Background Salivary gland biopsy is essential in primary Sjögren's syndrome (pSS) diagnostics. However, tissue analysis using traditional methodology has several limitations including inaccurate quantification of lymphocytic infiltration and poor correlation with dryness. To perform biomarker identification in the target organ, tissue would have to be sacrificed. By performing saliva proteomics the biopsy tissue can be saved, but hitherto, this technique has not yielded consistent biomarkers and is limited by the absence of saliva production by many sicca patients.
Objectives We aimed to explore whether Luminex analysis of a broad panel of cytokines in salivary gland biopsy supernatants (secretome) could provide biomarkers to stratify sicca patients and could give insights into pathogenesis.
Methods Labial salivary gland (LSG) tissues were rinsed after biopsy and incubated in 200μL of saline for 1h at room temperature. Tissue supernatants were rendered cell-free, frozen in liquid nitrogen and stored at -80°C. In supernatants from pSS and non-Sjögren's sicca (nSS) patients 104 targets were measured by Luminex. Eight pSS and 8 nSS patients were selected for analysis based on matched biopsy weights. Results from this discovery cohort were validated in an additional cohort (n=18 nSS, n=16 incomplete SS: iSS, n=26 pSS) and correlations with clinical parameters were assessed. Non-SS were defined as sicca patients without lymphocytic infiltration in the salivary gland biopsy or anti-SSA/SSB autoantibodies. Incomplete SS patients were defined as sicca patients having lymphocytic infiltration (lymphocytic focus score (LFS)>0) and/or anti-SSA/SSB autoantibodies but do not fulfill the AECG classification criteria and are not diagnosed as pSS.
Results Levels of 20 cytokines were significantly different between the nSS and pSS patients in the discovery cohort (p≤0.05). These 20 and 13 additionally selected cytokines based on a trend towards statistical significance and/or literature, were measured in a validation cohort. Weights of the biopsies did not significantly differ: 59.8±48.1mg in nSS vs 72.7±45.2mg in iSS vs 67.4±28.6mg in pSS. Fifteen out of these 20 cytokines were validated. From the 13 cytokines 7 were significantly elevated in pSS vs nSS. In iSS CXCL10 (IP-10) and CCL19 (MIP-3β) were significantly elevated. Cytokines correlating with LFS, ESSDAI, ESSPRI, % IgG and IgM+ plasma cells in LSG, Schirmer and/or serum IgG with Spearman r≥0.4 and p≤0.05 in pSS were selected for classification tree analysis, these were IL-2, IL-3, IFN-β, IL-21, CXCL13 (BLC), CXCL10 and CCL19. Using CXCL13 and IL-21 levels, 87.5% of pSS patients could be classified correctly. Based on the used cut off levels, 5 nSS and 9 iSS patients would be classified as pSS. Follow up of these patients may reveal development of pSS.
Conclusions Elevated levels of numerous cytokines were found in LSG biopsy secretomes from pSS patients versus non-autoimmune sicca patients correlating with clinical parameters. This method represents a novel tool to provide insights in pSS immunopathology and to identify therapeutic targets and biomarkers for diagnosis, prognosis and treatment response.
Disclosure of Interest None declared