Compared imaging of the rheumatoid cervical spine: prevalence study and associated factors

Joint Bone Spine. 2009 Jul;76(4):361-8. doi: 10.1016/j.jbspin.2008.10.010. Epub 2009 Mar 19.


Introduction: Cervical spine involvement is common and potentially severe in patients with rheumatoid arthritis (RA). The objectives of this study were to compare the prevalences of cervical spine abnormalities detected by standard radiography, computed tomography (CT), and magnetic resonance imaging (MRI) in patients with RA; and to identify factors associated with cervical spine involvement.

Methods: We studied 40 patients who met American College of Rheumatology criteria for RA and had disease durations of 2 years or more. Each patient underwent a physical examination, laboratory tests, standard radiographs (anteroposterior, lateral, open-mouth, flexion, and extension views), MRI with dynamic maneuvers in (if not contraindicated), and CT.

Results: Cervical spine involvement was found by at least one imaging technique in 29 (72.5%) patients (standard radiography, 47.5%; CT, 28.2%; and MRI, 70%) and was asymptomatic in 5 (17.2%) patients. C1-C2 pannus was the most common lesion (62.5% of cases), followed by atlantoaxial subluxation (AAS, 45%). The most common AAS pattern was anterior subluxation (25%), followed by lateral subluxation (15%) then by vertical, rotatory, and subaxial subluxations (10% each). Erosions of the dens were seen in 67.5% of patients by MRI, 41% by CT, and 12.5% by standard radiography. Of the 10 cases of anterior AAS by any modality, 9 were detected by standard radiography and 7 by MRI. CT was the best technique for visualizing atypical rotatory or lateral AAS. MRI was best for assessing the C1-C2 pannus, dens erosions, and neurologic impact of the rheumatoid lesions. The comparison of patients with and without cervical spine lesions suggested that higher modified Sharp score and C-reactive protein values predicted cervical spine involvement (P=0.002 and P=0.004, respectively).

Conclusion: Cervical spine involvement is common and may be asymptomatic, indicating that routine cervical spine imaging is indicated in patients with RA. Standard radiography including dynamic views constitutes the first-line imaging method of choice. Sensitivity and comprehensiveness of the assessment are greatest with MRI. MRI and CT are often reserved for selected patients. Cervical spine involvement is associated with disease activity and with rapidly progressive joint destruction.

Publication types

  • Comparative Study

MeSH terms

  • Adult
  • Aged
  • Aged, 80 and over
  • Arthritis, Rheumatoid / blood
  • Arthritis, Rheumatoid / diagnostic imaging*
  • Arthritis, Rheumatoid / pathology*
  • C-Reactive Protein / metabolism
  • Cervical Vertebrae / diagnostic imaging*
  • Cervical Vertebrae / pathology*
  • Cross-Sectional Studies
  • Female
  • Humans
  • Magnetic Resonance Imaging
  • Male
  • Middle Aged
  • Prevalence
  • Prospective Studies
  • Radiography
  • Sensitivity and Specificity
  • Tomography, X-Ray Computed


  • C-Reactive Protein