Distinct age-associated molecular profiles in acute myeloid leukemia defined by comprehensive clinical genomic profiling

Katherine Tarlock,Eric Severson,Mark Bailey,Tariq Mughal,Jeffrey S Ross,Philip J Stephens,Rachel Erlich,Vincent Miller,Soheil Meshinchi,Samantha Morley,E Anders Kolb,Doron Lipson,Jo-Anne Vergillo,Sohail Balasubramanian,Jie He,Yuting He,Geoff A Otto,Shan Zhong,Rhonda Ries

doi:10.18632/oncotarget.25443

Abstract

Large scale comprehensive genomic profiling (CGP) has led to an improved understanding of oncogenic mutations in acute myeloid leukemia (AML), as well as identification of alterations that can serve as targets for potential therapeutic intervention. We sought to gain insight into age-associated variants in AML through comparison of extensive DNA and RNA-based GP results from pediatric and adult AML. Sequencing of 932 AML specimens (179 pediatric (age 0–18), 753 adult (age ≥ 19)) from diagnostic, relapsed, and refractory times points was performed. Comprehensive DNA (405 genes) and RNA (265) sequencing to identify a variety of structural and short variants was performed. We found that structural variants were highly prevalent in the pediatric cohort compared to the adult cohort (57% vs. 30%; p < 0.001), with certain structural variants detected only in the pediatric cohort. Fusions were the most common structural variant and were highly prevalent in AML in very young children occurring in 68% of children < 2 years of age. We observed an inverse trend in the prevalence of fusions compared to the average number of mutations per patient. In contrast to pediatric AML, adult AML was marked by short variants and multiple mutations per patient. Mutations that were common in adult AML were much less common in the adolescent and young adult cohort and were rare or absent in the pediatric cohort. Clinical CGP demonstrates the biologic differences in pediatric vs. adult AML that have significant therapeutic impacts on prognosis, therapeutic allocation, disease monitoring, and the use of more targeted therapies.

Highlights

The complex genomic landscape of acute myeloid leukemia (AML), with several recurrent cytogenetic and molecular abnormalities, is an important factor in understanding the variable outcomes achieved in patients [1,2,3]
A total of 571 genomic variants of all types involving 131 genes were detected in pediatric cohort and 3,018 variants involving 219 genes in the adult cohort
Our results confirmed previous large-scale sequencing initiatives demonstrating that AML is a very heterogeneous disease with multiple distinct oncogenic drivers and many patients harboring multiple www.oncotarget.com mutations, suggesting distinct mechanisms of cooperation [4, 16,17,18]

Summary

Introduction

The complex genomic landscape of acute myeloid leukemia (AML), with several recurrent cytogenetic and molecular abnormalities, is an important factor in understanding the variable outcomes achieved in patients [1,2,3]. Large scale generation sequencing (NGS) efforts including the Therapeutically Applicable Research to Generative Effective Treatments (TARGET) initiative and The Cancer Genome Atlas (TCGA) have made significant contributions to our understanding of the complexity of the genomic landscape of AML [4,5,6]. Comprehensive genomic profiling (CGP), that utilizes DNA and RNA sequencing, has led to an improved understanding of oncogenic mutations in AML, as well as alterations that can serve as targets for potential therapeutic intervention. Some of the more recently detected mutations in AML have much higher frequencies in adults compared to children, and this can have significant clinical implications as targeted agents are developed against these mutations [9,10,11]

Methods

Results

Conclusion