Clinical Utility of a Next-Generation Sequencing Panel for Acute Myeloid Leukemia Diagnostics

Abstract

Next-generation sequencing (NGS) has redefined the genetic landscape of acute myeloid leukemia (AML), providing new molecular markers for diagnostic and prognostic classifications. However, its application in the clinical setting is still challenging. We hypothesized that a 19-gene AML-targeted NGS panel could be a valid approach to obtain clinically relevant information. Thus, we assessed the ability of this panel to classify AML patients according to diagnostic and prognostic indexes in a cohort of 162 patients. The assay yielded a median read depth >2000×, with 88% of on-target reads and a mean uniformity >93% without significant global strand bias. The method was sensitive and specific, with a valid performance at the clinical variant allele frequency cutoff of 3% for point mutations and 5% for insertions or deletions (INDELs). Three-hundred thirty-nine variants were found (36% INDELs and 64% single nucleotide variants). Concordance between NGS and other conventional techniques was 100%, but the NGS approach was able to identify more clinically relevant mutations. Finally, all patients could be classified into one of the 2016 World Health Organization diagnostic categories and virtually all into the recently proposed prognostic indexes (2017 European LeukemiaNet and Genomic classification). To sum up, we validate a reliable and reproducible method for AML diagnosis and demonstrate that small, well-designed NGS panels are sufficient to guide clinical decisions according to the current standards.