Prevalence and associated factors of anterior atlantoaxial luxation in a nation-wide sample of rheumatoid arthritis patients

A. Naranjo1, L. Carmona2, D. Gavrila3, A. Balsa4, M.A. Belmonte5, X. Tena6, C. Rodríguez-Lozano1, R. Sanmartí7, I. González-Álvaro8, and the EMECAR Study Group9*

1Rheumatology Department, Hospital Dr. Negrín, Las Palmas; 2Rheumatology Department, Hospital Clínico San Carlos, Madrid; 3Epidemiology Unit of the Spanish Society of Rheumatology, Madrid; 4Rheumatology Department, Hospital La Paz, Madrid; 5Rheumatology Department, Hospital General de Castellón, Castellón; 6Rheumatology Department, Hospital Germans Trias i Pujol, Badalona; 7Rheumatology Department, Hospital Clinic i Provincial, Barcelona; 8Rheumatology Department, Hospital de la Princesa, Madrid; 9Sociedad Española de Reumatología, EMECAR Study Group

ABSTRACT
Objective
To estimate the prevalence of anterior atlantoaxial subluxation (AAS) in patients with rheumatoid arthritis (RA), and to analyse its association with disease markers.

Methods
Cross-sectional analysis of a cohort of RA patients randomly selected from the clinical registries of 34 centres. AAS, defined as an atlantoaxial displacement in cervical spine X-rays greater than 3 mm on flexion films, was actively searched for. Bivariate and multivariate analysis was performed to examine its association with clinical, functional, and treatment variables. 

Results
AAS was found in 88 out of 736 patients with available cervical radiographs, (prevalence and 95% confidence interval [CI]: 12% [9.7-14.2]). The presence of AAS was highly associated with a Larsen score (0-150) over 50 (OR and 95% CI: 5.31 [2.68-10.55]), RA duration of more than 10 years (4.48 [2.70-7.44]), disease onset before age 50 (4.15 [2.42-7.12]), eye involvement (3.93 [1.63-9.46]), and previous RA related surgery (3.90 [2.46-6.19]). No association was found with rheumatoid factor. Multivariate analysis showed that a disease onset before the age of 50, the number of previous DMARD, and, above all, a Larsen score greater than 50 were important independent factors associated with AAS. There is a 33% increased risk for AAS every 10 units up in the Larsen score.

Conclusion
AAS is frequent in RA patients, particularly in those with markers of erosive disease.

Key words
Rheumatoid arthritis, cervical spine, atlantoaxial luxation.


*The members of the EMECAR Study Group are listed at the end of this paper.
This study was supported by the medical department of Aventis Pharma, SA.
Please address correspondence and reprint requests to: Antonio Naranjo, MD, Hospital Dr. Negrín, Rheumatology Department, 35020 Las Palmas, Spain. 
E-mail: anarher@gobiernodecanarias.org

Clin Exp Rheumatol 2004; 22: 427-432.
© Copyright Clinical and Experimental Rheumatology 2004.