Neurologic outcome of surgical and conservative treatment of rheumatoid cervical spine subluxation: a systematic review.

OBJECTIVE Rheumatoid arthritis commonly involves the upper cervical spine and can cause significant neurologic morbidity and mortality. However, there is no consensus on the optimal timing for surgical intervention: whether surgery should be performed prophylactically or once neurologic deficits have become apparent. METHODS A systematic review of the literature was performed to analyze neurologic outcome (Ranawat) and survival time (Kaplan-Meier) after surgical or conservative treatment using the MOOSE (Meta-analysis Of Observational Studies in Epidemiology) and GRADE (Grading of Recommendations, Assessment, Development and Evaluation system) criteria. RESULTS Twenty-five observational studies were selected. No randomized controlled trials (RCTs) could be found. All of the studies had a high risk of bias. Twenty-three studies reported the neurologic outcome after surgery for 752 patients. Neurologic deterioration rarely occurred in Ranawat I and II patients. Ranawat III patients did not fully recover. The 10-year survival rates were 77%, 63%, 47%, and 30% for Ranawat I, II, IIIA, and IIIB, respectively. The Ranawat IIIB patients had a significantly worse outcome. Another 185 patients treated conservatively were described in 7 studies. Neurologic deterioration rarely occurred in Ranawat I patients, but was almost inevitable in Ranawat II, IIIA, and IIIB patients. The Kaplan-Meier analysis showed a 10-year overall survival rate of 40%. CONCLUSION There are no RCTs that compared surgery with conservative treatment. In observational studies, surgical neurologic outcomes were better than conservative treatment in all patients with cervical spine involvement, and in asymptomatic patients with no neurologic impairment (Ranawat I) the outcomes were similar; however, the evidence is weak. Survival time of surgical and conservative treatment could not be compared.