Management of Myeloproliferative Neoplasms: A Focus on Polycythemia Vera and Myelofibrosis - Episode 11

Diagnosis and Risk Stratification in Myelofibrosis


Harry P. Erba, MD, PhD: Let’s turn our attention to the other MPN [myeloproliferative neoplasm] we were going to talk about, and that’s myelofibrosis [MF]. Mary Frances, I’m going to look to you to discuss how these patients present and what the diagnostic evaluation is.

Mary Frances McMullin, MD, FRCP, FRCPath: These patients present in many different ways. They tend to have low blood count symptoms and splenomegaly. Splenomegaly may be very prominent in patients with myelofibrosis. It can turn up with somebody who’s just been discovered to have a slightly low blood count requiring investigation. I’ve seen patients turning up to surgical clinics because of this big lump in their abdomen, which is the spleen. They may certainly turn up complaining because of their symptoms. These people usually feel very tired; night sweats are a prominent issue, as well as early satiety. They can come down any route for investigation to find out what is wrong.

What is the pathway to investigation? We’ll repeat the full blood count. Examination of a blood film is extremely important. We’re looking for the classical teardrop cells. These patients all need a bone marrow biopsy, because you’ve got to have the bone marrow to look for the myelofibrosis. We also look in the bone marrow for other reasons, because this is a group of patients where the examination of the bone marrow looking for other things causing the fibrosis is extremely important. Obviously, we conduct an assessment of them clinically, including splenomegaly.

Harry P. Erba, MD, PhD: I find it interesting that it’s really primarily in terms of marrow morphology, or the hyperplasia and atypia of the megakaryocytes, that we make the diagnosis. You may not even see fibrosis.

Mary Frances McMullin, MD, FRCP, FRCPath: Yes, fibrosis is a secondary event. Although we call it myelofibrosis, you have to remember that it’s a secondary event that the abnormal megakaryocytes produce the myelofibrosis. You can get fibrosis in many other conditions, which is why you have to look and make sure that there’s not something else going on.

Harry P. Erba, MD, PhD: In the cellular prefibrotic phase of MF, do you approach it the same way as MF?

Mary Frances McMullin, MD, FRCP, FRCPath: There’s still some dispute as to whether there is such a thing as prefibrotic myelofibrosis. That is the pathology diagnosis, which has been defined by a few people, and they have to be experts to look and see it. However, I do think there are some patients who have quite abnormal megakaryocytes. You can almost look at them and say, “Well, if there is a prefibrotic myelofibrosis, this is it.” It’s a difficult diagnosis to make.

Also, there is nothing out there as to how these prefibrotic myelofibrosis patients will progress and how they should be treated. No one has ever done a trial in prefibrotic myelofibrosis. I think it’s very difficult, because if you put this label on somebody, the word that will predominate is the fibrosis word. They can live for many years without a problem.

Harry P. Erba, MD, PhD: Morphologically, it seems to be similar to ET [essential thrombocythemia]. I’m not sure how they make that distinction.

Mary Frances McMullin, MD, FRCP, FRCPath: Yes. That’s the issue.

Harry P. Erba, MD, PhD: Rami, tell us about risk stratification in MF.

Rami Komrokji, MD: It’s getting very complicated. They have to memorize around 10 prognostic models now. The idea is, why do we do risk stratification in those patients? Obviously, it has had a prognostic value for the patients and then their families, but we tailor the treatment based on those risk factors. Namely we are thinking, for patients who are candidates for allogeneic stem cell transplant, when is the disease risk high enough to justify a procedure like allogeneic stem cell transplant, because it remains the only cure for those patients?

A subset of those patients with myelofibrosis are obviously eligible for transplant. As we, nowadays, can find alternative donor sources and the age of transplant is extending, more and more patients are eligible for the procedure. That’s one of the main goals when we risk stratify patients. It’s to decide whether the patient should proceed to transplant immediately or not.

This evolved with historically used clinical variables. We’ve had the IPSS, the International Prognosis Scoring System, where we looked at variables such as age, anemia, and leukocytosis. Symptoms, actually, had always come as a prognostic factor in patients with myeloproliferative diseases.

Then we give a score based on those clinical variables and we stage patients into a group labeled low, intermediate 1, intermediate 2, or high risk. Based on that, those patients who are in intermediate 2 or high-risk group, if they are eligible for transplant, we usually recommend for them to proceed to transplant.

There are different variations of that clinical model. The DIPSS [Dynamic International Prognosis Scoring System], the dynamic model, weighs anemia and transfusion. Then there is the DIPSS Plus that also weighs thrombocytopenia, as well as some cytogenetic information. Then, we started realizing the importance of the molecular data and somatic mutations. We had different versions of that. We have a model called GIPSS [Genetically Inspired Prognosis Scoring System], which is based solely on genetic information; no clinical information appeared in that model.

Now we have a model particularly for patients who we consider for transplant below the age of 70 called Molecular IPSS, or Molecular IPSS70, with a couple of versions.

Again, it gets complicated, but the idea is that there are certain clinical variables we’ve known a long time, like anemia, the presence of circulating blasts, and leukocytosis symptoms, which are always important. Then there is molecular information, such as regular G-banding [Giemsa banding], and now, molecular information—mutations such as SRSF2 and ASXL1. Those are associated with worse outcomes.

It gets complicated, but I think the idea is to look at those factors, and even as Ruben mentioned earlier, many of those are continuing. They cannot be just thought of as arbitrary staging systems or cutoffs. Then weigh those to decide how to tailor the treatment.

In terms of where the patients come in those staging systems, the majority will be in the intermediate and higher-risk groups. The patients who are truly very low risk are probably 10% to 20%. Most of the patients will be in intermediate or higher risk. If patients are not going to transplant, I think our decision for treatment is still based on symptoms, not on the stage, per se.

We also try to do this exercise looking at which one wins, out of all those models, and I don’t think there is a winner, per se. We looked at patients where there was discrepancy in 2 stages between models. For example, the genetic-inspired model of the GIPSS did do a little bit better than the clinical ones, but I don’t think there is really 1 winner or 1 model that tells you all. There is a value in using a transplant-specific model, because the other models, for example, had weight on age. If somebody who was 70 got a risk factor, and he was thinking of transplant, it was always puzzling for me. Now, because the patient is 70, I would upstage him and recommend a transplant for him. I think, for example, in deciding for transplant, we have to take age out of the equation.

Harry P. Erba, MD, PhD: That’s what the new model does. It takes age out up to 70, but adds in these molecular features.

Rami Komrokji, MD: Exactly.

Harry P. Erba, MD, PhD: Do all of those models apply to PV [polycythemia vera] and ET that have progressed to MF, or are there different models?

Rami Komrokji, MD: They are applied. There had been specific models developed. We actually contributed patients to the Italian group. We give them information from our database, as well. They’ve had a big collaboration where they looked at patients that had secondary MF from ET or PV, so they have a specific model called the MYSEC [Myelofibrosis Secondary to PV or ET] that looks particularly at that. I think the other models also have been used for patients with secondary MF, as well.

Harry P. Erba, MD, PhD: I want to jump to this transplant issue. How do you identify a patient for a transplant? Then, one of the challenges I find in transplant for these patients is, they have a big spleen. How do you take care of that spleen before you send the patient for transplant?

Mary Frances McMullin, MD, FRCP, FRCPath: Our team attempts to find patients for transplant. It’s the only curative option, and I think the possibility has to be considered upfront in each and every patient. The first group you run across that you throw out are the low-risk ones because they have a median survival of 15 years. Are you going to send them for a transplant with a 20% mortality? After that, you have to take into consideration the patient and what they’re like, and we’ve discussed age.

However, transplant in these patients is a major procedure. Many people will agree that patients with myelofibrosis do badly with transplant, so it’s a big decision. Then, do you wait until they’re at a stage where it’s not a benefit to them?

It has to be considered, and there’s not an easy answer. I’ve heard people saying that it only becomes an option for about 5% of myelofibrosis patients. That’s probably creeping up as older people get transplanted with different issues, but it’s difficult.

I think this issue of the big spleen is easy, in a way. Now, everybody would put them on ruxolitinib first, and try and shrink the spleen before they get to transplant. You will often improve their condition greatly by doing that, so I think it’s standard practice, and I tell them ruxolitinib until just before the transplant.

Harry P. Erba, MD, PhD: Have other panel members used that? Is that very successful, or do you have to go to splenectomy? Ruben?

Ruben A. Mesa, MD, FACP: It’s an excellent question. I’d say that the standard certainly is that these patients are already on ruxolitinib by the time they go to transplant. It certainly is in the United States, where the majority of eligible patients for ruxolitinib are probably on ruxolitinib. Even if a patient were diagnosed tomorrow and you were thinking transplant, it is always a 3-to-6-month process, at a minimum, in terms of typing, getting them on a schedule, and pretransplant work-up. It gives you that run-in time to have some response to ruxolitinib, hopefully improve their spleen size, and also improve the state of the patient—their activity level, their physical robustness—as well as the spleen.

Transcript Edited for Clarity