Advances in the Treatment of Myelofibrosis - Episode 3
Criteria used to risk stratify and classify patients with newly diagnosed myelofibrosis and establish treatment paradigms.
John Mascarenhas, MD: We're going to shift gears now and talk about risk stratification and patient classification. When you see patients with MF [myelofibrosis], how are you generally classifying them? I'll start with you again, Srdan. When we see oncology patients, the first question you get asked is, “What stage of cancer do I have? Is it stage I or is it metastatic? Is it stage IV?” Of course, with myelofibrosis, it doesn't work that way. MF is a little bit more complicated and nebulous, but how do you divide patients, or segregate them, and for what purpose?
Srdan Verstovsek, MD, PhD: This is actually a really good point because many are fixed on the grade of fibrosis in the bone marrow and say, “Hey, I have a grade 1 or 2 or 3. This is bad, right?” Patients think that if they have grade 3 then they don't have 2 or 1, and that is not the case. The fibrosis grade may have some significance on the outcome of the patient but there are many other factors, as we said before. Prognostication, in terms of life expectancy, does not depend on the degree of fibrosis solely it may be just one of the many factors. We look at patients’ overall characteristics, genetic characteristics, chromosomal characteristics, and certainly comorbidities that you brought up also matter in deciding when to intervene. The number 1 factor would be assessing the quality of life and seeing whether we can control that quality of life and the […decisions] usually associated with that, are if there is bone marrow failure, or anemia is present and whether we can provide patients medications to control the signs and symptoms. However, the most important part is to see whether the patients should be referred to a bone marrow transplant doctor. The prognostication part is used in our clinic specifically, really to assess the risk of dying and if the assessment by any of these prognosis-scoring systems is less than 5 years, we would refer patients to the transplant. That is not the best measure; the risk of dying would affect how I treat patients. Even the patients who may have a good prognosis, but suffer from bad quality of life, I treat the patients who have bad quality of life.
John Mascarenhas, MD: Aaron, when you think of myelofibrosis, which I think at this point, is obvious, it’s a very heterogeneous disease. Patients can present in various ways and their clinical courses can be quite different, but generally speaking, other than the risk scoring systems that we've talked about, and as Srdan pointed out, there's a number they've gotten more complicated as we've integrated cytogenomics into these risk scoring systems… Generally speaking, are there other paradigms of segregating patients in myelofibrosis or separating them? I guess what I'm leading into is the concept of a patient who might have more of a cytopenic bone marrow failure type-picture that might be somewhat like a myelodysplastic syndrome that Raajit mentioned before versus someone who might have more of a proliferative component. Almost like there are 2 types of CML [chronic myelogenous leukemia] where 1 could be proliferative, and 1 could be dysplastic. In myelofibrosis, does that ring true to you? How do you envision that?
Aaron T. Gerds, MD: I have a slide that I often throw in my slide deck where it shows all the myeloid disorders are kind of a smear, a palette of color, all over the place. We like to put things into discrete bins, but that certainly doesn't happen. MDS [myelodysplastic syndromes] abuts MPNs [myeloproliferative neoplasms], the overlaps, [and] acute leukemias are all in the mix. As you would imagine, patients with myelofibrosis come in all kinds of different shapes and sizes, and I'm not meaning physical size, but rather their disease course and tempo and the presentation you may see. Certainly, there's an emerging concept of this kind of proliferative myelofibrosis; patients that have high JAK/STAT allele mutations and allele burdens are often patients that have myelofibrosis out of PV [polycythemia vera] and ET [thrombocythemia] that come in with higher white counts, higher platelet counts, maybe different cytokine profiles, and this more dysplastic or cytopenic or empty kind of myelofibrosis patients that often have lower allele burdens of their JAK/STAT mutation, or have a more empty or less cellular marrow and these other features with cytopenias as their hallmark. There are certainly 2 bins of patients that we see. They may have different disease courses. When we talk about all these prognostic models, as much as we love to add genomics and cytogenetics in there, good old-fashioned risk variables such as anemia and thrombocytopenia don't really seem to fall away over time because they are probably really good predictors of disease. It's probably not the anemia itself, but the anemia stands for something that's going on with the disease, so I do think it's powerful in prediction. I think from a more practical standpoint, patients with more cytopenic myelofibrosis are just more difficult to treat. You’ve got more “moles” you're whacking in the “whack-a-mole.” You're not only whacking the symptoms and the spleen, but you're also trying to keep up their blood counts at the same time. As Srdan mentioned, if you also get a whiff of some high-risk mutations, you may be more tempted to send that patient off to a transplanter because that's the thing that may fix it all. Certainly, we think about these things. No 2 patients are alike.
John Mascarenhas, MD: I think you brought up a good point though, that the patients with cytopenic MF, and these are the ones that are typified by cytopenias like anemia, which in most prognostic models is one of the heavier weighted risk factors, as well as thrombocytopenia. They often track together, and of course, these are the patients who tend to score higher on the risk scores because of anemia and thrombocytopenia. There is an intimate link, often between the clinical picture, as you said, which is the simplest without even getting cytogenomic data, and you can kind of already get a sense of where a patient falls and the risk score that they're likely going to have.
Transcript edited for clarity.