Prognosis and Risk Stratification of Myelofibrosis
Moshe Talpaz, MD: Myelofibrosis has a very heterogeneous course and a very heterogeneous course of outcome. Patients can live anywhere from 1 year to 15 years or longer. And it’s important to differentiate between patient groups who have a better outlook and patient groups who have a worse outlook because that will influence the decision on treatment, experimental treatment, bone marrow transplant, and so forth. Several clinical tools have been made available to try to identify a clinical course for the patients. Such models are known as IPSS [International Prognostic Scoring System], DIPSS [Dynamic IPSS], and DIPSS-Plus, and we commonly use either the IPSS for patients who come to see us for the first time or DIPSS, which is a model that can be applied to patients at any time they are being seen.
These models are based on clinical features, specifically age of the patients. Patients older than 65 years have a higher risk and a worse outcome. Patients who have anemia with hemoglobin of less than 10 tend to have a more aggressive course. Patients with immature cells in the blood known as blasts are also patients with a higher risk of progression, and patients with an elevated white count of greater than 25,000 also represent a high-risk group of patients. So using those clinical parameters, we are able to identify patients who have a better outlook and patients who have a worse outlook and assign treatment according to these prognostic models.
Ruben Mesa, MD, FACP, director of the UT Health San Antonio MD Anderson Cancer Center: The risk stratification for myelofibrosis is something that continues to evolve. Having started with the IPSS, which included primarily clinical features such as cytopenias and symptoms, and any increase in blast, it really has evolved to a great degree. It moved to the Dynamic International Prognostic Scoring System and now includes many new scores with many new features. The one I would highlight is, in particular, the MIPSS [Mutation-Enhanced IPSS], or the MIPSS70. This is available online for people to be able to use. I share it with my trainees. There are now so many prognostic scores across not only hematologic malignancies but other cancers. The key is to know where to find these scores as opposed to memorizing them all.
The important parts are…the features that tell us about the disease. Once cytopenias have an adverse impact to increase blast is just moving toward acute leukemia. Third, symptoms and fibrosis both have a negative prognostic impact. And fourth, the incremental part is the impact of high-risk molecular features. Increasingly we are well aware that ASXL1 and EZHS1/2, as well as a variable additional group of additional somatic mutations, are important in confirming a negative prognostic score.
So being able to weave all those different elements into the score is helpful. You can log on to that, the website, for that score. The utilization of those scores is particularly beneficial as we consider the treatment option of the stem cell transplantation, as this is likely the best predictor of survival in patients with myelofibrosis.
When I visit with a patient with myelofibrosis, I’m mindful of 2 different sets of parameters. First, what is their risk? Or, what is their expected survival? And the scores as they stand currently are specifically aiming at measuring survival, whether that is eventual potential death through myelofibrosis, that way the patients who may pass away, can include debilitation, risk of pneumonia, heart failure, worsening spleen symptoms, etc.
The other way that patients can pass away from the disease is to transform to acute myeloid leukemia. That probably has a different natural history. It’s a much more abrupt life-threatening period if that occurs. That is, I think it is more heavily influenced by the presence of increased blasts and the high-risk molecular features present in the marrow. So I think the prognostic scores really separate us in 2 different ways.
The second set of parameters that I am mindful is: What is the burden that the patient faces as I make these treatment decisions? The survival of the patient is not the only factor that really needs to be considered. Patients with myelofibrosis may suffer from difficult disease-related symptoms; they may suffer from [splenomegaly]. They may suffer from cytopenias and the consequences of those cytopenias. And they clearly can suffer from risk of progression that can lead to a decrease in survival.
I am mindful of both sets of parameters as I try to make a treatment decision, and that’s solely the number of calendar days that a patient may live is not the only parameter to be considered.
The distribution of patients by risk in myelofibrosis depends largely on the score. It’s a relatively equal distribution between low, the intermediate categories, and high risk; it depends somewhat on the actual instrument. I would say the majority of patients are intermediate and high risk as we evaluate them by DIPSS or the more up-to-date methodologies.
I believe that clinicians sometimes underestimate the risk that patients with myelofibrosis are experiencing. This can be influenced, I think, by several factors. In a busy hematology/medical oncology practice, they may not initially appear as the patients with advanced solid tumors or aggressive more acute illnesses such as acute leukemia, diffuse large-cell lymphoma, or very active myeloma.
Several of the features that we have identified that are high risk—increased blasts, cytopenias, transfusion dependence, high-risk molecular features—I think, sometimes are not recognized for the severity of the impact on survival early enough by treating clinicians. What this translates to is, I think, frequently a delay in either initiation of therapy or a delay in consideration of stem cell transplants in appropriate candidates.