14-3-3ETA BIOMARKER DYNAMICS IN CIA MARK DISEASE SEVERITY IN A NON-INTERVENTIONAL STUDY

Background: Biomarkers are increasingly used in RA interventional studies as indicators of efficacy/safety/prognosis. We previously reported that serum 14-3-3η (eta) and its auto-antibodies (AAbs) are mechanistic markers that inform RA therapy response. 14-3-3η has been identified as a drug target because it potently induces inflammatory and joint damage factors. Prospective clinical studies demonstrate that 14-3-3η may predispose to, and its AAbs may protect from radiographic progression. Such non-interventional biomarker dynamics may strategically guide therapeutic development programs. Objectives: To examine serum dynamics of 14-3-3η markers along the arthritis disease course in a non-interventional study using a collagen-induced arthritis (CIA) murine model. Methods: Thirteen DBA/1LacJ mice were induced with 100μl of RA agent; 0.1mg of collagen+0.1mg of M. tuberculosis administered once, intradermally at the base of the tail for a 52-day disease course. Serum was collected on induction, sacrifice day and every 3 days in between. Animals were sacrificed either due to study conclusion (Day 52) or severity of arthritis symptoms measured by visual arthritis score (0-4) and left/right paw caliper measurements (mm). Mice were divided into 2 severity groups, 'severe' defined by early sacrifice Days 35/38 (N=5) or 'mild' by late sacrifice, Day 45/52 (N=8). 2-tailed paired t-tests were used to compare mean differences in biomarker expression between induction and sacrifice days. A Pearson correlation was run between 14-3-3η protein and AAbs in the two severity groups. Due to 14-3-3η protein and AAb expression range disparity, [protein/(Log10 AAbs)] x 100 was used to reach a biomarker unit ratio. For survival analysis, the groups were defined by above or below the mean 14-3-3η ratios of the induction and sacrifice date. A Kaplan-Meier curve was generated, the Log-Rank Mantel-Cox test was used for curve comparisons and a Hazard ratio determined. Results: At induction, 14-3-3η mean (SD) serum levels were similar in the severe and mild arthritis groups, 0.21 (0.11) and 0.28 (0.10)ng/ml, p=0.26, and decreased significantly by sacrifice date in the mild group to 0.04 (0.03)ng/ml, p=0.00008 but not in the severe 0.14 (0.14)ng/ml p=0.27. Both had a significant increase in 14-3-3η AAb titres (U/ml) by the sacrifice date; 53 to 364, p=0.01 and 28 to 278, p=0.001. The 14-3-3η protein:AAb ratios by paired t-test were significantly lower at sacrifice vs. induction date in the mild group (1.8 vs. 19.4 p=0.00008) while the change was not significant in the severe group (5.5 vs. 14.3, p=0.07). 14-3-3η AAb levels in the mild group correlated inversely and strongly with 14-3-3η protein levels (r=-0.84, p=0.002), but not in the severe (p=0.75). The survival groups were defined as mice achieving a protein:AAb ratio reduction of >10 (N=8) or <10 (N=5) by sacrifice date. The group that had >10 ratio reduction (and therefore lower protein burden) had longer median survival of 45 versus 36 days (95%CI:0.47-1.1). The Mantel-Cox test returned a Chi-sq of 6.6 p=0.01 and the Hazard ratio was 9.6 (95%CI:1.7-54.0). Conclusions: Consistent with 14-3-3η expression in RA, higher serum levels are deleterious in CIA and clearance of the protein by 14-3-3η AAbs results in less severe arthritis and longer survival. 14-3-3η may assist with drug development programs and should be tested further for personalized medicine strategies. Disclosure of Interest: A. Abulrob: None declared, W. Maksymowych Consultant for: Augurex Life Sciences Corp, A. Marotta Employee of: Augurex Life Sciences Corp DOI: 10.1136/annrheumdis-2014-eular.3427Citation: Annals of the Rheumatic Diseases, volume 73, supplement 2, year 2014, page 853Session: Rheumatoid arthritis - etiology, pathogenesis and animal models (Abstracts accepted for publication )