Performance of current guidelines for diagnosis of macrophage activation syndrome complicating systemic juvenile idiopathic arthritis

Abstract

Results The study sample included 362 patients with systemic JIA and MAS, 404 patients with active systemic JIA without MAS, and 345 patients with systemic infection. The best capacity to differentiate MAS from systemic JIA without MAS was found when the preliminary MAS guidelines were applied. The 3/5-adapted HLH-2004 guidelines performed better than the 4/5-adapted guidelines in distinguishing MAS from active systemic JIA without MAS. The 3/5-adapted HLH-2004 guidelines and the preliminary MAS guidelines with the addition of ferritin levels ≥500 ng/ml discriminated best between MAS and systemic infections. Conclusion The preliminary MAS guidelines showed the strongest ability to identify MAS in systemic JIA. The addition of hyperferritinemia enhanced their capacity to differentiate MAS from systemic infections. The HLH-2004 guidelines are likely not appropriate for identification of MAS in children with systemic JIA. Objective To compare the capacity of the 2004 diagnostic guidelines for hemophagocytic lymphohistiocytosis (HLH-2004) with the capacity of the preliminary diagnostic guidelines for systemic juvenile idiopathic arthritis (JIA)-associated macrophage activation syndrome (MAS) to discriminate MAS complicating systemic JIA from 2 potentially confusable conditions, represented by active systemic JIA without MAS and systemic infection. Methods International pediatric rheumatologists and hemato-oncologists were asked to retrospectively collect clinical information from patients with systemic JIA-associated MAS and confusable conditions. The ability of the guidelines to differentiate MAS from the control diseases was evaluated by calculating the sensitivity and specificity of each set of guidelines and the kappa statistics for concordance with the physician's diagnosis. Owing to the fact that not all patients were assessed for hemophagocytosis on bone marrow aspirates and given the lack of data on natural killer cell activity and soluble CD25 levels, the HLH-2004 guidelines were adapted to enable the diagnosis of MAS when 3 of 5 of the remaining items (3/5-adapted) or 4 of 5 of the remaining items (4/5-adapted) were present.