Many different surgical techniques have been described for the treatment of AAS. Dorsal and ventral approaches have been advocated, with neither having been shown to be superior to the other with similar complication and success rates reported [1]. When comparing risk factors affecting outcome, similar success rates for dorsal (88.9%) and ventral (85.3%) procedures were noted; however, dogs were also noted to have a higher incidence of postoperative neurologic deficits with dorsal procedures than with ventral procedures. Acute onset of clinical signs is a known positive predictor of a success [1, 2]. In some studies, age of onset of clinical signs was predictive of outcome, whereas in others it was not [1–3]. Severity of neurologic deficits at presentation may affect outcome [1, 3]; however, most studies show even those with the most severe neurologic dysfunction can recover well [2, 4–6]. Postoperative AA reduction and radiographic positioning of the dens have not been shown to correlate with outcome [1]. This supports the suggestion that the shearing motion of the instability is the critical insult to the spinal cord, not the static position of the vertebrae (see biomechanics below). The postoperative fatality rate is 5% with ventral procedures and 8% with dorsal procedures [7].

Dorsal techniques rely on fibrosis to stabilize the AA joint, as there is no access to the joint space with these techniques. With ventral techniques, the joint space can be arthrodesed, allowing for long‐term fusion of the joint to combat the instability.

Of the ventral techniques described, screws, with or without wire, and polymethylmethacrylate (PMMA) is likely the most common technique utilized. The use of PMMA has been shown to provide a strong, solid fixation. However, there are several disadvantages, including thermal damage, increased risk of infection, pressure necrosis of adjacent structures, and inconvenience when revision surgery is required [1, 8–12]. In addition, screw placement with these techniques requires a very shallow angle of placement, which can be difficult to navigate in such small anatomic spaces.

In several studies, locking plates have been promoted over compression plates due to a loss of stabilization from screw loosening and pullout from the vertebra with compression plates [13, 14]. Additionally, monocortical screw pullout was shown in humans to be similar to bicortical screws, supporting the use of locking plates with monocortical screws [15]. Biomechanical tests have shown locking plates to be more stable than nonlocking plates in human studies [15, 16]. As early as 1995, the biomechanical advantages of locking plates were shown in experimental models comparing locking and nonlocking implants in the human cervical spine, particularly in flexion [16]. In the canine cervical spine, both biomechanical studies and clinical applications of locking plate fixation have been published for treatment of cervical spondylomyelopathy and vertebral fractures [13, 17–19]. The soft, small vertebrae in these dogs can predispose to vertebral and spinal cord damage [10]. The only evaluation of locking plate treatment of AAS published to date was in three dogs in 2011 [20], all of which recovered to near‐normal neurologic function and were pain free. Two of the three dogs were grade 4 tetraparetic preoperatively, with the third being grade 2 tetraparetic. Application of the locking plate as described later in this chapter resulted in adequate arthrodesis of the C1–C2 space and improved neurologic function within two weeks of surgery in all three dogs, along with high owner satisfaction.