Background Prescription of recombinant biological agents in rheumatoid arthritis (RA) patients in Belarus is limited to the cases of permanent high disease activity and severe extra-articular manifestations including secondary Sjögren's syndrome (sSs). The possibility of sSs prediction could ensure more reasonable use of expensive biological agents (e.g., rituximab) in RA treatment at the early stage of the disease. It has the particular importance since Sjögren's syndrome (Ss) is highly associated with malignant lymphomas development [1].

Objectives To determine clinical and genetic predictors of sSs development in RA patients as well as to assess their prognostic value.

Methods RA and Ss were diagnosed according to the ARA criteria [2,3]. Pre-test probability (P_pre) of sSs development in RA patients was determined retrospectively in 590 RA patient sample (482 women and 108 men; 2625 visits), hospitalized into the rheumatology department of 9th Minsk clinical hospital in 1981-1996 and further observed at Republican rheumatologic center up to 2005 inclusive. sSs was diagnosed in 49 RA patients (8.3%; CI₉₅ 6.3-10.8%). The frequency of sSs was 10.0% (7.6-13.0%) in women and 0.9% (0.2-5.1%) in men with RA. 95% confidence interval (CI₉₅) of sSs frequency was calculated by Wilson method. Power, direction and statistical significance of marker associations with sSS were determined by the correlation coefficient gamma (R_G). Prognostic odds ratio (pOR), positive and negative likelihood ratios (LR⁺, LR^–) for the biomarkers selected by R_G were estimated by fourfold table and two-sided statistical significance of pOR- by Fisher exact test (p_2–t). Post-test probability (P_post) of sSs was assessed by formula based on the Bayes theorem [4]. 14 clinical symptoms presented before the sSs manifestation were analyzed as possible clinical predictors and A (8), B (17), C (8), DR (10), DQ (4) locus specificities of HLA system were analyzed as possible genetic markers of sSs development in RA patients.

Results Primary selected clinical biomarkers with R_2–t<0.05) clinical markers of sSs development: sex (pOR=11.8), AIT (4.3), Rs (4.3), CIN (2.9), RV (2.8), RN (2.4) and genetic markers of sSs risk: B14 (6.0), A10 (3.3), DQ2 (3.1), DR3 (2.9), B18 (2.4), B8 (2.1) as well as protective genetic markers of sSs: A9 (-6.4), B35 (-3.9).

Conclusions The obtained results ensure sufficiently convincing prediction of sSs development in RA patients. sSs P_post increases 3,1 times (from 10.0% to 30.8%) in a woman with RA and AIT and 9.0 times (from 10.0% to 90.2%) in the case of Rs and HLA-B14 presence, exceeding the required significance level (90.0%) for sSs prediction.

References

1. EULAR compendium on rheumatic diseases. Milan, 2009. P.314-28.

2. Arnett FC et al. Arthr & Rheum 1988;31(3):315-24.

3. Fox RI et al. Arthr & Rheum 1986;29:577-85.

4. Rudwaleit M et al. Arthr & Rheum 2006;54(2):569-78.

Disclosure of Interest None declared

DOI 10.1136/annrheumdis-2014-eular.4613