Abstract: Numerous stabilization techniques have been described for use after spinal fracture/luxation. The purpose of this case report is to describe the use of an external fixator system in three dogs with traumatic spinal instability. The technique described here offers several advantages over conventional methods: a) secondary adjustments are possible postoperatively, if needed; b) implant removal is simple and does not require general anesthesia; c) there is minimal disruption of endosteal and periosteal blood supply.

Editors' Note: This paper describes a method for external fixation of fracture/luxations in the thoracolumbar region of the spine of dogs. In only one of the three cases reported did the patient survive long enough to allow evaluation of all aspects of the treatment. However, in all cases, adequate alignment and stabilization of the spine was accomplished. The reviewers believe the method has potential merit for dealing with a vexing clinical problem and should be published so that others can apply and evaluate it in their own practices.

Many surgical techniques have been described for repair and/or stabilization of the canine spinal column following trauma^1-8. Each technique has its own unique set of reported advantages and disadvantages^1,4,8. The spinal external fixator system described here employs positive-profile end-threaded pins that are inserted into the bodies of each of four adjacent vertebrae after pilot holes are drilled with an appropriate sized bit. Eight pins are used, with one pin being placed on each side of each vertebral body. The pins in each vertebra are then attached to a metallic spinal fixator arch with fixation bolts and the four arches are then linked together using threaded connecting rods. This technique is relatively easy to use and has several advantages. Less soft tissue dissection is needed since the bodies of the vertebrae are not fully exposed during the surgery, thus reducing the danger of damaging vital structures and preserving surrounding muscles and their attachments that contribute to the stability of the spine.^5,8 Other advantages of this technique include: ease of implant removal; no interference if a dorsal laminectomy is necessary; provision of adequate stabilization for fracture healing to occur; and the ability to adjust the fixator in the postoperative period without surgical intervention⁸. The external fixator can be applied to both the lower thoracic and lumbar spine and is well tolerated by the patient.

Application of the technique requires a thorough knowledge of vertebral anatomy to ensure proper pin placement. Careful attention to skin placement is necessary when pins are inserted into the vertebral body to reduce tension and the possibility of postoperative pin-tract sepsis. Owners must be instructed and diligent in proper postoperative pin management.

The purpose of this case report is to describe the use of an external fixator system in three dogs with traumatic spinal instability.

Materials and Methods

Dogs:All dogswere referred to the Teaching Hospital of the Virginia-Maryland Regional College of Veterinary Medicine for suspected spinal instability following trauma. The diagnosis of spinal cord dysfunction secondary to trauma and vertebral canal instability was based primarily on findings from the neurologic examination and routine radiography. In dog 1, flexion/extension radiographs were also obtained to demonstrate dynamic instability. In dogs 1 and 2, spinal cord compression was visualized using computed tomography.^a. Reformatted CT images were generated using a remote diagnostic workstation.^b

Surgical technique: Dogs were positioned in sternal recumbency and prepared for a dorsal approach to the spine, with the incision centered over the area of instability. After surgical exposure, the fracture/luxation was held in reduction temporarily with Kirschner wires placed across the articular facets of the affected vertebral bodies. A dorsal laminectomy was performed if necessary to assure reduction and/or remove disc or bone fragments from the vertebral canal. A pilot hole was then drilled into the cranial portion of the body of the vertebrae at a 45⁰ angle to the dorsal spine. The hole passed through both vertebral cortices, using a drill bit that approximated the inner diameter of the positive profile pin to be used. In the lumbar vertebrae the accessory processes and the transverse processes were used as landmarks and in the thoracic spine the accessory processes and the rib heads were used as landmarks. A depth gauge was used to measure the depth of the pilot hole and this measure was marked on the positive profile end threaded pin with a sterile marker. A pilot hole was made in the contralateral side of the vertebral body in a similar fashion. The pilot hole on the contralateral side was made in the caudal portion of the vertebrae in order to keep the pins from interfering with each other during their placement.

Upon completing all the pilot holes, the skin and surrounding musculature were pulled to the center of the incision and a towel forceps was used to temporarily hold the incision closed. A number 15 scalpel blade was used to make a small skin incision to serve as an entry portal for the pin. A 6.5 mm drill guide was inserted into the entry portal and through the surrounding musculature to keep the threads of the pin from entangling the surrounding soft tissues in all cases. The pin was placed into the vertebral body using a low speed, high torque drill. This procedure was repeated until all eight pins were placed. Proper placement of the pins through the skin and musculature allowed closure of the skin with little tension on the incision line and soft tissues surrounding the pins.

The most cranial fixator arch was then secured to the two most cranial pins by the use of fixation bolts, followed by placement of the most caudal spinal arch to the two most caudal fixation pins. Initially, one threaded connecting rod on each side of the spinal column was used to connect the cranial and caudal spinal fixator arches. Before the cranial and caudal spinal arches were secured to the threaded connecting rod, the remaining two fixator arches were passed through the connecting rod, but not secured to the fixation pins. This allowed the incision to be closed in a routine fashion. Following skin closure the remaining two spinal arches were secured to the fixation pins using fixation bolts. One additional threaded connecting rod was placed on either side of the spinal column to complete the fixation. Postoperative radiographs were made, to assess pin placement and alignment of the vertebral canal. The sponge portions of disposable surgical scrub brushes were collected and placed around the skin/pin interface to decrease micromotion of the skin surrounding the pins. The spinal arches were then wrapped with cast padding and Vet Wrap^d. Post operatively, pin management consisted of twice daily pin cleaning with a 0.05% chlorhexidine solution and placement of surgical sponges around the pins to prevent micromotion at the pin/skin interface until granulation tissue formed around the pin. Dogs also were administered oral antibiotics for 10 days to prevent pin tract sepsis.

Results

Dog 1

A 2 yr old 27.3 kg female intact German Shepherd dog was presented to the referring veterinarian after being hit by a car. Upon presentation the dog was paraplegic with deep pain perception present in the rear limbs. Segmental spinal reflexes in the pelvic limbs were exaggerated and hyperpathia occurred on palpation of the region of the thoracolumbar (T-L) junction. Lateral spinal radiographs revealed a subluxation at T12-T13. The animal was administered 30mg/kg of methylprednisolone succinate^e (MPSS) IV and referred to our hospital.

Neurologic examination revealed paraplegia, absent cutaneous trunci reflex caudal to the T-L junction and hyperreflexia in the pelvic limbs. Anal tone was normal and crossed-extensor reflexes were not evident in the pelvic limbs. Pain perception was present in the digits of the pelvic limb, perineum and tail. These findings were interpreted as indicating a lesion somewhere in the T3-L3 spinal cord segments.

The dog was secured to a backboard and lateral and VD (horizontal beam) radiographs of the spine were obtained. These revealed a collapsed disc space at T12-T13 and slight ventral displacement of T13 relative to T12. (Figure 1a). Displacement improved with extension and worsened with flexion of the spine (Figure 1b) Computed tomography (CT) revealed a hyperdense soft tissue mass within the ventral vertebral canal presumably causing ventral compression of the spinal cord. (Figure 2a) Displacement of the T12 lamina into the spinal canal was evident also. (Figure 2b)

At surgery, a dorsal laminectomy revealed disc material and large amounts of hemorrhage within the spinal canal. Spinal fixation of T11, T12, T13 and L1 was performed, as described above. Postoperative radiographs demonstrated correct fixator placement and improved alignment of T12-13. (Figure 3a, 3b) Nine days following surgery the dog's motor function in the pelvic limbs had improved and the dog was only moderately ataxic. The dog was discharged from the hospital ten days following surgery and the owners were instructed to tighten the fixation bolts and nuts of the threaded connecting rods once weekly. Six weeks following surgery the dog's neurologic status had greatly improved, with only mild ataxia visible at a walk. The pin tracts had healed with no evidence of drainage or loosening. Spinal radiographs revealed that reduction had been maintained at T12-T13 and no evidence of pin loosening was visible. The owners reported that the dog tolerated the external fixator device for the six week period without any complications. (Figure 3c, 3d)The animal was sedated and the external fixator was removed. Postoperative radiographs demonstrated good alignment of the T12-13 vertebrae. (Figure 4) The owners were instructed to maintain the dog under cage rest for an additional 3 weeks. Six months following surgery the referring veterinarian reported that the dog appeared to be free of any neurologic deficits.

Dog 1, preoperative radiographs showing a collapsed disc space at T12-T13 (arrow) and slight ventral displacement of T13 relative to T12 (Figure 1a). Displacement improved with extension (Figure 1b, upper) and worsened with flexion of the spine (Figure 1b, lower). Lanz, et al., Figures 1a and 1b. Click on an image to see a larger view

Figure 2a. Computed tomography (CT) revealed a hyperdense soft tissue mass within the ventral vertebral canal presumably causing ventral compression of the spinal cord.	Figure 2b. Displacement of the T12 lamina into the spinal canal was evident also.

Figure 3a and 3 b. Postoperative lateral and dorsoventral radiographs showing correct fixator placement and improved alignment of T12-13.

Figure 3c, 3d. Post operative views of Dog 1 with spinal fixator in place.

Figure 4a and 4b. Six weeks postoperatively. There was good alignment of the T12-13 vertebrae. Note lucencies where pins had been placed and removed. Lanz, et al.

Dog 2

A 4 yr old 39.3 kg male intact German Shepherd dog was presented to the referring veterinarian after being hit by a car. On presentation the animal was paraparetic with exaggerated segmental spinal reflexes in the pelvic limbs. Thoracic radiographs revealed severe pulmonary contusions and subluxation of vertebra L1-L2. The dog was given one liter of Lactated Ringer's Solution^f IV, 500 mg of MPSS^e IV, 1000 mg of Cephazolin^g IV and 12 mg of butorphanol^h SQ and referred to our hospital the following day.

Upon presentation to the VMTH the animal was in severe respiratory distress and was immediately placed in an oxygen cage. It responded well to oxygen. Lateral thoracic films revealed severe pulmonary contusions along with subluxation of the L1-L2 vertebral bodies. Neurologic deficits were restricted to the pelvic limbs. Abnormalities noted included a nonambulatory paraparesis, exaggerated segmental spinal reflexes in the pelvic limbs and hyperpathia in the TL region. These signs indicated a lesion in the T3-L3 region. The dog was immobilized on a backboard and treated for pulmonary contusions until surgical stabilization of the vertebral bodies L1-L2 could be performed. Four days after initial presentation to the VMTH the dog had slight improvement in motor function in the pelvic limbs and was deemed able to withstand general anesthesia.

Pre-operative radiographs demonstrated subluxation of the L1-2 vertebral bodies, with ventral displacement of L2 vertebral body relative to L1. (Figure 5) Dorsal planar CT revealed left lateral displacement of L2. (Figure 6) Sagittal planar CT revealed L1-2 vertebral canal stenosis due to cranioventral displacement of the L2 lamina (Figure 7)

At surgery, the subluxation was repaired using two 6 hole 3.5 mm limited contact dynamic compression plates (LCDCP), one plate placed on either side of the L1 andL2 vertebral bodies. A dorsal laminectomy was performed for spinal cord decompression. The incision was closed in a routine fashion and the dog recovered uneventfully from surgery.

Postoperative radiographs revealed improved alignment of the L1-2 vertebral bodies. Six days following surgery the dog displayed improved motor function in the pelvic limbs and could walk with towel support. Ten days following surgery the dog was discharged from the hospital and the owners were instructed to continue cage rest of the dog for 6 weeks. Three weeks following surgery the owner reported that the dog had been running in the back yard and suddenly lost motor function to the back legs. Upon presentation to the VMTH a neurologic examination revealed exaggerated segmental reflexes and loss of motor function of the pelvic limbs and absence of deep pain perception in both pelvic limbs. Radiographs revealed T13-L1 subluxation and loosening of the 2^nd and 3^rd screws of the right LCDCP. (Figure 8) The owners were informed of the poor prognosis but elected to have surgical stabilization performed.

At surgery a dorsal approach was made to expose T12-L4 vertebral segments and the subluxation was temporarily held in reduction with Kirschner wires driven across the articular facets of vertebrae T13-L1. Positive profile pins were placed in the vertebral bodies of T12, T13, L3 and L4. All screws of the LCDCP were re-tightened. Four spinal fixator arches and 4 threaded connecting rods were used to stabilize the subluxation at T13-L1. Postoperative radiographs revealed adequate alignment of T13-L1 and L1-2. (Figure 9) The animal recovered uneventfully from surgery. Ten days following surgery the animal still did not have deep pain perception in the pelvic limbs and the owners elected to have the animal euthanized due to financial constraints. There were no complications associated with the spinal external fixator at the time of euthanasia.

Click on an image to see a larger view
Figure 5. Dog 2. Preoperative films. There is subluxation of the L1-2 vertebral bodies, with ventral displacement of L2 vertebral body relative to L1. Lanz et.al.	Figure 6. Dog 2. Preoperative dorsal planar CT showing left lateral displacement of L2 relative to L1. Lanz, et al.

Figure 7a and 7b. Dog 2. Sagittal planar CT showing L1-2 vertebral canal stenosis due to cranioventral displacement of the L2 lamina. Lanz, et al.

Figure 8a and 8b. Dog2. Radiographs made three weeks after repairing the original fracture luxation with a limited contact dynamic compression plate (LCDCP) at which time the dog suddenly had become paraplegic with exaggerated segmental reflexes and loss of pain perception in both pelvic limbs. There is T13-L1 subluxation and loosening of the 2nd and 3rd screws of the right LCDCP. Lanz et al.

Figure 9. Dog 2. Radiographs made after failure of the LCDCP and subsequent installation of an external fixator. There is adequate alignment of T13-L1 and L1-2. Lanz et al.

Dog 3

A 4 yr old 18.4 kg male neutered Shar-Pei was presented to the referring veterinarian after being hit by a car. At presentation the dog was ambulatory but ataxic in the pelvic limbs with exaggerated segmental spinal reflexes. Physical exam was unremarkable, except for the aforementioned neurologic deficits. Lateral radiographs of the spine revealed a subluxation of vertebrae L1-L2. No other abnormalities were found. The dog was treated with 0.2 mg/kg of butorphanol IV, 30 mg/kg of MPSS^e IV and l liter of Lactated Ringer's solution^f IV. The dog was referred to our hospital the following day for further evaluation.

At presentation, the dog had a non-ambulatory paraparesis, but had slight motor function and exaggerated segmental spinal reflexes in the pelvic limbs. Lateral and ventrodorsal radiographs of the spine revealed a fracture/subluxation of the L1-L2 vertebrae (Figure 10). Hypoplasia of the T13 ribs was noted also. Thoracic and abdominal radiographs revealed no abnormalities.

The dog then was anesthetized, positioned in sternal recumbency and a dorsal approach to vertebral bodies L1-L2 was made. Positive profile end threaded pins were placed in vertebral bodies T13, L1, L2 and L3. Four spinal arches and four threaded connecting rods were used to complete the fixator. Postoperatively, there was correct fixator placement and improved alignment of L1-L2 (Figure 11).

Two days following surgery the dog could walk with the aid of towel support. On the third day following surgery it was noted that the dog had significant swelling of his inguinal area. The dog seemed depressed and abdominal ultrasonography revealed a caudal abdominal hernia with incarcerated bladder and intestine. An exploratory celiotomy revealed a prepubic tendon rupture. The incarcerated segment of bowel was resected and an intestinal anastomosis was performed. The prepubic tendon was repaired by suturing the prepubic tendon to the pubic bone. The incision was closed in a routine fashion and the dog recovered uneventfully from surgery. Three days following his exploratory celiotomy the dog was discharged from the hospital. The referring veterinarian reported that two days following discharge the dog died of gastric dilatation and volvulus.

Click on an image to see a larger view
Figure 10a and 10b. Dog 3. Preoperative lateral and ventrodorsal radiographs showing fracture/subluxation of the L1-L2 vertebrae and hypoplasia of the T13 ribs. Lanz, et al.

Figure 11a and 11b. Dog 3. Postoperative radiographs showing correct fixator placement and improved alignment of L1-L2. Lanz, et al.

Discussion

External fixation for stabilization of vertebral instability/fracture is indicated to treat instability of the caudal lumbar spine, fracture/luxations secondary to discospondylitis and fractures not easily stabilized by internal fixation^{5, 8-10}. Internal skeletal fixators, "fixateurs internes", have been described for use in human and veterinary medicine^9-11. The disadvantages of internal skeletal fixators include premature loosening of clamps and major corrosion if the clamps are not of the same composition as the fixator pins or connecting rod. In addition this procedure requires large implant volume, postoperative manipulations of the fracture/luxation can not occur, and lastly, a second surgery is required for removal of the implants^9-12.

The use of external fixators for stabilization of spinal column fractures has been reported previously in the veterinary literature for the repair of caudal lumbar fractures in dogs^13,8. One technique uses a combination of a type II Kirschner-Ehmer device and spinous process plating and the other reported technique uses only a type II Kirschner-Ehmer device for stabilization of caudal lumbar fracture/luxations. The major postoperative complication that was encountered was pin tract sepsis, however; this problem was resolved following appropriate antibiotic therapy and pin management. The device used in this report proved simple to perform, but a possible disadvantage was that this technique required much preoperative planning.

The pin-bone interface is the most critical site of all external fixation⁵. In the cases reported here all pins were placed in the vertebral body after a pilot hole was drilled. A pilot hole approximating, but not exceeding the inner diameter of the fixation pin was created in the vertebral body. Pre-drilling a pilot hole has been shown to improve initial pin stability and reduce damage that may lead to excessive bone resorption and premature pin loosening². Loosening of the pin was not encountered in dog 1, which had the spinal external fixator in place for a total of 6 weeks.

Pin tract sepsis also has been associated with the use of external fixators. In order to decrease pin tract sepsis, placement of the pins through the skin and musculature needs to be properly performed to prevent tension from occurring around the pins. This was accomplished in these cases by careful intraoperative planning. Post-operative pin hygiene as described above produced satisfactory results in the one dog surviving long enough to be informative.

The external spinal fixation device used in the cases presented here provided sufficient stability and allowed proper realignment and stabilization of the vertebrae. Postoperative manipulation of the fixator was not required to improved alignment of the spinal column in any of the cases, but could have been performed easily. The fixator was placed in the caudal thoracic and cranial lumbar vertebrae without complications. All dogs tolerated the external fixator in the postoperative period and the unique design of the spinal arches allowed the dogs to be comfortable while in lateral recumbency. The disadvantage of the previous reports, in which a type II Kirschner-Ehmer device was used [e.g., ^{8, 13}], was that the animal is forced to lie on the connecting rods when positioned in lateral recumbency. The spinal arches currently come in two different sizes, 220 and 140 mm arches of the would-be circle. The 220mm spinal arch has 19 evenly spaced holes and the 140mm spinal arch has 15 evenly spaced holes. This configuration allows placement of fixation bolts to connect the fixation pins, allowing fixation pins to be placed in different dorsoventral planes. The fixation pins should be placed in the same craniocaudal plane in order to decrease tension and bending of the fixation pins after they have been secured to the spinal fixation arch. The holes in the arches also allow placement of the threaded connecting rods to complete the fixator. The fixator also can be used in compression fractures, by spanning the involved vertebral body. This principle was used in dog 2, in which the fixator was placed to span the luxation at vertebral bodies T13-L1 since a vertebral body plate had been used to repair the previously subluxated vertebral bodies L1-L2.

Using an unstable canine cadaver spine model, Walker, et al. (Walker TM, et al Proc. Veterinary Orthopedic Surgery 27^th Annual Conference, 2000) showed that a four or eight pin spinal arch external fixation was significantly (p<0.05) more stiff than intact spines in flexion and extension. In that same study, there was no significant difference in stiffness in the spinal arch constructs and polymethylmethacrylate (PMMA) internal fixation. The fixation device used in this report is comparable to a Type Ib (unilateral-biplanar) external skeletal fixator. The latter configuration consists of two Type I half-pin single connecting rods oriented at 90 degrees to each other and connected at their proximal and distal ends forming a bi-planar frame. The type Ib construct is more resistant to shear and bending forces than a type II configuration^14-16. In the previous veterinary reports, type II external skeletal fixators were used with or without spinous process plating. Dorsoventral bending is one of the main forces that needs to be counteracted following spinal stabilization, therefore; a type Ib would be more advantageous than a type II external skeletal fixator. The pins and connecting rods share the force in type Ib external fixators, whereas in type II the pins are almost all that is subjected to the entire load^14,15. Therefore, type Ib fixators are more applicable to fractures that are axially stable but unstable in bending.

Although only one dog had the fixator in place for six weeks, all dogs tolerated it in the post-operative period and the owners reported that dog 1 tolerated it well for the entire six weeks. Reduction and fixation of vertebral subluxation with the external spinal fixator proved to be an effective method of stabilization. Postoperative management of these cases was labor intense, which may deter its use in some patients. The fixator is available in two sizes that can accommodate both large and small breeds of dogs. Removal of the fixator after subluxation has healed can be performed using sedation alone. We conclude the spinal fixator is useful for stabilization of fracture/luxations in the region of the thoracolumbar junction.

Footnotes

^a Imex- Longview, TX
^b IQ/T, Picker International, Cleveland, Ohio
^c Voxel Q Visualization Station, Picker International, Cleveland, Ohio
^d Vet Wrap- 3M Animal Care Products, St. Paul, MN
^e MPSS - UpJohn Company, Kalamazoo, MI
^f Lactated Ringer's Solution - Baxter Health Corporation, Deerfield, IL
^g Cephazolin - Bristol-Myers Squibb Company, Princeton, NJ
^h Butorphanol - Fort Dodge Animal Health Company, Fort Dodge, IA

References

1.  Bruecker, K.A. and H.B.d. Seim, Principles of spinal fracture management. Semin Vet Med Surg (Small Anim), 1992. 7(1): p. 71-84.

2.  Clary, E.M. and S.C. Roe, In vitro biomechanical and histological assessment of pilot hole diameter for positive-profile external skeletal fixation pins in canine tibiae. Vet Surg, 1996. 25(6): p. 453-62.

3.  Garcia, J.N., et al., Biomechanical study of canine spinal fracture fixation using pins or bone screws with polymethylmethacrylate. Vet Surg, 1994. 23(5): p. 322-9.

4.  Lewis, D.D., et al., Repair of sixth lumbar vertebral fracture-luxations, using transilial pins and plastic spinous-process plates in six dogs. J Am Vet Med Assoc, 1989. 194(4): p. 538-42.

5.  Oberli, H., R. Frigg, and R. Schenk, [External fixator: surgical technique, pinless fixator, change in procedure]. Helv Chir Acta, 1994. 60(6): p. 1073-80.

6.  Richter-Turtur, M., et al., [Fractures of the spine]. Orthopaede, 1989. 18(3): p. 164-70.

7.  Shaffrey, C.I., et al., Surgical treatment of thoracolumbar fractures. Neurosurg Clin N Am, 1997. 8(4): p. 519-40.

8.  Shores, A., et al., Combined Kirschner-Ehmer device and dorsal spinal plate fixation technique for caudal lumbar vertebral fractures in dogs. J Am Vet Med Assoc, 1989. 195(3): p. 335-9.

9.  Ullman, S.L. and R.J. Boudrieau, 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. Vet Surg, 1993. 22(1): p. 11-7.

10. Bednar, D.A., Experience with the "fixateur interne": initial clinical results. J Spinal Disord, 1992. 5(1): p. 93-6.

11. Sim, E. and P.M. Stergar, The fixateur interne for stabilising fractures of the thoracolumbar and lumbar spine. Int Orthop, 1992. 16(4): p. 322-9.

12. Sim, E., [Spinal internal fixator--analysis of problems in 28 cases]. Wien Klin Wochenschr, 1992. 104(12): p. 349-55.

13. Phillips, L. and J. Blackmore, Kirschner-Ehmer Device Alone to Stabilize Caudal Lumbar Fractures in Small Dogs. Vet Comp Ortho and Traum, 1991. 4: p. 112-5.

14. Egger, E.L., et al., Type I biplanar configuration of external skeletal fixation: application technique in nine dogs and one cat. J Am Vet Med Assoc, 1985. 187(3): p. 262-7.

15. Palmer, R.H., et al., Principles of bone healing and biomechanics of external skeletal fixation. Vet Clin North Am Small Anim Pract, 1992. 22(1): p. 45-68.

16. Rudd, R.G. and J.G. Whitehair, Fractures of the radius and ulna. Vet Clin North Am Small Anim Pract, 1992. 22(1): p. 135-48.