PO:01:025 Comparison of classification criteria and clinical diagnosis in antiphospholipid syndrome: insights from a tertiary centre cohort

Objectives To compare the performance and agreement between APS classification criteria and their concordance with clinical diagnosis.Methods Retrospective cohort study of patients with APS followed at Rheumatology Department, Coimbra’s Hospital. Inclusion required clinical diagnosis by rheumatologists. Data were obtained from medical records (1999-2025). The proportion fulfilling classification criteria and reasons for discordance were assessed.Results We included 86 patients with clinical diagnosis of APS (83% female; median age at diagnosis 39). Primary APS affected 55%, while the remainder had APS associated with systemic lupus erythematosus (37%) or other autoimmune diseases (8%), such as undifferentiated connective tissue disease, rheumatoid arthritis or microscopic polyangiitis.Overall, 70% (n=60) met the Sapporo criteria, whereas 34% (n=29) met the 2023 ACR/EULAR criteria. All fulfilling the 2023 criteria also met the Sapporo criteria, while 36% (n=31) met only the latter. Concordance between the two systems was weak to moderate (Cohen’s kappa=0.36, p5 years between the clinical event and aPL positivity, or unknown temporal relationship between these (n=20). Other causes included low-titer aPL (n=2), loss of lupus anticoagulant positivity on repeat testing (n=3) and insufficient obstetric events (n=1). Despite not fulfilling classification criteria, 81% were on anticoagulants and 15% on antiplatelet agents.Conclusions These findings show only fair agreement between the two criteria, suggesting the new set identifies a narrower subset. The apparent 100% specificity reflects overlap with clinical diagnosis (both rely on similar features). As classification serves research purposes, clinical judgment remains essential, since patients not meeting criteria may still require antithrombotic therapy.