Hello and thank you for joining us today for this 

, which is focused on myeloproliferative neoplasms, or MPNs. In today’s presentation, my colleagues and I will review 4 clinical cases. We will discuss an individualized approach to treatment for each patient and will review key clinical trial data that affect our decisions.

I am Dr Srdan Verstovsek, from the University of Texas MD Anderson Cancer Center in Houston. Today, I’m joined by my colleagues and friends, Dr Prithviraj Bose, from the University of Texas MD Anderson here in Houston as well and Dr Rami Komrokji, from Moffitt Cancer Center in Tampa, Florida, as well as Andrew Kuykendall, from Moffitt Cancer Center in Tampa, Florida.

Thank you all for joining us today. Let’s get started with our first case.

Good morning. Our first case is a 68-year-old lady, who presents to her physician with symptoms of fatigue and abdominal pain lasting over the last 4 months. She also had some increased bruising. On physical examination, the spleen was palpable at 8 centimeters below the costal margin.

A CBC [complete blood count] showed a hemoglobin level of 9.1 g/dL; platelets, 98; and a white blood cell count of 32,000. She also was noted to have circulating peripheral blasts at 2%. She eventually undergoes a bone marrow aspirate biopsy that shows megakaryocyte proliferation and atypia, with some evidence of reticulin fibrosis. The peripheral blood smear shows leukoerythroblastosis, and the genetic testing shows JAK2 V617F mutation. This patient obviously meets the criteria for myelofibrosis. She does have the presence of the megakaryocyte proliferation on the bone marrow and the clonal markers. There are no other hematological diseases. She does have anemia, spleen, and a leukoerythroblastic picture.

Case 1: A 68-Year-Old Woman With Primary Myelofibrosis

Once we established the diagnosis, our next step is risk stratification because we are going to tailor the treatment based on that. In myelofibrosis over the past few years, there have been several models that have evolved in risk stratification. Some of them are purely clinical, and some of the newer models incorporate some of the molecular and genetic material within the risk stratification. The classical models have been the IPSS, the International Prognostic Scoring System, or the Dynamic IPSS or the Dynamic IPSS-Plus. All of those will weigh on tentative factors such as patient age, anemia, leukocytosis, and presence of blasts—constitutional symptoms. Some of them, in addition, like the DIPSS will allow for the model weighing on the anemia and the blood cell transfusion. The DIPSS-Plus will account for the thrombocytopenia as well as the karyotype. We can generally divide patients into 4 risk groups, from low, intermediate-1, intermediate-2, to highest, and the median survival that is according to those risk groups, from somewhere between 1 to 2 years for somebody who is labeled as a high risk, to someone that has a median overall survival exceeding 10 years if they are low risk.

So our patient, she’s above the age of 65, she’s anemic, has leukocytosis, some constitutional symptoms, and the platelets are below 100, she’s definitely going to be at higher risk using the IPSS or DIPSS-Plus.

In certain cases also as I mentioned, nowadays, we have some molecular models. There is the genetically inspired prognostic scoring system, or the MIPSS70, that had evolved particularly to allow risk stratification for patients we are considering for allogeneic stem-cell transplant, because it’s geared for those patients at the age that we think of transplant. It does not count the age as a risk factor. It incorporates presence of molecular mutations, whether the patient has calreticulin mutation or not, other high-risk mutations defined by presence of ASXL1, EZH2, SRSF2, or IDH1 or IDH2, and the newer version is actually also related to AF mutation as well. So based on clinical variables and molecule variables, we would put the patients in 1 out of 3 risk groups, the low, intermediate, or high, and the survival of these 3 will vary based on that group. If we take our patient here again, she probably would be high risk with the clinical variables and incorporating the absence of the calreticulin mutation and the bone marrow fibrosis.

There is another model that’s used mostly or primarily for patients for post-ET [essential thrombocythemia] or post-PV [

] myelofibrosis. This is the MYSEC Prognostic Model that was particularly developed in a cohort of patients that had secondary myelofibrosis. So this is myelofibrosis from either essential thrombocythemia or polycythemia vera. We see here the variables used there are the age, hemoglobin level, platelets, presence of circulating blasts, the calreticulin mutation status, as well as constitutional symptoms, and we can divide patients into 4 groups based on that. This model is available online and one can plug in the numbers and see the outcome for the patients nicely with that model.

Case 1: Risk Stratification Models for Myelofibrosis

Now, once we assess the risk, basically, the next step is really that we’re going to tailor the treatment, and there are 2 things we are thinking of. Obviously, presence of symptoms, and when we think of symptoms, we think what are the big dominant symptoms. Most of the time, we are dealing with splenomegaly and constitutional symptoms, and sometimes we are dealing with cytopenias. Sometimes, those are concomitant, so we try to decide on what is the most important or symptomatic for the patient. The other major issue we are thinking of is the allogeneic stem cell transplant. If the disease risk is high enough, then we’re going to be thinking of allogeneic stem cell transplant.

If we think of patients that may have low-risk disease, if those patients are asymptomatic, I think observation is a very appropriate first step for those patients or to consider a clinical trial. If those patients happen to be symptomatic, then based on the symptoms, particularly if they have splenomegaly and constitutional symptoms, ruxolitinib is probably a reasonable option. Interferon-based treatments can be considered in this early low-risk group, and in my opinion, that’s probably the best subtype of myelofibrosis that would benefit from interferon-based therapy. If patients are intermediate risk, intermediate-1 risk, again, we go by if they are symptomatic. Mostly, we are discussing splenomegaly and constitutional symptoms, then ruxolitinib at this point is standard of care for most of those patients or a clinical trial. If patients are asymptomatic, then we observe them. In a subset of patients with intermediate stress group, if they have other risk features, they are younger, we may entertain or discuss the role of allogeneic stem cell transplant for those patients.

Now, when we move to the intermediate or high-risk group, again, those are the patients we are thinking from the get-go, considering allogeneic stem cell transplant. If they are a candidate, we proceed to allogeneic stem-cell transplant. Sometimes, we will bridge those patients with JAK2 inhibitors because they could have a large spleen, they are symptomatic, and improving their splenomegaly and symptoms prior to transplant may improve their outcome. If patients are not candidates for allogeneic stem cell transplant, then we go again asking the question, what is the predominant symptom and what’s affecting the patients? If it’s the cytopenia, then we go through management of cytopenias, predominantly anemia, with limited options that we have in that area. If the dominant symptoms are again, splenomegaly, or constitutional symptoms, we are going to be considering JAK2 inhibitors. There is a subset of patients that will have thrombocytopenia, that it becomes a little bit more difficult to manage. However, when discussing how to dose ruxolitinib in that group, and nowadays, we have other new approved agents such as fedratinib, that can be considered in that subset of patients with thrombocytopenia.

Case 1: NCCN Guidelines for Treating Myelofibrosis

These are just some of the data with ruxolitinib from the COMFORT-II study. We know that there were 2 trials, the COMFORT-I and COMFORT-II, that looked at use of ruxolitinib in patients with intermediate- and high-risk myelofibrosis, and compared to Hydrea, our best available therapy. In reality, if you look at this, the majority of the patients will achieve a spleen response. Now, in the study, there was a requirement of 35% reduction by MRI [magnetic resonance imaging] volume of the spleen, which correlates typically to 50% between the spleen size, and that was achieved probably in half of the patients. But the majority of the patients will achieve some spleen reduction with ruxolitinib. There is no doubt that it’s a treatment that improves the functional status for the patients, their quality of life as well.

Now, what we’ve learned is that if patients have adverse mutations, that the response to ruxolitinib may not be as good. If you think of patients that have somatic mutations such as ASXL1, EZH2, or DNMT3A1, those patients did not do as well in terms of durability of response. That also matters by the number of mutations. However, if you look compared to Hydrea, the patients treated with ruxolitinib, even with high-risk mutations, tend to do better than those treated with Hydrea. However, it’s not as durable a response as obtained in those patients that don’t have those adverse somatic mutations.

When we think of the overall survival for patients on ruxolitinib, I think clearly those patients will drive survival benefit. This is pooled analysis from the COMFORT-I and COMFORT-II trials, showing the median overall survival around 5 years for patients treated with ruxolitinib versus 3.8 years for patients that were in the control arm, keeping in mind that there was a crossover in those studies. I always discuss with patients that the goal of therapy is not improving overall survival. There is no evidence that the ruxolitinib treatment will probably reverse the natural history of the disease. However, those patients’ symptoms improve, their spleen is reduced, their functional status is improved, their performances improve, and thus they drive survival advantage. So, if I have a patient who is intermediate or high risk, that is symptomatic with splenomegaly and constitutional symptoms, I do think that they derive survival advantage from the treatment.

When we look at the responses and the survival, correlating the spleen response with survival, we know that there is correlation between a spleen response and survival. When we look at the spleen response at 6 months and overall survival, there is correlation with that and correlation between those patients that have a stable response or an unstable response when compared to nonresponders. So, there is correlation between spleen volume reduction and overall survival for those patients.

One thing to keep in mind, it’s important to titrate the dose early for those patients. In many occasions, we shy away from using the appropriate doses of ruxolitinib. Sometimes, we are worried about the myelosuppression, and we start on a lower dose. In cases of borderline cytopenia, maybe that’s acceptable. However, I think the message is if the patients are tolerating the treatment, one should escalate quickly. You definitely do observe a dose-dependent spleen response. The higher the dose of the Jakafi, the higher the spleen reduction or the response. The symptom improvement typically starts at around the dose of 10 milligrams twice a day. What I do, if patients’ counts are appropriate, I start with the full dose. If the counts are borderline and I start them at the lower dose, within the first month or 2, based on their blood counts, I’ll try to escalate the dose to the maximum tolerated by the patient.

Now, obviously the major concern is the myelosuppression, where almost half of the patients will develop some anemia on treatment. Typically, the hemoglobin level will drop between 1 to 2 grams; around week 8 is where the nadir is. Thrombocytopenia can be observed as well, but less often, we will be dealing with neutropenia in those patients.

Now, there have been studies trying to look at patients with platelets between 50 to 100 x 109/L— the COMFORT-I and COMFORT-II study—platelets above 100 x 109/L. Subsequently, there were patients that had thrombocytopenia in the range of 50 to 100 x 109/L, that use of ruxolitinib was tested in those patients. It was determined that probably a 10-mg starting dose is safe in those patients, with the same expected thrombocytopenia and anemia that we observed in patients when their platelets were above 100 x 109/L. Sometimes, we would start at a 5-mg twice-daily dose in this group and escalate up to the 10-mg twice-daily dose, and if they are tolerating that, we could consider even further escalation in those patients.

When you look at the spleen response and the symptom improvement in those patients, the magnitude of benefit driven if one would go with this dose escalation policy is similar to what we expect or see in patients with platelets above 100 x 109/L.

Case 1: Ruxolitinib Treatment for Myelofibrosis

Now, recently, a second JAK2/FLT3 [Janus kinase/fms-like tyrosine kinase 3] inhibitor, fedratinib, was approved by the FDA based on 2 trials, the JAKARTA, that was an up-front study, and the JAKARTA2, that looked at patients after ruxolitinib failure. In the JAKARTA study, patients were stratified as intermediate 2 or high risk, and randomized between fedratinib at 2 different dose levels—the first being placebo; this led to the approval of this drug in this population. Fedratinib has a little bit of a different profile: It’s a JAK2/FLT3 inhibitor, with a toxicity profile a bit different from ruxolitinib.

When we look at the response rate in those patients treated with fedratinib at the 400 mg level, 36% of those patients achieved the primary endpoint which is a 35% reduction in spleen volume from baseline at 24 weeks. When you look at those patients that had platelets between 50 to 100 x 109 per liter, the response rates were similar—around 36%—and then around 1/3 of the patients had symptom improvement.

So regarding what we expect in terms of adverse effects, because of the FLT3 inhibition, we do observe more GI [gastrointestinal] toxicity. Those patients will have diarrhea, nausea and vomiting; that’s typically manageable in most of the patients. The rate of anemia and thrombocytopenia, when you look at the numbers, look equivalent or similar to ruxolitinib; however, one has to keep in mind that the study allowed patients below a platelet count of 100 x 109 per liter, with more cytopenia. So there is probably less—a little bit of cytopenia observed with fedratinib compared to ruxolitinib in those patients.

The fedratinib has a black box warning of Wernicke encephalopathy (WE). Originally, this was a valid concern; however, over meticulous review of those cases, it turned out that only a few handfuls, 3 cases or so, were true Wernicke encephalopathy. In many cases, this had been linked to thiamine deficiency—patients having nausea and vomiting, which leads to dehydration and malnourishment. Typically, we do check a baseline thiamine level in all of those patients, and we would consider thiamine replacement prior to starting fedratinib. It is an option for patients that may have borderline cytopenias, to consider in the up-front setting, and it’s an option for patients after ruxolitinib failure.

Case 1: Fedratinib Treatment for Myelofibrosis

Let’s all together discuss from the beginning, the diagnosis first because diagnosis appears to be problematic in some cases, particularly when the sampling of the bone marrow is not good enough, when the molecular testing for the driver mutations is not good enough. We have to understand that the diagnosis actually combines different aspects of what one sees in a patient. It’s not just the bone marrow. Perhaps Prithviraj, you can help us here. What does it actually entail to get the diagnosis of myelofibrosis? What needs to be ruled out that we wonder when we have fibrosis in the bone marrow? Is it always myelofibrosis or can it be something else? What are the diagnostic criteria and what needs to be ruled out?

Sure, Srdan, thank you. The primary myelofibrosis diagnostic criteria, there’s quite a few. There’s major and minor criteria. But 1 point I’d really like to make is you’re looking for megakaryocytic atypia. You’re not just looking for proliferation, because you see that in ET [essential thrombocythemia] as well. But the atypia is really, in my view, the hallmark of primary myelofibrosis. Of course, you have several minor criteria such as anemia, splenomegaly, increased LDH[lactate dehydrogenase], leukoerythroblastosis, etc. But you certainly need the clonal marker, and you touched on this, sometimes we don’t have a driver mutation. In about 10% of patients, we don’t have JAK2, MPL, or CALR mutated, and that’s what we call 

. But there, too, using a myeloid mutation panel, an NGS [next-generation sequencing] panel, which is getting increasingly prevalent nowadays, can really help because you find these nondriver mutations—mostly epigenetic and splicing mutations, ASXL1, TEK2, EZH2, DNMT3A, SRSF2, and SF3B1—which can really help you show that this is, in fact, a clonal disease.

Other diseases, sure, can cause myelofibrosis without being clonal—infections, autoimmune conditions, certain drugs even, certain recreational drug toxicities. But those are not going to have a clonal marker. So, yes, it’s really a combination of a lot of factors, pathologic, genetic, laboratory, as well as clinical.

So it seems that it is a clinician who actually makes a final decision, because you have to combine the bone marrow exam, including the molecular and cytogenetics, you combine the physical exam, you have a chemistry such as LDH [lactate dehydrogenase], and you have a blood count. All this needs to come together and the clinician many times is the one who makes the final decision. But you already mentioned the mutation analysis beyond the driver mutations, testing for the JAK2 mutations, calreticulin or MPL mutations is part of the diagnostic process and it’s 1 of the factors accounted for. But other mutations you already mentioned have a role in a differential diagnosis if there are questions. Andrew, the mutation analysis gets handed up only in these borderline cases but also for prognostication. Is this something that we should be doing quite often, expensive molecular testing? What’s your take on this—where there is a question of diagnostic differential? Where else should we do it and how often?

I think that certainly the NGS has allowed us to do molecular panels much more easily than we were able to do in the past. Certainly, with diseases like myelofibrosis, we typically do broad molecular panels on virtually every patient that comes in. I think that they provide a lot of information, diagnostics, and we touched on prognostics as well. Increasingly, we’re starting to have the potential for providing some therapeutic benefit as well.

I think Dr. Komrokji talked a little bit about some of the prognostic gene mutations such as ASXL1, SRSF2, EZH2, U2AF1, and IDH1/2. Certainly, I think those are helpful and give us a sense of where exactly we think this disease might go, and it may act a little bit more aggressively than we would have thought otherwise. But diagnostically, you may suggest or think you’re seeing something like myelofibrosis, but you could see some combination of the CSF3R mutation or a combination of tattoo, ASXL1, SRSF2, that makes you rethink the diagnosis and maybe consider something like chronic neutrophilic leukemia or atypical CML [chronic myelogenous leukemia] or CMML[chronic myelomonocytic leukemia], which can sometimes present some overlapping features with myelofibrosis.

I think the idea of potentially being able to use these as therapeutic is really where we need to get to and where we’re kind of going, and certainly with the IDH-1 and -2 inhibitors, there becomes the potential to use that. And increasingly, we’re starting to come out with more molecular-targeted treatments that could be used in some of these cases.

Case 1: The Diagnosis of Myelofibrosis

It seems to me that we are moving to putting a lot of weight for these additional mutational abnormalities in terms of prognosis. Rami, you have shown a couple of new prognostic scoring systems that incorporate molecular factors and even dissect the patients with mild fibrosis in different groups from those that are secondary to ET [essential thrombocythemia] or PV [polycythemia vera] to mild fibrosis or primary. How do we choose among all these? You showed only a couple and there are many more. Why not also talk about the risk of transformation to acute myeloid leukemia [AML] as a valid point for prognostication? How do we go with choosing the best and expanding to utilizing them perhaps to see who is going to transform to AML?

I think that’s a good point. Obviously, there is a lot of overlap between those risk models, and obviously, there are common variables that we can use in those models. Some of them are depending on a clinical phenotype, some of them are depending more on molecular phenotype. To keep in mind also that part of the clinical phenotype, sometimes it’s more flexible than molecular phenotype as well. Having said that, I think in practice, we usually start with something like the DIPSS [Dynamic International Prognostic Scoring System] or DIPSS-Plus because it is more comprehensive, and we do complement it with 1 of the genetically based models like the MIPSS70. Actually, Dr Kuykendall published some of our work looking at when we had a discrepancy between 2 stages difference, between somebody that in a clinical-based model, had, let’s say, a low risk, while a genetically inspired model was intermediate-2. It seems that the genetically inspired model did find the prognostic value. So, especially in the younger patients, as after doing the clinical model, I tend to use the MIPSS70-Plus, for example, in deciding particularly if those patients have enough risk features of the disease to consider allogeneic stem cell transplant.

Your point is well taken. I think we have to start looking more also about the risk of AML transformation. There are obviously certain predictors that are clinical, as well as molecular, that can help us to predict the risk of transformation in AML. Maybe as Andrew was alluding, that could be a population down the road that we would be considering for an allogeneic stem cell transplant earlier or intervention. For example, the group that has IDH-1/IDH-2 mutations is known to be at higher risk for transformation, and maybe those are the patients who we should be considering to receive some treatments earlier.

Case 1: Transformation of Myelofibrosis to AML 

We’re going to move on to talk about therapies. It was a very nice review that Dr Komrokji gave us on the possibilities of what to utilize in our patients with myelofibrosis. Do all the patients require therapy? Is that ultimately what happens in everybody? Prithviraj, what is your take here? How do we go about the therapy? Should we treat everybody at the time of diagnosis, or shall we wait? When is the time to intervene if not right at the beginning?

We would love to have a drug that really reverses the natural history of the disease and one that can fundamentally reverse the disease process. Unfortunately, we don’t have that. So, I think it’s appropriate to observe as Rami was saying earlier, in a patient who is asymptomatic, low risk—I think I would favor observation. I would start ruxolitinib in patients who are symptomatic and have substantial splenomegaly certainly, because the drug provides some undeniable benefits in those regards. I would probably not start it purely because of its survival improvement. Again, going back to what Rami was saying earlier, we completely believe that it improves survival in patients who require it. So, for patients who are symptomatic, who have big spleens, it really reverses their symptoms; it makes them feel better, eat better, be more energetic, etc. Perhaps it has an indirect effect on survival, which has really been shown in the long-term follow-up of the COMFORT trials. So that is there, but I would use it for symptoms and spleen, and that’s the drug I normally would start with. Regarding what percentage will ultimately require therapy, I don’t think we quite know that percentage, but I think eventually, it will be everyone.

In terms of initiating therapy on time, how would we go about defining when is the right time? We have a questionnaire that we all hopefully use, which is the MPN-10 questionnaire, to objectivize the quality of life of the patients and see how bad the quality of life is, which may help us define the timing of introduction. We can certainly measure the spleen by palpation, ultrasound, or CT [computed tomography] scan. I don’t think we use MRI [magnetic resonance imaging]. When is the best time or should we talk about utilizing the therapy early on to prevent the progression in a way, or to have the prolongation of life early? Andrew, what’s your take on that?

I think it’s a great question because I don’t think we know the answer exactly. I think we have bits of information that give us a better feel for that. But what we know about ruxolitinib therapy is we saw the waterfall plots that Rami presented from the COMFORT studies, and virtually everyone who is on that, so they got some degree of benefit. While it might not have reached the 35% threshold to call it a swimmer plot, virtually everyone got some improvement in their spleen volume.

Multiple studies have been looking at what predicts response versus nonresponse, and things that predict a less-than-adequate response or less of a response would be more molecularly complex disease, larger spleens, and those are things that happen throughout the course of the disease and happen in the later stages. So by holding off on therapy, it may be that we’re allowing some of these factors to arise that prevent the response, so we’re getting the best response we could.

We know that patients maybe earlier on in the disease course, with a spleen that’s not massive—with maybe just a 

 mutation and not a significant complex molecular profile—maybe those are the ones who are more likely to benefit, and we can acquire that response early on and prevent it from getting to an area where we’re not able to get that. But I think that’s a challenging question, to know the right time to start a therapy in someone. Oftentimes, I’d say that you can talk to a patient and patients will tell you, based on their quality of life, when is the right time to start a therapy. But it’s a challenging discussion.

No question about it. It does seem, as you mentioned, that early intervention in terms of management is preferred because it’s easier to manage possible adverse events with anemia and thrombocytopenia if you start earlier in patients who do not have already significant anemia or thrombocytopenia. You have to account for the ultimate goal, which should be the decrease of the spleen to the larger degree that was, as Rami showed, connected with the prolongation of life. So, we should choose earlier intervention in not-so-sick patients, easier management, better dose of ruxolitinib, better spleen response, and perhaps then prolongation of life.

Just to conclude, 1 more question before we finish discussion of this topic. Rami, you had mentioned the role of interferon in myelofibrosis. Where do you see the role of interferon to move from JAK inhibitors to another possible therapy, perhaps earlier on in the disease management?

 Obviously, interferon has been around in different versions and tested over many years. There is probably now approval for the newer formation, the ropeginterferon in Europe for ET [essential thrombocythemia], mPV [mean platelet volume], and myelofibrosis basically that have been studied, suggesting that interferon could alter the natural history of the disease, lead to reversal of the fibrosis or molecular responses. However, those were smaller studies, and the major limitation in most of those studies has been the toxicity and the patients not being able to stay on interferon for a long time to observe that benefit. Typically, it needs a couple of years of treatment with the interferon-based regimens to see some benefit.

So, in my opinion, based on the data, it’s that subset of patients that are early on in the disease, that they don’t have extensive fibrosis yet. They are younger, they are into that low, maybe intermediate-1 risk disease, that could benefit from interferon treatments. I typically would discuss a pegylated interferon like Pegasys or if there was a trial with ropeginterferon, in younger patients, early on in the disease, not much symptomatic. Once the patients have massive splenomegaly or constitutional symptoms or profound cytopenias, I typically have not observed clinical improvement or responses with interferon. So I think if one is going to think of interferon-based therapy, it should be done early on in patients that have lower-risk disease, not extensive fibrosis—not that high-symptom burden.

Very good. I will agree from my clinical practice, it’s the same type of experience, perhaps in the highly proliferative early phase where we need to control the blood cell count, it’s much better than in the later phase.

I thank you all very much for a lively discussion on the variety of the topics on myelofibrosis and thank you all for participating. We’re going to move on with the second case.

Case 1: Initiating Therapy for Myelofibrosis

 For case 2, we’re going to present a case of uncontrolled polycythemia vera. We have a 67-year-old male, who in February 2018 presents with frequent headache and dizziness. His hemoglobin level is very high, 20.5 grams g/dL, and white blood cell count and platelets are kind of high-normal— just a little bit elevated at 9.9cells per cubic millimeter of blood and 380 per microliter. Other results showed a hematocrit level of 63% and some microcytosis, suggestive of possible iron deficiency. The patient has a JAK V617F mutation and allele burden of 65%. A bone marrow biopsy was performed, which shows trilineage proliferation and pleiomorphic megakaryocytes. His prior medical history is notable for smoking. Based on a diagnosis of polycythemia vera, the patient is started initially on phlebotomy and aspirin.

Now, a few months later, the patient has undergone 3 phlebotomies, essentially monthly, and complains of continued dizziness, headaches, and nausea. At this time, he is continued on phlebotomy as needed and aspirin, but now hydroxyurea is added, a dose of 1,000 milligrams per day. A few months later, he’s had a couple more phlebotomies, and the dosage of hydroxyurea once again is increased now to 1500 milligrams per day.

Now it’s November, the patient is still dependent on phlebotomies and he also complains of itching. The dose of hydroxyurea is increased up to 2 grams per day. And in February, 1 year postdiagnosis, he’s had a couple more phlebotomies since November, he’s averaging about a phlebotomy per month, and he’s also experiencing abdominal fullness and some taste changes. His spleen is now palpable at 6 centimeters below the costal margin.

Overall, as a summary, this is a patient with polycythemia vera and over the course of a year, required phlebotomies on average once a month despite increasing doses of hydroxyurea. He also had some symptoms of moderate splenomegaly and itching, as well as headaches and some taste changes.

Case 2: A 67-Year-Old Male With Polycythemia Vera

When we look at the NCCN [National Comprehensive Cancer Network] guidelines for polycythemia vera, we start by risk stratifying patients as either low risk or high risk. This is pretty rudimentary. Lower-risk patients are considered those who are younger than 60 years of age and have no prior history of thrombosis. These patients are monitored for any new thrombotic events or bleeding. We make sure to focus on other cardiovascular risk factors and make sure that those are managed appropriately. We recommend a low-dose aspirin regimen as well as phlebotomies to maintain hematocrit levels less than 45%.

Typically, we’ll keep an eye on these patients every 3 to 6 months to ensure that no other events are happening and that symptoms are controlled. In patients who remain asymptomatic and are otherwise low risk, we continue with just aspirin and phlebotomies. In patients who develop some other disease-related issues, certainly a cytoreductive therapy can be considered, and there’s a list of those here from the NCCN guidelines.

Obviously, new thrombosis or disease-related bleeding would shift the patient over into the high-risk category. But also things like frequent need for phlebotomy or poor tolerance of phlebotomy, an enlarged spleen, uncontrolled blood counts (other than hematocrit, such as thrombocytosis or leukocytosis and disease-related symptoms) could all be considerations or factors that would indicate a good time to start someone on some cytoreductive therapy. And then certainly if the disease progresses over to myelofibrosis or acute leukemia, that pushes a patient down a different pathway.

For high-risk patients, which is either someone age 60 or older or has any prior history of a thrombotic event, the management is very similar in the sense that we’re still monitoring for new thrombotic events, managing risk factors, aspirin at a low dose, phlebotomy. But, now we consider the addition of cytoreductive therapy, and by this, we often mean either hydroxyurea or an interferon-based therapy. Keeping an eye on patients every few months, assessing for response and management of disease-related factors, and if everything is going well, we continue on the same treatment.

If there’s any kind of inadequate response or loss of response, which is a few items here listed such as intolerance to the hydroxyurea or interferon, a new event such as a thrombotic event or bleeding, any continued need for phlebotomy or poor tolerance, and then once again splenomegaly, thrombocytosis, leukocytosis, or disease-related symptoms, that might push someone to change therapy or alter therapy in some fashion. The options for altering therapy—once again, you have hydroxyurea and interferon, using an agent we haven’t used before, but now you have the addition of ruxolitinib, which has been studied in this setting and clinical trials are always an option as well. And certainly if the disease progresses, we go down a different pathway.

Case 2: NCCN Guidelines for Treating Polycythemia Vera

We’ll talk a little bit about ruxolitinib, which I mentioned is an option in the second-line setting in polycythemia vera, and this was based on the RESPONSE trial, which is a phase 3 trial of patients with polycythemia vera, requiring phlebotomies who had previously been on hydroxyurea. Patients were randomized 1-to-1 to either ruxolitinib or standard therapy, and this is a study that now has 5-year data that’s been presented and published recently.

The primary end point of this study was a composite end point between hematocrit control and spleen response. As you can see here, about 21% of patients in the ruxolitinib group were able to achieve both hematocrit control and spleen response. Broken down into spleen response and hematocrit control, you can see that those percentages are both superior to standard therapy, with 40% of patients being able to achieve a spleen response and 60% of patients being able to achieve hematocrit control. Always a key end point in myeloproliferative disease trials is symptom reduction. You can see here that ruxolitinib showed very impressive symptom benefit looking at it as a whole, with a 14-point symptom score on the MPN-SAF [Myeloproliferative Neoplasm Symptom Assessment Form],as well as clustered into cytokine clusters, hyperviscosity, and splenomegaly.

As I mentioned, this has long-term results that have been published recently. These were presented in 2019 and published earlier this year. These are the 5-year data in the RESPONSE study, and what you can see here is the response rates were impressive, but also the durability of response is impressive here. At 224 weeks, it was expected that over 70% of patients would be able to maintain the hematocrit control and maintain the spleen volume reduction. Looking at complete hematologic response, this was also maintained in over 50% of patients at the 224-week mark.

I’d be remiss to not mention the RESPONSE-2. The RESPONSE study enrolled patients that had large spleens, and so the RESPONSE-2 was an answer to the question of whether or not this was a drug that was predominantly good for patients who had enlarged spleens, and the answer unequivocally was yes. You can see here in the RESPONSE-2, patients without splenomegaly were able to achieve a very similar degree of hematocrit control and complete hematologic response as well as symptom improvements, compared to what was seen in the RESPONSE setting, and that was superior to what was seen with best available therapy. So, this is not a drug that is just for those with enlarged spleens, as it is able to give benefit to those who did not have splenomegaly as well.

Interestingly, we think about the toll that phlebotomies take on a patient, and by repeatedly phlebotomizing patients, we make them iron deficient, and that comes with a whole set of symptoms that can be quite debilitating and problematic for patients. So it’s challenging when that’s the 1 treatment option that we offer, is to solve 1 problem, but creates others.

So, this is an interesting study looking at the effect of ruxolitinib therapy on iron markers and iron deficiencies, and if you focus on the MCV [corpuscular volume], these are broken down into 4 lines and it looks at patients who started with normal and abnormal iron levels, whether or not they were on ruxolitinib or best available therapy. But you can see in the patients that had low MCVs and low iron saturations, suggestive of significant iron deficiency, those numbers improved over time while on the ruxolitinib therapy to a degree not seen in the best available therapy arm.

Then looking at the exposure-adjusted rates of common adverse effects with ruxolitinib, you can see in the polycythemia vera setting, the rates of anemia and thrombocytopenia are quite low. While there are many different adverse effects listed here, the rates of grade 3/4 adverse effects are quite low, but we need to mention that there is an increased risk of herpes zoster reactivation or shingles reactivation that’s seen with ruxolitinib at a higher rate than what was seen with the best available therapy. That’s certainly something that we counsel patients on and talk to them about—especially the need for the new shingles vaccine and keeping an eye out for this potentiality.

While the primary end point or secondary was not really included, there are thromboembolic events that were not really included as a primary end point within the study. That is often the goal of treatment in patients with polycythemia vera—to prevent further thromboembolic events. But looking at this, you can see that the rates of thromboembolic events with ruxolitinib were quite low, occurring in 1% of patients, compared to a higher rate that was seen with best available therapy. And then the patients that crossed over from best available therapy over to the ruxolitinib arm, the rate was quite low as well. So, I think that’s very reassuring to know that this is often what we’re trying to prevent from happening, and that while not included as a primary end point, this was something that was seen in a very small amount of patients.

As far as safety goes, we worry about the potential for secondary malignancies in patients who are treated with ruxolitinib, and here you can see that the rate of this was about 7 per 100 patient-years in patients originally randomized to ruxolitinib. That was a little bit more than what we saw with best available therapy, where that rate was 4.1 per 100 patient-years, and a little bit in between that in the crossover group. And also, the rates of nonmelanoma skin cancer were 5.1 in the original patients who were randomized to ruxolitinib, and that was about double than what was seen with the best available therapy arm as well. So, while I know the 4 of us are in Texas and Florida, we always recommend that patients see their dermatologists pretty regularly. But this is something to counsel patients on while they are on ruxolitinib.

Case 2: The RESPONSE Trial of Ruxolitinib for PV

Let’s have a nice discussion here on a number of topics. Let’s start with the diagnosis here again. Rami, we had a case here where the hematocrit level was 63%. The patient didn’t have a bone marrow biopsy, but is it absolutely necessary?

No. So you bring a good point. I don’t think you need a bone marrow for establishing the diagnosis in those cases. If the hematocrit is high enough, with the presence of a clonal marker like JAK2, I think one can make a diagnosis of polycythemia vera [PV]. However, we do get bone marrows at baseline typically, and I do them for more prognostic value and as a baseline to have as the patient’s disease progresses down the road to compare to. There are studies that look at the presence of fibrosis, for example, on the bone marrow in p vera, and that could be prognostic. I like to get a baseline because if there’s a question about transformation, I would like to compare at baseline. However, in certain cases, when the hematocrit is high enough, you have a clonal marker positive, you do not need the bone marrow for establishing the diagnosis.

So in this case, for example, the very high hematocrit level, with a JAK2 mutation positivity on blood, with the low serum erythropoietin, perhaps even iron deficiency, you would have enough there to say it is a PV.

Now, in terms of bone marrow, so you have some prognosis there, 20% of the patients may have a grade 2 fibrosis already at the time of diagnosis. You can also do the cytogenetic analysis to see occasionally whether there are some patients with chromosomal abnormalities. Prithviraj, is there a role of a genetic testing—expensive genetic testing—other than the JAK mutation? There is some information, but is it practical to test for it? Do we need it in short order? Yes? No? Maybe?

That’s a great question as well. I think the role of that is, right now, more of a research role because there are some mutations, SRSF2, ASXL1, and IDH, that are adverse in PV. But, again, you’re not going to act on them right now. So this is quite often a debated topic, whether everyone should be tested. Now, in research institutions such as ours, we do that, and it’s important from a research standpoint I think. But right now, from a patient care standpoint, perhaps not in a universal manner.

When it comes to therapy, hydroxyurea comes up as a drug of choice, although the guidelines that, Andrew, you just showed very well, would suggest hydroxyurea or interferon. Is hydroxyurea preferred in your hands, and are there cases for interferon, and once you start the cytoreductive therapy, what are the goals? Is it just hematocrit or is it something else? Is it everything else in blood cell counts, is it symptoms with the spleen? Give us a little bit of a summary here quickly. What do you choose and what do you aim for?

I think when starting patients on therapy, we mentioned the hydroxyurea and interferon are both first-line options. Typically speaking, I think that we often consider interferon more for younger patients—that we think there’s maybe a little bit of a better role for using this long term in those, and hydroxyurea more for the older patients. Now, I think they are both options, and certainly they’re very different and there’s a lot of discussion as far as which one is more appropriate in different circumstances. As far as hydroxyurea, we know it’s been used for a long time. It’s easy, we have expected adverse effects that we know how to manage, it’s relatively cheap, it’s oral and easy to take. Interferon, for the most part we’re using pegylated interferon, which is a subcutaneous injection given weekly. The cost of that is a little bit different, so sometimes, it factors into that as well.

But the goal of this is to really improve the quality of life of the patient and prevent thrombosis. In general, polycythemia vera has a favorable prognosis, so we have to think about not just improving quantity of life or reducing the risk of thrombotic events, but also maintaining a quality of life as well. So, when you mention the aspects of what do we consider uncontrolled disease in the setting of cytoreductive therapy, I think there are easy things. We want to prevent thrombosis, we want to be able to control blood counts, we want to keep people functional and active. But there’s also a lot of other things that go into this. Hydroxyurea, ruxolitinib was very good at controlling symptoms of the disease, and symptoms of the disease are something that is what bothers patients on a day-to-day basis. In someone whose blood counts are controlled, but have debilitating symptoms or symptoms that are bothersome, I think that’s a failure of our therapy and we need to consider other options. In the same manner, if we’re on some sort of treatment that is requiring the patient to have phlebotomies every month, I think that can also be considered a failure of therapy.

Case 2: The Diagnosis of Polycythemia Vera

So, we are moving from looking at the hematocrit control alone, to controlling the white cell platelet symptoms and the spleen. Rami, in your hands, if those factors are not controlled well, and certainly you would like to primarily have control of hematocrit and then the white cells and the spleen and the symptoms and the platelets, perhaps then there would be a number of cases where you are unable to control the 5 factors that I mentioned, with hydroxyurea alone. Therefore, there is room for more therapies, and the targeted therapy that we’re talking about here is ruxolitinib. It seems that it’s very valuable in the second-line setting. We have 60% normalization of the hematocrit, we have about 25% of the patients controlling all blood cell counts and it’s very durable. I think durability is particularly impressive to me. What’s your take on the utility of ruxolitinib in the second-line setting?

I totally agree. Obviously, as Andrew mentioned, we are looking at different dimensions. We are looking at the long-term disease outcome, short-term outcome. Most of those patients have a good prognosis when we’re looking at either 

events as well as the symptoms. We typically don’t think of p vera [polycythemia vera], that patients are symptomatic, but when you start talking to patients and listening, many of them had itching, that sometimes can interfere with activities, fatigue, some of them will have symptoms from the spleen, vasomotor symptoms. Hydrea sometimes will not do a great job controlling those aspects, even if the hematocrit is better controlled.

The other thing I always talk about is basically when we think about phlebotomies and what is acceptable and how often should we do phlebotomies on those patients. Is it really acceptable that we are doing a phlebotomy every month on those patients? Although there may not be evidence that phlebotomy per se is harmful, but I think a patient who requires phlebotomies on a regular basis has inadequate disease control or a disease that’s really active, that we may need to think of changing therapy based on that. The other thing, it also depends how often do you look. If we are seeing those patients and we are just checking hematocrit levels every 4 months, sometimes in between we don’t know how long they are staying about that target of 45 that we aim for. So, maybe sometimes we are also looking more often.

I think ruxolitinib offers a great option of controlling disease for patients who are sometimes difficult to control with Hydrea, especially when you start going on the higher doses. In my experience, toxicity with Hydrea starts to become very prominent when the dose is more than 2 grams per day. Patients will have nausea, sometimes sores, and you start seeing more myelosuppression with the treatment. If patients have symptoms, especially on a higher dose, if they are still requiring phlebotomies, if I’m not able to maintain their hematocrit in that range below 45 on a regular basis, for those particularly that have drenching night sweats, fevers, and itching, I find ruxolitinib as a very useful option for patients.

Let me ask Andrew 1 more question about the other side of the coin. We briefly talked about herpetic infections and squamous cell carcinoma of the skin. Is there a main concern here, or we are able to control those by perhaps monitoring patients’ skin and providing perhaps advice on the management of herpetic infections? Briefly.

I think these are the main adverse effects that we talk to patients about, that I think are concerning—certainly, the shingles reactivations and the nonmelanoma skin cancers. Recently, there was a presentation that suggested that the patients who were at a higher risk for these non-melanoma skin cancers were patients that had previous nonmelanoma skin cancers. I think that helps to certainly identify patients who are at higher risk, and often those patients already have dermatologists and we can counsel them to see their dermatologists more frequently to be able to prevent these things from occurring. As far as the herpetic and shingles infections, we’re not really sure how the new shingles vaccine affects that, but certainly it’s something that’s recommended from our standpoint. It’s not a live virus vaccine, and it has shown efficacy that’s preferable to the prior shingles vaccine.

But what I counsel patients on, oftentimes this is a group of patients 65 years and older, they’re at risk for shingles reactivations already, and shingles reactivations can be something that when managed early in the course of the disease, some of the problematic symptoms can be mitigated. Sometimes, the trouble is not recognizing this. So if counseling patients, they can become more aware of this and hopefully present to medical care earlier on in the process.

I agree. So, we balance here with the occasional adverse effects that you mentioned by carefully monitoring patients, but we count on a tremendous improvement not only in blood cell count, but in quality of life with ruxolitinib, and perhaps even improvement in the iron levels, eliminating that possible adverse effect or otherwise problematic issue in some of the patients. So, I thank you all very much for this lively discussion. We’re going to conclude this case and move on to the next one.

Case 2: Ruxolitinib Treatment for Polycythemia Vera

Case number 3 is a case of essential thrombocythemia [ET]. This is a 39-year-old woman with elevated platelets in the range of 475 × 109/L that were detected on an annual physical exam. The patient was, at the same time, found to have a hemoglobin level of 12.7/dL

and white blood cell count of 9.9 × 104/L, which are within the normal range. The bone marrow biopsy was then done, revealing proliferation of megakaryocytes with elevated numbers of enlarged, mature megakaryocytes in the bone marrow, without any evidence of reticulin fibrosis. The molecular testing that can be done not only on bone marrow, but also in blood, did not reveal a BCR-ABL fusion, excluding chronic myeloid leukemia, and there was no 

 mutation either, the most common one found in the BCR-ABL negative MPNs [myeloproliferative neoplasms]. This patient actually had a type 1 calreticulin mutation detected.

If you look at the diagnostic process for essential thrombocythemia, which is outlined here, we start with the collection of the criteria that would be major criteria, and that would include the bone marrow biopsy. The bone marrow biopsy in patients with high platelets serves the purpose of documenting the disease, looking at the size, shape, and colors of megakaryocytes as we noted before. It also makes sure that we are not facing another disease that many times, patients present similarly with high platelets, which is the early or prefibrotic myelofibrosis. In that case, there would also not be much fibrosis, but there will be some other findings that would be perhaps dependent on the eye of the hematopathologist, to dissect between the 2 types of the diseases in the bone marrow, looking basically at the megakaryocytes themselves.

In each disease and in ET particularly, we would have about 60% of the patients with the 

 mutation, about 20 or 25 with calreticulin mutation like this patient, and a smaller percentage of patients with MPL mutations. This is to say that there are patients who have no molecular marker at all in our testing.

For ET, all 4 major criteria or the first 3 major criteria and 1 of the minor criteria need to be fulfilled, and when we come to dissect the case between the ET or differentiate between ET and prefibrotic myelofibrosis, the key is to see whether there are any of the minor criteria that would help us take away the significance of differentiation from the bone marrow alone, by looking at the blood cell count, LDH [lactate dehydrogenase], and splenomegaly. If these abnormalities are present, then it will make it easier for us to diagnose patients with prefibrotic myelofibrosis.

Case 3: A 39-Year-Old Female With Essential Thrombocythemia

Now, going back to ET [essential thrombocythemia]and prognostication, in patients with polycythemia vera, we traditionally use 2 factors. On the left side of the slide are patients over 60 years of age and those with previous thrombosis that would dissect the patients in 2 groups: the low-risk group without any of these factors and the high-risk group with at least 1 of the 2 factors. The difference here in thrombosis incidence is the key to understand why we say that somebody is low risk for thrombosis and why the other patient with ET is a high risk for thrombosis.

The significance of platelet number, as you can see on this slide, as it relates to the risk of thrombosis, does not really exist in multiple different studies. But there is a correlation between the extreme thrombocytosis when platelets are above a million and a half and the connection with increased risk of bleeding due to acquired von Willebrand factor deficiency. When the patients have such a high number of platelets, a million and a half, 2 million, we would actually test the patient for the presence of a von Willebrand factor deficiency. And if present, we would stop the therapy with aspirin and try to decrease those platelets or consider decreasing the platelets for the risk of bleeding, not for the risk of thrombosis.

Now, the development of the molecular testing that can be done on blood has confirmed the spread of the 3 different driver mutations among the patients with ET as I mentioned before, but it also led us to realize that there is a significant difference in thrombotic risk between the patients with ET who have different driver mutations. What we see here easily on the left side of the slide, that the patients with the calreticulin mutation have a much lower risk of thrombosis than others. The patients with wild-type disease do not have any of the 3 markers (or we cannot find any of them on our testing) also have a lower risk for thrombosis.

The consequence of this particular testing has been introduced into enhancement of our risk assessment. This is endorsed by NCCN [National Comprehensive Cancer Center] guidelines now, not to only account for the age and previous thrombosis but also for the presence or absence of the 

 mutation. We did not really see specification in the NCCN guidelines of what mutation the patient should have to contribute to the prognostication for thrombosis, but simply just to say is there a 

 mutation—yes or no. With that, then we have 3 factors that would divide patients in 4 groups rather than in 2. For simplicity, in the lower part of the slide, we divided the so-called elderly low-risk patients (again, just based on age and thrombosis) into now a very low-risk group and a low-risk group, where the very low-risk group would have none of the 3 factors. The question is, do they need even aspirin? In the low-risk group (patients younger than 60 years) who never had thrombosis and had the 

 mutation … aspirin would be proper therapy. Similarly, for patients in the elderly high-risk group, we now have 2 groups: the intermediate-risk group for patients 60 years and older and that’s the only factor—the question is whether we really need cytoreduction therapy in them—and the high-risk patients of all the other ones that have 1 of the 3 factors in them, where we do provide aspirin and cytoreduction?

Now, this is a very simplified view of the NCCN guidelines. If we want to go in much more detail, then we would come up with this summary on 1 slide of multiple other slides that are available online when you go to the NCCN guidelines on the management of PV [polycythemia vera], ET, and myelofibrosis. This is a summary on ET. Baby aspirin in the very low or low-risk group and perhaps in those that have a very low risk and no cardiovascular risk factors may not be valuable at all. In fact studies suggest that in low-risk patients with the calreticulin mutation in particular, there is no value of taking baby aspirin because there is no high-risk thrombosis in these patients justifiably enough to even give the aspirin. It may even produce the countereffect of increasing risk of bleeding.

But the intermediate group (patients older than 60 years and with no 

 mutation and no thrombosis but takes baby aspirin) is good enough without cytoreductive therapy. But in high-risk groups one would prescribe cytoreductive therapy, and here hydroxyurea would be first-line therapy. Interferon is also mentioned as a possibility and an anagrelide, but this is usually second-line therapy. Hydroxyurea has been standard front-line therapy. The twice-a-day aspirin can be sometimes tried in patients who have a consistent problem with symptoms, in particular because of the rapid turnover of the platelets, 1 daily aspirin is not good enough to decrease the thickness and the symptomatology that comes from that.

The patients in the low-risk group may occasionally require cytoreduction therapy for progressive leukocytosis, progressive or worsening of disease-related symptoms, or progressive splenomegaly, but none of this is very common. But it needs to be noted in every practice that even for a low-risk patient, there are exceptions in terms of prescribing cytoreductive therapy.

Now, who are those patients with wild-type disease that don’t have any of the 3 driver mutations? Do they really have a disease? That is the usual question. This elevation in platelets in these particular patients, is it reactive or is it really a disease? Molecular dissection in almost half of the patients would show alternative mutations in 1 of the driver genes, not only in those that we commonly test for. So there is a possibility of having a molecular marker by extensive more detailed genetic testing in those who are usually labeled as having wild-type ET.

Case 3: Prediction of Thrombosis in Essential Thrombocythemia 

This is 1 of the studies that I like to highlight when we talk about the therapy. We talk about cytoreductive therapy as something that is standard practice for high-risk patients, but many of us in the community setting are nervous when we see, let’s say, a 45-year-old gentleman with high platelets of 1 million or 1.2 million, and say, “Wow, that is not a good number, you are at a high risk of a blood clot and we should treat that patient with cytoreductive therapy.” This is a study published not long ago in 1 of the leading journals, randomizing patients aged 40 to 59 years who had ET [essential thrombocythemia], with high platelets up to 1.5 million, without the high-risk features. In this setting, randomization that was between hydroxyurea or nothing—just of duration—proved the point that normalization of the platelets, which was accomplished with hydroxyurea, did not provide any substantial benefit. In other words, high platelets, up to 1.5 million, in patients with an otherwise low risk, younger than 60 years, no history of thrombosis, does not really provide the benefit.

One more point, which is perhaps important for every practice, is the issue of progression from ET to myelofibrosis. Now, we all say that there should be no fibrosis in bone marrow and diagnosis for patients with ET, and that’s correct. They may have some grade of fibrosis later on in their life, if you repeat the bone marrow biopsy after 10 or 20 years. But some presence of fibrosis on its own, it’s not good enough to call the patients transformed. As you see on this slide, there needs to be in addition to change to high-grade fibrosis, anemia or leukoerythroblastic or a big spleen or systemic symptoms or high LDH [lactate dehydrogenase]—all the known factors that are part of the diagnostic criteria for myelofibrosis.

But this is 1 of the practical points, that later on in the life of patients with ET, there might be some fibers in the bone marrow, and maybe these patients are at the risk of transformation to myelofibrosis, but they are not yet transformed because you need to fulfill those factors here to really call the patient transformed. But that’s the other end point of the life of patients with ET, who worry about progression to myelofibrosis: that may happen in a small proportion of the patients, perhaps 5% to 10% of patients do that.

Case 3: Low- to Intermediate- Risk Essential Thrombocythemia 

Let’s open this case for discussion. I will start with the question about the need for the bone marrow biopsy. I try to make a case that it is needed, but there is a counterargument many times that the management of patients with prefibrotic myelofibrosis or ET [essential thrombocythemia]does not really differ much. It is really about the thrombotic risk assessment and yes or no of cytoreductive therapy. Andrew, what do you think about the need for bone marrow biopsies in patients, let’s say a 45-year-old with high platelets of 800,000 mm3? What would you look for and decide?

I think you’re right in the sense that it’s unlikely to change management, and I think that Rami talked about this a little bit in the discussion of polycythemia vera [PV]. But with ET, the problem is that just with the clinical sign of high platelets and maybe some molecular information, presuming a diagnosis of ET may be kind of looking at the rosy side of things. There’s a lot of things that can be seen on a bone marrow biopsy that will give you a better sense of what’s going on. So, I do prefer to get bone marrow biopsies on these patients, maybe not emergently, but certainly within the first year of diagnosis, to make sure that we’re truly dealing with ET, and it’s unlikely to change management. But these are patients who you’re likely going to be following (especially in your example of a 45-year-old) over the course of decades, and they’re likely to undergo different bone marrow biopsies and you’d like to have a baseline to compare it with.

Just quickly, is there really a need for periodic bone marrow biopsies in otherwise clinically stable patients with ET?

Absolutely not. I don’t recommend doing regular bone marrow biopsies in these patients. I often tell them the reason we’d ever do a bone marrow biopsy is if the disease changes or if there’s questions that we can’t answer just from the clinical situation and the blood work. So, no, not regularly, but certainly if the white blood cell count goes up or if there’s progressive anemia or other blood count concerns, then oftentimes, we will do another bone marrow biopsy.

Now, the introduction of the presence of 1 or the other driver mutations in our prognostication for thrombosis is relatively new. Prithviraj, is this something that we should really embrace, and should we also test for other mutations, not driver mutations in ET?

Yes. To the first part, Srdan, I think so. I think the risk of thrombosis is largely restricted to JAK2, and at least the JAK2 CALR differentiation is important. I think less is known about the MPL mutation, we tend to cluster it with JAK2 and attribute similar characteristics. But it’s clear that the patients with the CALR mutation have a much lower risk, and it’s really the JAK2 that’s driving the thrombosis. So I think, yes, that is important and it affects management.

Regarding the second question, again, I think it goes back to what we discussed with PV. There’s been a number of genes, actually even more in ET than in PV, that have been implicated as being somewhat adverse. SRSF2 is the primary one. But, again, I don’t know that that affects management at this time, and I’m not sure that that should be a routine test for everyone. Again, it’s very important for research, though.

Case 3: Management of Essential Thrombocythemia 

We are all talking about the risk assessment for thrombosis, and that can change over time because it’s not only that we age and you get to 60 and then we always joke about giving a patient a prescription for a cytoreductive therapy when they have a birthday. But there are other factors that do evolve, other comorbidities. How do we go about proper assessment of the patients during their life? Rami, how would we go about it? We said no bone marrow biopsies. What other factors are we looking for to say, “You changed, you need to do something different?” It’s not only platelets, right?

Definitely, yes. Again, we would be thinking the same thing in 2 terms: a risk of thrombotic events and a risk of disease progression. Obviously, if somebody had a thrombotic event, then that will change the risk. I think it’s a good idea to also keep an eye on the cardiac risk factors for those patients. If there is anything changing there, then the risk of arterial events is probably going to increase. There is also a change in some of the blood counts. For example, I think leukocytosis could sometimes be predictive or correlates with thrombotic events more than a platelet count, per se. So, I would say from the risk of thrombotic events, those are the things I look at—if the patient obviously has had a thrombotic event, if they are having any other cardiac risk factors, and if there is change in other blood counts or inflammatory conditions that could set a thrombotic event.

In terms of the disease outcome, obviously if we start seeing cytopenias, anemia, circulating blasts, which are not common in ET[essential thrombocythemia]—but the anemia, high LDH [lactate dehydrogenase]—those are probably flags to start thinking: Is this disease progressing to post-ET MF [myelofibrosis]? Should we reassess? Disease risk is changing at that point.

Excellent. Then, in many of these cases, you would then say, either because the white cells are high or the thrombotic event happened or the spleen is increasing, we may need to start the cytoreductive therapy, or the patient is just older and it’s time to start therapy. Hydroxyurea is the first-line therapy, but interferon still has a role. Prithviraj, what is your take on interferon here? Unlike PV [polycythemia vera], in ET guidelines, it says hydroxyurea first, interferon later on.

Right. As we were discussing earlier, I tend to reserve interferon for the younger patients who obviously have a much longer duration of therapy that lies ahead of them. But I put a lot of faith in the MPD-RC trial that actually found no difference between hydroxyurea and interferon at 2 years in the front-line setting in PV or ET. So, in general, I stick with hydroxyurea, but I do use interferon in my younger patients where you have more of a concern for the long-term toxicities of hydroxyurea.

Andrew, how about anagrelide? There are studies of anagrelide versus Hydrea. Some show that Hydrea might be better for some particular characteristics that happen in patients with ET; others say it’s about the same.

I think we haven’t used anagrelide that much, but it is effective in managing platelet counts. Early on, there were some studies—1 major study showed that hydroxyurea was superior to anagrelide, especially for preventing arterial events. But there was a concern that potentially using different diagnostic criteria, that was enrolling a more high-risk patient population, and then more recent studies suggested noninferiority between the 2. But I would say overall, hydroxyurea has more potential benefits of its use and it’s largely tolerated better than anagrelide, but certainly anagrelide can be considered in some patients who are either not able to tolerate Hydrea and interferon, and it can reduce the platelet counts.

Once we start the cytoreductive therapy, I think we all agree that we are not only looking at the control of the platelets, but white cells if they are elevated, the symptoms if they are present, and the spleen if it is present. The goal of therapy has broadened to include multiple factors as in polycythemia vera. Once you perhaps control the platelets, but not the others, Rami, is there a time then to change, or you are satisfied with just controlling the platelets? How do you manage the need for a second therapy?

That’s a very challenging question because I really don’t think there is a good correlation between the platelet count, and we don’t even know the goal that we should get the counts to. I don’t think there is good correlation between the baseline or getting the counts controlled, but we all try to get the platelet count as normal as possible. I do pay attention to leukocytosis, I think that’s an important element. Also, as we mentioned, we are watching to see if there is any change in terms of the risk factors, the risk of thrombosis for those patients. But definitely, platelet count alone is not what I look at because I really don’t know what is the target or the goal in those patients. So, leukocytosis is definitely relevant to look at and obviously patient tolerability as well.

Case 3: Risk Assessment of Essential Thrombocythemia

The progression of ET to myelofibrosis [MF] or acute myeloid leukemia [AML] is not seen very often. But it is seen, and obviously it’s a very important problem because that means a shorter life expectancy. Prithviraj, in terms of management of progression, we all agree on bone marrow biopsy, and then what happens when you have a case of elevated blasts or fibrosis with progression in the cytopenias or a big spleen? How do we go about managing post-ET [essential thrombocythemia]MF or AML? Is there any specific difference, in other words?

I think the management of post-ET MF would be similar to that of primary MF or post-PV [polycythemia vera] MF. I think you’re still looking primarily at the spleen and symptoms as you make a decision for ruxolitinib, for example, or as you make a decision for an anemia-directed therapy. Again, you’re really looking at the picture of the whole patient, who can have not just 1 thing, but you try to choose the predominant problem as Rami alluded to earlier during his talk—to try and manage the disease. So, I think it’s either anemia-directed therapy or spleen-and-symptoms-directed therapy, which, of course, tends to be ruxolitinib. But, of course, the prognosis changes in a significant way, and I usually have that discussion with the patient when it is post-ET MF.

AML, of course, is a whole different cup of tea. When the blasts are increasing, you’re getting into an accelerated phase even before the formal post-MF, post-ET AML, you want to introduce the hypomethylating agents, for example. Those are the cornerstones right there for those patients, while perhaps continuing ruxolitinib if indicated for spleen symptoms, etc.

Andrew, Prithviraj was talking about when you change to myelofibrosis, then you manage basically like any other patient with myelofibrosis—looking at the symptoms, the spleen, anemia, and you may include a therapy with JAK inhibitors. Is there a role for JAK inhibitors in ET? I know there are some studies and it’s not officially approved, but is there a role or possible path forward for development? What’s your feeling about that?

I think that it’s less clear, obviously, than with polycythemia vera and with myelofibrosis, and probably because ET is just a little bit lower on the spectrum as far as the symptom burden. There’s less incidence of splenomegaly, and there’s less need to control the hematocrit level. So, the things that ruxolitinib have been shown to do effectively aren’t present as frequently in those patients with ET. But that doesn’t mean that none of those things is present in every patient with ET, and certainly there are some patients with ET who have significant constitutional symptoms and splenomegaly. Especially even in the patients with ET who have a 

 mutation, they may have hemoglobin levels that are getting close to that, that would be considered to be polycythemia vera. So, I think there is a subset of patients who could certainly benefit from ruxolitinib because we know that it’s a very good drug in controlling symptoms, reducing spleen size, and controlling hematocrit. Although the majority of patients with ET don’t need all of those things done, some do.

That was a really nice discussion. Thank you all very much for the review of ET. We’ll move on to the next case.

Transformation of Essential Thrombocythemia

The fourth case talks about post-MP [myeloproliferative disease] and AML [acute monocytic leukemia], or the blast transformation from primary myelofibrosis [MF]. We’ll start with the case here, a 68-year-old man who was diagnosed with primary MF about 3 years ago but is otherwise healthy. He has a fully matched donor, a sibling donor. He has mild night sweats, weight loss over the last 3 months of 10 pounds, some early satiety, some spleen-related symptoms in the left-upper quadrant, a lot of bone pain. So there are quite a few symptoms and splenomegaly as you see here, some aquagenic pruritus, fatigue, and some wasting.

The spleen is enlarged on exam; it’s 8 cm below the left costal margin. The white blood cell count is elevated at 35, hemoglobin is less than 10 g/dL, it’s 9.2 g/dL, platelets are quite low, less than 50 per mm3. Also leukoerythroblastosis, the teardrop cells that are fairly classic for primary myelofibrosis, with 5% circulating blasts in the blood. The bone marrow shows 80% cellularity, trilineage hematopoiesis, dysplastic megs, and a grade 3 fibrosis on a scale of 0 to 3. Eight percent blasts by immunohistochemistry and osteosclerosis. The cytogenetics show the relatively benign del20q, which is actually the most common chromosomal abnormality in primary myelofibrosis, and not a bad one to have. But then you also have the 

 mutation, which is an adverse prognostic one that we’ve been mentioning, and also this case is triple negative.

The patient begins RUX [ruxolitinib]. After 1 month, as you commonly expect, the symptoms and spleen are better. The drug does work very quickly, particularly with symptoms, and this patient is better in a month. However, 8 weeks later, the anemia has worsened. Hemoglobin was 9.2 g/dL; now it’s 8.6 g/dL. The splenomegaly has worsened, and, unfortunately, the blasts have climbed from 5% to 48% in the circulating blood.

Case 4: A 68-Year-Old Man With Primary Myelofibrosis

With that background, this slide tries to give you an overall snapshot of the rates at which PV [polycythemia vera] and ET [essential thrombocythemia] progress to myelofibrosis [MF] at 10 years, 15 years. We were talking about this earlier, that with time these diseases will evolve. In fact, ET can evolve into PV as well, and then to AML [acute myeloid leukemia], also known as the blast phase of the MPN [myeloproliferative neoplasm]. Not usually, these diseases will pass through a myelofibrosis phase before they get into AML; however, if you look at the outside of the arrows, you can go straight from PV to AML and from ET to AML at the rates of 10 and 15 years that are shown.

With primary myelofibrosis, progression to AML is actually even quite a bit more common, about 20% in this 1 study from Spain with about 100-month follow-up. So around 20% at that time period obviously getting higher as time goes on.

A number of studies have tried to delineate the risk factors for progression to blast phase for ET, PV, and myelofibrosis. The problem is that these are a lot of different studies, sometimes with even conflicting conclusions, and this table is kind of a collection of some of the findings. As an example, does splenectomy increase the risk of AML? This was reported but has been mostly discredited. It’s a little difficult: investigations in this area have not given us the clearest of findings. But some things are very clear, like exposure to multiple cytotoxic drugs in PV patients, serial exposure—that’s a clear risk factor. And your usual things: high blasts, anemia, low platelets, high white blood cell count. The things that make sense, the things that are bad anyway, are also the ones that give you a higher risk of AML transformation. And of course karyotype and genetics.

We get into that a little bit, showing you that triple negative patients who we already know do worse than 

 mutations with primary myelofibrosis from a survival standpoint, also have a worse leukemia-free survival, a higher rate of leukemic transformation. CALR, again, seems to have the lowest. It’s not clearly evident on this slide, but clearly the triple negatives are the worst.

Then we get into our nondriver mutations. As we know, these are referred to as the HMR, or high molecular risk mutations, as Rami showed earlier when discussing the MIPSS70 [Mutation-enhanced International Prognostic Score System]. But here you have 

, which all have…an increased risk of leukemic transformation in primary myelofibrosis.

I just want to point out a couple of things that are very striking. If you look at this slide and you particularly focus on 

, you see how less than 5% of patients in chronic phase MF have these mutations. But look at what happens when they are in blast phase. The 

 numbers actually have been quoted to be even higher in some studies—up to 30%, 

 also close to 30%. These 2 become very prominent in the post-MPN AML phase or the blast phase, which is very different from what we see in the chronic phase.

This slide just goes to show that there’s a lot that’s different between post-MPN AML’s genomic signature and that of de novo AML. As is very well known, 

 is mutated in up to 30% of patients with de novo AML. 

 also tends to be very common. But here in post-MPN AML, you have a much lower incidence of 

, underscoring the point that this is a fundamentally different, biologically different disease.

Case 4: Risk Factors and Genetic Abnormalities of MPN

With therapy, unfortunately, the story is quite dismal. This shows you work from Ruben Mesa from 15 years ago, showing that the survival is really very abysmal with conventional AML [acute myeloid leukemia]–type chemotherapy. In fact, the only ones that can have any long-term chance of survival are those who have a remission and then an allogeneic stem-cell transplant, not shown on this slide. But this just shows you that in general, chemotherapy is very poorly effective.

For azacitidine, the French looked at this about 10 years ago. This was a group of about 50 patients that either had AML or MDS [myelodisplastic syndrome] arising from MPN [myeloproliferative neoplasm]. MPN can also transform to MDS, about half and half. You see here that the azacitidine had a decent response rate of about 52%. Obviously, this is a highly established drug in both AML and MDS, so here too you have a reasonable response rate, 24% CR [complete response], 11% PR [partial response], and some MLFS [morphologic leukemia-free state]–type responses.

Just to remind you about this, there is something we call accelerated phase. It’s not particularly well established in the literature, but the only thing that seems to be clear to define it is this blast percentage, whether in the blood or in the marrow, of 10% to 19%. It seems to define this intermediate phase, which has an intermediate prognosis between chronic and blast phase but clearly is heading in the wrong direction. Again, blasts pretty much define accelerated phase in the 10%-to-19% range.

This shows you that most MPN patients who go from chronic phase to blast phase actually transit through the accelerated phase, with very few going straight from the chronic phase to the blast phase. This is work from our institution from a number of years ago.

Regarding the NCCN [National Comprehensive Cancer Center] Guidelines, we saw that chemotherapy doesn’t work very well at all. Transplant is the goal if you can get there, so you need to induce remission.I don’t think it’s very important how you achieve remission. But whether it’s an HMA [hypomethylating agent] or chemotherapy, if you can get them to remission and get them to allogenic stem cell transplant, that is really your only realistic hope of long-term survival. For those for whom that’s really not an option, frankly, chemotherapy doesn’t seem to have much of a role because you’re going to invite toxicity without much benefit. In those patients, a number of groups, including us, like to use ruxolitinib with the HMAs in part to control the spleen symptoms, etc. For those patients, it’s a very deliverable outpatient, relatively easy regimen, and that’s probably the way to go for those who are not going to a transplant. Of course, needless to say, a clinical trial is probably your best bet.

This is what I was touching on a little earlier. Our group at The University of Texas MD Anderson Cancer Center and our friends in New York have looked at ruxolitinib and decitabine in somewhat different dosing strategies. But unfortunately we came up with the same disappointing survival outcome of around 7 months median survival. This is a drug combination that is easier to deliver and perhaps gives patients some quality of life above what chemotherapy might achieve.

Case 4: MPN Transition From Chronic- to Blast-Phase MPN

Let’s start the discussion with what are the risk factors for the transformation. Many of these risk factors overlap to a great extent with risk factors that are accounted for in prognostication of overall survival of the patients, and the prognostications for overall survival primarily is used to define the patients that could be going to a transplant as if they have a life expectancy of less than 5 years.

It seems to me that there is a good case to prognosticate for transformation to myeloid leukemia, but it doesn’t really appear that we can do much about it. Andrew, what do you think about that? Is there a real need for everyday practice to prognosticate for not only survival but also acute myeloid leukemia risk?

Yeah, I think so. Certainly, if we were able to identify patients who, in the chronic phase, were at high risk of transformation to acute leukemia, those would be the patients we would rapidly take to transplant to see if we could prevent that. As Prithviraj discussed, once the transformation occurs, it’s really challenging to get those patients back to a state of their disease that’s then transplantable with any kind of success rate. I think it’s important to take that into account, and certainly our prognostic models, in myelofibrosis specifically, look at overall survival. But certainly some of those risk factors that we include in prognostic models exert their impact on overall survival by increasing the risk of transformation, and we touched on some of those, like thrombocytopenia, cytogenetic abnormalities, and certain mutations as well. So I think it is important to do that.

Prithviraj, in ET [essential thrombocythemia] and PV [polycythemia vera], you mentioned at the beginning that the risk of acute myeloid leukemia [AML] is very small, but some patients obviously will have it. While we focus so much on thrombotic risk assessment, we don’t really talk much about the prognostication in terms of transformation to myelofibrosis or acute myeloid leukemia in ET and PV. Should we start doing that?

You know, Srdan, I think the problem with it is that in the absence of a therapy that’s going to retard or delay or prevent that from happening, it perhaps has less practical relevance. Certainly patients will progress either through an MF phase or directly to AML. But when you are starting out treating them for PV or ET, all you can really do is focus on thrombotic risk—symptoms, spleen, things like that. That’s perhaps the reason that we have tended not to do that, and fortunately, it is still a very small risk for PV and ET, which obviously have long life expectancies otherwise. That’s really the driver of that. If we had a drug or a strategy that could reverse this process, certainly it would be more relevant.

Apart from possible implementation of prognostic factors in myelofibrosis, perhaps more than in ET and PV, for transformation to acute myeloid leukemia, there is not much else we can do. We know that the JAK inhibitors do not prevent progression from acute myeloid leukemia, but they don’t enhance the risk either. They’re just controlling the proliferation and inflammation, but they don’t change the possibility of a change.

Case 4: Assessing Risk for Transformation to AML

Rami, you have seen in this slide that it seems that the characteristics of a patient who transforms, at the time of transformation, indicates that the molecular evolution is the main reason. But cytogenetics have a role. There are other possible biological parameters. What are they, and how can we use them for perhaps therapeutic decision-making in that setting?

I think that’s a good question. Obviously, in most cases, what we observe in transformation is clonal evolution. However, there are cases where we, for example, see the 

 mutation disappearing, indicating that probably subclones have emerged. I think it will be important to try to identify, as you were alluding to, who those patients are and if can we intervene earlier in those patients to try to alter that. In my mind, the accelerated phase is just a transition, and those patients will get leukemia eventually. For example, as Prithviraj showed, there is a subset enriched with the 

 mutation, even those can happen in almost 20%, 30% of the patients at the time of transformation. Maybe if we are starting to see those mutations, an intervention with an IDH inhibitor would have a role. But in most of the cases, the disease is clonal evolution. Many of those mutations so far are not targetable so that we can do anything about it.

To your point of thinking of identifying those patients, the only therapy we have is really allogeneic stem-cell transplant. To your point, if we can identify ET [essential thrombocythemia] patients who have a high risk, should we be thinking of transplanting those patients if we have a solidarity of those models predicting the disease progressing to leukemia, or at the point where we see a molecular change that will suggest that the transformation is coming and inevitable?

That’s a very good point, but unfortunately, we and others are not really allowed, because of the cost issue, to monitor patients’ molecularity, to do molecular testing and see whether there is evolution to intervene earlier. To me, it seems that the standard way of looking at the people changing is just basically looking at the blasts. Can you look at the blasts? Prithviraj showed in the slide set that most people slowly evolve over time and get to 20%. If 20% is a cutoff for calling patients transformed, but the patients are transforming as you pointed out through accelerated phase, that appears to be the time to intervene, not to wait, and that percentage of blasts, 20%, is relatively insignificant in a sense that we don’t need to wait for it. Prithviraj, you pointed out hypomethylating agents as what NCCN [National Comprehensive Cancer Network] Guidelines would be suggesting over chemotherapy. Where would you use chemotherapy? Is there a role for chemotherapy at all in the management of patients with accelerated and blastic phase?

I think the role is very limited. Again, it’s in your young, fit patient where you have a donor lined up and where you know that transplant is your goal. That’s where the role is. It’s there, but it’s very small.

The JAK2 inhibitors were mentioned as well. Andrew, we use them to control the signs and symptoms, and we utilize perhaps 10 mg twice a day of ruxolitinib. And we stick with it, to control what we can, for quality of life. What’s your take on that? Is that how it should be done?

In general, we do that. But it’s a case-by-case basis. Certainly I don’t think the JAK inhibitor is there to induce remissions, and it’s unclear if it provides any role in that. Early on, ruxolitinib was looked at in AML [acute myeloid leukemia], and I think there was less than a handful of responders. But in general, it’s there to improve the symptoms related to the disease. Because as Rami mentioned, this is a molecularly complex disease that has subclones that are responsible for the progression to blast phase and AML, but certainly the clones that are there that led to the constitutional symptoms and splenomegaly are still there, and those are still symptoms that the patient has to deal with despite receiving treatment for acute leukemia. So we typically do that as well. We try to keep that on board in patients who were either benefiting from that previously or have signs or symptoms that would suggest they would benefit from that. We certainly titrate that up and down to manage the cytopenias that come along with that to maximize the patient quality of life.

This is really a good summary. If I may, it seems to me that the theme here is earlier intervention, either thinking with a bone marrow transplant when the blasts are coming up, or intervening with hypomethylating agents with or without JAK inhibitors, and not to wait for 20% blasts because most people will slowly transition. So we need earlier intervention, something akin of early intervention in myelofibrosis, not to wait for them to be high risk and in bad shape. Same kind of thoughts here with the transformation process.

Thank you very much for this lovely discussion. It was very valuable for our participants. Thank you for your thoughts and ideas how to manage this terrible condition when people are transforming.

I’d like to thank our panel for your case presentations and for a lively, informative discussion. It was a real pleasure. Thank you very much. To our viewing audience, thank you for joining us for this 

® presentation. We hope today’s discussion was a valuable use of your time and that you acquired some practical knowledge that you can take back to your clinic. Thank you, and have a good day.

Myeloproliferative Neoplasms