Background Elderly-onset rheumatoid arthritis (EORA), starting after the age of 60, is a subset of disease with different characteristics from the classic younger-onset RA (YORA). There are conflicting results regarding disease outcomes in both groups.
Objectives We aimed to compare clinical, pathological and ultrasound (US) features between these two disease subsets in the Barts Early Arthritis Cohort (BEAC).
Methods We included 140 BEAC patients fulfilling ACR 2010 criteria (99 YORA, 41 EORA) and assessed clinical, US and synovial histology data at baseline and after 6 months of DMARD therapy. US synovitis was defined according to the OMERACT definitions: synovial thickening (ST) and Power Doppler (PD) signal were scored using a previously reported semi-quantitative 0-3 score on a data set of 10 metacarpophalangeal and both wrist joints (range 0-36). We assessed the correlation between DAS28, US scores and histological parameters (Krenn score and staining for B cells, T cells, lining and sublining macrophages) using Spearman correlation and assessed for potential interactions with EORA.
Results At baseline, disease activity parameters were similar between both groups, including inflammatory markers, joint counts, DAS28 and CDAI. However, EORA patients had significantly higher US-ST (p=0.03) and US-PD (p=0.002) scores. This did not translate at the histological level, with similar Krenn synovitis (p=0.76) and individual cell scores (p>0.5). At baseline, US-PD scores correlated significantly with Krenn score (r=0.28, p=0.007) in both groups, but US-ST scores were only correlated with Krenn score in EORA patients (r=0.42, p=0.03), with a trend for interaction between EORA and US-ST (p=0.08). At 6 months, DAS28 and CDAI were not significantly different between EORA and YORA, but elderly-onset patients still presented higher US-ST (p=0.03) and US-PD (p=0.009) scores. At this time point, histological and US synovitis scores were not significantly correlated and EORA was not an effect modifier. DAS28 was significantly correlated with US-PD score at baseline (r=0.36, p=0.0001) and 6 months (r=0.26, p=0.02) and with Krenn score at baseline (r=0.36, p=0.0001), always without significant interaction with EORA. Notably, at 6 months EORA was an effect modifier of the correlation of DAS28 and Krenn score (interaction p=0.05), with a strong significant correlation in the EORA group (r=0.56, p=0.003) and a mild non-significant correlation in YORA (r=0.14, p=0.35). A similar pattern was observed in 6-months' individual cell scores, with significant correlations with DAS28 in EORA but not YORA for sublining (r=0.64, p=0.0005; interaction p=0.09) and lining macrophages (r=0.81, p<0.0001; interaction p=0.009), B cells (r=0.52, p=0.007; interaction p=0.04) and T cells (r=0.46, p=0.019; interaction p=0.09).
Conclusions EORA patients presented similar clinical and histological activity scores to the YORA group, but higher US synovitis, both at baseline and at 6 months. EORA was an effect modifier of the correlation between DAS28 and Krenn synovitis score at follow-up, which was significant only in this subgroup of patients. Thus, after institution of DMARD therapy, DAS28 seems to better translate the degree of synovial inflammation in elderly-onset disease than in younger-onset patients.
Disclosure of Interest None declared