Risk Factors of Cervical Spine Involvement in Patients With Rheumatoid Arthritis in the Imaging Era: A Cross-Sectional Study.

Background: Cervical spine involvement remains one of the more serious complications of rheumatoid arthritis (RA). Several factors interfere with the occurrence of this complication. The aim of this work was to identify the risk factors associated with cervical involvement, specifying the contribution of MRI in screening for rheumatoid cervical spine and discussing the value of standard X-rays.

Methods: A cross-sectional study included 100 patients with RA who underwent both standard radiography and MRI of the cervical spine. Clinical, paraclinical and therapeutic data were collected.

Results: Cervical pain was the predominant functional sign. It was mechanical in 44% of cases and inflammatory in 29%. Physical signs revealed on clinical examination included: pain on cervical spine mobilisation (49%), cervical stiffness (37%) and abnormal neurological examination in 3% of cases. Standard radiographs revealed C1-C2 instability in 52% of cases: vertical displacement (28%), rotatory displacement (22%), anterior displacement (19%) and lateral displacement (1%). MRI confirmed the presence of the following types of atlantoaxial subluxation: vertical (34%) and anterior subluxation (27%). Other lesions were identified on MRI: synovial thickening (31%) with contrast enhancement (17%), hyperintensity on diffusion-weighted imaging (DWI; 6%) and odontoid erosion (16%). Cervico-Medullar Angle (CMA) was pathological (CMA < 135°) in 3% of patients. A study of the correlation between patient-related parameters, RA-related parameters, cervical symptomatology-related parameters and rheumatoid C1-C2 involvement concluded that age was positively correlated with C1-C2 instability (p = 0.036), as was the presence of Rheumatoid Factor (RF) (p = 0.038), coxitis (p = 0.042) and dry eye syndrome (p = 0.039). Cervical spine stiffness was positively correlated with C1-C2 instability (p = 0.01). The presence of inflammatory cervical pain (p = 0), and painful mobilisation on examination (p = 0.03) were correlated with C1-C2 synovitis.

Conclusions: The 'RA and cervical spine' entity is associated with phenotypes of seropositive, destructive RA, with peripheral synovitis and systemic manifestations. Systematic clinical and radiological screening for this condition is essential to preserve functional prognosis. Diffusion imaging of the craniocervical region appears to be a valuable adjunct for the quantitative assessment of synovitis.