Delayed Early MRD Clearance and Outcomes in Pediatric Acute Myeloid Leukemia: A Retrospective Cohort Study

Abstract Background: Measurable residual disease (MRD) is an established prognostic marker in pediatric acute myeloid leukemia (AML), but whether the trajectory of MRD clearance provides prognostic information beyond MRD status at a later treatment time point remains uncertain. Methods: We conducted a retrospective cohort study of 418 children with newly diagnosed AML in the TARGET-AML database who had paired quantitative multiparameter flow cytometry-based MRD measurements at the end of induction 1 (EOI1) and EOI2. Using a 0.05% threshold, patients were classified as Neg/Neg, Pos/Neg, or Pos/Pos according to EOI1/EOI2 MRD status; 7 patients with a Neg/Pos pattern were included in the Pos/Pos group because of the small subgroup size. The primary outcome was 5-year cumulative incidence of relapse (CIR), with non-relapse mortality as a competing event. Results: Compared with Neg/Neg, Pos/Neg was associated with higher CIR after adjustment for age, ELN 2022 risk, white blood cell count, and treatment protocol (subdistribution hazard ratio, 1.66; 95% CI, 1.14–2.42; P=.008). Pos/Neg was also associated with inferior overall survival (hazard ratio, 2.03; P=.003) and event-free survival (hazard ratio, 1.89; P<.001). Among patients who were MRD-negative at EOI2, Pos/Neg remained associated with higher CIR (subdistribution hazard ratio, 1.70; 95% CI, 1.16–2.47; P=.006). Continuous ΔMRD did not improve time-dependent discrimination beyond categorical MRD kinetics. Conclusions: Early MRD trajectory was associated with relapse risk beyond EOI2 MRD status alone. Categorical MRD clearance pattern, but not continuous ΔMRD, may refine risk stratification in pediatric AML. Prospective studies are warranted to evaluate longitudinal MRD assessment for risk stratification in pediatric AML.