Session V Spine

A Changing Perspective of Odontoid Fractures: Implications for Evaluation and Management

William M. lannacone, MD, PhD, John Naranja, MD, William G. DeLong, Jr., MD, Christopher T. Born, MD, Robert M. Dalsey, MD, Lawrence S. Deutsch, MD, John B. Catalano, MD

Cooper Hospital/UMC, Camden, NJ

Introduction: Odontoid fractures have historically been classified according to the scheme proposed by Anderson and D'Alonzo in 1974: Type I fractures involve the tip of the dens; Type II fractures occur at the base, but do not extend into the body; and Type III fractures extend into the vertebral body. Management and subsequent prognosis has been largely based on this classification system. Most reports in the literature, however, apply this classification system based on the context of plain radiographs. The use of CT scans and tomograms has essentially been ignored in the acute evaluation of these injuries. Accordingly, the purpose of this study was to classify and evaluate traumatic fractures of the odontoid not only with plain radiographs but also in conjunction with the CT scan and/or tomograms.

Methods: Fifty-seven consecutive odontoid fractures occurring between April 1987 and November 1995 at a major Level I Trauma Center were retrospectively evaluated. An analysis of plain radiographs, CT scans and /or tomograms of acute odontoid fractures was completed, along with an evaluation of demographics, mechanism of injury, associated injuries, neurologic sequellae, mortality, treatment outcomes and rate of nonunion.

Results: By classifying the odontoid fractures, using CT scans and tomograms as well as plain radiography, we found a distinctly different incidence of the different subtypes of odontoid fractures. Our series had no Type I fractures, 16 (28%) Type II fractures, and 41 (72%) Type III fractures. Several fractures, initially felt to be Type II with a plain radiograph, were re-classified as Type III injuries after review of the CT scan. The primary mechanism of injury was a motor vehicle accident (73%) and multiple injuries were present in 51% of the patients evaluated. Neurologic sequellae were present in 7% of the patients and 11 died (19%). The method of treatment used was halo immobilization in 44 cases (77%), Minerva jacket in 3 (5%), Philadelphia collar in 3 (5%), and a SOMI brace in 3 (5%). Four patients underwent posterior spinal fusion (7%). The distribution of treatment outcomes, excluding those who expired, was as follows: 42 (91%) healed within 13 weeks; there was 1 (2%) asymptomatic nonunion; and 3 (7%) were lost to follow-up. We were unable to find a significant difference in age between those who healed (average age -45 ) and those with a nonunion (average age - 75 ). There was, however, a significant difference between those who expired (average age - 69) versus those who healed (p<0.05).

Discussion and Conclusion: Our data indicates that, with the use of CT scans and tomograms, a much higher incidence of Type III odontoid fractures were identified. This implies a much better prognosis for patients with odontoid fractures. As demonstrated with this series, the use of halo immobilization should result in high union rates for those who survive this injury. We were unable to correlate the traditional negative prognostic determinators of healing, such as age, probably because of the limited number of non-unions in this series of patients.