Background: Opinions differ on the timing of surgery for rheumatoid arthritis patients with atlanto-axial subluxation. Some clinicians wait for development of neurological signs; others favour prophylactic fusion and decompression. We examined the results of surgery in relation to neurological state at the time of operation.
Methods: 134 patients underwent surgery for rheumatoid involvement of the cervical spine, after development of objective signs of myelopathy. Surgical outcomes were examined prospectively in two groups-patients who were still ambulant at the time of presentation (Ranawat class III A) and patients who had lost the ability to walk (Ranawat class III B)-by means of neurological and functional grading systems in conjunction with standard measures of postoperative morbidity and mortality.
Findings: 58% of the ambulant patients attained Ranawat neurological grades I or II compared with only 20% of the non-ambulant patients (p<0.0001). The non-ambulant group also fared worse in terms of postoperative complication rate, length of hospital stay, functional outcome, and ultimately survival. Radiologically, the non-ambulant patients were characterised by a smaller cross-sectional spinal cord area.
Interpretation: The strong likelihood of surgical complications, the poor survival, and the limited prospects for functional recovery in non-ambulant patients make a strong case for earlier surgical intervention. At a late stage of disease most patients will have irreversible cord damage.