Identifying Factors Predictive of Atlantoaxial Fusion Failure in Pediatric Patients

Lessons Learned From a Retrospective Pediatric Craniocervical Society Study

Douglas L. Brockmeyer, MD; Walavan Sivakumar, MD; Marcus D. Mazur, MD; Christina M. Sayama, MD; Hannah E. Goldstein, MD; Sean M. Lew, MD; Todd C. Hankinson, MD; Richard C.E. Anderson, MD; Andrew Jea, MD; Philipp R. Aldana, MD; Mark Proctor, MD; Daniel Hedequist, MD; Jay K. Riva-Cambrin, MD


Spine. 2018;43(11):754-760. 

In This Article


Patient Characteristics and Surgical Details

We identified 131 patients who underwent a posterior C1-2 fusion for atlantoaxial instability (Table 1 and Table 2). Of these, 49 cases (37%) were performed at Utah, 21 (16%) at Boston, 15 (11%) at Columbia, 15 (11%) at Wisconsin, 14 (11%) at Texas, 8 (6%) at Colorado, and 9 (7%) at Jacksonville. The mean age was 10.6 ± 4.3 years (range 1.5–19 yr). Male patients comprised 53% of the study population. The most common cause of instability was os odontoideum (48%) followed by trauma (21%), Down syndrome (11%), rotatory subluxation (8%), congenital anomaly (7%), and skeletal dysplasia (2%), rheumatologic (2%), and tumor (2%).

Of the 131 operations, 120 (92%) were performed as the initial fusion operation and 11 (8%) were performed as a revision fusion operation. The method of fixation was transarticular screws (57%), Harms construct (i.e., C1 lateral mass and C2 pars screws) (34%), or a combination construct of these two methods (9%). Biological adjuncts other than bone morphogenetic protein (BMP), included demineralized bone matrix and synthetic bone graft extenders, were used in 54% of cases. BMP was used in 17% of cases.

After 82 of the 131 operations (63%), patients were instructed to wear a brace for 4 weeks or longer. Patients were placed in a halo vest or Minerva brace in 14 cases (11%). The remaining 49 patients (37%) were prescribed a brace to wear for comfort only.


Of the 131 C1-2 fusion operations, 117 (89%) resulted in a successful fusion and 14 (11%) were a surgical failure that required reoperation for revision of the arthrodesis or instrumentation. Every case with failed fusion had reoperation. Instrumentation failure occurred in 3 patients (2%) and graft failure occurred in 11 (8%).

Risk Factors for Surgical Failure

The results of the univariate analysis of risk factors associated with surgical failure after C1-2 fusion are presented in Table 3. Down syndrome (P = 0.003), unilateral fixation constructs (P = 0.04), and placement in postoperative Halo or Minerva bracing (P = 0.04) were associated with surgical failure on univariate analysis. Of the 14 patients with Down syndrome, 7 (50%) had a surgical failure, 5 of which were graft failures, and 2 were instrumentation failures. Of the 14 patients with unilateral fixation constructs, 4 (29%) had a surgical failure. Postoperative bracing in a hard collar, halo, or Minerva vest was considered by the investigators as a confounder because, in general, bracing might be more likely to be prescribed after surgery for patients with a high risk for surgical failure because of poor bone quality, insufficient screw purchase, or challenging surgical anatomy. Thus, we did not consider bracing to be an independent risk factor for surgical failure and did not include it as a variable in our multivariate analysis. (Interestingly, however, 10/14 patients [71%] who required a halo vest or Minerva and 60/68 patients [88%] who were instructed to wear a rigid collar for at least 4 weeks did go on to achieve a successful fusion.)

The multivariate logistic regression analysis (Table 4) included the significant variables identified with univariate testing (i.e., Down syndrome, unilateral fixation, previous fusion operation), as well as specific variables of interest that were identified a priori by the study investigators (i.e., sex and age). After controlling for these risk factors, only Down syndrome had a significant association with surgical failure (odds ratio 14.6, 95% confidence interval [3.7–64.0]).

Complications occurred in two patients, one with a superficial wound infection and one with a vertebral artery injury during transarticular screw placement. The wound infection was managed with local wound care and oral antibiotics and the patient had successful fusion. The vertebral artery injury was in a patient with Down syndrome; it was investigated with cerebral angiography, and vertebral artery occlusion was found. There was no neurological compromise, and the patient was managed with aspirin for 3 months. This patient had atlantoaxial fusion failure but ultimately underwent successful revision surgery.