Background/Purpose: The differential diagnosis of fever of unknown origin (FUO) is a major challenge in pediatrics especially for differentiation of systemic-onset juvenile idiopathic arthritis (SJIA) and infectious diseases. In a pilot study the analysis of MRP8/14 serum has been demonstrated as an excellent tool for the diagnosis of SJIA, allowing early differentiation of patients with autoinflammatory diseases SJIA or Familial Mediterranean Fever (FMF) versus those with other diseases with a specificity of 95%. Based on our study published in 2009, the analysis of MRP8/14 serum levels has been offered to paediatric rheumatologists nationwide in the last years. Therefore, we aimed to validate the findings from our pilot study in samples from daily clinical practice and to evaluate their relevance in clinical practice.

Methods: The study was designed as a retrospective analysis. The study group comprised 984 patients from 44 centers who presented with FUO. Patients were selected from our database between January 2009 and October 2012. Data collected included signs (laboratory parameters CRP, BSG, leucocytes) and symptoms as well as the final diagnosis made by the caring physicians. In all samples, concentration of MRP8/14 was determined by sandwich enzyme-linked immunosorbent assay (ELISA) with a cut-off of 9200 ng/ml (for SJIA versus other diseases). A questionnaire about the relevance of the MRP8/14 result for the final diagnosis was completed.

Results: Final diagnoses made by physicians were SJIA (n=301), FMF (n=135) and other inflammatory diseases (including infections, vasculitis and other autoinflammatory diseases) (n=548). MRP8/14 serum levels of patients with SJIA or FMF (10.090±1.930 ng/ml, mean±SEM) were elevated compared to other diagnoses (3.140±570 ng/ml) irrespectively of the presence of fever and anti-inflammatory treatment. In the group of untreated patients with fever (n=213) MRP8/14 levels of SJIA patients (18.685±4.130 ng/ml) were even higher compared to other diagnoses (5.285±1.535 ng/ml). In this group, the sensitivity and specificity of MRP8/14 to differentiate between patients with SJIA vs. other diseases (excluding FMF) were 75% and 89% respectively. The sensitivity of the test increased to 83% in the presence of fever, joint pain and CRP>1mg/dl. There was a significant correlation (p<0,001) between MRP8/14 and CRP or leucocyte counts in the total group of patients with SJIA. The clinicians reported that MRP8/14 results were helpful (56%), decisive (11%), not important (32%) and distracting (1%) in the total cohort.

Conclusion: Measurement of MRP8/14 levels is a helpful tool for the diagnosis of SJIA in FUO. In daily clinical practice the marker has a sensitivity of 75% and specificity of 89% to detect SJIA in untreated patients with fever. These results have to be replicated in an independent cohort and additional markers might be useful to improve the detection of SJIA in patients presenting with FUO.