BackgroundPrevention of relapse for adults with Acute Myeloid Leukemia (AML) in first remission (CR1) is the most common indication for allogeneic hematopoietic cell transplantation (alloHCT). Measurable residual disease (MRD) detection has been associated with higher relapse rates, but testing is not standardized. Establishment of validated AML MRD criteria would allow harmonization across centers facilitating clinical trials and generalizable practice guidelines.MethodsPatients aged 18 or older who underwent first alloHCT forFLT3, NPM1, IDH1, IDH2and/orKITmutated AML in CR1 were eligible for this multicenter study. Residual mutations detected in remission using ultra-deep error-corrected next-generation DNA-sequencing (NGS-MRD) associated with increased relapse risk were validated in a second independent cohort. The impact of baseline characteristics on NGS-MRD results was evaluated.ResultsPre-conditioning CR1 blood samples from 1075 patients were tested. Detection of residualNPM1and/orFLT3-ITD mutations before alloHCT was associated with worse outcomes after transplantation in both discovery (patients transplanted 2013-2017) and validation (patients transplanted 2018-2019) cohorts. In multivariate Cox regression analysis detection of residualNPM1m (relapse; HR: 4.9, 3.5-7.0, P<0.0001, survival; HR: 2.9, 2.2-4.0, P<0.0001), residualFLT3-ITD (relapse: HR: 4.2, 3.1-5.7, P<0.0001, survival; HR: 2.6, 2.0-3.6, P<0.0001), and receipt of non-melphalan containing reduced intensity conditioning were associated with inferior clinical outcomes.ConclusionsDetection of residualNPM1and/orFLT3-ITD mutations in the blood of patients undergoing alloHCT for AML in CR1 is associated with increased relapse and worse survival. In those undergoing reduced intensity conditioning, melphalan may reduce the risk associated with testing NGS-MRD positive prior to transplantation.