Background Rapid radiographic progression (RRP) is considered as a representative poor outcome in patients with RA. Previous studies reported that the swollen joint counts, inflammatory indices, autoantibodies and plain radiographic erosion at baseline are associated with the development of RRP. Recently, EULAR recommendations for the use of imaging of the joints in the clinical management of RA state that joint damages detected by MRI involve in the subsequent erosive progression of RA. However, there are few clinical investigations searching the prognostic factors toward RRP by combining MRI findings with other indices such as physical findings and serum variables.
Objectives To examine whether the MRI findings are superior to other indices to predict the subsequent development of RRP in patients with early-stage RA.
Methods This is a 1 year observational study from seventy-six early-stage RA patients recruited consecutively from Nagasaki University Early Arthritis Cohort in which the subjects received Gd-enhanced MRI of both wrists and finger joints every 6 months. All of the patients had been received DMARDs at baseline. The presence and degree of synovitis, bone marrow oedema and bone erosion in MRI were examined by OMERACT-RAMRIS. Plain radiographic damages of both wrists and finger joints was scored by Genant-modified Sharp score and the development of RRP at 1 year was identified as a score >3/year. We have tried to examine the factors to predict the development of RRP at 1 year including baseline data of physical synovitis, serum variables, MRI findings, plain radiographic damages and early clinical response at 3 months.
Results Median age and disease duration were 54.5 y.o and 3 months. Median DAS28-CRP, CRP (mg/dl), matrix metalloproteinase 3 (MMP-3) (ng/ml) was 4.31, 0.54 and 72.6, respectively. The rate of ACPA-positive and RF-positive were 82.9% and 79.0%. Median RAMRIS synovitis score, bone marrow odema score, bone erosion score and Genant-modified Sharp score at baseline were 9, 1, 0 and 0, respectively. Fifty-two out of 76 patients showed moderate to good clinical response at 3 months determined by DAS28-EULAR response criteria, however, RRP was developed in 12 patients at 1 year. Multivariate logistic regression analysis has identified that RAMRIS bone marrow oedema score at baseline (5 increase, Odds ratio 2.18, 95% C.I. 1.319-3.590, p=0.0023) is the only independent predictor toward the development of RRP at 1 year (Table 1).
Conclusions Our present data suggest that MRI bone marrow oedema is closely associated with poor radiographic outcome in patients with early-stage RA. Physicians should especially consider the tight control of disease activity if MRI bone marrow oedema is obvious in early RA patients.
Disclosure of Interest None declared