Harry Erba, MD, PhD: The other agent that can be used in this disease is ruxolitinib based on the results of the RESPONSE and RESPONSE-2 trials. These were patients with hydroxyurea intolerance or hydroxyurea resistance. Basically, by ELN [European LeukemiaNet] guidelines that were defined. Both studies—1 was in the study of patients with splenomegaly, RESPONSE; RESPONSE-2 was without splenomegaly—showed the same thing: a statistically significant decrease in phlebotomy requirements, higher rates of complete hematologic remission.

I’ve summarized those studies, but something I just didn’t mention was improvement in symptoms couldn’t be shown because it wasn’t a blinded study, so it didn’t go into the FDA label. But I’m going to bring this back to you, Robyn. You mentioned earlier how patients who receive cytoreductive therapy with hydroxyurea may continue to have quite a burden of symptoms, despite ruxolitinib not being approved in PV [polycythemia vera] based on symptom control. Is this something that you go to in PV, everything else being equal?

Robyn Scherber, MD: I do. I have to admit, when I see patients—especially if they’ve interferon or hydroxyurea and haven’t done well with it because of their symptom burden—I do like to go to ruxolitinib. I certainly have seen symptom responses with ruxolitinib that you don’t see with interferon or with Hydrea. As much as I think there are extremely promising data with interferon, I think 1 of the few caveats is the idea that a large portion of patients won’t be able to tolerate the drug because of the flu-like affects.

Especially for the highly symptomatic patients who have already tried a first-line therapy, ruxolitinib is a second line for symptom burden improvement and quality-of-life improvement. These are very reasonable things.

Mary Frances McMullin, MD: We have, in the UK, had several investigator studies where they’re all under the umbrella of magic with ET [essential thrombocythemia] and PV. We’ve had an experience of using ruxolitinib versus best-available therapy in these groups of patients for a prolonged period of time. Certainly in PV, you see a group of patients getting a very major symptom response. In fact, from the clinical experience you can almost define the patient who is likely to do well on ruxolitinib. As a second-line agent we would know from time to time. We have difficulties with accessing drugs. But requesting it for somebody who has a particular problem—in my book, that is really bad itch and sweats—is life transforming for some of these patients.

Harry Erba, MD, PhD: It’s often interesting to me the disconnect between trial eligibility and agency approval of drugs. Patients got into the trial because they had ELN-defined resistance or intolerance to hydroxyurea, but it wasn’t just based on symptoms from their disease. Splenomegaly, yes, but not the constitutional symptoms. Because it wasn’t a blinded trial, as we said, the FDA, at least in the United States, would not allow that to be part of the label that improved symptoms.

I think we all clearly have the same experience in PV that we do in MF [myelofibrosis]. That symptom control can be quite dramatic with ruxolitinib and should be considered for those patients.

I’m going to come back to you, Mary Frances, and ask, what do you think would be the impact of failing to control the disease on thrombotic events in overall survival? This is a disease that practicing oncologists see in their clinics all the time. The hematocrit is running a little above 45% percent. Are we doing patients a disservice by not being aggressive in the management of at least the thrombotic risk? What do you think?

Mary Frances McMullin, MD: If we’re talking about high-risk patients, then we can have the discussion of what constitutes high risk patient. In that group, there’s clearly evidence that controlling the hematocrit brings about better outcomes. In particular, the CYTO-PV study shows that a hematocrit below 0.4 or 0.5 L/L was desirable. That was a good randomized study.

We used this number for many years without that evidence, but we now have the evidence of what level we should be aiming for. I think you do have to argue that for the patient you want to have hematocrit below 45%.

Harry Erba, MD, PhD: How often do you look, though? Monthly, every 3 months, or does it depend on where they are in their treatment?

Mary Frances McMullin, MD: Where they are and how much the doctors are being paid for the visits.

Harry Erba, MD, PhD: All important.

Mary Francis McMullin, MD: That’s certainly an issue because your dose will vary between visits. We would check them frequently at the beginning. But once the hematocrit is controlled, we don’t see them—maybe every 12 weeks, depending on the practice and what they need. Yes, there’s probably times that they’re up. But then if they’re coming back at the 12 weeks way above 45%, you probably should be seeing them more frequently.

Harry Erba, MD, PhD: When you first make the diagnosis, what is the frequency of your visits? How often do you do phlebotomies? How aggressive are you in getting that hematocrit down?

Mary Frances McMullin, MD: For somebody who has a very high hematocrit level, I would see them weekly. If they’re an inpatient and they present with something—you may end up with, say, somebody presenting with a stroke and phlebotomizing them every day until you get the hematocrit down. For an outpatient, I would say weekly until their hematocrit is dying.

Robyn Scherber, MD: I agree. I’m usually pretty quick on the uptake. I had a recent experience, just to be totally honest. I had a patient who had a little bit of a wait time to see me, and just a few weeks before she could have the visit, she stroked because of a high hematocrit. More than ever, even the day that I’m seeing them for an initial visit, they are going to get that phlebotomy the same day, and depending on the risk factor potentially started…

Mary Frances McMullin, MD: And one of those patients with a high hematocrit do very obviously...

Moshe Talpaz, MD: Do we have a subset of patients with an unsolved problem? At this point, arterial clots happen regardless of hematocrit. We have no tools to address it, and it’s very frustrating if we lose a young patient because of this problem with a mesenteric artery thrombosis or things of that nature, and regretfully it has happened.

Robyn Scherber, MD: Is that not a slightly different group of patients? We seem to have this group of women presenting very often with only a very low JAK2 level in which there’s something different genetically presumably.

Moshe Talpaz, MD: We agree, but we don’t have the tools to define them.

Transcript Edited for Clarity