Risk Stratification in MDS


Mikkael Sekeres, MD, MS: We talked a little bit about the distinction between MDS [myelodysplastic syndrome] and AML [acute myeloid leukemia]. How do you stage MDS?

Rami Komrokji, MD: We talked about the WHO [World Health Organization] classification. That’s more pathological, and sometimes you could just simplify and say, “I’m going to take patients that have excess blasts or no excess blasts and use the WHO classification.” But the WHO classification doesn’t capture the whole disease, and the risk stratification or staging in MDS is really based on 3 variables, one of the blood counts—so obviously cytopenia; the blast percentage on the bone marrow…

Nowadays we started integrating some of the molecular data from somatic mutations as well. The standard had been the IPSS, or International Prognostic Scoring System. That usually gives a lump score based on those variables and based on that we classify patients into 4 stages. As hematologists, we always like to call them low, intermediate, intermediate-2. I never knew why it wasn’t just called stage 1 to 4, but that’s how we name them. Basically, the survival advantage is from somebody that has less than 1-year survival to somebody that has more than 5-year survival. We all knew that the IPSS had some shortcomings, such as how it doesn’t account for the depth of cytopenia. So, if somebody’s platelets were 10 or 90, they got the same scoring. The cytogenetics were simple groups of lumping the patients into cytogenetics.

The IPSS-R, or the revised IPSS, tried to address some of those shortcomings, accounting for the depth of cytopenia more for the blast percentage, but also very detailed cytogenetic information. Now we can put the patients into 5 categories. Again, the survival will vary from less than a year to somebody that has very low risk, almost an 8.8-year survival. I look at the risk stratification in a way that it’s gauging the risk of the disease.

It’s important information for the patients, obviously, but we also tailor the therapy based on the risk. In reality, we are just thinking, are we going to be able, or can we justify, taking those patients to allogeneic stem cell transplant? Is there enough disease risk that will allow us to take those or justify taking those patients to transplant? This is because the transplant, unfortunately, still have mortality and morbidity associated with it, although it’s the only curative option.

In the next step, we started integrating some of the molecular data. We know that there are certain mutations that add to the clinical variables, and there are several papers that look at several gene mutations. I think some of the established ones, for example p53 is always a bad mutation. SF3B1 is the only good mutation, when it’s present by itself. Everything else could be somewhere in between. But things like ETV6, EZH2, and ASXL1 tend to have bad prognoses. I think nowadays we do the clinical models. We complement that by some of the molecular data, and we come with a risk assessment for the patients. I also try to look at the disease tempo. I look how stable the counts are. Somebody who is progressively becoming cytopenic and symptomatic may behave differently from somebody who, although their disease is becoming intermediate risk, has had stable disease for a long time. The disease tempo could be an important part of it as well.

Based on that, obviously, we tailor the treatment according to that risk stratification. There are different models used; more sophisticated. I would say maybe the IPSS-R [Revised International Prognostic Scoring System] is the most-used model. It’s important to keep in mind that the disease affects even what we call lower risk. We looked at this in the MDS Clinical Research Consortium together, and even for patients that we called lower risk, when we look at that, 25% of those patients unfortunately died within a couple of years. So, even what we call a very-low risk is a disease that’s going to probably impact the patient’s survival and quality of life.

Ellen K. Ritchie, MD: I think you bring up a good point and the idea of a tempo which is not captured actually by any other risk stratifications. You can see a low-risk patient but their progressive cytopenias over a 2-month time period are really impressive, you know that patient is going to do worse than a patient who has a relatively stable tempo. We don’t have a good way to prognose for that when we first see the patient. Many times, patients and risk categories surprise us. They don’t behave by the rules, and I think the tempo is one of those ways that we’re able to determine who those patients are.

Jamile M. Shammo, MD: Unfortunately, there are some shortcomings to both the IPSS that may have been addressed by the IPSS revised, but still for both of them I think you cannot utilize that for therapy-related disease, and I see this get done all the time. Yet, those patients were excluded from the analysis. I also think that it’s a problem because it’s not dynamic nor can you use it in people who have been treated previously.

Ellen K. Ritchie, MD: Right.

Jamile M. Shammo, MD: You’re left on your own if someone, let’s say, failed a therapy and then how do you go about…Although I do think that the French group had an idea actually for perhaps risk stratifying patients after HMA [hypomethylating agents] failures, which is, I think, an interesting way of thinking about it. But I don’t know that it’s been utilized that much.

Ellen K. Ritchie, MD: That stratification after HMA failure is always dismal stratification. For those patients that fail HMAs, we are at a loss for how we’re going to treat those patients and extend their overall survival.

Mikkael Sekeres, MD, MS: Let’s say I’m a general practitioner, and I’m seeing patients every week with a bunch of different diagnoses. I’m seeing some with breast cancer, lung cancer, colon cancer, chronic leukemias, and by the way an MDS patient here and there, is there an easier way to think about risk stratification? Because, personally, I’m almost embarrassed to admit this, since I do this for a living, I have the revised IPSS up on the wall of my work room with all the different cytogenetic risk categories because I can’t memorize all of that. I’ve got to go and check it and calculate a score or go on to an online application like the MDS Foundation has where you can actually put in information about this. What’s a quick and dirty way to think about risk stratification for patients with MDS?

Jamile M. Shammo, MD: I think the IPSS and the IPSS-R are a good place to start. There’s no doubt.

Mikkael Sekeres, MD, MS: But if you can’t memorize that. Let’s say you’re in a busy clinic, and you can’t remember that. You’re seeing somebody with MDS. Quick and dirty, how do you risk stratify?

Jamile M. Shammo, MD: I have it on one of the clinic walls, and there’s an app, actually.

Rami Komrokji, MD: I think you look at the components. Like if the blasts are increased, cytopenias, chromosomes look bad more than one line, I don’t think you need to calculate the score, basically. So really, you’re looking at the components. Maybe nowadays just thinking of complementing that by asking for molecular testing, but I think I’ve heard you talk about this before; in simple way, for blasts more than 10%, it’s not going to be good, right?

Jamile M. Shammo, MD: Right.

Rami Komrokji, MD: That could be a simple starting point using the depth of cytopenias, blast percentage, and cytogenetics. If they are complex, you don’t even have to calculate the score.

Mikkael Sekeres, MD, MS: Let’s say you have a patient who has high blast percentage, more than 10%, we all agree on that; multiple cytopenias; and…poor-risk cytogenetics, and the easy ones to remember are chromosome 7 abnormalities that are always bad. If a patient comes in with a paragraph of cytogenetic abnormalities, you know they’re going to have very poor prognosis. That person is high risk and should be treated as such.

Transcript Edited for Clarity

Related Videos
Claire Roddie, MD
Guillermo Garcia-Manero, MD
Amer Zeidan, MBBS
Expert on DLBCL
Tycel Phillips, MD
Expert on ALL
Michael Andreeff, MD, PhD
Brian C. Ball, MD
Mitchell E. Horwitz, MD
Related Content