How We Handle Patients With Myelofibrosis in the Clinic

Andrew Kuykendall

Natasha, please walk us through how we generally approach treating patients with myelofibrosis when we see them in the clinic.

Natasha L. Johnson

When these patients are first seen in clinic, it’s by the physician, who reviews if they’ve been accurately diagnosed; performs risk stratification and determines if their disease is low-, intermediate-, or high-risk; and then we evaluate symptoms. When we look at patients who have myelofibrosis, it's common for them to have an enlarged spleen and associated symptoms; it’s also common for them to have cytopenias. These risk and symptom evaluations help to guide our treatment options and discussions with our patients.

Andrew Kuykendall

I think that we focus on two factors: we think about patients’ risk in terms of whether their disease is low-, intermediate-, or high-risk; I view that as a separate question than when we look at their spleen and symptoms.

So initially, we're thinking about how risky their disease is, and the main reason we're doing that is to evaluate whether they'd be a candidate for an allogeneic stem cell transplant or a curative option. Allogeneic stem cell transplant certainly offers the only curative option we have, but it's fairly dangerous as far as the risks of the procedure. Oftentimes, we're quoting risks of a 15% to 20% mortality rate from the transplant within the first year for non–disease-related reasons, which can be quite overwhelming for patients. In patients who are intermediate- to high-risk based on one—or hopefully a few—of these prognostic models, those are patients where the disease warrants at least consideration of transplant. However, patient-related factors must be considered as well. So many of patients with myelofibrosis are older and have multiple comorbidities; they may not be great transplant candidates. Of course, there’s a degree of risk, which is why we have prognostic models—to assess riskiness of the disease as to whether a transplant would be an appropriate thing to consider. Typically, we're reserving transplant for patients who have an estimated survival of less than 5 years—where that risk that I mentioned earlier may be worth trying to attain that long-term benefit.

As you mentioned, the next things we’re thinking about are spleen symptoms or anemia, and that's where we try to assess what types of treatment options may be more important. Historically, when we've thought about spleen and symptoms, we have favored a Janus kinase (JAK) inhibition approach. Right now, we have four approved JAK inhibitors in our treatment armamentarium, and all of them are very good at improving spleen volume as well as symptoms.

But then there's this subset of patients who don’t have concerning splenomegaly, symptomatic splenomegaly, or constitutional symptoms, where we're looking more at treating anemia, and that's where we don't really have as good of treatment options. Historically, JAK inhibitors like ruxolitinib and fedratinib have worsened anemia and thrombocytopenia, so they haven't been great options for those patients.

Here’s a question for you: what proportion of the patients that you're seeing have splenomegaly or disease symptoms as their most prominent issue compared with how many struggle with anemia as their chief complaint?

Natasha L. Johnson

About 40% of patients with myelofibrosis are presenting with anemia and virtually all will develop anemia throughout the disease course. About 25% are transfusion dependent at the time of their diagnosis; within a year, that increases to about 50% becoming transfusion dependent. However, the majority of patients are presenting with spleen and symptom burden over cytopenias in clinic.

Andrew Kuykendall

I agree with you. When you look at the data, you see that splenomegaly, to some extent, is present in probably 90% of patients with myelofibrosis. Not all of those patients have massive splenomegaly or symptomatic splenomegaly, but it really is a key feature of the disease.

As you mentioned, anemia is also quite common at diagnosis, but also something that we need to think about as being part of the natural history of this disease. Although it may not be something you have to deal with immediately, it’s probably going to be something you have to deal with down the road.

It's easy to say, “Oh, do you have splenomegaly and symptoms, or do you have anemia? Let’s go after one of those.” But the problem is many of these patients have both. If I’m saying 90% of patients have splenomegaly and 100% of patients are going to have anemia, that means that there’s a huge overlap between the two. Historically, that’s been a challenge. So I’d ask you: prior to recent approvals of medications—in the era where we had just ruxolitinib and even fedratinib—what was our approach to patients that we started on ruxolitinib or fedratinib who were anemic? How did we go about treating those folks?

Natasha L. Johnson

I’ve actually been thinking about this recently! Before the recent JAK inhibitor approvals that really benefit the patients with cytopenias, our course would be, for example, to use ruxolitinib and keep them on treatment despite worsening anemia to get the most benefit possible, because we really didn’t have other options. So regardless of whether they had anemia to begin with when they started ruxolitinib or fedratinib, or if they developed it after initiating treatment, we would look at adding on alternative therapeutic options to improve that anemia and/or supporting them with transfusions.

For example, many times with our patients who had anemia, whether they were on or off ruxolitinib or fedratinib, we could add on things like erythropoiesis-stimulating agents—they would be off-label, of course. We would also add on danazol, an androgen, to help improve hemoglobin in those patients. We might try immunomodulating agents with thalidomide or lenalidomide, given with a steroid. Those historically have been our main approaches when trying to improve anemia in our patients with myelofibrosis.

Andrew Kuykendall

It really was kind of a hodgepodge of things we could try to add on to get some sort of benefit for the anemia. Historically, we've known that ruxolitinib—the treatment we gave most often—worsens anemia. But there are publications that demonstrate that ruxolitinib-induced anemia doesn’t impact survival, suggesting we should differentiate between disease-specific and treatment-induced anemia, since the latter clearly has a detrimental impact on survival, correct?¹

Natasha L. Johnson

That is correct. Anemia has been shown to be an inferior prognostic variable in all of the risk-stratification scoring systems, and anemia and transfusion dependence are associated with worse overall survival in this patient population. Some of the risk scores actually put increased weight when patients are more severely anemic and transfusion dependent.

Andrew Kuykendall

It's interesting to go back and look at the development of these prognostic models. The International Prognostic Scoring System (IPSS), the Dynamic International Prognostic Scoring System (DIPSS), and the DIPSS plus, give equal weight to various clinical factors: age, leukocytosis, peripheral blasts, constitutional symptoms, and anemia—everything gets one point, although the DIPSS assigns an extra point if you're anemic. The DIPSS plus adds in unfavorable karyotype, thrombocytopenia evidenced by a platelet count less than 100 X 10⁹, and transfusion dependence. You get two points for being anemic, and you get an extra point if you require transfusion, which brings the total potential point assignment to three plus one for thrombocytopenia. We know that anemia is quite important and prognostically significant for patients with myelofibrosis. I think the question has been: if we induce anemia with a treatment, is that different than disease-related anemia? Fortunately, it seems that it’s somewhat benign, as ruxolitinib-induced anemia doesn't seem to worsen prognosis. The flip side of that, however, is: what if we can improve anemia? What if we are able to make a patient’s anemia better—does that get rid of the prognostic significance of the anemia? Are we able to improve outcomes? I think that's kind of been the unanswered question, because we didn't know if we were doing anything to improve outcomes necessarily by adding danazol or erythropoiesis-stimulating agents or lenalidomide or thalidomide or even luspatercept. We felt like it was probably improving patients’ quality of life because they didn't have to get transfusions as often, and maybe they were getting some improvements in fatigue, but we really didn't know how that translated to long-term outcomes. That was kind of our approach for a long time: let's focus on improving splenomegaly and symptoms and we'll kind of punt the anemia. If it develops as a result of treatment, we’ll try various things that may provide some benefit.

That's kind of a discordant, conflicting approach, because on one hand, we're doubling down on the prognostic significance of anemia with each subsequent prognostic model, but on the other hand, we accept that our most commonly used JAK inhibitors provide benefit at the expense of anemia and thrombocytopenia. I think that that's been something we have to carefully consider moving forward.