Jamile Shammo, MD: Well, it’s been shown in the Greenberg IPSS paper—if you take them collectively—that all MDS patients had about a 30% chance of progression to acute myeloid leukemia (AML) and, hence, that’s the most rare complication of this entity. But, if you split them up by low-risk and high-risk, it’s about 25% in the low-risk group and maybe 40% to 50% risk of dying of AML. This is sort of an older registry. But what about something newer? There was just a publication not too long ago about close to 4000 patients in the Düsseldorf MDS registry, and there, they went back and collected data on people who had actually died. They had about a 21-month follow-up of those patients. So, they didn’t have data on everyone that died, but pretty much 60% of the people evidently died; I would say 45% from AML progression—so kind of similar to what you saw in the IPSS paper—and then about 30% to 35% from infection or bleeding. When we talk about AML progression, granted, it’s about half of those patients; but remember, a very comparable percentage of patients will die from consequences of bone marrow failure. And I think that’s why it’s so important for us to not just work on perhaps controlling progression to leukemia, but also deal with consequences of bone marrow failure. And, a smaller percentage, perhaps 16%, died of things that were totally unrelated, cardiac failure among other things. So, this is sort of an interesting take on more recent causes of death for MDS. Do you find that to be similar to what you see in practice?

Rami S. Komrokji, MD: Yes, absolutely. I think the message is that MDS definitely can impact the survival for the patients, unfortunately. Even what we call a very low-risk or low-risk patient, their survival is going to be affected by the disease. As you mentioned, roughly one-third die from transformation to AML, but mostly of complications from bone marrow failure, infection, bleeding. I have to say that we see that less commonly in our practice than what’s reported. But I also think MDS becomes like a comorbidity. Most of the patients under the age of 70, they probably have other medical problems like cardiopulmonary diseases, being anemic all the time, cytopenic, which probably also contributes to the other medical problem. So, we see patients dying from other things like coronary artery disease, CHF (congestive heart failure), but, in reality, they probably are linked to the MDS.

Ellen K. Ritchie, MD: I think that’s true, that comorbid illnesses in the elderly patient are really magnified by myelodysplastic syndrome or having bone marrow failure. We also see a lot of cardiac complications in our patients, whether it’s a development of new atrial fibrillation in the setting of having a low hemoglobin, or whether it’s a high output heart failure type situation. But, we see real exacerbation of problems that might not be so severe for those patients if they didn’t have myelodysplastic syndrome. And a patient who has comorbidities who gets pneumonia, may be at higher risk of dying of that infection rather than a patient who doesn’t have myelodysplastic syndrome who gets an infection of some kind. So, I think it really just sort of strengthens the problems with other comorbid illnesses when a patient has myelodysplastic syndrome.

Mikkael A. Sekeres, MD, MS: So, to summarize, it sounds like most of the reasons our patients are dying are due to infections or bleeding complications. These are consequences of having MDS and cytopenias, but we can’t forget about competing risks in an older population where the average age of diagnosis is 70 years. In the United States, there’s going to be things like heart disease or pulmonary problems. And we don’t know what the contribution of MDS is to those comorbidities. I wonder if I could ask just one more question as we finish up this section on diagnosis and classification. We touched earlier in some of the next-generation sequencing and molecular characterization of MDS. Ellen, what do you consider some of the good risk molecular abnormalities versus some of the poor risk molecular abnormalities, and how does that influence your thinking about your patients?

Ellen K. Ritchie, MD: Well, when I think about the molecular abnormalities, it’s a really hard thing to discuss with patients because we really don’t have all the information that we need to have a real important discussion. I tell them if I went to 2nd Avenue and I actually decided to test everyone over the age of 60 who walked by, I would find that a good proportion of those patients had at least one molecular abnormality. And sometimes there are molecular abnormalities—which when I see them in my MDS patients, I really worry about—like an ASXL1 mutation, for example, which a certain proportion of regular people who are elderly walking around happen to have.

There are some better risk molecular problems, particularly I think those problems that have a drug that might work for them. For example, if you have an IDH2 or an IDH1 mutation, it may be possible that there is a class of drugs that might be really active against your disease. And I think it’s really important for us to be able to identify that for patients and say that there’s a drug that we might be able to use that might improve your outcome. But a lot of times you come up with combinations of mutations that you’re really not sure what to say to the patient, or you might have a mutation or a combination of mutations that you think is relatively poor prognosis, only to see the patient act like a low-grade MDS for quite a while. So, you painted a rather grave picture when they had a spliceosome mutation, they had an EZH2 mutation, and they had ASXL1 mutation but they did relatively well. I think we’re really at the beginning of really understanding the implications of these molecular mutations on the outcome of patients with myelodysplastic syndrome. I hope that we’re going to be able to collect a lot more data, and the normal aging population, as well as the patient population with hematologic malignancies as to what the time course in development is over disease over time.

Mikkael A. Sekeres, MD, MS: We’re at the cusp of understanding what to do with these molecular abnormalities. There are so many of them, it’s dizzying. I’ll sometimes joke that just when I think I’ve gotten all of these molecular abnormalities on a slide, every 2 weeks Blood Journal publishes another molecular abnormality that has some kind of relevance. So, I try to simplify it in thinking SF3B1 is a good risk, and basically everything else is either neutral or bad risk. P53s are obviously bad risk. In general, the more mutations you have, the worse your outcome is going to be. One mnemonic I remembered is that as you have some of these additional abnormalities—ASXL1, EZH2, RUNX1—in my mind I will upstage a patient from one risk group in the IPSS or the IPSS-R. For example, if ordinarily I called the patient intermediate-1 and that patient has a bunch of these abnormalities, I would then in my mind think that this patient is going to behave more like an intermediate-2. But, as you mentioned, Ellen, I’ve been surprised at how well some of my patients have done despite these molecular abnormalities.

Ellen K. Ritchie, MD: Even in patients with p53 mutations who I feel are poor outcome patients, I’ve been surprised a couple of times by older patients in their 80s and 90s who’ve had a p53 mutation who really weren’t candidates for anything other than growth factor therapy, who persisted for quite a while, and actually had a period of transfusion independence. So, you can be surprised, and I think we need to know a lot more about the course of these molecular mutations in the older population.

Rami S. Komrokji, MD: I think the first look we took at those mutations was almost very simplistic because first, we just looked at the mutations in a dichotomous way; positive, negative. We know that even what to call a mutation is very challenging. We get those tests, and we look at them as, is this real, is this not real? Is the mutation a missense, nonsense? There are really much more details in what we thought originally. For example, one thing when we looked at the p53, we looked at our patients that we sequenced, and it really matters also the size or the clone burden. So, patients that have a variant infrequency of more than 40% with a p53 clone, those are the ones that really had the worse outcome. While patients that had less than 10% clone size did not. It’s really not a simple way of yes or no, the mutation is present or not. It’s the type of the mutation, its implication in the biology, was it a founder mutation, driver mutation? It’s really much more complex than we think. I think, as Mikkael said, by data that we know now, we definitely know there is one that’s favorable, the SF3B1. That can be a confident statement I think. And then everything else, especially the number of mutations, does better. And the others may have either bad outcome or like neutral outcome. But this has been looked at in their presence. Some of those mutations will upstage the patients—as you said, one step. But, again, I think it’s going to be a work in progress, and we’ll learn in the coming few years much more about their significance.

Transcript Edited for Clarity