Fig. 3

CircRNF220 had potential diagnostic value and was related to the prognosis of pediatric AML patients. A The expression level of circRNF220 in patients with different subtypes of pediatric AML from M1 to M7. Every subtype was compared to the healthy control group. B Receiver operating characteristic (ROC) analysis for circRNF220 in the BM of AML patients. The area under the ROC curve (AUC) for distinguishing AML patients from healthy controls and patients with other hematological diseases was 0.9260. C qRT-PCR validation of circRNF220 expression in PB samples from patients with pediatric AML. D ROC analysis of circRNF220 in AML patient PB samples. The AUC for distinguishing patients with AML from normal controls was 0.9601. E CircRNF220 expression was measured by qRT-PCR in samples from 187 AML patients with different stages of disease (P, n = 87; NR, n = 20; CR, n = 75; Re, n = 5). F The average expression levels of circRNF220 before and after therapy (n = 13) in the paired samples from pediatric AML patients. G A total of 87 primary AML patient samples were divided according to the clinical outcomes of the corresponding patients. Patients who relapsed had significantly higher circRNF220 expression levels than those who achieved CR after therapy. H Relapse-free survival (RFS) curves for a cohort of 87 AML patients. The patients were dichotomized based on the median expression level of circRNF220 at diagnose. Statistical differences between the curves were calculated by using the log-rank test, and the two-sided P-value is indicated below the graph. HC, healthy control; BM, bone marrow; PB, peripheral blood; P, primary; NR, nonremission; CR, complete remission; Re, relapse; P (CR), primary patient with complete remission; P (Re), primary patient with relapse. * P < 0.05, **P < 0.01, *** P < 0.001. ns: not significant