The Hematologist

May-June 2016, Volume 13, Issue 3

Next-Generation Sequencing Multi-Gene Mutation Panels in Myeloid Malignancies

Fei Yang, MD, MS Fellow, Departments of Pathology, Genetics, & Knight Cancer Institute
Oregon Health & Science University, Portland, OR
Richard D. Press, MD, PhD Professor, Departments of Pathology, Genetics, & Knight Cancer Institute
Oregon Health & Science University, Portland, OR

Published on: April 14, 2016

The rapid adoption of next-generation (massively parallel) DNA sequencing (NGS) methods into clinical diagnostic laboratories has created an unprecedented opportunity to profile the multiple actionable driver genes in patients with known and/or suspected myeloid malignancies. The clinical availability of this abundance of information now presents practicing hematologists with several unique genomic-era challenges that could not have been imagined in the “single-gene” era of only a few years ago. Toward the goal of demystifying some of the unique complexities of NGS-based testing, this review will address some common practical questions that arise regarding “hotspot” gene mutation panels for myeloid malignancies. The same generic issues are equally applicable to other hematologic malignancies as well as nonhematopoietic solid tumors. This discussion will not touch on the use of gene panels in the context of clonal hematopoiesis of indeterminate potential, idiopathic cytopenias of undetermined significance, idiopathic dysplasia of undetermined significance, or clonal cytopenias of undetermined significance, which are extensively reviewed elsewhere,1-4 and which are becoming a very common diagnostic outcome after gene panel work-ups for cytopenia.

Acute myeloid leukemia (AML), myelodysplastic syndrome (MDS), and myeloproliferative neoplasms (MPN) are clonal disorders of hematopoietic stem cells. The simultaneous NGS-enabled detection of somatic mutations in dozens or hundreds of candidate target genes as markers of the neoplastic clone, can provide “actionable” information in four broad clinical utility categories, as outlined in the Table: 1) diagnosis, 2) prognostic risk stratification, 3) eligibility for targeted therapy, 4) and minimal residual disease (MRD) detection and monitoring.

Diagnostic Utility

A common diagnostic dilemma that is often favorably informed by NGS-based mutation profiling is the evaluation of a cytopenic patient for possible MDS. The presence of a pathologic MDS-associated driver mutation (Table) can establish the existence of clonal hematopoiesis and thus support an MDS diagnosis versus a nonclonal benign cause of cytopenia.5,6 The overwhelming majority of patients with MDS (approximately 70%-90%) will have one or more oncogenic mutations identified by NGS.6 Other common scenarios in which NGS can be diagnostically informative (Table) include distinguishing clonal MPNs from benign pathologic mimics (by detecting JAK2, CALR, or MPL mutations) and detecting a particular gene mutation that has a high positive predictive value for a specific diagnostic entity (often with a specific targeted therapy). One example is detection of the CSF3R T618I mutation that is found in approximately 80 percent of patients with chronic neutrophilic leukemia (and more rarely in atypical CML); ruxolitinib is now being tested in these patients since CSF3R signals through JAK2.

Prognostic Utility

The presence of certain recurrent somatic mutations in myeloid malignancies can provide useful prognostic risk stratification and inform optimal treatment (Table). In cytogenetically normal AML, consensus risk stratification recommendations include the routine assessment of mutations in the FLT3 (poor risk), NPM1, and CEBPA (favorable risk) genes to inform treatment decisions, including the use of hematopoietic stem cell transplantation.7,8 In patients with a normally favorable risk core binding factor AML, the presence of a KIT mutation conveys a relatively poorer prognosis, such that these patients should be considered intermediate risk.8 Less-mature data might suggest the concomitant use of other molecular markers for AML risk stratification, including TP53, IDH1/2, DNMT3A, TET2, and RUNX1 (all conferring poor risk).9 Multigene sequencing panels are also often used to stratify risk in MDS, with several common gene mutations conferring a poor prognosis (DNMT3A, U2AF1, ZRSR2, ASXL1, RUNX1, EZH2, NRAS, and TP53).5,10 Otherwise-low-risk MDS patients by International Prognostic Scoring System criteria, but with an EZH2 or TP53 mutation, have a worse prognosis than expected.

Therapeutic Utility

Perhaps the most direct patient care benefit of multigene sequencing panels is the detection of specific mutations that define a likely response (or nonresponse) to a targeted therapy, either with an approved drug, or one being evaluated in an ongoing clinical trial. Proven or promising examples include JAK2 inhibitors in JAK-STAT pathway-activated MPNs,11 second-generation FLT3 kinase inhibitors in FLT3-ITD-positive AML,9,12 inhibitors to the mutant IDH1/2-generated oncometabolite in AML,9,13 BRAF inhibitors in hairy cell leukemia,14 hypomethylating agents in TET2/DNMT3A/ASXL1-mutant MDS,15 and many other examples in ongoing clinical trials (Table).

MRD Utility

Another promising area of clinical utility for NGS-based profiling is monitoring the persistence of AML-associated mutations (“minimal residual disease”) after therapy. The persistent detection of these pathogenic mutations after standard AML induction chemotherapy is predictive of future relapse,16,17 and may indicate the need for more aggressive therapy.

How Many Genes Should Be Tested, and with What Sensitivity?

The mutation profiling tests available within clinical laboratories are predominantly NGS-based “hotspot panels” interrogating all exons or selected commonly mutated regions within 10 to 500 different “cancer genes.” The lab-specific choice of gene targets and correlated tumor types is heterogeneous and must be carefully assessed by the referring clinician. Assays specific for “hematopoietic malignancies” or “myeloid malignancies” are the common consequence of such decisions. The typical analytical sensitivity (lower detection limit) for such NGS-based tests is typically a variant allele fraction (VAF) of 1 to 10 percent (depending on sequencing coverage), which is slightly better than traditional Sanger sequencing. Although highly heterogeneous, most of these assays are able to reliably detect nucleotide substitutions and short insertion-deletion variants, and some can also detect copy number alterations, complex rearrangements, and gene fusions/translocations.

How Are Cancer-Associated Somatic Mutations Defined and Other Variants Filtered Out?

The initial long list of variants from the raw NGS instrument is carefully filtered to ultimately exclude, from the clinical diagnostic report, those variants that are either 1) nonsomatically acquired benign germline polymorphisms, or 2) nonpathogenic (for the cancer of interest). The former are typically defined as variants with a population allele frequency (in large normal population screening databases) greater than some defined threshold (typically 1%). Otherwise uncharacterized variants at a germline-compatible 50 percent or 100 percent allele frequency can often be clarified as “likely benign germline.”

Determining the “pathogenicity” of a particular variant is a much more difficult process, requiring rigorous investigations into 1) the association of the variant (how frequently?) with the tumor of interest (or other tumors); 2) in vitro or animal studies of variant function/oncogenicity; 3) in silico tools for predicting the biochemical (dys)function of the variant; and 4) clinical trials of possible targeted therapies. Invaluable database/tools for these variant annotation tasks include COSMIC (catalogue of somatic mutations in cancer), ClinVar, My Cancer Genome, cBioPortal, HGMD, OMIM, SIFT, Provean, PolyPhen-2, MutationTaster, CADD, and others.

How Are Cancer-Associated Somatic Mutations Interpreted As Potentially Clinically “Actionable?”

Although all NGS-based diagnostic reports will include a list of prefiltered variants/mutations, their annotation with respect to clinical “actionability,” whether the mutation is likely to directly impact patient care with respect to diagnosis, prognosis, or therapy, is highly variable.18 This assessment of actionability requires a manual (or semiautomated) curation of published literature (PubMed); professional guidelines; and comprehensive clinical, genetic, biochemical, and functional database searches (detailed above). Although real-world heterogeneity in the art and science of mutation interpretation and reporting remains substantial, stakeholders in the pathology, hematology, and oncology communities are advocating for and drafting a standardized system of discrete categories of actionability evidence. These four to five actionability categories may ultimately include terms such as: “actionable,” “potentially actionable,” “variant of unknown significance (VUS),” and “benign/likely benign.” Although any such “actionable” variant should, of course, be emphasized in the final report, many pathologists do not report benign or likely benign variants. The grey-zone VUS category presents a particular challenge to both laboratories and pathologists (“should we report it?”) and clinicians (“what does it mean?”). In our opinion, however, a reportable mutation list that is a little too long (i.e., includes VUS but not benign variants) is usually preferable, given that a “VUS” with today’s knowledge may become an “actionable” mutation in the future. The contrary opinion, to keep reports short and clinically focused, is certainly respected, particularly when the list of reported mutations is long, and the truly actionable mutations thus risk being “lost in the weeds.” However, since the average diagnostic MDS or AML case in our laboratory yields only two reportable variants (out of 42 or 76 genes tested), this risk of “over-reporting” would seem to be rather small.

Practical Issues: Test Availability, Timeline, Costs, and Reimbursement

NGS-based mutation profiling tests are now widely available in academic medical centers, commercial laboratories, and large community hospitals. However, since all of these tests are non-FDA approved “laboratory developed tests,” the technical details of each test — genes and tumors covered, mutation detection limits, reporting algorithms — are nonstandardized and unique to the individual laboratory. The turnaround time to a clinical report is typically one to three weeks, depending on test complexity. As with any complex laboratory test, knowing the analytical and clinical performance characteristics of the assay, perhaps best accomplished by direct consultation with the laboratory director or pathologist, is then essential for optimized clinical care.

As the instrumentation, bioinformatics, and workforce required for initiating and maintaining expertise in this rapidly evolving field are considerable, NGS mutation profiling tests are, from a clinical laboratory (but not pharmaceutical) perspective, quite expensive, ranging from a few hundred to a few thousand dollars. Perhaps more problematic to the patient than the cost of the test, however, is the poor reimbursement of these tests by many government and private insurance payers, perhaps due to the misconception that these expensive tests are for research purposes only and are of unproven benefit for direct patient care. As the data required to refute these misleading economic/medical contentions will require a large number of patients, continued patient access to these complex laboratory services may well depend on our collective ability to track and share clinical outcomes on our patients who undergo NGS-based mutation profiling.


  1. Steensma DP, Bejar R, Jaiswal S, et al. Clonal hematopoiesis of indeterminate potential and its distinction from myelodysplastic syndromes. Blood. 2015;126:9-16.
  2. Kwok B, Hall JM, Witte JS, et al. MDS-associated somatic mutations and clonal hematopoiesis are common in idiopathic cytopenias of undetermined significance. Blood. 2015;126:2355-2361.
  3. Cargo CA, Rowbotham N, Evans PA, et al. Targeted sequencing identifies patients with preclinical MDS at high risk of disease progression. Blood. 2015;126:2362-2365.
  4. Steensma DP. Cytopenias + mutations - dysplasia = what?. Blood. 2015;126:2349-2351.
  5. NCCN Clinical Practice Guidelines in Oncology. Myelodysplastic Syndromes. National Comprehensive Cancer Network. 2016;1.
  6. Papaemmanuil E, Gerstung M, Malcovati L, et al. Clinical and biological implications of driver mutations in myelodysplastic syndromes. Blood. 2013;122:3616-3627.
  7. Döhner H, Estey EH, Amadori S, et al. Diagnosis and management of acute myeloid leukemia in adults: recommendations from an international expert panel, on behalf of the European LeukemiaNet. Blood. 2010;115:453-474.
  8. NCCN Clinical Practice Guidelines in Oncology. Acute Myeloid Leukemia. National Comprehensive Cancer Network. 2016;1.
  9. Döhner H, Weisdorf DJ, Bloomfield CD. Acute myeloid leukemia. N Engl J Med. 2015;373:1136-1152.
  10. Gangat N, Patnaik MM, Tefferi A. Myelodysplastic syndromes: contemporary review and how we treat. Am J Hematol. 2016;91:76-89.
  11. Patel AB, Vellore NA, Deininger MW. New strategies in myeloproliferative neoplasms: the evolving genetic and therapeutic landscape. Clin Cancer Res. 2016;22:1037-1047.
  12. Sasine JP, Schiller GJ. Emerging strategies for high-risk and relapsed/refractory acute myeloid leukemia: novel agents and approaches currently in clinical trials. Blood Rev. 2015;29:1-9.
  13. Dang L, Yen K, Attar EC. IDH mutations in cancer and progress toward development of targeted therapuetics. Ann Oncol. 2016;27:599-608.
  14. Dietrich S, Pircher A, Endris V, et al. BRAF inhibition in hairy cell leukemia with low dose vemurafenib. Blood. 2016;pii:blood-2015-11-680074. [Epub ahead of print].
  15. Nazha A, Radivoyevitch T, Thota S, et al. Somatic mutational model to predict response to hypomethylating agents in myelodysplastic syndromes. Leuk Res. 2015;39:s9. [Epub ahead of print].
  16. Klco JM, Miller CA, Griffith M, et al. Association between mutation clearance after induction therapy and outcomes in acute myeloid leukemia. JAMA. 2015;314:811-822.
  17. Ivey A, Hills RK, Simpson MA, et al. Assessment of minimal residual disease in standard-risk AML. N Engl J Med. 2016;374:422-433.
  18. Sukhai MA, Craddock KJ, Thomas M, et al. A classification system for clinical relevance of somatic variants identified in molecular profiling of cancer. Genet Med. 2016;18:128-136.

Conflict of Interests

Dr. Yang and Dr. Press indicated no relevant conflicts of interest. back to top