Staging Models in Diagnosis and Risk Assessment of MDS

Gail Roboz, MD: There’s no doubt about it in my mind, that when a patient hears something is low risk, that is synonymous with, “I live forever.” So, “Low risk, alright, I’m not going to worry about it." And I think we do worry about it. Rami, we have published on this as part of the MDS CRC [Myelodysplastic Syndrome Clinical Research Consortium]. Low risk is not actually so low risk. I think when you’re assessing a patient who might technically meet low-risk criteria, but as Azra said, there are other things making you think, “Hmm,” how do you distinguish between somebody who you’re going to hang out and watch on some transfusions and see what’s going on, versus something different? You come up, you use the app, you use all the data that you have. It says low risk, but how do you know for sure whether you are going to sit tight and watch or wait, or transfuse? How do you make that decision?

Rami Komrokji, MD: Absolutely. As you mentioned, we looked at this in the setting of the consortium, and we looked at patients who we labeled lower risk by even all the risk models. Still the reality was that unfortunately, 25% of those patients died within 2 years. It’s really not trivial, and that’s not all lower risk. Obviously, it gets more complex because I think in lower risk, there is so much interplay between the MDS [myelodysplastic syndrome] and the comorbidities of the patient, and other problems. If somebody has heart disease, and now they are anemic, yes, the anemia is not maybe the cause of the mortality, but the anemia is definitely going to be contributing to the other comorbidities. We look at the traditional things like the risk models, the mutations, or the depth of the cytopenia. How profound is their cytopenia is very important. I think if they are clearly cytopenic, certain mutations make me worried. Then there is something none of the models capture, that those diseases are not static. There are the kinetics of the disease, or the dynamics. Sometimes you observe the patients, and within maybe a short period of 6 months or a year, you will figure out if their disease is starting to act up or not.

I think the complexity in the lower risk is this interaction with the, I call them sometimes host-related factors. It is the patient-related factors: their comorbidities, functional status, other diseases, how symptomatic they are. There are some fascinating studies looking at fatigue, and that it can predict the outcome as much as the IPSS [International Prognostic Scoring System], or those symptoms of patients if they are symptomatic. All those things factor in. People can go the way of looking at the revised IPSS, or this is intermediate risk, then they may do worse. I would say all those staging models are the starting point, just to get the initial gauge. You cannot generalize the treatment for every patient. You will see somebody you label as intermediate who will do very well for a long time. And you see somebody who is intermediate who will go to higher risk shortly. I have been trying recently to look at a project where I look at the natural history of the patients we see with lower-risk MDS. It’s very complex; many patients will stay lower risk. But still, as we mentioned, unfortunately probably more than half of the patients will die directly or indirectly from the disease. There are patients who go to higher risk who never cross to AML [acute myeloid leukemia]. There are patients who go to higher risk, then AML, and there are patients with lower risk who go to AML. It’s really very complex. I think the general guideline is to get an assessment of the risk as a starting point, look at the nature of the disease, how profound are the cytopenias, and the interplay with other comorbidities. That is what I usually do.

Gail Roboz, MD: If somebody comes to you in January, and you do your risk assessment, and the components include things like extensive neutropenia, what is the hemoglobin, what are the platelets, what are the blasts. You have your mutations, you make an assessment, and then 9 months later they have the exact same assessment. Is the risk the same? This is the question that we get. I get the consult, I get the same thing, but they were diagnosed, make it easy, a year ago. Is it the same?

Rami Komrokji, MD: I do not know if we have studies to tell that, because all the classical teaching says those risk models are not dynamic. They are at time of diagnosis; you cannot predict the outcome. But I would say, yes. If the kinetics of the disease, and that is what I was trying to refer to, are stable, those are the patients who are going to be stable. There is uncertainty about when is the disease going to turn around. But it is never looked at, to my knowledge, in a systemic fashion. But I would think, yes, if somebody is stable over a year, they are probably going to be better than somebody who is progressing.

Gail Roboz, MD: One of the bugaboos that I think we all deal with is, what is the hemoglobin threshold number? What do you transfuse at? Patients are driven crazy by this because depending on which office they land in, it is either 8.2, or 7.8, or 7.295 g/dL, with 4 decimal points afterwards, and this can drive you nuts.

Transcript edited for clarity.