Risk factors for atlantoaxial dislocation in rheumatoid arthritis: A comparative study

The objective of the study was to assess clinical and imaging features of rheumatoid arthritis associated with atlantoaxial dislocation (RA-AAD) in comparison to RA without AAD (RA-C) and to identify the associated factors to AAD.

This was a retrospective comparative study including RA patients over 18 years old. The control group of RA-C was matched according to age (±2 years), gender, and RA duration (±2 years). General data, RA characteristics, AAD features, and treatment modalities were recorded. Statistical analysis was performed to determine the predictive factors of AAD.

A total of 120 patients were included (78 RA-AAD and 42 RA-C); sex-ratio was 0.15. Mean age was 58.55 ± 9.14 years (RA-AAD) versus 60.43 ± 10.78 years (RA-C), (p = 0.31). Regarding RA features, RA-AAD patients have significantly a higher rheumatoid factor (RF) rate (p = 0.002), extra-articular manifestations (p = 0.009) especially osteoporosis (p < 10^-3 ), eye involvement (p = 0.02), coxitis (p < 0.0001), Disease Activity Score (DAS28) (p < 10^-3 ), and global health status scale (HAQ) (p = 0.003). X-ray analysis showed pathologic PADI (<14 mm) in 51.6% of cases. The AADI was pathological (>3 mm) in 100% of patients with the following distribution: 67.9% between (3-6 mm) and 23.3% between (6-9 mm). Atlantoaxial dislocation (AAD) was presumed anterior (85.4%), lateral (7.3%) and rotatory (7.3%). The abnormalities observed on magnetic resonance imaging (MRI) were: C1-C2 synovitis (61%), confirmed AAD (34.1%) and basilar impression (4.9%). AAD diagnosed by cervical MRI was anterior in 89.8% of cases, lateral (5.3%) and with a double ascending and anterior component in 4.9% of cases. An immobilisation with a type C4 cervical collar was indicated for all the patients and five of them had boluses of corticosteroid (methylprednisolone 1 g/day) for three consecutive days. C1-C2 arthrodesis by trans-articular screwing was performed in four patients with neurological signs or pathological measurements on cervical MRI. The outcome was favourable in 100% of cases after surgical treatment with improvement in neck pain and neurological condition within a mean decline of 4 ± 3 years [1-7]. Multivariate analysis identified coxitis, osteoporosis, high baseline DAS28, and a high rate of RF positivity as predictive factors of AAD.

Our results confirmed that predictive factors of AAD were related to higher disease activity and structural damage. These modifiable factors may be prevented by tight control, early treatment, and educating RA patients to improve treatment adherence.