Profiling of somatic mutations and fusion genes in acute myeloid leukemia patients with FLT3-ITD or FLT3-TKD mutation at diagnosis reveals distinct evolutionary patterns

Abstract

Background: The receptor tyrosine kinase FLT3 with internal tandem duplications within the juxtamembrane domain (FLT3-ITD) is a poor prognostic factor; however, the prognostic significance of missense mutation in the tyrosine kinase domain (FLT3-TKD) is controversial. Furthermore, the accompanying mutations and fusion genes with FLT3 mutations are unclear in acute myeloid leukemia (AML).

Methods: We investigated FLT3 mutations and their correlation with other gene mutations and gene fusions through two RNA-seq based next-generation sequencing (NGS) method and prognostic impact in 207 de novo AML patients.

Results: FLT3-ITD mutations were positive in 58 patients (28%), and FLT3-TKD mutations were positive in 20 patients (9.7%). FLT3-ITD was associated with a higher white blood cell count (WBC, mean 72.9 × 10⁹/L vs. 24.2 × 10⁹/L, P = 0.000), higher bone marrow blasts (mean 65.9% vs. 56.0%, P = 0.024), and NK-AML (normal karyotype) (64.8% vs. 48.4%, P = 0.043). NPM1 and DNMT3A mutations were enriched in FLT3-ITD (53.5% vs. 15.3%, P = 0.000; 34.6% vs. 13%, P = 0.003). However, the mutations of CEBPA were excluded in FLT3-AML (3.8% vs. 0% vs. 19.8%, P = 0.005). Mutations of Ras and TP53 were unlikely associated with FLT3-ITD (1.9% vs. 20.6%, P = 0.006; 0% vs. 6.1%, P = 0.04). The common fusion genes (> 10%) in FLT3-ITD had MLL-rearrangement and NUP98-rearrangement, while the common fusion genes in FLT3-TKD had AML1-ETO and MLL-rearrangement. Two novel fusion genes PRDM16-SKI and EFAN2-ZNF238 were identified in FLT3-ITD patients. Gene fusions and NPM1 mutation were mutually excluded in FLT3-ITD and FLT3-TKD patients. Their patterns of mutual exclusivity and cooperation among mutated genes suggest that additional driver genetic alterations are required and reveal two evolutionary patterns of FLT3 pathogenesis. Patients with FLT3-ITD had a lower CR (complete remission) rate, lower 3-year OS (overall survival), DFS (disease-free survival), and EFS (event-free survival) compared to FLT3_wtAML. NK-AML with FLT3-ITD had a lower 3-year OS, DFS, and EFS than those without, while FLT3-TKD did not influence the survival in whole cohort and NK-AML. Besides, we found that FLT3-ITD/TET2 bimutation defined a poor prognostic subgroup.

Conclusions: Our study offers deep insights into the molecular pathogenesis and biology of AML with FLT3-ITD and FLT3-TKD by providing the profiles of concurrent molecular alterations and the clinical impact of FLT3-ITD and FLT3-TKD on AML patients.

Keywords: Acute myeloid leukemia; FLT3-ITD; FLT3-TKD; Next-generation sequencing; TET2.

Fig. 1

OS (a), DFS (b), and EFS (c) curve of FLT3-TKD (n = 16), FLT3-ITD (n = 56), and FLT3 wild type (n = 132) AML patients; OS (d), DFS (e), and EFS (f) curve of normal karyotype AML patients with (n = 18) or without (n = 74) FLT3-ITD mutation; OS (h), DFS (i), and EFS (g) curve of FLT3-ITD AML patients with (n = 6) or without (n = 44) TET2 mutation

Fig. 2

Relationship between gene mutations and fusion genes of FLT3-ITD and FLT3-TKD AML. a, b Represent fusion genes by targeted NGS and its exclusive relationship with NPM1 mutation in FLT3-ITD positive AML (n = 60). c, d Represent fusion genes by targeted NGS and its exclusive relationship with NPM1 mutation in FLT3-TKD positive AML (n = 16)

Fig. 3

Distribution of somatic mutations and fusion genes in 82 AML patients with FLT3-ITD and FLT3-TKD. Each column displays an individual sample. White highlights in the top FAB subtype indicate that the information is not available (n.a.). Blue highlights indicate the presence of a gene mutation; grey highlights indicate wild-type status. CEBPA mutation is an allele double mutation in this panel. Mutated genes are clustered according to their pathways or family

Fig. 4

Circos of mutated genes and gene fusions in AML patients with FLT3 mutation. Ribbon widths are proportional to the frequency of a molecular event

Fig. 5

Molecular heterogeneity of AML exemplified by mutational and fusion genes profiling in FLT3-AML. Each spoke radiating from the central FLT3-ITD or FLT3-TKD hub represents the molecular pattern of a single patient. Cooperating mutations are grouped into three tiers according to the function and color-coded according to the figure key, and white space indicates no mutation or fusion. Overall, based on molecular combination, patients are segregated into different subgroups

Fig. 6

Schematic model for the two paths of evolution of FLT3 mutant AML. The first step is the occurrence of mutations or fusions, and the second step is the hit of FLT3-ITD or FLT3-TKD mutations