Search

Acute myeloid leukemia (AML) is a heterogeneous disease. Even within the same NPM1-mutated genetic subgroup, some patients harbor additional mutations in FLT3, IDH1/2, DNMT3A or TET2. Recent studies have shown the prognostic significance of minimal residual disease (MRD) in AML but it remains to be determined which molecular markers are the most suitable for MRD monitoring. Recent advances in next-generation sequencing (NGS) have provided the opportunity to use multiple molecular markers. In this study, we used NGS technology to assess MRD in 31 AML patients enrolled in the ALFA-0701 trial and harboring NPM1 mutations associated to IDH1/2 or DNMT3A mutations. NPM1 mutation-based MRD monitoring was performed by RTqPCR. IDH1/2 and DNMT3A mutations were quantified by NGS using an Ion Torrent Proton instrument with high coverage (2 million reads per sample). The monitoringof IDH1/2 mutations showed that these mutations were reliable MRD markers that allowed the prediction of relapse in the majority of patients. Moreover, IDH1/2 mutation status predicted relapse or disease evolution in 100% of cases if we included the patient who developed myelodysplastic syndrome. In contrast, DNMT3A mutations were not correlated to the disease status, as we found that a preleukemic clone with DNMT3A mutation persisted in 40% of the patients who were in complete remission, reflecting the persistence of clonal hematopoiesis.

Minimal Residual Disease (MRD) detection can be used for early intervention in relapse, risk stratification, and treatment guidance. FLT3 ITD is the most common mutation found in AML patients with normal karyotype. We evaluated the feasibility of NGS with high coverage (up to 2.4.10(6) PE fragments) for MRD monitoring on FLT3 ITD. We sequenced 37 adult patients at diagnosis and various times of their disease (64 samples) and compared the results with FLT3 ITD ratios measured by fragment analysis. We found that NGS could detect variable insertion sites and lengths in a single test for several patients. We also showed mutational shifts between diagnosis and relapse, with the outgrowth of a clone at relapse different from that dominant at diagnosis. Since NGS is scalable, we were able to adapt sensitivity by increasing the number of reads obtained for follow-up samples, compared to diagnosis samples. This technique could be applied to detect biological relapse before its clinical consequences and to better tailor treatments through the use of FLT3 inhibitors. Larger cohorts should be assessed in order to validate this approach.