Distinguishing acute lymphoblastic leukemia from juvenile idiopathic arthritis in children with musculoskeletal complaints: a retrospective cross-sectional study

To identify routinely available clinical and laboratory features that differentiate acute lymphoblastic leukemia (ALL) from juvenile idiopathic arthritis (JIA) among children presenting with musculoskeletal complaints and no peripheral blood blasts at initial evaluation. A retrospective cross-sectional study was conducted among hospitalized children younger than 16 years at Bahrami Children’s Hospital (Tehran, Iran) between 2013 and 2023. Eligible children presented with musculoskeletal symptoms and were subsequently diagnosed with ALL based on bone marrow examination with immunophenotyping or with JIA by a pediatric rheumatologist according to International League of Associations for Rheumatology criteria. Children who received systemic corticosteroids before a definitive diagnosis, had blasts on the peripheral blood smear, or had > 30% missing data were excluded. Demographic characteristics, symptoms, physical examination findings, and initial laboratory results were extracted from the first hospital presentation. Multivariable logistic regression was used to identify independent predictors, with bootstrap internal validation. Sixty-six children met the inclusion criteria (29 ALL; 37 JIA). Symptom duration before diagnosis was markedly shorter in ALL than in JIA (median, 3 vs. 90 days). Systemic features were more frequent in ALL, including fever (83% vs. 27%), weakness (45% vs. 5%), and weight loss (35% vs. 5%). Bone pain was more common in ALL (93% vs. 35%), whereas objective arthritis—particularly knee involvement—was uncommon (14% vs. 87%); absence of clinical arthritis was frequent (62% vs. 3%). Splenomegaly (41% vs. 3%) and hepatomegaly (38% vs. 0%) occurred more often in ALL. Laboratory testing showed lower hemoglobin and platelet counts in ALL (median hemoglobin 7.5 vs. 11.7 g/dL; platelets 75 × 10³ vs. 317 × 10³/µL), with lower neutrophil counts and higher inflammatory markers (ESR and CRP) and LDH levels. In multivariable analysis, lower hemoglobin concentration and shorter symptom duration were independently associated with ALL; discrimination was 0.994 (bootstrap-corrected 0.990), as assessed by the area under the receiver operating characteristic curve. In children with musculoskeletal complaints and no peripheral blood blasts, rapid symptom onset accompanied by systemic features, predominant bone pain, organomegaly, cytopenias, and elevated LDH strongly suggests ALL rather than JIA. Because inflammatory markers may also be elevated in ALL, these features should prompt early evaluation to exclude leukemia before initiating corticosteroids or other immunosuppressive therapy.