Background Thymic Stromal Lymphopoietin (TSLP) is a potent immunomodulatory IL-7 related cytokine involved in Th2 mediated immune responses and homeostatic T-cell expansion. Reduced TSLP expression by intestinal epithelial cells was recently shown to lead to reduced Th2 responses and development of Th1 mediated experimental colitis. In addition, TSLP is described as a proinflammatory factor in rheumatoid arthritis, which is driven by Th1/Th17 responses. A Th1/Th17 polarized environment is also present in the salivary glands of patients with Primary Sjögren’s syndrome (pSS).

Objectives To investigate TSLP expression in salivary glands of pSS patients as compared to non-SS Sicca (nSS) patients and to study the relationship to local and systemic disease parameters.

Methods Tissue sections from minor salivary glands of 38 pSS and 18 nSS patients were stained with a monoclonal antibody against human TSLP or an isotype control. Cells stained positive for TSLP were quantified per mm²of tissue. In addition, TSLP was quantified at sites with intact epithelium where no infiltrating lymphocytes were present. Quantified TSLP levels were correlated to local (lymphocyte focus score; LFS, %IgA positive plasma cells) and systemic (erythrocyte sedimentation rate; ESR, serum IgG levels) disease parameters.

Results TSLP expression was almost exclusively expressed by acinar cells in both pSS and nSS patients. The amount of TSLP-expressing cells per mm²was significantly decreased in pSS patients as compared to nSS patients (462±42 vs. 773±84, p<0.01). The number of TSLP-producing cells correlated negatively to LFS (r= -0.48, p<0.001) ESR (r= -0.41, p<0.01), serum IgG levels (r= -0.41, p<0.01) and correlated positively to the percentage of local IgA producing plasma cells (r=0.35, p<0.05). At sites with intact epithelium lacking cellular infiltrates, TSLP expression tended to be reduced in pSS patients as compared to nSS patients (840±75 vs. 1064±72, p=0.079). Furthermore, pSS patients with a LFS equal to or higher than 3 showed significantly decreased amounts of TSLP-producing cells per mm²at these sites compared to nSS patients (677±105 vs. 1064±72, p<0.01). Also, in intact epithelium, the amount of TSLP-producing cells correlated negatively to LFS (r= -0.40, p<0.01), ESR (r= -0.32, p<0.05) and serum IgG levels (r= -0.27, p<0.05)

Conclusions Reduced TSLP expression in pSS patients is associated with local and systemic inflammatory markers, including increased lymphocytic infiltration. Considering the described role of TSLP in promoting Th2 responses at mucosal sites, we hypothesize that TSLP is constitutively expressed in salivary glands and promotes a protective Th2 milieu, whereas loss of TSLP expression may contribute to Th1/Th17 associated immunopathology in pSS.

Disclosure of Interest None Declared