[Establishment and validation of prognosis predictive model using m⁶A RNA methylation regulators in children acute myeloid leukemia]

Objective: To figure out the association between the expression of m⁶A RNA methylation regulators and the prognosis of children AML, and provide genetic markers for monitoring the progression and recurrence of AML. Methods: Twenty two m⁶A RNA methylation regulators were firstly analyzed using the data from Therapeutically Applicable Research To Generate Effective Treatments(TARGET) database and The Genotype-Tissue Expression(GTEx) database, Wilcoxon rank test was performed to analyze the differentially expression of m⁶A RNA methylation regulators between the AML and normal tissue, 296 AML children were divided into training cohort and validation cohort by simple random sampling method, Lasso regression was used to screen out the risk factors and the multivariate Cox regression was applied for establishing prognosis predicting model in training cohort. Kaplan-Meier survival curve, time-dependent ROC curve and multivariate Cox regression were used to estimate the efficiency of risk score calculated by predictive model in validation cohort. Results: Twenty one m⁶A genes were up regulated in AML compared to Normal patients. Five m⁶A RNA methylation regulators(ZC3H13, YTHDC2, HNRNPA2B1, METTL3, METTL5) were included in final predicting model. Risk score could independently predict the survival of AML patients in training cohort(HR:2.72, 95%CI: 1.54-4.81, P=0.000 6) and validation cohort(HR:2.01, 95%CI:1.14-3.50, P=0.016). Low-risk patients had better prognoses than high-risk patients both in training cohort(P=0.001 9) and validation cohort(P=0.023). Conclusion: This prognosis predicting model constructed by m⁶A RNA methylation regulators could independently predict the survival prognosis in AML children, and should be helpful for clinical therapy.