NGS for Myeloid Neoplasm Evaluation
Originally posted October 24, 2022
David Viswanatha, M.D.
Professor of Laboratory Medicine and Pathology
Division of Hematopathology
Department of Laboratory Medicine and Pathology
Mayo Clinic, Rochester, Minnesota
Hello, my name is David Viswanatha. I am a consultant hematopathologist, and I co-direct the Molecular Hematopathology Laboratory here at Mayo Clinic in Rochester, Minnesota. Today, I am going to be discussing some updates to our next generation sequencing, or NGS offering, for myeloid neoplasm evaluation here at Mayo.
I have no disclosures.
So, the outline I will follow is as shown here. I will recap our current NGS panel offering for myeloid neoplasms. The test ID, or mnemonic, is NGSHM. I will then go over the new version of NGSHM. It's an update rather than a complete overhaul. As part of this, I'll discuss the new targets, and some of the new aspects of the chemistry, a brief rationale for why we have introduced new genetic regions or targets. I'll discuss briefly the alignment with the WHO in 2022 and ICC 2022 consensus guidelines and talk a little bit about germline predisposition targets. And then, just go over a couple of questions that continue to arise with next generation sequencing testing, and then conclude.
The current NGSHM panel consists of 42 genes, as well as two smaller, AML-focused subpanels. The test IDs for these are NGAML and NGAMT, and I'll talk about those in a little bit more detail. We also have a reflex testing option, which we call NGSFX, which can be ordered, if initially, a smaller subpanel, the NGAML or AMT, were ordered and the client would like to uncover the remaining genes in the parent panel, the 42-gene panel. That's a reflex option that can be used when an initial, smaller panel was ordered and completed. The current panel has a focus on detecting single nucleotide variants, or changes in DNA, as well as small, insertion/deletion duplication events, which we collectively refer to as indels. The analytical sensitivity for our current offering is 5%, meaning that we can detect a somatic genetic mutation in DNA down to a level of 5% in a background of normal DNA. The NGS testing involves, or incorporates, a very dedicated and highly experienced system, with individuals who focus specifically on this type of testing, and all clinical reports are reviewed and signed by consultant hematopathologists.
This table just shows the distribution of genes, and regions for NGSHM, for the current version. You can see just by looking through the various genes, and the exons that are covered, that some genes have complete exon coverage, others will have more broad, regional coverage. And some genes, such as the IDH1/2 genes, or the RAS genes for instance, will have very focused, targeted hotspot type of coverage. We do that depending on the clinical utility and the biology of the genetic disorder.
This slide shows in red boxes the composition of the NGAML 11-gene panel, and in the yellow or orange boxes, the NGAMT, or mini targeted-type AML panel, that we also have as a subpanel.
The new version of NGSHM, so it will retain the same mnemonic, will go up to 47 genes. This is a net increase of five genes in that one genetic target has been removed, and six new ones have been added. The new parent panel will also include the ability to order the same two AML-focused, or directed, subpanels, the NGAML and NGAMT, and the content there is not changing. And the reflex testing option will also remain as well if the smaller panels were ordered at first and completed.
Importantly, the new genes and gene regions on the updated version are shown here in the new genes line. And these include the BCORL1 gene. So, both BCOR and BCORL1 have increasingly been shown to associate with adverse prognostic outcome in acute myeloid leukemia and other myeloid neoplasms. So, we felt it was important to include the counterpart to BCOR, which is BCORL1. And so that will be a new addition. The BRAF gene was added because even though it is present at low frequency in AML cases, it presents possible therapeutic targets. NF1, or neurofibromatosis 1 gene, was added to cover the entirety of genetic alterations that can be seen in juvenile myelomonocytic leukemia, or JMML, and make our coverage for RASopathy-type diseases more complete. And then finally, the PPM1D gene, which encodes a tumor suppressor, was felt to be important because this gene is more frequently mutated or found mutated in patients who have secondary AML, or therapy-related myeloid neoplasms. The other two genes on the list includes STAT3, which we felt important to include because some presentations of T large granular lymphocytic leukemia can mimic presentations of myeloid dysplastic syndrome. And of course, STAT3 mutations are the most common genetic alteration found in TLGL, lymphoproliferative disease. Finally, the UBA1 gene, in which somatic mutations were identified very recently, in a multiorgan inflammatory condition called VEXAS syndrome. And we are seeing increasing incidents of this syndrome, which can overlap with bone marrow manifestations, and felt it was important to add the UBA1 target onto our panel. So, that will also be available.
The new NGSHM panel includes an update to the chemistry. It's actually a new type of chemistry. And the bioinformatics processing is also updated and upgraded with a number of important features, including enhanced error correction, a contamination module, because we do high throughput sequencing with multiple samples at a time and improvements for indel detection. In addition, the analytical sensitivity has improved slightly to 2% now.
In tabular format, this is showing the new panel. The red font genes are the subpanel NGAML genes, and the green font genes are the new ones that have been added.
So recently, this past summer, two sets of papers emerged from the international consensus classification and the WHO working groups, which provide updates to the categories of myeloid and histiocytic neoplasms. And though it's unfortunate that there has been a split in the working groups in terms of classifications, there are some common highlights shared between the groups, including the following.
So acute myeloid leukemia with single gene mutations in the CEBPA gene, specifically in the C terminal basic Leucine zipper domain, or bZIP domain, and these are typically in frame insertions or deletions, tend to be associated with highly favorable prognosis very similar to the classic biallelic CEBPA mutation pattern. And so, this has been called out as a separate entity in the new classification themes. AML with myelodysplasia-related genetic alterations has also been identified as a more distinct type of entity, and the associated genes that are found mutated are shown in the parentheses. Again, all of these genes are covered by the NGSHM panel. MDS/AML with TP53 mutation, also identified as a more distinct entity. And then there's been a revised regrouping of myeloid neoplasms that are associated with germline genetic alterations.
While we await some consensus on merging the revised classifications from these two groups, it's important to note that the updated version of the NGSHM that is coming, and in fact, our current version, handily meet the requirements for identifying these genetically defined myeloid neoplasm categories and continue to provide additional information for correct classification and prognosis. And I've provided the references for these papers below.
So, a few words about germline predisposition. The ICC and WHO have reorganized the germline predisposition classifications into three major groups. The first being those that occur without any evidence of organ dysfunction, or systemic kind of disease, or a previous chronic platelet disorder. And so, examples of germline alterations in these genes would be CEBPA, DDX41, and TP53.
The second category includes myeloid neoplasms that are associated with a prodromal phase history of associated platelet disorder. And so, examples of this would be RUNX1 and an ANKRD26 associated familial platelet disorder, or patients who present with germline ETV6 and also may have platelet abnormalities earlier in life, prior to developing a myeloid neoplasm later.
And finally, germline predisposition to myeloid neoplasms with associated organ or system dysfunction. And examples of this would be genetic alterations in the germline of GATA2, or genes associated with the RASopathy.
One of the questions that comes up is, Why do NGS testing? And I wanted to provide some key points here for why there are substantial advantages for next generation sequencing. First of all, NGS provides comprehensive genetic profiling. We can detect different variant classes. We can detect these across multiple genes, and we can do this across different myeloid neoplasms all at once. This aids in the diagnosis, prognosis, and therapeutic management of these neoplasms. Trying to generate similar information from multiple linked genetic tests is simply not a scalable solution to the expanding knowledge and complexity at the genetic level, in myeloid neoplasms.
Second, NGS provides a fairly definitive assessment in patients who present with idiopathic cytopenias of uncertain significance, or ICUS. So, the identification of somatic alterations in patients with unknown cytopenias can reclassify these individuals as having clonal cytopenias of uncertain significance, which we termed CCUS. And of course, these patients have a variable risk of progression to overt myeloid neoplasms. The whole field of clonal cytopenias is now expanding and becoming more interesting and relevant in that there are maybe patterns of mutations occurring in patients that have different risks for progression to acute myeloid leukemia, for instance. Again, our NGS panel covers this possibility and provides this type of information.
In contrast, the presence of a negative result by NGS in a patient with an unexplained cytopenia provides relatively strong evidence against a neoplastic myeloid process. It doesn't completely exclude it, but the probability of having a myeloid neoplasm with a negative NGS assay result, and also a normal cytogenetics finding, very much argues against the possibility of a myeloid neoplasm. So there's a good negative predictive value in this situation.
We can concurrently assess for the possibility of specific germline predisposition conditions for the development of myeloid neoplasms, but of course, because our panel is targeted for somatic alterations, if we do find a potential germline alteration of interest, this will have to be called out in the report and may require additional testing to confirm its presence.
Finally, we have better technical resolution with NGS for complex genetic alterations, so better accuracy, based on the digital nature of NGS testing. It provides much better resolution relative to other testing modalities like Sanger sequencing, and this is consistent over different variant classes. The nice thing about NGS, and panel-based NGS in particular, is that it's easier and more cost effective to upgrade as new knowledge emerges. So we can more readily switch in a genetic target or two or three, or remove genetic regions that are of less value.
So, some of the challenges that we continue to face with NGS, I've tried to express in these relatively common questions that we get. So one of these is that NGS is relatively more expensive compared to single gene molecular assays. And that is true at the moment. However, as I mentioned, combining testing over many single gene assays to achieve the same purpose, dramatically increases the cost. And at some point, very quickly, that tips in favor of doing targeted NGS testing. In addition, sequencing and reagent costs continue to decline, and improvements in sequencing platforms make the cost per base DNA sequencing cheaper and cheaper, with each passing week or month. So this is favorable. Reimbursement has remained variable for NGS testing, but as more consensus guidelines incorporate mutation testing as part of diagnosis or management considerations, including the WHO, ICC, NCCN, for example, reimbursement is becoming less of an issue for NGS.
The second major issue we hear about is that NGS testing takes too long. And again, there's some truth to that, but the fact of the matter in the current state is that NGS is a relatively complex methodology and includes some fixed time points, which are difficult to significantly alter. So, these would include things like the time taken to prepare high-quality DNA library for sequencing, the actual on sequencer time, and some aspect of post-processing in the bioinformatics pipeline. However, having said that, we remain continually focused on significantly reducing the turnaround time. And we do this currently by prioritizing AML and other urgent cases, for example, but we are always reevaluating the end-to-end process and looking for ways to trim the turnaround time through faster platforms or automated analytics that will help us out in that regard.
Having said that, for acute myeloid leukemia in particular, we still recommend doing standalone FLT3 gene mutation testing to permit the early administration of targeted TKI therapy, such as gilteritinib, for instance, as part of protocols for AML and the FLT3 single gene assay has a more rapid turnaround time in the current state. So, we recommend that, and we will very soon have separate IDH1/2 single gene testing available to address that particular genetic therapeutic target as well.
So, in conclusion, I have presented the newest version of our NGSHM panel, which will launch later in 2022. It is a 47 gene complete panel with two small AML-oriented subpanels. And an in silico reflex testing option that can be ordered, if one of the subpanels was performed initially. The enhanced content reflects a significant number of additional genes of value, including the UBA1 gene and VEXAS syndrome. This new panel has better analytical sensitivity and accuracy. And of course, as always, we strive to provide this with the highest quality clinical and genetic interpretations, provided by our consultant hematopathologists and genetic counselors who have extensive experience in the diagnosis of myeloid neoplasms.
Thank you very much for your attention.
Contact us: email@example.com
Image credit: Getty Images