 We now have a second FDA-approved drug, fedratinib, in myelofibrosis. I’m going to turn to Ruben to discuss the clinical trial data, and then I’m going to ask Srdan to talk about, since they’re both indicated in the first-line setting, how you choose between these 2. Let’s start with Ruben.

There are 2 main studies. First, the pivotal study, which was the JAKARTA study that looked at 2 different doses of fedratinib: 400 mg and 500 mg versus placebo. It showed without question vast superiority in terms of control of spleen and symptoms. To remind folks again, this was several years ago, so this was concurrent during the approval time of ruxolitinib. This is a different study from what might have been done if it was much more recent.

It clearly showed benefit, without question. Adverse effects included some GI

adverse effects, potential for cytopenias, and in a small number of individuals, a neurological adverse effect that was potentially related to Wernicke encephalopathy.

A close look at that still puts that somewhat in question, but the agent is approved with a black box warning to monitor and replace thiamin before patients begin on the therapy and monitor for Wernicke encephalopathy. It functionally should not be a large issue in that this was a rare adverse effect and should be prophylaxed with thiamin. That is the basis of that frontline approval. The delay in approval was in resolving this whole issue of thiamin, Wernicke encephalopathy, as well as the drug development pipeline and delay.

There is also second-line data with a second-line trial led by Dr Harrison from the United Kingdom, myself, and many of our colleagues who we’ve recently updated at ASH and EHA, and it was recently published. That found a good 30% to 35% response rate for spleen and symptoms in the second-line setting for individuals who had failed ruxolitinib by relatively rigorous criteria in terms of resistance or intolerance. It was clearly active in the second-line setting for people who had failed ruxolitinib.

We’re pleased to see the commercial availability of a second agent, and it’s certainly something people should be considering for their patients who have failed ruxolitinib, for whom you do not have a good alternative clinical trial to consider. Fedratinib is approved and is an option for these individuals.

I will make 1 comment on dose as well. In the pivotal study, there was both the 400-mg and 500-mg dose. The 400-mg dose was felt to optimize efficacy, and there was more toxicity in the 500-mg arm, so 400 mg is the dose.

 Krisstina, the other toxicity is GI toxicity with diarrhea. Is there anything special you do or tell your patients?

Absolutely. Setting the stage for the potential adverse effect of diarrhea as a toxicity so they’re anticipating it is extremely important. Then do coadministration with some of our great antiemetics, which we now have available.

There is subduraconin, which is oral, which I have been using particularly for patients who have a history of being sensitive to medications and are potentially at high risk for significant nausea, but diarrhea is a problem as well. Managing the symptomatic diarrhea with Imodium, Lomotil, I have even gotten some GI input for patients who potentially may need more advanced therapies for managing the diarrhea.

Videos

Fedratinib for Primary Myelofibrosis

 Jamile, for the sake of time, I want to ask a few directed questions about practical management of patients on ruxolitinib. I often get asked, “What do you do with the drug prior to a surgical intervention?” Or “How about if the patient is admitted for some other inflammatory disorder: pneumonia, infection? How do you manage ruxolitinib during that time? Do you abruptly stop it? Do you taper it? How do we do this?”

I try to avoid abrupt stoppage of the drug as much as possible, but if someone is having surgery, and they’re going to be unable to take PO medications, then you can hold the drug for a few days. I try not to extend this over a week if I can help it. I try to avoid dose interruption as much as possible.

 Sometimes when a patient gets admitted and they’re sick or they’re intubated, I cannot continue ruxolitinib. And you can’t taper it, so I’ll give them a low dose of prednisone. Does anyone else do that? Is there any science behind it?

There’s no question that prednisone, in certain circumstances, can provide a bit of additional relief, but there can be low-grade rheumatologic conditions and other things they might have that might benefit from that as well. Prednisone certainly has some impact, without question, with all the limitations that exist in terms of chronic use of corticosteroids.

Ruxolitinib and Surgical Interventions

 Let’s move on now. I want to discuss the clinical trial data that has led to first the approval of ruxolitinib: the COMFORT-1 and COMFORT-II trials. On our panel is Srdan, who has clearly played a major role among other faculty members here in the development of ruxolitinib for myelofibrosis. Srdan, tell us about the data.

The studies that we are talking about were done a long time ago. These are the COMFORT-1 and COMFORT-II studies that led to ruxolitinib approval.

As time goes on, we learn a lot about the overall outcome of patients, and that has led us to not only realize that the drug is very good for control of the large spleen, symptomatic splenomegaly, and for improving the quality of life, but also to be able to extend life of the patients by virtue of changing the metabolism in the body: the albumin goes up, cholesterol goes up, performance status goes up, the ability to walk improves, and improvement in the kidney function has been reported last year.

Organ function and metabolism improves, and you control the disease signs and symptoms, which translates into prolongation of life for a few years. It’s also recognized by regulatory bodies, and it’s on the label for its use, but that’s not the reason to use it, right? The reason to use it is to intervene for immediate benefit: controlling the signs and symptoms, and with that, hope that life will be prolonged.

It’s evident in other studies that came along later, for example the JUMP study that was published recently in a final form, suggesting that early intervention matters and that we should not wait for the patients to be in a dying mode with high risk, a very big spleen, bad blood cell counts, and bad quality of life.

Early intervention is leading us to be able to provide optimal dosing, meaning the highest dose in a safe way, which will then decrease the spleen the most because it seems that the decrease of the ultimate size of the spleen that you can achieve on therapy dictates the durability of the benefit, and in many studies, it has been associated with life extension.

We learned that earlier intervention gives you the opportunity to provide a better dose; you have less myelosuppression, which is a toxicity related to the ruxolitinib in particular, and that will translate in the overall outcome that is beneficial, so the patient is not only feeling well but also living longer. This is where the major issue is: How do you initiate? You should follow the guidelines on how to initiate best in those cases regarding when to initiate: earlier on, don’t wait for patients to be advanced.

You then optimize the dose to get the maximum safe dose in the early few months to get that response to the maximum. That is the tricky part because anemia happens in about 50% of the patients. Patients may already be anemic; by early intervention, perhaps you have less of a problem with that.

Many worry about providing a high dose, and as you know, last year, there was a very interesting report with the European Hematology Association of a study suggesting that you should start with a lower dose of ruxolitinib, 10 mg twice a day, and then increase. Every month, you would go from 10 mg twice a day, to 15 mg, to 20 mg, to 25 mg. In that way, you avoid unnecessary myelosuppression, yet you can optimize the dose to the maximum safe dose. The results were similar to results reported in COMFORT-1 and COMFORT-II trials 10 years ago.

If there is a point to learn from most recent data—which is not published in papers; it was the poster last year—it is that we learn as we go. We can get to the same benefits on the spleen symptoms and life prolongation by alternative dosing regimens. Go from low to high, but don’t stay low. Go from low to high in a few months.

 That’s an important point. Without that data that you just mentioned from Europe, for years I’ve been starting at a lower dose and increasing, and quite frankly, it’s to enhance patient compliance. If you scare them away with complications in the first month, it might be hard to keep them on it. There might be some market data to suggest that many patients got 1 month and then didn’t get any more therapy.

Starting at a lower dose makes sense from a patient compliance standpoint, and they often feel better right away, even at the lower dose. The point you make is critical, both in PV and in MF, but this drug is not 1 dose fits all. You need to optimize the dose to maximize benefit. Thank you for that.

The other thing that’s interesting about the data is this: We’ve all seen that the survival curves from COMFORT-1 and COMFORT-II are superimposable early on and cross over, suggesting that the benefit in long-term outcome may be in controlling symptoms sooner rather than later. Those patients in the COMFORT-1 study, as I remember as 1 of the many investigators, were sick patients; they didn’t have anything else. They had big spleens and constitutional symptoms.

You’ll remember that 66%, two-thirds, of the patients in these studies were high-risk, and one-third were intermediate-2. There are more modern studies, even from analysis of general practice in Europe, that suggests that early intervention makes it easier for us to manage the patients, and the survival benefit is also enhanced because you can maintain patients much longer on the therapy.

It’s all a circle that we are trying to put together; the data that would help us achieve most of the benefit for our patients: early intervention, better dose and spleen as the target, in addition to quality of life. That is the way we are learning over the last 10 years about how to optimally use JAK inhibitors in the community settings.

 Correct me if I’m wrong, but I don’t know of any data with ruxolitinib, fedratinib, or interferon that says intervention with those agents decreases the risk of leukemic transformation. If that’s the case, it suggests that it’s symptom control, improvement in quality of life, nutritional status, muscle mass, and all those things that may be most important with the treatment.

You’re absolutely correct. There is no evidence for improvement in the risk of a transformation for progression to acute myeloid leukemia, which is an important point, not only to emphasize the benefit but also to say the disease is still there, and it can transform. Early intervention with a transplant, as we talked about transplant quite a bit, is encouraged.

I encourage my patients, even though they may have excellent response to ruxolitinib, to look at transplant as an option and potentially do the transplant if they have a donor when they feel the best. There is now evidence that it means that there’s possibly a better outcome with the transplant itself. As Ruben described, these 2 issues—the medical management and the transplant—go hand in hand. There is no comparison between the 2; they should be combined.

Ruxolitinib for Primary Myelofibrosis

 Jamie, let’s come back to that question about intervention for intermediate and high-risk patients versus low-risk or asymptomatic patients. We all know we see the 66-year-old with 2% blasts 1 day, and right then, that would make that patient intermediate-2: a candidate for ruxolitinib or fedratinib, the 2 FDA-approved agents. However, they’re completely asymptomatic, or at least they tell you so. We can ask Ruben to comment on whether that’s true if they’re completely asymptomatic.

Let’s say you have somebody who is happy with their quality of life, and they’re 66 years old, and they have 2% blasts 1 day or they have a sweat every 3 months. They’ve made it to the intermediate-risk group. Do you rush to put all those patients on therapy? What kind of conversation do you have with patients about that?

It’s a question, Harry, because it’s a commitment when you start a patient with myelofibrosis on ruxolitinib. There is a lot of blood testing that you have to do, there is follow-up, and there is a lot more that they need to do. They have to be convinced that they are in need of added treatment.

Your point about making sure that we have totally validated their symptoms in some kind of measure, as opposed to just, “Yeah, I’m feeling fine,” is extremely important. What about their spleens? Have they grown larger over time? Those questions are all important. There are a couple of studies, even though the initial COMFORT studies, which I know Srdan will talk about, looked at intermediate-2 and high-risk disease.

There are some data to support the use of ruxolitinib in some intermediate-1 patients that have splenomegaly or symptomatology. We have to be flexible in understanding each patient and think of other symptoms they are dealing with: Do they justify the initiation of treatment that will require frequent monitoring? Once they start, they’re set to continue, right? It’s a conversation we need to have with the patient.

 Yeah, and sometimes I don’t convince my patients. I clearly have intermediate-risk patients, and after talking with them and looking at their family, I say “This patient is symptomatic,” and the patient says, “I don’t want to start a chronic medication yet.” It’s the fun part about being a doctor.

Ruben, let’s come back to NCCN Guidelines. What is recommended for the intermediate- and high-risk patients? How do you choose between these? We’ll then get into the clinical data.

 Excellent. The guidelines are certainly based on the most broadly used risk score. We talk about intermediate and high risk by which criteria? Those have still been based off the DIPSS, which are the most broadly available and used criteria, but it does bring up some other more granular issues in prognosis that Jamile had brought up.

For intermedaite-1-risk patients, they have a heterogenous prognosis because we realize that some of those might be young individuals with anemia and a high-risk molecular mutation that by MIPSS70 might be much more high risk. But by DIPSS, they might still be intermediate-1.

The full gamut exists under the NCCN Guidelines from JAK inhibition for splenomegaly and symptoms to allogeneic transplant for people who might be higher risk, to even just observation if they were intermediate-1 solely on the basis of age. That was probably the most mixed group. If your patient falls in the intermediate-1, I strongly advise physicians—particularly of patients of transplantable age—to get an NGS panel to further clarify their prognosis.

For intermediate-2 and high-risk patients, that is your most standard group. That is the largest group of patients with MF. That initial decision is this: Do you consider transplant in the near term? Building on that discussion you had with Krisstina regarding who the candidates are, anyone who says that it’s an easy decision is not the person facing the transplant.

Comparing a transplant with medical therapy is comparing 2 different animals altogether. It’s different both philosophically and in terms of scope. The frontline medical therapy includes ruxolitinib, which has been the frontline therapy for many years, but it now also fedratinib under the NCCN Guidelines since September 2019.

JAK inhibition is used with the goal of improving spleen, symptoms, and aspects of the disease. Importantly, we then have to monitor for efficacy and decide on whether frontline therapy has had sufficient efficacy. Or one considers perhaps switching from 1 of those agents to another or alternative approaches.

Treating Intermediate/High-Risk Primary Myelofibrosis

 I want to come back and take us away from the brink of mortality and with advancing risk to circle back to the low-risk patient. We’ll talk about the Cervantes criteria. A very small percentage of patients, about 10% in that analysis, had low-risk disease. Krisstina, do you see these patients? Would you ever consider therapy for a low-risk patient, symptomatic or asymptomatic?

Patients definitely present with low-risk myelofibrosis. In those patients, this is where we’re assessing whether they are associated with a symptom burden. In those patients who have low-risk disease by the DIPSS [dynamic international prognostic scoring system] or the MIPSS70 [mutation-enhanced international prognostic scoring system]—the more sophisticated prognostication tools—still have a significant symptom burden. Those are patients for whom we can consider intervening.

I recently worked with the CIBMTR [Center for International Blood and Marrow Transplant Research] and a group of academic centers looking at the role of stem cell transplantation. Interestingly, there was even a group within that analysis with low-risk disease who proceeded to stem cell transplantation. Some physicians would say that, for the treatment of myelofibrosis, the only true curative measure is allogeneic stem cell transplantation, and in patients who are wishing for a more aggressive and disease-modifying approach even in low-risk disease, some physicians are proceeding with that.

There’s not significant evidence to support that approach, and the best evidence is in the intermediate-2 and higher, and more recently considering intermediate-1a disease. We look at the patient’s goals, their risk status, and their symptom burden.

In general, though, for low-risk disease, we monitor these patients’ counts to see if they are transitioning to a higher-risk phenotype, a higher DIPSS. Are they having molecular evolution? Are there more high-risk mutations recurring over time? We’re taking all these pieces of the puzzle to decide when to initiate therapy and how aggressive to initiate therapy.

 Thank you for that, Krisstina. You prompted a question I was going to ask you later because you are the 1 member of our panel who is a believer in stem cell transplant. You are a transplant physician, and I also think it’s because you’re probably the youngest, and it is a young person’s job to take care of these patients.

I’m going to come to a question for you: Our challenge in picking patients for the curative option of transplant is the data are intermediate-2 and high risk, but those are often the patients who are cachectic, have huge spleens, and have other issues that may preclude them from being an ideal transplant candidate. On the other hand, you have low-risk patients who may be healthy and able to undergo stem cell transplant, even with its risk.

My question for you is this: Is there anything you use in that low-risk patient to talk with them more aggressively about transplant, like mutation analysis or cytogenetics? Does that influence your opinion?

Absolutely. That’s a great question. At the University of Arizona Cancer Center are using these molecular characteristics to help guide us. We know that there are particularly high-risk phenotypes within myelofibrosis, particularly the triple-negative patients, the 

. These are extremely high-risk phenotypes: genotypes that can certainly have a higher risk of transformation to acute myeloid leukemia down the line.

The idea of intervening when patients are a bit younger, a bit more robust with higher performance status, fewer comorbidities, less splenomegaly, and fewer cytopenias is certainly attractive for a conversation with the patient. There are certain approaches, so for patients who are developing more splenomegaly, we can certainly use ruxolitinib as a bridge to control the splenomegaly symptoms before we take them to allogeneic stem cell transplant.

This is such an institutional approach, so every institution and every academic center for transplant will have a different approach and a different discussion. As a community, we’re moving toward an earlier discussion. For community physicians, I would like to impart the importance of early academic referral.

Often, we see that these patients are referred to academic centers when they’re already symptomatic from their disease. They have intermediate-2 or high-risk disease, and they’re saying, “Now it’s time to transplant.” No. It could have been a lot smoother. The donor selection takes time. All these things take time, so if it’s early within the disease course it can be better for the patient.

 I agree. I’m going to remind our viewers that you opened the transplant door, so I’m jumping ahead in our agenda, and I’m going to keep you on the spot as our local expert. Let’s say you have a myelofibrosis patient. I’ve heard that if you don’t do a myeloablative preparative regimen, you might as well not do it in a patient with myelofibrosis. It’s not going to work. Any truth to that? How do you select the regimen?

I’d like to clarify, I’m not an allogeneic stem cell transplanter.

I have done some research in this area. In general, this is a polarizing issue within the transplant community, and there are not significant data that would sway one to myeloablative versus reduced intensity conditioning.

This is an institutional preference. This needs to be looked at in the literature. It’s a challenge because myelofibrosis is a rare disease, and to do prospective analysis in this is very difficult. I don’t know if we’ll ever achieve that, so continued retrospective analysis will help move the ball forward. To answer your question: It’s polarizing, and it’s controversial.

Management of Low-Risk Primary Myelofibrosis

We can continue to talk about polycythemia vera for quite some time, which is different from 10 years ago, or maybe 20 years ago, even before the 

However, it’s time to move on, and I’m going to ask Jamile to start with a discussion of patient classification and risk stratification in myelofibrosis. For the sake of time, we will not spend a lot of time on the diagnostic criteria, which are well-published by the WHO and will move on to risk stratification and how that affects management of patients. Jamile?

There are multiple prognostic scoring schema that have been developed over the years. The first was published in 2009 by Cervantes, and it’s called the IPSS [International Prognostic Scoring System], which was a tool utilized at the time of diagnosis to risk stratify patients relative to their survival. There were 5 variables that were utilized: age over 65, white blood cell count over 25,000 per mm3, peripheral blast count over 1%, presence of constitutional symptoms, and hemoglobin below 10 g/dL.

That gave you 4 risk groups with median overall survival ranging from 130-some months down to 27 months, so you can see a tremendous variation of variability. In general, people think low-risk to INT-1 [intermediate-1 risk] is lumped together as 1 entity and then INT-2 and high-risk disease as another.

Mostly, stem cell transplant patients were offered to intermediate- and high-risk disease, but it then became clear that, as people went on to live with this disorder, they acquired other features that may also correlate with worse outcomes, namely the development of anemia transfusion and transfusion dependence. This is why the Dynamic International Prognostic Scoring System [DIPSS] was created to account for the development of anemia in patients who may not have been at the outset.

That was the next advance, and later, the DIPSS plus also accounted for a few other variables, including the platelet count below 100,000 per mm3, the presence of adverse cytogenetic abnormalities, and the development of transfusion dependency. The bottom line is that, whichever score you utilize, remember that the IPSS is good for the time of diagnosis, and the dynamic scores are good throughout the disease scores.

When the mutations became available with 

, people started to look at the impact of those mutations on the outcome of patients who had myelofibrosis. There was a trend to think to include those mutations into the scoring schema, so there’s something called mutation-enhanced international prognostic scoring system [MIPSS70], which encompasses the presence of certain mutations that are labeled as high risk. That evolved into the MIPSS70-plus, which includes cytogenetic abnormalities.

All that culminated in the genetically inspired prognostic scoring system, and that includes both molecular and cytogenetic abnormalities. I know this is a lot of information, and all the little bullets that you have in there to construct a scoring schema are interesting, and it may help in selecting people who may be candidates for stem cell transplant patients. I will stop there.

 Thanks, Jamile. That was an excellent summary, because it has evolved and now includes the mutational and cytogenetic data, which are prognostically important.

Primary Myelofibrosis: Risk Stratification

 I completely agree with those comments about symptom control outside phlebotomy control, hematopoietic control, and splenic volume reduction. We now have more tools for treating patients, and we alluded to 1 earlier: pegylated form of interferon, and Ruben alluded to the updated EHA. Srdan, do you want to tell us about the PROUD-PV study?

We are talking about the major development in the polycythemia vera arena in terms of new therapeutic options. Perhaps many colleagues in the United States do not know that a year ago, there was approval of ropeginterferon, what I call a super long-acting interferon, in Europe for polycythemia vera patients. It’s given under the skin every 2 weeks, and after a year, you can even switch to once a month. It was approved to be given to patients with polycythemia vera for control of the blood cell counts.

The goal is complete hematopoietic response, not in patients with a big spleen, because it’s perhaps not applicable when given to patients with big spleens. In that setting, the drug has been superior, after about 3 or 4 years of therapy, to hydroxyurea. That was the PROUD-PV study that you mentioned: the randomized study between hydroxyurea and ropeginterferon in patients who were either newly diagnosed, not previously treated with cytoreductive therapy, or already on hydroxyurea for some time but not optimally managed.

They were not refractory and not resistant, but there was room for improvement. In that randomized study, the analysis showed that, after 1 year, there was no superiority of interferon, but we know that interferons take time to work. After several years, 3 or 4 years, superiority was seen.

Regardless of the comparison to hydroxyurea, the label allows its use for polycythemia vera, and it doesn’t talk about first- or second-line, or about it being superior or inferior to hydroxyurea, which is in accordance with the European leukemia guidance and the NCCN Guidelines as well. Ruben described this well: The guidelines would say cytoreductive therapy and hydroxyurea or interferon.

At least in Europe, and hopefully soon in the United States, we’ll have a ropeginterferon to offer to our patients. It is much less toxic, it has better efficacy when you have less toxicity, and I’m looking forward to embracing these in our own patients here at The University of Texas MD Anderson Cancer Center.

 Before we move on, any other comments about what’s coming or what’s new in PV? That was the major highlight coming out of EHA. One final question about the use of interferon: It’s been touted as something that is more of a disease-modifying agent because some earlier studies show decreases in, for example, the 

 allelic burden. Has that been shown? Can you tell us about that from the PROUD-PV or other randomized trials, Srdan?

This particular information is quite interesting because this is the only therapy that may, in some patients, give an eliminated 

 clone by virtue of not being detected by our sensitive testing. That has been noted in the past with commercially available long-acting interferon Pegasys [peginterferon alfa-2a] in clinical studies.

This is offline use of that medication, but in ropeginterferon studies at the last ASH, there was a similar report of about 10% to 15% of the patients achieving the molecular response, not detectable 

We hope that means a lot for these people, but perhaps it’s not a cure. We don’t associate this with a cure, but if you have no detectable 

 mutation, that would mean that there may be no or minimal malignancy there anymore. That would perhaps cancel it in the long-term benefit for these patients in the future. It is something to follow; it’s not the goal of the therapy, but it’s an interesting correlative study that is evidence of the biological ability of these medications to alter, possibly, the natural course of the disease.

Polycythemia Vera: PROUD-PV Trial

 When I think about PV patients and controlling their hematocrit and splenic volume, which is not a problem in many patients with PV and is why the RESPONSE-2 trial was done, the other thing I’m focusing on is quality of life. This disease affects their quality of life in many ways with other symptomatology related to this myeloproliferative neoplasm.

I’d like Jamile to comment on this. What has been your clinical experience with using ruxolitinib? Quite frankly, do you ever switch to ruxolitinib for something other than hematologic control or splenic volume reduction?

I’m glad you brought this up because, as I was hearing everyone talk, I’ve been thinking about the last few patients I ended up switching to ruxolitinib. It wasn’t because of a lack of hematologic optimization with cytoreduction that they were on before hydroxyurea, namely. It’s because they developed night sweats and pruritus that was difficult to control, and they weren’t able to sleep. Imagine if you have 1 night after the other of not being able to sleep because of pruritus that didn’t respond to any other measures.

That was the impetus for a subset of my patients who had PV: to improve their quality of life. My experience mirrors the clinical trial data. Ruben is here, so he can speak to that because there was a 50% reduction in half the patients. Close to half the patients had greater to or equal to 50% reduction in their constitutional symptoms. It’s very similar to what I have experienced.

 This is an important point, because what we’re talking about is a bit outside the label of where the drug has been approved based on splenic volume reduction and improvement in phlebotomy or hematocrit control. Ruben, you have been involved in a symptom control study in PV. What can you tell us about the quality of life for these patients who are on ruxolitinib?

Without question, it has been significantly favorably impacted both in aggregate and looking at all the symptoms together: a total symptom score model and with individual symptoms. In PV, there are certain symptoms that are particularly difficult: pruritus perhaps being 1 of the most difficult to live with on a chronic basis, many of whom were quite refractory with night sweats, spleen-related symptoms, and others.

It is helpful with that kind of a target symptom view in terms of the most problematic symptom and quality of life, as well as in aggregate. My colleague Robyn Scherber and I presented at last year’s ASH Meeting, showing how centrally important the symptoms were to the MPN patient’s quality of life. You can imagine, if you were itching 24-7, how disruptive that would be for your quality of life. It can be a difficult thing to live with.

Utilization of Ruxolitinib for Polycythemia Vera

 You make a very good point. Until recently, for decades our management of PV has been aspirin, hydroxyurea, and phlebotomy; we didn’t have anything else. Worrying about where hematocrit is and symptoms and all these things, we didn’t really have any other tools. People didn’t want to use interferon, but there are better-tolerated interferons and more data coming, and now there’s ruxolitinib. Srdan, I’m going to turn to you to review the data for ruxolitinib in PV patients.

You set the stage very well, and only now do we have a good reason to review our practice and see how well we are doing with hydroxyurea. In the retrospective chart review studies, a number of which have been done, 20% to 25% of patients are not benefiting and should be candidates for something else. We have the opportunity to introduce ruxolitinib as the JAK1/2 inhibitor that Ruben mentioned as an alternative in these people who are not responding well.

There were 2 studies that led to its approval. First there was the RESPONSE study, then REPSONSE-2. They should be looked at together. They clearly describe the value of ruxolitinib in this setting well. The first study, RESPONSE, was for patients who did not do well on hydroxyurea, required phlebotomy, and had a big spleen. They were randomized between the ruxolitinib and control arm, which in the majority of cases was hydroxyurea by itself, because that’s the evidence of not having many other choices.

The RESPONSE-2 study was a copy of that study, so patients were refractorily resistant to Hydrea [hydroxyurea] with the need for phlebotomy, but without a big spleen. The benefits are easy to understand: there was a significant difference in ability to control the red blood cell count. Without going through the numbers themselves, the majority of people had controlled hematocrit on the ruxolitinib and a minority on the control arms.

In the study that requires a big spleen, the spleen was significantly reduced with ruxolitinib and not with the control arm. When we look beyond just the red blood cell count in the spleen, we also look at the control of the white blood cells, the platelets, and the symptoms. This is how I look at the benefit to any therapy: There are 5 factors that are built into the guidelines for what the response should be or how to assess response. That should be control of the red blood cells, white blood cells, platelets, spleen, and symptoms.

We talk about primary goals: eliminating phlebotomy, normalizing the hematocrit at 45%, controlling myeloproliferation as Dr Shammo described, white blood cells and platelets to the extent possible. Then for the quality-of-life issues, control the big spleen that may be symptomatic and general systemic symptoms. These are the benefits of the new therapies we cherish, and ruxolitinib is now an option in the second-line setting.

 How often do you have to adjust the dose of the ruxolitinib? We at the Duke Cancer Institute start at 10 mg twice daily. How do you do that? What do you use to adjust the dose?

For myelofibrosis, you’re right: the starting dose for every PV patient is suggested to be 10 mg twice a day, and the expectations are that you would go up. About 65% of patients need to increase: from 10 mg, you go to 15 mg twice a day, then to 20 mg twice a day, and even some patients need 25-mg-plus a day. It’s all adjusted based on the need to control the blood cell count and the symptoms that we described.

Only about 5% of patients need less than 10 mg twice a day, and that would be because of too much myelosuppression. This is where we simplify things for patients when looking at the numbers because other nonhematologic toxicities are rare. There are some, but they’re not common. The guidelines would be to look at the blood cell count and optimize the dose to derive those benefits and control the symptoms, spleen, and blood cell count to the maximum.

 I have a question. Ruben, I need your help on this. I see the data from the RESPONSE trials showing superiority in terms of decreased phlebotomy requirements, better erythroid control, and reduction of splenic volume in RESPONSE. The goal of therapy in PV is to prevent thrombotic events. Is there any evidence that, by switching to ruxolitinib, you may prevent thrombotic events?

In short, yes. The studies were not powered with thrombosis as a primary end point, but there have clearly been many fewer thrombotic events on the ruxolitinib arm versus the best alternative therapy arm. There’s good mechanistic reason to believe that that is true. First, there was better, more consistent control of counts as well as likely improvement in inflammation and other aspects that might create a pro-inflammatory milieu.

I’m mindful that, for patients who are inadequately controlled with supplemental phlebotomies on other medical therapy, they’re going to be spending a good amount of their time above the target hematocrit of 45%. Every time you phlebotomize them for their hematocrit of 49%; how long have they been above 45%? Maybe they’ve been above 45% since the last time you phlebotomized them; we don’t know. It’s an uneven control of the hematocrit. There are multiple mechanistic reasons why stable control with a JAK inhibitor like ruxolitinib is probably superior for thrombosis prophylaxis.

 Krisstina, the primary end point of RESPONSE was at 32 weeks, and we’re talking about a disease that in patients who had similar life expectancies to age and gender-match control. What do we know about long term? Are there longer-term data from the RESPONSE trial about tolerability and durability of these responses?

Absolutely. We see at 156 weeks that these responses—hematologic, splenic, and symptom responses—were sustained, and patients maintained activity on the ruxolitinib. The RESPONSE data have certainly given us another option to treat patients, particularly if they have had progressive disease on first-line therapy, and it has been a valuable contribution to our therapeutic armamentarium.

Polycythemia Vera: Ruxolitinib Trial Data

 One of the things I’ve noticed in the NCCN Guidelines that I’ve had questions about concerns a patient on a cytoreductive therapy, whether it’s interferon or hydroxyurea. And let’s assume that they’re having adequate control of their blood counts and tolerating therapy. Having another thrombotic event requires a switch to therapy, and I’m wondering: Is that necessary? How do you make that choice about what you switch to? Ruben?

It’s an excellent question. It depends a bit on the circumstance. Is it a new thrombotic event in a different vascular distribution? Is it extension of a prior clot? If it was a DVT, sometimes it’s difficult to know if it is an extension of a prior clot or a new clot in a different area. Were the counts controlled at the time that they had the second event?

I try to see if it is a failure of the therapy or if it is a failure of some other thing. Was it a true event? Without question, in terms of second-line therapy, we now have an FDA-approved second-line therapy of ruxolitinib for individuals with PV [polycythemia vera] who also can be indicated clearly in that second-line setting, and it is particularly helpful if they have splenomegaly symptoms or more difficult disease-associated features.

In the studies, there is a suggestion of further protection against thrombotic events. Those studies were not necessarily powered for thrombosis as a primary end point, but there is some thought that it decreases inflammation and other aspects that may be improved by the JAK1/2 inhibition and might be beneficial in terms of thrombosis prophylaxis.

 Jamile, what would cause you to switch from whatever cytoreductive therapy you’re on to another—ruxolitinib or interferon if you’re on hydroxyurea—other than thrombotic events?

It has to do with the tolerability of the cytoreductive therapy and the results I’m looking for. Some patients are unable to tolerate hydroxyurea. I’ve had a patient who had significant allergic reaction to the drug, so I had no way to do it but to switch to ruxolitinib in this instance.

I would switch if somebody is developing unexpected or untoward [adverse] effects from the dose of Hydrea [hydroxyurea] that is required to control their blood counts or constitutional symptoms or if they start to develop splenomegaly despite being on Hydrea [hydroxyurea] or other forms of cytoreduction. There are a multitude of reasons, and in patients who have PV, every patient has to be dealt with individually. We have to ask ourselves: “Have we optimized the doses? Have we talked about compliance? Can we up the dose before we make the switch?”

It is a commitment on both the patient and my part to say, “Now we’re ready for part b.” All the issues that have been mentioned could lead to a shift in therapy. The issue is that you need to continue to monitor the treatment and response to treatment, and that is something that may not have been optimized because we think PV is a chronic disease. When patients have very little symptoms, some say “Take your Hydrea, see you later.”

It’s very important to monitor response to treatment, look at the blood counts and constitutional symptoms, and talk with the patient about whether we are optimizing this to the fullest. If not, then switch.

Polycythemia Vera: Switching Therapy

 Ruben, I’m going to turn to you. For our viewers, Ruben chairs the NCCN  MPN Guidelines committee. I would like you to comment on what the NCCN recommends for initiation of cytoreductive therapy. When do you consider that in terms of that choice? Then, follow-up on that with the comment of whether it is hydroxyurea or interferon: how do you use those?

First, to give proper credit, I did initiate that group as the initial panel, but it is now chaired by our friend and colleague Aaron Gerds. I am very familiar with the NCCN Guidelines. They’re a key piece. Srdan and I were on that initial wave helping to set these up. They were positive in that they brought some evidence-based standards to our US–based treatment.

Cytoreductive need for high-risk patients in PV is aimed at both decreasing that difficult risk of thrombosis and bleeding as well as trying to improve the disease burden with symptoms. 2 options are listed as initial therapy; we’re building on that base and assuming that people have already been on phlebotomy and aspirin. That’s a base: control of hematocrit and aspirin. The cytoreduction, then to try to achieve a European leukemia net response: phlebotomy independence, control of the white blood cell count, control of the platelet count, and resolution of splenomegaly if they’re in control of symptoms.

With that, either hydroxyurea or long-acting interferons: both are thought to be clinician’s choice, between them and the patients. There are increasing data; there are randomized data from earlier studies showing that hydroxyurea is superior to phlebotomy alone. It’s an older drug, but people are frequently comfortable using it.

Deficiencies can be difficulties with cytopenias, mouth ulcers, cutaneous malignancies, and sometimes other associated adverse effects. For interferons, there is an update coming up at this year’s EHA. We saw it in the late-breaking abstracts regarding the pegylated interferon given every 2 weeks. In the US most commonly, the pegylated interferon alpha 2A has been the available construct.

When trying to achieve a good control of the counts, on the downsides with interferon, it can cause mood disorders, or rarely, autoimmune problems such as hypothyroidism, or flu-like symptoms. Both have their pluses and minuses, with interferon being the preferred choice in women of childbearing age or individuals who are pregnant.

Treatment Options for High-Risk Polycythemia Vera

 Krisstina, coming back to you, I want you to concentrate now on a low-risk patient. They don’t have any cardiovascular risk factors, and they’re not as aged as most of the panel members today: over 60. They haven’t had a thrombotic episode. They might be having microcirculatory symptoms, dizziness, fatigue, or confusion, trouble finding words, whatever it might be. Migraine headaches, for example. Comment on the use of phlebotomy and low-dose aspirin in that group of patients. Does the patient have to have symptoms for you to initiate phlebotomy and aspirin therapy for that low-risk patient?

For low-risk polycythemia vera patients, we approach all patients with low-dose aspirin, 81 mg or 100 mg daily, with concurrent phlebotomy, with that important hematic profile that Dr Jamile Shammo mentioned, or that 4-fold increase in thrombotic events or cardiovascular death. As a blanket recommendation, for those who are diagnosed with low-risk PV, we approach those patients in that way.

Regarding the symptoms, certainly there are a group of patients who will have significant associated microvascular symptoms: pathognomonic aquagenic pruritus and very symptomatic fatigue. In fact, this is the most commonly associated symptom, which was beautifully shown in a landmark study that patients are afflicted with it, and it can affect their activities of daily living.

In these patients, although we are definitely treating them and optimizing their hematic profile and decreasing their thrombotic risks with recurrent aspirin, we need to be assessing their symptom burden as an end point of care in combination with the hematologic parameters we’re targeting.

Regarding the microvascular symptoms, you can increase the dose of aspirin in some patients and have some modest benefit. There are some additive, supportive care maneuvers that you can certainly entertain for the pruritus, and gabapentin has been used to some additional help. I’d be interested in some of the other panel members’ unique supportive care maneuvers that they’re doing for symptom burden.

Thinking about symptom burden as equally important as the hematologic profile. Direct your therapy for your low-risk disease, and if it’s not adequately controlled, then think about second-line therapies.

 Krisstina, give our viewers a bit more understanding: How often are you monitoring a CBC or a hematocrit in your patients with PV?

 There are different phases. As you’re initially meeting a patient, you’re in an induction phase. You’re trying to be more aggressive: assess more aggressively and treat more aggressively, perhaps with more frequent phlebotomies.

For those patients, I’m seeing them every 2 to 4 weeks and making adjustments and recommendations for phlebotomy until we obtain adequate control. Once they’ve attained adequate hematologic and symptomatic control, you can start stretching out that monitoring frequency to every 2 to 3 months.

Management of Low-Risk Polycythemia Vera

 Jamile, we have identified a specific hematocrit hemoglobin level above which we do not want to see our patients. Can you review for us the data that support that? How often do you have to check if the patient is above that level, and when do you intervene?

 There are data to suggest that optimization of the hematopoietic parameters in patients who have PV [polycythemia vera] are extremely important. In fact, let me start with the hematocrit, for example. There are data that came from the CYTO-PV study that suggests that maintaining a hematocrit level below 45% is associated with a 4-fold decrease in mortality from cardiac event and major thrombotic events, and that’s something we can easily do as practitioners, as hematologists: maintain the hematocrit below 45%.

How often should you look for that? It depends on the stage of the disease. If you’re starting a new patient on phlebotomy, for example, you might do it twice a week and have someone checking those numbers twice a week. As the disease goes on or as you’re getting closer to optimization, the frequency of checking the hematocrit becomes less and less. We have to try as much as possible to keep this below 45%. I don’t want to put my patients at risk by that parameter.

The other thing is the platelet count. Practitioners have to realize that, as you phlebotomize patients, the platelet count is likely to go up; we’re inducing a state of iron deficiency by doing phlebotomy. People get nervous, and then they start to think, “Now the platelet count is rising.” The only issue there is perhaps the platelet count interfering with the administration of antiplatelet therapy.

It’s OK to have a mildly elevated platelet count, and it’s OK to give aspirin. But if you’re concerned, you could check for acquired Von Willebrand factor deficiency. I always worry that perhaps trying to chase the platelet count in the setting of phlebotomy may be misleading to some in terms of initiation of other treatments, or even suggesting that maybe this isn’t a disease that’s responding.

Finally, leukocytosis is something we need to pay attention to, because it is an indicator of myeloproliferation. For some people, there’s a lot of debate about whether they may be associated with increasing thrombotic events, mostly with data pointing to retrospective analysis of the association between leukocytosis and thrombotic events.

We [at Rush University Medical Center] did 520 patients who had PV, and we looked at the risk and didn’t find it in this retrospective analysis. These were private institutions that submitted data to evaluate that piece, but what it was associated with is disease progression. It’s something we need to pay attention to. Perhaps if you have ongoing leukocytosis, maybe this would be the time to evaluate the patient fully and see exactly if there is progression. Are they compliant with their medication? We don’t even talk about that. We then move forward. The most important piece when it comes to blood counts is that.

Any time you talk about MPNs [myeloproliferative neoplasms], there’s also this piece about splenomegaly and a piece about constitutional symptoms. Every time you see patients in the clinic, those would be things that we need to pay attention to, to assess how is disease doing and what we need to pay attention to in terms of disease control.

Polycythemia Vera: Monitoring Patients

 Krisstina, I’d like to turn back to you and have you comment about risk assessment and polycythemia.

Risk assessment is based primarily on 2 factors with polycythemia vera. No. 1, how old is the patient? Greater than 60 years of age or less than 60 years of age? No. 2, did they have a prior history of a thrombotic event? If they are young and have not had a history of thrombotic events, then they are a low-risk polycythemia vera patient. On the flip side, if they are more advanced age—as I’m aging into the 40s, that advanced age at 60 is hard to state. But as they’re getting older, it’s not a magic cutoff. If they’re 59 years old and they turn 60, that does not mean that, all of a sudden, their thrombotic risk goes up. We have to take that into consideration as we’re treating these patients and look at some of the other factors that we’re thinking about when we’re assessing thrombotic risk.

What are some of those factors? Those are the cardiovascular factors that we would assess as a primary care physician. Do they have underlying diabetes? Underlying coronary artery disease? Hyperlipidemia? Are they obese? Are they a smoker? It’s a personalized approach when you’re assessing polycythemia vera and their thrombotic risk, and it leads to an opportunity for discussion with our patients in regard to these lifestyle measures.

That’s an area of my own personal passion; we’re treating patients. We’re not treating CVCs [central venous catheters] or hematologic profiles. As hematologists, it’s important to be assessing these risk factors, to be assessing lifestyles, and to be addressing them. Not just identifying them but talking about these things.

Polycythemia Vera: Risk Assessment

 I want to turn the attention of the panel to maybe the 2 most important and most frequent questions I get about testing in PV [polycythemia vera]. The first is if the 

 allelic burden at the time of diagnosis is important. The second is about reports of splice site variance: Are those artifacts, or are those clinically important? I’m going to turn it back to Srdan to start on that. The 

 allele burden; how about if it was really low? Could this just be CHIP [clonal hematopoiesis of indeterminate potential]?

The testing is usually done in the clinical setting, where you are suspicious of polycythemia vera. We are not randomly testing people to see whether they have a low level of the 

 mutation, which can then be achieved. We are testing in the setting of suspicion of polycythemia vera, where there would be a high-normal or above-normal red blood cell count. You would also look for high white blood cells, high platelets, iron deficiency, and erythropoietin level, and in that setting, you would also test for the 

The presence of mutations on its own is not in the diagnostic criteria for any of the MPNs [myeloproliferative neoplasms] because they might have the CHIP. If you have the mutation in the context of all these other tests, Dr Shammo has articulated that very well; on borderline cases for mass PV, you would do this testing and account for all these factors to make sure that you don’t miss PV because the therapeutic approach must be different than if there is no PV; phlebotomy is mandatory.

On the other hand, it may be important, but we are not routinely doing that testing to follow. There is good literature on high allelic burden being prognostic in terms of progression to myelofibrosis or exposing people to high-risk of thrombosis itself.

 If the allele burden at the time of diagnosis is less than 10% or less than 5%, could that be consistent with PV?

It can be, in the context of hyperproliferative bone marrow, high red blood cell count, low erythropoietin, and iron deficiency. However, that would be unusual because the median allelic burden in PV is about 45%. It would be an unusual case.

 Ruben, what do we do with these reports of exon splice site variance where exons are dropped out? Is that just an artifact of the assay, or are they clinically important?

We don’t fully know yet. As we continue to get more genetic information, it is a very rapidly moving target. What is an activating mutation and what really is genetic artifact? The annotation of our genetic reports is a very active process that is not fully baked. I’ll try to look them up as they arise, as well as whether they have the potential to alter our understanding of what we should be doing for those patients.

In terms of our decision support as hematologists moving forward, trying to keep track of all these things is quite difficult because there’s such a river of information as it relates to different types of mutations, allelic burden, and other sorts of pieces. As we get supported over time, that will be an important part as we’re trying to digest what the mutual impact of half a dozen or a dozen mutations is, some of which are in nontraditional splice sites. It becomes more and more difficult to interpret that information.

I fully agree with Ruben. This is the area of research, but specifically on the 

 gene variance, what is done in a community setting is you have a hot spot test that you do: a 

 V617F. We don’t look for the whole gene; only if you have a negative result do you then do the 

 exon 12. In some academic centers they analyze the whole gene.

In some cases, you will then find the variance in addition to the hot spot mutation that you are looking for. It seems that the presence of variance or other genetic abnormalities within those genes make a difference for the outcome for the aggressiveness, but it’s completely new area of research.

Polycythemia Vera: Prognostic Markers

 I want to point out 1 other thing in the small print. Because is the case for many pathologists on that committee, they say “If you don’t do it, you miss some important prognostic information on the presence of fibrosis.” Of course, I don’t know what to do with that prognostic information, so that doesn’t sway me 1 way or the other. How about you?

That’s a good point because the patients always worry about the long-term outcome of the disease. By default, they say it’s benign, but there is a proportion of the patients, about 20% to 25% of patients who, over their life span, may progress to myelofibrosis, and a certain small percentage who will develop leukemia.

Myelofibrosis in particular is the worry because, when you transform officially, the life expectancy can shorten. The bone marrow biopsy can provide some useful information to semi-intelligently prognosticate for the future. This is all relative, but if you have a cytogenetic abnormality, an abnormality of chromosomes, which is very rare but not impossible, that would perhaps prejudice most people to have a high risk of change.

The presence of fibrosis at the time of diagnosis of PV [polycythemia vera] is present in up to 20% of the cases; on a scale from 0 to 3, grade 2 is already seen in up to 20% of the patients. I repeat it because it’s important. These patients with PV are at a higher risk of a change to full-blown myelofibrosis during their life span.

This is the kind of information that we would like to have at the beginning for full understanding not only about diagnosis but also possible prognosis for transformation. Not for the thrombotic risk, which I already mentioned as a primary goal of therapy, but for the future. That needs to be highlighted: the presence of the fibrosis does not mean myelofibrosis by default. Fibrosis can to some degree be seen in PV.

 Thanks, Srdan. Ruben, do you want to add to that?

I just wanted to add an important point that I share with patients, which is that I agree that baseline bone marrow biopsy is helpful. The challenge is if we see a bit of fibrosis and they have that higher risk, we still don’t know what that means in absolute terms. If I compare 2 patients—1 without that fibrosis and 1 with—yes, the patient who has the fibrosis has a higher likelihood of progression through myelofibrosis, but that might be in 10 years compared with the other patient who will never progress during their life span.

As we think about these prognostic factors, it is worse as a population, but it’s always difficult to translate for a patient what that means. Additionally, all of that is in current science. We may well develop therapies, or some of our current therapies might alter that natural history further. So prognostic always has to be taken in appropriate consideration; sometimes patients can become very distressed by some of these prognostic findings as well.

 As everybody said, the bone marrow biopsy may or may not be important. The 1 piece I wanted to add in PV is for patients who may not meet the criteria. Let’s say you have a 

 V617F positivity, and you’re entertaining an MPN [myeloproliferative neoplasm]. Perhaps the hemoglobin isn’t 16.5 or 18.5 g/dL, where you don’t have to do a marrow, but perhaps it’s 13 g/dL. If you did the bone marrow and then you get an iron stain, which I highly recommend on anybody who has a marrow with the suspicion of an MPN, that has 0 iron in it. If you have absent iron stores, then your patient clearly is iron deficient and may have PV, even though the hemoglobin doesn’t necessarily meet the criteria.

In that sense, when the heme parameters do not fit the WHO [World Health Organization classification and diagnostic criteria for myeloproliferative neoplasms] 2016 as opposed to 2017, as it says in the NCCN [National Comprehensive Cancer Network] Guidelines, it has to be performed so that you can characterize the MPN a bit better and pay attention to the iron piece. I often see it done, yet nobody comments on it, and we could improve a lot on the prognostic accuracy of the characteristics of the MPN we’re looking for.

 I agree with that. The iron stain can be helpful in those borderline cases, and I often look for iron deficiency.

Polycythemia Vera: Bone Marrow Biopsies

 Let’s move on. Srdan, I’d like you to discuss how patients with PV [polycythemia vera] present and what the differential diagnosis of that presentation is.

Although polycythemia vera is the name, it implies that all the cells grow, which is true. 

 means high blood cell count: red blood cells, white blood cells, and lipids. Not everybody has high white cells and high platelets, but everybody does have high red blood cell counts. The cardinal feature of polycythemia vera is a high red blood cell count.

That can be in a setting of blood clots at the time of the presentation in about 20% of the patients, and it’s often associated with the systemic symptoms: itching, night sweating, low-grade fever, fatigue, flushing in the face, not feeling well overall, memory loss, and blurred vision are symptoms related to impaired circulation in the body because there are too many cells in the blood vessels. These are the characteristics of the patients we see, but not every patient who has a high red blood cell count has polycythemia vera.

As you alluded in your impression, there are other conditions that may resemble these kinds of presentations. Particularly in the setting of younger people, we need to be cautious when we look at the erythrocytosis associated with the use of certain medications like testosterone. Erythrocytosis is associated with problems with lung function—people who had sleep apnea, for example. There are also problems related to certain cancers—kidney cancer, for example.

Not everybody who has high a red blood cell count has polycythemia vera. The first goal is to make sure we are talking about the bone marrow neoplasm and not reactivated erythrocytosis, because the approach is different. With the disease, we have not only the number of cells increasing but also their activity. They are sticky, and that is the problem: too many cells that are sticking together, and that increases the risk of thrombosis. Therefore, the approach there is to decrease the thrombotic risk in polycythemia vera patients.

To go back to the diagnostic process, once we roll out the secondary diseases that would increase blood cell counts, which are usually benign, you would look at lung functions, medications, and other possible causes. You would focus specifically on the criteria for PV that, in 2016, has been simplified and modified to quite a significant degree. We now have 3 major criteria to fulfill, or you can fill at least 2 of them, and I’ll give you a brief summary of those.

The cardinal picture is high red blood cell count. We are talking about decreased hematocrit or hemoglobin, as is routine when you do CBC [complete blood count]. That should be associated with the testing you can do in blood for the presence of 

 V617F mutation, which is present in about 95% of the patients, or in those who don’t have that 1, testing for 

That will cover almost everybody: either of the 2 mutations, and there is still a possibility of having anecdotal cases of patients who do not have any of these, which may be a matter of the sensitivity of the test, or they have something else that can link mutation for which we cannot test.

 mutation, and then the question of the bone marrow biopsy: Do you need to have it done? In the synopsis for the diagnostic criteria it says No. 1, high red blood cell count; No. 2, bone marrow biopsy; and No. 3, the 

 mutation. If you have all 3, you have PV, no question about it, but there is a minor criterion there which is suboptimal serum erythropoietin level, meaning low erythropoietin level.

Harry P. Erba, MD, PhD, Duke Cancer Institute

Articles