Transcript:James M. Foran, MD, FRCP: I wanted to ask you, how do you classify patients when they come in, or how do you prognosticate for patients when they first meet you when you work them up?

Azra Raza, MD: We have all struggled with how to classify myelodysplastic syndromes since the late 70s. In 1976, the French, American, and British hematologists got together and proposed the FAB classification, which was prevalent for many years, until the World Health Organization (WHO) hematopathologists built upon that original FAB [French American British] classification and refined it, to a large degree. And this WHO classification then eventually replaced the FAB classification, but still the WHO or FAB was not giving us an idea about the prognosis of a patient.

So, then the International Prognostic Scoring System (IPSS) was developed. It was first developed in 1997 and this took into account, of course, three parameters. What is the percentage of blasts, what are the number of cytopenias, and what are the cytogenetics? Three parameters simply because no one parameter was able to accurately give us a prognosis. But, despite all of these scoring systems and classifications, the problem is that all of this is dependent on prognosticating for the population as a whole, which means you have a curve here.

So, even in this curve, you can say that 50% of the people will survive for 3.5 years. But then within that curve, there’s a third that will be dead before, and a third that will be living for up to 20 years. So, when an individual is sitting across my desk from me in the clinic, and I try to apply these systems to the individual, it becomes very challenging. Because then I don’t know which third of the curve this patient is going to be occupying. And so, to use any of these classification systems and tell the patient a very exact precise-sounding number like your survival is likely to be 2.3 or 2.7 years, I think is a gross disservice to the patient, because it in no way can reflect an individual’s prognosis.

James M. Foran, MD, FRCP: I completely agree. I don’t think that it’s always that meaningful for a person to hear an actual specific number, because I’ll say the confidence intervals around those are so wide. Obviously, within the revised IPSS, or the revised scoring system from three years ago, at least we get more information on the role of the CBC [complete blood count] itself where it appreciates how low the ANC [absolute neutrophil count] is, or it appreciates how low the platelets are, it helps discriminate a little more finely. The cytogenetic grouping is a little more discriminating as well, so I think we can hone it down a little better. But, I still feel that’s kind of a Holy Grail for us to determine for an individual rather than for a population, what their prognosis and what the right strategy would be.

Azra Raza, MD: I’d like to add to that that the revised IPSS system takes the same three parameters and looks at them with a little more detail. So, for example, the blast percentage is now divided into 0%-2%, 3%-5%, 5%-10%, and 10%-20%. Well, one man’s ceiling is another man’s floor. One pathologist counts 2% and another is going to count 3%. And that changes the score in the system, which I think is within the range of human error, that kind of counting.

James M. Foran, MD, FRCP: No, I agree. I do teach our Fellows that nothing magical happens when the blast count goes from 9% to 10%. That’s somewhat arbitrary. That’s an excellent point. That’s not a definitive thing.

Azra Raza, MD: So what I do in my actual practice, which is I think, practically speaking, that’s very important for our colleagues to understand, that the way you are going to assign prognosis to an individual can only be, right now, by following their disease and the natural history of their disease. So, when a patient is diagnosed, please do not rush into start instituting therapies, because what looks like a very serious anemia, with one transfusion, can improve and then remain improved for the next six months. And you don’t have to start a disease-modifying drug immediately for such a patient. So my advice, especially for lower risk MDS patients, at diagnosis just follow them for a while with a supportive care measure only, so that you get some idea about what their natural history is going to look like. And, again, apply that rule of punctuated equilibrium. You don’t know how long they’re going to last in that equilibrium state before the next event happens.

James M. Foran, MD, FRCP: I was going to use that very term. I’m going to borrow that term if that’s okay with you.

Azra Raza, MD: Well, I borrowed it from Stephen Jay Gould, so be my guest.

James M. Foran, MD, FRCP: At a recent presentation, somebody asked me, ‘What are the two pearls that you would want somebody to know in MDS?’ And the first pearl was mirroring what you said, which is that, first, you have to be patient. You have to be patient to confirm the diagnosis, see the evolution of the disease, and see if your interventions or transfusions are working. And the second is, you actually have to monitor transfusion support. And I don’t think that’s done consistently in practice where we know if there’s a change in the transfusion burden, for instance, or a change in ferritin over time, or a deterioration of the CBC, until somebody flags it in practice. So, that’s the other innovation I would have recommended. I really like your point about being patient. I think that’s a critical one.

Transcript Edited for Clarity