Abstract
BACKGROUND: Outcomes of patients with t(8;21)(q22;q22) acute myeloid leukemia (AML) after allogeneic hematopoietic stem cell transplantation (allo-HSCT) remain heterogeneous. METHODS: To identify the risk factors for relapse and survival after allo-HSCT in t(8;21) AML patients, we retrospectively evaluated the clinical and prognostic information of 142 patients with t(8;21) AML undergoing allo-HSCT between January 2002 and September 2018 at 15 hematology research centers in China. RESULTS: Twenty-nine patients (20%) relapsed after undergoing allo-HSCT. A > 1-log reduction in RUNX1/RUNX1T1-based minimal residual disease (MRD) directly before allo-HSCT and a > 3-log reduction within the first 3 months after allo-HSCT were associated with a significantly lower post-transplant 3-year cumulative incidence of relapse (CIR, 9% vs. 62% and 10% vs. 47%,all P < 0.001), whereas transplantation during the second complete remission (CR2, 39% vs. 17% during CR1, P = 0.022), during relapse (62% vs. 17% during CR1, P < 0.001) and KIT D816 mutations at diagnosis (49% vs. 18%, P = 0.039) were related to a significantly higher 3-year CIR. Multivariate analysis demonstrated that a > 1-log reduction in MRD directly before transplantation (CIR: hazard ratio(HR), 0.21 [0.03-0.71], P = 0.029; overall survival (OS): HR = 0.27 [0.08-0.93], P = 0.038) and a > 3-log reduction in post-transplant MRD within the first 3 months (CIR: HR = 0.25 [0.07-0.89], P = 0.019; OS: HR = 0.38 [0.15-0.96], P = 0.040) were independent favorable prognostic factors, and transplantation during relapse (CIR: HR = 5.55 [1.23-11.56], P = 0.041; OS: HR = 4.07 [1.82-20.12], P = 0.045) were independent adverse prognostic factors for post-transplant relapse and survival in patients with t(8;21) AML. CONCLUSION: Our study suggests that for patients with t(8;21) AML undergoing allo-HSCT, it would be better to receive transplantation during CR1 with a MRD directly before transplantation achieving at least 1-log reduction. MRD monitoring in the first 3 months after allo-HSCT might be robust in predicting relapse and adverse survival after allo-HSCT.