Fracture of the odontoid process. An experimental and clinical study

Abstract

Fractures of the odontoid process have a bad reputation because of their tendency for non-union. The cause has been considered to be insufficeint arterial blood supply and/or inadequate fixation of the fracture. The fracture mechanism has not been elucidated. In this study, the theories on hyperflexion, hyperextension, and horizontal shear as the cause of odontoid fractures were tested by experiments on cadaver specimens. However, no odontoid fracture could be produced by violent hyperflexion, hyperextension or horizontal shear; instead inuries occurred below the axis. Violent vertical compression also failed to produce odontoid fractures; instead such an impact resulted in fractures of the atlas. But, when an impact was used that combined horizontal shear and vertical compression, odontoid fractures could be produced, and they were similar to those seen in patients. The experimental and clinical odontoid fractures were classified into four types, A, B, C, and D, depending on their level; type A passing through the isthmus of the odontoid process and the others passing progressively at lower levels. All types were observed in the clinical study and all but tye C were produced in the experimental study. When a combined type of impact (horizonal shear and vertical compression) was directed straight in the sagittal plane through the axis (anterior or posterior impact), type D odontoid fractures were produced; when such an impact was directed 45° to the sagitall plane (anterolateral impact), type H odontoid fractures were produced, and when it was directed 90° to the sagittal plane (lateral impact), type A odontoid fractues were produced. The influence of different odontoid fractures were filled with contrast medium either before or after the production of the fracture. The arterial sources to the odontoid process are paired anterior and posterior ascending arteries, paired inferior and superior anterior horizontal arteries, and paired posterior horizontal arteries. Arteries entering the odontoid process at the base anastomose inside the bone with arteries entering at the apex. In some instances when the experimental fracture involved a part of the body of the axis, injuries to the ascending arteries were observed. In all instances, thus irrespective of the level of the fractures , arteries within the odontoid process were filled with contrast medium. When the fracture was situated in the isthmus, this filling was mediated by arteries entering at the apex of the odontoid process. The medical records and radiographs of 78 patients with odontoid fracture were studied. Most fractures were caused by a high velocity force. Three patients died as a consequence of the odontoid fracture. Eighteen patients had neurologic symptoms on admission to the hospital. In most of them these symptoms were minor, but 1 had tetraplegia. Forty-eight patients were examined with radiography at a follow-up investigation. Twenty-four had bone union. One had been operated on early with posterior fusion and 4 had undergone such an operation late (because of non-union). All fusions had healed, but the state of the fracture could not be determined on the radiographs. Nineteen patients had non-union, 1 of them with a spontaneous anterior fusion between the axis and the atlas. The rate of bone union was significantly (p<0.05) increased the more the fracture involved the body of the axis, when comparing the fracture types; it was significantly (p<0.05) increased in fractures displaced anteriorly compared to those displaced posteriorly, and it was significantly (p<0.05) increased after skull traction < six weeks compared to treatment in a collar or Minerva jacket.