Background MRI is increasingly used to objectively assess disease activity and damage in patients with rheumatoid arthritis (RA), especially in clinical trials. The preferred scoring method is the RAMRIS, which implies the use of gadolinium, an intravenous contrast agent, to assess synovitis.
Objectives We evaluated in a small preliminary study if the fusion of MRI and US Power Doppler (PD) images could avoid gadolinium.
Methods 12 patients (10 women) affected by RA, with at least one involved wrist, were studied. Mean age was 58.8±9.1 years and mean disease duration was 54.5±56.1 months. Disease activity (DAS 28-CRP) was evaluated on the day of the examinations.
MRI was performed on a extremity-dedicated machine (Oscan 0.3 T, Esaote, Genova Italy) and US by MyLab Twice ultrasound scanner with a virtual navigator software (Esaote, Genova, Italy). T3Dt1 sequences were used in the fusion images (parameters: TR/TE/NEX 38/16/1, matrix 192*192*39, FOV 160*160*60, thickness 0.6 mm). US scans were performed sagittally on the radio-lunate joint and axially on the proximal carpal bones of the dorsal wrist with B-mode and PD (PRF 500–750Hz; WF: 3; maximum gain with probe's calibration in air). Magnetic position sensor linked to the US transducer (13 MHz) and low magnetic field transmitters were used to allowed fusion images.
MRI synovitis was scored by the RAMRIS and with the contrast-enhanced dynamic method resulting in rate of early enhancement (REE) and relative enhancement, (RE). The US score was the Global OMERACT-EULAR Synovitis Score (GLOESS) and its individual parts.
Results Inter-reader agreement for PD by weighted kappa was 0.75 (75%CI 0.53–0.96) for the sagittal and 1 (75%CI 1–1) for the axial view. It was 0.85 (75%CI 0.69–0.98) for MRI synovitis.
There was a correlation between DAS28 and the MRI synovits score (r=0.59, p=0.04), REE (r=0.60, p=0.04), and RE (r=0.58, p=0.05), US Doppler in the long (r=0.75, p=0.005) and in the axial axes (r=0.86, p=0.01), but not with B-mode synovial effusion (r=0.56, p=0.056). The GLOESS correlated with DAS28 (r=0.66, p=0.019), but not with RAMRIS synovitis, erosions or BME nor with REE. The MRI wrist synovitis score did not correlate with the Doppler score on the sagittal and axial views, nor with US B mode score. This was true also for the three individual MRI slices of the wrist, the REE and the RE. The figure shows from the left the pre-contrast axial MRI image, the post-contrast and the fusion images.
Conclusions Although both MRI and Doppler are good indicators of disease activity in RA, they seem not to be interchangeable. This may be due to the fact that MRI and US show different features of synovitis or, alternatively, that MRI comprised the whole wrist whereas US visualized only its dorsal area. Although our preliminary data do not support the use of fusion images, these should be investigated on larger number of patients with a more extended PD approach.
Disclosure of Interest None declared