Stage-related surgery for cervical spine instability in rheumatoid arthritis.

Thirty-six consecutive patients with cervical spine instability due to rheumatoid arthritis (RA) were treated surgically according to a stage-related therapeutic concept. The aim of this study was to investigate the clinical results of these procedures. The initial change in RA of the cervical spine is atlanto-axial instability (AAI) due to incompetence of the cranio-cervical junction ligaments, followed by development of a peridontoid mass of granulation tissue. This results in inflammatory involvement of, and excessive dynamic forces on, the lateral masses of C1 and C2, leading to irreducible atlanto-axial kyphosis (AAK). Finally, cranial settling (CS) accompanied by subaxial subluxation (SAS) occurs. According to these three separate pathological and radiological lesions, the patients were divided into three therapeutic groups. Group I comprised 14 patients with isolated anterior AAI, who were treated by posterior wire fusion. Group II comprised 15 patients with irreducible AAK, who were treated by transoral odontoid resection. The fixation was done using anterior plating according to Harms in combination with posterior wire fusion according to Brooks. Group III comprised seven patients with CS and additional SAS, who were treated with occipito-cervical fusion. Pre- and postoperatively, evaluation was performed using the parameters pain (visual analog scale), range of motion (ROM), subjective improvement and Health Assessment Questionnaire (HAQ). The neurologic deficit was defined according to the classification proposed by Ranawat. Radiographs including lateral flexion and extension views, and MRI scans were obtained. The average clinical and radiographic follow-up of all patients was 50.7 +/- 19.3 months (range 21-96 months). No perioperative fatality occurred. Postoperative pain was significantly relieved in all groups (P < 0.001). In group II a slight improvement in the HAQ was obtained. In groups I and II the ROM of all patients increased significantly (average gain of motion in group I: 11.3 degrees +/- 7. 8 degrees for rotation; 7.8 degrees +/- 5.6 degrees for bending; average gain of motion in group II: 21.5 degrees +/- 14.0 degrees for rotation; 17.2 degrees +/- 5.5 degrees for bending), while it decreased significantly in group III (10.7 degrees +/- 18.1 degrees for rotation; 6.7 degrees +/- 18.5 degrees for bending). Preoperatively 27 patients had a manifest neurologic deficit. At follow-up four patients remained unchanged, all others improved by at least one Ranawat class. All patients, except one, showed solid bony fusion. According to the significantly improved postoperative subjective self-assessment and the clinical and radiological parameters, transoral plate fixation combined with posterior wire fixation after transoral odontoid resection represents an effective reliable and safe procedure for the treatment of irreducible AAK in rheumatoid arthritis.