Rami Komrokji, MD: Jeanne, one of the crucial steps after we establish the diagnosis is risk stratification. It probably has more implications in myelofibrosis than other MPNs [myeloproliferative neoplasms]. There are so many models out there that could be confusing. Can you walk us through how you risk stratify and how you’re going to tailor the treatment based on that?

Jeanne Palmer, MD: Absolutely. That’s a great question. One of the things that’s hardest when you meet these patients is making sure you set the stage appropriately. At the end of the day, especially if they’re younger, you want them to have a transplant somewhere down the road but maybe not at that time.

As I walk them through the different risk stratifications, I look at the DIPSS [Dynamic International Prognostic Scoring System] score, which was published in 2009. It can be used very easily. You can do it in your head in the clinic. That takes into account things like age, anemia, white blood cell count, blast percentage, and symptom burden. With that, you can stratify people to low risk, intermediate 1 or 2, or high-risk disease. This is one of the main things I look at.

For transplant, it has been proven in multiple areas that these patients, especially if they’re intermediate 2 or higher, are probably at the appropriate time for transplant. Although as we talk more about these risk stratifications, there are patients who are maybe intermediate 1 who’d be appropriate. But this risk stratification is tried and true. We have to remember is that this risk stratification—when we use those survival curves—was generated before ruxolitinib was commercially available. Chances are if we started to look back and redid the DIPSS modeling, we might find that the survival curves are a little different.

In the next level, there’s DIPSS Plus. That takes into account the DIPSS score and adds thrombocytopenia, cytogenetic abnormalities, and the need for red blood cell transfusions, all of which are critical. Sometimes the need for red blood cell transfusions alone, if somebody doesn’t necessarily have the high white blood cell count, is something that makes me consider moving forward with transplant. DIPSS Plus can be very helpful.

The risk of the driver mutation is important. We know that calreticulin is good risk, so patients who have the calreticulin mutation can have very prolonged survivals. JAK2 and MPL are in the middle. If they don’t have any driver mutation, that represents higher-risk disease, and we consider these triple-negative patients.

Finally, the next step is looking at the MIPSS [Mutation-Enhanced International Prognostic Scoring System] score, or the mutational profiling. This is where we look for somatic mutations. This is what we call next-generation sequencing [NGS], or IntelliGEN. There are all sorts of names for it. This test looks for any number of mutations, depending on the panel you use. These mutations can include ASXL1, EZH2, IDH1, and IDH2. These mutations can also convey risk. We know this based on some studies that were done.

The MIPSS score takes into account a number of variables, including these high-risk molecular mutations as well as cytogenetics. The MIPSS score can be very helpful, especially in a younger person you’re considering for transplant. They might look like they have intermediate 1, lower-risk disease, but then they have 2 or 3 high-risk mutations, and you go, “This is giving me a prediction model.” This is more for younger patients. Some of these mutations can also give potential therapies, especially if their disease progresses. But this mutation profile can be very helpful when counseling a patient, when talking about transplant.

It’s important with most patients to go through this entire model and say, “This is what we’re looking at.” Based on that, you can say what changes will make you say that we need to move to therapy or that we need to be thinking about proceeding with transplant. As both a transplant physician and a myelofibrosis physician, that ends up being a very nice segue and helps me set the stage so you aren’t waiting until they get to the point where they need transplant and are throwing it at them. Looking at the big picture is such a critical component when you first diagnose somebody with myelofibrosis.

The only time that I struggle with sending the next-generation sequencing—I’d be interested to see what my colleagues think—is in older patients who aren’t going to be transplant candidates. When I first meet them, do I want to say, “You have a couple of high-risk mutations, but I don’t know that I can do anything about it?” I often end up sending it. There’s enough knowledge about this that people want it sent. But I say that I struggle with it because it sometimes gives us information that generates more anxiety and stress and doesn’t necessarily give us a treatment option.

Ruben Mesa, MD: Let me chime in, because you brought up a couple of important points. First, in terms of treatment planning, the assessment of risk is the beginning but not the end, meaning that the disease burden and risk are key. There are individuals who may have a life expectancy of 20 years, but if they feel poorly, they still require therapy. Both are important.

Second, you brought up a very relevant point: don’t quote for any patient their expected survival based on the old DIPSS curves. They’re probably completely inaccurate at this point in terms of length of life. The relative differences may be important in terms of what’s a higher-risk feature vs not, but what one might expect for length of life is probably out the window. The third is what you raised about NGS. It’s an evolving science. It’s helpful. But you’re right, it certainly ranges from whether it impacts what you’re going to do now vs whether it might impact what you’re going to do later. It might be necessary later if we don’t have it now.

I try to share with patients that if you have a higher-risk marker, you may do worse, but we don’t necessarily know in comparison with what. If you’re a patient with ET [essential thrombocythemia] and you find that you have a TP53 mutation, it probably isn’t a good thing. But what does it mean exactly? Does it mean that your disease is going to progress in 15 years vs you would never progress if you didn’t have it? That’s possible. It’s a very delicate thing, because for some patients, it can create an overwhelming amount of anxiety and stress over this uncertainty. But we have great uncertainty. What does it mean in absolute terms?

Jamile Shammo, MD, FACP, FASCP: I tend to agree. Your question is a good one. Perhaps in an older individual where transplant isn’t considered, I might do it in accelerated phase. Then you can identify targetable mutations, maybe an IDH1 or IDH2, and that could be helpful therapeutically.

Rami Komrokji, MD: To summarize, for a non-MPN specialist, it’s confusing. There are 5 or 6 clinical models and 3 or 4 genomic models. But the idea that you’re hearing about is basically that there are certain clinical variables that are important in myelofibrosis symptoms. It’s one of the diseases that incorporates symptoms. Patients have constitutional symptoms, leukocytosis, anemia, and circulating blood. Those are features in all those models. Then you add some molecular profile.

It’s very hard to go through all those risk models, but the variables are common. We have clinical variables that are known to be associated with worse outcome. When somebody is anemic, they’re automatically almost intermediate risk by any model. The same is with certain mutations, as mentioned. None of those models test the tempo of the disease or the kinetics. Sometimes I’ll see patients, and I’ll say, “You‘re in this risk category, but let’s wait and see the coming 6 months.” Not all patients, even in the same risk category, have the same trajectory of the disease. Sometimes I’ll observe only to get a sense of how the patient’s disease is behaving. That’s key. As you mentioned, the disease burden is weighted in some of the clinical risk models, but not heavily.

Ruben Mesa, MD: That’s very true. I have many patients who are engineers who will derive the mathematical formula of what to expect. Having been an engineer before medicine and being from both that world and our world, I try to share with them that biology isn’t linear. You’re right. You might have 15 years, but you might have 14½ years where things are pretty good and then the last 6 months where the train comes off the rails. That’s very important, and it might have huge implications in terms of management. Some of those folks are better off being observed for a period of time and then intervening later. It’s very much not a linear disease.

Rami Komrokji, MD: I can spot those engineers in a second. I walk into the room and I say, “You’re an engineer.” They’re like, “How did you know?” Excel sheets.

Ruben Mesa, MD: The graph paper. The mechanical pencil.

Jeanne Palmer, MD: Yes. One thing I’ll also comment on is that for patients who have high-risk mutations, a lot of times it generates a huge amount of anxiety and stress. They go online and go on all these MPN blogs, and everyone talks about these mutations. The 1 thing I use them for, I say, “These tell me to jump faster.” I’m going to make a change, especially with regard to transplant if they have these high-risk mutations. I’ll probably jump faster at changes and say we need to go transplant rather than giving more time for them to think about it or to observe.

One thing that’s important is that the mutations themselves don’t make a decision for me. It’s the mutations in addition to the clinical parameters that we see. You can do the MIPSS, MYSEC-PM [Myelofibrosis Secondary to Polycythemia Vera and Essential Thrombocytopenia–Prognostic Model], DIPSS, GIPPS [Genetically Inspired International Prognostic Scoring System], and everything else. There are tons of them. But at the end of the day, it’s more of a big clinical picture. Sometimes I pull up the calculators online and I show them. I say, “Your blasts were 2% this day and 3% this day. Look at what it does to your risk score,” just to show them how changeable these can be.

Transcript edited for clarity.