[Anterior osteosynthesis of odontoid fractures].

PURPOSE OF THE STUDY The conservative treatment of an odontoid fracture with immobilization in a halo-vest or collar often results in pseudoarthrosis. Therefore, surgical treatment is preferred, and the Magerl-Böhler anterior osteosynthesis of the C2 dens is one of the options for achieving good bony union. The aim of this study was to show that the success of reducing an odontoid fracture is related to the direction of fracture lines and that of displacement, and that anterior osteosynthesis of the C2 dens provides sufficient stability when fixed either with one or two screws. MATERIAL Patients treated for odontoid fracture at the Department of Neurosurgery, Teaching Hospital of the Faculty of Medicine, Palacky University in Olomouc, were followed up and prospectively evaluated. From February 1994 to October 2006, 50 patients between 17 and 98 years of age (average age, 51.16 years) underwent surgery by the Magerl-Böhler method. Of them, 40 were men (average age, 46.0 years) and 10 were women (average age, 71.8 years). The minimum follow-up period was 1 year. METHODS In each patient, X-ray examination of the upper cervical spine in antero-posterior and lateral projections was done and a CT scan of the C2 vertebra was obtained. The radiographs were evaluated for location of the fracture, direction of the fracture line and direction of dens displacement. The fracture was then categorized according to the conventional classifications of Anderson- D'Alonzo (1974), Roy-Camille (1973) and White-Panjabi (1978). The patients with type II and some with type III fractures (shallow type) were indicated for surgery, regardless of fracture line direction and the direction and extent of displacement. Patients suspected of spinal cord injury were immobilized and the fracture was reduced by skull traction as soon as possible after injury. In patients without neurological deficit the fracture was reduced under general anaesthesia before surgery carried out by the Magerl-Böhler method from the anterior approach. The extent of antero-posterior displacement of the odontoid fracture was measured on lateral X-ray images at the first week after surgery and then at complete bony union. The results were statistically evaluated. RESULTS An anatomical position of the dens after reduction was achieved in 18 patients (38 %). In the remaining 32 patients, reduction was not complete but sufficient to permit screw insertion. Fractures with an anterior oblique fracture line were more difficult to reduce than fractures with a posterior oblique or a transverse fracture line. The average displacement values following reduction of the fractures were as follows: anterior displacement of 3.88 mm; posterior displacement of 1.86 mm; and anterior or posterior displacement of 1.08 mm. The differences were statistically significant. A recurrent displacement during bone healing occurred in 13 patients (26 %). It affected type A fracture in five of 17 patients (29.4 %), type B fracture in seven of 21 patients (33.3 %) and type C fracture in one patient out of 12 (8.3 %). The average extent of displacement was 1.53 mm in type A, 1.20 mm in type B, and 0.08 mm in type C fractures. The average displacement for the whole group was 1.04 mm. This implies that recurrent displacement of the dens was more frequent in fractures with an anterior oblique fracture line than in those with either posterior oblique or transverse fracture line, and this was statistically significant. Of the 43 patients with single-screw fixation, 12 (28 %) experienced recurrent displacement during healing, and of the seven patients with two screws one patient had displacement (14 %). Although the extent of displacement was higher in one-screw than in two-screw fixation (average, 1.17 mm and 0.29 mm, respectively), the difference was not statistically significant. In patients under 70 years of age, 21.6 % and, in patients over 70 years of age, 41.7 % of the fractures had recurrent displacement (average, 0.78 mm and 1.83 mm, respectively). This was not statistically significant. Bony union was achieved in all treated patients. DISCUSSION The success of reduction in displaced odontoid fractures depends on time between injury and treatment and, as reported in the literature, is more difficult after two weeks of injury. Anatomical reduction has been achieved in 65 % of displaced fractures regardless of the extent of displacement (5 mm or more). Fracture displacement after osteosynthesis of the dens with a single screw has been described in one of 17 patients (6 %), and union has been achieved in all cases. In our group recurrent displacement was found in 13 out of 50 patients (26 %). Exact data on the extent of recurrent odontoid displacement after anterior osteosynthesis is not available. It has been shown by biomechanical studies that one- or two- screw fixation provides comparable stability that, however, reaches only 50 % strength of an undamaged dens. CONCLUSION The anterior osteosynthesis of odontoid fractures was successful even in the patients in whom the fracture could not be sufficiently reduced. Union was achieved in all cases even though some fractures became displaced during bone healing and osteosynthesis therefore was not stable. Displacement and more difficult reduction were recorded more frequently in type A than in type B and C fractures.