Abstract
Introduction
Posterior instrumentation of C2 is important in treating occipitocervical, atlanto-axial, and subaxial spinal injuries. The variable C2 anatomy can make instrumentation challenging and prone to potentially severe and even life-threatening complications. The goal of this study was to (1) evaluate a large series of posterior C2 screws to determine accuracy as assessed by computed tomography (CT) scan, (2) assess dimensions of “safe bony windows” with CT, and (3) assess the perioperative complication rate related to errant screw placement. We hypothesize that the evolution of a greater variety of options for C2 fixation and the ability to tailor these techniques to patients' specific anatomy has resulted in decreased complications and increased accuracy.
Material and Methods
A retrospective review of a single Level I trauma center spine database was assessed to identify all patients with operative management for traumatic conditions with C2 instrumentation between December 2002 and September 2008. Radiographic analysis included evaluation of pre- and postoperative CT scans to quantify the patients' bony anatomy as well as to classify the accuracy of C2 screw placement. All C2 screws were graded using the following definitions: Type I: Screw threads completely within the bony cortex. Type II: Less than half the diameter of the screw violates the surrounding cortex. Type III: Clear violation of the transverse foramen or spinal canal.
Results
A total of 169 patients (99 males and 70 females) underwent posterior C2 screw fixation for traumatic conditions of the spine, including 23 atlanto-occipital dissociations, 16 unstable C1 ring injuries, 79 dens fractures, 9 Hangman fractures, 8 other C2 fractures, 3 C2 hyperextension injuries, 4 C2–C3 flexion–distraction injuries, 14 subaxial injuries, and 19 odontoid nonunions. Average CT measurements of pedicle height, axial width, and laminar width were 8.1 mm (standard deviation 2.1 mm), 5.8 mm (1.9 mm), and 5.7 mm (1.5 mm), respectively, with males having larger pedicle height (p < 0.001), pedicle width (p < 0.001), and laminar width (p < 0.022). A total of 324 screws underwent CT analysis and chart review. A total of 169 pedicle, 85 trans-articular, 50 laminar, and 20 short pars screws were placed with 80.6, 84.4, 96, and 94.4% accuracy rates (Grade I vs. II and III), respectively. Overall 85% of the screws were rated as being in the ideal location. There were no cases of known vertebral artery injury at the time of surgery. Two pedicle screws were unacceptably placed (Grade III). Both patients had screws encroaching on the vertebral artery foramen: one patient had a vertebral artery occlusion and the other had a Grade I dissection documented on CT angiogram without clinical signs of stroke or emboli as measured on transcranial Doppler. No neurological injuries resulted from instrumentation.
Conclusion
Our findings demonstrate a lower than previously reported incidence of complications associated with posterior C2 screw placement. In the management of 169 patients with 324 screws, only 2 patients had vascular injuries, neither of whom had any sequelae. The multiple techniques of posterior C2 fixation that are currently available allow for unprecedented flexibility in determining which technique is best suited for a given patient's anatomy.