Background 14-3-3eta (η) is a mechanistic biomarker that, as a soluble ligand, dose-dependently induces key joint damage and inflammatory factors, including TNF-alpha (α). There is a high clinical unmet need in RA to identify markers that inform the likelihood of response to biological therapies, given the heterogeneity of patient response. We have previously reported that 14-3-3η marks response to TNFα therapy and postulate that the potent induction of TNFα by 14-3-3η may diminish the efficacy of standard anti-TNF doses in some patients or over time in the same patient.

Objectives This study evaluates whether various 14-3-3η cut-offs better inform the likelihood of a good EULAR response to anti-TNFs.

Methods Serum 14-3-3η was measured in a cohort of 74 RA patients who were candidates for anti-TNF therapy; all patients were refractory to standard DMARDs. Median patient age was 59 years and 73% were female. Pre-treatment 14-3-3η levels were measured on the Augurex 14-3-3η ELISA and EULAR classification criteria were used to define Good response to anti-TNF therapy. Two-tailed Mann-Whitney U-tests were used to assess group differences between Good and inadequate responders. Three 14-3-3η concentration cut-offs were selected as follows: the reported manufacturer's reference range ≥0.19 ng/ml as well as 2 and 4 times that level, namely 0.40 and 0.80 ng/ml. To assess 14-3-3η's association with a Good EULAR response, Fisher's exact test and stepwise multivariate analyses were used. Variables included in the model were age, gender, disease duration, HAQ, CRP, and DAS28-ESR.

Results Median 14-3-3η levels were significantly lower in patients who achieved a Good EULAR response compared to inadequate responders, 2.52 ng/ml (0.79 - 20) vs 0.72 ng/ml (0 - 7.01), p<0.03. Fishers' exact test revealed that the two lower 14-3-3η cut-offs were associated with a higher odds (OR) of a achieving a Good EULAR response to anti-TNF therapy. For patients with 14-3-3η levels ≤0.19ng/ml, the OR (95%CI) was 9.3 (1.9 - 45.4), p<0.008; ≤0.40 ng/ml was 9.5 (2.4 - 37.1), p<0.002; ≤0.80 ng/ml was 3.4 (1.0 - 10.8), p<0.05. Multivariate analysis revealed that 14-3-3η, at all 3 cut-offs, was the only independent predictor of a Good EULAR response with the ≤0.40 ng/ml being the strongest predictor of response to anti-TNF therapy with a likelihood ratio of 10.2, p=0.0014 yielding an R² of 18%.

Conclusions Consistent with 14-3-3η's pathophysiological mechanism as an upregulator of TNFα, higher 14-3-3η levels are clinically associated with a lower likelihood of response to anti-TNF therapy. 14-3-3η may inform anti-TNF dosing/administration strategies to optimize response to therapy.

Disclosure of Interest A. Marotta Employee of: Augurex Life Sciences Corp, W. Maksymowych Consultant for: Augurex Life Sciences Corp

DOI 10.1136/annrheumdis-2014-eular.3426