Preliminary Decision Tree May Help Differentiate Acute Lymphoblastic Leukemia From Juvenile Idiopathic Arthritis

Hepatomegaly, splenomegaly, or lymphadenopathy are the most important features in distinguishing acute lymphoblastic leukemia (ALL) from juvenile idiopathic arthritis (JIA), according to research results published in Pediatric Rheumatology. In addition, the presence of bone pain, lowered neutrophils and platelets, and other general symptoms should prompt further evaluation.

Researchers conducted a multicenter, case-control study to identify the early clinical and laboratory features of ALL in children with isolated, persistent, osteoarticular pain to assess whether these criteria might efficiently distinguish between ALL and JIA diagnoses.

Study participants included children aged <18 years who were diagnosed with ALL based on bone marrow examination results after at least 1 month of persistent bone or joint pain, with a secondary diagnosis of JIA. Patients were enrolled in the study between October 2000 and March 2011 at multiple centers in France.

The total cohort included the data of 49 patients with ALL and 98 with JIA, of which the most frequent forms were oligoarthritis (43.9%), polyarthritis (29.6%), and enthesitis-related arthritis (18.4%). Among the 98 patients with JIA, 5.1% had systemic arthritis and 3.1% had psoriatic arthritis. Patients with ALL had B-form; 9 were treated by a rheumatologist, 4 had been treated for oligoarticular or polyarticular JIA before they were diagnosed with ALL, and 1 had arthritis at ALL diagnosis.

Median age range across both groups was similar, although more boys had ALL and more girls had JIA; median time to diagnosis was 2 times shorter in the ALL group.

Clinical examination at diagnosis indicated that only 1 patient in the ALL group had arthritis and only 2 in the JIA group had hepatomegaly, splenomegaly, or lymphadenopathy. In the JIA group, 18 patients had no arthritis at the time of initial physical examination; according to the investigators, the JIA diagnoses were made because the patients had medical histories consistent with JIA even if their arthritis had been resolved at diagnosis.

Prior to diagnosis, no differences were noted in the lymphocyte rate between the 2 groups. Patients in the ALL group had lower neutrophils, eosinophils, basophils, monocytes, platelets, hemoglobin, and hematocrit (P ≤.001), but a higher mean corpuscular volume (P =.009). Patients with ALL had higher rates of elevated C-reactive protein and erythrocyte sedimentation levels. At the time of diagnosis, these trends were the same, with the exception of lymphocytes, which were lower in patients with ALL compared with patients with JIA.

Results of a multivariate analysis demonstrated that the presence of hepatomegaly, splenomegaly, or lymphadenopathy were significantly associated with ALL (regression coefficient, 5.0; odds ratio [OR], 154; 95% CI, 30-793; P <.001). Similarly, nonarticular bone pain and fever were also associated with ALL (regression coefficients, 2.6 and 1.5; OR, 13.6, 95% CI, 3.2-57.4 and OR, 4.5, 95% CI, 1.9-10.5; P =.001 and P =.01, respectively).

When data relative to hepatomegaly, splenomegaly, or lymphadenopathy were excluded, nonarticular bone pain, general symptoms (including asthenia, anorexia, or weight loss), neutrophil counts, and platelet counts were associated with ALL compared with JIA.

Study limitations included the retrospective nature of the study, the small sample size, and the definition of the control group, which investigators noted could have been expanded to consider all children with chronic osteoarticular pain who may have had ALL, JIA, or other pathologies.

“Based on our findings, we propose a preliminary decision tree that could be tested in prospective studies,” the researchers concluded.

Reference

Louvigné M, Rakotonjanahary J, Goumy L, et al; on behalf of the GOCE group. Persistent osteoarticular pain in children: early clinical and laboratory findings suggestive of acute lymphoblastic leukemia (a multicenter case-control study of 147 patients). Pediatr Rheumatol Online J. 2020;18(1):1.