Radiology of the Spine
WSAVA 2002 Congress
Ruth Dennis, MA, VetMB, DVR, DipECVDI, MRCVS.
Animal Health Trust
Newmarket, Suffolk, UK


 Spinal pain

 Paraparesis, hemiparesis or tetraparesis/plegia

 Spinal deformity


 Pyrexia of unknown origin


Careful positioning is essential for most radiographic studies of the spine, and this necessitates the use of heavy sedation or general anaesthesia. However, in the traumatised patient in which vertebral instability may be present, the abolition of protective muscle spasm by anaesthesia may be dangerous and so survey lateral radiographs with the patient conscious should be taken first.

Due to the divergence of the primary X-ray beam and the importance of imaging the intervertebral disc spaces, the spine should be radiographed in short segments, using radiolucent foam wedges to make sure that it is aligned parallel to the table top. Good centring and collimation of the X-ray beam will ensure that geometric distortion is minimal and will reduce the amount of scattered radiation produced. In medium and large dogs where the thickness of the tissue to be radiographed exceeds 10cm a grid is needed to reduce the effects of scattered radiation.

Contrast studies

Contrast techniques may be used to outline the spinal cord and cauda equina, which are not visible on plain radiographs. Areas of cord swelling or compression can then be identified.


Opacification of the subarachnoid space using non-ionic, low osmolarity, water-soluble iodinated contrast media at about 300 mg iodine/ml (e.g., iopamidol, NIOPAM; iohexol, OMNIPAQUE). General anaesthesia is essential and the technique should be performed under aseptic conditions. Seizures occasionally occur on recovery, especially if contrast medium has entered the skull. Contrast is injected via either the cisternal or the lumbar route:

1.  Cisternal myelography-injection made between the back of the skull and the arch of C1 into the cisterna magna. Cerebrospinal fluid can be collected in ample amounts from this site, for analysis. Contrast medium warmed to body temperature to reduce viscosity is then injected slowly at a dose rate of up to 0.3ml/kg, depending on the expected site of the lesion.

2.  Lumbar myelography-injection is made between caudal lumbar vertebrae, ideally L5-6 in dogs and L6-7 in cats. The amount of CSF that can be collected from a lumbar puncture is very variable, although in cases of spinal disease lumbar CSF is more likely to be abnormal than CSF obtained from the neck.

Advantages and disadvantages of the two techniques

Cisternal myelography is easier to perform, especially in fat animals, and yields a reliable amount of CSF. However, if resistance is encountered in the subarachnoid space due to the presence of a significant lesion the contrast medium will instead pass cranially into the skull. Thus, contrast may not reach the sites of lesions causing marked cord swelling or compression. Lumbar myelography overcomes this since contrast medium is injected under pressure and has no "escape route" so it will be forced around lesions. Lumbar myelography is usually preferred for suspected thoracolumbar disc disease, to outline the caudal extent of a lesion found with cisternal myelography and to demonstrate the caudal cord if contrast fails to pass a more cranial lesion. However, lumbar puncture may fail to yield an adequate amount of CSF for analysis. Sometimes it is necessary to perform both injections in order to outline a lesion fully.

Myelographic lesions

Normally the spinal cord is outlined by smooth, regular columns of contrast medium. The subarachnoid space is widest at the cisterna magna; the cord widens slightly at the cervico- thoracic junction and the mid-lumbar area because of emerging nerve roots; the cauda equina is outlined with a tapering "fish-tail" appearance. Often the mid-thoracic area is slow to fill with contrast following cisternal puncture.

Lateral radiographs should be taken initially to follow the passage of contrast. When a suspected lesion is identified, VD and oblique views should be taken as necessary to provide 3-D information. At the cervicothoracic junction the DV view is preferred to the VD since this will encourage pooling of the contrast in the area of interest.

Myelographic abnormalities can be divided into the following 3 categories:

1.  Intramedullary-Spinal cord swelling results in divergence and thinning of the contrast columns on all projections.

2.  Extradural-Compression of the spinal cord by a lesion outside the meninges; the contrast column on the affected side of the spinal cord is displaced inwards and thinned or broken with obvious cord narrowing. The orthogonal projection often shows apparent cord widening.

3.  Intradural, extra-medullary-The lesion is between the layers of meninges and creates a filling defect in the subarachnoid space, producing a forked or "golf tee" appearance in profile. The spinal cord is compressed.


Opacification of the epidural space may helpful for assessment of the lumbosacral junction if the dural sac is short or lies too far dorsally over the L7-S1 space. Contrast medium can be injected either between caudal vertebrae in the tail or at the lumbosacral junction. The use of fluoroscopy with image intensification or spot films is usually necessary to check needle placement. The epidurogram is difficult to interpret since the contrast columns are irregular and scalloped.

Stress radiography

Following myelography, flexed, extended and traction views may be indicated in some cases to demonstrate dynamic lesions such as cervical spondylopathy, lumbosacral stenosis and Hansen Type II disc protrusions.


MRI is becoming increasingly available in veterinary practice, and is ideal for spinal imaging as it shows internal spinal cord architecture which radiography fails to do (neoplasia, oedema, haemorrhage, syringohydromyelia). Subarachnoid contrast medium injection is not required and so MRI is safer than myelography.


 Seven cervical vertebrae (C1=atlas, C2=axis). C6 has large transverse processes.

 Thirteen thoracic vertebrae, articulating with ribs.

 Seven lumbar vertebrae.

 Three fused sacral vertebrae.

 Variable number of caudal (coccygeal) vertebrae.

Vertebrae form from several centres of ossification, with separate central epiphyses called "end-plates". Fusion occurs at about 6 months in small animals. The odontoid peg or dens of C2 represents the vestigial vertebral body of C1 and has a separate centre of ossification.


Assess using "Roentgen signs"

 The number of vertebrae (in each section of the spine)

 The shape and size of vertebrae and the vertebral canal

 The alignment of adjacent vertebrae on neutral and stressed views

 Bone architecture; radiographic opacity, presence of osteolysis or new bone production

 The width and opacity of disc spaces and intervertebral foramina

 Myelographic lesions.

The principles of interpretation will be illustrated by showing examples of common spinal diseases including congenital, developmental, degenerative, inflammatory, traumatic, neoplastic and metabolic lesions.


1.  Radiographic examination of the canine spine; Dennis, R. (1987) Veterinary Record 121 31-35.

Speaker Information
(click the speaker's name to view other papers and abstracts submitted by this speaker)

Ruth Dennis, MA, VetMB, DVR, DipECVDI, MRCVS
Animal Health Trust
Newmarket, Suffolk, UK

MAIN : : Radiology of the Spine
Powered By VIN