Atlantoaxial instability in acute odontoid fractures is associated with nonunion and mortality

BACKGROUND CONTEXT: Odontoid fractures are the most common geriatric cervical spine fractures. Nonunion rates have been reported to be up to 40% and mortality up to 35%, and poor functional outcomes are common. Atlantoaxial instability (AAI) is a plausible prognostic factor, but its role has not been previously examined. PURPOSE: To determine the effect of severe AAI on the outcomes of nonunion and mortality in patients with acute odontoid fractures. STUDY DESIGN: Retrospective cohort/single institution. PATIENT SAMPLE: One hundred twenty-four consecutive patients with acute odontoid fractures. OUTCOME MEASURES: Rates of nonunion and mortality. METHODS: Two independent blinded reviewers measured AAI using postinjury computed tomography scans. Patients were classified as having ''severe'' or ''minimal'' AAI on the basis of greater versus less than or equal to 50% mean subluxation across each C1-C2 facet joint. Rates of nonunion and mortality were compared using independent samples t tests and adjusted for age, displacement, and subtype using binary logistic regression. RESULTS: One hundred seven patients had minimal AAI and 17 had severe AAI. Mean follow-up was 4.4 months (standard deviation=4.6). Patients with severe AAI were more likely to experience nonunion (29% vs. 10%, respectively; p=.03) and mortality (35% vs. 14%, respectively; p=.03) regardless of treatment modality. Fracture displacement correlated with AAI (r(2)=0.65). When adjusted for patient age, the odds ratio of nonunion with severe AAI approached significance at 3.3 (95% confidence interval [CI]: 0.9-11.7). Mortality prediction with AAI approached a twofold increased risk (odds ratio=2.1; 95% CI: 0.6-6.8). In patients with Type-II fractures, the odds of mortality with severe AAI approached a threefold higher risk (odds ratio=3.3; 95% CI: 0.9-12.3). CONCLUSIONS: Patients with acute odontoid fractures and severe AAI may be more likely to experience nonunion and mortality, suggesting the possibility that aggressive management could be warranted. Further investigation with a large prospective study including patient-important functional outcomes is justified. (C) 2015 Elsevier Inc. All rights reserved.