imaging dogs with suspected disc herniation: pros and cons of myelography, computed tomography, and...
TRANSCRIPT
IMAGING DOGS WITH SUSPECTED DISC HERNIATION: PROS AND CONS
OF MYELOGRAPHY, COMPUTED TOMOGRAPHY, AND MAGNETIC
RESONANCE
IAN ROBERTSON, DONALD E. THRALL
Myelography, computed tomography (CT), and magnetic resonance (MR) imaging are the diagnostic
modalities currently used in the evaluation of dogs with suspected disc herniation. Where high-field MR imaging
is available, it is considered the optimal modality for any myelopathy in dogs, including those with disc disease.
CT myelography may be the next best option, particularly in nonchondrodystrophoid dogs. In
chondrodystrophoid dogs, in which extrusion of mineralized disc material is common, plain CT will enable
diagnosis in most cases. Myelography is still considered adequate for diagnosis of disc herniation when MR and
CT are unavailable. r 2011 Veterinary Radiology & Ultrasound, Vol. 52, No. 1, Supp. 1, 2011, pp S81–S84.
Key words: computed tomography, dog, intervertebral disc disease, MR imaging, myelography.
Introduction
SUSPECTED INTERVERTEBRAL DISC herniation is a common
indication for spinal imaging in dogs. Although survey
radiography is often the modality used first, particularly in
first-opinion practice, it is an inaccurate method for diag-
nosis of disc herniation and is subject to marked interob-
server variation.1 Survey radiographic signs associated with
disc herniation provide only limited, indirect evidence
about the spinal cord injury responsible for the clinical
signs. Imaging dogs suspected of having intervertebral disc
herniation with a modality other than survey radiography
is a prerequisite to surgical intervention. Accurate assess-
ment of the location, laterality, and degree of spinal cord
compression facilitates surgical treatment and should en-
hance patient outcomes.
Myelography, computed tomography (CT), and mag-
netic resonance (MR) imaging are the diagnostic modal-
ities currently used in the assessment of dogs with
suspected disc herniation. Optimal imaging depends on se-
lecting a modality on the basis of various factors related to
the patient, modality attributes, and availability. Factors
not directly related to the suspected myelopathy affect the
choice of modality including presence of metal near the
region of interest2 (a contraindication to MR, but may also
adversely affect CT), presence of a cardiac pacemaker (a
contraindication to MR), history of a prior adverse reac-
tion to contrast medium (which may inhibit repeat use of
organic iodide or gadolinium-containing media), and a
clinical need for immediate imaging and treatment rather
than referral to a distant center. In patients requiring a
repeat examination (e.g., because of deterioration of signs
or to assess the effects of surgery), comparison is facilitated
by using the same imaging modality as for the original
examination.
Myelography
For many years, myelography was the principal imaging
modality used for diagnosis of disc herniation in animals,
and it is still adequate for this purpose. Nowadays, a de-
cision to use myelography for a dog with suspected disc
herniation is usually based on the lack of availability of
either CT or MR imaging (MRI).
Myelography is opacification of the subarachnoid space
around the spinal cord with the aim of identifying any
deviations that may reflect the site of a spinal cord lesion.
Problems relating to myelography include its invasiveness,
which can cause spinal cord damage,3 potential for adverse
reactions to contrast medium (e.g., exacerbation of mye-
lopathy, seizures), and the tendency for many spinal
lesions, including disc herniation, to narrow the sub-
arachnoid space, which excludes contrast medium and
hence may obscure the nature of the lesion. Myelography
can be completed within 30min, but difficulty making the
injection may greatly prolong the procedure. Similarly, any
epidural contrast deposition following inaccurate needle
placement will reduce the diagnostic quality of the study.
The two-dimensional nature of radiographs means that
superimposition by unrelated structures may make it diffi-
cult to visualize small lesions, particularly in the thoracic
spine. The superior latitude afforded by digital radiogra-
Address correspondence and reprint requests to Ian Robertson, at theabove address. E-mail: [email protected] November 10, 2010; accepted for publication November 12,
2010.doi: 10.1111/j.1740-8261.2010.01788.x
From the Department of Molecular Biomedical Sciences College ofVeterinary Medicine, North Carolina State University, Raleigh, NC 27606
S81
phy systems, particularly flat-panel detectors, enhances le-
sion conspicuity compared with film-screen radiography
and may be augmented by image manipulation at the
workstation.
CT
CT is an accurate imaging modality for dogs with sus-
pected disc herniation.4–7 The tomographic (slice) imaging
of CT eliminates the problem of superimposition that
affects radiographs and the improved contrast resolution
of CT enables detection of mineralized disc material or
hemorrhage in the vertebral canal. In chondrodystrophoid
patients, where disc mineralization is much more common,
conventional helical/multidetector CT will yield a correct
diagnosis of disc herniation in the majority of affected
dogs. Helical and multidetector scanners enable rapid,
thin-slice imaging of the entire spinal column and high-
quality multiplanar reconstructions, which can greatly en-
hance lesion conspicuity. While single slice scanners can
acquire thin slices, the acquisition time and tube heat load-
ing often make this impractical for large patient volumes.
With thicker slices (4–5mm), volume averaging can reduce
lesion conspicuity. If CT is available but MRI is not, plain
(noncontrast) CT is a logical first test to detect disc her-
niation in chondrodystrophoid dogs. For other breeds, or
for conditions other than disc herniation, myelography
may be a more suitable first choice if MRI is not available.
CT Myelography
CT imaging after opacification of the subarachnoid
space combines the benefits of CT and myelography. Con-
centrations or volumes of contrast medium in the sub-
arachnoid space that would not be visible radiographically
are readily detectable with CT. This means that a markedly
narrowed subarachnoid or uneven contrast medium dis-
tribution can be tolerated in a CT study when it might
render a myelogram nondiagnostic; however, the other
disadvantages of myelography relating to difficult injec-
tion, cord damage, and adverse reactions remain. The de-
cision to perform CT myelography often follows an
inconclusive myelogram or plain CT study. In the absence
of high-field MRI, CT myelography (þ /� intravenous
contrast medium administration) is often the next best op-
tion, particularly in nonchondrodystrophoid dogs.
CT Myelography Augmented with Intravenous ContrastMedium
Intravenous administration of contrast medium can
aid assessment of lesions affecting the epidural space and
intervertebral foraminae. This technique is more beneficial
for assessment of patients with neoplastic or inflammatory
disorders affecting the vertebrae or paravertebral tissues
than those with disc herniation.
MRI
MRI has revolutionized neural imaging and, if available,
is a suitable imaging modality for most neurologic patients.
In many veterinary institutions, MRI has largely replaced
myelography because it avoids the need for intrathecal
contrast medium administration and provides a relatively
complete anatomic depiction of the spinal cord and ver-
tebral column. Uniquely with MRI, lesions affecting the
spinal cord can be visualized directly.� Use of a heavily
T2-weighted single-shot turbo spin-echo MR sequence8
accentuates the high signal from cerebrospinal fluid
and provides a rapid, noninvasive evaluation of the sub-
arachnoid space that appears similar to a conventional
myelogram and provides similar insight into the morpho-
logy of the subarachnoid space.
There is considerable variation in MRI quality in vet-
erinary practice. It may be difficult to obtain diagnostic
studies of the spine with low-field MR, particularly in small
patients. Low-field permanent magnets have low signal to
noise ratio (SNR) and relatively small field of view, which
may necessitate repositioning the patient repeatedly to ob-
tain images of the entire region of interest.w Enlarging the
field of view, increasing slice thickness, and increasing the
number of acquisitions will increase SNR, but these ad-
justments reduce image resolution and prolong anesthesia
and magnet time, potentially exacerbating patient hypo-
thermia. If high-field MR is available, it is an excellent
first-line test to assess disc herniation. A comprehensive
high-field MR study of the spine takes approximately
40–45min, slightly longer on average than a myelogram.9
If only a low-field magnet is available, there may be an
advantage of performing myelography as the initial screen-
ing procedure, assuming that CT is not available, at least in
large dogs.
Myelography vs. CT
At veterinary facilities with routine access to both my-
elography and CT it may not be clear which test should be
performed first for a dog with suspected disc herniation.
Results of these modalities have been compared retrospec-
tively in 182 dogs with disc disease.4 Limitations of this
study arise because the choice of imaging modality was
clinician-dependant (rather than randomized), and few en-
rolled dogs were imaged using both modalities; however,
the number of dogs is large and the diagnosis was known in
each dog, thus some reasonable conclusions are possible.
�See accompanying article ‘‘Optimal Magnetic Resonance Imaging ofthe Spine,’’ pp. S72–S80.wSee accompanying article ‘‘Pros and Cons of Low Field Magnetic
Resonance Imaging in Veterinary Practice,’’ pp. S5–S14.
S82 ROBERTSON andTHRALL 2011
The overall sensitivity for disc herniation was similar for
myelography (84%) vs. CT (82%). CT was more sensitive
than myelography in chronically affected dogs, and my-
elography was more sensitive than CT in dogs weighing
o5 kg.
CT and myelography were compared prospectively in 19
chondrodystrophoid dogs with acute disc herniation.6
Dogs had CT followed by myelography and all lesions
were confirmed surgically. This study concluded that my-
elography and CT were equally sensitive for acute disc
herniation in chondrodystrophoid dogs.
CT and myelography were also compared prospectively,
along with contrast-enhanced and CT myelography, in 46
dogs with acute cervical or thoracolumbar myelopathy.4
This study included 25 chondrodystrophoid and 21
chondrodystrophoid dogs; the order of imaging was con-
sistent and the diagnosis known for all dogs. CT myelo-
graphy was the most sensitive technique for identification
of lesions compressing the spinal cord, enabling detection
of lesions in some dogs with normal myelograms. Also, in
some dogs with spinal cord swelling, lesion laterality could
be determined with CT when it was not evident myelo-
graphically. Plain CT was adequate for detection of her-
niation of mineralized disc in chondrodystrophoid dogs,
but was inadequate for this purpose in many non-
chondrodystrophoid dogs. In the latter group, the X-ray
attenuation by the herniated disc material was not suffi-
ciently different from the adjacent cord to be reliably de-
tected. Based on these results, plain CT should be the
imaging modality used first for chondrodystrophoid dogs
with suspected acute disc herniation, and myelography
should be used first for patients in which other spinal dis-
orders are considered more likely than disc disease.
That some dogs receiving myelography first would have
to undergo subsequent CT imaging for complete charac-
terization is understood and the practice of performing CT
myelography immediately after a conventional myelogram
is common, particularly if the myelographic findings are
equivocal or the study nondiagnostic for technical reasons;
however, even with CT, epidural contrast medium can be
difficult to differentiate from mineralized disc material if
there is no precontrast CT study available for comparison.
Epidural contrast medium can be avoided by performing a
cisternal injection, raising the head, and allowing the con-
trast medium to gravitate caudally. Cisternal injection is
not usually recommended for myelography in dogs with
suspected thoracolumbar disc herniation because flow of
contrast medium may be obstructed even by minimally
compressive lesions; however, the superior contrast reso-
lution of CT enables detection of the small amount of
contrast medium that may flow past the lesion. Similarly,
sufficient intrathecal contrast usually remains for CT
myelography several hours after injection, which is useful
if postoperative imaging is considered necessary.
Myelography and CT vs. MR
There have been few studies to compare MRI with my-
elography and/or CT myelography in dogs with suspected
disc herniation. When MRI and myelography were both
performed in a series of dogs with caudal cervical spondy-
lomyelopathy, MRI was more accurate in predicting the
site, severity, and nature of spinal cord compression.10
Myelographic or CT myelographic assessment of lesions
is difficult when there is circumferential attenuation of the
subarachnoid space that excludes contrast medium, for
example in dogs with acute disc extrusions and extensive
epidural hemorrhage.7 While the center of a region of sub-
arachnoid space attenuation on a myelogram often ap-
proximates the site of spinal cord injury, confirming the site
of a compressive lesion and its laterality, which is impor-
tant for surgical planning, may be impossible. The diag-
nostic accuracy of high-field MRI is not usually affected by
attenuation of the subarachnoid space. In such cases, MRI
may allow the subarachnoid attenuation due to spinal cord
swelling (e.g., as a result of edema associated with acute
disc herniation, infarct, myelitis, or a mass) to be distin-
guished from an extradural lesion.
Before MRI, it is likely many patients with acute non-
compressive nucleus pulposus extrusions were subjected to
unnecessary spinal decompression, justified on the basis
that a compressive lesion could not be ruled out without
surgical investigation. In addition to the acute manifesta-
tions of disc disease, MRI is more sensitive for chronic
spinal cord lesions such as atrophy or syringomyelia. Spi-
nal parenchymal lesions are apparent with myelography or
with CT myelography only if they cause swelling of
the cord or if contrast medium enters the cord, which is
uncommon.11
In facilities where myelography, CT, and MRI are rou-
tinely available, the image quality for each modality is
commensurate with modern expectations, and cost is not a
significant factor, we consider MRI to be first choice for
imaging canine spinal patients; however, there are many
occasions when the above prerequisites are not fulfilled or
the patient’s condition suggests that an alternative modal-
ity should be selected.
There are instances where myelography or CT myelo-
graphy provide information not readily available with
MRI. For example, it can be difficult to determine if the
origin of a large spinal lesion is dural or parenchymal using
either MRI or CT, and in such cases, myelography may be
the best modality for this determination.12,13
Dynamic instability associated with caudal cervical
spondylomyelopathy or lumbosacral disc disease is an-
other example of a condition for which myelography may
be preferred. Myelograms showing the effects of traction,
flexion, and extension are easier to obtain than corre-
sponding MRI, although such positional views can be ac-
S83IMAGING SUSPECTED DISC HERNIATIONVol. 52, No. 1, Supplement 1
quired with a dog in the bore of a magnet,14 with minimal
prolongation of magnet time.
CT and CT myelography may be preferred to MRI for
the assessment of vertebral fractures. The ability to acquire
1mm or submillimeter CT slices in a bone kernel (with
subsequent multiplanar reconstruction, including surface
and volume rendering) enables depiction of complex frac-
tures in a format that can assist in surgical planning. Thin
slice acquisition is also possible with high-field MR sys-
tems, but acquisition times may be long and/or limited to
gradient echo sequences, which have less tissue contrast. In
trauma patients with spinal fractures MRI is considered
superior to CT for assessment of soft tissue injuries asso-
ciated with fractures.
There is evidence that MRI can provide prognostic
information in spinal patients. For example, in studies
of 77 dogs,15 and 159 dogs,16 the extent of spinal cord
T2 hyperintensity (cross section and cord length) was
negatively associated with the probability of a full
functional recovery. Such studies suggest there may
be a benefit from MRI in addition to that associated with
more accurate diagnosis of spinal cord disorders. Even
so, it may be prove difficult to demonstrate improved
outcomes9 for spinal patients having MRI instead of
myelography.
ACKNOWLEDGMENT
Disclosure: The authors declare no conflict of interest.
REFERENCES
1. Lamb C, Nicholls A, Targett M, Mannion P. Accuracy of surveyradiographic diagnosis of intervertebral disc protrusion in dogs. Vet RadiolUltrasound 2002;43:222–228.
2. Parry AT, Tanner A, Chandler K, Lamb CR. What is your diagnosis?Foreign body. J Am Vet Med Assoc 2010;237:359–360.
3. Kishimoto M, Yamada K, Ueno H, Kobayashi Y, Wisner ER. Spinalcord effects from lumbar myelographic injection technique in the dog. J VetMed Sci 2004;66:67–69.
4. Israel SK, Levine JM, Kerwin SC, Levine GJ, Fosgate GT. The rel-ative sensitivity of computed tomography and myelography for identificationof thoracolumbar intervertebral disk herniations in dogs. Vet Radiol Ultra-sound 2009;50:247–252.
5. Hecht S, Thomas WB, Marioni-Henry K, Echandi RL, MatthewsAR, Adams WH. Myelography vs computed tomography in the evaluationof acute thoracolumbar intervertebral disk extrusion in chondrodystrophicdogs. Vet Radiol Ultrasound 2009;50:353–359.
6. Dennison SE, Drees R, Rylander H, et al. Evaluation of differentcomputed tomography techniques and myelography for the diagnosis ofacute canine myelopathy. Vet Radiol Ultrasound 2010;51:254–258.
7. Olby NJ, Munana KR, Sharp NJ, Thrall DE. The computed tomo-graphic appearance of acute thoracolumbar intervertebral disc herniations indogs. Vet Radiol Ultrasound 2000;41:396–402.
8. Pease A, Sullivan S, Olby N, et al. Value of a single-shot turbospin-echo pulse sequence for assessing the architecture of the subarachnoidspace and the constitutive nature of cerebrospinal fluid. Vet Radiol Ultra-sound 2006;47:254–259.
9. Parry AT, Harris A, Upjohn MM, Chandler K, Lamb CR. Doeschoice of imaging modality affect outcome in dogs with thoracolumbarspinal conditions? J Small Anim Pract 2010;51:312–317.
10. da Costa RC, Parent J, Dobson H, Holmberg D, Partlow G. Com-parison of magnetic resonance imaging and myelography in 18 Dobermanpinscher dogs with cervical spondylomyelopathy. Vet Radiol Ultrasound2006;47:523–531.
11. Lu D, Lamb C, Targett M. Results of myelography in seven dogswith myelomalacia. Vet Radiol Ultrasound 2002;43:326–330.
12. Kippenes H, Gavin PR, Bagley RS, Silver GM, Tucker RL, SandeRD. Magnetic resonance imaging features of tumors of the spine and spinalcord in dogs. Vet Radiol Ultrasound 1999;40:627–633.
13. Drost WT, Love NE, Berry CR. Comparison of radiography, my-elography and computed tomography for the evaluation of canine vertebraland spinal cord tumors in sixteen dogs. Vet Radiol Ultrasound 1996;37:28–33.
14. Penderis J, Dennis R. Use of traction during magnetic resonanceimaging of caudal cervical spondylomyelopathy (‘‘wobbler syndrome’’) inthe dog. Vet Radiol Ultrasound 2004;45:216–219.
15. Ito D, Matsunaga S, Jeffery ND, et al. Prognostic value of magneticresonance imaging in dogs with paraplegia caused by thoracolumbar inter-vertebral disk extrusion: 77 cases (2000–2003). J Am Vet Med Assoc2005;227:1454–1460.
16. Levine J, Fosgate G, Chen A, et al. Magnetic resonance imaging indogs with neurologic impairment due to acute thoracic and lumbar inter-vertebral disk herniation. J Vet Intern Med 2009;23:1220–1226.
S84 ROBERTSON andTHRALL 2011