Background There is an unmet need for RA markers that inform prognosis and treatment response. While RF and anti-CCP are useful, augmenting their predictive capacity with additional markers is desirable. CRP is routinely used to gauge response to therapy but remains low in a proportion of patients, limiting its clinical utility in that subgroup. 14-3-3eta, a soluble RA diagnostic marker, is involved in the disease process, complements RF and anti-CCP and is modifiable with anti-TNF therapy in RA and PsA.
Objectives We examined whether positive serum 14-3-3eta corresponds with RA disease profile, and whether baseline titres inform response to anti-TNF therapy, also in patients with a lower CRP.
Methods Serum 14-3-3eta was measured in banked samples of 112 patients with established RA (N=75; RAPPORT cohort) and early disease (N=37; TEACH and I-COBRA cohorts). Patients were biologically-naïve, median age 57 and 76% female. 14-3-3eta was considered +ve at >0.19 ng/ml and median differences in clinical variables between 14-3-3 eta +ve and -ve groups were evaluated using the Mann-Whitney U-test. For relationships between the clinical variables and 14-3-3eta, Pearson correlations were run. The 75 established RA patients were candidates for anti-TNF therapy and 14-3-3eta serum concentrations were additionally measured at ~15 weeks post-treatment. EULAR criteria were used to define Good response. Across the group and in patients with a low CRP (≤10mg/L), the significance of median differences in 14-3-3eta expression between Good EULAR responders and non-responders was evaluated using the Mann-Whitney U-test. A ROC curve and corresponding AUC was generated.
Results Of the 112 patients, 92 (82%) were 14-3-3 eta +ve and 20 (18%) were -ve. In the 14-3-3eta +ve vs -ve group the following medians were significantly higher; HAQ (1.9 vs 1.0, p=0.001), ESR (40.5 vs 25.0mm/hr, p=0.007), CRP (18.8 vs 6.6mg/l, p=0.018) and anti-CCP (54.6 vs 5.0U/ml), p=0.02); while DAS28 (6.7 vs 6.0, p=0.16) and RF (12.0 vs 8.0IU/ml, p=0.12) showed only a trend. No significant Pearson correlations emerged between any of the baseline variables and 14-3-3eta titres, especially CRP. Of the 75 established RA patients receiving anti-TNFs, 15 (20%) achieved a Good EULAR response and had significantly lower median baseline 14-3-3eta than those that did not (0.72 vs 2.52ng/ml; p=0.015). Low CRP did not correspond with treatment response and 27 of the 75 patients (36%) had a baseline CRP ≤ 10mg/L; 7 (26%) of which achieved a Good EULAR response and had a lower median 14-3-3 eta than the 20 that did not (0.0ng/ml vs 3.5ng/ml; p=0.0043). The corresponding ROC AUC for response in the low CRP group was 0.88 with an optimal 14-3-3eta cut-off of 0.48ng/ml; LR+ =14.3 and LR- =0.31.
Conclusions In RA, 14-3-3eta sero-positivity marks more severe disease and lower titres correspond with better response to anti-TNF therapy, also in patients with lower CRP.
Disclosure of Interest W. Maksymowych: None Declared, V. Bykerk: None Declared, K. Siminovitch: None Declared, M. Boers: None Declared, R. Landewé: None Declared, D. van der Heijde: None Declared, P. P. Tak: None Declared, M. Genovese: None Declared, M. Weinblatt: None Declared, E. Keystone: None Declared, K. Young: None Declared, A. Marotta Employee of: Augurex Life Sciences Corp