SAT0381 Neutrophil-related biomarkers discriminate septic arthritis from other inflammatory arthritides: data from the bossa trial

Background Septic arthritis is a life threatening purulent invasion of a joint by an infectious agent which produces arthritis. If untreated, septic arthritis causes structural damage to the joint. Unfortunately no relevant biomarkers are available for the diagnosis of this disabling condition. We aimed at determining whether calprotectin (S100A8/A9) and Human neutrophil alpha-defensins (HNP1–3) could discriminate septic from other inflammatory arthritides Objectives We aimed at determining whether calprotectin (S100A8/A9) and Human neutrophil alpha-defensins (HNP1–3) could discriminate septic from other inflammatory arthritides Methods Patients joint effusions for which septic arthritis was suspected were prospectively collected in Grenoble Hospital. Patients with inflammatory synovial fluid (i.e. with white blood cell >2000/mm3 and >80% polymorphonuclear neutrophils (PMN)) were included in this trial. Diagnosis of septic arthritis was retained if bacteria were cultured from inflammatory synovial fluid and/or blood samples. Diagnosis of pseudo gout was retained when pyrophosphate calcium crystals were observed in inflammatory synovial fluid. Diagnosis of rheumatoid arthritis was retained according to rheumatologist opinion. C Reactive protein (CRP), both neutrophil-related proteins calprotectin and human neutrophil alpha-defensins (HNP1–3) levels were assessed in synovial fluids. Threshold for biomarkers were determined by ROC curve analysis. Sensibility, Specificity, Positive (PPV) and Negative (NPV) Predictive Values at a pre specified threshold were calculated. Biomarkers with p value≤0.05 were included into a multivariate model. Multivariate logistic regression with stepwise selection was performed to build the final combined model. Results A total of 74 patients were included: septic arthritis (n=26), pseudo gout (n=28) and Rheumatoid arthritis (n=20). Patients with septic arthritis group were more likely to be male (69% vs. 31%, p=0.030), were younger (median age (range) 65.5 (27–84) vs. 72.5 (33–91), p=0.047)), displayed higher synovial fluid PMN count (9,600 (1,800–68000) vs. 6560 (750–22500, p=0.047))., LR+12.2 and LR- 0.26. Calprotectin was significantly increased in patients with septic arthritis (190 (0.1–247) vs. 62 (0–208) mg/L, p In the multivariate model, including calprotectin, HNP1–3, synovial fluid PMN count and gender, calprotectin was the only biomarker discriminating septic arthritis from non-septic inflammatory arthritides with 76% sensitivity, 94% specificity, 86% PPV and 88% NPV. Conclusions Our data show here that synovial fluid calprotectin is a relevant biomarker to discriminate septic arthritis from other inflammatory arthritides. This biomarker should be tested in an independent cohort. Acknowledgements The authors thank Sylvie Papacatzis for her help in the study. Disclosure of Interest None declared