Acute Myeloid Leukemia (AML) is a heterogeneous disease with poor overall five-year survival of less than 30%. Current risk stratification is largely based on cytogenetics, combined with information of the most commonly mutated genes in AML (e.g. NPM1, FLT3, DNMT3A). To improve clinical decision making and to increase our understanding of the mechanisms underlying AML it is essential to gain additional information about the mutational landscape of AML.

In this prospective study we perform comprehensive Next Generation Sequencing (NGS) to determine the mutational landscape of AML. Starting from September 2014, bone marrow samples, with matched skin biopsies, were collected from all newly diagnosed samples of AML at Skåne University Hospital, Sweden. So far, almost 40 AML samples have undergone whole-exome sequencing (WES) (100X coverage), targeted AML-gene panel sequencing (>100 genes with recurrent mutations in the TCGA AML data set) (400X), RNA-seq and low pass Whole Genome Sequencing (WGS) (1.5X). Additionally, clinical data such as age, treatment response and survival outcome are collected and samples are also cryopreserved for functional follow-up studies. The targeted AML-panel sequencing allows for high coverage data enabling identification of not only common but also rare variants present in subclones, while WES might reveal genes and pathways not previously associated with AML. Low pass WGS enables the detection of cytogenetic alterations, ranging from larger structural rearrangements to fusion gene detection. RNA-seq also makes the detection of fusion genes possible as well as providing global gene expression data.

So far our prospective study has identified 22 recurrently mutated genes (with mutations present in >5% of the reads). Out of these, DNMT3A (34%), NPM1 (29%), TET2 (21%), FLT3 (18%) and RUNX1 (18%) were the most commonly mutated genes. The corresponding mutation frequencies in TCGA AML data set are DNMT3A (26%), NPM1 (27%), TET2 (9%), FLT3 (28%) and RUNX1 (10%). More than 70% of the cases carry combinations of mutations in two up to seven of the genes included in our AML panel. Each patient also carries a private combination of unique exomic variants. RNA-seq data confirmed all clinically known fusion genes and principal component analysis revealed that cases with e.g. NPM1 mutations have a uniform gene expression pattern.

Although diagnostics has improved over the last years, information of the most commonly mutated genes has not largely improved risk stratification. A plausible explanation is the clonal complexity in AML and the joint risk combination of common and rare variants. NGS-based methods have greatly improved our possibility to detect genetic alterations and comprehensive NGS of AML has the potential to identify mutational patterns that can further improve diagnostics and risk stratification.

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution