Predictive markers of lymphomagenesis in Sjögren's syndrome: From clinical data to molecular stratification.

Sjögren's syndrome (SS) is a chronic systemic autoimmune disease, affecting predominantly the exocrine glands, a large array of systemic manifestations and high risk of lymphoma development. The latter constitutes the major adverse outcome of SS contributing in the increased morbidity and mortality of the disease. The vast majority of lymphomas in SS are B-cell non-Hodgkin's lymphomas (NHL), primarily indolent mucosa-associated lymphoid tissue (MALT) lymphomas, followed by nodal marginal zone lymphomas (NMZL) and diffuse large B cell lymphomas (DLBCL). In the last 3 decades and due to the adverse impact of NHL in disease outcome, an effort has been undertaken to identify markers and models predicting patients with SS at high risk for lymphoma development. Several epidemiological, clinical, laboratory and histological parameters, some of which are evident at the time of SS diagnosis, were proved to independently predict the development of NHL. These include salivary gland enlargement, skin vasculitis/purpura, glomerulonephritis, peripheral neuropathy, Raynaud's phenomenon, lymphadenopathy, splenomegaly, cytopenias, hypocomplementemia, cryoglobulinemia, rheumatoid factor, anti-Ro/La autoantibodies, hypergammaglobulinemia, serum monoclonal gammopathy, biopsy focus score and organization of lymphocytic infiltrates in the salivary glands into ectopic germinal centers. Prediction models combining some of the afore-mentioned predictors have also been described. However, the identification of specific and sensitive molecular biomarkers, related to the process of lymphomagenesis is still pending. Recently, we described a novel biomarker the miR200b-5p micro-RNA. Low levels of this miRNA in the minor salivary glands, appears to discriminate with high specificity and sensitivity the SS patients who have from those who do not have NHL. miR200b-5p, being expressed years before the clinical onset of NHL, independently predicts NHL development with a predictive value higher than the previously published multifactorial models and has a possible role in the monitoring of therapeutic response. Thus, it is a strong candidate for the identification and follow-up of patients at risk.