, C. (2001) The role and timing of
decompression in acute spinal cord injury: what do we know? What
should we do? Spine 26, S101110.
Fox, M., Onofrio, B., Hanssen, A. (1996) Clinical outcomes and radiological
instability following decompressive lumbar laminectomy for degenerative
spinal stenosis: a comparison of patients undergoing concomitant arthrodesis versus decompression alone. Journal of Neurosurgery 85, 793802.
Goffin, J., van Loon, J., Van Calenbergh, F., Plets, C. (1995) Long-term
results after anterior cervical fusion and osteosynthetic stabilization for
fractures and/or dislocations of the cervical spine. Journal of Spinal
Disorders 8, 500508; discussion 499.
Griffiths, I. (1970) A syndrome produced by dorso-lateral explosions of
the cervical intervertebral discs. Veterinary Record 87, 737741.
Griffiths, I.R. (1978) Spinal cord injuries: a pathological study of naturally
occurring lesions in the dog and cat. Journal of Comparative Pathology
88, 303315.
Griffiths, I.R., Duncan, I.D., Lawson, D.D. (1974) Avulsion of the brachial
plexus. 2. Clinical aspects. Journal of Small Animal Practice 15, 177183.
Harrington, M.L., Bagley, R.S. (1998) Realignment of a seventh lumbar
vertebral fracture/luxation using a Senn retractor in two puppies.
Hawthorne, J.C., Blevins, W.E., Wallace, L.J., Glickman, N., Waters, D.J.
(1999) Cervical vertebral fractures in 56 dogs: a retrospective study.
Hay, C.W., Muir, P. (2000) Tearing of the dura mater in three dogs.
Veterinary Record 146, 279282.
inappropriate standard of care. Journal of Neurosurgery 93, 17.
Jeffery, N.D., Blakemore, W.F. (1999) Spinal cord injury in small animals 2.
Current and future options for therapy. Veterinary Record 145, 183190.
Kriaras, I., Michalopoulos, A., Turina, M., Geroulanos, S. (2000) Evolution
of antimicrobial prophylaxis in cardiovascular surgery. European Journal
of Cardiothoraic Surgery 18, 440446.
Kuntz, C.A., Bonagura, J.D. (2000) Sacral fractures and sacrococcygeal
injuries in dogs and cats. In: R.W. Kirk and J.D. Bonagura (eds), Kirks
Current Veterinary Therapy XIII: Small Animal Practice, 10231026.
303
304
Kuntz, C.A., Waldron, D., Martin, R.A., Shires, P.K., Moon, M., Shell, L.
(1995) Sacral fractures in dogs: a review of 32 cases. Journal of the
Lanz, O.I., Jones, J.C., Bergman, R. (2000) Use of an external fixator to
correct spinal fracture/luxation and instability in three dogs. Veterinary
Neurology and Neurosurgery, http://www.neurovet.org/LanzSpinalFixator/
LANZ1final.htm#spinal%20instability,%20traumatic.
LeCouteur, R.A., Sturgess, B. (2003) Spinal fractures and luxations. In:
Lewis, D.D., Stampley, A., Bellah, J.R., Donner, G.S., Ellison, G.W.
(1989) Repair of sixth lumbar vertebral fracture-luxations, using transilial pins and plastic spinous-process plates in six dogs. Journal of the
MacMillan, M., Stauffer, E.S. (1991) Traumatic instability in the previously fused cervical spine. Journal of Spinal Disorders 4, 449454.
Magnier, J., Lavaste, F. (2002) In vitro biomechanical evaluation of unstable T13L1 stabilization procedures in dogs. Veterinary Surgery 31, 288.
Marcellin Little, D.J., Papich, M.G., Richardson, D.C., DeYoung, D.J. (1996)
Pharmacokinetic model for cefazolin distribution during total hip arthroplasty in dogs. American Journal of Veterinary Research 57, 720723.
Matthieson, D.T. (1983) Thoracolumbar spinal fractures/luxations: surgical
management. Compendium on Continuing Education for the Practicing
McAfee, P.C., Bohlman, H.H., Ducker, T., Eismont, F.J. (1986) Failure of
stabilization of the spine with methylmethacrylate. A retrospective
analysis of twenty-four cases. The Journal of Bone and Joint Surgery
American volume 68, 11451157.
McAnulty, J.F., Lenehan, T.M., Maletz, L.M. (1986) Modified segmental
spinal instrumentation in repair of spinal fractures and luxations in dogs.
McKee, W.M. (1990) Spinal trauma in dogs and cats: a review of 51 cases.
Millis, D.L., Martinez, S.A. (2003) Bone grafts. In: D. Slatter (ed.),
Textbook of Small Animal Surgery, 3rd edn, 18751890. Philadelphia:
Elsevier Science.
Moise, N.S., Flanders, J.A. (1983) Micturation disorders in cats with
sacrocaudal vertebral lesions. In: R.W. Kirk (ed.), Current Veterinary
Therapy VIII, 772776. Philadelphia: WB Saunders.
Murtaugh, R.J. and Ross, J.N. (1988) Cardiac arrhythmias: pathogenesis
and treatment in the trauma patient. Compendium on Continuing
Neer, T.M. (1992) A review of disorders of the gallbladder and extrahepatic
biliary tract in the dog and cat. Journal of Veterinary Internal Medicine
6, 186192.
Novelli, A. (1999) Antimicrobial prophylaxis in surgery: the role of pharmacokinetics. Journal of Chemotherapy 11, 565572.
Long-term functional outcome of dogs with severe injuries of the thoracolumbar spinal cord; 87 cases. Journal of the American Veterinary
Papadopoulos, S.M., Selden, N.R., Quint, D.J., Patel, N., Gillespie, B.,
Grube, S. (2002) Immediate spinal cord decompression for cervical spinal
cord injury: feasibility and outcome. Journal of Trauma 52, 323332.
Pare, B., Gendreau, C.L., Robbins, M.A. (2001) Open reduction of sacral
fractures using transarticular implants at the articular facets of L7S1: 8
consecutive canine patients (19951999). Veterinary Surgery 30, 476481.
Patterson, R.H., Smith, G.K. (1992) Backsplinting for treatment of thoracic and lumbar fracture/luxation in the dog: principles of application
and case series. Veterinary and Comparative Orthopaedics and
Perrouin-Verbe, B., Lenne-Aurier, K., Robert, R., Auffray-Calvier, E.,
Richard, I., Mauduyt de la Greve, I., Mathe, J.F. (1998) Post-traumatic
syringomyelia and post-traumatic spinal canal stenosis: a direct relationship: review of 75 patients with a spinal cord injury. Spinal Cord 36,
137143.
Platt, S.R., Radaelli, S.T., McDonnell, J.J. (2001) The prognostic value of
the modified Glasgow Coma Scale in head trauma in dogs. Journal of
Ramon, S., Dominguez, R., Ramirez, L., Paraira, M., Olona, M., Castello, T.,
Garcia Fernandez, L. (1997) Clinical and magnetic resonance imaging
correlation in acute spinal cord injury. Spinal Cord 35, 664673.
Rouse, G.P. (1979) Cervical spinal stabilization with methylmethacrylate.
Rouse, G.P., Miller, J.I. (1975) The use of methylmethacrylate for spinal
stabilization. Journal of the American Animal Hospital Association 11,
418425.
Salisbury, S.K., Cook, J.R., Jr (1988) Recovery of neurological function following focal myelomalacia in a cat. Journal of the American Animal
Sandman, K.M., Smith, C.W., Harari, J., Manfra Maretta, S., Pijanowski, G.J.
(2001) Comparison of pull-out resistance of Kirschner wires and Imex
miniature interface fixation pins in polyurethane foam. Veterinary and
Schulz, K.S., Waldron, D.R., Grant, J.W., Shell, L., Smith, G., Shires, P.K.
(1996) Biomechanics of the thoracolumbar vertebral column of dogs
during lateral bending. American Journal of Veterinary Research 57,
12281232.
Schulz, K.S., Waldron, D.R., Fahie, M. (1997) Application of ventral pins
and polymethylmethacrylate for the management of atlantoaxial instability: results in nine dogs. Veterinary Surgery 26, 317325.
Schwarz, T., Crawford, P.E., Owen, M.R., Stork, C.K., Thompson, H.
(2001) Fatal pulmonary fat embolism during humeral fracture repair in
a cat. Journal of Small Animal Practice 42, 195198.
Selcer, R.R., Bubb, W.J., Walker, T.L. (1991) Management of vertebral column fractures in dogs and cats: 211 cases (19771985). Journal of the
Selden, N.R., Quint, D.J., Patel, N., dArcy, H.S., Papadopoulos, S.M.
(1999) Emergency magnetic resonance imaging of cervical spinal cord
injuries: clinical correlation and prognosis. Neurosurgery 44, 785792;
discussion 792793.
Sharp, N.J.H., Gilson, S.D., Kornegay, J.N., Wheeler, S.J., Hopkins, A.L.,
Lane, S.D. (1998) Long term follow-up of spinal fracture fixation using
screws or pins and bone cement in 30 dogs. Veterinary Orthopedic
Society 1998; Snowmass, CO.
Shores, A., Nichols, G., Rochat, M., Fox, S.M., Burt, G.J., Fox, W.R.
(1989) Combined KirschnerEhmer device and dorsal spinal plate fixation technique for caudal lumbar vertebral fractures in dogs. Journal of
Skeen, T.M., Olby, N.J., Munana, K.M., Sharp, N.J.H. (2003) Spinal
arachnoid cysts in 17 dogs. Journal of the American Animal Hospital
Smeak, D.D., Olmstead, M.L. (1985) Fracture/luxations of the sacrococcygeal area in the cat. A retrospective study of 51 cases. Veterinary
Surgery 14, 319324.
Smith, G.K., Walter, M.C. (1985) Fractures and luxations of the spine.
In: C.D. Newton and D. Nunamaker (eds), Textbook of Small Animal
Orthopedics, 307322. Philadelphia: JB Lippincott.
Snyder, P.S., Cooke, K.L., Murphy, S.T., Shaw, N.G., Lewis, D.D., Lanz, O.I.
(2001) Electrocardiographic findings in dogs with motor vehiclerelated trauma. Journal of the American Animal Hospital Association 37,
5562.
Stone, E.A., Betts, C.W., Chambers, J.N. (1979) Cervical fractures in the
dog: a literature and case review. Journal of the American Animal
Swaim, S.F. (1971) Vertebral body plating for spinal immobilization.
Swaim, S.F. (1975) Evaluation of four techniques of cervical spinal fixation
in dogs. Journal of the American Veterinary Medical Association 166,
10801086.
Trauma
Tator, C.H., Fehlings, M.G., Thorpe, K., Taylor, W. (1999) Current use and
timing of spinal surgery for management of acute spinal cord injury in
North America: results of a retrospective multicenter study. Journal of
Neurosurgery 91 (1 Suppl), 1218.
Taylor, R.A. (1981) Treatment of fractures of the sacrum and sacrococcygeal region. Veterinary Surgery 10, 119124.
Tidwell, A.S., Mahony, O.M., Moore, R.P., Fitzmaurice, S.N. (1994)
Computed tomography of an acute hemorrhagic cerebral infarct in a
dog. Veterinary Radiology and Ultrasound 35, 290296.
Turner, W.D. (1987) Fractures and fracture-luxations of the lumbar spine:
retrospective study in the dog. Journal of the American Animal Hospital
Ullman, S.L., Boudrieau, R.J. (1993) Internal skeletal fixation using a
Kirschner apparatus for stabilization of fracture/luxations of the lumbosacral joint in six dogs. A modification of the transilial pin technique.
Vaccaro, A.R., Madigan, L. (2002) Spinal applications of bioabsorbable
implants. Orthopedics 25 (10 Suppl), s11151120.
Vaccaro, A.R., Venger, B.H., Kelleher, P.M., Singh, K., Carrino, J.A.,
Albert, T., Hilibrand, A. (2002) Use of a bioabsorbable anterior cervical
plate in the treatment of cervical degenerative and traumatic disk disruption. Orthopedics 25 (10 Suppl), s11911199; discussion s1199.
Viguier, E., Petit-Etienne, G., Magnier, J., Lavaste, F. (2002) In vitro biomechanical evaluation of unstable T13L1 stabilization procedures in
Waldron, D.R., Shires, P.K., McCain, W., Hedlund, C., Blass, C.E. (1991)
The rotational stabilizing effect of spinal fixation techniques in an unstable vertebral model. Progress in Veterinary Neurology 2, 105110.
Walter, M.C., Smith, G.K., Newton, C.D. (1986) Canine lumbar spinal
internal fixation techniques. A comparative biomechanical study.
Waters, D.J., Wallace, L.J., Roy, R.G. (1994) Myelopathy in a dog secondary to scar tissue (cicatrix) formation: a complication of vertebral articular facet fracture. Progress in Veterinary Neurology 5, 105108.
Weisman, D.L., Olmstead, M.L., Kowalski, J.J. (2000) In vitro evaluation
of antibiotic elution from polymethylmethacrylate (PMMA) and
mechanical assessment of antibiotic-PMMA composites. Veterinary
Surgery 29, 245251.
Weisse, C., Aronson, L.R., Drobatz, K. (2002) Traumatic rupture of the
ureter: 10 cases. Journal of the American Animal Hospital Association
38, 188192.
Wheeler, J.L., Cross, A.R., Rapoff, A.J. (2002) A comparison of the accuracy and safety of vertebral body pin placement using a fluoroscopically
guided versus an open surgical approach: an in vitro study. Veterinary
Surgery 31, 468474.
Willer, R.L., Egger, E.L., Histand, M.B. (1991) Comparison of stainless
steel versus acrylic for the connecting bar of external skeletal fixators.
Wong, W.T., Emms, S.G. (1992) Use of pins and methylmethacrylate in
stabilisation of spinal fractures and luxations. Journal of Small Animal
Yarrow, T.G., Jeffery, N.D. (2000) Dura mater laceration associated with
acute paraplegia in three dogs. Veterinary Record 146, 138139.
Yeh, L.S., Hou, S.M. (1994) Repair of a mandibular defect with a free vascularized coccygeal vertebra transfer in a dog. Veterinary Surgery 23,
281285.
Yerby, S.A., Toh, E., McLain, R.F. (1998) Revision of failed pedicle screws
using hydroxyapatite cement. A biomechanical analysis. Spine 23,
16571661.
Zindrick, M.R., Wiltse, L.L., Widell, E.H., Thomas, J.C., Holland, W.R.,
Field, B.T., Spencer, C.W. (1986) A biomechanical study of intrapeduncular screw fixation in the lumbosacral spine. Clinical Orthopaedics and
Related Research 203, 99112.
PROCEDURES
Non-surgical (13.3713.47)
Cage confinement can give excellent results and is the only non-surgical option for most cats (Carberry et al., 1989;
Selcer et al., 1991). A splint is preferred for dogs if certain criteria are satisfied including intact deep pain sensation,
an intact ventral buttress and no pelvic, thoracic or soft tissue injuries (Patterson and Smith, 1992) (see page 290).
It is essential that as much of the area under the splint as possible along with the skin of the groin and axilla be
examined daily. Urine soiled straps must be replaced. Areas at risk from urine scalding or fecal soiling benefit from
application of Desitin ointment (Pfizer Inc., New York). If there is any doubt, the straps or splint must be loosened
sufficiently to allow proper inspection and then reattached. If doubt remains the splint must be removed.
Useful indicators of potential problems are the animals mental attitude, its appetite, and the presence of pain
or fever (13.36A). Most animals are surprisingly comfortable in a splint unless problems such as decubital ulcers
develop. The splint should be maintained for a minimum of 4 weeks (3 weeks for immature animals), and is
preferably then followed by a further 24 weeks of strict cage confinement.
305
306
13.37 Diagram to illustrate the construction of an

external splint that is to be applied to a dog with
a spinal injury. The splint is made from sheet
aluminum, 0.54 mm in thickness, double or
triple layered if necessary for added rigidity. In
animals less than 10 kg, materials such as
Orthoplast (Johnson and Johnson, New
Brunswick, NJ) may be substituted. To make
the splint, measure and cut the splint to the
dimensions AB, 2AC and 2BD (Patterson
and Smith, 1992).
13.37
13.38 The splint extends from between the

scapulae to the tail base, measured directly
along the spine. Most animals have marked
kyphosis following spinal injury, which must be
considered during measurement.
13.38
13.39 The material is bent along its longitudinal axis

to form a ridge; the splint is flattened slightly
over the pelvic area. At the sides, the splint
should extend laterally to the mid-scapula and
the hip joint (13.46). The splint should be well
padded with cotton, especially at the edges,
and covered with Elastoplast (Johnson and
Johnson, New Brunswick, NJ).
13.39
13.40 Some animals tolerate application of the splint

while fully conscious, but most require
sedation and a few need general anesthesia
(Patterson and Smith, 1992). With the animal
on its side, the splint is slid under the dog with
the ridge aligned along the animals dorsal
midline.
13.40
Trauma
13.41 The animal and splint are then rotated
simultaneously until the animal is lying on its back in

the splint (Patterson and Smith, 1992). The dog
usually relaxes and the fracture may reduce to
some degree. The axillae and groin are padded
heavily with cotton wool prior to attaching the splint
using Elastoplast (Johnson and Johnson, New
13.41
Brunswick, NJ). This has the advantage of being
light, porous and adhesive. Velcro straps are much
easier to re-adjust but more likely to cause abrasions. Velcro is preferred when the splint needs to be
removed frequently.
13.42 To secure the dog to the splint, the caudal half

of the animal is first extended over the end of
a table.
13.42
13.43 Elastoplast (Johnson and Johnson, New

Brunswick, NJ) is applied in a cruciate pattern
over the pelvis, avoiding the anus and vulva. In
male animals it is important that any strap
crossing the base of the penis is not too tight,
or pressure necrosis of the urethra is possible
(Patterson and Smith, 1992).
13.43
13.44 In male dogs, a cod-piece roughly the shape

of a half flowerpot can be made from
aluminum, Orthoplast (Johnson and Johnson,
New Brunswick, NJ) or an empty saline bag to
protect the base of the penis and to try to
funnel urine away from the abdomen and
inguinal straps (Bagley et al., 2000).
13.44
307
308
13.45 In the pectoral region, the straps are applied as a figure of eight,
extending in front of the shoulder and under the neck like a
harness. The straps around the chest must be snug but must
not impede ventilation. Radiographs can be taken through the
splint to check fracture alignment.
13.45
13.46 The main advantages of an external splint are

that it is inexpensive and strong (Patterson and
Smith, 1992). It also facilitates lifting the
animal, turning it from side to side, moving it to
clean the kennel, or taking it out to void.
Handles made from aluminum rods can also
be incorporated into the splint or may be
taped on afterwards (Bagley, 2000).
13.46
13.47 The main disadvantage is the potential for

pressure sore formation along the edges and
underneath the splint (13.36A). In addition, the
straps can cause skin excoriation in the axilla
or inguinal region. They may also hinder
manual expression of the bladder, and
prevention of urine scald can be challenging.
13.47
Trauma
Metal and bone cement (13.4813.60)

This is the internal fixation technique of choice. Threaded pins or screws should be used (Sandman et al., 2001); the
implant diameter should be as large as possible (Beaver et al., 1996), each implant should penetrate two cortices
(Zindrick et al., 1986), and ideally three implants should be used on each side of the lesion (Garcia et al., 1994).
The cement bar must also be strong enough (Rouse and Miller, 1975; Willer et al., 1991) (13.59). Variables in using
implants include:
The pattern in which implants are placed (13.49, 13.54).
The angle or trajectory at which they penetrate the vertebrae (13.52).
Their entry points (13.55).
Intravenous antibiotic effective against staphylococci (such as a cephalosporin) is given and repeated every
12 h during surgery (Marcellin Little et al., 1996; Novelli, 1999; Kriaras et al., 2000).
Fracture reduction can be very difficult in some injuries without some type of mechanical assistance (Boudrieau,
1997; Schulz et al., 1997; Bagley et al., 2000) (13.64). A laminectomy spreader, Senn retractor, or a curved
hemostat can be used for reduction of L7 fractures or manual traction can be put on the tail (Blass and Seim,
1984; Beaver et al., 1996; Harrington and Bagley, 1998; Bagley et al., 2000). Cervical injuries can be reduced
using traction on the maxilla or a Gelpi retractor can be placed in holes drilled in the middle of a vertebra (Blass
and Seim, 1984; Blass et al., 1988) (11.38). A ScovilleHaverfield or similar retractor can also be placed in adjacent
disc spaces or between a disc space and the base of the skull for C2 fractures (Boudrieau, 1997).
Initial stabilization is provided by bone-holding forceps (13.48, 13.60) while K-wires are placed across the articular facets to provide additional stability and maintain alignment (Blass and Seim, 1984; Beaver et al., 1996; Berry
et al., 1999) (13.28A, 13.48). When a hemilaminectomy must be performed, when a facet is fractured, and when
a ventral approach is used for cervical fractures, the K-wire can instead be placed across a disc space to maintain reduction (Blass et al., 1988). Additional methods for reducing a fracture are discussed in 13.64.
A small K wire is used initially to make a point of purchase for the drill bit. It can also be used to make a test hole
to evaluate bone quality over the chosen path of the implant. If bone quality is poor this small test hole can easily
be redirected without prejudicing the final pilot hole. Pins should be placed using a low power setting to reduce
bone necrosis; a pilot hole is recommended for threaded pins (Egger et al., 1986; Walker et al., 2002). Muscle
should be protected during drilling with a drill guard. If brisk hemorrhage arises from the pilot hole then a finger
or bone wax should be placed over the hole, which should be tapped and screwed as soon as possible (1.19).
A flatter implant trajectory increases implant holding strength due to the greater bone contact (Garcia et al.,
1994). Such an angle is also necessary if a hemilaminectomy has been performed so that the cement can be positioned well away from the spinal cord. Landmarks for lumbar vertebrae are the ventral portion of the base of the
accessory process and the junction of the transverse process with the vertebral body (Rouse and Miller, 1975; Blass
and Seim, 1984; Wong and Emms, 1992; Bagley et al., 2000; LeCouteur and Sturgess, 2003). Entry points should
be between these landmarks and are discussed in 13.55. A skeleton should always be available for reference.
The implant entry point should be no higher than the base of the accessory process. If a flat trajectory is to be used
then the entry point should ideally be no higher than the floor of the vertebral canal otherwise the vertebral canal may
be entered inadvertently (Walker et al., 2002). There is some variation based on the exact vertebra (1.211.23, 1.25,
13.56). Also, due to the concave ventral surface in the middle of each vertebral body, implant placement must be
more precise centrally. A CT scan is invaluable for planning the entry point as well as the trajectory.
The bone cement is applied in a cylindrical pattern on each side of the spine. Wound closure is then much easier than when cement is applied in a doughnut pattern. Implants must be encased fully by the cement (13.57).
The wound is irrigated with saline to dissipate heat generated by the curing of the cement. The spinal cord and
nerves must not contact the cement, especially following hemilaminectomy. Gelfoam (Pharmacia, Kalamazoo,
MI) has good insulating properties and should be used to protect exposed spinal cord (Roosen et al., 1978;
Boker et al., 1989). Excision of epaxial muscle facilitates wound closure (Blass and Seim, 1984). Potential causes
of cement failure include a fold or thin region of the cement bar. Cement can be reinforced with a pin provided
that this is encased fully by the cement (Bagley et al., 2000) (13.25, 13.28, 13.48). Antibiotic should not be added
to the methylmethacrylate unless done in a very carefully controlled fashion as it tends to reduce strength (Ethell
et al., 2000; Weisman et al., 2000). It is preferable to give a second dose of intravenous antibiotic just prior to
bone cement application.
309
310
13.48 A: Counter pressure is

applied by bone clamps on
the spinous processes
during the initial approach,
K-wire fixation of facets
(shown here) and while the
cement hardens (Blass and
13.48
Seim, 1984) (13.24). Note
the torn supraspinous
A
B
ligament (arrow). B: All K-wires
must be incorporated fully into the cement to prevent migration (13.63). If this is not possible then
small screws should be used. In this Pomeranian, three screws and three positive profile pins were
used for fixation; several hairline fractures dictated the implant pattern. The dog (4.38) was almost normal
6 months later.
13.49 Diagrams to
illustrate one
possible implant
pattern. The
preferred pattern is
13.49
to drive implants
perpendicular to
A
the vertebral body
(13.54A), which allows placement of three implants in each vertebra
instead of the two shown here (Garcia et al., 1994). A: Dorsoventral
view. B: Transverse view.
13.50 Steinmann pins must be cut and notched and

the ends then bent to provide purchase for
the bone cement. Positive profile pins are
preferred due to the tendency for smooth pins
to migrate (Wong and Emms, 1992) (13.29).
They also have a roughened surface for better
cement attachment and do not need to be
notched (5.32).
13.50
Trauma
13.51 A: Lateral, and B: ventrodorsal radiograph

of pins and bone cement fixation. This dog
developed a late postoperative infection
(13.34). Alignment is suboptimal. In theory this
dog is at risk for late onset syringomyelia
(Perrouin-Verbe, et al., 1998).
13.51
A
13.52 Range of angles that

can be used for
vertebral implants
(same dog as in
13.10). A 30 angle
from horizontal has
been recommended
as standard (Walker
13.52
et al., 2002;
Wheeler et al., 2002). A
A: Transverse CT scan
B
of T12 vertebra to show the range of possible implant trajectories. The
angles have to be more vertical in thoracic vertebrae because of the ribs. Landmarks for thoracic vertebrae
are the ventral portion of the base of the accessory process and the tubercle of the rib. B: Transverse CT
scan of L2 vertebra to show the range of possible implant trajectories (1.211.23).
311
312
13.53 A: This 3D reconstruction

is from the dog shown in
13.3113.33. The screws
are visible in blue and were
placed using the more
vertical angle shown in
13.49B than the flat angle
shown in 13.54B. B:
Corresponding ventrodorsal
radiograph to show screw
position.
13.53
A
13.54
13.54 Diagrams to show the preferred implant

pattern. A: Dorsoventral view. The advantage of
this pattern is that it makes it easier to place three
implants in each vertebral body (Garcia et al.,
1994). B: Transverse view to show a flat trajectory
for screws. Implants can also be put in using the
more vertical angle shown in 13.49 and 13.53 but
a flatter trajectory provides slightly more bone
purchase as well as more room for cement (13.52).
Trauma
13.55 Illustration of potential

entry points in the
vertebral bodies. A, B:
T12 vertebra. C, D: L2
vertebra. Implants should
enter bone between the
two lines; the higher the
entry point, the more
vertical the trajectory and
vice versa. If further
clarification of anatomy is
needed, the floor of the
vertebral canal can be
palpated with care at the
intervertebral foramen
using a fine, curved
dental instrument.
13.55
A
13.56 The entry points and

implant trajectory will
vary with the anatomical
location of the fixation.
A: Entry points and
potential implant
trajectories are shown on
a CT scan of C6 vertebra
(11.4111.43). The disc
spaces and intervertebral
13.56
foramina should be
avoided (Rouse, 1979;
A
B
Wong and Emms, 1992).
B: Entry points and potential implant trajectories are shown on a CT scan of L6 vertebra. The range of
possible angles is less here than can be used at L2 and is even less for L7 vertebra (1.25). Variations
may also be marked between different types of dog.
313
314
13.57 Fractures or luxations involving

L7 vertebra may
warrant different implant
patterns. Radiographs of this
13.57
dog are shown in 13.27 and
A
B
13.65. A: Implants can be
placed in L7 down the pedicle
and into the base of an articular facet as shown here. Note that there is insufficient cement around the
screws; the technique should also not have been used with a compromised ventral buttress. B: Screw
placement in the sacrum (two most lateral screws on each side) is also discussed on page 204.
Implants can also be placed into the wings of the ilium, but should be placed vertically (13.65) rather
than horizontally (13.21, 14.12) for better bone purchase.
13.58 Bone cement is best applied when it no longer

sticks to the surgeons gloves. It can be used
when more liquid, but then bone wax wrappers
should be used as a mold. Wrappers are first
coated with bone wax to retard cement
adhesion and they must be removed before it
hardens.
13.58
13.59 Vertebral bodies A: before, and B: after

bone cement application (13.33, 13.53,
13.57, 13.60). The cement bar must be of
sufficient thickness; as a rough guide the
bar should be between 1.25 and 2.5 cm in
diameter depending on the weight of the
animal. This usually requires 40 g of cement
for large dogs and 20 g for smaller animals
(Rouse and Miller, 1975; Willer et al., 1991).
Ideally vacuum mixing should be used to
prevent bubble formation and to lessen
exposure of personnel (Anderson, 1988)
(13.33). Steinmann pins of various sizes
can be incorporated into the cement to
reinforce it (13.25). This is useful for large
dogs or at sites subject to excessive
forces (11.17C).
13.59
A
Trauma
13.60 Intraoperative view A: after implant placement,

and B: after bone cement application. Bone
wax has been used to fill the screw heads to
facilitate implant removal if necessary. Kirschner
wires should ideally be placed through the
articular facets once the fracture has been
apposed to maintain alignment (13.28, 13.48).
Bone holding forceps maintain reduction as
cement hardens (13.24, 13.25, 13.48).
13.60
A
Modified segmental fixation (13.6113.63)

The original spinal stapling technique should only be used for cats or for dogs weighing less than 7 kg (Matthieson,
1983) (13.63). The improved modified segmental fixation has been used in dogs over 40 kg (McAnulty et al., 1986)
(13.28B). Both techniques should ideally be used only when the ventral buttress is intact.
13.61 Diagram to show application of modified

segmental fixation. The orthopedic wire
penetrates the base of the articular facet and is
twisted around the Steinmann pins. Wires can
also be placed through the base of each spinous
process.
13.61
315
316
13.62 Close-up of an articular facet joint to show the

position of the wire in modified segmental
fixation. Additional pins are used depending on
the estimated stability of the fracture and on the
weight of the dog (McAnulty et al., 1986).
13.62
13.63 Spinal stapling as shown in this 1.6 kg dog

(13.16) and modified segmental fixation provide
useful alternatives for fixation in the lumbosacral
region (McAnulty et al., 1986). A: The bone
fragment visible at L6/7 could not be reached
after a dorsal laminectomy. B: The dog was
normal and there was good fracture healing
6 weeks later although one K-wire had
migrated. The dog was normal 6 months after
surgery.
13.63
A
External fixation (13.6413.66)

Although dependent on the exact design of the fixator and on the type of implants used, this can be an extremely
reliable method for use in even very large dogs (Walker et al., 2002) (13.65, 14.13). It is also a useful rescue procedure (13.27) and can be used at infected sites (14.12). Placement of pins under fluoroscopic guidance may
give more consistent insertion angles and better bone purchase; a closed technique makes fracture stabilization
during implant insertion of paramount importance (Wheeler et al., 2002).
Trauma
13.64 Reduction of an overridden L7 fracture

using a laminectomy spreader (same dog
as in 13.27) (Bagley et al., 2000). The
positions of the laminectomy spreader jaws
are marked by white arrowheads.
* articular facet of L7. A: Before
reduction. A short dorsal laminectomy has
been made in the center of the caudal
lamina of L7; the laminectomy spreader has
been anchored cranially at this point and
caudally under the roof of S1. B: Fracture
reduced. Torn articular facet joint capsule is
now visible (black arrow).
13.64
A
317
318
13.65 After failure of the implant

shown in 13.27, an
external apparatus was
applied as shown. A:
Lateral view. B:
Dorsoventral view.
Implants in each ilial wing
13.65
are angled from dorsal to
A
B
ventral to obtain better
purchase into soft bone. Skin and lumbodorsal fascia must be pulled towards the midline prior to pin
penetration or there will be excessive tension at wound closure (Shores et al., 1989; Lanz et al., 2000).
13.66 External fixator (SK External Fixation System,

Imex Veterinary Inc., Longview, TX) in place
(14.13). Postoperative management can be
labor intensive. There is also a risk of
traumatic removal; this dog panicked and
removed the entire device as it tried to enter its
kennel during a thunderstorm 5 weeks after
fixation. Incredibly, the dogs neurological
status, which was almost normal, did not
deteriorate (Lanz et al., 2000; Walker et al.,
2002).
13.66
Miscellaneous conditions
Degenerative 319
Degenerative myelopathy (chronic degenerative
radiculomyelopathy, CDRM) 319
Synovial cyst 320
Facet joint pain 320
Sacroiliac joint pain 320
Leukodystrophies 320
Arachnoid cyst 321
Spondylosis deformans 321
Anomalous 321
Congenital vertebral anomalies
Sacrocaudal dysgenesis 322
Spina bifida 322
Tethered spinal cord 322
Spinal dysraphism 322
Syringomyelia and hydromyelia
Cartilaginous exostoses 323
Dermoid sinus 323
Epidermoid cyst 323
Arachnoid cyst 323
Metabolic 326
Lysosomal storage diseases
Osteoporosis 326
Nutritional 326
Hypervitaminosis A
321
331
Vascular 332
Fibrocartilaginous embolism (ischemic
myelopathy) 332
Ascending myelomalacia 332
Ischemic neuromyopathy (aortic embolism, iliac
thrombosis) 332
Spinal cord hemorrhage, hematoma 333
Intermittent claudication 334
334
334
Further reading
337
322
326
326
Infectious/inflammatory 326
Discospondylitis 326
Epidural empyema 328
Inflammatory CNS diseases 329
Epidural steatitis 331
331
14
Psoas muscle injury
References
Idiopathic 326
Tumoral calcinosis 326
Disseminated idiopathic skeletal hyperostosis
(DISH) 326
Trauma
Chapter
This chapter covers various other important conditions of

the spine that are likely to be encountered but are not
covered specifically elsewhere in this book. Surgical treatment is not indicated in many of these conditions, thus
they must be diagnosed correctly to avoid unnecessary
operations (Braund and Sharp, 2003). Key references
are given, but the list is not intended to be encyclopedic.
See Dewey et al. (2003), Braund (Further reading) and
LeCouteur et al. (2000), which are comprehensive and
have excellent bibliographies. The diseases are listed
according to the DAMNIT scheme (Table 3.1).
DEGENERATIVE
Degenerative myelopathy (chronic
degenerative radiculomyelopathy, CDRM)
This is a degenerative condition of the spinal cord of older
dogs; it is mostly seen in large breeds and particularly
German shepherd dogs. A similar disorder has been
described in a cat (Mesfin et al., 1980). There are probably several different types of canine degenerative
myelopathy that are grouped together and cannot be differentiated at present. Phenotypes in the various breeds
of dog may vary somewhat from that of the classic condition found in the German shepherd dog (Matthews et al.,
320
1985; Salinas and Martinez, 1993). Onset of signs in

German shepherd dogs is from 5 years of age. Progressive
pelvic limb ataxia, loss of proprioception, and paraparesis
occur. The neurological examination usually indicates a
T3L3 lesion, but some dogs lose their patellar reflex
because of dorsal nerve root involvement. Urinary and
fecal function is normal. Spinal pain is not seen. Signs in
affected dogs progress over several months, and eventually the thoracic limbs will be involved (Averill, 1973;
Griffiths and Duncan, 1975).
The diagnosis is confirmed by the absence of structural spinal cord disease on myelography or MRI. Even if
other lesions are found, for example a disc extrusion, the
possibility of degenerative myelopathy also being present should be considered (10.53). Analysis of lumbar
CSF often reveals a moderate elevation in protein. The
cause is unknown and there is no known treatment.
Degenerative changes are found at necropsy throughout
the spinal cord and in nerve roots, mainly affecting the
thoracolumbar region. There is a loss of both myelin and
axons, with cellular infiltration in the most severely
affected regions (Johnston et al., 2000).
animals (Cook et al., 2002). Diagnosis is not simple

and depends largely on nerve blocks in humans
(Manchikanti, 1999). Signal intensity changes in lumbar
pedicles on MRI may be a diagnostic tool that is applicable to animals (Morrison et al., 2000).
Synovial cyst
Leukoencephalomalacia is a degenerative CNS disease,

reported in Rottweilers in the USA, the Netherlands and
Australia. There is malacia throughout the spinal cord,
particularly in the cervical region, due to demyelination
and cavitation. The etiology is unclear.
Affected dogs show clinical signs from 1 to 4 years of
age. There is pelvic limb ataxia and hypermetria, progressing through paraparesis to tetraparesis over a period
of months. Proprioceptive deficits are marked. Limb
reflexes are intact or hyperactive.
Differentiation from neuroaxonal dystrophy and other
CNS conditions is important. All routine diagnostic tests
are normal in leukoencephalomalacia. The major differential features are the presence of proprioceptive deficits
in leukoencephalomalacia, and tremor or dysmetric head
movements together with decreased menace responses in
neuroaxonal dystrophy (see below).
There is no treatment and the prognosis is poor;
most dogs with leukoencephalomalacia are euthanized
within 1 year of presentation (Chrisman, 1992).
These cysts arise from the synovial joints between the

articular facets. Often the affected joint(s) also show(s)
degenerative changes on survey radiographs. The cysts
tend to occur mainly in the cervical area of giant-breed
dogs but can be found in the thoracolumbar area
(Dickinson et al., 2001); they have also been reported in a
German shepherd dog presenting with lumbosacral pain
(Webb et al., 2001). Synovial cysts in humans are thought
to be an under-diagnosed problem in the elderly and this
problem will probably be recognized more frequently in
dogs with the increased use of cross-sectional imaging and
especially MRI (Charest and Kenny, 2000). CT myelography and MRI reveal discrete, round, cystic structures
(Levitski et al., 1999a,b; Lipsitz et al., 2001) (11.8).
Imaging by myelography usually shows dorsolateral compression of the spinal cord due to soft tissue rather than
bone. Cysts in the caudal neck may be located more
laterally or even ventrolaterally within the vertebral canal
(4.25, 11.8). Dorsal laminectomy and removal of the cyst
or distractionstabilization (page 219) can be performed.
Prognosis has been good over a mean follow-up of 17
months although it is not known how young dogs do over
a longer period of time (Dickinson et al., 2001). In
humans, synovial cysts are usually associated with facet
joint degenerative disease and so there may be overlap
with facet joint pain (see below) (Howington et al., 1999).
Facet joint pain

Facet joint pain is a common cause of both neck and low
back pain in humans and is a potential source of pain in

This is another common cause of low back pain in
humans that can mimic disc pain and that might have a
counterpart in animals (Pang et al., 1998; Swezey, 1998;
Hodge and Bessette, 1999). CT evidence of osteoarthritis in the joint is suggestive but MRI will probably prove
to be more useful as an indicator of physiological significance (Morrison et al., 2000).
Leukodystrophies
Several breed-specific disorders are described in Afghan
hounds, Dalmatians, Miniature poodles and Dutch
kooiker dogs. The most common examples are seen in
Rottweilers.
ROTTWEILER LEUKOENCEPHALOMALACIA
ROTTWEILER NEUROAXONAL DYSTROPHY

Neuroaxonal dystrophy is a degenerative CNS disease
seen mainly in Rottweilers, but has also been reported
occasionally in cats. Axonal dystrophy with spheroids is
seen in parts of the CNS, including the dorsal horn gray
matter of the spinal cord, and in the gracile, cuneate and
dorsal spinocerebellar tract nuclei. Cerebellar atrophy
may also be seen. The etiology is unknown and there
may be overlap in some dogs between this condition and
leukoencephalomalacia.
Affected dogs show clinical signs from puppyhood,

but these may not be noticed until the dog is adult.
There is pelvic limb ataxia and hypermetria of the thoracic limbs. Proprioception is normal. Limb reflexes are
intact or even hyperactive. Head incoordination, tremor,
positional nystagmus, and loss of menace response (due
to cerebellar involvement) develop later, often after several years. Weakness is not seen.
Differentiation from other CNS conditions is important (see above). Routine diagnostic tests are normal.
There is no treatment and the long-term prognosis is
poor, although affected dogs may survive as active pets
for several years (Chrisman, 1992).
Arachnoid cyst
In an arachnoid cyst there is a focal accumulation of
CSF in the subarachnoid space, which compresses the
spinal cord (Dyce et al., 1991; Skeen et al., 2003). These
can be acquired secondary to some type of injury (13.32)
although they are usually congenital lesions (see below).
14.1 Spondylosis deformans was an incidental finding in this

7-year-old Boxer dog.
Spondylosis deformans
Spondylosis deformans is a common radiographic finding
in older dogs, but it is rarely associated with clinical signs
(Morgan et al., 1989). Generally the osteophytes develop
ventrally and laterally on the vertebral body, and they
may grow to the point that they bridge the intervertebral
space (Larsen and Selby, 1981) (14.1). Osteophytes
around the articular facet joints may be of more significance (see Facet joint pain, above). The thoracolumbar
junction and lumbosacral joint are particularly affected.
Where these changes are seen at the lumbosacral joint
they may be related to clinical signs of lumbosacral disease. However, this diagnosis should not be reached on
the basis of survey radiographs alone, as such changes
are seen in many normal dogs (see page 188). It is
important to differentiate spondylosis deformans from
the changes seen in discospondylitis (14.1114.13).
ANOMALOUS
Congenital vertebral anomalies
Vertebral malformations are common findings in dogs
and are also seen occasionally in cats. Some do cause
compressive myelopathy or are associated with anomalies of the spinal cord (Bailey, 1975; Bailey and Morgan,
1992). Many cases of atlantoaxial subluxation also have
an underlying congenital vertebral malformation (see
Chapter 9). Anomalies such as hemivertebrae, butterfly
vertebrae and block vertebrae are relatively common
but they rarely cause clinical signs. Hemivertebrae are
seen usually in the small, brachycephalic breeds. They
14.2 Dog with severe ataxia and paraparesis due to a hemivertebra at T7. Threaded pins and bone cement were used to stabilize
the site prior to performing a dorsal laminectomy. Neurological function was unchanged postoperatively and the dog was walking
well within a few days.
321
322
are wedge-shaped deformities that may cause spinal

deviation in the lateral or dorsoventral plane depending
on the orientation of the wedge (14.2). When the deviation is marked they tend to cause a chronic, progressive
myelopathy. Skeletal abnormality may be apparent on
physical examination but the exact site should be confirmed by neuroimaging (Dewey et al., 2003). Animals
that have severe spinal cord compression may show a
marked deterioration of neurological deficits after invasive tests such as myelography, so MRI is preferred.
Deterioration is also common when animals with these
lesions are subject to decompressive laminectomy (page
126). Stabilization of the vertebral column should be the
primary aim of surgery and decompression should not be
performed without concomitant fixation (Aikawa, 2001).
Nevertheless, the prognosis is guarded in dogs with
severe neurological deficits. Vertebral anomalies would
be an ideal condition in which to use electrophysiological
monitoring during spinal cord surgery (see Chapter 4,
page 60).
Sacrocaudal dysgenesis
Cats and dogs with congenital tail defects often have vertebral abnormalities of the sacrum and caudal vertebrae.
Manx cats, Pugs and Bulldogs are affected most often
and the condition is inherited in the Manx. The vertebral
abnormalities may themselves cause neurological
deficits that affect the pelvic limbs, lower urinary tract
and anus or there may be malformations of the spinal
cord such as spina bifida. Diagnosis is suspected from the
clinical signs. Radiography will demonstrate vertebral
abnormalities and myelography or MRI may reveal spinal
cord malformations. Treatment is not possible and the
prognosis is poor.
Spina bifida
This is a developmental defect resulting from failure of
the embryonic vertebral arch to fuse normally. There
may be protrusion of the meninges or spinal cord into
a meningocele, a myelocele or a meningomyelocele; this
is termed spina bifida aperta. Alternatively, there may
be no protrusion of nervous tissue, which is termed
spina bifida occulta (Wilson et al., 1979; Wilson, 1982).
There is a high incidence of these conditions in English
bulldogs and Manx cats.
Spina bifida usually involves the caudal lumbar spine.
Clinical signs indicative of L4S3 spinal cord dysfunction
occur. Radiography may reveal defects in the dorsal
vertebral arch, such as paired spinous processes, and
myelography or MRI may demonstrate a meningocele.
Treatment is not possible but untethering may help (see
below). When there is an opening in the skin, closure
should be considered to prevent meningomyelitis.
Tethered spinal cord

This condition is usually associated either with spina
bifida and meningo(myelo)cele or with an intradural
lipoma (Fingeroth et al., 1989; Plummer et al., 1993;
Huttmann et al., 2001; Shamir et al., 2001). It can also
be caused by previous surgery that leaves a residual
defect in the dura mater with subsequent spinal cord
herniation (Henry et al., 1997).
Spinal dysraphism
Malformations of the spinal cord have been described in
several breeds of dogs, particularly Weimaraners (Broek
et al., 1991). Various lesions of the central canal, gray
matter, dorsal sulcus and ventral fissure have been
described and a syringohydromyelia may also be present.
A bunny-hopping pelvic limb gait, abnormalities of the
hair coat, a depression of the sternum, and head tilt may
be seen. Neurological deficits localize to the T3L3 spinal
cord and signs are often evident in affected puppies. The
signs are usually non-progressive. Abnormalities may be
detected using MRI (Dewey et al., 2003).
Syringomyelia and hydromyelia

Syringomyelia and hydromyelia are fluid-filled cavitations
of the spinal cord and central canal respectively.
Syringomyelia can be a congenital or an acquired (11.12)
spinal cord lesion (Cauzinille and Kornegay, 1992;
Perrouin-Verbe et al., 1998), whereas hydromyelia is
more often associated with congenital malformations.
Hydromyelia can also be a complication of myelography
(Kirberger and Wrigley, 1993), and has been described in
a cat secondary to feline infectious peritonitis (FIP)
(Tamke et al., 1988). Clinical differentiation is often
impossible and these two conditions may occur together.
In humans these conditions probably result from overcrowding within the caudal fossa of the skull, such as
occurs in Chiari type-1 malformations. The overcrowding
is thought to obstruct CSF flow and lead to the syringohydromyelia (Nishikawa et al., 1997). Early surgical intervention to decompress the caudal fossa and restore the
flow of CSF in susceptible individuals might prevent the
development of syringomyelia (Fischbein et al., 1999).
The onset of clinical signs is usually in adult dogs.
Diagnosis is aided considerably by MRI (Kirberger et al.,
1997; Levitski et al., 1999b; Taga et al., 2000). Associated
changes such as spinal hyperesthesia, torticollis or scoliosis are common (Child et al., 1986; Bagley et al., 1997;
Dewey et al., 2003). Syringohydromyelia has been
reported in Cavalier King Charles spaniels, which usually
present with a characteristic scratching at the shoulder
region (14.3A). Other signs include apparent neck, thoracic limb or ear pain and thoracic limb lower motor
neuron (LMN) deficits. In these dogs the disorder is
ribs and limb bones. It is caused by abnormal differentiation of cartilage cells in bones that develop by endochondral ossification, leading to the production of large
masses composed of a thin cortex lined by cartilage and
with a core of cancellous bone. The etiology is unclear
(Gambardella et al., 1975; Meomartino et al., 1997).
The masses continue to grow until skeletal maturity is
reached and may also continue thereafter in some dogs.
Diagnosis is by radiography (14.3B). Surgical decompression of compressive lesions may be necessary but the
prognosis is guarded in dogs and is considered to be poor
in cats. Neoplastic transformation can result in clinical
signs in adult animals and this may actually be more
common than was thought previously (Jacobson and
Kirberger, 1996; Dewey et al., 2003).
Dermoid sinus
B
14.3 A: Syringohydromyelia (arrow), mild hydrocephalus and
caudal occipital malformation in a Cavalier King Charles
spaniel with persistent scratching at the shoulder. B: Multiple
cartilaginous exostoses affecting the last rib and L4 vertebra.
thought to develop secondary to compression at the

level of the foramen magnum (Rusbridge et al., 2000).
Anti-inflammatory medications may improve the clinical
signs temporarily but early surgical intervention is
recommended in humans with Chiari type-1 malformations (Rusbridge, 1997; Fischbein et al., 1999). There is
a trend away from shunting of the syrinx in humans
towards decompressing the caudal fossa by an occipital
craniectomy (Sgouros and Williams, 1995; Nishikawa
et al., 1997; Sakamoto et al., 1999). This approach
has also been used with success in a small number of
dogs with syringomyelia (W.B. Thomas, personal communication). When syringomyelia develops following
trauma, a decompressive laminectomy with subarachnoid space reconstruction is recommended (Sgouros and
Williams, 1996).
Cartilaginous exostoses
Cartilaginous exospores (osteochondromatosis) may
cause spinal cord compression at any site of the vertebral
column of dogs and cats. Lesions may also occur on the
In dermoid sinus (pilonidal sinus) the skin over the dorsal midline is inverted and, in some dogs, the invagination communicates with the dura mater. Rhodesian
ridgebacks and Shih tzus have a high incidence
(Tshamala and Moens, 2000). Infection from the cyst
may extend to the spinal cord, causing meningitis and
myelitis with associated clinical signs.
Diagnosis is based on physical examination and clinical signs. If a cyst is suspected to be in communication
with the vertebral canal, myelography or MRI is
preferable to fistulography. Infected lesions are treated
by antibiotics and surgical excision; it may be necessary
to perform a laminectomy to retrieve all the tissue.
Careless exploration of this type of lesion, without a
full appreciation of its extent, can lead to the development of marked neurological deficits.
Epidermoid cyst
This is a rare cystic lesion that arises from entrapment of
epithelial cells within the neural tube. This lesion can be
congenital or it can be acquired secondary to mechanical
implantation of epithelial cells, such as by puncture of
the spinal cord by a needle (Tomlinson et al., 1988).
Myelography or MRI will show an intramedullary
lesion, which is likely to be mistaken for a spinal cord
tumor such as nephroblastoma of young dogs (see
page 249).
Arachnoid cyst
These are not true cysts as they do not have an epithelial
lining (Dyce et al., 1991). Rather there is a focal accumulation of CSF, probably due to adhesions within the
subarachnoid space (Dyce et al., 1991; Moissonier et al.,
2002) (14.6). The natural pulsation of CSF is sufficient
to produce a bulbous or tear-drop enlargement of the
subarachnoid space and this can then cause marked spinal
323
324
cord compression (Dewey and Coates, 2003) (14.4).

Dysraphism or syringomyelia may accompany this condition (Dyce et al., 1991; Galloway et al., 1999). The condition occurs at two main locations: the cervical area or
the thoracolumbar junction. It should be suspected in
a young dog with progressive signs of myelopathy, but
which is pain free. Rottweilers may be over-represented
(Moissonier et al., 2002; Rylander et al., 2002; Skeen
et al., 2003). The condition is also described in cats
(Shamir et al., 1997). Arachnoid cysts can also be
acquired secondary to trauma or disc lesions (Rylander
et al., 2002; Skeen et al., 2003) (13.32). Diagnosis is by

myelography or MRI (Galloway et al., 1999) (14.4).
Sonography is helpful during surgery to provide intraoperative orientation (Galloway et al., 1999). Surgical
decompression may be effective. A dorsal laminectomy is
usually needed in order to gain adequate exposure of the
cyst (11.55, 14.5). Seven of 11 dogs improved in one
series after fenestration and resection of the cyst. However, three dogs with cervical lesions required postoperative ventilatory assistance, two because of hematoma
formation (Rylander et al., 2002). Successful long-term
14.4 Ten-month-old Retriever with progressive paraparesis for 2 months. A: Well-defined, bulbous dilation of the dorsal subarachnoid
space at C2/3. This communicates with the rest of the subarachnoid space as contrast flowed caudally within a few minutes
(14.8B). B: Dorsoventral view confirms the discrete border of this lesion. Surgical appearance of the lesion is shown in 14.8A.
14.5 Seven-month-old Labrador with tetraparesis for 1 month. Myelography revealed a subarachnoid cyst at C2/3. A: The dura
has been opened over the cyst; the wall is still intact and compressing the spinal cord (arrow). B: The cyst wall (arrow) has been
incised to reveal the spinal cord beneath a pool of CSF. The dog was normal 2 years after surgery (Skeen et al., 2003).
14.6 Three-year-old Toy poodle with a cervical arachnoid cyst.

The dura has been opened to reveal an extensive network of
fine adhesions (arrow) in the subarachnoid space. The dura
mater and arachnoid adhesions are being manipulated (arrowhead) with two, modified 25-gauge needles (5.19). This dogs
gait improved but it remained fecally incontinent 4 months later.
It then developed urinary incontinence and its gait
deteriorated 26 months postoperatively (Skeen et al., 2003).
outcomes were obtained in 8 out of another 13 dogs that

were followed for more than 1 year. There were no significant predictors of a good outcome although there was
a trend towards a good outcome in dogs of less than
3 years of age, those that had a duration of signs of less
than 4 months, and when marsupialization was used as
the surgical technique (Skeen et al., 2003).
Marsupialization is one potential technique for treating arachnoid cysts (McKee and Renwick, 1994).
Durotomy is performed and the dural edges are sutured
to periarticular soft tissues using magnification and nonabsorbable suture material (4/0 to 8/0 depending on the
breed). A wide laminectomy gives excellent access but
runs the risk of putting undue tension on the dura, especially at the cranial and caudal ends of the durotomy. If
this happens a release incision should be made (14.7A).
Another technique used to treat arachnoid cysts is to
make a wide fenestration in the dura. Here the dura is
excised over the cyst (14.8) and ideally visible subarachnoid adhesions (14.6) are also removed using meticulous
hemostasis (Frykman, 1999; Rylander et al., 2002). A rare
14.7 A: Marsupialization at T12/13 in a 9-year-old Westie with progressive paraparesis and fecal incontinence for 2 months.
A release incision may be needed if marsupialization puts tension on the dura (arrow). The dog was doing well 22 months after
surgery (Skeen et al., 2003). B: The laminectomy in this dog is not as wide but still gave enough access to suture the dura accurately.
14.8 A: A wide dural fenestration has been made after hemilaminectomy at C2/3. The dog was much improved 24 months after
surgery (Skeen et al., 2003). B: Radiograph taken 10 min after the images shown in 14.4. Contrast has flowed from the previous
point of obstruction (arrow) into the remainder of the subarachnoid space. CSF flow dynamics have been studied in one dog
with an arachnoid cyst and showed normal communication with the subarachnoid space (Moissonier et al., 2002).
325
326
potential complication of not closing a dural incision is

that the spinal cord could herniate through the defect;
this has only been reported in humans (Osterholm,
1974; Henry et al., 1997; Watters et al., 1998).
METABOLIC
Lysosomal storage diseases
Lysosomal storage diseases occur where there is a defect
of metabolism caused by dysfunction in a specific
enzyme pathway. They are relatively common disorders
in dogs and cats and frequently cause neurological signs,
which are usually seen from early in life and are progressive. Most cause signs of intracranial disease but cats
with mucopolysaccharidosis type VI can present with
paraparesis. Neurological deficits are caused by vertebral exostoses that can resemble cartilaginous exostosis
and hypervitaminosis A (Haskins et al., 1980, 1983).
Mucopolysaccharidosis type IIIA in dogs can present as
an adult onset pelvic limb ataxia or hypermetric gait
(Fischer et al., 1998; Jolly et al., 2000). Cytoplasmic
inclusions may be detected in hepatocytes but definitive diagnosis requires specialized techniques. There is
no treatment and the prognosis is poor.
Osteoporosis
This is a common problem in humans, the incidence of
which increases with age as well as in women after menopause or ovariectomy (Riggs, 2002). Other risk factors
include chronic renal disease, type I diabetes, corticosteroid use and Cushings disease (Khanine et al., 2000;
Vestergaard et al., 2002). Vertebral fractures are a common sequel to osteoporosis in humans (Khanine et al.,
2000); diagnosis is made by radiography, scintigraphy,
CT or MRI (Cook et al., 2002; Tan et al., 2002). Even
though vertebral fracture due to osteoporosis is a common problem in humans it is nevertheless underdiagnosed (Gehlbach et al., 2000). Osteoporosis has been
produced experimentally in canine lumbar vertebrae as a
result of ovariectomy or steroid administration (Norrdin
et al., 1990; Yamaura et al., 1993); it has also been recognized as a clinical problem in dogs with Cushings disease
as well as in routine post-mortem studies (Pellegrini et al.,
1979; Huntley et al., 1982; Schleithoff, 1984). Vertebral
fracture secondary to osteoporosis is therefore likely to be
an under-recognized cause of spinal pain in dogs, especially in older, neutered animals with Cushings disease.
14.9 Hypervitaminosis A in a cat showing massive vertebral

exostoses.
parts of the spine and the limbs may also be involved

(Goldman, 1992). Clinical signs related to nerve root and
spinal cord compression are seen. Neck pain and rigidity,
ataxia, paresis of the thoracic limbs, and lameness may
occur. The diagnosis is suspected in a cat fed exclusively
on liver, and is confirmed by radiography (14.9).
Treatment is difficult. Stopping vitamin A intake can
arrest the development of further exostoses, and antiinflammatory drugs may relieve clinical signs.
IDIOPATHIC
Tumoral calcinosis
Also known as calcinosis circumscripta, this condition
has been described as a cause of compressive spinal cord
dysfunction in young dogs. The most common site is
between C1 and C2 but it has also been reported in the
thoracolumbar region (McEwan et al., 1992; de Risio
and Olby, 2000). Diagnosis is by radiography; a mineralized mass is visible at the site of cord compression
(14.10). The cause is not known. Surgical decompression may be successful.
Disseminated idiopathic skeletal

hyperostosis (DISH)
This is thought to be an exaggerated proliferation of bone
as a result of minor stresses, which results in extensive
ossification throughout the body. A characteristic feature
in the spine is a flowing pattern of ventrolateral new
bone formation that extends over at least four adjacent
vertebrae. Clinical signs usually just reflect the mechanical limitations to movement (Morgan and Stavenborn,
1991; Dewey et al., 2003).
NUTRITIONAL
Hypervitaminosis A
INFECTIOUS/INFLAMMATORY
Discospondylitis
Cats fed a diet with excessive vitamin A may suffer from

a skeletal condition characterized by severe exostosis of
the vertebrae, particularly in the cervical spine. Other
Discospondylitis is an inflammatory condition centered

on the intervertebral disc and involving the vertebral end
plates and adjacent bone of the vertebral body. Large
breeds of dogs are affected most often. The condition is

rare in cats (Malik et al., 1990; Watson and Roberts,
1993). The disease is usually caused by Staphylococcus
intermedius but other organisms including Streptococcus
sp., Escherichia coli and Brucella canis are also found.
Fungal infections are uncommon but should be considered particularly in German shepherd dogs; hyphae may
be detected in urine sediment (Butterworth et al., 1995;
Watt et al., 1995; Greene, 1998).
The bacteria usually gain access to the disc by
hematogenous spread from other foci in the body, of
which the bladder is most common. Occasional cases
associated with foreign body migration are seen, usually
grass seeds (awns). Iatrogenic cases following surgery
are also recognized, especially when an implant has been
used (13.34, 14.11, 14.14).
The first clinical sign is usually pain, with concurrent
or later development of neurological deficits. Certain
sites are predisposed: lumbosacral, caudal cervical, thoracolumbar and mid-thoracic vertebrae. Multiple discs
may be involved, either adjacent or distant to the disc
14.10 Tumoral calcinosis in a young Labrador retriever with

progressive paraparesis. The myelogram reveals a large,
mineralized epidural mass located dorsally between C1 and C2.
Following surgical removal of the mass the dog returned to
normal within a few weeks (Lewis and Kelly, 1990).
first identified. Systemic illness is common, typically

with pyrexia, lethargy, inappetence, and dysuria related
to cystitis.
Diagnosis is by radiography (14.11, 14.12). Generally
the radiographic changes are seen clearly, but occasionally the radiographs will be normal even though infection
is present (occult discospondylitis). When the radiographs are inconclusive, a bone scan may reveal the
lesion (Stefanacci and Wheeler, 1991) (see page 59).
Blood or urine cultures are positive in about two thirds
of patients (Kornegay, 1986). Brucella canis titers should
be analyzed in endemic areas.
Treatment of bacterial discospondylitis is initially
by use of antibiotics. In view of the predominance of
S. intermedius, clindamycin, cepazolin or cloxacillin are
the best initial choices. Ampicillin and amoxycillin are
not useful because the organism is often resistant to these
drugs through the mechanism of -lactamase production.
Further treatment is governed by clinical progression
and results of bacteriological testing. Treatment must be
continued for at least 6 weeks, even when the response
is good. Most dogs respond well to antibiotics alone
(Gilmore, 1987), or combined with surgical curettage
(Kornegay and Barber, 1980). It is not known how many
dogs suffer a recurrence of signs.
Treatment for B. canis is with minocycline (25 mg/kg
PO q24h for 2 weeks) and streptomycin (5 mg/kg IM or
SQ q12h for 1 week) or gentamycin (2 mg/kg IM or SQ
q12 h for 1 week) (Greene, 1998). There is potential for
zoonotic spread, and recurrence is common. It may be
wise to castrate male dogs with B. canis infection as the
testes can act as a reservoir of infection (Kerwin et al.,
1992).
Surgical curettage of solitary, readily accessible
lesions often leads to rapid resolution of signs
(Kornegay and Barber, 1980). It also provides material
for culture and promotes blood supply to the affected
disc. Surgical intervention should always be considered
in patients with an unsatisfactory response to antibiotics. The main risk is that surgery could destabilize the
14.11 Ten-year-old Doberman with
S. intermedius discospondylitis at C6/7.
It had undergone total hip replacement
9 months previously, which became
infected and was removed after 4
months. A: Destructive lesion at C6/7.
There is sclerosis and destruction of C6
and C7 end plates (arrowhead). B: CT
scan through caudal C6 confirms bony
destruction within the vertebral body
(arrowhead).
327
328
14.12 Rottweiler with uncontrollable

pain due to discospondylitis at L7/S1
that did not respond to 1 week of
antibiotics and opioid analgesics. An
external fixator was applied and the
dog was pain-free the next day. It
was normal 3 months later although
L7/S1 had not fused; no bone graft
had been used at surgery.
14.13 A: The dog shown in 14.12 at 6-week follow-up. B: Dorsal view of the external fixator (SK External Fixation System, Imex
Veterinary Inc., Longview, TX). A potential limiting factor in large dogs is the length of pin; these were only just long enough to attach
the clamps.
affected intervertebral space further; loss of stability

can also occur spontaneously in animals with advanced
bony destruction. Subluxation or collapse of the intervertebral space then results together with severe pain
and neurological deficits. When collapse occurs in the
lumbosacral area the animal often develops intractable
pain (14.12). Fixation of the spine is then best provided by an external device that can be removed once
the infection has resolved and the site has stabilized
(Auger et al., 2000; Walker et al., 2002). In the lumbosacral area threaded pins or screws may be driven
through the articular facets into the sacrum (McKee
et al., 1990). Pins or bone cement have also been used
successfully in the face of infection (Lavely et al.,
2002). An autogenous bone graft placed into the
affected interspace probably enhances complete fusion
(Auger et al., 2000) although this does not seem to be
crucial for a successful outcome (McKee et al., 1990)
(14.13). Decompression of discospondylitis lesions by
laminectomy without stabilization is not recommended

as it may lead to marked instability.
In addition to instability and collapse of the disc
space, another serious complication of discospondylitis
is the development of epidural abscessation (see below)
(Lobetti, 1994; Remedios et al., 1996).
Fungal discospondylitis is a systemic infection that
occurs in German shepherd dogs and occasionally in
other breeds of dog; this disorder is not restricted to tropical or semi-tropical regions as was initially thought. It is
caused usually by one of several Aspergillus species and
has a poor prognosis. Diagnosis can often be obtained by
identifying fungal hyphae in urine sediment. Itraconazole
may control the disease but the long-term prognosis is
poor (Butterworth et al., 1995; Watt et al., 1995).
Epidural empyema
This is an emergency situation that requires rapid diagnosis and treatment (14.14). Clinical signs include fever,
14.14 A: Doberman that underwent a ventral slot and distraction stabilization using screws and bone cement for a dynamic lesion
at C5/6. The dog did well for 4 days but then became febrile and suffered progressive neurological deterioration. B: A repeat
myelogram 2 weeks after surgery revealed small erosions of the C6/7 end plates and a diffuse epidural mass extending from the
mid-body of C5 vertebra to mid-C7 (arrowheads). The dog died; epidural abscess was found at necropsy.
lethargy, pain and neurological deficits along with

peripheral neutrophilia and a neutrophilic CSF pleocytosis. Less than half of dogs described have had concomitant discospondylitis. Myelography can provide an
accurate diagnosis but is no longer recommended in
humans if MRI is available because the latter permits
earlier identification of lesions (Angtuaco et al., 1987;
Reihsaus et al., 2000). Lumbar puncture is also no longer
recommended in humans as it carries the risk of introducing infection into the subarachnoid space. Surgical
decompression and drainage together with antibiotics
are the treatment of choice in humans. The biggest problem is one of early diagnosis before massive neurological
damage has occurred (Reihsaus et al., 2000). Results in
dogs have been mixed. One study reported a good outcome in five of seven dogs whereas others report a poor
outcome even after aggressive therapy (Dewey et al.,
1998; Lavely et al., 2002).
Inflammatory CNS diseases

STEROID-RESPONSIVE MENINGITIS
ARTERITIS
Several aseptic meningitis syndromes have been described
in dogs. Meningitis and polyarteritis has been described in
a colony of young research Beagles; Bernese mountain
dogs also suffer from a similar syndrome (Meric, 1988).
These disorders are similar to necrotizing vasculitis of the
meningeal arteries (Meric et al., 1986).
The clinical signs are typical of meningitis, with depression, cervical pain and stiff gait; pyrexia is also seen in this
condition. The disease may be acute or it can have a
relapsing pattern. Analysis of CSF reveals marked pleocytosis (with CSF white blood cell counts in the hundreds or
even thousands per microliter), mainly comprising neutrophils along with an increase in protein concentration.
Infectious agents are not seen in the CSF and culture is
negative; an immune-mediated mechanism is suspected.
The CSF may be relatively normal between bouts of the
disease.
Treatment with long-term corticosteroids is recommended (prednisolone 24 mg/kg/day initially, reducing
to anti-inflammatory doses) until the clinical signs and
CSF pleocytosis resolve. The prognosis is fair, although
some dogs experience relapses. Older dogs with high
IgA levels in their CSF require a longer duration of corticosteroid therapy and have a less favorable prognosis
(Tipold and Jaggy, 1994; Cizinauskas et al., 2000).
CANINE DISTEMPER VIRUS INFECTION

Canine distemper virus (CDV) infection is the most
common infectious cause of neurological disease in dogs.
Demyelination and inflammation occur in various sites
throughout the CNS. The virulence of the virus strain
and the immunocompetence of the dog are important
factors in determining the severity of disease. Other
CNS infections may be seen in association with the
immunosuppression related to CDV. Systemic signs of
disease may occur, although this is not a consistent feature (Tipold et al., 1992; Thomas et al., 1993). The neurological signs may be multifocal although in one third of
cases they suggest a focal lesion. Signs may be acute in
onset. CSF analysis usually reveals mild to moderate,
lymphocytic pleocytosis (see page 45). Diagnostic methods include testing for local production of antibody in
the CSF; fluorescent antibody testing of conjunctival
cells, CSF leukocytes, or skin biopsy; or for viral RNA by
329
330
polymerase chain reaction (Tipold et al., 1992; Thomas

et al., 1993; Haines et al., 1999; Sharp et al., 2000).
Confirmation of a diagnosis of CDV infection is difficult
ante mortem. Skin biopsy is easy and may be the most
accurate method (Haines et al., 1999).
Treatment is restricted to managing the clinical signs
and providing supportive care. Prognosis is guarded.
FELINE INFECTIOUS PERITONITIS VIRUS

INFECTION (FIP)
Neurological signs, including spinal cord syndromes, may
be seen in cats with FIP virus infection. Multisystemic
signs, particularly ocular involvement, are seen in most
affected cats. Neurological signs are most often associated with the dry form of the disease. CSF is usually
abnormal with increased protein and a mixed, predominantly neutrophilic pleocytosis (Foley et al., 1998) (see
page 45). There is no definitive treatment and the prognosis is poor (Kline et al., 1994).
OTHER INFECTIOUS AGENTS
GRANULOMATOUS
MENINGOENCEPHALOMYELITIS (GME)
This is an inflammatory disease of unknown etiology; any
part of the CNS may be involved and the lesions can
either be focal or diffuse in nature. The retina and optic
nerves may also be affected. Perivascular accumulations
of mononuclear cells are present throughout the CNS,
14.15 T2-weighted MRI of a 9-year-old Boxer dog with an

extensive high signal area (arrow) in the cervical spinal cord.
Histopathological diagnosis was GME.
which can lead to the formation of large granulomas

(Cordy, 1979).
Clinical signs are typical of meningoencephalomyelitis;
onset is usually between 3 and 7 years of age. Spinal cord
syndromes may be seen alone or as part of a multifocal
presentation. The signs of spinal disease may be restricted
to spinal hyperesthesia or may include neurological
deficits. The course is usually chronic, but some cases
show a rapid decline.
Analysis of CSF reveals moderate, mainly mononuclear pleocytosis with increased protein, but the
findings are non-specific. Imaging in dogs with spinal
cord involvement may reveal an intramedullary lesion
(14.15). Treatment with immunosuppressive doses of
corticosteroids may lead to improvement (prednisolone
12 mg/kg/day) (Munana and Luttgen, 1998). The longterm prognosis used to be poor, although excellent longterm results have now been obtained after treatment
with cytosine arabinose (Cuddon et al., 2002).
Meningomyelitis due to other agents is relatively rare in

dogs and cats; infections are usually associated with brain
involvement and therefore present with multifocal neurological signs. Many organisms have been implicated,
including bacteria, rickettsial, fungal and helminth
organisms; the distribution of some of these is regional
(Munana, 1996; Stiles, 2000; Gasser et al., 2001). Some
of the organisms that may affect the spinal cord in dogs
and cats are listed in Table 14.1.
The signs are typical of inflammatory CNS disease as
described above. Some infections show mainly signs of
intracranial disease, typically rickettsial infections when
seizures, dullness, depression, and vestibular signs occur.
Pelvic limb hyperextension is often seen in young dogs
with toxoplasmosis or neosporosis (Nesbit et al., 1981).
Systemic signs of infection may be apparent, for example gastrointestinal disturbance or respiratory signs,
which may indicate the portal of entry into the CNS.
Patients may be pyrexic, but this is variable.
Confirmation of the diagnosis is usually by CSF
analysis. Pleocytosis is variable, but very high cell counts
(in the thousands per microliter) may be seen, mostly
Table 14.1 Known organisms associated with spinal cord disease

Viruses
Bacteria
Rickettsia
Fungi
Protozoa
Helminth
Canine distemper
Feline infectious
peritonitis
Rabies
Staphylococcus sp.
Brucella canis
Rickettsia sp.
Ehrlichia sp.
Bartonella sp.
Cryptococcus neoformans
Coccidioides imitis
Blastomycoides dermatidis
Histoplasma capsulatum
Toxoplasma gondii Baylisascaris sp.

Neospora caninum
composed of neutrophils; eosinophils may also be present.

Organisms may be seen in the CSF. Culture may be
attempted but is often unrewarding. Serological testing may be useful in some infections, for example in
neosporosis, rickettsial disease or cryptococcosis.
Those rare patients where bacterial infections are identified should be treated with antibiotics. The bacterial
sensitivity and CNS penetration of the drug must be considered in choosing an antibiotic (Greene, 1998). Fungal
infections are difficult to treat, but amphotericin B, itraconazole and fluconazole may be effective. Rickettsial
infections are treated with tetracyclines, preferably doxycycline or chloramphenicol. Protozoal infections may be
treated with clindamycin, azithromycin or trimethoprimsulfonamide combined with pyrimethamine (BoschDriessen et al., 2002).
The use of corticosteroids in meningomyelitis is controversial. They are contraindicated in fungal infection,
but may have a role in acute bacterial infections of the
CNS (Vandecasteele et al., 2001; Gijwani et al., 2002).
The potential consequences of corticosteroid use must
be considered before their administration. Their use at
the same time as antibiotics can also make it difficult to
determine which is responsible for an improvement in
the animals condition.
Epidural steatitis
This is a common complication to epidural empyema but
could also occur secondary to infection with Brucella,
Bartonella or Mycobacterium sp. (Postacchini and
Montanaro, 1980; Nas et al., 2001). Epidural fibrosis and
fat necrosis are also seen secondary to vascular compromise in lumbosacral disease (see page 190).
TRAUMA
The most common disorders are discussed on pages
30 and 283. Brachial plexus injuries are a common
consequence of trauma and it is imperative that they be

recognized on admission of the patient (2.15, 2.26,
2.28, 13.1, 14.16B).
Re-implantation of avulsed ventral nerve rootlets into
the cervical spinal cord has recently been proposed as a
new treatment for avulsion injuries in dogs and cats
(Moissonnier et al., 1996, 2001). Although an effective
treatment is needed desperately, surgical repair is
unlikely to be useful for most animals with plexal
injuries due to the difficulty in producing an early
return of function to the distal limb and the likelihood
of residual sensory deficits (Rodkey and Sharp, 2003).
Trauma can occasionally produce neurological deficits
by affecting the blood supply to the spinal cord, peripheral nerves or muscles. Acute pelvic limb paralysis is
described in cats after severe abdominal trauma. It is
thought to be a vasospasm or thrombosis of radicular
blood vessels (Summers et al., 1994). Paralysis of a
thoracic limb has been described after injury to the subclavian artery in a dog (MacCoy and Trotter, 1977).
Gunshot injuries may involve the vertebral column or
spinal cord and are usually devastating (Fullington and
Otto, 1997).
Psoas muscle injury

This unusual condition can cause acute onset of pelvic
limb lameness with severe pain on extension of the hip
and palpation of the psoas muscles along with reluctance to stand (Krawiecki and Puignero, 1995; Breur
and Blevins, 1996). If inflammation and hemorrhage is
severe there may also be unilateral femoral nerve
deficits (N. Olby, personal communication). The clinical signs are caused by an injury or a hematoma within
the psoas muscle; femoral neuropathy is due to compression within the muscle compartment (Robinson
et al., 2001). Diagnosis is made by ultrasound or MRI.
Most dogs improve with rest but a tenomyectomy may
be necessary to resolve the pain.
14.16 A: Post-mortem appearance of a

brachial plexus avulsion. Note the absent
nerve roots and rootlets on the side with
subdural hemorrhage. B: This dog was
paraplegic due to an L1 vertebral fracture;
its brachial plexus avulsion was almost
overlooked before repairing the vertebral
fracture.
331
332
VASCULAR
Fibrinocartilaginous embolism
(ischemic myelopathy)
Fibrocartilaginous embolism (FCE) is a syndrome of
acute, severe neurological dysfunction of dogs. It is an
important differential diagnosis in cases of disc disease,
trauma, and other acute spinal conditions (Cauzinille
and Kornegay, 1996). The emboli are composed of fibrocartilage, identical to the material within the nucleus
pulposus. In pathological studies of FCE, emboli are
found in the vasculature of the spinal cord substance or
nerve roots. The exact mechanism by which they gain
access to these areas is not clear.
Adult dogs of large and giant breeds are affected most
often along with Miniature schnauzers, but this condition can affect any dog. FCE can occur in dogs as early
as 8 weeks of age, although it is seen most often in
adults between 3 and 7 years (Cauzinille and Kornegay,
1996; Junker et al., 2000; Hawthorne et al., 2001). The
condition appears to be very rare in cats (Scott and
OLeary, 1996; Abramson et al., 2002).
Peracute, severe neurological presentations occur,
often following vigorous exercise or mild trauma.
Owners may note a progression of the signs over a
period of several hours, often from an initial lameness to
eventual paralysis. However, progression beyond the
first 24 h is very unusual and this helps to differentiate
the condition from other myelopathies. Spinal hyperesthesia is not usually present on clinical examination,
but sometimes severe discomfort is apparent during
the development of the condition and this can persist
for several hours after onset (Cauzinille and Kornegay,
1996). Any part of the spinal cord may be affected, and
often the signs are markedly asymmetrical.
Confirmation of the diagnosis is by elimination of

other causes. Myelography or MRI should be performed
to rule out compressive spinal cord lesions (14.17).
Generally, the myelogram is normal, but in some cases of
FCE, it will show an intramedullary pattern of spinal
cord swelling. MRI reveals spinal cord edema and
swelling in the acute stage. Changes in CSF are often
non-specific although xanthochromia is very suggestive
of FCE (4.1).
Treatment is by supportive care of the patient. Use
of methylprednisolone sodium succinate (MPSS), as in
spinal trauma, may be useful in the first few hours following onset (see page 83). Other corticosteroids are
not indicated.
The prognosis is variable and dependent largely on the
size of the area that undergoes necrosis. Discrete upper
motor neuron (UMN) lesions often improve, but loss of
nociception or extensive LMN deficits carry a poor prognosis (Cauzinille and Kornegay, 1996). Dogs that recover
may be left with residual neurological deficits unless the
lesion is small (Hawthorne et al., 2001).
Ascending myelomalacia
This is probably an ischemic lesion induced by the
diffuse spread of extruded disc material; it is illustrated in 8.5 and 14.18 and discussed on page 128.
Extensive malacia may also be evident following
trauma (13.10).
Ischemic neuromyopathy (aortic

embolism, iliac thrombosis)
Ischemic neuromyopathy is a common cause of acute
paraplegia in cats. Involvement of either one pelvic limb,
14.17 10-year-old cat that was normal prior to an acute onset of symmetrical tetraparesis localizing to the C6T2 spinal cord. T2weighted MRIs show high signal within the spinal cord (arrows) that extends A from the level of C2 to C7 vertebral bodies and that
B occupies much of the cross-sectional area of the spinal cord. Histological diagnosis was FCE.
or of one thoracic limb, has been reported in 13 and 14%

of cats, respectively (Smith et al., 2003). Hypothermia,
tachypnea and signs of cardiac disease are common in
affected cats and renal or gastrointestinal dysfunction
may also be apparent. Affected limbs are usually
arreflexic, hypothermic and hypalgesic; femoral pulses
are often absent and the gastrocnemius muscles
swollen and painful in pelvic limb involvement. The nail
beds are characteristically cyanotic and the toes do not
bleed after needle prick. Diagnosis is based on the
clinical signs (Griffiths and Duncan, 1979; Laste and
Harpster, 1995). Pre-existing cardiomyopathy underlies
the thromboembolic episode, but the presence of a
thrombus does not entirely explain the clinical signs.
There also appears to be a failure of collateral circulation
caused by release of vasoactive substances from the area
of the thrombus.
The prognosis for long-term survival is guarded. Rectal

temperature of 99F (37.22C) or higher was the best
predictor of survival to discharge; 76% of 39 cats with a
temperature of 99F (37.22C) or more survived to be
discharged (Smith et al., 2003). Median long-term survival of 44 cats was 117 days. Recurrence occurred in 11
cats (25%). Cats with congestive heart failure at presentation had significantly shorter survival times than cats
without heart failure (77 days vs 223 days). Low-dose
aspirin therapy is recommended as an inexpensive, safe
treatment that is at least as effective as high-dose aspirin
or warfarin (Smith et al., 2003).
This condition also occurs in dogs although it is less
common than in cats. The presenting signs are also distinct from those in cats; the onset tends to be gradual
and most dogs do not appear to be in discomfort.
Hyperadrenocorticism, neoplasia and cardiac disease
are predisposing factors (van Winkle et al., 1993).
The prognosis depends on the underlying cause;
the prognosis for dogs surviving the acute episode is
favorable (Boswood et al., 2000).
Spinal cord hemorrhage, hematoma
14.18 Typical myelographic appearance of a dog with

myelomalacia secondary to intervertebral disc extrusion.
There is infiltration of contrast material into the parenchyma
of the spinal cord (arrow) (Lu et al., 2002).
Hemorrhage and hematoma formation may occur within

the spinal cord, in the subarachnoid space or in the
epidural space. Epidural hemorrhage is a common complication of acute thoracolumbar disc extrusion (Olby et
al., 2000). Hemorrhage can also cause marked deterioration after ventral slot surgery for cervical disc extrusion
(Seim and Prata, 1982). Bleeding can occur with various
coagulopathies such as hemophilia, rickettsial disease,
trauma, anticoagulant toxicosis, or as a complication to
any severe bleeding disorder (Stokol et al., 1994). If spinal
hemorrhage is suspected in the absence of trauma, tests
for clotting function should be performed. Analysis of CSF
may reveal xanthochromia (4.1) or erythrophagocytosis.
14.19 T2-weighted MRI of a 9-month-old dog with neck pain and tetraplegia for one day caused by a hematoma (arrowheads).
Platelet numbers, coagulation profile and a buccal mucosal bleeding time were normal. The dog failed to improve and was
euthanized after 3 weeks; necropsy showed that the hematoma was subarachnoid; no cause was identified. A: Sagittal image.
The dotted line shows the level of the transverse image. B: Transverse image. For MRI of blood, see page 58. C: Necropsy.
333
334
If fresh blood is seen in a CSF sample, it is most likely to

be caused by puncture of a dural vessel rather than
reflecting an underlying blood disease.
Intramedullary hematoma formation is an unusual
cause of tetraparesis in young, large-breed dogs (Martin
et al., 1986) (14.19). The cause is not known; there may
or may not be an associated coagulopathy. Neck pain is a
variable finding. Surgical decompression is indicated
in compressive lesions not associated with clotting
disorders. Any underlying coagulopathy must also be
addressed.
This is recognized in both humans and dogs with
lesions in the lumbar and lumbosacral regions (Tarvin
and Prata, 1980; Markwalder, 1993; Porter, 1996). It is
characterized by circulatory impairment of neural
structures that occurs secondary to exercise. Intermittent claudication has also been recognized in humans
with cervical or thoracic spinal cord compression where
it can complicate cervical spondylotic myelopathy
(Kikuchi et al., 1996).
1. Is decompression indicated for severe vertebral
deformities and if so, should it be performed at the same
time as stabilization or at a second surgery?
2. What is the long-term recurrence rate following medical
treatment for discospondylitis? Is this influenced by the
duration of antibiotic therapy?
3. Is MPSS of any value in FCE?
4. Is any treatment regime useful in degenerative myelopathy?
REFERENCES
Abramson, C.J., Platt, S.R., Stedman, N.L. (2002) Tetraparesis in a cat
with fibrocartilaginous emboli. Journal of the American Animal Hospital
Aikawa, T. (2001) Surgical management of vertebral deformity in four
dogs. Journal of the American Animal Hospital Association (Japan) 16,
3944.
Angtuaco, E.J., McConnell, J.R., Chadduck, W.M., Flanigan, S. (1987) MR
imaging of spinal epidural sepsis. American Journal of Roentgenology
149, 12491253.
Averill, D.R., Jr (1973) Degenerative myelopathy in the aging German
Shepherd dog: clinical and pathologic findings. Journal of the American
Bagley, R.S., Silver, G.M., Seguin, B., Lincoln, J.D., Britt, L.G. (1997)
Scoliosis and associated cystic spinal cord lesion in a dog. Journal of the
Bailey, C.S. (1975) An embryological approach to the clinical significance

of congenital vertebral and spinal cord abnormalities. Journal of the
Bailey, C.S., Morgan, J.P. (1992) Congenital spinal malformations. Veterinary
Bosch-Driessen, L.H., Verbraak, F.D., Suttorp-Schulten, M.S., van
Ruyven, R.L., Klok, A.M., Hoyng, C.B., Rothova, A. (2002) A prospective,
randomized trial of pyrimethamine and azithromycin vs pyrimethamine
and sulfadiazine for the treatment of ocular toxoplasmosis. American
Journal of Ophthalmology 134, 3440.
Boswood, A., Lamb, C.R., White, R.N. (2000) Aortic and iliac thrombosis
in six dogs. Journal of Small Animal Practice 41, 109114.
Braund, K.G., Sharp, N.J.H. (2003) Neurological examination and localization. In: D. Slatter (ed.) Textbook of Small Animal Surgery, 3rd edn,
10921107. Elsevier Science, Philadelphia.
Breur, G.J., Blevins, W.E. (1996) Traumatic injury of the iliopsoas muscle
in 25 dogs. Veterinary Surgery 25, 262.
Broek, A.H.M.v.d., Else, R.W., Abercromby, R., France, M. (1991) Spinal
dysraphism in the Weimaraner. Journal of Small Animal Practice 32,
258260.
Butterworth, S.J., Barr, F.J., Pearson, G.R., Day, M.J. (1995) Multiple discospondylitis associated with Aspergillus species infection in a dog.
Cauzinille, L., Kornegay, J.N. (1992) Acquired syringomyelia in a dog.
Cauzinille, L., Kornegay, N. (1996) Fibrocartilaginous embolism of the
spinal cord in dogs: review of 36 histologically confirmed cases and retrospective study of 26 suspected cases. Journal of Veterinary Internal
Medicine 10, 241245.
Charest, D.R., Kenny, B.G. (2000) Radicular pain caused by synovial cyst: an
underdiagnosed entity in the elderly? Journal of Neurosurgery 92
(1 Suppl), 5760.
Child, G., Higgins, R.J., Cuddon, P.A. (1986) Acquired scoliosis associated
with hydromyelia and syringomyelia in two dogs. Journal of the
Chrisman, C.L. (1992) Neurological diseases of Rottweilers: neuroaxonal
dystrophy and leukoencephalomalacia. Journal of Small Animal Practice
33, 500504.
Cizinauskas, S., Jaggy, A., Tipold, A. (2000) Long-term treatment of dogs
with steroid-responsive meningitis-arteritis: clinical, laboratory and therapeutic results. Journal of Small Animal Practice 41, 295301.
Cook, G.J., Hannaford, E., See, M., Clarke, S.E., Fogelman, I. (2002) The
value of bone scintigraphy in the evaluation of osteoporotic patients with
back pain. Scandinavian Journal of Rheumatology 31, 245248.
Cordy, D.R. (1979) Canine granulomatous meningoencephalomyelitis.
Veterinary Pathology 16, 325333.
Cuddon, P.A., Coates, J.R., Murray, M. (2002) New Treatments for
Granulomatous Meningoencephalomyelitis. Proceedings of the ACVIM
Veterinary Forum 2002; 319321. American College of Veterinary Internal
Medicine, Dallas, TX.
de Risio, L., Olby, N.J. (2000) Tumoral calcinosis of the thoracic spine:
Veterinary
Neurology
Neurosurgery.
http://www.neurovet.org/
DeRisioandOlby1/Tumoralcalcinosis.htm
Dewey, C.W., Hoffman, A., Coates, J.R. (2003) Miscellaneous spinal conditions. In: D. Slatter (ed.) Textbook of Small Animal Surgery, 3rd edn,
Dewey, C.W., Kortz, G.D., Bailey, C.S. (1998) Spinal epidural empyema
in two dogs. Journal of the American Animal Hospital Association 34,
305308.
Dickinson, P.J., Sturges, B.K., Berry, W.L., Vernau, K.M., Koblik, P.D.,
LeCouteur, R.A. (2001) Extradural spinal synovial cysts in nine dogs.
Dyce, J., Herrtage, M.E., Houlton, J.E.F., Palmer, A.C. (1991) Canine
spinal arachnoid cysts. Journal of Small Animal Practice 32, 433437.
Fingeroth, J.M., Johnson, G.C., Burt, J.K., Fenner, W.R., Cain, L.S. (1989)
Neuroradiographic diagnosis and surgical repair of tethered cord syndrome in an English Bulldog with spina bifida and myeloschisis. Journal
Fischer, A., Carmichael, K.P., Munnell, J.F., Jhabvala, P., Thompson, J.N.,
Matalon, R., Jezyk, P.F., Wang, P., Giger, U. (1998) Sulfamidase deficiency in a family of Dachshunds: a canine model of mucopolysaccharidosis IIIA (Sanfilippo A). Pediatric Research 44, 7482.
Foley, J.E., Lapointe, J.M., Koblik, P., Poland, A., Pedersen, N.C. (1998)
Diagnostic features of clinical neurologic feline infectious peritonitis.
Frykman, O.F. (1999) Spinal arachnoid cyst in four dogs: diagnosis, surgical treatment and follow-up results. Journal of Small Animal Practice 40,
544549.
Fullington, R.J., Otto, C.M. (1997) Characteristics and management of
gunshot wounds in dogs and cats: 84 cases (19861995). Journal of the
Galloway, A.M., Curtis, N.C., Sommerlad, S.F., Watt, P.R. (1999)
Correlative imaging findings in seven dogs and one cat with spinal arachnoid cysts. Veterinary Radiology and Ultrasound 40, 445452.
Gambardella, P.C., Osborne, C.A., Stevens, J.B. (1975) Multiple cartilaginous exostoses in the dog. Journal of the American Veterinary Medical
Gasser, A.M., Birkenheuer, A.J., Breitschwerdt, E.B. (2001) Canine rocky
mountain spotted fever: a retrospective study of 30 cases. Journal of the
Gehlbach, S.H., Bigelow, C., Heimisdottir, M., May, S., Walker, M.,
Kirkwood, J.R. (2000) Recognition of vertebral fracture in a clinical setting. Osteoporosis International 11, 577582.
Gijwani, D., Kumhar, M.R., Singh, V.B., Chadda, V.S., Soni, P.K., Nayak, K.C.,
Gupta, B.K. (2002) Dexamethasone therapy for bacterial meningitis in
adults: a double blind placebo control study. Neurology India 50, 6367.
Gilmore, D.R. (1987) Lumbosacral diskospondylitis in 21 dogs. Journal of
Goldman, A.L. (1992) Hypervitaminosis A in a cat. Journal of the
Greene, C.E. (1998) Infectious Diseases of the Dog and Cat, 2nd edn.
Griffiths, I.R., Duncan, I.D. (1975) Chronic degenerative radiculomyelopathy in the dog. Journal of Small Animal Practice 16, 461471.
Griffiths, I.R., Duncan, I.D. (1979) Ischaemic neuromyopathy in cats.
Haines, D.M., Martin, K.M., Chelack, B.J., Sargent, R.A., Outerbridge, C.A.,
Clark, E.G. (1999) Immunohistochemical detection of canine distemper
virus in haired skin, nasal mucosa, and footpad epithelium: a method for
antemortem diagnosis of infection. Journal of Veterinary Diagnostic
Investigation 11, 396399.
Haskins, M.E., Aguirre, G.D., Jezyk, P.F., Patterson, D.F. (1980) The
pathology of the feline model of mucopolysaccharidosis VI. American
Journal of Pathology 101, 657674.
Haskins, M.E., Bingel, S.A., Northington, J.W., Newton, C.D., Sande, R.D.,
Jezyk, P.F., Patterson, D.F. (1983) Spinal cord compression and hindlimb
paresis in cats with mucopolysaccharidosis VI. Journal of the American
Hawthorne, J.C., Wallace, L.J., Fenner, W.R., Waters, D.J. (2001)
Fibrocartilaginous embolic myelopathy in miniature schnauzers. Journal
of the American Animal Hospital Association 37, 374383.
Henry, A., Tunkel, R., Arbit, E., Ku, A., Lachmann, E. (1997) Tethered
thoracic cord resulting from spinal cord herniation. Archives of Physical
Medicine and Rehabilitation 78, 530583.
Hodge, J.C., Bessette, B. (1999) The incidence of sacroiliac joint disease in
patients with lowback pain. Canadian Association of Radiologists
Journal 50, 321323.
Howington, J.U., Connolly, E.S., Voorhies, R.M. (1999) Intraspinal synovial cysts: 10-year experience at the Ochsner Clinic. Journal of
Neurosurgery 91 (2 Suppl), 193199.
Huntley, K., Frazer, J., Gibbs, C., Gaskell, C.J. (1982) The radiological
features of canine Cushings syndrome: a review of forty-eight cases.
Huttmann, S., Krauss, J., Collmann, H., Sorensen, N., Roosen, K. (2001)
Surgical management of tethered spinal cord in adults: report of 54
cases. Journal of Neurosurgery 95 (2 Suppl), 173178.
Jacobson, L.S., Kirberger, R.M. (1996) Canine multiple cartilaginous exostoses: unusual manifestations and a review of the literature. Journal of
Johnston, P.E.J., Barrie, J.A., McCulloch, M.C., Anderson, T.J., Griffiths, I.R.
(2000) Central nervous system pathology in 25 dogs with chronic degenerative radiculomyelopathy. Veterinary Record 146, 629633.
Jolly, R.D., Allan, F.J., Collett, M.G., Rozaklis, T., Muller, V.J., Hopwood, J.J.
(2000) Mucopolysaccharidosis IIIA (Sanfilippo syndrome) in a
New Zealand Huntaway dog with ataxia. New Zealand Veterinary
Journal 48, 144148.
Junker, K., van den Ingh, T., Bossard, M.M., van Nes, J.J. (2000)
Fibrocartilaginous embolism of the spinal cord [FCE] in juvenile Irish
Wolfhounds. Veterinary Quarterly 22, 154156.
Kerwin, S.C., Lewis, D.D., Hribernik, T.N., Partington, B., Hosgood, G.,
Eilts, B.E. (1992) Diskospondylitis associated with Brucella canis infection in dogs: 14 cases (19801991). Journal of the American Veterinary
Khanine, V., Fournier, J.J., Requeda, E., Luton, J.P., Simon, F., Crouzet, J.
(2000) Osteoporotic fractures at presentation of Cushings disease: two
case reports and a literature review. Joint Bone Spine 67, 341345.
Kikuchi, S., Watanabe, E., Hasue, M. (1996) Spinal intermittent claudication
due to cervical and thoracic degenerative spine disease. Spine 21, 313318.
Kirberger, R.M.M., Wrigley, R.H. (1993) Myelography in the dog.
Kirberger, R.M., Jacobson, L.S., Davies, J.V., Engela, J. (1997) Hydromyelia
in the dog. Veterinary Radiology and Ultrasound 38, 3038.
Kline, K.L., Joseph, R.J., Averill, D.R. (1994) Feline infectious peritonitis
with neurologic involvement: clinical and pathological findings in 24
cats. Journal of the American Animal Hospital Association 30, 111118.
Kornegay, J.N. (1986) Discospondylitis. In: R.W. Kirk (ed.), Current
Veterinary Therapy, 810814. Philadelphia: WB Saunders.
Kornegay, J.N., Barber, D.L. (1980) Diskospondylitis in dogs. Journal of the
Krawiecki, J.M., Puignero, V. (1995) Avulsion of the tubercle of the psoas
minor muscle in a dog. Point Veterinaire 26, 11671169.
Larsen, J.S., Selby, L.A. (1981) Spondylosis deformans in large dogs
relative risk by breed, age and sex. Journal of the American Animal
Laste, N.J., Harpster, N.K. (1995) A retrospective study of 100 cases of
feline distal aortic thromboembolism: 19771993. Journal of the
Lavely, J., Vernau, K.M., LeCouteur, R.A. (2002) Spinal epidural empyema
in seven dogs. Journal of Veterinary Internal Medicine 16, 369.
LeCouteur, R.A., Grandy, J.L., Ettinger, S.J., Feldman, E.C. (ed.) (2000)
Diseases of the spinal cord. Textbook of Veterinary Internal Medicine:
Diseases of the Dog and Cat, Vols 1 and 2. Philadelphia: WB Saunders.
Levitski, R.E., Chauvet, A.E., Lipsitz, D. (1999a) Cervical myelopathy
associated with extradural synovial cysts in 4 dogs. Journal of Veterinary
Levitski, R.E., Lipsitz, D., Chauvet, A.E. (1999b) Magnetic resonance
imaging of the cervical spine in 27 dogs. Veterinary Radiology and
Lewis, D.G., Kelly, D.F. (1990) Calcinosis circumscripta in dogs as a cause
of spinal ataxia. Journal of Small Animal Practice 31, 3537.
Lipsitz, D., Levitski, R.E., Chauvet, A.E., Berry, W.L. (2001) Magnetic resonance imaging features of cervical stenotic myelopathy in 21 dogs.
Lobetti, R.G. (1994) Subarachnoid abscess as a complication of discospondylitis in a dog. Journal of Small Animal Practice 35, 480483.
MacCoy, D.M., Trotter, E.J. (1977) Brachial paralysis subsequent to traumatic partial occlusion of the right subclavian artery. Journal of the
Americal Animal Hospital Association 13, 625630.
Malik, R., Latter, M., Love, D.N. (1990) Bacterial discospondylitis in a cat.
Manchikanti, L. (1999) Facet joint pain and the role of neural blockade in
its management. Current Review of Pain 3, 348358.
Markwalder, T.M. (1993) Surgical management of neurogenic claudication
in 100 patients with lumbar spinal stenosis due to degenerative spondylolisthesis. Acta Neurochirurgica (Wein) 120, 136142.
Martin, R.A., Shell, L., Dodds, W.J. (1986) Focal intramedullary spinal cord
hematoma in a dog. Journal of the American Animal Hospital Association
22, 545550.
335
336
Matthews, N.S., de Lahunta, A. (1985) Degenerative myelopathy in an

adult Miniature Poodle. Journal of the American Veterinary Medical
McEwan, J.D., Thomson, C., Sullivan, M., Callanan, S., Park, M. (1992)
Thoracic spinal calcinosis circumscripta causing cord compression in two
German Shepherd dog littermates. Veterinary Record 130, 575578.
McKee, W.M., Renwick, P.W. (1994) Marsupialisation of an arachnoid cyst
in a dog. Journal of Small Animal Practice 35, 108111.
McKee, W.M., Mitten, R.W., Labuc, R.H. (1990) Surgical treatment of
lumbosacral discospondylitis by a distraction-fusion technique. Journal
Meomartino, L., Fatone, G., De Vico, G. (1997) Osteochondromatosis
(multiple cartilaginous exostoses) in a dog: report of a clinical case.
Veterinaria Cremona 11, 9599.
Meric, S.M. (1988) Canine meningitis. Journal of Veterinary Internal
Medicine 2, 2635.
Meric, S.M., Child, G., Higgins, R.J. (1986) Necrotizing vasculitis of the
spinal pachyleptomeningeal arteries in three Bernese Mountain dog littermates. Journal of the American Animal Hospital Association 22, 459465.
Mesfin, G.M., Kusewitt, D., Parker, A. (1980) Degenerative myelopathy in
a cat. Journal of the American Veterinary Medical Association 176, 6264.
Moissonnier, P., Duchossoy, Y., Lavieille, S., Reviron, T., Ye, J.H., Horvat, J.C.
(1996) Reimplantation of ventral rootlets into the cervical spinal cord
after their avulsion in the dog: a new surgical treatment of brachial plexus
injuries? Veterinary Surgery 25 (3), 263.
Moissonnier, P., Duchossoy, Y., Lavieille, S., Horvat, J.C. (2001) Evaluation
of ventral root reimplantation as a treatment of experimental avulsion of
the cranial brachial plexus in the dog. Revue de Medecine Veterinaire
152, 587596.
Moissonier, P., Carozzo, C., Gnirs, K., Ruel, Y., Blot, S., Begon, D., Rault, D.,
Delise, F., Boulouha, L., Colle, M.-A. (2002) What is the exact nature of
spinal subarachnoid cysts? Imaging and surgical investigation in 13 dogs.
Morgan, J.P., Stavenborn, M. (1991) Disseminated idiopathic skeletal
hyperostosis (DISH) in a dog. Veterinary Radiology 32, 6570.
Morgan, J.P., Hansson, K., Miyabayashi, T. (1989) Spondylosis deformans
in the female Beagle dog: a radiographic study. Journal of Small Animal
Morrison, J.L., Kaplan, P.A., Dussault, R.G., Anderson, M.W. (2000) Pedicle
marrow signal intensity changes in the lumbar spine: a manifestation of
facet degenerative joint disease. Skeletal Radiology 29, 703707.
Munana, K.R. (1996) Encephalitis and meningitis. Veterinary Clinics of
North America Small Animal Practice 26, 857874.
Munana, K.R., Luttgen, P.J. (1998) Prognostic factors for dogs with granulomatous meningoencephalomyelitis: 42 cases (19821996). Journal of
Nas, K., Gur, A., Kemaloglu, M.S., Geyik, M.F., Cevik, R., Buke, Y., Ceviz, A.,
Sarac, A.J., Aksu, Y. (2001) Management of spinal brucellosis and outcome of rehabilitation. Spinal Cord 39, 223227.
Nesbit, J.W., Lourens, D.C., Williams, M.C. (1981) Spastic paresis in two
littermate pups caused by Toxoplasma gondii. Journal of the South
African Veterinary Association 52, 243246.
Nishikawa, M., Sakamoto, H., Hakuba, A., Nakanishi, N., Inoue, Y. (1997)
Pathogenesis of Chiari malformation: a morphometric study of the posterior cranial fossa. Journal of Neurosurgery 86, 4047.
Norrdin, R.W., Histand, M.B., Sheahan, H.J., Carpenter, T.R. (1990) Effects
of corticosteroids on mechanical strength of intervertebral joints and vertebrae in dogs. Clinical Orthopaedics and Related Research 259, 268276.
Olby, N.J., Munana, K.R., Sharp, N.J.H., Thrall, D.E. (2000) The computed tomographic appearance of acute thoracolumbar intervertebral
disc herniations in dogs. Veterinary Radiology and Ultrasound 41,
396402.
Osterholm, J. (1974) The pathophysiological response to spinal cord injury.
The current status of related research. Journal of Neurosurgery 40, 533.
analysis of 104 cases. Acta Anaesthesiologica Sinica 36, 7174.
Pellegrini, N., Braca, G., Pozzo, F., Pambianco, L., Poli, A. (1979) Diabetic
osteopathy in the dog. Post-mortem findings and pathogenesis. Annali
della Facolta di Medicina Veterinaria di Pisa 31, 271280.
Perrouin-Verbe, B., Lenne-Aurier, K., Robert, R., Auffray-Calvier, E.,

Richard, I., Mauduyt de la Greve, I., Mathe, J.F. (1998) Post-traumatic
syringomyelia and post-traumatic spinal canal stenosis: a direct relationship: review of 75 patients with a spinal cord injury. Spinal Cord 36,
137143.
Plummer, S.B., Bunch, S.E., Khoo, L.H., Spaulding, K.A., Kornegay, J.N.
(1993) Tethered spinal cord and an intradural lipoma associated with a
meningocele in a Manx-type cat. Journal of the American Veterinary
Porter, R.W. (1996) Spinal stenosis and neurogenic claudication. Spine 21,
20462052.
Postacchini, F., Montanaro, A. (1980) Tuberculous epidural granuloma
simulating a herniated lumbar disk: a report of a case. Clinical
Orthopaedics and Related Research 148, 182185.
Reihsaus, E., Waldbaur, H., Seeling, W. (2000) Spinal epidural abscess:
a meta-analysis of 915 patients. Neurosurgical Review 23, 175204;
discussion 205.
Remedios, A.M., Wagner, R., Caulkett, N.A., Duke, T. (1996) Epidural
abscess and discospondylitis in a dog after administration of a lumbosacral epidural analgesic. Canadian Veterinary Journal 37, 106107.
Riggs, B.L. (2002) Endocrine causes of age-related bone loss and osteoporosis.
Novartis Foundation Symposium Series 242, 247259; discussion 260264.
Robinson, D.E., Ball, K.E., Webb, P.J. (2001) Iliopsoas hematoma with
femoral neuropathy presenting a diagnostic dilemma after spinal decompression. Spine 26, E135138.
Rodkey, W.G., Sharp, N.J.H. (2003) Surgery of the peripheral nervous system. In: D. Slatter (ed.), Textbook of Small Animal Surgery, 3rd edn,
Rusbridge, C. (1997) Persistent scratching in Cavalier King Charles Spaniels.
Rusbridge, C., MacSweeny, J.E., Davies, J.V., Chandler, K., Fitzmaurice, S.N.,
Dennis, R., Cappello, R., Wheeler, S.J. (2000) Syringohydromyelia in
Cavalier King Charles Spaniels. Journal of the American Animal Hospital
Rylander, H., Lipsitz, D., Berry, W.L., Sturges, B.K., Vernau, K.M.,
Dickinson, P.J., Anor, S.A., Higgins, R.J., LeCouteur, R.A. (2002)
Retrospective analysis of spinal arachnoid cysts in 14 dogs. Journal of
Sakamoto, H., Nishikawa, M., Hakuba, A., Yasui, T., Kitano, S., Nakanishi, N.,
Inoue, Y. (1999) Expansive suboccipital cranioplasty for the treatment
of syringomyelia associated with Chiari malformation. Acta Neurochirurgica
(Wein) 141, 949960; discussion 960961.
Salinas, E.M., Martinez, N.L. (1993) Description of a case compatible with
degenerative myelopathy in an English Mastiff dog. Veterinaria Mexico
24, 159162.
Schleithoff, N.S. (1984) Frequency of occurrence of skeletal lesions among
adult dogs Berlin: Fachbereich Veterinarmedizin der Freien Universitat
Berlin.
Scott, H.W., OLeary, M.T. (1996) Fibro-cartilaginous embolism in a cat.
Sgouros, S., Williams, B. (1995) A critical appraisal of drainage in
syringomyelia. Journal of Neurosurgery 82, 110.
Sgouros, S., Williams, B. (1996) Management and outcome of posttraumatic syringomyelia. Journal of Neurosurgery 85, 197205.
Shamir, M., Rochkind, S., Johnston, D. (2001) Surgical treatment of tethered spinal cord syndrome in a dog with myelomeningocele. Veterinary
Record 148, 755756.
Shamir, M.H., Shahar, R., Aizenberg, I. (1997) Subarachnoid cyst in a cat.
Sharp, N.J.H., Vitek, S., Steingold, S., Olby, N.J., Munana, K. (2000)
A rapid, antemortem diagnostic test for canine distemper virus. Journal
Skeen, T.M., Olby, N.J., Munana, K.M., Sharp, N.J.H. (2003) Spinal arachnoid cysts in 17 dogs. Journal of the American Animal Hospital
Smith, S.A., Tobias, A.H., Jacob, K.A., Fine, D.M., Grumbles, P.L. (2003)
Arterial thromboembolism in cats: acute crisis in 127 cases (19922001)
and long-term management with low-dose aspirin in 24 cases. Journal of

Stefanacci, J.D., Wheeler, S.J. (1991) Skeletal scintigraphy in canine discospondylitis. American College of Veterinary Radiology, 66.
Stiles, J. (2000) Canine rickettsial infections. Veterinary Clinics of North
Stokol, T., Parry, B.W., Mansell, P.D., Richardson, J.L. (1994)
Hematorrhachis associated with hemophilia A in three German Shepherd
dogs. Journal of the American Animal Hospital Association 30, 239243.
Summers, B.A., Cummings, J.F., de Lahunta, A. (1994) Veterinary
Neuropathology. St Louis: Mosby.
Swezey, R.L. (1998) The sacroiliac joint. Nothing is sacred. Physical
Medicine and Rehabilitation Clinics of North America 9, 515519.
Taga, A., Taura, Y., Nakaichi, M., Wada, N., Hasegawa, T. (2000) Magnetic
resonance imaging of syringomyelia in five dogs. Journal of Small Animal
Tamke, P.G., Peterson, M.G., Dietze, A.E., de Lahunta, A. (1988) Acquired
hydrocephalus and hydromelia in a cat with feline infectious peritonitis:
a case report and brief review. Canadian Veterinary Journal 29,
9971000.
Tan, D.Y., Tsou, I.Y., Chee, T.S. (2002) Differentiation of malignant
vertebral collapse from osteoporotic and other benign causes using
magnetic resonance imaging. Annals of the Academy of Medicine
Singapore 31, 814.
Tarvin, G., Prata, R.G. (1980) Lumbosacral stenosis in dogs. Journal of the
Thomas, W.B., Sorjonen, D.C., Steiss, J.E. (1993) A retrospective evaluation of 38 cases of canine distemper encephalomyelitis. Journal of the
Tipold, A., Jaggy, A. (1994) Steroid responsive meningitis-arteritis in
dogs: long-term study of 32 cases. Journal of Small Animal Practice
35, 311316.
Tipold, A., Vandevelde, M., Jaggy, A. (1992) Neurological manifestations
of canine distemper virus infection. Journal of Small Animal Practice 33,
466470.
Tomlinson, J., Higgins, R.J., LeCouteur, R.A., Knapp, D. (1988) Intraspinal
epidermoid cyst in a dog. Journal of the American Veterinary Medical
Tshamala, M., Moens, Y. (2000) True dermoid cyst in a Rhodesian ridgeback. Journal of Small Animal Practice 41, 352353.
Vandecasteele, S.J., Knockaert, D., Verhaegen, J., Van Eldere, J.,
Peetermans, W.E. (2001) The antibiotic and anti-inflammatory treat-
ment of bacterial meningitis in adults: do we have to change our strategies in an era of increasing antibiotic resistance? Acta Clinica Belgica
56, 225233.
van Winkle, T.J., Liu, S.M., Hackner, S.G. (1993) Clinical and pathological
features of aortic thromboembolism in 36 dogs. Journal of Veterinary
Emergency and Critical Care 3, 1321.
Vestergaard, P., Lindholm, J., Jorgensen, J.O., Hagen, C., Hoeck, H.C.,
Laurberg, P., Rejnmark, L., Brixen, K., Kristensen, L.O., FeldtRasmussen, U., Mosekilde, L. (2002) Increased risk of osteoporotic
fractures in patients with Cushings syndrome. European Journal of
Endocrinology 146, 5156.
Watson, E., Roberts, R.E. (1993) Discospondylitis in a cat. Veterinary
Watt, P.R., Robins, G.M., Galloway, A.M., OBoyle, D.A. (1995)
Disseminated opportunistic fungal disease in dogs: 10 cases (19821990).
Watters, M.R., Stears, J.C., Osborn, A.G., Turner, G.E., Burton, B.S.,
Lillehei, K., Yuh, W.T. (1998) Transdural spinal cord herniation: imaging
and clinical spectra. American Journal of Neuroradiology 19, 13371344.
Ultrasound 42, 913.
Wilson, J.W. (1982) Spina bifida in the dog and cat. Compendium on
Continuing Education for the Practicing Veterinarian 4, 626632, 636637.
Wilson, J.W., Kurtz, H.J., Leipold, H.W., Lees, G.E. (1979) Spina bifida in
the dog. Veterinary Pathology 16, 165179.
Yamaura, M., Nakamura, T., Nagai, Y., Yoshihara, A., Suzuki, K. (1993)
Reduced mechanical competence of bone by ovariectomy and its preservation with 24R,25-dihydroxyvitamin D3 administration in beagles.
Calcified Tissue International 52, 4956.
FURTHER READING
Braund, K.G. (2003) Clinical Neurology in Small Animals: Localization,
Diagnosis and Treatment. http://www.ivis.org/special_books/Braund/
toc.asp
337
Postoperative care
Chapter
15
Analgesia 339
Opioid analgesics 340
Non-steroidal anti-inflammatory drug (NSAID)
analgesia 341
Analgesia in cats 343
Other agents 343
Box 15.1 Important clinical signs to recognize in

the postoperative neurosurgical patient
Nursing care 343

Moving patients 343
Myelography 343
Treatment plans 343
Cleanliness 343
Urinalysis 344
Recumbency 344
Hydration and nutrition
344
Flooring for recumbent patients

Physiotherapy
345
346
Control of urinary function 350

Anatomy and physiology 350
Disorders of micturition 351
Pharmacological manipulation of micturition
Assisted emptying of the bladder 352
351
Postoperative complications 355

Urinary tract infections 355
Gastrointestinal disturbances 355
Pancreatitis 357
Wound complications 357
Urine scald 357
Decubital ulcers 358
Miscellaneous 359
References
360
The surgeon should accept that some form of postoperative complication will develop in many neurosurgical patients. This may range from a simple urinary tract
infection (UTI) to less common but life-threatening
Post-myelographic seizures
Inappetance
Depression
Persistent fever
Wound discharge
Diarrhea or vomiting
Abdominal pain
Melena or coffee ground vomit
Change in ability to void, or in the ease of manual

expression
Blood or floccular material in urine
Odiferous urine
Urine scald
Decubital ulcers
Increase in spinal pain
Deterioration in neurological status
conditions such as pneumonia or pancreatitis. If complications are viewed as something to be expected the
surgeon is more likely to have the patient monitored
appropriately (Box 15.1).
Care of neurosurgical patients can be labor intensive
(Table 15.4). It is important that hospital nursing staff
have a high standard of training and are properly
informed about what parameters to monitor, how often
to monitor them, and what the most likely complications are going to be in each individual patient. Clear
written instructions should also be given to the client at
the time of discharge.
ANALGESIA
Neurosurgical procedures often cause a great deal of
pain for human patients. It is likely that in the past veterinarians have tended to overlook the degree of pain
suffered by our patients (Hansen and Hardie, 1993;
Dohoo and Dohoo, 1996; Lascelles et al., 1999). Doubt
as to whether an animal really is in pain often caused a
340
clinician to withhold analgesics, yet the same clinician

would administer antibiotics without documented evidence of infection (Crane, 1987). When in doubt, a
dose of opioid should be tested for effect.
Opioid analgesics
Opioids are the most effective analgesic for severe pain.
The approximate duration of action of these drugs,
together with their advantages and disadvantages, are
outlined in Table 15.1. Preoperative or intraoperative
opioid administration is recommended to reduce the
requirement for postoperative analgesia. Fentanyl patches
are a convenient way to provide long-lasting analgesia
(15.1). As these patches have a delayed onset of action
(see Chapter 6, page 84), they should be applied the
night before elective surgeries and are supplemented
with additional intravenous opioid agonists as needed.
The pure opioid agonists are the most potent analgesics but they also have more potential side-effects.
They should be given on a fixed schedule every 24 h;
the interval can be decreased if greater analgesia is
needed. The response varies markedly between patients.
Therefore small, intravenous doses should be given
every 5 min (for dogs and cats 0.1 mg/kg morphine;
0.05 mg/kg oxymorphone; 0.05 mg/kg hydromorphone)
until a satisfactory level of analgesia is achieved (stop
administration if mydriasis occurs in cats). The total dose
required for that patient is then repeated on the fixed
schedule thereafter. A continuous intravenous infusion
can also be used by delivering the cumulative dose for
each time interval as a constant rate infusion over
that same period, but the animal must be re-evaluated
frequently in case adjustments are needed to the
rate. Common side-effects include ileus and nausea.
Table 15.1 Narcotic analgesic agents1

Drug*
Dose
Interval
Advantages
Disadvantages
Morphine (pure agonist)
Dog: 0.21.0 mg/kg IM, SQ

Dog: 15.0 mg/kg PO
Cat: 0.10.4 mg/kg IM, SQ
4h
46 h
4h
Inexpensive, sedative
Respiratory depression
Bradycardia**
Caution IV**
Emesis
Oxymorphone (pure
agonist)
Dog: 0.050.2 mg/kg

IV, IM, SQ
Cat: 0.020.1 mg/kg
IV, IM, SQ
24 h
Less vomiting than

with morphine
As for morphine
Expense, availability
Hydromorphone
(pure agonist)
Dog: 0.020.4 mg/kg

IV, IM, SQ
Cat: 0.020.2 mg/kg
IV, IM, SQ
24 h
Similar to oxymorphone
As for morphine
Expense
Butorphanol
(agonist/antagonist)
Dog: 0.21.0 mg/kg

IV, IM, SQ
Cat: 0.10.6 mg/kg IV, IM, SQ;
0.51.0 mg/kg PO
12 h
Reduced respiratory or
cardiovascular effects
Less potent analgesic in

severe pain, sedation
lasts longer than
analgesia
Reduced respiratory
or cardiovascular
effects
If respiratory depression
does occur it can be
difficult to reverse
26 h
68 h
Buprenorphine
(partial agonist)
Dog: 0.0050.02 mg/kg

612 h
IV, IM, SQ
Cat: 0.0050.01 mg/kg IV, IM, SQ
Fentanyl transdermal
patch (pure agonist)
Dog: 10 kg: 25 g
1029 kg: 50 g
30 kg: 75150 g
Cat: 25 g patch
Lasts up to
Convenient, well
3 days; longer tolerated, apply before
in cats
surgery (see page 84)
Delay in onset 1224 h

dogs, 612 h cats.
Ineffective in some
animals
Codeine 60 mg with
300 mg acetaminophen
Dog: Dose at 10 mg/kg of

the acetaminophen PO
8h
Sedation, GI sideeffects. Not for cats
Inexpensive
Useful orally
1
B. Hansen, personal communication.
* Consult appropriate source for contraindications and adverse reactions and to verify doses. Many opioids are controlled substances.
** Concomitant administration of atropine may be required to overcome bradycardia; this is only a problem if hypovolemic or under anesthesia.
GI, gastrointestinal.
Postoperative care
Histamine release occurs in dogs after IV administration

of morphine; this can cause problems for animals under
general anesthesia or those with hypovolemia but not
usually for hemodynamically stable, conscious dogs.
The dose-dependent respiratory depression of opioid
agonists means that they must be used with care in
animals that have decreased ventilatory function, such
as those with cranial cervical spinal cord lesions. Opioid
agonists can increase intracranial pressure and should be
used with care after severe head trauma unless paCO2
can be monitored and preferably intracranial and mean
arterial blood pressure as well (Marik et al., 1999; de
Nadal et al., 2000; Lam and Warner, 2001). Judicious use
of these agents is justified in animals with less severe
head injuries provided that their neurological status can
be monitored closely and ideally that the effect on
paCO2 can also be assessed. Animals that become
hypercapnic will need to be ventilated in order to
restore paCO2 to between 25 and 30 mmHg.
Epidural morphine is a useful technique for some neurosurgical patients such as those with severe nerve root or
cancer pain. Respiratory depression has not been a problem in normal dogs. It does not have central opioid
effects and is not associated with sensory, sympathetic
or motor blockade, so the patient can still walk and analgesia can be effective as far cranial as the thoracic limbs.
Epidural morphine is equivalent both to intravenous morphine and to local bupivacaine nerve block for controlling
pain after lateral thoracotomy (Pascoe and Dyson, 1993;
Popilskis et al., 1993). The onset of action is from 20 to
60 min and mean duration of analgesia is 6 h; this
15.1 Skin reaction to a fentanyl patch. The patch has been

removed as dermatitis can increase the rate of absorption.
Fever can have the same effect (Bagley et al., 2000).
increases in dogs to a mean of 13 h with the addition of

epidural medetomidine at 5 g/kg (Branson et al.,
1993). Buprenorphine and morphine are equally effective when given epidurally to dogs; the dose for morphine is 0.1 mg/kg and for buprenorphine is 4 g/kg,
diluted in warm saline to a volume of 1 ml per 5 kg and
injected into the lumbosacral epidural space (Smith et
al., 2001). Urinary retention can occur after epidural or
systemic opioid (Herperger, 1998). Epidural abscessation
has been reported as a rare complication of epidural analgesia (Remedios et al., 1996). In general, transdermal
fentanyl lasts much longer than epidural morphine and is
easier to use (Robinson et al., 1999).
Non-steroidal anti-inflammatory
drug (NSAID) analgesia
The NSAIDs are indicated either as a supplement to
opioid analgesia or when opioids are unsuitable (Table
15.2). Carprofen is particularly useful as it can also be
given intravenously; it is most effective when given preoperatively (Lascelles et al., 1998). All NSAIDs provide
analgesia by blocking the cyclooxygenase enzymes that
produce prostaglandins. Cyclooxygenase (COX) activity
is either primarily constitutive (the COX-1 enzyme constantly produces prostaglandins necessary for normal
metabolic functions in many tissues) or primarily
inducible (the COX-2 enzyme is induced by inflammation to produce prostaglandins locally). Prostaglandins are
integral to the inflammatory process and they enhance
nociception both peripherally and centrally. Some
prostaglandins, particularly prostaglandin E (PgE) are
important for homeostasis in tissues like kidney and
intestine.
Newer NSAIDs have now been developed that only
inhibit COX-1 weakly. Carprofen, ketoprofen, etodolac
and meloxicam are COX-1 sparing; carprofen and ketoprofen being somewhat less so than meloxicam (KayMugford et al., 2000; Brideau et al., 2001). The analgesic
effects of these newer NSAIDs are good and they are
roughly equivalent to each other. Carprofen provides
superior analgesia to the opioid pethidine in dogs and
cats (Lascelles et al., 1994; Balmer et al., 1998); higher
doses of most NSAIDs provide analgesia comparable to
low doses of opioids in most studies.
The COX-1 sparing NSAIDs in general cause less
gastrointestinal ulceration than more potent COX-1
inhibitors. The gastrointestinal side-effects of etodoloc
and carprofen are significantly less than aspirin (Reimer
et al., 1999). Carprofen is known to preserve protective
levels of PgE in the gut and no significant gastrointestinal problems were observed in more than 200 dogs with
degenerative joint disease (Holtsinger et al., 1992).
However, meloxicam in higher doses can cause severe
341
342
Table 15.2 Non-steroidal anti-inflammatory drug analgesia1

Drug*
Dosedog
Dosecat
Comment**
Aspirin
1025 mg/kg PO q12 h
10 mg/kg PO q3648 h
Coagulopathy
Ketoprofen
1 mg/kg PO, IV, IM, SQ q24 h

for 5 days
1 mg/kg PO, IV, IM, SQ q24 h

for 5 days
Meloxicam
Acute pain:
0.2 mg/kg PO, IV, SC once, then
0.1 mg/kg q24 h
Acute pain:
0.2 mg/kg PO, IV, SC once,
then 0.1 mg/kg PO q24 h for
34 days
Long-term therapy: up to
0.1 mg per cat q4872 h
Long-term therapy:
0.1 mg/kg q24 h
Etodolac
1015 mg/kg PO q24 h

Lower dose for long-term use
N/A
Meclofenamic acid
1.1 mg/kg PO q24 h

Lower dose for long-term use
N/A
Deracoxib
34 mg/kg q24 h
N/A
Carprofen
2 mg/kg PO, IV q12 h
24 mg/kg q5 days PO, IV, SC
Rare, severe hepatic effects.

Do not repeat within 5 days
or use long-term for cats
1
B. Hansen, personal communication.
* Consult appropriate source for full list of contraindications and adverse reactions and to verify doses.
** All NSAIDs have the potential to cause GI side-effects.
N/A, not available.
ulceration despite its greater selectivity for COX-2.

Acute renal failure has been reported in humans taking
COX-1 sparing drugs (Goldstein et al., 2001; Ahmad
et al., 2002). Owing to the tendency for neurosurgical
patients to develop gastrointestinal disturbances,
NSAIDs should not be used for more than a few days
and this probably applies to COX-1 sparing drugs as
well (Ahmad et al., 2002). A specific contraindication
to NSAID use is when the animal has recently had
corticosteroids, including methylprednisolone sodium
succinate (MPSS), because of the high risk of gastrointestinal bleeding and even perforation (Toombs et al.,
1986; Strombeck and Guilford, 1990; Hinton et al.,
2002). Care in the use of NSAIDs is also needed in any
animal with renal or hepatic dysfunction; especially during periods of hemodynamic instability. Furthermore,
COX-1 inhibition by some of these drugs, such as
aspirin and ketoprofen, prevents thromboxane A2 synthesis and causes decreased platelet aggregation with
significantly increased bleeding times (Grisneaux et al.,
1999; Mathews et al., 2001; Kerwin and Maudlin,
2003) (15.2). A synthetic form of prostaglandin E
15.2 Postoperative appearance of a dog given 12 mg/kg of

aspirin 24 h before ventral slot. Severe bruising was probably
due to poor platelet aggregation. Aspirin blocks COX-1
irreversibly, thereby inactivating platelets. New platelets
must then be generated and so the effect on coagulation
can last 7 days or more. Aspirin should not be used
preoperatively.
Postoperative care
(misoprostol) can be used prophylactically to counter

gastrointestinal side-effects of NSAIDs (Walt, 1992;
Murtaugh et al., 1993). Misoprostol is also protective to
the kidney and could be used for this purpose should
NSAIDs be deemed necessary (Paller and Manivel,
1992; Shield, 1992; Davies et al., 2001).
Analgesia in cats
Selected opioids can be used in cats (Table 15.1).
Meloxicam and carprofen are also tolerated well by cats
using oral (both drugs) and IV (carprofen) routes (Parton
et al., 2000; Slingsby et al., 2000; Lascelles et al., 2001)
(Table 15.2). Acetaminophen should not be used in cats.
Other agents
These include:
Gabapentin is an anticonvulsant that has also
been used to alleviate trigeminal pain in humans.
It can be an effective analgesic for animals with
pain caused by neoplasia or inflammation of a
nerve root. The dose range starts at 5 mg/kg q8h
and extends in dogs up to 50 mg/kg q8h given to
effect; potential side-effects include sedation and
ataxia.
Medetomidine can provide useful supplemental
analgesic at low doses (13 g/kg/h). The drugs
cardiovascular effects are much shorter in duration
when used at such low doses but are quite
dramatic with an IV bolus. The dose should
therefore be given over 10 min to avoid
hypotension; it can also be delivered by syringe
pump or in IV fluids.
The skeletal muscle relaxants methocarbamol
(55132 mg/kg PO in divided doses) or diazepam
( Table 15.7) are useful to relieve muscle spasm
after spinal surgery.
Acupuncture may also be a helpful adjunctive
means of providing analgesia (Haskins, 1987;
Still, 1989; Janssens, 1992).
Physiotherapy such as heat, ultrasound (15.13B)
and laser (15.14A) can provide additional pain
relief. Pain relief after laser therapy is probably
mediated mainly through serotonin release
(Walker, 1983; Clokie et al., 1991).
NURSING CARE
Moving patients
Proper support must be provided for the spine when
moving an animal with a spinal lesion, especially an
anesthetized patient. Use of a board or stretcher is
indicated (13.1).
Myelography
It is important that the animals head is kept elevated
at all times following the injection of contrast into the
subarachnoid space, both during the myelogram and
during recovery from anesthesia. This simple precaution is easily overlooked as a means of preventing postmyelographic seizures.
Treatment plans
The best way to nurse a neurological patient, especially
one with multiple problems, is to make a plan for each
day as illustrated in Tables 15.3 and 15.4. In this manner, voiding requirements, physiotherapy needs, medications, laboratory work and routine tasks are planned
out and not forgotten. A specific plan for physiotherapy is also useful (Table 15.5).
Cleanliness
Animals that soil themselves repeatedly may be easier
to manage if their entire hindquarters are shaved.
Soiled areas should be bathed as necessary; bathing of
the perineal area may be needed several times a day in
some animals. It is often suggested that patients with
surgical wounds should not be bathed until a minimum
of 5 days after surgery. However, in one study bathing
Table 15.3 Nursing plan for a neurological patient*
Task
0800
TPR
(X)
Express bladder
(X)
Feed 1/4 can
(X)
Check drinking water
(X)
Phenoxybenzamine
5 mg PO
(X)
( )
Diazepam 2.5 mg PO
(X)
( )
Trimethoprim-sulfa
120 mg PO
(X)
Urinalysis
1200
1600
2000(etc)
( )
( )
( )
( )
( )
( )
( )
( )
( )
( )
*Taken from the hospital record of a 4-year-old male Dachshund that

had undergone a hemilaminectomy for thoracolumbar disc extrusion
4 days previously. The dog was paraplegic and incontinent, with
reduced deep pain sensation. An E. coli, sensitive to trimethoprim/
sulfa, had been cultured from his urine on admission.
TPR, temperature, pulse and respiratory rate; (X), task completed;
( ), task still to be completed.
343
344
and swimming did not influence the incidence of wound

infection, even when performed in the first 3 days after
surgery (Hosgood, 1992).
Urinalysis
If urinary incontinence is present, a urinalysis should be
done every 23 days, regardless of whether the patient
is on antibiotics. Most animals with neurological deficits
that are severe enough to interfere with motor function
will also have inefficient bladder emptying that predisposes them to retention cystitis.
Recumbency
Recumbent, tetraparetic animals need to be turned every
2 h because sustained hypostatic congestion predisposes
the patient to pneumonia; alternating between lateral
and sternal recumbency will also help to improve
ventilation. The chest should be auscultated twice

daily to facilitate early detection of problems. Patients
should be fed while supported in a sternal position to
reduce the risk of aspiration (Nicoll and Remedios,
1995). They should be kept on a water bed if possible
to minimize the development of decubitus.
Hydration and nutrition

Tetraparetic animals dehydrate easily and should be
offered water from a sternal position every 4 h and most
also require intravenous fluids. Total protein, hematocrit, urine specific gravity and serum electrolyte concentrations should be monitored daily in these patients.
Nutritional support for the neurosurgical patient
should also be evaluated carefully. Stress, particularly
from trauma or surgery, results in an increased metabolic
rate and so enhanced nutritional intake is important
Table 15.4 Nursing plan for a neurological patient*

Task
8 am 10 am 12 noon 2 pm 4 pm 6 pm
8 pm 10 pm 12 midnight 2 am 4 am 6 am
TPR b.i.d.
(X)
( )
Express bladder
q 6 h; take outside
(X)
Nil per os
except water
Offer water q4 h
and place sternal
(X)
Sucralfate 500 mg
per os q12 h
(X)
( )
375 mg Amoxycillin
clavulanic acid q12 h
(X)
( )
Rebandage feet
(X)
Turn q4 h
(X)
( )
Chest radiograph
( )
( )
( )
( )
( )
( )
( )
( )
( )
(X)
( )
Coupage and turn q4 h
Place in sling for 1/2 h
( )
( )
( )
( )
( )
( )
( )
( )
( )
( )
( )
* Taken from the hospital record of a 6-year-old male Doberman (same dog as shown in 15.7) with cervical spondylomyelopathy that had
undergone a ventral decompression 1 week previously. The dog was recumbent, tetraparetic, and although continent, he needed manual
expression of his bladder to initiate voiding. Food was being withheld as he had suffered severe bloody diarrhea the previous day, for which he was
being treated with sucralfate. The dog also tended to chew his feet if they were left unbandaged. The chest radiograph was scheduled because
the dog appeared depressed and auscultation of the lungs suggested early pneumonia. Coupage and turning the animal effectively every hour is
preferable for pneumonia; propping the animal into sternal recumbency is also helpful. The nursing care for this patient is particularly complex, and
illustrates the support needed for certain types of neurosurgical patients.
TPR b.i.d., temperature, pulse and respiration rate, twice a day; (X), task completed; ( ), task still to be completed.
Postoperative care
to minimize immunosuppression and the tendency to

deplete body protein. The patient should be given
palatable, high-quality meals several times daily to cope
with their increased requirement; ideally, their exact
caloric requirements should be calculated. Conversely,
an inactive, overweight, paraplegic dog with no serious
complications may need a reduction in its nutritional
intake (Strombeck and Guilford, 1990).
FLOORING FOR RECUMBENT

PATIENTS (15.315.7)
Different flooring materials and their indications are illustrated below. In general, a raised grate is best for most
small paraparetic or paraplegic patients. The advantage
of a grate is the reduced tendency for an animal to lie in
its own urine and feces, particularly if non-retentive
bedding is provided. A grate does not provide padding
15.5 A waterproof foam mattress; in general the foam should

be as thick as possible. Non-retentive bedding is recommended
as urine tends to pool in the depression made by the patient.
Foam does distribute body weight more evenly but decubitus
will still develop without additional precautions.
15.3 Paraplegic dog on a raised grate. This flooring allows urine

and other fluids to pass through it keeping the dog cleaner and
drier. It is suitable for most small patients and for larger patients
with minimal neurological deficits.
15.6 An inflatable airbed also provides good protection against

decubitus. A useful, cheap and disposable alternative is to use
plastic bubble packing material; the largest-sized bubbles seem
to work best.
15.4 Non-retentive bedding. Artificial fleece (Unreal Lamb Skin,

Alpha Protech, North Salt Lake, UT) is an excellent material to
use with a raised grate as it stays dry yet allows liquids to pass
through (Swaim et al., 1996). It may make some patients
uncomfortably hot (Nicoll and Remedios, 1995).
15.7 A waterbed is the ideal surface for a large recumbent

patient. Problems include cost and the risk of cooling the animal
unless the water is heated. Punctures can be a problem just as
they can for airbeds. The patient also tends to roll off the edge
unless the bed is surrounded by a barrier. See Table 15.4.
345
346
and is not suitable for large, severely tetraparetic patients

or those with established decubitus. In such cases, a variety of padded flooring materials can be employed; each
has its own advantages and disadvantages.
Nursing staff must be aware that patients with disturbed pain perception are at extreme risk for the development of thermal burns. A thermostatically controlled
circulating water blanket is the only safe way to warm

a neurological patient (15.8). Electric heating blankets
can be unreliable and dangerous (Swaim et al., 1989).
Direct heat must not be applied to patients with a fentanyl patch (page 85).
PHYSIOTHERAPY
However good the flooring material, the aim is to minimize the recumbency period, because this is when
the animal is most susceptible to complications. This
period can be shortened by physiotherapy and early
mobilization. Towel walking of paraplegic dogs or
supporting a tetraparetic patient in a sling are often
possible within 2448 h of surgery (15.9, 15.10).
15.8 Neurological patients are susceptible to cooling excessively

as they may have difficulty avoiding draughts or generating their
own heat. Heat can only be supplied safely by warming the
surrounding air (15.20), or via a thermostatically controlled water
blanket.
B
15.9 A: Walking a paraplegic dog by supporting its
hindquarters with a towel placed under the abdomen just in
front of the pelvic limbs. The tail can also be used to provide
support provided that it is held at the base to avoid injury.
B: A Walkabout sling (Walkabout Harnesses, Santa Cruz, CA)
can also be useful for early mobilization.
B
15.10 These two slings with frames provide an excellent way
to rehabilitate tetraparetic dogs. A: A canvas sling and metal
frame; this sling tends to rub more than the one shown in B.
B: This sling is made from Neoprene (Dupont-Dow, Wilmington,
DE); it is shown in 15.16 supporting a dog in a whirlpool bath.
Postoperative care
Tetraparetic animals with good motor function can be

walked with assistance using a tracking harness
(15.11). Assistants must take care not to injure themselves when lifting heavy patients (15.12).
Standard physiotherapy techniques are often effective when modified for animals (Taylor, 1992;
Bockstahler et al., 2002) (15.1315.15). Massage and
15.11 A tracking harness gives effective support to a tetraparetic

dog with good motor function. This is the same dog illustrated in
15.38 and 15.39, 4 months after surgery for cervical spondylomyelopathy (11.22). The decubital ulcer visible in 15.38 has
contracted but is still open and eventually required surgical closure.
passive range of motion can usually begin almost immediately after recovery from surgery. Massage of the
limbs for 15 min once or twice a day is usually well
tolerated especially for recumbent animals. Massage
should be performed in a distal to proximal direction
in order to promote venous return (Berry and Reyers,
1990). Hot packing is useful to reduce swelling, pain
and muscle spasm (7.12). Hot packs should be insulated from the animals skin; treatments should last
1020 min repeated every 812 h (Jerram et al., 1997).
Ultrasound is another way of applying warmth to
deeper tissues and is useful to prevent and treat muscle
spasm (Taylor, 1992) (15.13B). Ultrasound is converted
to heat, mainly at the bonetissue interface; it causes
little temperature rise in superficial tissue. Intensities
from 0.5 to 4 W/s are used; pain indicates excessive heat
generation. Creative thinking can adapt other human
techniques to veterinary patients (Taylor, 1992) (15.14,
15.15). Neuromuscular stimulator packs (Respond Dual
Channel Neuromuscular Stimulator, Medtronic) can
build specific muscle groups and can be useful (Taylor,
1992), for example after brachial plexus injury has
caused shoulder muscle atrophy (see Chapter 2). As for
nursing care, it is recommended to develop a specific
plan for each animal (Table 15.5). An external splint (see
15.12 A: A mechanical hoist is

invaluable for lifting large, recumbent
dogs in a sling. The hoist is raised
and lowered using a simple lever
mechanism (arrowhead). B: This
dog has been suspended from the
same hoist but this time using two
Walkabout slings (Walkabout
Harnesses, Santa Cruz, CA).
B
15.13 A: Passive range of motion; this
is done for 510 times in two or three
sessions a day (Jerram et al., 1997).
B: Ultrasound should not be used over
nerves or directly over a laminectomy
defect as it generates heat (Taylor, 1992).
Here the energy is being directed away
from the bony defect.
347
348
15.14 A: Low power, cold laser is used

immediately after surgery to stimulate
healing by evoking an intracellular,
photochemical response; it also causes
release of serotonin to provide pain relief.
B: A treadmill is invaluable for tetraparetic
dogs; it can be combined with a sling
as here.
15.15 Balance training stimulates

recovery of proprioceptive
responses. A: Tetraparetic dog on
a ball. B: Same dog on a balance
board.
Table 15.5 Physical therapy plan for a neurological patient*

Task
8 am
Range of motion
10 am
12 noon
2 pm
4 pm
6 pm
8 pm
10 pm
12 midnight
Proprioceptive sensory stimulation
Massage (effleurage)
Heliumneon laser; scan surgical site

Standing balance in sling
(15.10, 15.12). 10 min increasing
to 30 min
X
X
* Taken from the hospital record of an 11-year-old dog (similar to the dog shown in 15.10A) with cervical spondylomyelopathy that had undergone
a ventral decompression 1 week previously. Active postoperative physical therapy begins postoperative Day 1. Active mobilization and sensory
stimulation continues every 4 h until the dog is discharged.
X, task completed.
13.46) can be applied as a temporary precaution to

make physiotherapy safer for an animal with an unstable
thoracolumbar lesion. Swimming in a whirlpool bath or
a bathtub is invaluable (15.1615.19) but the animal
must be dried thoroughly afterwards (15.20).
In addition to the physical well-being of the patient,

the mental attitude of the animal often has a significant
effect on outcome, particularly when recovery is delayed
(15.39). For dogs facing protracted recovery periods the
temporary and intermittent use of a paraplegic cart can
Postoperative care
15.16 The sling shown in 15.10B has been used to support

this dog inside a whirlpool bath. While in the pool, the patients
limbs are flexed and it is encouraged to swim. Chlorhexidine
has been added to the water, which should be at 3840C.
The tank is either filtered or it must be drained daily.
15.17 A small dog can be suspended by hand in a shallow

bath. Water levels can be varied to provide support along with
some degree of weight bearing. At the end of the session, the
patient must be dried with towels and a warm air drier to
prevent hypothermia (15.20).
15.18 Life jackets are necessary

when an animal is in a deep pool as
it can still get into difficulty, especially
once tired. A: The jacket has a
handle on top and must fit properly
so that the dog cannot slip out of it
(K-9 Float coat, Ruffwear, Bend,
OR). B: Hydrotherapy should begin
slowly, an attendant must always be
present and the animal must wear a
life jacket (Taylor, 1992; Jerram et al.,
1997).
A
15.19 Here the patient is being made to swim against a jet of

water; it can also be made to swim around the tank to retrieve a
ball. Animals with cervical lesions may be reluctant to maintain
their head in this somewhat extended position.
15.20 A warm air blower is invaluable as neurological patients

are inefficient at drying themselves. The drier must not be too
hot, as the patient may be unable to move away or may not
even feel the heat on areas with poor nociception.
349
350
play a role in rehabilitation. In general, the risk of the dog

coming to rely on the cart is usually outweighed by the
stimulus of enhanced mobility although they should
probably not be used for the first month in order to
encourage the animal to walk. In special circumstances,
some dogs will do well in a cart long-term (15.21). The
permanent use of carts is suitable for selected patients
in which there is no hope for return of neurological
function. In such cases the owner must:
Be informed of the final prognosis.
Be able to cope with the physical demands of
getting the dog into and out of the cart.
Be able to be with the dog while it is in the cart.
Be able to manage the incontinence and tendency
for urine scald along with repeated infections.
Insure that the bladder is emptied at least three

times daily.
Monitor the urine regularly using a dipstick at home.
The dog must be of a suitable temperament.
Provided that these criteria are satisfied, many dogs
can have an excellent quality of life and may live for
years in this way.
CONTROL OF URINARY FUNCTION

The clinician is usually aware of the animals preoperative
urinary status, but the status can change several times in
the days after surgery. For example, an incontinent dog
that could be expressed easily before surgery might
become impossible to express a few days after surgery.
Several days of intermittent catheterization might then
be necessary before manual expression again becomes
possible using drugs to modify sphincter tone (page 24).
Anatomy and physiology

Parasympathetic fibers run in the pelvic nerves, which
originate from S1 to S3 spinal cord segments and cause
the detrusor muscle of the bladder wall to contract
(15.22). The detrusor muscle normally contracts in a
co-ordinated fashion to empty the bladder. Each individual smooth muscle cell in the bladder wall is joined
to its neighbors by so-called tight junctions. These
junctions have low electrical resistance so that the wave
of membrane depolarization spreads rapidly from cell to
cell over the whole detrusor muscle. Over-distension of
the bladder can disrupt the tight junctions so preventing the normal co-ordinated contraction of the detrusor
muscle fibers. This should be prevented by keeping
bladder volumes low at all times and paying strict attention to assisted emptying of the bladder (see below).
15.21 A cart (K-9 Cart Company, Oak Harbor, WA; Deweys

Wheelchairs for Dogs, Prineville, OR) designed for use in
paraplegic dogs. These are made in a variety of sizes to
suit different body weights.
Cortical
neurons
UMN-sphincter
L
1 2 34
Motordetrusor
Sensorydetrusor
Sympathetic
S
1 23
Sensorysphincter
Motorsphincter
15.22 The physiological control of

micturition is complex and involves the
integration of spinal cord reflexes with
the modulating effect of higher centres.
Any lesion in the spinal cord is likely to
alter either the reflexes, their higher
control, or both. This can then change
the bladders ability to store urine
efficiently and to empty itself completely.
Postoperative care
Sympathetic fibers run in the hypogastric nerves,

which originate from L1 to L4 or L5 spinal cord segments, and cause the urethral smooth muscle to contract.
Somatic nerve fibers run in the pudendal nerves, which
originate from S1 to S3 spinal cord segments and cause
the urethral striated muscles to contract. Both the
smooth and the skeletal muscle tone of the urethra contribute to maintaining a functional urethral sphincter
mechanism. (These will be referred to collectively as the
urethral sphincter mechanism unless stated otherwise.)
out when the bladder fills to the point where intravesicular pressure exceeds sphincter pressure causing urinary
retention and overflow. After approximately 1 month, a
reflexive emptying of the bladder develops.
In LMN lesions, the sphincter tone is decreased and
so urine will tend to leak continuously (Table 15.6).
Occasionally some animals, such as cats with sacrocaudal
injuries, have an LMN bladder that is difficult to express
because of increased sympathetic tone (OBrien, 1988).
Disorders of micturition
Pharmacological manipulation
of micturition
Micturition disorders are divisible broadly into either

upper motor neuron (UMN) deficits affecting the spinal
cord proximal to the sacral segments, or lower motor
neuron (LMN) deficits affecting the sacral spinal cord
or nerve roots. In mild or moderate UMN spinal cord
lesions (grade 2 or 3), increased urethral tone may prevent the patient from emptying the bladder fully. Initially,
in severe UMN lesions (some grade 4 and all grade 5), the
detrusor muscle is paralysed and urine is retained. It leaks
The most common problem is excessive urethral sphincter tone in UMN bladder dysfunction. The patients voluntary efforts to void, or attempts at manual expression
of the bladder, may be unable to overcome this excessive
urethral tone. The result is an increase in the residual
volume of urine in the bladder. As it is not usually possible to distinguish smooth from striated muscle effects
in any individual patient, the simplest approach is to block
the activity of both sphincters (15.23, Table 15.7). Dogs
Table 15.6 Differences in urinary function associated with UMN or LMN lesions
UMN (cranial to
sacral segments)
paraplegicacute
UMN (cranial to
sacral segments)
paraplegicchronic
LMN lesions
(sacral segments)
Detrusor function
Striated muscle sphincter tone
or
or
Smooth muscle sphincter tone
or
or
Normal; increased; decreased.

LMN, lower motor neuron; UMN, upper motor neuron.
15.23 Innervation of the bladder and urethra.

Detrusor muscle
Urethral sphincter
(smooth and striated muscle)
Somatic sympathetic
Contracts and closes
Parasympatheticcontracts
Sympathetic relaxes
351
352
with UMN lesions that have minimal motor function

are therefore usually started on diazepam, along with
either phenoxybenzamine or prazosin. This prophylactic
approach is preferable to waiting for an UTI to develop;
the drugs can easily be discontinued once the animal is
able to void unaided. Drugs acting on the urinary tract
are shown in Table 15.7:
Diazepam is used to reduce the striated muscle
tone. It is not recommended in cats; dantrolene
can be used as an alternative.
Phenoxybenzamine is used to reduce the smooth
muscle tone (sympathetic antagonist). There
is usually a delay of 23 days before
phenoxybenzamine takes effect. The drug
is potentially carcinogenic and may be
discontinued.
Prazosin is now the preferred alpha antagonist. It
is inexpensive but can cause marked hypotension
and so animals should first be tried on half the
calculated dose and ideally blood pressures are
measured for the first few days of treatment (Lane
and Thomas, 2000). An alternative is terazosin,
which has also been used safely in dogs and cats
(R. Schueler, personal communication).
Assisted emptying of the bladder

The three crucial roles of assisted voiding are to:
Prevent leakage and urine scald by keeping
intravesicular pressure and volume low.
Empty all urine from the bladder periodically to
reduce the risk of retention cystitis.
Prevent overstretching of tight junctions by
avoiding bladder distension.
Retention cystitis will develop in any patient that
does not empty its bladder completely. No antibiotic,
regardless of its potency or spectrum, can prevent this.
Use of antibiotic cover in these circumstances only
succeeds in selecting the most resistant organism
(Lees, 1986). Incontinent animals with either UMN or
LMN bladder lesions require assisted emptying at least
three times daily, with pharmacological assistance recommended for UMN lesions. Some clinicians prefer
indwelling catheters to avoid urine leakage and the
patient stress that manual expression can cause.
Repeated cystocentesis is not recommended as a means
of keeping the bladder empty. Bethanechol and possibly cisapride (Lane and Thomas, 2000), may improve
voiding in animals with LMN lesions but are not recommended in UMN lesions.
Table 15.7 Drugs acting on the lower urinary tract1

Use
Drug*
Species
Dose
Notes and side-effects
Alpha antagonist
Phenoxybenzamine
Dog
Cat
0.250.5 mg/kg PO q12 h

Per cat1.257.5 mg PO q1224 h
Hypotension
Start at lowest dose
Carcinogen?
Availability?
Alpha antagonist
Prazosin
Dog
Cat
1 mg /15 kg PO q1224 h
Per cat0.250.5 mg PO q1224 h
Hypotension
Care in CRF
Seizures
Alpha antagonist
Terazosin
Dog
Cat
Per dog12 mg up to 11 kg,

25 mg up to 50 kg; PO q12 h
Per cat0.51 mg PO q12 h
Hypotension
Priapism, rare, needs
immediate treatment
Skeletal muscle
relaxant
Diazepam
Dog
Cat
Per dog210 mg PO q8 h
Not recommended
Give 1020 min before

expression
Skeletal muscle
relaxant
Dantrolene
Dog
Cat
15 mg/kg PO q812 h
0.52.0 mg/kg PO q8 h
Alternative to diazepam
Potential for hepatic toxicity
Urinary antiseptic
Methenamine
mandelate
Dog
Cat
1020 mg/kg PO q812 h

Same or 250 mg PO q812 h
Urine pH must be below 6.5

Liver function must be normal
Lane and Thomas, 2000.

*Consult appropriate source for full list of contraindications and adverse reactions.
CRF, chronic renal failure.
Postoperative care
MANUAL EXPRESSION (15.24, 15.25)

Probably the best method to try initially for assisted
voiding is simple manual expression of the bladder.
Drugs that decrease sphincter tone usually make expression easier. It can sometimes be difficult to estimate the
completeness of bladder emptying without subsequent
catheterization or ultrasound scanning. Occasional use
of one of these objective methods for determining
bladder volume is useful, but with experience the effectiveness of manual expression can usually be assessed
quite well by palpation. The advantage of manual expression is that, in contrast to catheterization, it avoids the
introduction of bacteria into the normally sterile bladder. Even if only performed twice daily, with the third
emptying done by catheter, this approach is preferable
to catheterization each time. The disadvantages of
manual expression are that it can be difficult to perform
effectively in some animals, especially those with large
15.24 Manual expression of the bladder relies on gentle,

continuous, caudal abdominal pressure to overcome urethral
resistance.
15.25 Excessive force should not be used to overcome high

sphincter resistance; this might damage the bladder wall. It is
better to maintain moderate pressure to cause the sphincter
to fatigue.
or tense abdomens, those with external splints, or before

increased urethral tone has been modified by drugs.
Some animals become very difficult to express due to
the development of UTI, presumably because it often
induces urethral spasm. Animals managed with manual
expression are prone to urine leakage and need frequent bathing of their hindquarters to prevent urine
scald and decubitus (15.35).
INTERMITTENT ASEPTIC
CATHETERIZATION (15.2615.28)
This may prove necessary in some difficult or fractious
patients, or in dogs whose sphincter tone has not been
modified successfully. Strict attention must be paid to
aseptic technique. Even with these precautions, bacteria
may still be introduced into the bladder because the
15.26 Aseptic technique is crucial when catheterizing the bladder.

The tip of the penis or the vulva is irrigated with sterile saline and
then dilute iodophor or chlorhexidine solution. In females a sterile
speculum or a gloved finger is used to locate the urethral opening.
15.27 A sterile, soft rubber urinary catheter with sterile

lubricating jelly is advanced using a no touch technique,
holding it instead by a short length of sterile catheter wrapping.
The catheter is protected inside its wrapping to prevent
contamination prior to insertion.
353
354
15.28 Once the catheter is in the bladder, a sterile syringe is

used to aspirate urine. If large volumes are anticipated, then a
sterile three-way tap or stopcock is useful to allow urine to be
emptied from the syringe without repeated disconnection and
potential contamination.
15.29 A Foley catheter can be sutured to the vulva and the

bulb inflated as shown. The catheter can then be connected
to a closed collection system (15.30, 15.31).
distal urethra has a normal flora that includes Staphylococcus intermedius and Escherichia coli in particular
(Stone and Barsanti, 1992). Nosocomial organisms
may also be introduced by poor technique. Urinary
catheters are the second most common source of nosocomial bloodstream infection in people (Maki and
Tambyah, 2001); the risk of infection also increases
with the number of catheterizations (Lees, 1986;
Lulich and Osborne, 1995). Irrigation of the prepuce
or vestibule with 0.02% chlorhexidine prior to each
catheterization may help to reduce the infection rate
(see Closed collection system, below).
CLOSED COLLECTION SYSTEM (15.2915.31)

This system is useful in incontinent patients with severe
UTI, in males that are hard to express manually, and
15.30 A commercially available urine collection container is

preferred for the closed technique. This apparatus prevents
reflux of urine into the bladder and has a much lower likelihood
of introducing bacteria into the urinary tract than IV drip tubing
and an empty fluid bag.
15.31 Drip tubing connects this catheter to an empty IV fluid

bag. The bag should always be kept lower than the animal; if
the animal is moved the tubing should be clamped to prevent
reverse flow. These bags are difficult to empty aseptically; laying
the bag on the floor increases the likelihood of contamination.
The preferred collection apparatus is shown in 15.30.
to measure urine output in patients with reduced renal

function. Closed collection has the advantage of permitting diuresis while keeping the bladder empty.
Diuresis is useful if the patient has a severe infection as
it helps to flush debris and bacteria from the bladder.
This technique is also useful for the short-term management of incontinent animals as it keeps the animal clean
and dry, which reduces both the development of urine
scald and the chances for direct contamination of a
surgical wound. Closed collection should ideally be
replaced by another means of assisted voiding within
3 days. If a catheter is to be used for longer, it must be
replaced every 3 days and use of urinary antiseptics
should be considered. The risk of infection can be kept
Postoperative care
15.32 Tube cystostomy is best performed using a Bard Urological

Catheter (C.R. Bard, Inc., Covington, GA) with a mushroom tip.
low for the first 4 days if good catheter management is

practiced (Smarick et al., 2002); but increases significantly thereafter. The vestibule or prepuce should be
irrigated every 8 h with sterile saline and then a very
dilute solution of chlorhexidine (0.02%) (Smarick et al.,
2002). Nosocomial bacteriuria or candiduria develops
in up to 25% of humans requiring a urinary catheter for
7 days or more (Maki and Tambyah, 2001); corresponding rates for animals can be as high as 50% after 4 days
(Barsanti et al., 1985).
TUBE CYSTOSTOMY
In certain circumstances a prepubic tube cystostomy
provides an alternative to repeated catheterization
(15.32). An interlocking box suture pattern is recommended to anchor the tube (Daye et al., 1999). Tubes
are left in place for a minimum of 7 days; animals are
able to void normally after tube removal once the initiating problem has resolved (Williams and White, 1991).
Tube cystostomy can also be used occasionally as a longterm option for animals that are permanently incontinent and difficult to express (Smith et al., 1994).
consequence of UTI is transient bacteremia, which

may lead to wound infection or sepsis (Barsanti and
Finco, 1984). Repeated UTI is common; it may be the
presenting problem for animals that fail to recover
nociception after a severe injury even though they
recovered good motor function (Olby et al., 2003).
The potential for increased residual urine and an
associated UTI should be recognized in all patients
with a lesion severe enough to disturb motor function.
In such animals a urinalysis should be performed every
23 days on urine collected by cystocentesis. Evidence
of inflammation on urinalysis, such as the presence of
inflammatory cells or bacteria, warrants:
Urine culture.
Re-assessment of the method for assisted voiding.
Specific antibiotic therapy.
In an animal with marked neurological deficits and a
severe, established UTI, a 1224 h period of diuresis
combined with continuous evacuation of the bladder
into a closed drainage system is recommended. Before
urine culture results are available, the initial antibiotics
of choice are trimethoprim-sulfa, amoxycillinclavulanic
acid or a cephalosporin. Treatment is maintained for 14
days and repeat culture is performed once the animal
has been off antibiotics for 7 days. Multidrug-resistant
organisms may be treatable with subcutaneous amikacin
or imipenem (Barker et al., 2002). Candida sp. are an
important cause of UTI in humans and may account for
infections in animals where bacterial culture is negative,
especially those on long-term antibiotic therapy (Maki
and Tambyah, 2001).
Methenamine mandelate (Table 15.7) is a urinary antiseptic agent that is useful in chronic UTI or in patients
undergoing intermittent catheterization (Kevorkian et al.,
1984; Krebs et al., 1984). The drug is hydrolyzed in the
bladder to ammonia and formaldehyde at a pH 6,
which usually requires the mandelic acid supplementation (Pearman et al., 1978). It should only be used in
animals with normal liver function.
Gastrointestinal disturbances
POSTOPERATIVE COMPLICATIONS
The most important potential complications are UTI,
gastrointestinal disturbances, pancreatitis, surgical
wound complications, urine scald and decubital ulcers.
Urinary tract infections

Stagnant urine remaining in the bladder after voiding will
predispose the patient to UTI. Animals with Cushings
disease or diabetes mellitus are also predisposed to UTI
and animals given corticosteroids may be at increased
risk of pyelonephritis (Barsanti et al., 1992). An important
Up to 15% of dogs with disc disease will develop gastrointestinal problems and the mortality rate has been
reported to be as high as 2%. The major risk factor in one
study was use of dexamethasone; dose and duration of
therapy were not important (Moore and Withrow, 1982).
Vomiting can have a variety of causes including corticosteroids, antibiotics, NSAIDs or pancreatitis. If vomiting
occurs, food, water and non-critical medications should
be withheld for 24 h and intravenous fluids given to
replace losses. A high index of suspicion should be maintained for pancreatitis (see below) and aspiration pneumonia (especially in recumbent animals). Drugs used to
355
356
treat gastrointestinal disturbances are listed in Table 15.8.

Persistent vomiting warrants anti-emetic therapy and suggests an underlying cause such as pancreatitis.
Diarrhea will tend to increase the risk of UTI or
wound infection and every attempt should be made to
prevent it, or at least to shorten its severity and duration.
Food should be withheld for 1224 h and then a high
fiber diet offered. If this fails, opioids are the preferred
anti-diarrheal drugs (Strombeck and Guilford, 1990).
Bleeding into the gastrointestinal (GI) tract, presenting
as coffee ground vomit or melena, should be treated
aggressively (8.4). This problem has a high potential mortality rate; compromise of the GI barrier can also lead to
bacteremia and sepsis (Epstein et al., 1992). All oral
intake and non-essential drugs are stopped and symptomatic therapy is started. Misoprostol is indicated for
NSAID-induced GI ulceration. It would also seem to be
the logical therapy for corticosteroid-induced bleeding.
However, it does not reduce the incidence of GI hemorrhage caused by corticosteroids; omeprazole, cimetidine and sucralfate are also ineffective (Hanson et al.,
1997; Neiger et al., 2000). The lack of a suitable pharmacological agent emphasizes the importance of prevention, supportive care and dietary management. Barium
sulfate may work to stop GI bleeding or diarrhea when
all else fails.
Colonic and gastroduodenal perforation carry the
highest mortality rate of all GI complications. Furthermore,
Table 15.8 Drugs acting on the gastrointestinal system1

Use
Drug**
Species
Dose
Notes and side-effects
Antiemetic
Metoclopramide
Dog and cat
0.20.5 mg/kg SQ q8 h
Not if GI bleeding, seizures
Antiemetic
Chlorpromazine
Dog and cat
0.5 mg/kg IM q8 h
Hypotension, seizures
Ulceration
Ranitidine
Dog
Cat
2 mg/kg PO, IV, q812 h

2.5 mg/kg PO, IV, q12 h
More effective than cimetidine
Ulceration
Famotidine
Dog
0.51.0 mg/kg PO,

SQ, IV q24 h
0.5 mg/kg PO, SQ q24 h
Once daily dosing
Can bind other drugs
Cat
IV in cats not recommended
Ulceration
Sucralfate
Dog and cat
0.251.0 g per
animalPO q8 h
Ulceration
Omeprazole
Dog and cat
0.50.7 mg/kg
PO q24 h
NSAID-induced ulceration
Misoprostol
Dog
Cat
15 g/kg PO q8 h
Unknown
Diarrhea, abortion
Diarrhea
Loperamide
Dog and cat
0.08 mg/kg PO q8 h
Narcotic overdose
Diarrhea
Bismuth salicylate
Dog
0.25 ml/kg PO q6 h
For enterotoxic diarrhea, not

with NSAIDs or corticosteroids
Cat
Not recommended
Diarrhea and ulceration
Barium sulfate
Dog and cat
0.5 ml/kg q 24 h,
max. 3 doses
Constipation
Aspiration
Constipation
Bisacodyl
Dog and cat
510 mg PO q24 h
per animal
Laxative
Constipation
Psyllium
Dog and cat
210 g PO in food q1214 h
Laxative and stool softener
Strombeck and Guilford, 1990; Hart et al., 1997; Plumb, 1999.

*Consult appropriate source for full list of contraindications and adverse reactions.
GI, gastrointestinal. NSAID, non-steroidal anti-inflammatory drug.
Postoperative care
affected dogs often show none of the classic signs of an

acute abdomen making early detection very difficult.
Major risk factors are multiple doses of corticosteroids,
particularly dexamethasone, or combination of corticosteroids with NSAIDs; enemas may also play a role in
colonic perforation (Toombs et al., 1986; Hinton et al.,
2002; Reed, 2002).
Control of defecation rarely causes a problem, as
reflexive emptying will occur periodically even in
animals with functional transection of the spinal cord.
However, constipation can become a problem. Good
hydration status must be maintained but enemas should
be used with caution due to the potential risk of colonic
perforation associated with this procedure (Toombs
et al., 1986). High fiber diets and stool softeners such as
psyllium derivatives are safer alternatives (Hart et al.,
1997). One potential late complication in dogs that fail
to recover after intervertebral disc disease is fecal
incontinence; this applies to dogs with UMN as well as
LMN injuries (Olby et al., 2003).
Pancreatitis
Acute pancreatitis has a high mortality rate. It should
be considered along with GI perforation in any neurological patient that develops a sudden onset of vomiting,
collapse and pyrexia during its postoperative course.
Some dogs show a more insidious onset of signs. High
doses of corticosteroids do seem to predispose dogs to
pancreatitis, particularly those with neurological disease (Strombeck and Guilford, 1990; Williams, 1995).
Diagnosis can be difficult; serum amylase and lipase
are not specific but may help if markedly elevated and
abdominal ultrasound can also be very useful (Hess et al.,
1998). Treatment entails withholding all oral intake for
5 days until the clinical signs are in remission. Intravenous
fluid therapy must keep up with fluid losses, which can
be dramatic, and urine output should also be monitored. Metabolic acidosis and electrolyte derangements
are common and warrant regular blood gas and electrolyte measurements. An anti-emetic such as chlorpromazine should be used at the lowest possible dose to
control vomiting. Metaclopramide is not recommended.
Procaine penicillin and an aminoglycoside are the antibiotics of choice (Strombeck and Guilford, 1990). The
prognosis is guarded.
complication was swelling of the wound (7.12), which

affected 7.5% of the dogs. Wound discharge occurred
in another 5% but it was not determined what proportion of these was infected (Hosgood, 1992).
Many factors play a role in wound infection (Hosgood,
2003) (see Box 6.2). Spinal surgery is classified as a clean
surgical procedure; the infection rate for clean procedures is reported as 2.5% (Vasseur et al., 1988). However, a number of factors predispose spinal surgeries to
a higher infection rate compared to other clean procedures, including frequent preoperative use of corticosteroids or NSAIDs, obesity, GI ulceration and UTI
(Epstein et al., 1992; Hosgood, 1992) (8.4, 15.2).
Routine use of intraoperative antibiotic is therefore recommended, especially when surgery lasts longer than
90 min (Vasseur et al., 1988; Novelli, 1999). Postoperative antibiotic is only recommended when intra- or postoperative infection is documented (Kriaras et al., 2000).
A sterile adhesive dressing is useful to protect the
wound from gross contamination (15.33). This should be
kept in place for several days, but the wound must also
be checked daily for seroma formation, swelling, redness or discharge. Large seromas should be drained and
the wound repaired again to close dead space (15.34).
If infection is documented, a culture should be taken
from the depths of the wound after aseptic preparation
of the skin. Antibiotic therapy should be instituted, initially directed at staphylococcal involvement (cefazolin
515 mg/kg IV, IM q8 h, or cefadroxil 20 mg/kg PO
q12 h). Severe wound infections should be subject to
surgical debridement and irrigation.
Urine scald
Urine scald is an important cause of dermatitis and also
predisposes to decubital ulcer formation. Any dog that
soils itself with urine or feces must have the affected
Wound complications
Wound complications have been recorded in as many
as 14% of dogs undergoing spinal surgery (36/264).
Surgical time over 90 min and use of multifilament,
absorbable suture material increase the complication
rate; monofilament, absorbable material is therefore
recommended for wound closure. The most common
15.33 A sterile, self-adhesive dressing. Use of a bandage does

not reduce the incidence of wound complications; it is not known
if it lowers the actual wound infection rate (Hosgood, 2003).
357
358
15.34 A: Seroma after cervical dorsal

laminectomy. This area and the lumbosacral region are at high risk of seroma or
hematoma formation and so closure of
dead space must be perfect. B: Closed
suction apparatus draining a seroma.
The protective bandage has been
removed for illustration.
15.35 Early urine scald around the vulva in a dog incontinent

due to an L6 fracture managed using an external splint. Regular
replacement of the straps and washing and drying the skin is
indicated. Desitin (Pfizer Inc., New York) is helpful although
ingestion of large amounts can cause zinc toxicity.
area bathed and dried. Emollient cream and a water

repellent ointment such as Desitin (Pfizer Inc., New York)
are particularly useful (15.35). Animals that are unable
to move can be placed on human infant absorbent diapers,
which collect urine and prevent it from soaking the skin;
these can be weighed if necessary to quantify urine output. An indwelling catheter may be needed to prevent
leakage until the urine scald resolves.
Decubital ulcers
Decubital ulcers result mainly from unrelieved compression of tissue between a hard surface and a bony
prominence (15.3615.39). Even small paraplegic animals can develop decubital ulcers (15.37).
The skin should be kept clean and dry at all times and
in recumbent animals the bony prominences should be
examined at least daily for the onset of decubitus. The
15.36 The area over the ischiatic tuberosity in the early stages of
decubitus formation, to show edema and the onset of hair loss. At
a more advanced stage, areas of decubitus may appear simply as
a wet area on the hair coat due to the exudation of serum.
skin overlying the shoulder, elbow, rib cage, pelvis, hip,

and lateral stifle are the areas most at risk. The hair over
high-risk areas must be parted and inspected regularly
to look for moisture and hyperemia (Swaim et al.,
1996). The early appearance of decubitus is development of erythema, edema and tenderness, followed by
serum exudation and alopecia. Skin and subcutaneous
tissue loss then develops rapidly. Decubital ulcers may
give rise to episodes of bacteremia, with the potential
risk of infecting the surgical wound or implant.
Postoperative care
15.37 Some paraplegic dogs, like this Miniature schnauzer, sit

compulsively with their pelvic limbs extended rigidly. This posture
puts constant pressure on the ischiatic tuberosity leading to
decubitus. Foam rubber pads or aluminum splints can be
fashioned to prevent continued contact (Coates et al., 1995;
Swaim et al., 1996).
An appropriate flooring material (15.315.7) is essential to prevent, or at least retard, the onset of decubitus.
Resolution of established decubitus is obviously difficult
until the inciting cause is eliminated. Relief of pressure
is an important principle of treatment, which has been
reviewed in detail (Swaim et al., 1996). Neurological
patients are at particular risk due to the high potential
for urine and fecal soiling, which must be minimized
by regular bathing and drying (15.17, 15.20). Prompt
removal of devitalized tissue and regular irrigation with
an antiseptic solution (such as 0.4% chlorhexidine or
Domeboro solution (Domeboro Astringent Solution,
Bayer, Morristown, NJ)) are recommended. Surgical
debridement and primary closure, possibly using skin
flaps or grafts, may be required to resolve an indolent
ulcer, even after neurological function returns (Swaim
et al., 1996) (15.11).
Miscellaneous
15.38 An advanced decubital ulcer over the greater trochanter

that will require aggressive management.
15.39 The same dog illustrated in 15.11 and 15.38 is shown

lying on inflatable rings, one placed under the shoulder and one
under the hip. An alternative is to fashion a bandage into a
similar doughnut shape (Nicoll and Remedios, 1995). These
devices are difficult to keep in place for long; they were used to
allow this dog to be moved from his waterbed into an area with
lots of activity.
A number of other problems can arise during the postoperative period. These include:
Self-mutilation can occur in some animals with
paresthesia or absent deep pain (2.26, 15.7).
Some male dogs with severe UMN lesions develop
permanent erections; this can also lead to selfmutilation or the exposed penis may simply get
traumatized (Olby et al., 2003). Gabapentin
(page 343) has been used to treat spasticity in
humans with chronic spinal cord injuries (Gruenthal
et al., 1997); it may help to reduce the muscle
spasm contributing to erection in dogs. Elizabethan
collars or other restraint devices may also be helpful.
In humans, prolonged recumbency combined with
inactivity increases significantly the risk of deep vein
thrombosis and pulmonary thromboembolism. These
are almost certainly under-recognized as causes of
morbidity and mortality in veterinary neurosurgical
patients (Feldman, 1986). Treatment is difficult so
every effort should be directed toward avoiding
circulatory stasis by providing adequate
physiotherapy and intake of fluids (LaRue and
Murtaugh, 1990). Aspirin may be protective.
Some problems are more likely to develop after
surgery for specific disorders. Myelomalacia
develops in some dogs with thoracolumbar disc
disease (see page 128); pathological fractures can
occur in dogs with spinal tumors; pneumonia and
gastric dilation or torsion are more likely in
recumbent, tetraparetic dogs after cervical surgery.
Some complications can be prevented by a
thorough presurgical evaluation (15.40).
Finally, an animals mental status should not be overlooked in the recovery process. Affection from nursing
359
360
15.40 Cervical swelling illustrated by barium esophagram.

This was a hematoma that arose 10 days after surgery in a
Doberman with von Willebrand disease. The dog presented
with intermandibular edema, Horners syndrome, laryngeal
paralysis, megaesophagus and aspiration pneumonia.
staff, multiple physiotherapy sessions throughout the

day, putting the animal in a location where it can see a
lot of activity, regular trips outside and frequent contact with owners extending to overnight stays at home
where feasible, can mean the difference between success and failure in some animals (15.39).
REFERENCES
Ahmad, S.R., Kortepeter, C., Brinker, A., Chen, M., Beitz, J. (2002) Renal
failure associated with the use of celecoxib and rofecoxib. Drug Safety:
an International Journal of Medical Toxicology and Drug Experience 25,
537544.
Bagley, R.S., Silver, G.M., Connors, R.L., Harrington, M.L., Cambridge,
A.J., Moore, M.P. (2000) Exogenous spinal trauma: surgical therapy and
aftercare. Compendium on Continuing Education for the Practicing
Balmer, T.V., Irvine, D., Jones, R.S., Roberts, M.J., Slingsby, L., Taylor,
P.M., Waterman, A.E., Waters, C. (1998) Comparison of carprofen and
pethidine as postoperative analgesics in the cat. Journal of Small Animal
Barker, C.W., Stevenson, M.A., Zhang, W., Sanchez, S., Boudinot, F.D.,
Budsberg, S.C. (2002) Pharmacokinetics and pharmacodynamics of
imipenem in dogs. Veterinary Surgery 31, 475476.
Barsanti, J.A., Finco, D.R. (1984) Bacteremia of urinary tract origin
(urosepsis). In: R.W. Kirk (ed.), Current Veterinary Therapy, 11501153.
Barsanti, J.A., Blue, J., Edmunds, J. (1985) Urinary tract infection due to
indwelling bladder catheters in dogs and cats. Journal of the American
Barsanti, J.A., Shotts, E.B., Crowell, W.A., Finco, D.R., Brown, J. (1992)
Effect of therapy on susceptibility to urinary tract infection in male
cats with indwelling urethral catheters. Journal of Veterinary Internal
Medicine 6, 6470.
Berry, W.L., Reyers, L. (1990) Nursing care of the small animal neurological
patient. Journal of the South African Veterinary Association 61, 188193.
Bockstahler, B., Lorinson, D., Grossliner, K., Lendl, S. (2002) Effects of a
rehabilitation program for dogs undergoing decompressive spinal cord
surgery. Veterinary Surgery 31, 293.
Branson, K.R., Ko, J.C.H., Tranquilli, W.J., Benson, J., Thurmon, J.C.
(1993) Duration of analgesia induced by epidurally administered morphine and medetomidine in dogs. Journal of Veterinary Pharmacology
and Therapeutics 16, 369372.
Brideau, C., van Staden, C., Chan, C.C. (2001) In vitro effects of
cyclooxygenase inhibitors in whole blood of horses, dogs, and cats.
Clokie, C., Bentley, K.C., Head, T.W. (1991) The effects of the heliumneon
laser on postsurgical discomfort: a pilot study. Journal of the Canadian
Dental Association 57, 584586.
Coates, J.R., Sorjonen, D.C., Simpson, S.T., Cox, N.R., Wright, J.C.,
Hudson, J.A., Finn Bodner, S.T., Brown, S.A. (1995) Clinicopathologic
effects of a 21-aminosteroid compound (U74389G) and high-dose
methylprednisolone on spinal cord function after simulated spinal cord
trauma. Veterinary Surgery 24, 128139.
Crane, S.W. (1987) Perioperative analgesia: a surgeons perspective. Journal
Davies, N.M., Longstreth, J., Jamali, F. (2001) Misoprostol therapeutics
revisited. Pharmacotherapy 21, 6073.
Daye, R.M., Huber, M.L., Henderson, R.A. (1999) Interlocking box
jejunostomy: a new technique for enteral feeding. Journal of the
de Nadal, M., Munar, F., Poca, M.A., Sahuquillo, J., Garnacho, A.,
Rossello, J. (2000) Cerebral hemodynamic effects of morphine and fentanyl in patients with severe head injury: absence of correlation to cerebral autoregulation. Anesthesiology 92, 1119.
Dohoo, S.E., Dohoo, I.R. (1996) Postoperative use of analgesics in dogs and
cats by Canadian veterinarians. Canadian Veterinary Journal 37, 546551.
Epstein, M.D., Banducci, D.R., Manders, E.K. (1992) The role of the
gastrointestinal tract in the development of burn sepsis. Plastic
Reconstructive Surgery 90, 524531.
Feldman, B.F. (1986) Thrombosis-diagnosis and treatment. In: R.W. Kirk
(ed.), Current Veterinary Therapy IX, 505509. Philadelphia: WB Saunders.
Goldstein, J.L., Correa, P., Zhao, W.W., Burr, A.M., Hubbard, R.C.,
Verburg, K.M., Geis, G.S. (2001) Reduced incidence of gastroduodenal
ulcers with celecoxib, a novel cyclooxygenase-2 inhibitor, compared to
naproxen in patients with arthritis. American Journal of Gastroenterology
96, 10191027.
Grisneaux, E., Pibarot, P., Dupuis, J., Blais, D. (1999) Comparison of ketoprofen and carprofen administered prior to orthopedic surgery for control of postoperative pain in dogs. Journal of the American Veterinary
Gruenthal, M., Mueller, M., Olson, W.L., Priebe, M.M., Sherwood, A.M.,
Olson, W.H. (1997) Gabapentin for the treatment of spasticity in
patients with spinal cord injury. Spinal Cord 35, 686689.
Hansen, B.D., Hardie, E.M. (1993) Prescription and use of analgesics in
dogs and cats in a veterinary teaching hospital: 258 cases (19831989).
Hanson, S.M., Bostwick, D.R., Twedt, D.C., Smith, M.O. (1997) Clinical
evaluation of cimetidine, sucralfate, and misoprostol for prevention of
gastrointestinal tract bleeding in dogs undergoing spinal surgery. American
Hart, R.C., Jerram, R.M., Schulz, K.S. (1997) Postoperative management
of the canine spinal surgery patientPart II. Compendium on Continuing
Haskins, S. (1987) Use of analgesics postoperatively and in a small animal
intensive care setting. Journal of the American Veterinary Medical
Herperger, L.J. (1998) Postoperative urinary retention in a dog following
morphine with bupivacaine epidural analgesia. Canadian Veterinary
Journal 39, 650652.
Hess, R.S., Saunders, H.M., van Winkle, T.J., Shofer, F.S., Washabau, R.J.
(1998) Clinical, clinicopathologic, radiographic, and ultrasonographic
abnormalities in dogs with fatal acute pancreatitis: 70 cases
213, 665670.
Hinton, L.E., McLoughlin, M.A., Johnson, S.E., Weisbrode, S.E. (2002)
Spontaneous gastroduodenal perforation in 16 dogs and seven cats
(19821999). Journal of the American Animal Hospital Association 38,
176187.
Postoperative care
Holtsinger, R.H., Parker, R.B., Beale, B.S., Friedman, R.L. (1992) The
therapeutic efficacy of carprofen (RimadylVTM) in 209 clinical cases of
canine degenerative joint disease. Veterinary Comparative Orthopaedics
Hosgood, G. (1992) Wound complications following thoracolumbar
laminectomy in the dog: a retrospective study of 264 procedures.
Hosgood, G. (2003) Wound repair and specific tissue response to injury.
In: D. Slatter (ed.), Textbook of Small Animal Surgery, 3rd edn, 6686.
Janssens, L.A.A. (1992) Acupuncture for the treatment of thoracolumbar
and cervical disc disease in the dog. Problems in Veterinary Medicine 4,
107116.
Jerram, R.M., Hart, R.C., Schulz, K.S. (1997) Postoperative management
of the canine spinal surgery patientpart I. Compendium on Continuing
Kay-Mugford, P., Benn, S.J., LaMarre, J., Conlon, P. (2000) In vitro effects
of nonsteroidal anti-inflammatory drugs on cyclooxygenase activity in
dogs. American Journal of Veterinary Research 61, 802810.
Kerwin, S.C., Maudlin, G.E. (2003) Hemostasis, surgical bleeding, and
transfusion. In: D. Slatter (ed.), Textbook of Small Animal Surgery, 3rd
edn, 4465. Philadelphia: Elsevier Science.
Kevorkian, C.G., Merritt, J.L., Ilstrup, D.M. (1984) Methenamine mandelate with acidification: an effective urinary antiseptic in patients with
neurogenic bladder. Mayo Clinic Proceedings 59, 523529.
Krebs, M., Halvorsen, R.B., Fishman, I.J., Santos-Mendoza, N. (1984)
Prevention of urinary tract infection during intermittent catheterization.
Journal of Urology 131, 8285.
Kriaras, I., Michalopoulos, A., Turina, M., Geroulanos, S. (2000) Evolution
of antimicrobial prophylaxis in cardiovascular surgery. European Journal
of Cardiothoracic Surgery 18, 440446.
Lam, A.M., Warner, D.S. (2001) Opiates, intracranial pressure, and autoregulation. Anesthesiology 94, 177178.
Lane, I.F., Thomas, W.B. (2000) Diagnosis and management of urinary
retention. Veterinary Clinics of North America, Small Animal Practice
30, 2557.
LaRue, M.J., Murtaugh, R.J. (1990) Pulmonary thromboembolism in dogs:
Lascelles, B.D.X., Butterworth, S.J., Waterman, A.E. (1994) Postoperative
analgesic and sedative effects of carprofen and pethidine in dogs.
Lascelles, B.D.X., Cripps, P.J., Jones, A., Waterman Pearson, A.E. (1998)
Efficacy and kinetics of carprofen, administered preoperatively or postoperatively, for the prevention of pain in dogs undergoing ovariohysterectomy. Veterinary Surgery 27, 568582.
Lascelles, B.D.X., Capner, C.A., Waterman-Pearson, A.E. (1999) Current
British veterinary attitudes to perioperative analgesia for cats and small
mammals. Veterinary Record 145, 601604.
Lascelles, B.D.X., Henderson, A.J., Hackett, I.J. (2001) Evaluation of the
clinical efficacy of meloxicam in cats with painful locomotor disorders.
Lees, G.E. (1986) Risks of urinary catheterisation. In: R.W. Kirk (ed.),
Current Veterinary Therapy, 11271129. Philadelphia: WB Saunders.
Lulich, J.P., Osborne, C.A. (1995) Bacterial infections of the urinary tract.
In: S.J. Ettinger and E.C. Feldman (eds), Textbook of Veterinary Internal
Medicine: Diseases of the Dog and Cat, Vols 1 and 2, 4th edn,
Maki, D.G., Tambyah, P.A. (2001) Engineering out the risk for infection
with urinary catheters. Emerging Infectious Diseases 7, 342347.
Marik, P., Chen, K., Varon, J., Fromm, R., Jr, Sternbach, G.L. (1999)
Management of increased intracranial pressure: a review for clinicians.
The Journal of Emergency Medicine 17, 711719.
Mathews, K.A., Pettifer, G., Foster, R., McDonell, W. (2001) Safety and
efficacy of preoperative administration of meloxicam, compared with
that of ketoprofen and butorphanol in dogs undergoing abdominal surgery. American Journal of Veterinary Research 62, 882888.
Moore, R.W., Withrow, S.J. (1982) Gastrointestinal hemorrhage and pancreatitis associated with intervertebral disk disease in the dog. Journal of
Murtaugh, R.J., Matz, M.E., Labato, M.A., Boudrieau, R.J. (1993) Use of
synthetic prostaglandin E1 (misoprostol) for prevention of aspirininduced gastroduodenal ulceration in arthritic dogs. Journal of the
Neiger, R., Gaschen, F., Jaggy, A. (2000) Gastric mucosal lesions in dogs with
acute intervertebral disc disease: characterization and effects of omeprazole or misoprostol. Journal of Veterinary Internal Medicine 14, 3336.
Nicoll, S.A., Remedios, A.M. (1995) Recumbency in small animals: pathophysiology and management. Compendium on Continuing Education for
the Practicing Veterinarian 17, 13671374.
Novelli, A. (1999) Antimicrobial prophylaxis in surgery: the role of pharmacokinetics. Journal of Chemotherapy 11, 565572.
OBrien, D. (1988) Neurogenic disorders of micturition. Veterinary
Long-term functional outcome of dogs with severe injuries of the thoracolumbar spinal cord. Journal of the American Veterinary Medical
Paller, M.S., Manivel, J.C. (1992) Prostaglandins protect kidneys against
ischemic and toxic injury by a cellular effect. Kidney International 42,
13451354.
Parton, K., Balmer, T.V., Boyle, J., Whittem, T., Machon, R. (2000) The
pharmacokinetics and effects of intravenously administered carprofen
and salicylate on gastrointestinal mucosa and selected biochemical measurements in healthy cats. Journal of Veterinary Pharmacology and
Therapeutics 23, 7379.
Pascoe, P.J., Dyson, D.H. (1993) Analgesia after lateral thoracotomy in
dogs: epidural morphine vs. intercostal bupivacaine. Veterinary Surgery
22, 141147.
Pearman, J.W., Peterson, G.J., Nash, J.B. (1978) The antimicrobial activity
of urine of paraplegic patients receiving methenamine mandelate.
Investigative Urology 16, 9198.
Plumb, D.C. (1999) Veterinary Drug Handbook, 3rd edn. Ames, Iowa:
Iowa State Press.
Popilskis, S., Kohn, D.F., Laurent, L., Danilo, P. (1993) Efficacy of epidural
morphine versus intravenous morphine for post-thoracotomy pain in
dogs. Journal of Veterinary Anaesthesia 20, 2125.
Reed, S. (2002) Nonsteroidal anti-inflammatory drug-induced duodenal
ulceration and perforation in a mature rottweiler. Canadian Veterinary
Journal 43, 971972.
Reimer, M.E., Johnston, S.A., Leib, M.S., Duncan, R.B., Jr, Reimer, D.C.,
Marini, M., Gimbert, K. (1999) The gastroduodenal effects of buffered
aspirin, carprofen, and etodolac in healthy dogs. Journal of Veterinary
Remedios, A.M., Wagner, R., Caulkett, N.A., Duke, T. (1996) Epidural
abscess and discospondylitis in a dog after administration of a lumbosacral epidural analgesic. Canadian Veterinary Journal 37, 106107.
Robinson, T.M., Kruse Elliott, K.T., Markel, M.D., Pluhar, G.E., Massa, K.,
Bjorling, D.E. (1999) A comparison of transdermal fentanyl versus
epidural morphine for analgesia in dogs undergoing major orthopedic surgery. Journal of the American Animal Hospital Association 35, 95100.
Shield, M.J. (1992) Misoprostol: new frontiers; benefits beyond the gastrointestinal tract. Scandinavian Journal of Rheumatology Supplement
92, 3152.
Slingsby, L.S., Waterman Pearson, A.E. (2000) Postoperative analgesia
in the cat after ovariohysterectomy by use of carprofen, ketoprofen,
meloxicam or tolfenamic acid. Journal of Small Animal Practice 41,
447450.
Smarick, S., Aldrich, J., Foley, J., Kass, P., Ling, G., Fudge, M., Haskins, S.
(2002) Catheter-associated urinary tract infections in a small animal
intensive care unit. Veterinary Emergency and Critical Care 12, 195.
Smith, J.D., Stone, E.A., Gilson, S.D. (1994) Permanent cystostomy
catheter as a palliative treatment for urine outflow obstruction from
presumed transitional cell carcinoma. American College of Veterinary
Surgeons Annual Conference 1994; 426.
Smith, L.J., Yu, K.A.J. (2001) A comparison of epidural buprenorphine
with epidural morphine for postoperative analgesia following stifle surgery in dogs. Veterinary Anaesthesia and Analgesia 28, 8796.
Still, J. (1989) Analgesic effects of acupuncture in thoracolumbar disc disease in dogs. Journal of Small Animal Practice 30, 298301.
361
362
Stone, E.A., Barsanti, J.A. (1992) Urologic Surgery of the Dog and Cat.
Malvern, PA: Lea & Febiger.
Strombeck, D.R., Guilford, W.G. (1990) Small Animal Gastroenterology,
2nd edn. Davis, CA: Stonegate Publishing Company.
Swaim, S.F., Lee, A.H., Hughes, K.S. (1989) Heating pads and thermal
burns in small animals. Journal of the American Animal Hospital
Swaim, S.F., Hanson, R.R., Jr, Coates, J.R. (1996) Pressure wounds in
animals. Compendium on Continuing Education for the Practicing
Taylor, R.A. (1992) Postsurgical physical therapy: the missing link.
Compendium on Continuing Education for the Practicing Veterinarian
14, 15831593.
Toombs, J.P., Collins, L.G., Graves, G.M., Crowe, D.T., Caywood, D.D.
(1986) Colonic perforation in corticosteroid-treated dogs. Journal of the
Vasseur, P.B., Levy, J., Dowd, E., Eliot, J. (1988) Surgical wound infection
rates in dogs and cats. Data from a teaching hospital. Veterinary Surgery
17, 6064.
Walker, J. (1983) Relief from chronic pain by low power laser irradiation.
Neuroscience Letters 43, 339344.
Walt, R.P. (1992) Misoprostol for the treatment of peptic ulcer and antiinflammatory-drug-induced gastroduodenal ulceration. New England
Journal of Medicine 327, 15751580.
Williams, D.A. (1995) Exocrine pancreatic disease. In: S.J. Ettinger and
E.C. Feldman (eds), Textbook of Veterinary Internal Medicine: Diseases
of the Dog and Cat, Vols 1 and 2, 4th edn, 13721392. Philadelphia:
WB Saunders.
Williams, J.M., White, R.A.S. (1991) Tube cystostomy in the dog and cat.
Index
Note: Numbers in bold refer to illustrations or tables

A
Abdominal palpation 2324
Abrasions 167
Abscess formation 51, 184, 122, 212,
225, 229
Accessory process 141, 156
ACE inhibitors 85
Acetaminophen 343
Achilles tendon rupture 20
Acupuncture 96, 343
ADCON-L 87
Addisons disease 20, 38
Adjacent segment disease 225
Adverse drug reactions
analgesics 8485, 340342
anesthetics 85
anti-coagulants 82, 99, 333, 342
anti-convulsants 343
anti-emetics 356
anti-infectives 85
anti-inflammatories 83, 84, 85,
341342
cardiovascular system agents 85,
352
gastrointestinal system agents 356
sedatives 282
Albuminocytological dissociation 45
Allograft 117, 220, 221, 274, 292
see also Fat graft
Alpha-antagonists 352
Amikacin 355
Aminoglycoside 357
Amoxycillin 327, 344, 355
Amphotericin B 331
Ampicillin 327
Amputation, tail 297
Analgesia
opiates 8485, 340341, 343, 356
palliative 254
postoperative care 131, 339343
preoperative assessment 8485
thoracolumbar disc disease 123,
131
Anal reflexes 183
Anal sac adenocarcinoma 184
Anaphylaxis 85
295297
Anatomical specimens 73, 74
Anatomy
blood supply 1417
nervous tissue 16
skeleton 614
urinary system 350351
Anesthesia
complications 86
CSF collection 64
induction 85
maintenance 85
premedication 85
recovery 8586
side-effects 85
Angiotensin converting enzyme
(ACE) inhibitors 85
Angled bur guard 77
Anomalous spinal disease 36
Antibiotics see also specific drugs
drug side-effects 85, 355
prophylactic 83, 309, 327, 331,
355, 357
wound infection 357
Anticlinal vertebra 9
Anti-coagulant drugs 82, 99, 333,
342
Anti-convulsants 343
Anti-emetics 356, 357
Anti-infective agents 85
Anti-inflammatory drugs see also
Corticosteroids; Non-steroidal
anti-inflammatory drugs
cervical disc disease 96
CSM 218
hypervitaminosis A 326
lumbosacral disease 188
palliative 254
side-effects 83, 85, 341342
thoracolumbar disc disease 123
ventral slot surgery 102
Antiseptic, urinary 352
Anulus fibrosus
approach to ventral neck 115
bulging 110, 124125, 219

CSM 219, 234, 236
fenestration 154, 202
mini-hemilaminectomy 159
pediculectomy 159
removal 234
spinal fracture 288289
thoracolumbar disc disease
124125
Aorta 10
Aortic embolism 332333
Apnea 82
Arachnoid cyst 321, 323326
diagnosis 29, 36, 122
localization of signs 37
marsupialization 244
myelographic abnormalities 51
postoperative complications 297,
300, 301
spinal 212
subarachnoid 324
Arrhythmia, heart 297
Arteritis 329
Articular facet fracture 287289
Articulations, synovial 911
Ascending motor tracts 6
Ascending myelomalacia 36, 37, 332
Ascending sensory tracts 5
Aseptic catheterization (intermittent)
353354
Aseptic necrosis 259
Aspergillus infection 42, 328
Aspiration 167
Aspirin
analgesia 342
bleeding disorders 99, 342
ischemic neuromyopathy 333
platelet dysfunction 82, 342
Assessment, clinical 8182
brachial plexus 3031
cranial nerves 20, 21, 25
DAMNIT formula 21, 35, 36
differential diagnosis 20
etiology 2028
eyes 21
functional 2022, 2529, 184
364
Index
Assessment, clinical (contd.)

grading 124, 125
history taking 19
lesion localization 21, 3839
lesions 2832
lumbosacral plexus 3031
neurological 2028
observation 20, 21
palpation 21, 2225, 26
physical 1920
Ataxia 167, 169
Atlantoaxial joint 4647
Atlantoaxial subluxation (fracture)
161180
bone cement 177178
cats 169170
clinical signs 37, 161163
complications of treatment
166167
cross-pins 180
CSF collection 43
decision-making algorithm 165
diagnosis 36, 37, 94, 163164, 212
dorsal stabilization 166167
fixation 166180, 171176
multiple ventral implants
177178
non-surgical treatment 164165,
167
postoperative care 169
prognosis 169
ventral fusion 166168
wire fixation 178180
Atlantooccipital dysplasia 37
Atlantooccipital space 71
Atlas (C1) 7, 14, 106, 161
Atropine 85
Axis (C2) 7, 14, 174
Azithromycin 331
Azotemia 82
Azygous vein 15, 16
B
Bacterial infections see also specific
organisms
CSF analysis 45
postoperative complications 355
spinal cord disease 330
Balance training 348
Barium sulfate 356
Bartonella 42, 330, 331
Basset hound (atlantoaxial
subluxation) 161
Bathing (postoperative) 343, 349
Baylisascaris 330
Beagle
chondrodystrophy 13
meningitis 329
Bedding (postoperative) 346
BenceJones proteinuria 42
Bernese mountain dog (meningitis) 329
Bethanechol 352
Bile duct rupture 19, 282
Biomechanics of spinal fixation
286290
Biopsies
core 79
needle 6162, 275
tumors 6162, 251252, 275, 276
Bisacodyl 356
Bismuth salicylate 356
Bladder
assisted emptying 131, 343, 344,
352, 353
dysfunction 351352
innervation 351
leakage 20
neurological screening 20
rupture 282
Blastomycoides dermatidis 330
Bleeding disorders see also Von
Willebrand disease
aspirin-induced 99
bleeding time test 217
blood clots 99
cervical disc surgery 99
cervical myelopathy 82
CSF analysis 65
CSM 217, 224, 225, 229, 234
epidural 57
gastrointestinal 83, 84, 229, 342,
356
Hemoclips 78
hypoventilation 82
laminectomy 98, 138, 146
neurovascular bundles 241
preoperative assessment 213, 217
prevention 138
respiratory 82
spinal cord 51, 94, 281, 333
subdural 57
surgical complications 98, 128,
224, 225, 229
thyroid hormone supplementation
217
venous plexus 99, 146, 234
Von Willebrand factor 42, 82, 217,
229, 360
Blood cell counts
CSF 44, 45
routine laboratory analysis 41
white blood cells 42, 329
Blood contamination 44
Blood gas analysis 42
Blood supply 1417
see also Vascular system
Body position
CSF collection 66
head position 29, 94
imaging 46, 47, 54, 55, 56
patient examination 2025, 2528
paw position 20, 22
perception of 5, 6, 22, 23, 348
position for radiography 46
postoperative 225, 228229, 343,
344, 345346
Body weight (cement bar thickness)
314
Bone diseases
CSM 211, 212
DISH 326
lumbosacral 183, 184
mimicking spinal disease 20
neurological patient 81
osteoarthritis 20
osteochondritis dissecans 20, 36,
37, 183
osteochondritis 20, 36, 37, 183
osteoporosis 38, 122, 326
scintigraphy 5960
tumors 94, 248, 249, 260
Bone physiology
composition 7
demineralized matrix 292
production 209
Bone surgery
allograft 117, 220, 221, 274, 292
atlantoaxial subluxation 168,
177178
autograft 117
bleeding from cancellous bone 142
bone vessel plug 77
bone wax 77, 142, 313, 314
cement 238, 314
CSM 223, 232, 235, 238240,
245
decompression 117, 203, 243
external fixation 290292
facetectomy 203
fragments 51
fusion 234
hemilaminectomy 1, 143, 268,
269
infection risk 301
instrumentation 77, 79
laminectomy 195, 198, 200, 208,
272
lumbosacral disease 191, 198,
200
Index
Bone surgery (contd.)

metal and cement fixation 313, 314
methylmethacrylate 78
rongeurs 76
trauma 291292
tumors 255256
vertebrectomy 274
Botulism 38
Boxer dog
proprioception disorder 22
spondylosis deformans 320
Brachial plexus
anatomy 2
assessment 3031
avulsion 30, 37 331
miosis 25
myelopathy 36
nerve sheath tumor resection 275,
276
spinal trauma 282, 283, 302
Brain herniation 65, 66
Brainstem lesions 29, 212
Brain tumor 38
Brucella canis 42, 327, 330, 331
Buccal mucosal bleeding time 217
Bulldog 322
Bupivacaine 341
Buprenorphine 340, 341
Bursitis 94
Butorphanol 340
Buttress, vertebral 287289
C
Cage rest 125, 126, 131, 228, 288, 305
Calcinosis circumscripta 212
Calcinosis (tumoral) 37, 122, 326
Calcium metabolism 42, 212, 326
Calculi 38, 122
Callus encroachment 297, 299
Candida 355
Canine distemper virus (CDV) 42,
329330
Carcinoma 248
Cardiovascular disorders
arrhythmia 282, 297
bradycardia 86
cardiac arrest 167
cardiomyopathy 213
cardiopulmonary arrest 99
disorders mimicking spinal disease
20
endocarditis 20
hypertension 20
insufficiency 20
tachycardia 85
vascular diseases 20, 36, 329,
332333
Carprofen 341, 343
Cartilaginous exostoses 36, 323
Cart (paraplegic) 350
Catheters 353354
Cats
analgesic drugs 340, 343
anti-inflammatory drugs 342
atlantoaxial subluxation 169170
cervical disc disease 35, 103
81, 262
feline infectious peritonitis 35,
330
feline leukemia virus 81, 262, 263
gastrointestinal tract 356
Horners syndrome 25
hypervitaminosis A 326
ischemic myelopathy 37, 38
ischemic neuromyopathy 281,
332333
lymphoma 263
Manx 322
pharmacology 340, 342, 343, 352,
356
sacrocaudal dysgenesis 322
spinal trauma 36, 37, 169, 287,
297, 301302, 305
127, 128, 129, 133
tumors 261263
urinary tract 302, 352
Cauda equina
compression 181182
disc fenestration 202
functional anatomy 3
lumbosacral laminectomy 199,
200
nerve roots 3
neuritis 36, 37, 184
Caudal vertebra
tail injuries 296297
Cauterization 75, 83, 275
Cavalier King Charles spaniel
(syringomyelia) 322323
Cefadroxil 357
Cefalosporin 83, 85, 355
Cefazolin 83, 327, 357
Cellulose surgical spears 77
Cement plugs
CSM 218, 219, 220223, 226,
237239
ventral decompression 117
vertebral distractionstabilization
221223
Central cord syndrome 29
Cerebellar disease 29
Cerebellomedullary cistern (CMC)
CSF collection 64, 6568, 85
injection of contrast medium 53, 71
Cerebrospinal fluid (CSF)
cell counts 44, 45
collection 43, 6469, 85
examination 44, 45
meningitis 329
neutrophils 43
sedimentation chambers 44
ventral recumbancy 69
xanthochromia 45
Cervical disc disease 93120 see also
cats 103104
diagnosis 3537, 9496
clinical signs 37, 9394
palpation 24
dorsal decompression 98
laryngeal nerve paralysis 99
non-surgical treatment 96
pain 37, 94
prognosis 102103
surgical treatment
approach to ventral neck
106109
complications 98102
distractionstabilization 117118
dorsal decompression 98
dorsal laminectomy 120
hemilaminectomy 119, 266270
indications 96
locking plate 220
paramedian approach to ventral
neck 232233
ventral decompression 9698,
111118
Cervical fibrotic stenosis 212
Cervical spine
blood supply 15
diagnosis 3537
anatomical location of injury
295297
myelography 38, 70, 72
radiography 46, 49
ligaments 14
365
366
Index
Cervical spine (contd.)

spinal cord segments 3
stenosis, fibrotic 36, 37
Cervical spondylomyelopathy (CSM)
211246 see also Wobbler
syndrome
cement plug 218, 237239
CSM 218
diagnosis 35, 36, 94, 7172,
212216, 281
decision-making algorithms
223224
neurological examination 20
hypothyroidism 217
surgical treatment 218224
complications 224228
dorsal decompression 223,
241245
dorsal laminectomy 218
indications 219
laminoplasty 223224, 245246
metal and cement fixation 218,
239240
paramedian approach to ventral
neck 232233
presurgical evaluation 216217
prognosis 229
ventral decompression 218,
219221, 233235
vertebral distraction 221223,
235237
Cervical tumors 35, 37
Cervical vertebrae 78
C1/C2 7, 14, 106, 161,174
C1/C5 15, 2829, 36, 3738, 212
C1/C8 1
C5/C7 8
C5/C6 212
C6/C7 52, 211
C6/T2 15, 28, 28, 36, 212
Chemical shift artifact 2
Chemotherapeutic agents 254,
262263
Chest harness 237, 239
Chihuahua (atlantoaxial fracture) 161
Chloramphenicol 331
Chlorhexidine 355
Chlorpromazine 356, 357
Chondrodystrophoid breeds 213, 93
Chondroid metamorphosis 12
Chondrosarcoma 248
Chronic degenerative
radiculomyelopathy (CDRM) 319
Cicatrix formation 287
Cimetidine 356
Cisapride 352
Cisplatin 254
Claudication,
intermittent 38, 334
neurogenic 36
Clavulanic acid 344, 355
Cleanliness, postoperative 343344
Client communication 88
Client consent form 88
Clindamycin 327, 331
Closed collection system 354355
Cloxacillin 327
Coccidioides imitis 330
Cocker spaniel
kyphosis 131
thoracic tumor 250
Codeine 340
Cod-piece 307
Collar (head) 237, 239, 295
Column splitting 52
Computerized tomography (CT)
imaging 5557 see also Magnetic
resonance imaging
body position 55, 56
comparison to MRI 57
CSM 214216
disc extrusion 56, 57, 94
dorsal recumbency 55
neoplasia 251
spinal cord expansion 57
thoracolumbar disc disease 56, 123
trauma 284285
Confinement (cage) 125, 126, 131,
228, 288, 305
Congenital diseases
arachnoid cyst 122,
atlantoaxial instability 281
myopathy 38
vertebral anomalies 37, 212,
321322
Consent form 88
Conservative treatment
166167
CSM 218
decompression 124, 125
immobilization 282
lumbosacral disease 188, 192
recovery times 124, 125
rest 123, 124, 125, 192, 282
Constipation 356
Contrast meda
absorbtion 53
discography 50
epidural leakage 71
epidurography 50
hand position 70
iohexol 43, 49, 86
myelography 49, 72
poor filling 53
spinal cord parenchyma 54
sterile meningitis 4243
Coonhound paralysis 38
Cord dorsum potentials 61
Corpectomy 58, 149, 147
Corticosteroids see also Antiinflammatory drugs
canine distemper virus infection
330
CSM 218
drug side-effects 355, 356, 357
fungal infections 331
GME 330
meningitis 329
meningomyelitis 331
MPSS 8384, 133, 217, 257, 282,
293, 332, 342
osteoporosis 326
126, 131, 133
wound infection 83, 357
Costs of surgery 88
Coupage 344
Coxofemoral arthritis 184
Cranial nerves 20, 21, 25
Creatine kinase (CK) 42
Creatinine 82
Crossed extensor reflex 32
Cross-pin fixation see also Pin
fixation; Screw fixation
dorsal 180
spinal trauma 282
Cruciate ligament rupture 20, 184
Cryoprecipitate, bleeding disorders
217
Cryptococcus neoformans 42, 330
Curettage 236
Curettes 75, 79
bone graft collection 79
House 75, 144
Shea 75
Cushings disease
osteoporosis 326
wound infection risk 83
Index
Cushings syndrome 81
Cutaneous trunci reflex 20, 24, 2627
Cuticle bleeding time 217
Cyclosporin 85
Cystocentesis, postoperative 297
Cysts
arachnoid 29, 37, 51, 122, 212,
297, 300, 301, 321, 323326
cystocentesis 297
dermoid sinus 184, 212, 323
epidermoid 36, 212, 323
kidney 55
subarachnoid 324
synovial 36, 37, 52, 94, 122, 183,
215, 245, 320
tube cystostomy 355
Cytosine arabinoside 254
D
Dachshund
cervical disc extrusion 93
chondroid metamorphosis 12
hemilaminectomy 344
L4/S3 lesions 38
treatment 126
thoracolumbar mid-bodies 10
ventral slot decompression 116
Dalmatian (L2/L7 defects) 2829
DAMNIT scheme 21, 35, 36
Dantrolene 352
Decision-making algorithms
cervical spondylomyelopathy 223
neoplasia 256
trauma 293, 294
Decompression (dorsal) for CSM
218, 219
see also Ventral decompression
Decubital ulcers 339, 347, 358359
Decubitus 127, 167, 297
Deep pain sensation see also
Nociception; Pain
after durotomy 293
after hemilaminectomy 150
lack of 293
malacia 293294
spinal fracture 87
trauma 293295, 301302
Deep vein thrombosis (DVT) 127, 359
Degenerative disc disease 36, 183,

184, 281, 320
Degenerative leukoencephalomalacia
320
Degenerative myelopathy 36, 37,
209, 212, 319, 319320
Degenerative neuroaxonal dystrophy
320321
Dehydration (fluid intake) 85, 343,
344345, 355
Demineralized bone matrix 292
Dens
hypoplasia 164
ligaments 14
malformation 164
normal CT 7
Dental instruments 76
Depression (post-surgical) 339
Dermatitis 83, 167
Dermatofibrosis 38
Dermoid sinus 36, 37, 184, 212, 323
Descending motor tracts 56
Desitin 358
Desmopressin (DDAVP) 217
Detrusor function 351
Dexamethasone 65, 83
Diabetes mellitus 81, 83
Diagnosis, differential
cervical spine 3539
CSM 281
lumbar spine 15, 37
lumbosacral spine 3539, 184, 184
myelopathy 184
spinal trauma 184
synovial cyst 36, 94
thoracolumbar spine 3539, 122
Diagnostic techniques 4172
diagnostic imaging
CT scan 5557
MRI 5759
myelography 5154, 7072
radiography 4550
radiology 5051
scintigraphy 5960
ultrasonography 5455
electrophysiology
cord dorsum potentials 61
electromyography 6061
F waves 61
spinal cord evoked response
6061
laboratory analysis
biochemistry 4142
biopsy 6162
CSF analysis 4245, 6469
hematology 41
microbiology 42, 45
serology 42
urinalysis 42
Diaphragm
herniation 282
paralysis 82
Diarrhea 339, 356
Diazepam
analgesia 85, 96, 102, 343
lower urinary tract effects 352
micturition disorders 352
post-myelographic seizure
prevention 216
Disc anatomy 1213
Disc degeneration 12, 36, 183, 184,
281, 320
Disc extrusion (bulge) see also
Intervertebral discs; specific
spinal regions
age related 59
atlantoaxial 163
cervical 93, 94, 99
definition 13
diagnosis 56, 57, 94, 212
dorsolateral hemilaminectomy
144145
Hansen type 13
intervertebral foramen 96
protrusion 59
131
surgical indications 219
thoracolumbar 122, 127, 129, 131,
136
Disc herniation see also Intervertebral
discs
CSM 211, 219
dorsal hemilaminectomy 119
Hansen type 12, 37
surgical indications 219
terminology 58
thoracic 123
thoracolumbar 36
traumatic 285
Disc mineralization 59
Discogenic pain 184
Discography 50
Discospondylitis 326328
CSM 212, 225, 228
differential diagnosis 36, 122
fungal infection 328
367
368
Index
Disc trauma 36, 51

Disseminated idiopathic skeletal
hyperostosis (DISH) 36, 326
Distractionstabilization 117118,
235237
Doberman
bleeding disorders 217
chronic active hepatitis 213
CSM 36, 211212, 214215,
219220, 222, 243
epidural mass 329
hypothyroidism 217
intramedullary lesion 53
MRI of cervical spine 2
seizures during myelography 86
Staphylococcus intermedius 327
von Willebrand disease 359360
Dogs (general) see also specific breeds
anticonvulsants 343
anti-inflammatory dose 342
carts 350
gastrointestinal tract pharmacology
356
narcotic analgesic dose 340
physical therapy 348
postoperative nursing plan 343,
344
urinary tract pharmacology 352
water therapy 349
Domino lesions
after fracture repair 300
after ventral slot surgery 117, 220,
221, 228
cement plug 221222
Wobbler syndrome 101
Dorsal funiculi 1
Dorsal longitudinal ligament 14
Doxorubicin 254
Doxycycline 331
Draping 74, 7879
Dressings 344, 357
Drug interactions 85, 342
Drug side-effects
ACE inhibitors 85
alpha-blockers 352
anesthetics 85
anti-coagulants 82, 99, 333, 342
anti-convulsants 343
anti-emetics 356
anti-infective agents 85
anti-inflammatory agents 83, 85,
341342
barium sulfate 356
Baytril 85
bisacodyl 356
buprenorphine 340
butorphanol 340
codeine 340
dantrolene 352
fentanyl 8485, 341
glycopyrrolate 85
halothane 85
iohexol 86
ivermectin 85
ketoconazole 85
ketoprofen 342
loperamide 356
metronidazole 85
misoprostol 356
morphine 340, 341
MPSS 84
nitrous oxide 85
non-steroidal 83, 85, 341342
opiates 8485, 340, 341
oxymorphone 340
phenoxybenzamine 352
prazosin 352
sedatives 282
steroids 83, 85, 326, 357
sucralfate 356
terazosin 352
trimethoprimsulfonamide 85
Dura
dorsolateral hemilaminectomy
incision 148
herniation 191
lesions 51
stay suture 279
tears 36, 122, 281
Durotomy
dorsolateral hemilaminectomy 148
patients with no deep pain
sensation 293
149
Dysgenesis, sacrocaudal 322
Dyspnea 82, 99, 167
Dysplasia
antlantooccipital 36, 37
myelodysplasia 36, 212, 322
Dysraphism (spinal) 36, 212, 322
F waves 61
gastrocnemius muscle 60
neoplasia 250251
spinal cord evoked response
6061
Electrosurgery
cautery 83
instruments 75
Embolism
aortic 332333
fat 297
fibrocartilaginous 35, 37, 212, 281,
332
Empyema, epidural 36, 328329
Endocarditis 20
Endotracheal tube 85
End-plate failure 225, 227, 238
Enrofloxacin 85
Ependymoma 51, 248, 252
Epidermoid cyst 36, 212, 323 36
Epidural abscess 51, 122, 212, 225
Epidural empyema 328329
Epidural fibrosis 183
Epidural hemorrhage 57
Epidural leakage 49, 53, 71
Epidural lipomatosis 36
Epidural steatitis 330331
Epidurography 50
Epineurium 276
Erections 359
Erythrophagocytosis 45
Escherichia coli 327
Esophagus disease 98, 99
Etodolac 341
Euthanasia 229, 293
Euthyroidism 217
Evoked response, spinal cord 6061
Exercise intolerance 3637, 222
Exostoses, cartilaginous 36, 323
Extensor postural thrust 23
External fixation 237, 316318,
328
External splints 239, 290, 305, 306
Extradural tumors 248250,
259263
Eye function 21
Edema 99
Ehrlichia 38, 42, 330
Electric drill 77
Electrocardiogram (ECG) 213, 282
Electrocautery 83
Electromyography (EMG) 6061
Electrophysiology
cord dorsum potentials 61
Facetectomy 192, 202

Facet joints
fracture 191, 207
pain 36, 122, 184, 320
removal 130
Facial sensation 20
Famotidine 356
Fat embolism 297
Index
Fat graft
autogenous fat 8687
dorsolateral hemilaminectomy 149
fenestration of disc extrusion 201
foraminal decompression and
facetectomy 202
free fat 8687
laminectomy 208
minimization of scarring 8687
necrosis 38
128, 129
128, 129
Fecal incontinence 131, 183
Feline diseases see also Cats
128, 129
FeLV-associated lymphoma 262
(FIV) 81, 262
feline infectious peritonitis (FIP)
35, 330
feline leukemia virus (FeLV) 81,
263
Femoral nerves 127
Fenestration and disc removal
bulging disc 202
comparison to ventral slot 97
CSM 221, 227
hemilaminectomy 149, 159
instrumentation 75
laminectomy 201
pediculectomy 159
221, 227
127, 130, 154159
Fentanyl 84, 340, 341
Fever 339
Fibrocartilaginous embolism (FCE)
35, 36, 37, 212, 281, 331332
Fibroid metamorphosis 12
Fibrosarcoma 38, 248, 249, 260
Fibrosis
cervical 36, 37, 212
epidural 183, 190
peridural 131
restrictive 225, 273
Fine needle aspiration 61
FIP (feline infectious peritonitis) 35,
330
Fistula, cutaneous 127
Fixation devices
biomechanics 289290
external 290291
internal 291
multiple implant 166
Fixationstabilization
caudal disc extrusion 96
Flexor (withdrawal) reflex 27
Floating limb gait 28, 212
Flooring (cage) 345346, 359
Fluconazole 331
Fluid intake 85, 343, 344345, 355
Focal granulomatous
meningoencephalomyelitis
(GME) 248
Food intake 343, 344, 356
Forage (CSM surgery) 221, 232,
235
Foramen magnum 179
Foraminal decompression 192, 202
Foreign body migration 36
Free radicals 84
Fungal infection 42, 328, 330
Fusionfixation
167168
new bone production 209
sacral subluxation 204
F waves 61
G
Gabapentin 343
Gadoteridol 58
Gagging 20
Gait abnormalities
disconnected 212
floating limb 28, 212
Gallstones 38, 122
Gastrocnemius muscle 60
Gastrointestinal tract disturbances
bleeding 83, 84, 342, 356
diarrhea 339, 356
drug-induced 83, 84, 341, 342,
356
melena 339, 356
parasites 38, 122
perforation 83
pharmacology 356
postoperative 339, 355357
ulceration 127, 297, 341, 356
vomiting 339, 356
Gelatin sponge 78
Gelfoam 36, 78, 86, 99, 269
Gentamicin 327
German shepherd
chronic degenerative
radiculomyelopathy 319320
fungal discospondylitis 328
lumbosacral disease 38, 185, 186,
190
nuclear bone scan 60
sacral subluxation 182
thoracic tumor 250
German short-haired pointer 4
Giant breed dogs (CSM) 211, 222,
223
see also Large breed dogs; specific
breeds
Glioma
poorly differentiated 252
tumor classification 248
Glucocorticoids 217
Glucose metabolism 20, 38
Glycopyrrolate 85
GME (focal granulomatous
meningoencephalomyelitis) 248
Golden retriever
bulbous dilation of dorsal arachnoid
space 324
left hemiparesis 29
myelomalacia 61
thoracolumbar spinal cord segments
3
vertebral bone loss 61
Golf-tee pattern 52
Gracilis contracture 20, 184
Grafts
bone 117, 220, 221, 274, 292
fat 38, 8687, 127, 128, 129, 149,
201, 202, 208,
Granulomatous
meningoencephalomyelitis
(GME) 330
Great Dane
CSM 36, 211, 213, 222224, 228,
246
dyspnea after imaging 82
hypothyroidism 217
Gunshot injury 36, 331
H
Halothane 85
Halti head collar 237, 239, 295
Hansen-type disc herniation 12, 13,
182, 183
Harnesses 218, 237, 239, 347
Headlight 73, 74
Head movements 29, 94
369
370
Index
Heart disease
arrest 167
cardiomyopathy 213
cardiopulmonary arrest 99
rhythm disorders 85, 86, 282, 297
Heat therapy 347
Helminths 330
Hemangiosarcoma 248, 261
Hematoma
CSM 212, 225, 229
extradural 99
225, 229
prognosis 284
spinal cord 36, 37 212, 333
spinal trauma 284
Hemilaminectomy see also Minihemilaminectomy
cervical disc disease 98, 119,
266270
dorsolateral 136150
neoplasia 266270
nerve sheath tumor 275279
spinal trauma 294
126, 151154
Hemiparesis 94
Hemivertebra 320321
Hemoclips 78
Hemocytometer 44
Hemogram 41
Hemorrhage see also Bleeding
disorders
aspirin-induced 99
cervical 82, 99
CSF analysis 65
CSM 224, 225, 229, 234
diaphragm paralysis 82
epidural 57
gastrointestinal 229
Hemoclips 78
hypoventilation 82
prevention 138
psoas muscle 38, 281
respiratory 82
spinal cord 51, 94, 281, 333
subdural 57
surgical 98, 128, 138, 146, 224,
225, 229
thoracolumbar 128
venous plexus 99, 146, 234
Hemostasis 81, 82
Hemostat 77
Hepatic disease 81,

hepatitis (chronic) 213
hepatotoxicosis 85
Histoplasma capsulatum 330
History taking 15, 19
Hock flexion 27, 30, 182
Hopping 20, 22
Horners syndrome
cats 25
cervical tumor 251
CSM 224
Horses (Wobbler syndrome) 218
Hounsfield units (CT numbers)
283
Hydration 85, 343, 344345, 355
Hydrocephalus 162163
Hydromorphone 340
Hydromyelia 212, 322323
Hydrosorb TS 87
Hydroxyurea 254
Hyoid venous arch 171
Hypercalcemia 42
Hyperesthesia 20, 21, 183, 184
see also Nociception; Pain
Hyperflexion injury 99, 227
Hypergammaglobulinemia 42
Hyperkalemia 20, 38
Hyperreflexia (pseudo) 183
Hypertension 20
Hyperthyroidism 20, 38
Hypertrophic osteodystrophy 20
Hypervitaminosis 36, 37, 326
Hypocalcemia 42
Hypodermic needles 76
Hypoglycemia 20, 38
Hypokalemia 20
Hypothyroidism 81, 82, 213, 217
Hypoventilation 82, 99
I
Iatrogenic injury
CSM 224, 225, 226
intraoperative complications 224
neural 225, 226
225, 226
Idiopathic disorders 38, 326
Iliac thrombosis 38, 332333
Imipenem 355
Immobilization 282
Implants see also Locking plates;
Metal and cement fixation
CSM 220, 224, 225, 226, 238,

239240
failure 167, 191, 192, 224, 225,
226, 258259, 299, 318
intervertebral 220
lumbar spinal injury 296
metal and cement fixation
239240, 296
migration 300
multiple ventral implants
177178
screw fixation 192
225, 226
thoracic spinal injury 296
threaded 292
titanium 255
wound infection risk 83
Infections see also Wounds and
wound infections; specific
organisms
antibiotic prophylaxis 83
discospondylitis 36, 326328
epidural empyema 328329
infectious agents 330331
inflammatory disorders 35,
330331
surgical 83, 99, 131, 191, 357
Inflammatory disorders 35, 36
atlantoaxial subluxation 36, 163
canine distemper virus infection
329330
cauda equina neuritis 36
central nervous system 122,
329331
discospondylitis 36, 326328
epidural empyema 36, 328329
epidural steatitis 330331
FIP 330
foreign body migration 36
Gelfoam reaction 36
GME 330
meningo(encephalo)myelitis 36
steroid-responsive
meningitisarteritis 329
Infraspinatus contracture 20
Instrumentation, surgical 7379
Intercapital ligament 13
Intercostal artery 15, 278
Intermittent aseptic catheterization
353354
Intermittent claudication 38, 334
Internal fixation devices 291
see also Metal and cement
fixation
Interspinous ligaments 14
Index
Intertransverse ligaments 14
Intervertebral discs see also Disc
herniation
anatomy 12
disc removal 75
fusion 220, 221
Intervertebral foramen 3, 11, 96
Intracranial disease 163
lesions 94
raised pressure 64
Intraduralextramedullary lesions
29, 51, 52, 53, 248
Intraduralextramedullary tumors
248252, 259263
CSM surgery 224226
trauma surgery 297, 297299
tumor surgery 258
Intubation 85
Intumescence 2, 3, 258
Iohexol 43, 49, 86
Ischemic injury 225
Ischemic myelopathy 35, 36, 51,
122, 184, 281
Ischemic neuromyopathy 20, 38,
122, 212, 281, 332333
Ischemic neuropathy 281,
332333
Ischemic spinal cord 86
Isoflurane 85
Itraconazole 331
Ivermectin 85
J
Jamshidi needle 79
Jaw tone 20
Joint capsule proliferation 211
Joint pain 36, 122, 184, 320
Jugular vein 69, 278
Junctionopathy 20
K
Ketoconazole 85
Ketoprofen 341, 342
Kidney disease
calculi 38
cysts 55
nephroblastoma 249250
pain 122
worms 38, 122
Kirschner wires 178
K-wires 117, 167168, 235

Kyphosis 131, 306
L
Labrador
disc herniation 123
displacement of L2, 284
dorsal hemilaminectomy 119
spinal cord hematoma 36, 333
spinal cord mass 58
subarachnoid cyst 324
tumoral calcinosis 327
Lameness (pelvic limb) 183, 184,
204
Laminectomy
CSM 218, 219, 227, 243244
dorsal 98, 120, 126, 189190,
195202, 218, 223, 272
healing 8687
lumbosacral 188, 189190, 192
neoplasia 270273
spinal trauma 294
technique
compared to pediculectomy 153
dorsal 126, 189190, 218, 219,
227
House curette 75
spreader 74
thoracolumbar 126, 270273
Laminectomy membrane 228
Laminoplasty 219, 223224,
245246
Large breed dogs see also specific
breeds
cardiomyopathy 213
CSM 211
spondylosis deformans 185
thoracolumbar disease 132133
Laryngeal nerve 98, 99, 108
Laryngoscope 85
Larynx 171
Laser therapy 348
Lasix 65
Laxatives 356
Lesions (examination)
classification 50
decision-making algorithm 223,
224
localization 21, 2831
severity 3132
Leukodystrophy 36, 320, 320321
Leukoencephalomalacia 320
Lhasa Apso 94
Ligamentous hypertrophy 211
Ligaments
cruciate 20
dorsal longitudinal 14
ligamentum flavum 197, 199
nuchal 242
ventral longitudinal 14
vertebral 1314
Limb disorders
functional assessment 2526
gait, attitude and posture 212
lameness 183, 184, 204
Lipomatosis, epidural 36
Liver disease 81
Locking plates 219, 220, 220, 221,
223
Locomotor status 20, 20, 22
Lomustine 254
Longissimus tendon 271
Longus colli muscles 109, 112, 173
Loperamide 356
Lower motor neurons (LMN)
bladder function 24, 351
diagnostic pitfalls 29
differentiation of abnormalities 26
EMG 60
lesion severity 3132
limb function 2526
motor system function 26
muscle atrophy 26
reflexes 5, 26
Lumbar arteries 16
Lumbar cord segments 3
Lumbar disc disease see also
Lumbosacral disease
anatomical location of the injury
296
CSF collection 6869
differential diagnosis 15, 35, 37
lumbar puncture 68, 69, 72
myelography 71
radiography 4748, 4950
Lumbar nerve roots 3
Lumbar nutrient foramen 9
Lumbodorsal fascia 154
Lumbosacral anatomy 11, 182
Lumbosacral disease 181209
see also Lumbar disc disease
diagnosis
clinical signs 183
CT 186188
differential 36, 184
371
372
Index
Lumbosacral disease (contd.)

discography 186
electrophysiology 184
epidurography 186
examination 183184
hyperesthesia 184
L4/S3 lesions 37
myelography 185
survey radiography 185
surgical treatment
complications 191192, 192
dorsal fusionfixation 190191,
204209
dorsal laminectomy 189190,
195202
facetectomy 190, 202204
foraminal decompression 190,
202204
prognosis 192
Lumbosacral junction 296
Lumbosacral myelography 4849, 72
Lumbosacral plexus 4, 30, 148
Lumbosacral reflexes 183
Lumbosacral vascular system 183
Lumbosacral vertebrae
anatomy 89, 10, 11
L1/L7, 1
L2, 284
L3, 3637
L4/S3, 28, 3738
L6/L7, 296
L7, 12, 296
Lymphoma see also Neoplasia;Tumors
extradural 249
FeLV-associated 81, 263
spinal 262
Lysosomal storage disease 36, 326
M
Magnetic resonance (MR) imaging
comparison to CT 57
CSM 216
diagnostic imaging 5759
trauma 285286
Malacia, focal 148, 286, 293294
Mannitol 65
Manx cat
spina bifida 322
Marsupialization 244, 325
Massage (effleurage) 348

Mastiff
CSM 215, 245
dens hypoplasia 164
Hansen-type disc herniation 182
Medetomidine 343
Medullary lesions 51
Melena (coffee ground) vomit 339,
356
Melanoma 248
Meloxicam 341342, 343
Menace deficits 20, 29
Meninges 34
Meningioma see also Neoplasia;
Tumors
CSM 212
diagnosis
intradural 249
prognosis 260
surgical dissection 273
treatment outcome 261
Meningitisarteritis 329
Meningitis (FIP) 330
Meningocele 212
Meningoencephalitis 38
Meningomyelitis
CSM 212
neck pain 94
Meningomyelocele 36, 212
Mental status
depression 339
neurological screening 20
Metabolic disease 36
Metaclopramide 356, 357
Metal and cement fixation, see also
Implants
CSM 218, 219, 221, 228,
239240
failure 258259
implant pattern 310, 312
implants 239240, 296
trauma 309316
Methenamine mandelate 352
Methocarbamol 96, 102, 343
Methoxyfluorane 85
Methylmethacrylate bone cement 78
see also Bone cement
Methylprednisolone sodium succinate

(MPSS)
see also Corticosteroids
drug dose 84
drug interactions 342
drug side-effects 84
neuroprotective effect 83
spinal trauma 282, 293
treatment
CSM 217
decision-making 293
fibrocartilaginous embolism
332
outcome 133
preoperative 8384, 217, 257
spinal cord tumors 257
spinal trauma 282
133
Metronidazole 85
Micturition see also Urinary system;
Urine
neurological damage 24
physiological control 350
Miniature dogs see also Small breed
dogs
Chihuahua 161
pinscher 118
poodle 177, 325
schnauzer 332, 359
toy breeds 165
Mini-chuck 76
Mini-hemilaminectomy
combined with pediculectomy
154
dorsolateral 143
lateral approach 151
126127, 136, 151154
Minocycline 327
Miosis 25
Misoprostol 343, 356
Mobilizers, surgical 75
Morphine 340, 341
Motor system 56
Motor neurons (LMN and UMN) 5,
20, 2426, 29, 3132, 60, 121,
351
Movement disorders 183 see also
Walking ability
Moving of patients, postoperatively
343
Mucopolysaccharidosis type VI 326
Mucosal bleeding time (buccal) 217
Multiple cartilaginous exostoses
(osteochondromatosis) 212
Index
Multiple ventral implants 169, 177

Muscle 146
Muscle activity see also
Electromyography
EMG 6061
gastrocnemius muscle 60
urinary sphincter 351
Muscle atrophy 26, 183
Muscle masses 20, 26
Muscle miosis 25
Muscle palpation 21, 24
Muscle relaxants 352
Muscles and muscle groups
epaxial 147, 152, 159, 196, 197
iliocostalis 155, 156, 157, 158
intercostal 278
levator costarum 157, 158
longissimus 141, 155
longus colli 109, 112, 173, 233
multifidus 271
omotransversarius 276
pectoral 276
psoas 37, 38, 331
rhomboideus 241
scalenus 277, 278
spinalis 242, 243, 267
sternocephalicus 107, 233
sternohyoid 107,171
sternothyroid 172
strap 233
trapezius 241
Musculocutaneous nerve 31
Mycobacterium 331
Myelitis 184
Myelodysplasia (spinal dysraphism)
36
Myelography
abnormalities
common causes 51
CSM 213216
interpretation 5051
neoplasia 251
nerve sheath tumor 51
neurodegeneration 5354
spinal cord expansion 57
trauma 283284
complications 5154, 217
contrast media 49, 70, 86
indications 4849
spinal region
cervical 49, 7072, 95, 164
lumbar 4950, 71
lumbosacral 72
normal lateral 49
thoracolumbar 122123
Myeloma 42, 61, 248
Myelomalacia
malacia 148
postoperative 127, 359
secondary to disc extrusion 333
thoracolumbar disc disease 127, 128
Myelopathy
brachial plexus 36
cervical 38, 82, 212
degenerative 37, 184, 209
fibrocartilaginous embolic 212
hemorrhage 82
hypoventilation 82
hereditary 36
ischemic 35, 37, 51, 184
Myelotomy 87
Myesthenia gravis 38
Myopathy 20, 38
Myositis 20, 38
Myxoma 260
N
Narcotic agents 85, 96, 340
Necrosis
aseptic 259
fat graft 36
Needle biopsy 275
Neoplasia 247279 see also specific
tumor types
diagnosis
biopsy 251252, 262
classification 248
clinical signs 247248
differential 35, 163, 281
electrophysiology 250251
imaging 250251
staging 252254
extradural 248249, 259
extramedullary 249250, 259261
fibrosarcoma 38
intradural 38, 249250, 259261
intramedullary 250, 261
lumbosacral 184
nerve root tumors 251, 275, 279
nerve sheath tumor 257, 275279
neurological patients 81
prognosis 259261, 263
spinal 248, 261263
spinal cord 59, 257258
treatment
262263
surgical treatment 255259,

266279
tumor biology 248250
Neospora caninum 42, 330
Nephroblastoma 51, 248, 249250,
261, 273
Nerve block 341
Nerve compression 6, 29
Nerve hook 76
Nerve roots 3, 12, 21, 131
nerve root compression 29
nerve root signature 93, 94, 103,
212, 279
nerve root tumors 251, 275,
279
Nerve sheath tumor
clinical signs 248
diagnostic imaging 252
prognosis 260261
resection 275279
surgery 257
tumor biology 248
Nerves
cranial 20, 21, 25, 277, 278
femoral 127
laryngeal 98, 99, 108
musculocutaneous 31
pelvic 30
pudendal 30
radial 31, 275
saphenous 31
sciatic 30
superficial peroneal 31
thoracic 277, 278
tibial 31
ulnar 31
Nervous tissue, functional anatomy
16
Neuritis
cauda equina 36, 37, 184
polyradiculoneuritis 38
Neurogenic muscle disease 24, 26
Neurological disease
CSM 212
ischemic 38, 332333
neuroaxonal dystrophy 320321
neurogenic atrophy 26
neuromyopathy 37, 38, 184,
212
neuropathy 38, 184, 332333
patient management 81
peripheral 38, 184
trauma 282283
373
374
Index
Neurological examination see also

Reflexes
brachial plexus 3031
cranial nerves 20, 21, 25
DAMNIT formula 21, 35, 36
eyes 21
functional assessment
attitude, posture and gait 20, 22,
29
hyperesthesia 20, 21, 184
limbs 28, 29
motor function 20, 22, 26
pain 21, 25, 2728
proprioception 22
urinary function 20, 21
grading 124, 125
lesion localization 21, 3839
lumbosacral plexus 3031
observation 20, 21
palpation
abdomen 2224
muscle 21, 26
pain assessment 24
spine 2425
Neuromuscular disorders 20
Neuromyopathy (ischemic) 20
Neuropathy 38, 184, 332333
Neurovascular bundle 139, 141
Neutrophils 43, 45
Nitrous oxide 85
Nociception (pain perception) see also
Pain
decision-making algorithms 293,
294
deep pain sensation 20, 21,
87, 125, 127, 293295,
301302
localization 37, 38
pathophysiology 286
prognosis 32, 301302
recurrent 150
trauma 87, 293295
withdrawal reflex 28
Non-steroidal anti-inflammatory
drugs (NSAIDs) see also Antiinflammatory drugs
effect on thyroid function tests
217
pharmacology 341
postoperative care 192, 341343
preoperative assessment 82, 83
side-effects 83, 85, 341342, 355,
356
wound infection 357
166167
CSM 218
decompression 124, 125
immobilization 282
recovery times 124, 125
rest 125, 126, 131, 192, 228, 282,
288, 305
124, 125
trauma 282, 289290, 305309
Nuchal ligament 241, 242
Nucleus pulposus 13, 111, 152, 156
Nursing care, postoperative 343345
see also Postoperative complications
Nutrient foramen 9
Nutrition see also Food intake
hypervitaminosis A 36, 37, 326
postoperative nursing care
344345
vitamin disorders 36, 37, 217, 326
O
Obesity 83
Occipital artery 15
Occipital protuberance 178
Oculovestibular response 20
Odontoidectomy 171, 177
Oligodendroglioma 252
Omeprazole 356, 356
Operating loupes 73, 74
Operating microscope 73, 74
Operation duration 83
Opioid analgesics 340341, 343, 356
side-effects 8485, 340, 341
Organophosphate toxicity 38, 42
Oropharyngeal pain 94
Orthopedic disease
CSM 211, 212
DISH 326
fracture 122
mimicking spinal disease 20
tumors 248, 249, 260
Osteoarthritis 20
Osteochondritis dissecans (OCD)
20, 36, 37, 183
Osteophytes 211
Osteoporosis 38, 326

122
Osteosarcoma 248, 249, 260
Otitis externa 167
Otitis media 94
Oxymorphone 340
P
Packed cell volume (PCV) 82
Pain see also Deep pain sensation;
Nociception
abdominal 38, 122
discogenic 184
facet joint 36, 122, 184
neck 94
oropharyngeal 94
with paraplegia 32
pleuritic 94
postoperative
after hemilectomy 149
after spinal trauma 301302
after ventral slot surgery 103
sacroiliac joint 184
125
Pain syndrome (thalamic) 38
Palliative therapy 254
Palpation (spine) 2425
Palpebral reflex 20
Pancreatitis 38, 122, 357
Panniculus reflex 21
Panosteitis 20
Paraparesis 29, 32
Paraplegia 26, 32, 130
Paralysis
Coonhound 38
progressive myelomalacia 127
tick 38
Patellar luxation 20
Patellar reflex 20, 2627
Pathophysiology, trauma 290292
Patient examination
etiology 2028
history taking 19
lesions 2832
neurological 2028
physical 1920
Paw position 20
Pedicle grafts 87
Pediculectomy 127, 151154
Pelvic nerve 30
Penicillin derivatives 83
Penile erection 359
Penrose drains 276
Perineal reflex 20
Index
Periodontal disease 83
Periosteal elevator 75, 136, 197, 235
Peritonitis 330
Peroneal nerve 31
Pharynx pain 94
Pheochromocytoma 20
Phenobarbitone 217
Phenoxybenzamine 343, 352
Physical examination 1920
see also Neurological examination
Physical therapy 131, 343, 346350
Pilonidal sinus (dermoid sinus) 36,
37, 184, 212, 323
Pin fixation 78 see also Cross-pin
fixation; Screw fixation
angles for fixation 311, 312
entry points and trajectory 312
external fixation 292
failure rates 169
insertion 76
metal and cement fixation 309
stapling technique 315316
Steinmann pins 221, 290292,
296, 298, 310, 314, 315
vertebral distractionstabilization
221
Pinscher, miniature 118
Placing test 23
Plasma cell tumor 259, 260
Platelets 82, 342
Pleocytosis 44, 45, 49, 329, 330
Pleuritic pain 94
Pneumatic system 77
Pneumomediastinum 225
Pneumonia 99, 297
Pneumothorax 127, 130, 158, 297
Polyarthritis 20, 38, 94, 122
Polymyositis 38, 94, 122, 163
Polyradiculoneuritis 38, 184
Poodle
atlantoaxial subluxation 161, 163,
168
dens hypoplasia 164
Position (perception of)
anatomy 5
nerve compression 6
postoperative sensory stimulation
348
reflex step 23
Positive profile pins 78
Postoperative care 339360
analgesia 339343
cleanliness 343344
CSM 228229
flooring and bedding 345346

hydration and nutrition 344345
moving patients 343
myelography 343
neoplasia 259
neurosurgery 339
physical therapy 346350
recumbency 344
slings and harnesses 346, 347, 349
trauma 297
treatment plans 343, 344
urinary function 344, 350352,
353355
Postoperative complications 224
see also Intraoperative
complications; Preoperative
complications
CSM 225, 226228
decubital ulcers 358359
deep pain sensation (loss of) 150
deformity 258
fenestration 130
gastrointestinal 127, 297, 355357
implant failure 258259
infections 83, 258, 259, 354355,
357
myelomalacia 359
osseous fusion 97
penile erection 258, 359
respiratory 168
self-mutilation 359
seroma formation 358
spinal cord tethering 149
127132
thromboembolism 359
trauma 299, 299301
tumor recurrence 259
urine scald 357358
Posture, attitude and gait 20, 20, 23,
181
Prazosin 352
Prednisolone 96, 218, 329, 330
Prednisone 254
analgesia 8485
anesthesia 8586
client communication 87
clinical assessment 8183
durotomy 87
imaging 82
laminectomy healing 8687
myelotomy 87
pharmacological considerations
8385
recovery after spinal cord injury 87
surgical considerations 8687

Pressure sores 308
Procaine penicillin 357
Proprioception (perception of
position)
anatomical considerations 5
effects of compression 6
postoperative sensory stimulation
348
reflex step 23
Prostate disease 38, 81, 122, 184
Protein analysis 42, 45
Protozoal infections 38, 330
Pruritus 183
Pseudohyperreflexia 27, 183
Psoas muscle 37, 38, 122, 184, 281,
331
Psyllium 356
Ptosis 25
Pudendal nerve 30
Pug
tetraparesis 101
Pulmonary thromboembolism (PTE)
359
Pupillary light reflex 20
Pyoderma 81
Pyometra 81
Pyrimethamine 331
R
Rabies 330
Radial nerve 31, 275
Radiation injury 36
Radiation therapy 254255
Radicular arteries 16
Radiculopathy 20, 38
Radiography
imaging techniques
discography 50
epidurography 50
interpretation 5051
myelography 4850, 5154
normal spinal 46
positioning 46, 47
principles 5051
survey radiography 4546, 283
x-rays 82, 99
lesions
atlantoaxial joint 4647
cervical 46, 49, 94
extradural 51
intradural 51
intramedullary 51
375
376
Index
Radiography (contd.)
lesions (contd.)
lumbar 4950
neoplasia 250
thoracolumbar 122
trauma 283
Ranitidine 356
Recovery times, thoracolumbar
surgery 125
Recumbency
CSF collection 59
imaging 47, 55
postoperative 225, 228229, 344
Red blood cell counts 44, 45
Reflexes 2627
anal 183, 185
crossed extensor 32
cutaneous trunci 20, 24, 24, 2627
hock flexion 27, 30, 182
LMN 5, 26
palpebral 20
panniculus 21
patellar 20, 2627
perineal 20
postural 20
pseudohyperreflexia 27, 183
pupillary 20
reflex step 23
SchiffSherrington sign 32
spinal 20, 21, 87
UMN 6, 26
withdrawal 20, 27, 28, 31
Renal disease
calculi 38
cysts 55
nephroblastoma 249250
pain 122
worms 38, 122
Respiratory disorders
dyspnea after imaging 82
fentanyl-induced 84
hypoventilation 82
pneumomediastinum 225
pneumonia 99, 297
pneumothorax 127, 130, 158,
297
postoperative 168
pulmonary edema 167
respiratory arrest 167
upper airway disease 20, 38
Rest see also Non-surgical treatment
cage confinement 125, 126, 131,
228, 288, 305
lumbosacral surgery 192
spinal trauma 282, 288

124, 125
Restrictive fibrosis 225, 273
Reticulosis (focal granulomatous
meningoencephalomyelitis) 248
Retractors
AdsonBaby 74
Army Navy 152, 233
blunt 74
Gelpi 74, 108, 138, 235, 236, 237,
238
Gosset 74
Langenbeck 138
multi-toothed 74
pediatric Balfour 74
self-retaining 74, 197
Senn 138
Weitlander 74
Retriever
bulbous dilation of dorsal arachnoid
space 324
left hemiparesis 29
myelomalacia 61
thoracolumbar spinal cord
segments 3
vertebral bone loss 61
Rhizotomy 148
Rhodesian ridgeback 222
Rib (thirteenth) 140, 155
Rickettsia 42, 330
Rongeurs 76
Root signature 93, 94, 103, 212, 279
Rottweiler
CSM surgery 244
discospondylitis 328
laminectomy 201, 244
leukodystrophy 320321
leukoencephalomalacia 320
lumbar tumor 254
neuroaxonal dystrophy 320321
spinal compression 216
S
Sacral vertebrae
S1/S3, 1
S3, 37, 3738
Sacrocaudal spine
dysgenesis 36, 37, 322
tail injury 296297
Sacroiliac joint pain, cervical 36, 184,
320
see also Lumbosacral disease
Sacrum 3
Saphenous nerve 31
Sarcoma 248, 261

Scalenus muscle 277, 278
Scalpel 75
Scar formation 8687, 191, 192
SchiffSherrington sign 26, 32
SchiffSherrington syndrome 32,
284
Schmorls node 36, 94, 184
Schnauzer (miniature)
decubital ulcers 359
fibrocartilaginous embolism 332
Sciatic nerve 31
Scintigraphy 5960
Scoliosis 61, 127
Screaming (in dogs) 94
Screw fixation see also
Fusionfixation; Pin fixation
174176
CSM 220, 223, 225, 226,
237240
failure rates 169
lumbar spine 274, 296
lumbosacral spine 190, 191, 192,
296
pelvic limb lameness 204
225, 226
techniques
cement plug 237, 238
laminectomy 191, 192, 208
locking plates 223
metal and cement fixation
239240
placement 207, 208
threaded screws 240, 292, 309
ventral slot 220
vertebrectomy 274
trauma 296
Sedatives 282
Sedimentation chambers 44
Seizures
after myelography 5354, 86, 339
after neurosurgery 339
prazosin 352
Self-mutilation 127, 183, 359
Sensory function see also Neurological
examination; Nociception;
Proprioception
ascending sensory tracts 5
facial sensation 20
Sepsis (postoperative) 99, 229, 259,
297
Septic arthritis 20
Septic shock 83
Seroma formation 99, 191, 229, 358
Sevofluorane 85
Index
Sharpei 60
Shih Tzu
chondrodystrophy 12
disc extrusion 94
L4/S3 neurological defects 30
neck pain 94
Shock
septic 83
spinal 32, 282
Silky terrier (disc extrusion) 102
Sinus (pilonidal) 36, 37, 184, 212, 323
Skeletal hyperostosis 36, 326
Skeleton 614
Slings and harnesses 346, 347, 349
Small breed dogs see also Miniature
dogs; Toy breed dogs
Soft tissue tumors 94
Spears 77
Specific gravity (CSF) 45
Sphincter tone 191
Spina bifida 36, 37, 184, 322
Spinal arteries 16
Spinal cord
anatomy 1, 16, 24
blood supply 16
evoked responses 6061
nerve fibers 6
segments 13
ultrasound examination 55
white matter tracts 56
Spinal cord disease
compression 227, 228
expansion 57
hematoma 36, 37, 94, 212,
333334
hemorrhage 36, 51, 94, 281,
333334
herniation 147, 273
infection 330
injury 87
ischemia 86
liquification 132
malacia 287
swelling 127
tethering 36, 37, 149, 184, 322
tumors 59, 257258
Spinal dysraphism (myelodysplasia)
36, 212, 322
Spinal ganglion 3
Spinalis muscle 267
Spinal nerves 3, 183, 191, 277,
278
Spinal palpation 2425
Splints, external 239, 290, 305, 306
Spondylitis 326328
Spondylomyelopathy (cervical) 20,

35, 36, 94, 71, 211246, 281
see also Wobbler syndrome
Spondylosis deformans 185, 320, 321
Sponges 77, 78, 79, 141
Stabilization
caudal disc extrusion 96
Staphylococcus 327, 330
Stapling (spinal) 316
St Bernard 238
Steatitis (epidural) 330331
Steinmann pins 221, 290292, 296,
298, 310, 314, 315
Stenosis
cervical fibrotic 36, 37, 212
vertebral canal 183
Sternocephalicus muscles 107
Sternohyoid muscles 107
Sternothyroid muscles 172, 172
Steroid-responsive meningitisarteritis
329
Steroids 83, 326, 329, 357
see also Corticosteroids; Nonsteroidal anti-inflammatory
drugs
Stilette 65
Stranguria 184
Streptococcus 327
Streptomycin 327
Subdural hemorrhage 57
Subdural space 4
Sucralfate 344, 356
Suction tip and bulb syringe 77
Supraspinous ligaments 14
Surgical instruments 7379
Surgicel 77
Surgifoam 77
Survey radiography 163164, 213,
283
Suturing 276
Swallowing 20
Swimming therapy 348349
Synovial articulations 911
Synovial cyst 320
cervical 215
column splitting 52
CSM 212, 219, 245
differential diagnosis 36, 94, 122
lumbosacral 183
Synthes locking plate 219, 220, 221
Syringes
collection of CSF 67
hypodermic needles 76
suction tip and bulb syringe 77
Syringohydromyelia 163
Syringomyelia 36, 37, 51, 94, 163,
212, 297, 311, 322
T
Tail defects
Tail injuries 296297
Tartar scrapers 76
Technical errors 258
Temperature, pulse, respiratory rate
(TPR) 343, 344
Temporal muscle mass 20
Temporomandibular joint lesion 94
Tendonitis 20
Tendons 20, 271
Terazosin 352
Tetanus 38
Tethered spinal cord 36, 37, 149,
184, 322
Tetracyline antibiotics 331
Thalamic pain syndrome 38
Thiamine deficiency 36
Thoracic spine
anatomy 89
aorta 10
blood supply 15
localization of injury 296
mid-bodies 10
radiography 4748
diagnosis
clinical signs 121122
CSF analysis 122
differential 122
imaging techniques 56, 122123
kyphosis 131
prognosis 132133
surgery
complications 127, 129,
130132
disc extrusion 127, 129, 131
decompression 124, 126127,
130
hemilaminectomy 126,
126127, 136150,151154
kyphosis 131
laminectomy 126, 270273
pediculectomy 126, 127,
151154
recovery times 124125
377
378
Index
Thoracolumbar vertebrae
T1/T13, 1
T2, 36
T3/L3, 15, 28, 36, 37
Thrombin 99
Thrombocytopenia 82
Thromboembolism 359
Thrombosis (iliac) 332333
Thyroid gland 20, 172, 217
Thyroid hormone supplementation
217
Thyroid-stimulating hormone (TSH)
217
Tibial crest avulsion 20
Tibial nerve 31
Tick paralysis 38
Toxicity
organophosphate 38, 42
toxic disease 36
toxic myopathy 38
zinc 358
Toxoplasma gondii (toxoplasmosis)
42, 330
Toy breed dogs 165, 325
see also Miniature dogs
TPR (temperature, pulse, respiratory
rate) 343, 344
Trachea 108, 171
Traction (response to) 213, 214,
223, 224
Transverse foramen 7, 8
Transverse ligament of the atlas
14
Trauma 281318 see also
bile duct rupture 19
biomechanics 286289
diagnosis
anatomical location of the injury
295297
decision-making algorithm 293,
294
differential 36, 184
imaging techniques 283,
283286
initial assessment 281282
neurological signs 28, 282283,
293295
fracture 21, 36, 87, 286289
gunshot injury 36, 331
neurology
deep pain sensation 293295
examination 21
pathophysiology 289292
psoas muscle 331
prognosis 301302
recovery 87
spinal region
cervical 295297
lumbar 296
lumbosacral 184
sacral 36, 296297
thoracic 282, 296
treatment
bone grafts 292
choice 293295
fixation 2890291, 316318
metal and bone cement
291292, 309316
modified segmental 291
305309
urinary tract 282
Traumatic feline ischemic myelopathy
36, 37, 281
Trimethoprimsulfonamide 85, 331,
343, 355
Tube cystostomy 355
Tumoral calcinosis (calcinosis
circumscripta) 36, 37, 122, 212,
326
Tumors 252254 see also Neoplasia;
specific types of tumor
biopsy 6162
classification 248, 253
seeding 276
spinal 59, 257258, 261263
U
Ulcer formation
decubital 339, 358359
gastrointestinal 127, 297, 341,
356
Ulnar nerve 31
Ultrasonography 5455, 250
Ultrasound therapy 347
Upper airway disease 20, 38
Upper motor neurons (UMN); see
also Lower motor neurons
bladder function 20, 24, 351
crossed extensor reflex 32
EMG 60
motor evaluation 2526
reflexes 6, 26
severity 3132
Ureter calculi 38, 122
Urethra
innervation 351
neoplasia 184
tumor 38
Urinary system see also Bladder

antiseptic 352
function 21, 22, 2324, 351352
incontinence 24, 183, 302
infection 42, 81, 82, 83, 131, 297,
355
micturition disorders 351
neurological damage 24
pharmacology 352
physiological control 350
Urine
urine retention 182
urinalysis 42, 339, 343, 344
urine scald 127, 297, 339,
357358
voiding change 339
V
Vascular disease 20, 36, 329,
332334;
see also Cardiovascular disorders
Vascular system see also Blood supply
arteries
aorta 10, 332333
intercostal 15, 278
occipital 15
omocervical 276
superficial cervical 278
blood supply
spinal regions
cervical spine 15
lumbar spine 16
lumbosacral spine 183
spinal cord 16
thoracic spine 15
vertebral column 1417
veins
azygous 15
jugular 278
venous drainage 17
vena cava 10
venous plexus 114, 145146,
148
Ventilation (mechanical) 101
Ventral decompression (slot)
comparison to fenestration 97
complications 98, 99, 227, 228
CSM 218, 218, 219221, 227,
228
laryngeal nerve 108
postoperative care 100, 101,
102
recovery 103
Index
Ventral fissure 16
Vertebrae (general)
anticlinal 9
blood supply 1417
column splitting 52
congenital anomalies 321322
intervertebral discs 1112
ligaments 1314
synovial articulations 911
Vertebral buttress 287289
Vertebral canal 7, 8
cauda equina compression 181
CSM 211, 219
tumors 253
type I disc extrusion 14
Vertebral column 1417
Vertebral distraction/stabilization
218, 219, 221223
see also Fixationstabilization;
Internal fixation devices; Metal
and cement fixation
Vertebral foramen 141
Vertebral fracture 225, 331
Vertebral instability 211
Vertebral plexus 15, 269, 273
Vertebral tumors 253, 255257
Vertebrectomy 255, 273274
Vestibular disease 85
Viral infections 330
Visual function
blindness 85
neurological examination 20, 21
placing test 23
ptosis 25
Vitamin A 36, 37, 326
Vitamin E 217, 257
Vomiting 339, 355356

Von Willebrand (VW) factor/disease
cervical hematoma 360
CSM surgery 229
diagnosis 217
hypothyroidism 217
W
Walking ability
CSM surgery 229
lumbar vertebrectomy 274
pain 32
paraparesis 32
paraplegia 32
spinal reflex 87
supported 346
thoracolumbar disc surgery 124,
130
ventral slot surgery 102, 103
Warfarin 333
Washing (postoperative) 343
Water bath therapy 348349
Water intake 85, 343, 344345,
355
Waterproof drapes 74
Wedges (foam) 46
Weimaraner
spinal dysraphism 322
Westie (marsupialization) 325
Wheelbarrowing test 23
White blood cells (WBC)
CSF analysis 44, 45
meningitis 329
urinalysis 42
White matter tracts 56

Wire fixation
178180
dorsal 178180
failure rates 169
K-wires 117, 167168, 235
Withdrawal reflex 20, 28, 31
Wobbler syndrome see also Cervical
spondylomyelopathy
domino lesions 101
drug-induced seizures 86
horses 218
MRI 58
surgery 77, 218
Working dogs (lumbosacral disease)
183, 188
Wounds and wound infections
discharge 339
dressings 344, 357
obesity 83
postoperative 99, 127, 128, 297,
339, 357
predisposing factors 83, 357
prevention 357
septic shock 83
X
Xanthochromia 45
Y
Yorkshire terrier 43
Z
Zinc toxicity 358
379

Small Animal Spinal Disorders (Second Edition)

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Small Animal Spinal Disorders (Second Edition)

Uploaded by

Copyright:

Available Formats

Dedication

Professor Ian R. Griffiths, Glasgow University Medical School

2005, Elsevier Limited. All rights reserved.

Copyright 1994 Times Mirror International Publishers Limited

First edition 1994

The second edition of Small Animal Spinal Disorders

Vancouver and Hertfordshire 2005

Knowledge of functional anatomy is important both

The spinal cord lies within the vertebral canal, fitting

Small Animal Spinal Disorders

1.2 Position of the spinal cord segments

The rest of the cervical spinal nerves exit the vertebral

Relationship of spinal cord segments

Some spinal cord segments lie in the vertebra of the

The nerves of the cauda equina have a typical peripheral

The meninges (1.10) surround the central nervous

Small Animal Spinal Disorders

1.8 A: Lumbosacral plexus: femoral (a);

1.11 Section of spinal cord to show the subdural space (between

cord within the CSF. The denticulate ligaments also

sacrocaudal vertebrae. The meningeal sac is outlined by

1.12 The reflex arc of the LMN is influenced

Upper motor neuron

Lower motor neuron

leptomeninges (pia-arachnoid mater) and the ependymal lining.

Spinal cord white matter tracts

carried by non-myelinated fibers, particularly in the

DESCENDING MOTOR TRACTS

Small Animal Spinal Disorders

1.13 The UMN system is the sum effect of the

ASCENDING MOTOR TRACT

Spinal cord nerve fibers and the

bear weight, loss of voluntary movement and, finally,

1.14 The atlas C1 (a) articulates with the skull (b)

1.15 CT scans of a normal dog through

surgical landmarks (8.198.21). The total number of

lateral to the vertebral canal and of the lamina dorsally.

Small Animal Spinal Disorders

being composed of lateral masses. Caudally on the body

C6 are particularly large and project ventrally; they are

THORACIC AND LUMBAR VERTEBRAE

1.18 Photograph of the floor of the vertebral canal of C4, C5

1.19 Nutrient foramen (arrows) in the center of

Small Animal Spinal Disorders

1.24 Lumbar vertebrae. The vertebral bodies are long (particularly

1.26 Lumbosacral vertebrae. L7 differs from the other lumbar

1.27 Sagittal MRI of a dog to show that the intervertebral

are ventral, and between the fused sacral vertebrae).

The vertebral bodies are joined by intervertebral discs

Small Animal Spinal Disorders

1.28 The L7 nerve root runs in a lateral recess in the floor of

Degeneration of the intervertebral discs occurs with

DEFINITION OF DISC LESIONS

1.30 Hansen type I disc extrusion is most common in

et al., 1995). The classification scheme used most

1.31 Hansen type II disc herniation. This occurs mainly

classical like the Beagle, to less obvious like the Basset

The nuchal ligament extends from the dorsal arch of the

Small Animal Spinal Disorders

1.32 The cut-away portion reveals the weak ventral longitudinal

contributes to the low incidence of intervertebral disc

1.35 The most significant ligament between C1 and C2 is

Small Animal Spinal Disorders

1.38 The lumbar spine is supplied by spinal