Prognostic Impacts of D816V KIT Mutation and Peri-Transplant RUNX1-RUNX1T1 MRD Monitoring on Acute Myeloid Leukemia with RUNX1-RUNX1T1

Abstract

The prognostic significance of KIT mutations and optimal thresholds and time points of measurable residual disease (MRD) monitoring for acute myeloid leukemia (AML) with RUNX1-RUNX1T1 remain controversial in the setting of hematopoietic stem cell transplantation (HSCT). We retrospectively evaluated 166 high-risk patients who underwent allogeneic (Allo-HSCT, n = 112) or autologous HSCT (Auto-HSCT, n = 54). D816V KIT mutation, a subtype of exon 17 mutations, was significantly associated with post-transplant relapse and poor survival, while other types of mutations in exons 17 and 8 were not associated with post-transplant relapse. Pre- and post-transplant RUNX1-RUNX1T1 MRD assessments were useful for predicting post-transplant relapse and poor survival with a higher sensitivity at later time points. Survival analysis for each stratified group by D816V KIT mutation and pre-transplant RUNX1-RUNX1T1 MRD status demonstrated that Auto-HSCT was superior to Allo-HSCT in MRD-negative patients without D816V KIT mutation, while Allo-HSCT was superior to Auto-HSCT in MRD-negative patients with D816V KIT mutation. Very poor outcomes of pre-transplant MRD-positive patients with D816V KIT mutation suggested that this group should be treated in clinical trials. Risk stratification by both D816V KIT mutation and RUNX1-RUNX1T1 MRD status will provide a platform for decision-making or risk-adapted therapeutic approaches.

Keywords: AML; D816V KIT mutation; RUNX1–RUNX1T1; hematopoietic stem cell transplantation; measurable residual disease.

Figure 1

Cumulative incidence of relapse according to KIT mutations (A–C) or D816V KIT mutation (D–F) in all patients (A,D) or patients in Allo-HSCT (B,E) and Auto-HSCT (C,F) groups.

Figure 2

RUNX1–RUNX1T1 expression kinetics during peri-transplant period and optimal time points. (A) RUNX1–RUNX1T1 expression kinetics according to transplant type (Allo-HSCT vs. Auto-HSCT) before HSCT and at 1 or 3 months after HSCT (numbers of Allo-HSCT vs. Auto-HSCT; before, 112 vs. 54; 1 month, 70 vs. 13; 3 months, 86 vs. 16). (B) RUNX1–RUNX1T1 expression kinetics according to the occurrence of relapse (numbers of relapsed patients vs. non-relapsed patients; before, 21 vs. 145; 1 month, 12 vs. 71; 3 months, 11 vs. 91). (C) Receiver operating characteristic (ROC) curve analysis of RUNX1–RUNX1T1 levels at each time point. RUNX1–RUNX1T1 levels were normalized to the number of ABL1 transcripts and expressed as copy numbers per 10⁵ copies of ABL1. Results are expressed as mean ± SEM. * p < 0.05, ** p < 0.01. AUC indicates area under curve.

Figure 3

Survival outcomes according to MRD positivity defined by 3 log reduction in RUNX1–RUNX1T1 levels at pre-HSCT (A) and at 1 month (B) or 3 months (C) after HSCT.

Figure 4

Cumulative incidence of relapse (CIR) according to both D816V KIT mutation and MRD positivity defined by 3 log reduction in RUNX1–RUNX1T1 levels at pre-HSCT (A) and effects of transplant types on CIR in each stratified group (B–E).