Background Advanced imaging has contributed to improve the understanding of the pathogenesis of gout and of its diagnosis. MRI bone marrow oedema (BME) is relatively frequent in the joints affected by gouty arthritis, but its significance is still controversial. The bone adjacent to tophi is frequently oedematous as a possible result of the osteoclast-activating property of urate crystals. By contrast, BME due to acute arthritis has been considered to be rare and mainly associated with concomitant septic arthritis.

Objectives To evaluate the frequency of BME of the wrist in a consecutive series of patients affected by gout with no clinical and laboratory evidence of local infection.

Methods Forty-nine consecutive patients affected by gout according to the preliminary ACR criteria were studied. In 31, urate crystals were identified in synovial fluid or tophi. Forty-one were men, mean age was 65.1±15.6 years at the time of MRI, and median disease duration was 84 months (range 3-492 months). Twenty-one patients had an ongoing or recent (less than 1 month) acute gout attack in the studied wrist and 3 more had had previous wrist involvement. Clinical characteristics related to gout (number of attacks, number of involved joints, presence of tophi, type of treatment) were recorded together with serum uric acid (SUA) and CRP.

MRI of the involved wrist using STIR sequences in the coronal and axial planes was obtained on a extremity-dedicated 0.2 T machine (Artoscan, Esaote, Genova, Italy). The wrist was chosen because it is frequently involved in gout and BME can be scored with a standardized system (the RAMRIS) developed for rheumatoid arthritis, which was applied to the bases of the five metacarpals, the wrist bones, radius and ulna (range 0-45).

Results Twenty-eight (57.1%) patients had at least one bone with BME RAMRIS score of one. By setting the cut off level at a score of 3, 21 patients (42.9%) still had BME. RAMRIS BME score of the wrist was higher in patients with wrist arthritis (median 6, range 0-32) than in those without (median 0, range 0-24; p=0.015). However, also 5 patients who did not recall wrist arthritis showed BME. The severity of BME did not correlate with the presence of tophi, SUA, number of involved joints, number of attacks, age, sex, but slightly correlated with disease duration (rho=0.320, p=0.028), CRP (rho=0.312, p=0.045), ESR (rho=0.328, p=0.044), and the use of colchicine (rho=0.329, p=0.023), which in turn was strongly correlated with disease duration (rho=0.404, p=0.005). All the patients had self limiting arthritis, which improved with colchicine. None of them had clinical and laboratory evidence of septic arthritis during a median follow up of 2 years.

Conclusions BME is frequent in patients with acute gout, is not related to septic arthritis or osteomyelitis, and is associated with more severe inflammation. Strenghts of this study are its prospective nature, with patients not selected for disease severity, and the choice of the wrist, a joint for which there is a standardized scoring for BME. Limitations are the low field of the magnet, which however, should have decreased, and not increased, the number of bones with BME identified. BME should be included among the typical MRI lesions of gout.

Disclosure of Interest None declared

DOI 10.1136/annrheumdis-2014-eular.4380