Background Interleukin-6 (IL-6) is one of the most important cytokines in rheumatoid arthritis (RA). IL-6, mainly produced in the inflammed joint, is involved in several features of structural damage. Although some authors suggested that IL-6 assessment may improve RA management, data is scarce in early arthritis
Objectives To determine whether IL-6 could be used to assess disease activity and structural damage in early arthritis. For this, we assessed the relationship between IL-6, CRP as a comparator, and several outcome measures during the first 36 months of an early arthritis cohort
Methods The early arthritis cohort ESPOIR is a French national observational cohort of patients with at least 2 joints affected by synovitis for >6 wks and <6 months at baseline. IL-6 was assessed at each timepoint from frozen sera using a chemiluminescence assay (Elecsys® IL-6, Roche Diagnostics). Thus, IL-6 and CRP were compared to repeated measurements through 36 months of inflammation (i.e. swollen joint count (SJC) and ESR-DAS28), function assessed by HAQ and radiographic severity by Total Sharp van der Heijde score (SHS). Progressors were defined as patients with a change of SHS >1 at 12 months. The hypothesis was that IL-6 at baseline and through follow-up would be better correlated than CRP to radiographic change and to persistent synovitis. To further assess the relationship between IL-6 and inflammation, the presence and the grade of synovitis was recorded in a sample of 126 patients of the cohort using ultrasonography (US). Last Observation Carried Forward imputation method was used. The relationship between the IL-6 level either at baseline or over time (AUC IL-6), CRP, and the above mentioned parameters was investigated by Spearman’s rank test
Results 812 patients of the 813 patients in the cohort had IL-6 levels. Of these 812 patients, 623 (76.7%) were female, mean (SD) age was 48.1 (12.5) years, and mean duration of joint swelling was 4.8 (5.8) months; 579 patients (71.3%) satisfied the ACR 1987 criteria for RA at baseline. Baseline IL-6 was 25.5 (59.0) pg/ml and was significantly higher in early RA (30.9 (67.6) pg/ml) than in early undifferenciated arthritis (12.2 (23.1) pg/ml, p<0.0001). IL-6 was twice higher in the 447 progressors (31.1 (75.0) vs. 15.0 (23.2), p<0.0001). The variation through follow-up of both IL-6 and CRP moderately correlated with the variation of HAQ (respectively r=0.220 and r=0.214, both p<0.0001) and with the change of ESR-DAS28 (r=0.279 and r=0.268, both p<0.0001). Variation of SJC was more correlated with variation of IL-6 (r=0.382, p<0.0001) than with variation of CRP (r=0.159, p<0.0001). SHS change was more associated with variation of IL-6 (r=0.330, p<0.0001) than with variation of CRP (r=0.077, p=0.031). At baseline, IL-6 significantly correlated with the number of US synovitis ((r=0.198, p=0.025 in B mode, r=0.259, p=0.003 in Doppler mode) while the correlation was not significant with CRP (r=0.087, p=0.330 in B mode and r=0.066, p=0.462 in Doppler mode)
Conclusions IL-6 monitoring is correlated with persistent synovitis and should be considered as a surrogate marker of RA and structural damage in early arthritis.
Disclosure of Interest A. Baillet: None Declared, L. Gossec: None Declared, S. Paternotte: None Declared, N. Hoyle Employee of: Roche Diagnostics GmbH, B. Combe: None Declared, O. Meyer: None Declared, M. Dougados: None Declared