Practical Acute Myeloid Leukemia (AML) Biomarker Testing Using Next-Generation Sequencing (NGS) Technology: A Comprehensive, Rapid, Inexpensive and Flexible Approach

The diagnostic category of AML includes a very heterogeneous group of neoplasms but progress has been made in recent years to identify biologically and clinically relevant genomic biomarkers that aid in further sub classification, provide more accurate prognostic information and supply targets for tumor-specific therapy and monitoring of residual disease. At this time, clinically relevant biomarkers include single nucleotide variants (SNVs), small insertions and deletions (indels) and a variety of translocations. They involve many different genes, including CEBPA, which is a highly G/C nucleotide-rich gene that is extremely difficult to evaluate using standard NGS methods. Because of the number and complexity of biomarkers, it has been a challenge for clinical laboratories to provide testing that is of practical clinical use; most approaches require multiple testing modalities, including less than satisfactory commercial sequencing panels, resulting in poor clinical testing service with respect to biomarker coverage, turn-around time and cost.

We have developed an integrated "laboratory-developed" NGS-based approach that evaluates all current, clinically relevant AML biomarkers simultaneously (including CEBPA and identity markers for post-transplant chimerism evaluation), in as little as 2.5 days (<6 hrs hands-on time) for a total cost of only a few hundred dollars per sample. The method is flexible with respect to future, custom addition of newly discovered biomarkers and can be used on any of the Life Technologies or Illumina NGS platforms. Briefly, the method involves extraction of both DNA and RNA, the use of 3 slightly different techniques for initial library preparation (one each for CEBPA, other small variants, and translocations), followed by addition of indexed sequencing ends for either of the major sequencing technologies, pooling of the libraries and sequencing. The data analysis pipeline is rapid and simple to use and was custom created using predominantly readily available software.

The method was used successfully to evaluate 50 patient blood/liquid bone marrow aspirate samples and 25 custom-designed DNA synthetics containing previously identified variants in ASXL1, BRINP3, CEBPA, DNMT3A, EZH2, FLT3, GATA1, GATA2, HNRNPK, IDH1, IDH2, KIT, KRAS, KMT2A(a.k.a. MLL), NPM1, NRAS, PHF6, PTPN11, RAD21, RUNX1, SMC1A, SMC3, STAG2, TET2, TP53, U2AF1, WT1, and translocations CBFB/MYH11, DEK/NUP214, KMT2A(a.k.a.MLL)/X=any partner, PML/RARA, NUMA1/RARA, STAT5B/RARA, ZBTB16/RARA, RBM15/MKL1, RPN1/MECOM, RUNX1/RUNX1T1 with 100% concordance and reproducibility. It has a reliable sensitivity of at least 1% and further evaluation is underway to establish the limit of detection for each target-type using ultra-deep sequencing, to determine acceptability for residual disease detection. The assay was shown to be 100% specific when challenged with normal samples.

This integrated approach is novel and makes maximal use of available sequencing technology to simplify clinical biomarker evaluation, as required for practical management of patients with AML.