Contrast-Enhanced CT for Localizing Compressive Thoracolumbar Intervertebral Disc Extrusion

Introduction

Spinal cord compression and injury by an extruded intervertebral disc may produce paresis, paralysis, and/or pain. Each of these sequelae can significantly affect the patient's quality of life. Hansen's type I disc (extrusion) in the thoracolumbar region is most common, accounting for approximately 86% of cases.¹ The majority of type I extrusions occur in chondrodystrophic breeds.

Treatment of intervertebral disc extrusion often involves surgical decompression via laminectomy or hemilaminectomy. Determining the level and laterality of disc extrusion is critical to preoperative planning. This requires advanced imaging procedures and modalities that include myelography, computed tomography (CT), and magnetic resonance imaging (MRI). When disc material and spinal cord compression are not found at the initially operated site, additional surgical procedures must be performed at adjacent sites in the spinal canal. This increases anesthesia time, which can lead to anesthetic complications and increases the rate of postoperative infection.² Increased surgical manipulation can result in additional spinal cord trauma and could result in a decline in neurologic function postoperatively.

Myelography is one of the most widely used imaging procedures for identification of spinal cord compression in dogs. It involves injecting contrast medium into the subarachnoid space followed by radiographic imaging to aid in lesion identification. Myelography has been reported to improve the accuracy of localizing intervertebral disc protrusions and extrusions to 97% compared with 72% for survey radiographs.^3,4 Myelography is invasive and associated with the risk of serious complications.^5–8 CT of the spine has long been used in human medicine to identify herniated disc material in the lumbar spine.^5,9–13 Presently, MRI is used most commonly in human medicine; however, CT has been shown to be superior to myelography with an accuracy of 76% and a sensitivity of 100% compared with myelography (in which the accuracy is 57–71% and the sensitivity is 89.6%).^5,11,14,15 In veterinary medicine, MRI is expensive and access is limited to teaching hospitals or large referral practices. CT is becoming more readily available in private referral practice and the price of a CT is similar to that of myelography.

Subarachnoid injection of iodinated contrast medium is frequently performed prior to spinal CT evaluation to assist in localizing the herniated disc material.^5,9,13 Using subarachnoid contrast with CT is equally as invasive as myelography and the potential for complications remains. Two veterinary studies have already shown that CT without subarachnoid contrast injection can be useful in localizing extradural spinal cord compression. One of these studies reported that CT was successful in identifying both mineralized disc material and acute hemorrhage in the vertebral canal without subarachnoid contrast medium injection.¹⁶ The authors were also able to differentiate between acutely and chronically extruded disc material based on measurement of Hounsfield units.¹⁶ In the second report, 20 dogs were diagnosed with intervertebral disc extrusion and CT alone was 90% accurate for determining the location and 96% accurate for determining the laterality of disc extrusion.¹⁷ This was compared with myelography, which was 88% accurate for location and 92% accurate for laterality of extruded disc material.¹⁷ The improved diagnostic results seen with CT can dramatically reduce perioperative risk and postimaging complications.¹⁷

Human studies of lumbar intervertebral disc extrusions and protrusions have reported similar results for CT compared with myelography. CT is generally superior to myelography, but is inferior to MRI.^5,9–14 The objective of this study was to compare CT with intravenous (IV) contrast medium injection (CTIVC) with CT myelography (CTM) for sensitivity of diagnosis of disc extrusion or protrusion in the thoracolumbar region. The hypothesis was that CTIVC would be as sensitive as CTM for identifying an extradural compressive spinal cord lesion.

Materials and Methods

Records of all dogs that had surgery for thoracolumbar intervertebral disc extrusion from 2004 to 2007 were reviewed from two specialty referral hospitals (Veterinary Specialists of Nevada in Reno, NV and Veterinary Medical Surgical Group, Ventura, CA). Eighty-four dogs were identified and information on age, sex, breed, imaging modality, contrast agent used, imaging results, and surgical findings were recorded. All animals were imaged under general anesthesia and all images were initially evaluated by board certified radiologists. For the purpose of this study, all images were reviewed a second time by a single, board certified radiologist who was blinded to both the previous radiologists' findings and to the surgical findings.

CTM cases had a myelogram performed, which was followed by axial CT scanning using a fourth generation scanner^a (Figures 1, 2). Patients were positioned in dorsal recumbency with the forelimbs extended caudally. Scanning was performed from the ninth or tenth thoracic vertebrae (T) to the third lumbar vertebra (L) using transverse CT slices obtained perpendicular to the vertebral canal. Technique settings varied based on patient size, but ranged between 100–160 kVp, 160–200 mAs, 15–20 cm field of view, and 1–3 mm axial slice thickness. All images were acquired using a soft tissue algorithm. Iohexol contrast medium^b was used in all cases at a dose of 0.3–0.4 mL/kg (72–96 mg I/kg).

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

CTIVC was performed using a third generation scanner^c (Figures 3, 4). The patients were positioned in dorsal recumbency with the forelimbs extended caudally. Scanning was performed from T2 to the second sacral vertebra (S). Technique settings varied based on patient size. For patients <6.8 kg, settings were: 120 kVp, 160 mAs, 15 cm field of view, and 2 mm helical slice thickness with a pitch of 1.5. For patients >6.8 kg, settings were: 160kVp, 160 mAs, 20 cm field of view, and 3 mm helical slice thickness with a pitch of 1.5. All images were acquired using a soft tissue algorithm. All scans were performed following a single IV bolus injection of iodinated contrast medium at 2.2 mL/kg (528 mg I/kg) to a maximum dose of 60 mL. Contrast was administered through either an 18 or 20 gauge catheter in either the cephalic or saphenous vein. Scans were performed within 5 min of contrast injection. The IV contrast agents used were diatrizoate sodium^d (29 cases), diatrizoate meglumine^e (4 cases), and iohexol^b (5 cases).

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

Level of disc extrusion was defined as the disc space underlying the spinal cord affected by the extradural compressive lesion. Laterality was defined as the side of the vertebral canal most affected by the extradural compressive lesion. Laterality determinations were made to aid in surgical planning and were defined as right, left, or ventral for the purposes of this study. Cases that had compressive material in a right ventral or left ventral location were designated right or left according to the side with the most compressive material. If the compressive material could not be designated right or left because the amount of material was not greater on one side, it was designated as simply ventral.

Hemilaminectomy was performed in all cases by a board certified surgeon. Surgery was performed at the location indicated by the imaging examination. The criteria for surgical intervention included clinical signs consistent with spinal cord compression (i.e., pain, neurologic dysfunction) in patients where compression of the spinal cord was confirmed by the imaging modality chosen.

Results

In total, 84 dogs that had surgery for intervertebral disc extrusion were recruited from two separate hospitals during the study period. Of these, 25 were excluded because the CT scans were not performed prior to surgical intervention. An additional nine dogs were excluded for incomplete data. Fifty dogs that had a CT scan performed following either subarachnoid or IV contrast injection with complete medical records available were included in the study (Table 1). Of these, 34 dogs had CTIVC and 16 dogs had CTM.

Table 1 Clinical Characteristics and Computed Tomography (CT) Examination Findings in the 50 Dogs Included in the Study

View Fullscreen

Table 1 (continued)

View Fullscreen

Dachshunds were the most common breed in each group comprising 52% of the CTIVC group and 50% of the CTM. Shih tzu dogs and mixed-breed dogs were the second most commonly represented breeds (4/50). Shih tzu dogs were more common in the CTIVC group, compromising 9% of the total cases (3/34). They were less represented in the CTM group at 6% (1/16). Welsh corgis, mixed-breed dogs, basset hounds, and pugs were equally represented in the CTIVC group at 6% (2/34) each. Pekingese, mixed-breed dogs, and cocker spaniels were equally represented in the CTM group at 12.5% (2/16) each. One Shih tzu dog was included in the CTM group and one of each of the following breeds was included in the CTIVC group: Labrador retriever, German shepherd, and Lhlasa apso.

In the CTIVC group, 38% of the cases were spayed female dogs and 50% were spayed females in the CTM group. In the CTIVC and CTM groups, 47% and 44% were neutered males, respectively. Fifteen and six percent were intact male dogs in the CTIVC and CTM groups, respectively. Mean age at diagnosis was 8 yr for the CTIVC group (range, 1–13 yr) and 6 yr for the CTM group (range, 3–9 yr).

One postprocedural complication was noted in one CTM case. That dog developed anisocoria postoperatively, which resolved after several hours without long-term effects. No postprocedural complications were noted in the CTIVC cases.

The sensitivity of detecting the level of disc extrusion in the CTIVC group was 97% (95% confidence interval [CI], 82.95–99.85) and 94% in the CTM group (95% CI, 67.71–99.67). Sensitivity for laterality detection was 100% in the CTIVC group (95% CI, 87.34–100) and 100% in the CTM group (95% CI, 71.66–100).

Discussion

The objective of this study was to compare CTIVC to CTM. CTIVC was as sensitive as CTM for detecting the level and side of extradural compression. Advantages of CTIVC include speed of image acquisition and lack of necessity for subarachnoid contrast injection, which can be difficult in either arthritic or obese patients.

The overall complication rate in this study was 2%. The rate of complications in the CTM group was 6%. This compares favorably with reported rates for lumbar injection of nonionic iodinated contrast agents.⁶ Seizures have been reported in 5–10% of dogs following lumbar iohexol injection and in up to 35% of dogs following cisternal contrast injection.⁶ Other reported complications of myelography include epidural or central canal contrast injection, apnea, meningitis, and cardiac arrhythmias.^5–8 Contrast medium injection can also cause deterioration of the animal's neurologic condition and iatrogenic damage to the spinal cord can occur during the procedure.^6,20 Inadvertent injection of the contrast agent into either the epidural or subdural space or into the central canal of the spinal cord can make interpretation of the results difficult or render the study nondiagnostic.⁸ Iohexol^b was the contrast agent used in all myelogram cases included in this study. Nonionic contrast agents have been associated with a lower rate of postinjection complications than ionic contrast agents.^6,21,22

Although not a complication of the imaging study, disc material was not found during surgery in one of the included cases. In that case, CTM was unsuccessful in localizing the extruded intervertebral disc material. CTM findings were suggestive of an extradural compressive lesion at T12–13, but at surgery, no disc material was found. One week later, the dog had an emergency MRI, which revealed an extradural compressive lesion at T11–12 with a dural tear and severe spinal cord edema. Intervertebral disc material was found during a hemilaminectomy at T11–12. Intramedullary extension of the extruded disc material through the dural tear was suspected. The severe edema made interpretation of the CTM study difficult.

CTIVC was not associated with any complications in this study. The IV injection of ionic iodinated contrast materials such as diatrizoate sodium or diatrizoate meglumine has been associated with adverse reactions in 6–8% of human patients.^20,23,24 Reactions can either be mild (such as pain on injection, fever, hives, and itching) or they can be severe, such as anaphylactic shock, nephrotoxicosis, or bronchospasm.^21,23 The reported rate of serious complications in humans ranges from 0.04% to 0.2%.²³ Three cases of adverse reactions to ionic contrast have been reported in dogs.^21,22 The rate of adverse reactions to nonionic contrast agents such as omnipaque in humans is lower than that of ionic contrast agents.²³ Serious reactions have been reported to occur in 0.04% of cases.²⁰ To the authors' knowledge, there are no reports of adverse reactions to nonionic IV contrast medium injection in small animals.

Injection of an IV contrast agent was selected because it has been shown to increase the sensitivity of CT for detecting extradural compression in dogs with spinal cord compression requiring surgical intervention.^19,20 The contrast material in the dural venous sinus and adjacent soft tissue structures can help delineate asymmetry in the spinal cord from an extradural compressive lesion.²⁰

Performing CT without subarachnoid contrast injection can prevent the complications frequently seen with myelography and can shorten the procedure, thereby decreasing the duration of anesthesia. The benefits of decreasing anesthetic duration include decreased postoperative infection rates and decreased risk of hypothermia or hypotension during the procedure.

Not all patients with clinical signs consistent with a transverse myelopathy are diagnosed with intervertebral disc extrusion. In such cases, additional imaging with MRI or additional diagnostic testing may need to be performed. For cases with intervertebral disc extrusion in the thoracolumbar region, the authors feel that CTIVC is a suitable method for diagnosing and localizing the affected disc level and side.

One limitation of this study is its retrospective nature. This prevented certain information from being routinely documented, such as duration of the imaging study, duration of the surgical procedure, and a complete surgical description of the exact location and amount of extruded material. A prospective study comparing the different imaging modalities is necessary to evaluate these differences. Further, the assumption that the compressive material was herniated intervertebral disc was made based on gross appearance. Histologic evaluation of the removed material would be necessary to rule out other disease processes. This could also be incorporated into a prospective study. Information that was not routinely recorded in the case record included number of calcified disc spaces on the radiographic images, and complications encountered while performing the myelographic or CT imaging procedures. All of these factors would be valuable for the final determination of a preferred method of imaging for suspected intervertebral disc extrusion.

Another limitation of the study is that the cases came from two different hospitals. As such, no direct comparison between the CTM and CTIVC images could be performed for each patient. A prospective clinical study directly comparing CTIVC to CTM in each patient would better evaluate this imaging modality compared with the currently accepted method for diagnosis.

Conclusion

CTIVC is as diagnostic as CTM for compressive intervertebral disc extrusion in the thoracolumbar vertebral canal. The benefits of CTIVC include an increased speed of image acquisition and decreased complication rate compared with CTM.