Long-term follow-up of the cervical spine with conventional radiographs in patients with rheumatoid arthritis

Objectives: To investigate the prevalence of cervical spine damage due to rheumatoid arthritis (RA) in the long term and to investigate which disease-specific factors are related to this damage. Method: Patients with early RA from the Nijmegen inception cohort with 6 to 12 years of follow-up were included. Conventional radiographs of the cervical spine were obtained at baseline, 3, 6, 9, and 12 years and scored for erosions of C1 and C2, anterior atlantoaxial subluxation (AAS) and atlantoaxial impaction (AAI). Disease-specific factors, such as disease activity, functionality, and peripheral joint damage, at baseline, 3, 6, and 9 years, were compared between patients with and without cervical spine damage at 9 years. Results: A total of 196 patients were included, of whom 134 had radiographs at 9 years. Cervical spine damage was present in 16% (22/134) of the patients at 9 years. During the total 12 years of follow-up, AAS and erosions of C2 were observed most frequently. Erosions of C1 and AAI were very rare. Patients with cervical spine damage at 9 years had a higher number of erosions of the peripheral joints and failed more disease-modifying anti-rheumatic drugs (DMARDs) at 3, 6, and 9 years. Patients without peripheral erosive disease at 3 years were unlikely to develop cervical spine damage within 9 years of disease duration. Conclusions: The prevalence of cervical spine damage due to RA was 16% at 9 years. Patients without peripheral erosive disease at 3 years were unlikely to develop cervical spine damage at 9 years.