Pilot Study Analysis of Serum Cytokines to Differentiate Pediatric Septic Arthritis and Transient Synovitis

Background: In pediatric patients, the presentation of the nontraumatic acutely painful joint/limb poses a diagnostic dilemma due to the similarity of presentations of the most likely diagnoses [septic arthritis (SA), transient synovitis (TS), osteomyelitis]. Current tools employed to differentiate these diagnoses rely on nonspecific inflammatory markers, radiologic imaging, and arthrocentesis. Diagnostic algorithms utilizing these clinical, radiographic, and biochemical parameters have produced conflicting results. The purpose of this study was to identify a serum-based inflammatory signature which can differentiate SA from TS in pediatric patients. Methods: Serum samples were collected from 22 pediatric patients presenting with joint/extremity pain whose working diagnosis included SA or TS. Each sample was analyzed for serum abundance of 72 distinct biomarkers and cytokines using enzyme linked immunosorbent assay based arrays. Linear discriminant analysis was performed to identify a combinatorial biomarker panel to predict a diagnosis of SA or TS. Efficacy of the biomarker panel was compared with definitive diagnoses as based on laboratory tests, arthrocentesis results, and clinical scenario. Results: At the time of presentation: (1) mean erythrocyte sedimentation rate in the SA group was 56.6 mm/h and 12.4 mm/h in the TS group (P<0.001), (2) mean C-reactive protein was 55.9 mg/dL in the SA group and 13.7 mg/dL in the TS group (P=0.12), and (3) mean white blood cell was 10.9 k/mm(3) in the SA group and 11.0 k/mm(3) in the TS group (P=0.95). A combined panel of 72 biomarkers was examined using discriminant analysis to identify a limited set of predictors which could accurately predict whether a patient was diagnosed with SA or TS. A diagnostic algorithm consisting of transforming growth factor alpha, interleukin (IL)-7, IL-33, and IL-28A serum concentration correctly classified 20 of the 22 cases with a sensitivity and specificity of 90.9% (95% confidence interval: 73.9%-100.0%). Conclusion: This study identifies a novel serum-based 4-cytokine panel that accurately differentiates SA from TS in pediatric patients with joint/limb pain.