Background:Mutations in DNA and chromatin regulation genes have been associated with clonal hematopoiesis (CH) in older healthy subjects, but are also involved in leukemogenesis occurring as early events. In AML patients, the prognostic role of CH‐mutations detected at diagnosis has been assessed mainly by investigating single mutations, with diverging results. Thus, the significance of pre‐treatment CH‐mutations as a group remains to be elucidated.Aims:To investigate the distribution and clinical impact of CH‐mutations detected at diagnosis in AML patients with normal karyotype and treated with chemotherapy eventually followed by allogeneic hematopoietic stem cell transplant (alloHSCT) within the prospective NILG trial 02/06 [ClinicalTrials.gov Identifier: NCT00495287].Methods:Out of 572 newly diagnosed AML patients enrolled into the trial, 270 had a normal karyotype. For 213 of them, a molecular profile was obtained at diagnosis with standard approach and targeted NGS on peripheral blood and bone marrow samples. CH‐mutations were defined as DNMT3A, TET2, ASXL1, IDH1 and IDH2. Based on the available data, 209 patients were re‐classified as per the 2017 European Leukemia Net (ELN) guidelines. All patients were treated with intensive chemotherapy eventually followed by consolidative alloHSCT.Results:At least one CH‐mutation was reported for 131 out of 213 evaluable patients (61.5%), while 82 patients had no CH‐mutations (38.5%). Patients carrying CH‐mutations were older in comparison with patients without CH‐mutations (median age 53 vs 47 years, P = 0.0003). No preferential association was found in respect of AML type (i.e. de novo vs secondary AML, P = 0.25). The most recurrent CH‐mutation was DNMT3A (83/131), followed by TET2 (32/131), IDH2 (26/131), IDH1 (20/131) and ASXL1 (18/131). A single CH‐mutation was observed in 65% of patients, while 33.5% had a double mutation and 1.5% a triple mutation. Overall, CH‐mutations were detected across all ELN risk groups. The likelihood of carrying a CH‐mutation was significantly increased in patients with a NPM1 mutation (P < 0.0001), while the opposite was reported for patients with a biallelic CEBPa mutation (P < 0.0001). Taken together, CH‐mutations did not affect the 5‐years disease free survival (DFS) and overall survival (OS) (46% vs 54%, P = 0.45; and 50% vs 55%, P = 0.21). However, within the ELN risk categories, the presence of CH‐mutations negatively affected the 5‐years DFS (86% vs 59%, P = 0.03) and OS (86% vs 71%, P = 0.09) in patients in the favorable risk group, while these outcomes were not affected in the intermediate and adverse risk groups. By analyzing each CH‐mutation separately, TET2 emerged as an independent risk factor for the achievement of CR [OR 0.25 (95% CI 0.08–0.7), P = 0.008], but not for OS [OR 1.56 (95% CI 0.89–2.75), P = 0.12]. Finally, 186 patients who received a consolidative alloHSCT in first CR were investigated, 107 of them carrying CH‐mutations. AlloHSCT provided a significant benefit in terms of OS for the whole study population (P = 0.001), irrespective of the presence of CH‐mutations (P = 0.87).Summary/Conclusion:CH‐mutations are frequently detected at diagnosis in AML patients with normal karyotype across all ELN risk groups, but more likely in NPM1 mutated patients. Overall, the presence of CH‐mutations is not associated with worse outcomes except for patients within the ELN favorable risk group. These findings suggest that an active search of these mutations might refine the risk of relapse in patients considered at favorable risk, while their role is limited in patients with more aggressive disease profile.image
Read full abstract