CGE19-065: Number of Somatic Mutations Is an Independent Predictor of Overall Survival in Acute Myeloid Leukemia

Abstract

Purpose: Acute myeloid leukemia (AML) is characterized by multiple somatically acquired mutations that affect genes of different functional categories. It has been well established in myelodysplastic syndrome (MDS) that the cumulative number of somatic mutations has an impact on overall survival. However, no such data exist for AML. In this study, we sought to determine the number of clinically significant somatic mutations for each cytogenetically defined risk group of AML and to determine whether this had an impact on overall survival (OS). Methods: In this retrospective, single-center study, all adult patients diagnosed with AML from August 2016–December 2017 were reviewed. Baseline characteristics, somatic mutations in the diagnostic bone marrow as detected by Next Generation Sequencing (NGS), and survival outcomes were analyzed. NGS panel was done in-house and could identify 94 genes. Patients were divided into favorable, intermediate, and poor risk groups based on cytogenetics, and molecular abnormalities using NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines) for AML, version 1.2018. Kaplan-Meier plots and Cox regression analyses were utilized. Results: A total of 105 AML patients were included; baseline characteristics and frequency of identified clinically significant (CS) mutations are described in the presentation. The FLT3 mutation occurred in the highest frequency (22%) followed by DNMT3A & ASXL1 (15%). 17 (16%) patients were favorable risk, 33 (31%) intermediate risk, and 55 (52%) were poor risk. 67.6% of patients were male, and the median age was 64 (20–79) years. There was a difference in the number of CS mutations between the intermediate risk group and favorable risk group (P=.007), but not between the favorable risk and poor risk groups (P=.221) or between the intermediate risk group and poor risk group (P=.093). Increased number of CS mutations (≥ 5) was seen with equal frequency across risk groups and predicted for shorter overall survival in both univariate (HR=2.80; P=.039) and by multivariate Cox regression analysis (P=.001) independently from assigned risk group. There were no differences in age, gender, smoke, geographic, and different risk groups by multivariate analyses. Conclusion: Our study shows that ≥ 5clinically significant somatic mutations were associated with adverse outcomes and decreased survival, independent of risk groups and induction regimen. Thus, it may be a useful prognostic factor. This finding needs to be validated using a larger sample size.