This week’s highlighted Anemia in MDS article can be viewed below. The approach to MDS increasingly involves genomics. Approximately 50% of patients with MDS carry somatic mutations in spliceosome genes, and SF3B1 is the most commonly mutated one.

Next-generation sequencing (NGS) is now more standard in the evaluation of myelodysplastic syndrome (MDS). Genomic features, such as mutations in TP53 and SF3B1, are being incorporated into MDS classification systems and prognostic tools, and they may have direct implications for treatment.

There has been a more widespread adoption of NGS in the community over the last few years, and I wonder if this, in part, reflects the influence of more reflex testing by pathology departments. The NGS panels that are commercially available and are used at the time of diagnosis are broader than before, which has also been a positive development.

It is worth emphasizing the value of repeat testing at the time of clinically significant changes in the patient’s disease status. There is an increased understanding of the polyclonal nature of myeloid neoplasms, with subclones potentially expanding at times of progression or loss of response. With any clinically relevant change in the patient’s disease status, it is important to retest with NGS to determine whether there is an expanded subclone that might be targetable. There are new targets that are actionable, and, while only a relatively small subset of patients with MDS have these targets, it is sufficiently common to warrant retesting.

Molecular assays that detect fusions are also evolving. Fusion assays might be more applicable in MDS/myeloproliferative neoplasm overlap syndromes, but the idea is to identify hard-to-find translocations that can be targetable. There are some tyrosine kinase inhibitors, for instance, to which the disease may be sensitive, depending on the type of fusion that is identified. So, molecular diagnostics are evolving hand in hand with therapeutics, and, increasingly, there are opportunities to identify actionable alterations.

In most patients with MDS, testing for somatic mutations should be standard, since genomics are being integrated more and more into the diagnosis, classification, risk stratification, and treatment of MDS. All of our classification systems are starting to integrate genomics. The World Health Organization classification is being updated and will include a reclassification based on molecular features (eg, SF3B1 and TP53). Whereas SF3B1 mutations are associated with a better overall survival and lower rates of leukemia transformation, TP53 mutations are, unfortunately, linked to some of the worst outcomes in MDS.

A variety of genomic prognostic systems have been developed and proposed. The Molecular International Prognosis Scoring System is one such system that integrates both clinical and molecular features and leads to improved discrimination compared with the Revised International Prognostic Scoring System score. Ultimately, we must ensure that these emerging tools become practical enough to be used easily by community oncologists.

When we see hypomethylating agent failure, I think that we should reevaluate the disease and incorporate genomic testing to identify actionable targets. Although each target may be present in only a small subset of patients, such testing could affect a much larger group of individuals when all targets are considered.

Many of the applications for targeted therapy are best done in the context of clinical trials, but the individualization of therapy in MDS based on molecular features has already begun. For instance, we have luspatercept for patients with lower-risk MDS with ring sideroblasts, which is strongly correlated with SF3B1 mutations. IDH mutations occur in a small subset of patients with MDS, and clinical trial protocols incorporate IDH inhibitors. And then, lenalidomide has been known for some time to have utility in patients with MDS with chromosome 5q31 deletion.

Just a few years ago, it was not a sure bet that a patient who was referred to an academic center would have an NGS profile performed. Conversely, today, it is very unusual for a patient with acute myeloid leukemia (AML) to be referred without having had an NGS panel, and the majority of patients with MDS have been tested as well. These panels can range from approximately 15 to 100 genes, but they all incorporate the key genes that are part of prognostication and treatment decisions. Typically, it is a commercial NGS test that is performed using a focused gene panel for myeloid malignancies, and it is being reflexively ordered for anyone with an underlying myeloid disorder.

Genomics in MDS can help to determine which patients are at a higher risk of progression to AML. This is important because we want to make the necessary preparations for transplant and follow these patients closely so that we can catch them before they progress. Mutational data can complement the Revised International Prognostic Scoring System, and there are mutations that have been associated with a higher likelihood of progression in MDS, such as IDH1, IDH2, SRSF2, and NPM1.

As Dr Komrokji mentioned, some of the mutations that are actionable in AML, such as IDH mutations, may also be relevant in MDS; however, the subset of patients having these targets in MDS is much smaller than in AML. Thus far, the targeted agents for IDH1, IDH2, and FLT3 mutations have been approved by the US Food and Drug Administration only for AML.