Outcomes and Complications Associated With Ventral Screws, Pins, and Polymethyl Methacrylate for Atlantoaxial Instability in 12 Dogs
Clinical outcomes and complications of a technique used for atlantoaxial stabilization were evaluated in a group of 12 dogs. At surgery, the atlantoaxial joint was realigned and rigidly fixated using cortical bone screws, K-wire, and polymethyl methacrylate. Results in nine dogs were graded as excellent. Results in two dogs were judged as good. One dog was euthanized 17 months after surgery for recurrent cervical pain. Eight dogs had no postoperative complications. The surgical technique described provided an adaptable method for the correction of atlantoaxial instability.
Introduction
Atlantoaxial instability is a neurological condition of the cranial cervical spinal cord that primarily affects young, small-breed dogs.1 Atlantoaxial instability usually develops from abnormalities of the odontoid process (dens) or its supporting ligamentous structures subsequent to a congenital malformation or agenesis of the dens.1–5 Trauma to the cranial cervical spinal cord may also cause atlantoaxial instability from fracture of the dens or rupture of its supporting ligaments. Regardless of the pathogenesis, the instability often leads to compression of the cervical spinal cord and various degrees of pain and neurological dysfunction.
A tentative diagnosis of atlantoaxial instability is often made with survey radiographs. Radiographic findings suggestive of atlantoaxial instability include absence, agenesis, or dorsal deviation of the dens; increased distance between the dorsal arch of the atlas and the dorsal spinous process of the axis; and dorsal deviation of the vertebral body of the axis.12 Ventroflexion of the neck during radiographic evaluation may make the instability and subluxation more apparent radiographically, but it should be performed with caution to avoid iatrogenic spinal cord injury. The use of advanced imaging techniques, such as magnetic resonance imaging (MRI), allows for assessment of the anatomy of the spinal cord proper and is helpful in identifying additional anatomical spinal cord defects (e.g., syringomyelia) in animals with atlantoaxial instability. Concurrent spinal cord abnormalities may contribute to the prognosis in dogs with atlantoaxial instability.
Nonsurgical treatment of atlantoaxial instability, including strict cage confinement and a rigid cervical brace, has been successful in some cases of atlantoaxial instability; however, nonsurgical treatment generally results in recurring or progressive clinical signs.34 The goal of surgical treatment is to stabilize the unstable articulation and subsequently prevent further spinal cord damage. Historically, surgical treatment of atlantoaxial instability has been associated with complication rates as high as 71% for dorsal fixation techniques and 53% for ventral fixation techniques.25 Dorsal surgical stabilizing techniques have included the use of wire, wire and polymethyl methacrylate (PMM), braided suture material, a portion of the nuchal ligament, a metallic retractor, and pins.146–11 Fixation methods using a ventral approach have included bone plating, pins alone, pins with PMM, and screws placed in lag fashion.12–14
Success of the surgical technique used for treatment of atlantoaxial instability is ultimately dependent upon the ability to achieve a rigid and persistent fixation of the first to second cervical (C1-C2) joint. It has been the authors’ experience that in many affected dogs, the ventral vertebral body of C2 is anatomically abnormal and the bones are relatively soft, making purchase of any apparatus difficult. Most difficulties with surgical treatment of this disease arise when the appropriate angulation necessary for direct transarticular pin or screw placement is not achieved. Numerous anatomical abnormalities of this area often require the surgical fixation technique to be adapted to the individual animal. For those animals that have poor bone strength or anatomical abnormalities that preclude conventional fixation techniques, adaptation of a ventral fixation technique using a combination of cortical bone screws, pins, wire, and PMM may be necessary. The purpose of this report is to present the clinical outcomes and complications associated with the use of a surgical technique employing these methods in 12 dogs with atlantoaxial instability.
Materials and Methods
Twelve dogs with suspected atlantoaxial instability were evaluated at Washington State University Veterinary Teaching Hospital from October 1996 to September 2000. All dogs underwent routine physical and neurological examinations and were tentatively diagnosed with either atlantoaxial instability or a C1 to the fifth cervical (C5) cord lesion. All dogs had a complete blood count, serum biochemical profile, survey radiographs of the cervical spine, and MRI of the cervical vertebral column and spinal cord. In certain cases, cerebrospinal fluid analysis (n=2) or computed tomography (CT) scans (n=2) of the cervical spine were also performed. All imaging procedures were performed under general anesthesia. Radiographs were taken using lateral [Figure 1], dorsoventral, and open-mouth positioning techniques. Computed tomographic imaging was performed in a ventral recumbent position, and MRI was performed in a dorsal recumbent position.
Surgical Technique
Only one dog (case no. 3) had surgery immediately following imaging. In all other cases, surgery was performed on a subsequent day. Dogs were induced with either thiobarbiturate or propofol and then maintained on isoflurane in oxygen anesthesia. All dogs were positioned in dorsal recumbency with towels placed between the surgical table and the neck to elevate the cervical spine and provide for a more stable operating platform. Thoracic limbs were extended caudally. Methylprednisolone sodium succinatea (30 mg/kg intravenously [IV]) was administered at the beginning of surgery and at 2 and 6 hours later at 15 mg/kg IV. Cefazolinb (22 mg/kg IV) was administered every 1.5 hours during surgery, beginning at the initial incision.
A ventral midline incision and a routine approach were made to the ventral aspects of C1 to the fourth cervical (C4) vertebra.15 The skin was incised from the cranial aspect of the larynx to approximately the C5. The sternohyoid muscle was divided along its fibrous median raphe using blunt dissection. Dissection was continued along the midline to the trachea. The sternothyroid muscle was incised 2 cm from its attachment to the thyroid cartilage and was repaired during closure of the incision. Muscles were retracted with self-retaining retractors (i.e., Gosset or Nelsonc) padded with moistened laparotomy sponges. The trachea, esophagus, carotid sheath, sympathetic trunk, and recurrent laryngeal nerve were retracted toward the left side to expose the longus colli muscles. The longus colli muscles and the rectus capitis ventralis muscles were elevated from the ventral aspect of C1 to the third cervical (C3) vertebra using blunt dissection. After dissection of the longus colli muscles, the Gosset retractors were removed, and Gelpid retractors were placed over the ventral vertebral bodies to retract the longus colli muscles from the vertebrae.
Once the ventral aspects of C1 to C3 were free of muscle, the atlantoaxial joint was incised. A Housee or similar curette was used to remove the articular cartilage from the articular surfaces of C1 and C2. The House curette was then used as a lever against the cranial aspect of the articular surface of the axis. With the atlas acting as a fulcrum, the axis was carefully repositioned to a normal anatomical alignment with the atlas. While the axis was held in position with the House curette, K-wires or cortical bone screws were placed across the atlantoaxial joint in a craniolateral direction, 35° to 40° from midline, toward the alar notch of the atlas. The placement of transarticular wires or screws was not always possible because of a high incidence of malformation of the cranial articular surface of the axis. Only six of the 12 dogs reported here (case nos. 1, 2, 5, 7, 11, 12) had transarticular implants.
Cortical bone screws were inserted into the medial aspect of each wing of the atlas caudal to the transverse foramen in a craniolateral direction. For placement of these screws, holes were predrilled using an appropriately sized drill bit prior to screw insertion. The proximal cortex was tapped with the corresponding, appropriately sized tap. The size of the screws chosen was based on the size of the bones. Screws ranged from 1.5 to 2.0 mm in diameter and 10 to 20 mm in length. Screws were then placed into the middle of the caudal aspect of the cranial articular surfaces of C2 at the insertion point of the longus colli muscle on each side of midline and were directed craniolaterally at 30° to 40°. The final pair of screws was placed at the base of the transverse processes of C2 or C3 and directed laterally at 30° to 40° from midline [Figure 2].
In all dogs, except for case no. 1, Steinman pins or K-wires were cut to bridge the distance between the screws. Pins or wires were placed parallel to the midline from the most caudal screw head to the most cranial screw head. A Jacobs hand chuck was placed on the end of each pin and used for leverage to bend the pins so they would fit snuggly around the exposed screw heads [Figure 3]. K-wires were bent with wire twisters. The pins or wires were secured to the protruding screw heads with cerclage wire. For dogs <5 kg, 22-gauge wire was used. For dogs >5 kg, 20-gauge wire was used. Wires were twisted under tension to bring about apposition of the screw heads and pins. Wires were left upright and cut three to four twists from the pin [Figure 3].
The entire apparatus was encased in PMM.f Cefazolin (500 mg) was added to every 20 grams of polymer prior to the addition of the monomer (10 mL). The amount of PMM used was dependent on the size of the dog. Typically, a half-dose of PMM (20 grams; approximately 4 cubic cm) was sufficient. The PMM was placed on the ventral aspect of the vertebral body, taking care to enclose the entire apparatus, including the ends of the pins [Figure 4]. The surgical field was irrigated with saline until the polymer was cured (approximately 5 minutes), and suction was used to remove all debris. Care was taken to ensure that the Gelpi retractors were not encased within the PMM. The Gelpi retractors were removed, and the trachea, esophagus, and carotid sheath were returned to their normal anatomical position. The sternothyroid muscle was sutured at the previous myotomy site with absorbable suture material. The sternohyoid muscle was sutured along its fibrous median raphe using a simple continuous pattern and absorbable suture material. The subcutaneous tissue and skin were closed routinely.
In all cases, lateral and dorsoventral postoperative radiographs were taken to assess implant position and the degree of reduction of the atlantoaxial joint [Figure 5]. Morphineg was administered (0.3 mg/kg intramuscularly [IM] or subcutaneously [SC], q 4 hours as needed) for pain control during the immediate 12 to 24 hours after surgery. A fentanyl patchh (2 to 4 μg/kg transdermal) was applied immediately after surgery for more sustained pain relief.
Grading of Outcomes
Follow-up questioning regarding clinical progression, complications, survival length, and, if appropriate, cause of death was conducted by telephone interview with either the referring veterinarian or the client. Animals were subjectively graded based on long-term clinical outcome, similar to methods previously used.14 Those animals with no postoperative complications, cervical pain, or neurological deficits were graded as having excellent results. Animals with manageable complications, mild pain, or ataxia were graded as a good result, and those animals with moderate ataxia, neurological deficits, and/or chronic pain were graded as a poor result.
Results
Breeds represented in the study included the Yorkshire terrier (n=5), miniature poodle (n=3), shih tzu (n=2), Pomeranian (n=1), and Japanese chin (n=1) [Table 1]. The mean age of the dogs was 3 years (range, 6 months to 8 years). There were eight females and four males. Presenting clinical signs included cervical pain (n=10), monoparesis (n=1), hemiparesis (n=1), paraparesis (n=3), tetraparesis (n=4), ataxia (n=3), head tilt (n=2), and abnormal turning of the head (n=2). Duration of clinical signs ranged from 2 days to 2 years. Five dogs (case nos. 1–5) had agenesis of the odontoid process with dorsal subluxation of C2 relative to C1. Four dogs (case nos. 6, 7, 10, 12) had dysgenesis or a rudimentary odontoid process with dorsal subluxation of C2. Three dogs (case nos. 8, 9, 11) had a normal odontoid process with dorsal subluxation of C2 relative to C1. Other than spinal cord compression associated with the instability, no additional spinal cord abnormalities, such as syringomyelia, were found in these dogs with MRI.
One dog (case no. 10) had been operated on previously (at a different hospital) for suspected instability with insertion of ventrally placed pins at the C2-C3 vertebrae. This dog was presented for ataxia and coughing. The dog was diagnosed with laryngeal paralysis at the time of thiobarbiturate induction. Because metallic implants were used in the previous surgery, a CT scan was performed (rather than an MRI) of the dog’s cervical region, and one of the pins used in the previous surgery was found to be within the spinal canal. A laryngeal tieback procedure was performed, followed by surgical removal of the previously placed implants and PMM. The dog returned 4 months later for stabilization of the C1-C2 area.
In all dogs, the articular cartilage was removed with a House curette, and the atlantoaxial joint was realigned manually as previously described. In all dogs except one (case no. 1), screws, pins, and wire were used with PMM to stabilize the atlantoaxial joint. The pins were placed parallel to the ventral vertebral bodies and were wired to the screws. In two of the dogs (case nos. 8, 9), a single pin was used and was bent with a 180° turn at the cranial end so both arms of the pin were parallel to the vertebral bodies. In one dog (case no. 12), a single K-wire was bent into a rectangle and secured to the protruding screw heads. Eight of the dogs had screws and pins bridging C1-C3. Three dogs (case nos. 4, 5, 11) had screws and pins in C1-C2 only. In one dog (case no. 1), only cortical bone screws and PMM were used to stabilize the atlantoaxial joint. The C1-C2 joint space was bridged in six of the dogs, using cortical bone screws in four of the dogs (case nos. 1, 2, 7, 12) and pins in two dogs (case nos. 5, 11).
Eight dogs had no complications [Table 2]. Four dogs (case nos. 2, 5, 6, 11) had complications following surgery. Thirteen days following surgery, case no. 2 presented for mild dyspnea. Upon survey radiographs of the cervical spine, the implanted apparatus was unchanged. With visual inspection of laryngeal function under thiobarbiturate sedation, the dog had decreased movement of the arytenoid cartilage upon inspiration. Because the dog still retained some arytenoid function, no treatment was instituted, and the dog eventually recovered. Based on follow-up questioning (41 months postoperatively), the owners judged the dog’s condition to be excellent; however, coughing still occurred when the dog became excited. Case nos. 5 and 6 became dyspneic during recovery and required temporary tracheostomies. Case no. 5 had a tracheostomy tube in place for 2 days and required no further treatment. In addition, this dog developed a mild right Horner’s syndrome following surgery that resolved within 2 days. Case no. 6 had a tracheostomy tube in place for 4 days and required no further treatment. Case no. 11 presented 6 months following surgery for migration of a transarticular K-wire. The dog was clinically normal; however, the owners could feel the tip of the wire under the skin. On lateral survey radiographs, it was evident that one of the K-wires used to cross the C1-C2 joint space had migrated cranially and dorsally. This wire was not bent at the end to secure it well in the PMM. The wire was removed under anesthesia. The dog was clinically normal at follow-up questioning.
The mean postoperative follow-up time for the dogs in this study was 20 months (range, 2 to 41 months). All owners thought the condition of their animals improved following surgery. Seven dogs were graded as excellent based upon follow-up questioning. Two dogs (case nos. 2, 6) were graded as good owing to occasional upper respiratory stridor when they became excited. Three dogs (case nos. 3, 5, 10) were graded as excellent until their death. Clinical signs of cervical pain returned 17 months after surgery in case no. 3. Cervical radiographs showed intact implants, and a myelogram did not reveal any spinal cord compression. The dog was euthanized after unsuccessful attempts to control the pain. Case no. 5 died 10 months after surgery while under anesthesia for a routine dental cleaning, and case no. 10 died 20 months following surgery from choking on a piece of carrot. Necropsies were not performed on any of these three dogs.
Discussion
Numerous techniques have been employed for the surgical stabilization of atlantoaxial instability. A report comparing various surgical techniques and complication rates documented a 48% failure rate with dorsal techniques compared to a 38% failure rate with ventral techniques.2 However, if the results of procedures using ventral pins and PMM were also included in the data, and the same criteria were applied to define success or failure, then the failure rate of all ventral procedures would be 44%.14 Risk factors, such as age of onset and duration of clinical signs, and degree of neurological impairment may also influence outcomes following surgery.16
Because the study reported here contained no control dogs, no direct comparisons can be made between complications with this surgical technique and others employed for atlantoaxial instability stabilization. However, subjectively the overall type and severity of complications in this group of dogs seemed comparable, if not somewhat less, than complications described in historical data from previously published case studies.1–1416
Fixation failure has been reported to be as high as 25% with dorsal fixation techniques and 18% with ventral fixation techniques.2 One dog (case no. 10) in this study had an implant failure, which was found incidentally on a routine follow-up radiograph. The failure consisted of fracture of the two most proximal screw heads. Although the implant had shifted cranially, the dog did not have clinical signs and was otherwise normal. A necropsy was not performed in this dog or any of the dogs in this study to confirm fusion of the atlantoaxial joint. The overall incidence of implant failure in this group of dogs was not determined radiographically.
Good results have been reported with the application of cortical bone screws across the C1-C2 joint space from a ventral approach.4 Additionally, ventrally placed pins have been used with PMM, also with good results.14 In the dogs of this study, pins or wires were used, in addition to cortical bone screws and PMM, to provide a rigid, reinforced framework. Because many dogs with atlantoaxial instability have malformed C1 and C2 vertebrae, this construct provides a stabilizing implant that may be specifically modified for each surgical application.
Conclusion
Atlantoaxial subluxation may be associated with a variety of anatomical deformities of the C1 and C2 vertebral bodies, making surgical stabilization challenging. The surgical technique described in this report allowed for reconstruction with realignment of this region. Overall, surgical mortality and morbidity was low using this technique.
Solu-Medrol; Pharmacia & Upjohn Corp., Kalamazoo, MI
Cefazolin for injection, USP; Bristol-Myers Squibb, Princeton, NJ
Sklar, West Chester, PA
Sklar, West Chester, PA
Sklar, West Chester, PA
Polymethylmethacrylate, Surgical Simplex P; Howmedica, Inc., Rutherford, NJ
Morphine sulfate injection, 15 mg/mL, USP; Elkins-Sinn, Cherry Hill, NJ
Duragesic; Janssen Pharmaceutica, Titusville, NJ
Citation: Journal of the American Animal Hospital Association 40, 3; 10.5326/0400204
Citation: Journal of the American Animal Hospital Association 40, 3; 10.5326/0400204
Citation: Journal of the American Animal Hospital Association 40, 3; 10.5326/0400204
Citation: Journal of the American Animal Hospital Association 40, 3; 10.5326/0400204
Citation: Journal of the American Animal Hospital Association 40, 3; 10.5326/0400204
Lateral survey radiograph of a dog with atlantoaxial instability/subluxation. There is increased distance between the dorsal arch of the first cervical (C1) vertebra and the dorsal spinous process of the second cervical (C2) vertebra (arrow).
Ventral aspect of a skeletal model of the atlantoaxial joint showing the approximate placement of cortical bone screws.
Ventral aspect of a skeletal model of the atlantoaxial joint showing the attachment of bent K-wires to the protruding screw heads using cerclage wire.
Ventral aspect of a skeletal model of the atlantoaxial joint with polymethyl methacrylate molded around the implanted apparatus.
Postoperative, lateral survey radiograph following application of ventral pins, screws, and polymethyl methacrylate in the same dog as in Figure 1. Because of anatomical abnormalities of the cranial aspect of C2, two sets of screws were inserted into the atlas. There is good alignment of the atlantoaxial joint and reduction of the subluxation.
Contributor Notes
