Increased lymphoma risk in patients with systemic manifestations of Sjögren's disease: a population-based study

To assess clinical and immunological characteristics, disease activity, and lymphoma development in a well-characterized cohort of patients with Sjögren's disease, with particular emphasis on differences between glandular and systemic phenotypes.

We conducted a retrospective observational study of adult patients (≥18 years) with Sjögren's disease not associated with any other systemic autoimmune rheumatic disease, defined as primary Sjögren's disease (pSjD), who were seen at a tertiary center in Madrid, Spain. Patients were classified according to the 2016 ACR/EULAR criteria or, alternatively, according to a modified version in which salivary gland ultrasonography (SGUS) was used as an alternative minor criterion when unstimulated whole salivary flow (UWS) and/or ocular staining score (OSS) were not available. For patients diagnosed before 2016, clinical data from medical records were used to apply these criteria retrospectively during the study period. SGUS findings were graded according to the Outcome Measures in Rheumatology (OMERACT) scoring system or based on expert radiologists' assessment. Systemic disease activity was assessed using the EULAR Sjögren's Syndrome Disease Activity Index (ESSDAI). Demographic, clinical, laboratory, histopathological, lymphoma occurrence, and treatment data were retrospectively collected. Patients were stratified into glandular (exocrine-limited) or systemic (glandular and systemic involvement) phenotypes.

Among 278 patients included, 148 (53%) had a glandular phenotype and 130 (47%) a systemic phenotype. The mean age at diagnosis was 54.6 (± 14.5 years); 95% were women. Sicca symptoms were highly prevalent in both groups (xerophthalmia 89%, xerostomia 84%). In the systemic group, musculoskeletal involvement was most common, particularly arthritis (38%), followed by cutaneous manifestations (26%), cytopenias (19%), pulmonary involvement (16%), and renal or central nervous involvement (5%). ESSDAI scores were significantly higher in patients with systemic involvement (13.9 ± 21.7 vs. 4.0 ± 7.9; p < 0.001). Immunological profiles were similar across phenotypes. Lymphoma occurred in 15 patients (5%), predominantly in patients with systemic involvement (10% vs. 1%; p = 0.001), mostly non-Hodgkin lymphoma. Immunosuppressive therapy was more commonly used in patients with systemic involvement.

Adequate stratification of pSjD according to presence or absence of systemic manifestations is mandatory. Our results encourage a close follow-up of systemic pSjD phenotype, particularly for early diagnosis of lymphoma.