Usefulness of magnetic resonance imaging of the hand versus anticyclic citrullinated peptide antibody testing to confirm the diagnosis of clinically suspected early rheumatoid arthritis in the absence of rheumatoid factor and radiographic erosions.

OBJECTIVE: The diagnosis of rheumatoid arthritis (RA) is sometimes difficult to establish early in the disease process, particularly in the absence of its classic hallmarks. Our aim was to compare the practical usefulness of magnetic resonance imaging (MRI) of the hand versus anticyclic citrullinated peptide (anti-CCP) antibody testing to confirm the diagnosis of clinically suspected RA in the absence of rheumatoid factor (RF) and radiographic erosions.
METHODS: We prospectively included patients with early inflammatory arthritis and strong clinical suspicion of RA, in whom initial complementary tests (RF and radiographs of hands, wrists, and feet) did not provide unequivocal confirmation of the diagnosis. In all patients, anti-CCP antibodies were assessed and contrast-enhanced MRI of the most affected hand was performed according to a specifically designed protocol. The MRI criterion for the diagnosis of RA was either the presence of synovitis with bone erosions or bone marrow edema, which is currently considered to be a forerunner of erosions.
RESULTS: In the 40 patients (28 women), the mean age at diagnosis was 54 +/- 6 years and the median duration of symptoms was 4 +/- 2.6 months (range 1.5 to 12). Final diagnoses at 1-year follow-up were RA in 31 patients, undifferentiated arthritis in 7 (5 self-limiting), and psoriatic arthropathy (PsA) and antisynthetase syndrome in 1 patient each. Anti-CCP antibodies were positive only in 7 patients, all of whom were finally diagnosed with RA. The prevalence of anti-CCP positivity in our series of seronegative RA patients was thus 23% (7/31); in these patients the anti-CCP antibodies had a specificity of 100% (95% CI: 71.7 to 100) and sensitivity of 23% (95% CI: 9.6 to 41.1). Use of the MRI criterion led to the correct diagnosis in 100% of patients with RA and to false-positive results (1 with PsA and 1 with antisynthetase syndrome). The MRI criterion had a specificity of 78% (95% CI: 40.0 to 97.2) and sensitivity of 100% (95% CI: 90.8 to 100) for identification of seronegative RA.
CONCLUSION: Although the tests are not mutually exclusive, in our experience MRI is more helpful than anti-CCP antibody determination in confirming the diagnosis of clinically suspected early RA in patients in whom the diagnosis cannot be confirmed using conventional methods.