Identifying acute lymphoblastic leukemia mimicking juvenile idiopathic arthritis in children

OBJECTIVE: Acute lymphoblastic leukemia (ALL) may present with arthritis implying the risk of being misdiagnosed as juvenile idiopathic arthritis (JIA). The aim of this study was to identify predictors for ALL based on clinical and laboratory information. METHODS: This cross-sectional, retrospective study compared clinical presentation and laboratory results of 26 children with ALL and arthritis versus 485 children with JIA (433 non-systemic, 52 systemic JIA). Using a Bayesian score approach the findings were evaluated by calculating odds ratios (OR) and lnOR as a measure of diagnostic weight. RESULTS: Distinction on clinical grounds was difficult, as even a high number of joints involved did not exclude ALL. One or more hematologic cell counts were low (Hb <10 g/dL, platelet count <100 x 10(9)/L, neutrophil count < 1.0 x 10(9)/L) in 92% with ALL, 25% with systemic JIA and 10% with non-systemic JIA. Neutropenia and thrombocytopenia had the highest ORs of 128 (95% CI 43–387) and 129 (95% CI 26–638), each giving a diagnostic weight of 4. The estimated risks of ALL were 0.2% with normal cell counts and 9%, 67% and 100% when one, two or three cell lines were affected. CONCLUSION: A simple count of cell lines with low counts can serve as a basic diagnostic strategy. Children with tri- or bilinear involvement should be referred to a bone marrow, and those with unilinear involvement a thorough screen for further evidence of ALL (organomegaly, ESR, LDH, uric acid, and blood smear).