Biology of myeloproliferative neoplasms (MPN); mutational landscape and implications for prognosis and therapy

DR LOVE: Welcome to Oncology Today — Myeloproliferative Neoplasms Edition, a special audio program focused on current management and emerging strategies in the treatment of MPNs. This is medical oncologist Dr Neil Love.

For this program, I met jointly with Professor Claire Harrison and Dr John Mascarenhas. To begin, I asked Professor Harrison to discuss key aspects of the biology of MPNs.

PROF HARRISON: For a general oncologist or a fellow, I think the first thing is about the use of mutations in terms of diagnostics. Obviously we’ve got our 3 driver mutations that we know about. And then that we really have to focus hard to be sure about our diagnosis in patients who have this so-called triple-negative disease, the term we borrowed from breast oncology. And then in the last 12 to 18 months, we’ve increasingly understood the role of, say, nondriver mutations in terms of prognostics and particularly in the field of myelofibrosis, when we’re looking to prognosticate a patient, about taking that risk of higher-risk therapy. Using the newer prognostic tools, such as the MIPS or the GIPSS, which incorporate this data for the patient, thinking, is this a high-risk patient I should be worried about, watching more closely and thinking about taking to transplant?

DR LOVE: John, can you maybe go into a little more granularity about the specific types of mutations that are seen? How often they’re seen and kind of what the implications are?

DR MASCARENHAS: Increasingly, as next-gen sequencing is being incorporated into the clinical care of our patients, we are realizing and appreciating that there are many mutations that often exist within a patient’s hematopoietic cells. Typically the driver mutations predominate, and that’s usually JAK2 V617F, calreticulin or MPL. And as Claire pointed out, there’s about 10% of patients with myelofibrosis that are triple-negative.

And then there are a number of subclonal mutations that we frequently see that involve a lot of the pathways that are also common in related diseases like myelodysplastic syndrome and acute myeloid leukemia. And these include spliceosome mutations, other signaling pathway mutations.

And there’s a group of mutations that are classified often as high molecular risk mutations, which include ASXL1, EZH2, IDH1 and 2, and SRSF2, and they are — whether you have 1 or more than 1 of those mutations, categorize patients who have poor outcomes typically. In fact, some data would suggest that patients who acquire or those mutations grow out while on ruxolitinib defines a patient population that is going to do poorly.

We know that there are mutations that have adverse prognostic impact. Whether they clearly influence treatment decisions I think is still an evolving area of research — for example, IDH mutations can be targeted now with IDH inhibitors, which are used both in the relapsed/refractory and up-front setting in acute myeloid leukemia. And now trials are embarking in advanced forms of myelofibrosis and MPNs combining ruxolitinib, a JAK1/2 inhibitor, with enasidenib, an IDH2 inhibitor, in patients who have IDH2-mutated MPNs, in which we think those patients, particularly in combination with a JAK2 mutation, do poorly.

We’re beginning to include or incorporate this data, which we’re increasingly we have at our disposal to both refine prognostication and risk stratification and hopefully influence treatment. But we still have, I think, a lot more work to do to better understand how to tailor the treatment to the mutational profile.

Diagnosis and risk stratification of MPN

DR LOVE: Claire, can you talk about what you consider to be the basic essential workup of a patient with a suspected MPN? And also, you referred to different classifications in terms of staging and risk. Can you talk about that?

PROF HARRISON: Yes. So, I mean, the basic workup for a patient with MPN, you have to start with your kind of ABC of clinical medicine, of course. You’ve got to take a look at the patient in front of you, think about the blood workup you’ve got, think about whether this could be due to a reactive cause. Typically in London we’re asking questions, of course, about alcohol and cigarettes if we’re dealing with an erythrocytosis. For a thrombocytosis, we’d be thinking about excluding, for example, connective tissue disease, another malignancy, iron deficiency.

And for a patient with myelofibrosis, usually that’s clear of the big spleen and the leukoerythroblastic film, but we have to bear in mind other malignancies that could be mirroring this. The kind of things we’ve seen in clinical practice would be myelodysplasia, mastocytosis, CML — all of those things need to be excluded.

You do your clinical examination. Usually in my practice in central London, we’ve been doing a molecular profile, usually looking at the driver mutations. For PV, an EPO level is helpful, examining the spleen and then taking a look at the patient. And if you really think you’re dealing with an MPN, then using the current WHO 2016 classification, you really have to do a bone marrow and try to get that in front of a hematopathologist who can exclude an underlying reactive cause.

For PV, if you’re lacking a JAK2 V617F, the EPO is low and the patient’s got a panmyelosis, you’d be thinking about doing an exon 12 mutation, but you might as well also go to your bone marrow. For ET now, really you have to be doing a bone marrow in this setting.

But I think it’s always important to exclude CML, particularly in a patient that’s got atypical features. Remembering to use BCR-ABL, thinking about doing cytogenetics, because many of our newer scoring systems utilize that as well. And then you have to go on and think about the risk stratification for your patient.

For ET and PV, we’re still using very much our older fashioned how old is the patient? Are they greater than 60 or less than 60? Have they had a thrombosis or not? Really the molecular data is still being explored in that setting. But for myelofibrosis, then we’re beginning to introduce their molecular markers. But in my practice in central London, I really wouldn’t be doing an extended gene panel unless I had a patient who had triple-negative atypical features or I was thinking about taking that patient down the transplant pathway. And that typically would be a patient with a prognosis of less than 5 years.

Pathophysiology of MPN and therapeutic approach to patients with MPN

DR LOVE: John, I was reflecting back on when the initial data came out in myelofibrosis with ruxolitinib and we were hearing these incredible stories of patients having dramatic relief of symptoms. And one of the things that was a little confusing in the beginning was the fact that JAK inhibitors work on patients whether they have JAK mutations or not in terms of symptom relief. Can you talk a little bit about your vision of the pathophysiology of the symptoms of myelofibrosis and how that correlates with some of the interventions we’re looking at?

DR MASCARENHAS: Yes. I think one thing that has clearly emerged is that a JAK inhibitor, whether it’s ruxolitinib or any of the inhibitors that are in clinical development, some of which are very far along, they really aren’t JAK2 V617F specific.

One thing that seems to predominate across all patients with myelofibrosis is hyperactivity of the JAK-STAT signaling pathway, whether it’s through activation of JAK2 V617F or exon 12 mutation or calreticulin or MPL. And probably also in patients with triple-negative disease, there seems to be hyperactivity of this pathway and, therefore, an opportunity for inhibiting the pathway with a JAK inhibitor. And even JAK inhibitors are a general term, because their selectivity can vary between selectively JAK2, JAK1 and then other pathways that also may be relevant to the pathogenesis of these diseases.

Ruxolitinib, for example, is a drug that has an effect regardless of the mutational status and does result in improvement, particularly in symptom burden, which we believe is related to depression and inflammatory cytokine signaling as a result of this hyperactive pathway. It also affords patients significant reduction in splenomegaly.

And these two components really probably drive why patients do better overall with ruxolitinib. But we know, particularly after all of these years that have gone by since the FDA approval in 2011, that with its success there are limitations in terms of our expectations. It’s not a curative therapy. And sometimes, as Claire knows, we will use ruxolitinib and then assess for each patient when the best time for that patient is to move on to hematopoietic stem cell transplant, if that’s in their future. Because the reality is that eventually we know that for any given patient that the time to discontinuation can vary, but the median time to discontinuation is approximately 3 years. Although ruxolitinib is a huge step forward in the treatment of myelofibrosis, we still have to be wary of the fact that we need to plan for second-line agents and/or the timing of transplant.

Case (Prof Harrison): A man in his early 60s with symptomatic primary myelofibrosis (MF) and mutations in JAK2, ASXL1, EZH2 and TET2 receives ruxolitinib as first-line therapy

DR LOVE: So let’s bring out a couple of cases, just to get more of a clinical flavor of how this plays out. And Claire, I see you have a 62-year-old man who’s actually heading to take a vacation in Florida, so maybe he’s going to come down here and check out the beaches maybe, I don’t know. But can you talk about how he presented and kind of what happened?

PROF HARRISON: Sure. This is a man I recently saw in my practice, actually. Still working but had been increasingly struggling with his work, which is, he’s got a fair degree of manual component to his work. He was lethargic. He’d lost weight. He was noted to have splenomegaly, although he hadn’t detected that himself. And he was found to have a mild anemia. We went through the diagnostic pathway, and he had a JAK2 mutation.

He started ruxolitinib for splenomegaly and symptoms, and looking at him, realizing that here’s a guy who’s 62, he’s fit enough to take to a transplant, we’re using ruxolitinib really here as a bridge until we can work out what we’re doing with this man. In the meantime, we find that he’s got a bunch of — a JAK — extra mutations, so ASXL1, EZH2, both of which John mentioned are high molecular risk, and then TET2, which is probably not of so much significance.

I started him on 15 mg BID and then titrated the dose up. But I know this man wants to go to Florida. It’s a really important holiday for him, so I would normally be pushing more towards a maximal benefit. And I think this is something that we’re increasingly understanding in the field, that we don’t just push until the patient gets a response. For these patients, we want to push to maximal response. But real life interjects. He wants to go to Florida. I don’t want him to have a blood transfusion or to miss his holiday. My plan is to titrate the dose up when he comes back from his holiday.

But this man is worrying me a bit, because although he’s got great relief in his symptoms, his spleen isn’t really going down. For me, that’s an alarm bell. When this man comes back, I’ll push the dose up to 20, probably 25. But if his spleen isn’t coming down, for me that’s an alarm bell, and I’ll be wanting to make this patient move along the pathway either with an experimental agent or to transplant.

Dosing and quality of life with ruxolitinib

DR LOVE: John, any comments on this case?

DR MASCARENHAS: No. I mean, I think that this is a scenario that we frequently we see. And I think that there’s 2 schools of thought, and although I totally agree with Claire in terms of optimizing the dosing of ruxolitinib, I tend to be personally more cautious in my titration up, as I’m trying to balance the cytopenias that can arise. And eventually if I or the patient is not happy with the response in terms of symptom burden or reduction of splenomegaly, it really starts to prompt the discussion regarding second-line options and clinical trial options, either in combination or in lieu of ruxolitinib or to reevaluate the role of transplant in an individual.

DR LOVE: How much of an impact did therapy have on this man’s quality of life and functional status, Claire?

PROF HARRISON: Significant, actually. Because when I first saw him, he was thinking about reducing his working hours. And he really doesn’t want to do that, because he wants to kind of maximize his income. And ruxolitinib so far has enabled him to be able to work more. He’s got more energy. But he’s not really that bothered by his spleen, which is something that I think is quite interesting. This guy’s got a spleen that’s 20 centimeters below the costal margin.

DR LOVE: Wow.

PROF HARRISON: And he’s not really bothered by it. But it’s improved his quality of life. But for me, I’m still looking at this man and thinking, I can see ASXL1. I can see EZH2. This is a guy I’m going to watch more carefully. And if I don’t see a spleen response in a 62-year-old, I’m really going to want to be perhaps more aggressive. But he’s quite happy. But, of course, like many of our patients, the difficult conversation is the conversation about mortality and risks of transplant. And for these patients, they’ve never really encountered that. They’ve not been in hospital before. And talking to them about a procedure that has a 10% to 15% mortality is quite tricky really.

Case (Dr Mascarenhas): A man in his early 60s receives ruxolitinib for symptomatic primary MF and experiences a return of symptoms 3 years later

DR LOVE: John, let’s bring in another patient. You have a 63-year-old man.

DR MASCARENHAS: The patient that I had was a 63-year-old gentleman who was placed on ruxolitinib for symptom burden and spleen reduction, who initially had a response. And then approximately 3 years of enjoying that response, there’s been some creep-back of his symptom burden and some depression in his platelet count.

And this is a common scenario that we see, where patients can have a tremendous improvement in their functionality. And I recall the days — and I know Claire knows it well — of the COMFORT studies, and in the US at least, it was a randomized, placebo-controlled study, double blinded, so we didn’t know as investigators, and the patients didn’t know, but it was so clear who was getting ruxolitinib, because the patients who were on investigational product were jumping out of their seats within a month to come in to see the study team, whereas the patients who were on placebo still were quite symptomatic and debilitated.

It’s a drug that really affords a lot of benefit and quality-of-life benefit and functionality. But when I start to see patients start to lose that — and sometimes it’s subtle. Sometimes it’s not so subtle. Sometimes you see creep-back of original symptoms like night sweats or fatigue, spleen-related complaints. That to me is a signal that one should not sit and watch the patient get worse before reacting. And that’s really a time when we start to reevaluate the clinical trial options or the necessity for transplant.

And again, one of the things that Claire and I have spoken about, and others, at conferences is, when is the optimal time when someone is enjoying the benefits of ruxolitinib to go to transplant? And it is extremely difficult to take someone who’s in their 60s, who’s engaged in life and a family and who’s feeling well on ruxolitinib and discuss admitting them for a month in the hospital with a mortality of 10% to 15% and then all the other morbidity that can be associated with transplant after transplant, even with a successful transplant. That’s an extremely difficult conversation. I like to use the mutations to help color that picture and give it some texture as one tries to make decisions that are often difficult to do.

Bone pain associated with MF; use of bone marrow biopsy to monitor disease progression

DR LOVE: Yes. For the rest of us in oncology, it was so amazing to hear when the COMFORT trials were presented, these stories of these patients. I don’t know if there’s ever been anything in oncology that’s been so dramatic in terms of symptom relief the way that story played out as you just described. What an amazing thing to have happen.

One other thing you put here in the workup I was curious about, John, is that he had diffuse bone pain. Maybe you can talk a little bit about that. And I know you marrowed him.

DR MASCARENHAS: Yes. We see bone pain as a frequent symptom with myelofibrosis. It’s often diffuse. Often it’s in the long bones, and patients will feel it usually the most at night. Sometimes it affects their quality of life and their ability to sleep and contributes to fatigue. And this is a symptom that is often addressed quite well with ruxolitinib and other JAK inhibitors.

But it’s one of the symptoms I often see that makes me nervous if patients are doing well on ruxolitinib and then start to develop worsening bone pain. That really is continuous and often not responsive to acetaminophen or NSAIDs, and that usually would prompt me to repeat a bone marrow biopsy to evaluate whether there’s been disease evolution. Because with myelofibrosis, although we look at the blood smear and you look for leukoerythroblastosis, patients can evolve within the bone marrow. And we’ve seen patients evolve from chronic to accelerated blast phase disease. And their peripheral smear is sometimes not even changed in a concerning way. So bone pain is one of those symptoms that — particularly if they’ve not had it and they’ve been enjoying relief of their symptom burden on ruxolitinib — that would prompt me to reevaluate with a bone marrow biopsy.

DR LOVE: And so when he initially was started on treatment, he was having bone pain? It went away and then it came back?

DR MASCARENHAS: Yes, it came back. I mean, he had a constellation of symptoms that really went away. And he was almost symptom free for quite some time. And then at each visit, we would tease out that some of the original symptoms were beginning to slowly return. And at first, actually, he wasn’t reporting them actually at first. And then as they got worse, he acknowledged that they had been coming back over a period of months.

DR LOVE: I see that his repeat bone marrow showed no progression, but no longer hypercellular and MF3. What was your take on that?

DR MASCARENHAS: Yes. So he was MF3 to begin with. He really didn’t reduce his fibrosis grade on ruxolitinib. And he had a bone marrow cellularity that was about 80% when he started out and was about 20-something-percent at the time of marrowing him. And this is a phenomena that we sometimes see. It’s almost like a bone marrow exhaustion or failure and often accompanied by low platelets and often a bigger spleen. And I think biologically there’s probably a shift in hematopoiesis, at least malignant hematopoiesis, from the bone marrow to the spleen.

It’s a finding that we see frequently. And at least the saving grace in this case was that he had not transformed to an acute myeloid leukemia, which is a whole different story.

DR LOVE: Claire, any thoughts about this case?

PROF HARRISON: I think it’s a really interesting case, and I think it illustrates a few points. I think the first thing is about the importance of talking to your patients about symptoms.

And we try to use a structured symptom assessment. And more and more of our patients are coming in with an app or whatever, and I think that’s really important, because it triggers the conversation, and sometimes it’s not the patient, I suppose, that talks about this.

I completely agree. Bone pain for me, that’s really a very worrying symptom, and that’s one in an elderly population that they might not mention, because they often think it’s because of arthritis or something else.

And I think your point, John, about blast transformation is also a really important one, but I would also say developing thrombocytopenia, for me, is a worrying sign. And I’m always really careful to make sure my patients, even if they look well on rux, we have a look at blood film at least a couple of times a year and watch their LDH, because early evolution of disease may be silent. These patients may still have good symptom and fairly good spleen response, but they may have an upturn in circulating blasts. And we all know, once they reach blast transformation, you know that’s pretty much unsalvageable and a big disaster for a patient that would otherwise have been suitable for more intensive therapy.

Design, eligibility criteria and results of the Phase III COMFORT-I and COMFORT-II studies evaluating ruxolitinib for patients with MF

DR LOVE: I’d like to reflect back on the COMFORT studies that we just referred to. And John, maybe we can begin just talking about specifically how they were done, where they were done, what the designs of these trials were.

DR MASCARENHAS: Sure. There were 2 randomized Phase III studies, COMFORT-I and COMFORT-II. COMFORT-I was conducted here in North America and COMFORT-II in Europe, and Claire can correct me — I think it was about 9 countries in Europe. And COMFORT-I was a randomized study to ruxolitinib versus a placebo, and it was double blinded in patients with intermediate-2 and high-risk disease by the International Prognostic Scoring System, with at least a 5-cm spleen palpable below the left costal margin.

And this was a positive study that met its primary endpoint of 35% spleen volume reduction, which equates to at least a 50% reduction in palpable splenomegaly. And was clearly superior to placebo after a 6-month course study period, and then patients on the placebo arm were able to switch over to study drug, become unblinded after 6 months of therapy.

COMFORT-II was similarly constructed, except it was best available therapy. I thought that was a much better design in many ways, much more compassionate, actually. Patients could get therapies that had some activity in myelofibrosis. And again, after 6 months of therapy were evaluated for spleen response as a primary response. And in both studies symptom burden improvement, quality-of-life improvement were also measured.

And in both studies it was clearly evident that ruxolitinib was superior to, obviously to placebo, but also to best available therapies in terms of reduction in splenomegaly and improvement in symptoms. And in symptom improvement, it was really across the board whether it was symptoms related to what we think are cytokines, like constitutional symptoms or spleen-related symptoms, energy level, fatigue level, appetite — it was pretty remarkable and obvious. And patients did very well on ruxolitinib.

Those 2 studies were really the pivotal studies. And Claire Harrison led the COMFORT-II study that led to the approval of ruxolitinib, our first real effective therapy in myelofibrosis in 2011, at least here in the US.

A pooled analysis of overall survival in the COMFORT-I and COMFORT-II trials of ruxolitinib for the treatment of MF

DR LOVE: Claire, you did publish a pooled analysis of COMFORT-I and -II looking at survival. Can you talk about what we know at this point about the impact on survival and also the impact on the biology of the disease, the amount of fibrosis, et cetera?

PROF HARRISON: Yes. We were already seeing striking benefits for our patients. And it’s interesting, John, that you feel that COMFORT-II was a more ethical study, that we were looking at therapies that were effective for the disease, because, sadly, where in reality the best available therapy arm of COMFORT-II, those patients didn’t fare any better than the patients you were treating in New York with placebo.

But because the 2 trials had very similar eligibility criteria, we’ve been able to do some really interesting analyses across this cohort of patients. And these were really important studies, so the first randomized studies we’ve ever done in myelofibrosis. And what was really apparent to us is we’ve heard, really well described by John, was that these patients almost had a Lazarus effect. They were getting their quality of life and their functionality back.

We didn’t anticipate that patients would have a survival benefit, but as soon as you were treating a patient with this drug, you could see that the patients were living a better quality of life. And quite quickly — I think it was about 15 months into the COMFORT-I — after the COMFORT-I study analysis, there was a survival benefit. And we saw it later in COMFORT-II because, for various reasons, we did a 2:1 randomization, et cetera.

We’ve done a pooled analysis, as you referred to, of the survival benefit for these patients. And this analysis has been done correcting for crossover, because both studies allowed a crossover, for obvious ethical reasons. And using this analysis, we’ve been able to show that approximately there’s a 30% prolongation of life for patients treated with ruxolitinib. But we also saw some other really interesting points that are useful in general practice.

Predictors of overall survival benefit with ruxolitinib

PROF HARRISON: The first is, these kind of ideas, how can you tell that your patient is benefiting? We can see symptoms. We can see spleen. But how do you know your patient’s getting a survival advantage?

The pooled analysis looked at a snapshot of how was the patient’s spleen response at 24 weeks? And for patients that had at least a 10% volumetric reduction, which equates to quite a modest palpable spleen length reduction, these patients were having a survival advantage. That’s one useful thing that you can use in clinical practice. If your patient is having a spleen reduction by 6 months, they’re likely to have a survival benefit. A corollary, interestingly, if your patient gains weight, they’re also likely to be having a survival benefit.

And then we’ve been able to do a number of molecular studies. For example, in COMFORT-II we had ethics to look at the driver mutation clearly show that response was independent of that, as we’ve already discussed. But we were also able, for example, able to look at these high molecular risk mutations and show that — for example, the patient case I discussed. This man has got 2 high molecular risk mutations, but from COMFORT-II we were able to see that despite this, he’s just as likely to have a response.

Data from the MD Anderson suggest that time to treatment failure may be shorter, but we shouldn’t be deterred from putting a patient on the drug just because he has these mutations. Similarly, we’ve looked at cytokine patterns, and we’ve been able to examine bone marrow responses for patients. But this data is really perhaps a little bit more disappointing in a way.

There are a handful of case reports that suggest patients may well have normalization of marrow morphology. We have one of those patients in my practice in central London. This patient also had a striking JAK allele burden reduction and a resolution of a cytogenetic abnormality. But, and, in fact, this patient actually continues on ruxolitinib, having started in in 2009, so he’s a very atypical patient.

We shouldn’t be expecting, really, nor monitoring in clinical practice for response, I think, in terms of marrow, and there’s no benefit, particularly in measuring cytokines. But we can see, of course — I’ll just return to the theme of we do expect to see survival benefit. But the trick is optimizing the dose to get the best advantage for your patient without costing them too much in terms of cytopenia. And then the real challenge in the field still, as John alluded to, is when is your patient losing that response and when should you do something about it?

Biologic rationale for the survival benefit with ruxolitinib in MF

DR LOVE: It’s really interesting about the survival benefit, John. Any thoughts about how much of it is slowing down the disease process and how much of it is just better nutrition, people with better performance status?

DR MASCARENHAS: There’s a couple of things that we’ve learned over the years from the COMFORT studies and subsequent studies, that JAK inhibitors, as a class, definitely affect an aspect of the disease pathogenesis that drives the myeloproliferation and the elaboration of cytokines that make patients feel ill.

None of the drugs at this point, in my opinion, and ruxolitinib included, what they don’t seem to actually achieve is MRD or measurable residual disease. They rarely, outside of a few cases, they rarely induce complete bone marrow pathologic remissions or molecular remissions. Clearly what we’re really looking at is modification of disease aspects that contributed to the clinical picture in terms of splenomegaly and symptoms but aren’t really reducing the burden of the disease in a significant way.

I think we have to change our expectations. My expectation when I treat patients with ruxolitinib or other drugs is that this is a drug that can profoundly improve your quality of life and your symptom burden and your functionality. You can gain weight back. And those things are probably, what I believe, leads to improvement in survival and outcomes of patients who don’t get it or who are even delayed in getting ruxolitinib treatment. Because if you look at the COMFORT study designs, based on the fact that these patients crossed over from either placebo or best available therapy, there was a delay in getting ruxolitinib. And even that delay appears in intention-to-treat analysis, really appears to affect survival.

Which begs a whole other question, which is, should we be treating patients with myelofibrosis earlier on and not waiting for them to become so sick? And that was a study called the ReTHINK study that was unfortunately stopped due to low accrual, which was trying to take patients with earlier disease, lower-risk disease — by DIPPS score, but perhaps with a high molecular risk feature — and taking them to ruxolitinib therapy earlier in their course.

I do think ruxolitinib improves survival. And those studies that have been presented by Claire and others, particularly using a statistical tool to show what would happen if patients didn’t cross over and their outcome, I think, really speak to the fact that ruxolitinib does a fantastic job in reversing a lot of the cachexia disability and wasting that you see in patients with myelofibrosis. And that probably is what’s predominantly driving the survival benefit.

DR LOVE: It was mentioned before about JAK allele burden. Is that useful in terms of predicting treatment benefit or any other type of allele assessment, John?

DR MASCARENHAS: At Mount Sinai and I think a lot of places, we do quantitative PCR or we get next-generation sequencing, in which we get a variant allele fraction. And the reality is, I don’t know that we know what to do with it. I mean, as Claire pointed out, we get a lot of things that I don’t think we know what to do with at this point and probably, depending on where you practice, you may get more or less data back.

There is a concept that variant allele frequency, those patients with a higher allele frequency, above 50, have more of a myeloproliferative-type phenotype — so higher white counts, bigger spleens — and may, in fact, be more amenable to JAK inhibition than patients with a lower frequency of the disease that might have more of a myelo-deplete, lower counts, smaller spleens, more pancytopenia. There may be a differential effect on treatment with ruxolitinib and, perhaps, with other JAK inhibitors, depending on the variant allele frequency of the driver mutation.

Perspective on the use of ruxolitinib in patients with MF in the absence of symptoms

DR LOVE: Claire, we were talking before about various staging systems that are used and how that was integrated into these trials. But from a practical clinical point of view, is the real issue whether the patient is symptomatic or not? For example, can you describe a situation where you would use ruxolitinib in a patient who was asymptomatic and didn’t have splenomegaly? Claire?

PROF HARRISON: Yes, that’s a great question, actually. Of course in London we’re a little bit limited by NICE, which is our kind of reimbursement authority. Thinking hypothetically, would I use ruxolitinib in a patient with myelofibrosis? He was completely asymptomatic without splenomegaly. I think my answer here would be no. I could imagine that patient might well have a leukocytosis or a thrombocytosis. Would I be worried about that in a patient otherwise?

Of course leukocytosis is an adverse prognostic marker, and if the leukocyte count was progressively increasing although a blast appearing, that’s a patient who needs a therapy. But I’m not sure that I would be using ruxolitinib in that context.

What agent would I be using? That’s a great question. And I think right now we don’t have many agents at our disposal. Of course if I felt the patient was progressing rapidly, I’d be thinking about transplantation, and I might use the drug to hold them.

Across the English Channel in France, they have published recently some interesting data about interferon in early phases of myelofibrosis. And this is a debate that still rumbles on in the myeloproliferative community: how good is interferon in early-phase disease? Might this actually be something that holds the disease or not? And I think honestly, we really don’t know. But to answer your question in short, a patient without splenomegaly and without symptoms, no. I wouldn’t be using ruxolitinib at the present time.

DR LOVE: It’s really interesting, this idea of early versus late, because, actually, the trial itself kind of looked at that. What fraction of people, Claire, actually did cross over to ruxolitinib?

PROF HARRISON: In Europe, unfortunately, the trial initially didn’t have crossover. So we did have some patients who — they were sitting in the waiting room next to a patient who’s on ruxolitinib, springing out of the chair, as John so nicely put it, and they were feeling worse and worse.

We have a number of patients in our control arm in the COMFORT-II study who stopped early. And the vast majority of our patients crossed over to receive ruxolitinib once that was permitted. But in the COMFORT-II trial, actually, that was later.

Just thinking about that theme there actually, we’ve got some nice evidence in the real-world setting looking at early versus late ruxolitinib, and that comes from the work of Francesca Palandri in Italy, where she’s gathered really impressive data in around 400 patients with myelofibrosis. And she did a multivariate analysis looking at factors that predicted response.

And her analysis actually suggested that if you were to start ruxolitinib within 2 years of diagnosis versus beyond 2 years, then the patients were more likely to get a spleen and also, interestingly, a symptom response. And there are a number of other factors that came up in that analysis. But just thinking about should you move your patient early? Yes, I think you should if they’ve got symptoms and spleen. After all, our standard therapies aren’t effective for those patients.

Side-effect profile of ruxolitinib

DR LOVE: I want to hear a little about the issue of cytopenias, particularly in patients on JAK inhibitors. But John, just also curious about other issues that come up, both in the trial and your own clinical experience in terms of side effects or toxicities with ruxolitinib. For example, have you seen hyperlipidemia?

DR MASCARENHAS: Yes. For the most part, ruxolitinib is a very well-tolerated drug. And I always preface the conversation about ruxolitinib with patients with the statement that most patients actually do very well on ruxolitinib and don’t have tolerability issues. But there are certain things that we’re aware of that we need to pay attention to.

Cytopenias are expected, to some extent. And within the first 2 to 3 months of using ruxolitinib — and of course in a dose-dependent fashion — you do expect to incur some treatment-related anemia and thrombocytopenia. And warning the patient ahead of time that that can happen so that they don’t look at that as a failure of the drug is very important, because what you don’t want is for patients to make the mistake of being disappointed or being afraid that the disease is progressing rapidly or the drug is failing them in some way to come off therapy prematurely.

Talking patients through that initial several months and expectations and tempering expectations about what the blood counts will do is very important. And even incurring some degree of transfusion requirements during the first several months of therapy to me is not treatment failure and as par for the course with ruxolitinib.

Most patients really tolerate it well. Do not have nausea/vomiting/diarrhea or other side effects of that sort.

We do appreciate that there’s probably an increased risk of certain infectious complications — zoster and shingles, for example, has been noted in probably about 5% of patients that are treated. At least in the US, we have access to a dead zoster vaccine, shingles vaccine. I recommend that to patients who have not been vaccinated and actually patients who’ve been vaccinated previously with the live attenuated vaccine.

And I do counsel patients on the fact that they could be at increased risk for having both bacterial and viral illness, and although infrequent, there have been some other rare, atypical forms of infections that have been picked up. And all patients with MPNs, regardless of treatment, but, perhaps, more so if they’ve been heavily pretreated with hydroxyurea previously and are on ruxolitinib, I insist on skin exams looking for surveillance for nonmelanoma skin cancer and general health maintenance.

DR LOVE: I had mentioned about hyperlipidemia. Do you see that?

DR MASCARENHAS: Yes and no. I mean, we see it if we look for it. I’ve got to be honest. I don’t really look for it.

There’s a clear reversal in a lot of the metabolic profile that you see with ruxolitinib treatment, which again, I think, speaks to the fact that there’s a fundamental change in the cancer cachexia that really plagued the patients. And you can get weight gain. And in most patients, that’s welcomed. And they look better. They feel better. They carry it well. And the polycythemia vera patient population, we’re beginning to note that sometimes that weight gain is not great, could be more than significant. And with that, you can see changes in the lipid profile and even, to some degree, like a metabolic syndrome: hypertension, et cetera.

We and others are interested in antidotes that may counteract those aspects. We have a trial that will open up looking at metformin as an antidote to weight gain with ruxolitinib in polycythemia vera patients, but we are beginning to pay attention, and we should pay attention also to the lipid profile and obviously glucose and blood pressure.

Management of cytopenias associated with ruxolitinib

DR LOVE: Claire, anything you want to add to that in terms of tolerability issues with ruxolitinib? And also some of the experiences you’ve had in terms of thrombocytopenia and anemia and how you manage those patients.

PROF HARRISON: I think that generally I completely agree with John. This is a very well-tolerated drug, but there are some predicted toxicities, as he mentioned.

As I look at my patient, I have this broad rule of thumb. If I dose them with a standard dose, I expect a 10% to 20% fall in their hemoglobin and roughly a 40% drop in the platelet count. If I’ve a patient whose got borderline anemia and they’re frail and elderly, I might use a lower starting dose. Because clearly these toxicities in the main are actually dose related. They’re also time related.

John mentioned very nicely, you warn your patient about anemia. Generally the hemoglobin is returning towards but not reaching baseline by about week 18 to 20.

I think the other thing is just looking at those patients who have had recurrent infections. We are careful in central London — all our patients get hep B, hep C, HIV, for example. And then mainly the big issue, I think, in dosing ruxolitinib is thrombocytopenia. And that can really limit effective dosing. We can manage anemia. We can maybe add danazol. We might be able to add an erythropoietin — curiously enough, that tends to work in some patients. But thrombocytopenia is a big challenge in terms of getting an effective dose.

Potential correlation between JAK inhibitor therapy and second primary cancers

PROF HARRISON: And then I think 1 more thing in terms of tolerability and unexpected side effects that’s come up recently is this issue and question about non-Hodgkin’s lymphoma for patients treated with a JAK inhibitor. Not only ruxolitinib, again, also in the paper that was published by the group in Austria and France, there were a cohort of patients who had received other JAK inhibitors, such as fedratinib. And what they showed was that in patients with a preexisting B-cell clone, there appeared to be in that cohort of patients roughly 15-fold increase in the risk of B-cell lymphoma.

Now, and that paper did have some quite interesting kind of mechanistic data with a mouse model, for example, looking at JAK1 inhibition. I have to say that in my practice in central London, I haven’t noticed that. And I think it’s an important discussion point, right here and now — should we be afraid of using ruxolitinib in a patient that might have CLL that’s inactive or a previous history of a non-Hodgkin’s lymphoma? Because there does appear to be an association with these malignancies and MPN.

And there are 2 subsequent publications looking at cohorts of patients not really picking up this issue. And so I think while it’s a live issue that’s been raised, for me it’s not an issue that would stop me using a drug or a class of drug that is effective for a patient in need.

DR LOVE: Just to clarify, though, in the COMFORT trials, was anything seen?

PROF HARRISON: No. This was not seen in these patients. No. But in fairness, the prevalence of these lymphomas, I think there were 6 patients who developed lymphoma in a cohort of about 400 patients. We might have expected it, if the data held true, 1 or 2 in the COMFORT studies, but we didn’t see them.

Another point that the authors of that paper make is that these lymphomas were very, very aggressive, and the patients actually very rapidly developed aggressive disease. It potentially could be missed. Another reason why I think it’s important. But for myself it doesn’t change, at the moment, how I treat or how I monitor my patients.

DR LOVE: John, agree or disagree?

DR MASCARENHAS: One hundred percent agree. I mean, any publication like that will cause concern in the physician and patient community, but like Claire, on my side I can’t really say that I’ve really seen this as a major issue. And we know historically that MPNs are associated with a higher risk of secondary primary malignances. It’s unclear to me what’s really contributing to that in any given individual. And I would never keep a beneficial drug like ruxolitinib or other JAK inhibitors from a patient out of a concern that right now remains somewhat unclear and needs more evaluation.

Spectrum of activity and ongoing investigation of novel JAK inhibitors in MF

DR LOVE: I want to go on and hear about a couple of more cases. In a second, Claire, I want to ask you about your 67-year-old man. But just to precede that a little bit, can you talk about what we know right now about other JAK inhibitors? And, of course, I want to ask you about fedratinib, but just maybe to begin with an overview of what are the other JAK inhibitors out there. And, at least from a basic pharmacologic point of view, how they compare to ruxolitinib.

PROF HARRISON: Sure. Actually, we’ve seen a host of other JAK inhibitors that have been tested in clinical trials. None of them so far have reached the end of the road and have been approved by the FDA and the EMA, but quite a few are quite close to that.

There are a host of JAK inhibitors that were halted in clinical development, largely due to either lipase-related toxicity or, interestingly, neurological toxicity. None of these drugs, relating to your comment about their pharmacology, are completely identical. None of them are mutation specific, so they all bind to the kinase domain. And they have variable selectivity for JAK2. So, for example, ruxolitinib, as we well know, is a JAK1 and 2 inhibitor versus fedratinib, which is probably a little bit more selective for JAK2 but may have some other activity — for example, FLT3 — and perhaps also bromodomain inhibitory activity.

Contrast momelotinib, which is a little bit more similar to ruxolitinib but may have an effect on the hepcidin pathway, which might explain that drug’s differential activity with regard to anemia. And the pacritinib, which again, seems to be more JAK2 selective but also has other effects on FLT3, IRAK, for example. Thinking about the JAK inhibitors and those that are close to being in clinical practice or those for whom colleagues may have actually had an experience of treating patients.

We have just a selection now of fedratinib, which was halted in clinical development because of concerns about Wernicke’s encephalopathy. And that’s right now in a couple of trials known as the FREEDOM studies, and I think there are a few patients, John, in the US who are now enjoying the benefits of fedratinib in the FREEDOM trials.

We have momelotinib, which is, again, entering a Phase III study. And then pacritinib, with its rather unique niche activity for patients with quite marked thrombocytopenia, which is also about to reenter a Phase III study.

Efficacy of fedratinib, pacritinib and momelotinib in patients with MF

DR LOVE: I’d like to hear about a couple of patients who got these other JAK inhibitors. But John, in general, have responses been seen in patients who progressed on ruxolitinib?

DR MASCARENHAS: Yes. So there’s 2 patient populations in which one could envision 1 or more of these other JAK inhibitors to find a niche. And as Claire pointed out, perhaps pacritinib would be unique it its ability to be delivered to patients with low platelets, where ruxolitinib is not FDA approved. The label by the FDA for the use of ruxolitinib is intermediate- and high-risk myelofibrosis with at least a platelet count of 50,000. There’s really an unmet need, which maybe is 10% to 15% of patients with myelofibrosis that have a low platelet count, a very low platelet count less than 50,000.

And the PERSIST-2 study, which was a randomized study of pacritinib to 2 different arms, 200 mg twice a day, 400 mg once a day and best available therapy, in which ruxolitinib was also used, was clearly a drug that had activity in this low platelet population, where the drug could be delivered for the most part safely and effectively.

This could be used up front, for example, in low platelets. But also could be used second line in patients who may be failing ruxolitinib or may be not even failing ruxolitinib but may be sensitive to ruxolitinib in terms of thrombocytopenia. And one could turn to pacritinib in that sense.

I think momelotinib is an interesting drug, because the data, particularly the Phase II data with the long-term follow-up, was quite impressive for anemia responses. As Claire pointed out, this is probably perhaps an effect of one of the other kinases that are inhibited in its profile. And if I had a patient who had significant transfusion-dependent anemia, even up front or, again, second line, I would consider momelotinib as an option.

And then fedratinib was really impressive in the JAKARTA-2 study in spleen reduction, second line after ruxolitinib failure. That would be a drug in my patients, particularly with progressive splenomegaly on ruxolitinib — whether they initially had a response and then lost it or never had a response — that then I would turn to fedratinib if it was commercially available. I could see a future, hopefully not too long from now, a future where there would be potentially multiple JAK inhibitors available for these niches, either up front or second line.

Activity and tolerability of fedratinib for MF

DR LOVE: I’d like to hear a little bit more about some of the data related to this and also a couple of these cases. Claire, can you kind of go through the fedratinib story, beginning with the issue that came up about the Wernicke’s that you referred to and then where we are right now in terms of data?

PROF HARRISON: Sure. I mean, fedratinib was really the second JAK inhibitor that was taken through clinical trials after ruxolitinib. And, in fact, there were 2 quite large studies conducted. Earlier studies with this drug show the similar profile of benefits to ruxolitinib. However, the drug has got a longer half-life, so it’s only once-a-day dosing and, slightly different to ruxolitinib, there’s clearly some gastrointestinal toxicity with this agent, probably because of its FLT3 activity.

In the Phase III studies, the Phase III study was, again, a placebo-controlled study. It was the last placebo-controlled study that we did in myelofibrosis, actually. Here, again, we were looking at 2 doses of this drug, 400 mg or 500 mg, almost an identical patient population to the COMFORT studies and completely identical endpoints: spleen volume and symptoms.

The drug at both doses was clearly effective for both parameters. The study was interesting, because it used a kind of confirmatory check that there was a spleen response a month later, which made it a little bit confusing. But broadly, the effectiveness was about the same as ruxolitinib, maybe a bit more. Maybe the symptom response a little bit less. But, of course, at this time, ruxolitinib was quite widely available. This study allowed crossover.

At the same time as we were seeing these benefits, which had to be tempered by managements of GI toxicity and a little bit more anemia, we were very excited to suggest the possibility of doing a second-line study.

Results of the single-arm, Phase II JAKARTA-2 study evaluating fedratinib in patients with MF previously treated with ruxolitinib

PROF HARRISON: Myself and a number of other investigators said, “We’ve got this bunch of patients. They’re starting to lose their response to ruxolitinib, or maybe they’re a bit thrombocytopenic. They’re on a rigid study. We’re having to take them off or reduce the dose to a less-than-effective level. Can we do a second-line study?” And, hence, the birth of JAKARTA-2, which we published just a couple of years ago, and then we presented a reanalysis of that data recently at ASCO and the EHA.

And so the JAKARTA-2 was the first kind of formal second-line study looking at patients after ruxolitinib. And in the initial analysis, it was very complicated, because right at the time when we were beginning to see all these exciting responses with around 50% of patients getting a spleen response and 20% to 30% of patients getting a symptom response, there were these concerns about Wernicke’s encephalopathy arising with fedratinib.

The drug went on a clinical hold. Disaster. We had to take all the patients off, give them all thymine replacement. And there, at the end of the day, there are 9 potential patients that may have Wernicke’s encephalopathy. And these cases were all very, very carefully evaluated.

In the JAKARTA-2 study, I think there was 1 patient who probably had hepatic encephalopathy. We have to remember, these patients are often rather frail, often got comorbidities. Maybe they’ve also got splenic vein thrombosis.

And the index patient for Wernicke’s encephalopathy actually was an oncology patient — because there’s been an interest, of course, in developing JAK inhibitors outside of the myelofibrosis field — and this was a lady who was cachectic, malnourished, developed GI toxicity, refused enteral feeding and developed Wernicke’s encephalopathy.

DR LOVE: What kind of cancer did she have, incidentally?

PROF HARRISON: This is a patient who had breast cancer, I think.

DR LOVE: Breast cancer. Wow. Interesting.

PROF HARRISON: I’m not completely certain. Yes.

DR LOVE: I know it was looked at in pancreas for a while.

PROF HARRISON: Yes. And it’s also been looked at in the breast cancer setting, and you can see why. Because if there are constitutional benefits for these patients with advanced cancers who are cachectic — in there’s centrality of a kind of JAK-STAT signaling, for example —

DR LOVE: Right.

PROF HARRISON: And the lack of toxicity in general of these compounds.

At the end of the day, there were 9 potential patients. And there was a very careful central review looking at all of these patients. I learned a lot about Wernicke’s encephalopathy. I learned that there were specific MRI features I didn’t even know about when I was a medical student. And at the end of the day, there was 1 definite, 2 possible. Many of the patients, their symptoms got better. Some hepatic encephalopathy. And there have been several studies now looking at high prevalence of thiamine deficiency, et cetera, et cetera.

Response to fedratinib in the JAKARTA-2 study; ongoing Phase III studies of fedratinib in the second-line setting for MF

PROF HARRISON: And so now we have the resurrection of fedratinib, which I think is important, because it is an effective drug. We know in the CML setting. We need more than 1 BCL-ABL kinase inhibitor. And that’s very likely to be true as well in myelofibrosis. These patients are very heterogeneous clinically and also at the molecular level.

Right now, interestingly, we don’t know anything about which type of patient, based on their molecular profile, might respond better. I think we could do a bit more work around that. Whether we’ll find something or not, I don’t know.

But now we are beginning to use fedratinib again in the context of these 2 studies, FREEDOM-1 in North America and FREEDOM-2 in Europe, where we’ll be looking at the drug second line. Meanwhile, we’ve reanalyzed the JAKARTA-2 data, still showing significant responses for patients. I think this potentially could be a very useful drug. And we are probably going to discuss a couple of cases that illustrate that very well.

DR LOVE: Yes. Actually, you have a patient who was on the JAKARTA-2 study. I’d like to hear about what happened to him. But before you do, just globally, what was the response rate in the second-line setting?

PROF HARRISON: In the initial paper that we published in 2017, we saw roughly a 55% response in spleen volume and about a 30% symptoms response. But there are some important nuances in thinking about that data.

First of all, because we had the study halted, in the primary analysis we used something called “last observation carried forward,” which means that we’d been doing an MRI scan at week 12. And in the statistical analysis plan, we were allowed to carry forward that week 12 measurement, and if week 24 was missing, we carried it forward.

In the reanalysis, using the more recent definitions of ruxolitinib resistance or intolerance that have been developed — although none are widely accepted — we still see roughly 30% to 35% of patients achieving spleen responses and roughly 25% of patients having symptom responses. It’s broadly in line, interestingly, with the degree of response we might see first line.

DR LOVE: Hmm, really interesting. Is there any experience with ruxolitinib after fedratinib?

PROF HARRISON: Yes. We did have that experience, because, unfortunately, we have this bunch of patients globally that we suddenly had to ring up in the middle of the night and say, “Are you walking in a straight line? Do you have any double vision? I’m going to ship you some vitamin pills” — lovely experiences for the patients —

DR LOVE: Wow.

PROF HARRISON: — and the investigators. A bunch of those patients, of course, went in the opposite direction from fedratinib to ruxolitinib. And John and I both have probably a handful of those patients. And yes, they do respond. But, interestingly, in my experience, there are some patients who may well respond a bit better to fedratinib. Quite what that patient looks like we haven’t defined yet. But yes, there are patients who respond to ruxolitinib after fedratinib.

DR LOVE: John, any experience with fedratinib on trial?

DR MASCARENHAS: No, I’ve seen a number of patients who’ve had fedratinib, but I was not an investigator on the fedratinib study. Although we’re participating in the FREEDOM study, I don’t have clinical trial experience with fedratinib. But to Claire’s point, there also may be a phenomena that may not necessarily be dependent on the specific JAK inhibitor, but there has been some postulation that even holiday from a JAK inhibitor and then reintroduction could salvage or induce some sort of response in a patient who may not have been responding or lost response previously.

Anecdotally, from my side, I can’t say that I’ve actually seen that, but I have talked to others, and perhaps Claire has seen this effect of stopping, for example, ruxolitinib and then trying to reintroduce it a period later to see if it has a response. And those would be things you would do in the setting where you didn’t have, for example, clinical trial options like fedratinib or pacritinib.

DR LOVE: Interesting.

Case (Prof Harrison): A man in his late 60s with triple-negative primary MF experiences a good response to fedratinib as second-line therapy on the JAKARTA-2 trial

DR LOVE: Claire, what happened with this 67-year-old man?

PROF HARRISON: This is a gentleman — again, interestingly, he was on the COMFORT study, classical clinical presentation of primary myelofibrosis. Interestingly, he was triple-negative. He was a little bit anemic. Hemoglobin was 95. Platelet count a bit borderline, 98. He had significant symptoms, very large spleen.

Initially, he responded well to ruxolitinib, but then we got into this situation, which we often get into clinically, where he was developing thrombocytopenia, so we had to reduce the dose. Then he was losing his response. Ultimately, we took him off trial and we maximized his ruxolitinib dose. And even on 25 mg BID, he still had residual symptoms and splenomegaly.

We were able to take this patient into the JAKARTA-2 study, and he did really well. He got 50% reduction in his splenomegaly. He got good symptom responses. And he was well maintained without dose reduction and had very little GI toxicity.

I have to say, an important clinical point is actually about understanding that this drug really does have GI toxicity and you need to give your patients prophylaxis. Standard prophylaxis is fine, but you do need to warn them about that.

This man did well on the study, but he’s one of the patients that I had to call up and say, “I’m really sorry. You have to stop the drug.” After this, we switched him back to ruxolitinib. And recently, we’ve been lucky enough to be able to get him compassionate-use fedratinib, because this is a man who actually was asking, “Can I please go back on that drug?” He’s right now just completed around 10 weeks back on fedratinib and is doing well again.

DR LOVE: Did he progress when he got the ruxolitinib again?

PROF HARRISON: He got a modest response, and then he started to get an increase in blasts and an increase in symptoms. And I think he is actually a patient in whom there’s another quite relevant clinical point.

Because symptoms are not always related to disease, this is a man who got quite bad arthritis. He’s a patient who’s got quite significant depression. Assessing his symptomatology was quite tricky. But he was definitely beginning to progress on the ruxolitinib. And being anemic, having the burden of transfusions, we haven’t had to transfuse him yet on fedratinib. Quite likely at some point in the future we will.

DR LOVE: I’d say he’s a poster child/person for being in clinical trials. Incredible. He was in the COMFORT trial, and then he goes into the JAKARTA-2 trial. And it looks like he —

PROF HARRISON: Yes. I think he probably did PERSIST-2 in the middle as well.

DR LOVE: Wow!

PROF HARRISON: I mean, this is another thing we need to be aware of. I mean, he’s a pretty atypical patient.

DR LOVE: Wow. Incredible.

PROF HARRISON: He’s lasted the stay. He always thinks we’ve got a new thing up our sleeve for him.

DR LOVE: Wow, I guess so.

Case (Dr Mascarenhas): A man in his early 70s with primary MF and cytopenias experiences an improvement in symptoms and reduction in spleen size after treatment with pacritinib on the PERSIST-2 study

DR LOVE: John, let’s hear about your 72-year-old man who ended up getting pacritinib. What happened there?

DR MASCARENHAS: So this was a patient who presented — and this is not an uncommon story — a person in their 70s who presents with what turns out to be probably primary myelofibrosis that was there for a number of years but not really picked up because the cytopenias were mild and weren’t really appreciated. And then by the time the patient comes to our attention, to the hematologist, the platelet count is quite low, in the 30,000 range. Hemoglobin was also low and was receiving transfusions at an increasing frequency, almost every 2 months. And was becoming quite symptomatic, with constitutional symptoms, and had a spleen that was about 10 centimeters, and leukocytosis.

Based on the fact that his platelet count was low and the FDA label for ruxolitinib is 50,000, this was a patient that we thought was most primed for the PERSIST-2 study in which he enrolled. And he had a clear benefit with pacritinib at 200 mg twice a day, in terms of improvement in his symptom burden and his spleen.

And as Claire pointed out with fedratinib, pacritinib is also a FLT 3 inhibitor, so one needs to be aware of the potential for diarrhea and have loperamide ready for the patient. And usually the patients seem to get through that within the first month and seem to be fine afterwards. Sometimes it just had to do with modifying the diet and the timing of the medication. And he benefited significantly.

His transfusion requirements didn’t really change significantly — red blood cell transfusion requirement. He didn’t drop his platelet count below his baseline platelet count. But he continued to feel well, and his functionality clearly improved on the drug.

Potential role of fedratinib, pacritinib and momelotinib in the management of MF

DR LOVE: Claire, just kind of curious. Assuming you could access fedratinib, and for that matter pacritinib, as freely as you can access ruxolitinib, ideally, how, if at all, would you incorporate these 2 agents into your practice?

PROF HARRISON: I think there’s a kind of evolving theme here. I think with both fedratinib, perhaps to a lesser extent than ruxolitinib, we do see evolving thrombocytopenia. We can try to temper that — for example, I have a patient where I might see some thrombocytopenia with ruxolitinib. I may choose to add danazol in the hope that I can maintain a platelet count enough to get an adequate dose of ruxolitinib. But I see that pacritinib as being potentially useful for these very tricky patients with thrombocytopenia, where we really have not much else to offer them except potentially adding steroids.

I think fedratinib is interesting, because it’s kind of evolving as a second-line drug. But, actually, the first-line data with fedratinib was pretty good. I see right now that the strategy is yes, we would develop this for a patient who’s got progressive splenomegaly and maybe also progressive symptoms.

But if I had freely available drug, is there a patient where I might choose to use fedratinib rather than ruxolitinib? I’m not sure, but twice a day burden of taking drug is quite significant versus once a day. If there’s emerging data on less immune suppression, which there isn’t at the present time with fedratinib, that might influence me. But right now I think it’s ruxolitinib first line, with the exception of the very thrombocytopenic patients, where we’d be using pacritinib, either first or second line. And then fedratinib appears to be really emerging as a kind of second-line candidate drug.

I think momelotinib is also interesting, and it’s also kind of coming along. That drug also didn’t really cause as much thrombocytopenia as ruxolitinib, and I think many of us have got some quite impressive experience of anemia response or stabilization with that agent. And we mustn’t forget, the anemia is a big burden for these patients. Having a transfusion carries with it a risk — iron overload — and it chews up your time. That could be an important drug, probably second line for the anemic patient.

DR LOVE: John, oncologists are now facing many situations with multiple agents available within the same class. There are 6 checkpoint inhibitors approved. Four PARP inhibitors. The list goes on. Again, if fedratinib were available and you could access it in the first line, are there situations where you’d use that rather than ruxolitinib?

DR MASCARENHAS: Unfortunately, there’s not data to compare up-front ruxolitinib to fedratinib. But if you look at the JAKARTA-1 study and look at the impact on splenomegaly, particularly with the JAKARTA-2, the patient profile where I would consider using fedratinib up front in a rux-naïve patient would be perhaps a patient who presents with a massive spleen, in which splenomegaly is really the driving reason to treat them, that would be a patient I could consider fedratinib treatment before ruxolitinib.

I do think that the availability of multiple agents would help us on a case-by-case basis tailor the treatment and perhaps cycle the therapies in a more effective way. Because right now we just have 1 drug commercially to play with.

Ongoing investigation of erythroid maturation agents in patients with MF

DR LOVE: I’m curious — I want to talk a little bit about PV and ET just very briefly in terms of particularly JAK inhibitors. But I am kind of curious, we’ve been hearing more and more about erythroid maturation agents. There was some great data presented at ASH on luspatercept. It seems really exciting to me, anyhow, just hearing about it. Obviously MDS, thalassemia is kind of where it started. But it’s also being looked at in MPNs, specifically myelofibrosis.

Claire, any thoughts about that and how that might change not only the management in terms of anemia but maybe even affect the use of JAK inhibitors?

PROF HARRISON: Yes. I think that’s a great thought, actually. As I mentioned, I think anemia is a huge issue for these patients. Yes, 10% to 15% of patients have this severe thrombocytopenia, but 80% of patients who have myelofibrosis get anemia, and the vast majority of patients treated with a JAK inhibitor develop anemia.

We’ve been watching the data from the MD Anderson on the sister compound to luspatercept, sotatercept, showing both as sotatercept alone and, interestingly, and I think very cleverly, in combination with ruxolitinib — that there do appear to be anemia responses. And we are working, actually, as a community in looking at luspatercept in patients with myelofibrosis, both alone and in combination with ruxolitinib. And I think that data is looking pretty exciting. And it’s likely to be maturing soon.

I think as a concept it’s a really interesting one. Because beyond just anemia, but the idea of further blocking the effect of cytokines on erythropoiesis and trying to allow that to normalize is really interesting.

Perspective on the role of luspatercept and sotatercept in the management of MF

DR LOVE: Hmm. John, any thoughts about this development and how it might affect management of myelofibrosis?

DR MASCARENHAS: We know that erythropoiesis-stimulating agents, ESAs, have a modest effect in patients with myelofibrosis that have anemia, particularly the patients with a low endogenous erythropoietin level, same as myelodysplastic syndrome, but often the response is not always durable in a given individual.

The sotatercept/luspatercept story is really interesting. I’m not sure that we really understand the biology well enough to invoke the implied mechanism of action, which is an activin ligand trap that essentially negatively regulates maturation of red blood cells. So it’s a different way of trying to improve anemia in these patients in which I totally agree with Claire is an urgent unmet need. Our patients are often anemic. Many of them are transfusion dependent. So it not only affects quality of life but contributes to the disease process and to the burden of the disease.

A therapy like luspatercept that could be given even in combination with ruxolitinib to offset anemia would be a real welcome addition, particularly if it’s well tolerated, which my perception at this point is that these drugs will likely be well tolerated. And the one point that I’d like to point out is, not all anemia is the same. Anemia that’s related directly to the disease, we know from multiple studies, and it basically factors into each risk prognostic scoring system, is an adverse prognostic factor. However, therapy-related anemia probably is not quite the same in terms of its clinical impact in terms of prognosis. And there have been analyses from the COMFORT studies that would suggest that patients with ruxolitinib-induced anemia still fare as well in terms of survival as patients who don’t develop the anemia. Having said that, to keep my patients out of the infusion suite would be a tremendous benefit.

Case (Prof Harrison): A man in his late 50s with severe pruritus associated with polycythemia vera (PV) experiences a dramatic relief of symptoms with ruxolitinib

DR LOVE: Let’s finish out talking about PV and ET. And Claire, you have a 57-year-old patient with PV. Can you kind of talk about what happened to him?

PROF HARRISON: My patient is a 57-year-old man who actually initially presented to the dermatologist with severe pruritis. And he was initially diagnosed with urticaria and cycled through all of the standard therapies for that, and it really took 2 years for them to actually think about an underlying hematological diagnosis here. And, actually, it was 4 years before he was diagnosed formally with polycythemia vera.

At that point then, he was treated with venesections and aspirin. He was below a kind of high-risk threshold, and he hadn’t had a thrombotic event. But all of the time this poor man’s pruritis was getting worse and worse.

He lives in England. It’s a pretty wet country. He couldn’t go out in the rain. He couldn’t swim. He couldn’t wash. It got to the point where he couldn’t even shower. And in my clinic room, he couldn’t even sit still in the chair.

He had already been treated with hydroxyurea, because his platelet count had become to go up. And at the time I saw him, he had had some PUVA therapy, and we were thinking about interferon. But, actually, as I use more interferon, I’m increasingly aware that for some patients it can make their pruritis a lot worse. And indeed, that’s what happened with this gentleman.

And so happily, after about 6 months of optimizing PUVA, using loads of antihistamines, lots of cream, et cetera, with a patient who’s really uncomfortable, not well controlled in terms of his hematology, loads of venesections, all of which necessitating about a half a day off work, we were able to access ruxolitinib for him. And as we’ve seen in myelofibrosis, interestingly, the patient’s pruritis switched off within a couple of days. He’s able to take a bath, which is a pretty important basic human need. But importantly, also, he’s been venesection free. Kind of illustrating the PV patient with the big symptom burden, which is fairly specific for PV pruritis, and this need for venesection, which has a big impact on hospital facilities but also on the patient’s quality of life.

Efficacy of ruxolitinib in managing pruritus associated with PV

DR LOVE: Wow. Just thinking about that and just imagining what it was like for him to go through the relief of these symptoms and for you, as his physician, to see that happen. The pruritis sounds terrible, and the fact that it was palliated is amazing, really.

PROF HARRISON: He’s a great advertisement for using a therapy that’s very effective for pruritis, and he’s a very happy man. That’s definitely true.

DR LOVE: John, any comments about the case?

DR MASCARENHAS: No. It’s, unfortunately, something I’ve also seen. Pruritis, particularly aquagenic pruritis, is quite uniquely associated with polycythemia vera. And I’ve had a number of patients over the years who’ve struggled so badly with it that it’s affected their social interactions, their family interactions, their employment. Some of these patients literally can’t bathe, and it’s a terrible situation. And the itching is so intense and so upsetting that it has a psychological effect on them. The exposure to ruxolitinib, for example, has been tremendous in dramatically changing their lives for the better.

DR LOVE: I never thought about the pathophysiology of pruritis. Is it a cytokine thing? Do we know what cytokine it is?

DR MASCARENHAS: I don’t know that we really understand the pathophysiology of pruritis in this setting. Presumably it’s due to hyper-reactivity of MAST cells, but I don’t know if Claire has better insight into it.

There have been a number of attempts to try to understand it, but to my knowledge there really isn’t a great understanding of what is truly mediating this effect. But what we can see is that, for example, ruxolitinib clearly impacts that in a positive way. I would have to believe that in some form or fashion it’s cytokine driven.

DR LOVE: Claire?

PROF HARRISON: Oh, I’ve nothing further to add, really, actually. It’s one of the areas where we haven’t really taken the science forward, but we clearly have a very effective therapy.

And I think it’s interesting, because it’s something that not all the patients equate with their disease. I’m always quite careful to ask the patient, “Do you get itching?” And turning to the beginning of our conversation about diagnostics, for me, a patient with an erythrocytosis, I ask them about aquagenic pruritis. Because I think it’s a clue. And it’s a symptom that clearly — in this case it was very bothersome, but for some other patients, it can be a wearisome symptom to be itching for up to an hour after you’ve had a shower, and it means that they avoid things like swimming, et cetera.

So simple management usually is effective, antihistamines, et cetera. But there are these cases where it is very severe and we don’t really understand pathophysiology. No.

Case (Dr Mascarenhas): A woman in her late 60s with essential thrombocythemia (ET) refuses cytoreductive therapy

DR LOVE: Let’s finish out, John, with your 67-year-old woman.

DR MASCARENHAS: This is an older woman who had had essential thrombocythemia, which tends to be more of an indolent MPN compared to the other ones we’ve discussed. A high platelet count of 850,000. And like a subset of my patients that I see, she’s quite intelligent, quite well read. Tries to keep up on the literature and has really been against the idea of taking any form of cytoreductive therapy, and particularly hydroxyurea, and has a concern that still prevails in the patient, and even in the physician community, of the potential leukemogenic effect of hydroxyurea.

I, for one, am less concerned about that in most of the patients I’m treating, as the data doesn’t truly support that risk. Although I completely understand the concern and the fear and the reservation amongst patients in starting a chemotherapeutic, particularly one that they would be on for years on end.

She had always been against the idea of cytoreduction. She is a quite fit and very young-looking person who has no cardiovascular risk factors. One of the discussions I had with her and other patients like this is, what are the goals of therapy in ET and PV? What are we trying to accomplish with the therapies we use? And does it make sense for that individual patient?

And in her case, although her age was advanced, over 60, which is considered a risk factor for thrombosis, and she had not had a thrombotic event, nor did she have the classic modifiable cardiovascular risk factors, we had decided to not cytoreduce her with hydroxyurea, although that really goes against the NCCN guidelines that we use here in the US that would suggest that high-risk patients with ET should be reduced. And we typically try to normalize the platelet count.

And the problem I have with this, and I’d be interested to hear what Claire thinks about this, is that we have these criteria for response as it relates to normalization of  the blood counts, spleen and symptoms, but we really don’t have data that supports that that actually is a great surrogate endpoint and clinical trial endpoint, for example, for risk reduction in thrombosis and improvement in overall survival or change in disease course.

There really remains a lot to be done, particularly in the ET and PV side, to better understand the pathophysiology, what are clinically relevant endpoints and how best to address 2 aspects of patient care — one is the thrombosis, which is usually what consumes our thoughts and our maneuvers early on in the disease course, and then eventually the concern for progression and disease and evolution of disease.

Effect of age and mutational status on the decision to treat ET

DR LOVE: Claire, we can close out with any comments you have about this case.

PROF HARRISON: Yes, I think that’s a great case. And I’ve a lot of patients who really resent the idea that when they reach their 60^th birthday we have to have this conversation about starting a therapy. And I think, though, there is some change coming probably in this setting, as we’ve been looking recently in the UK — we’ve uplifted, actually, our age for high risk in PV patients to 65, because I really think a 60-year-old nowadays is probably, on the whole, fitter than a 60-year-old was when the data that we’re basing our current practice on was collected.

I think in ET there is another important point, which is the role of the driver mutation, because we clearly see a different profile for the patients with the calreticulin mutation. These patients have a lower risk of thrombosis, maybe a high risk of bleeding. There’s a question about the use of aspirin, especially in lower-risk patients.

And for those patients, I’ve got this kind of creep in my clinical practice now where I’d be saying to that patient if they’re fit — you obviously have to look at your patient carefully as John described — I probably would be a bit happier leaving that patient until they’re 65-plus. On the converse, a patient with a JAK2 mutation and ET, a slightly higher white count, a bit of hypertension, a smoker I can’t persuade to stop smoking, then that patient is going to be heading towards a cytotoxic.

I think your point, John, about what is our target here is a really very relevant one. In the historical studies in the ET setting, we do see a correlation with lowering the platelet count with hydroxyurea and reduction in risk of thrombosis, but even though we lowered the platelet count to the same extent with anagrelide, perhaps that may well be a less effective drug in some patients.

We have to do quite a bit more work, I think, in these patients to avoid exposing them to nonessential drugs, but equally not be afraid of having the conversation and encouraging them to take a drug to reduce the risk of thrombosis in the first instance.