Serial Next-Generation Sequencing for Detection of Germline Predisposition in Acute Myeloid Leukemia

Somatic next-generation sequencing (NGS) panel are now routinely performed at diagnosis of acute myeloid leukemia (AML) to identify genetic variants for risk stratification and treatment decisions. The variant allele frequencies (VAF) suspected to be germline variants remain persistently high in complete remission (CR) state, allowing these variants to be identified as germline variants. The NGS panel used clinically for diagnosis includes representative genes associated with germline predisposition to myeloid neoplasms. Based on above information, we aimed to investigate the frequency of germline variants and their prognostic significance of AML in patients with matched samples at diagnosis and at CR state. We conducted a prospective multicenter study (KCT0004825) collecting serial samples from AML patients. The process of interpreting germline variants is performed according to the American College of Medical Genetics and Genomics (ACMG) variant interpretation guidelines. We classified germline variants as causality if they were either ‘Pathogenic’ or ‘Likely Pathogenic’ by the ACMG guideline or if they were accompanied by a somatic mutation with ‘Variants of Uncertain Significance’. This study enrolled 603 patients diagnosed with treatable AML, excluding acute promyelocytic leukemia, between Feb. 2017 and Jun. 2021 from 17 institutions. Among these, paired CR samples were available for 343 patients who were eligible for NGS at diagnosis and after chemotherapies. We detected germline variants in 19 patients (5.5%). The most frequent germline mutations were in DDX41 (n=11, 58%) followed by DNAH5 (n=3, 16%), CEBPA and TP53 (both n=2, 11%), and MPL and GATA2 (both n=1, 5%). One patient harbored both TP53 and DNAH5 germline mutations. The mean VAF at diagnosis and CR samples were 48.9% [inter-quantile range (IQR): 47.8%-49.9%] and 48.0% (IQR: 47.6%-49.7%) respectively. Compared with non-germline mutated patients, patients with germline mutations were more hypoplastic (50% vs 80%, P < 0.001), and low bone marrow blast counts (34% vs. 65%, P=0.004). Germline mutations were not observed in the favorable cytogenetic risk group (0% vs. 23.1%, P= 0.03), and were rarely detected in 2022 European LeukemiaNet favorable risk group (5.3% vs. 47.5%, P <0.001), compared with non-germline mutated patients. Intensive chemotherapy was more frequently administered in the non-germline mutated group (P=0.022) but, there was no difference in the rate of allogeneic transplants between the two groups (P>0.05). Comparing survival rates, there was no difference between two groups that received intensive chemotherapy, nor was there a difference for those who underwent allogeneic transplantation (all, P>0.05). In conclusion, NGS analysis using matched CR samples proves to be an invaluable tool in detecting germline mutations. Since patients with germline mutations are rarely classified into the favorable risk category, most patients in this group should consider allogeneic transplantation. Although patients with germline mutations did not show significant prognostic differences in treatment outcomes due to the small number of patients, more studies involving a larger cohort are needed to determine the role of allogeneic transplantation.