Transcript:Harry P. Erba, MD, PhD: Rami, I want to come back to something. We were talking about the diagnostic criteria. I don’t remember Mary Frances talking about next-generation panels looking at all these markers. Now, based on some of these new prognostic models, do you think we should be getting a next-gen panel looking for ASXL1 and EZH2 mutations, and so on?

Rami Komrokji, MD: They are not part of the diagnostic criteria. Obviously, a clonal marker at the JAK2 mutation, calreticulin, or MPL are now part of the diagnostic criteria. I do think there is a value in checking them—definitely a prognostic value and maybe, down the road, a therapeutic value. We know, in myelofibrosis, the presence of the ASXL1 or SRSF2 mutation is associated with worse outcome.

There are other mutations that are labeled as higher-risk mutations. Some mutations may predict the response to treatment. Even with ruxolitinib, there has been a look at patients who have higher-risk mutations versus not. I think it’s valuable to check for them, but they are definitely not in that part of the diagnostic criteria yet.

Harry P. Erba, MD, PhD: Mary Frances, I’m going to have Ruben go through the data of ruxolitinib for myelofibrosis [MF] in the COMFORT-I or COMFORT-II. Before he gets to that, I want to put you on the spot. It seems to me, from those very positive studies, it calls into question whether there is room for any initial therapies other than ruxolitinib in myelofibrosis. Is that really your go-to, based on the data?

Mary Frances McMullin, MD, FRCP, FRCPath: At the moment, it has to be the go-to, based on the data. Apart from anything else, it’s the license therapy. I worked in a part of the world where it took quite a long time, because we were behind in getting access to it. It was interesting watching that, and when suddenly a therapy became available, you wanted to give it to everybody. It almost seems as if there were more people with myelofibrosis, but really, that was because there’s something you can do. Therefore, everybody picks up the patients and refers them to you. I think, in the absence of something better, where you could do an up-front clinical trial—and that’s probably unlikely to happen— it has to be the starting point, at the moment.

Harry P. Erba, MD, PhD: OK. Ruben, why don’t you go through the clinical data that have convinced us of this.

Ruben A. Mesa, MD, FACP: Sure. The COMFORT studies were the first large, randomized phase III trials done on myelofibrosis of any agent. Now, over 8 years, both these studies have been highly impactful. They were ruxolitinib versus placebo. The COMFORT-I study was a North American and Australian study. COMFORT-II was done in Europe versus best alternative therapy. Combined, they clearly showed several things.

One, ruxolitinib was clearly superior to either placebo or best alternative therapy. It didn’t really matter what else you chose to use for control of splenomegaly in an under-6-month time frame. It had relatively rapid response and significant improvement in the disease-associated symptoms. Again, this was both in a placebo-controlled way, as well as with best alternative therapy. What we’ve learned over time is that the consequence of improving splenomegaly and symptoms was very impactful, both in terms of that short-term benefit that you feel—the patient is clearly feeling better, and they’re under better control—and in the longer term.

We’ve had analysis that showed that with the combined data set, as well as individually, when 1 accounts for the issue of crossover, there is a survival advantage. That was not prebuilt into the study, so there’s always been controversy and discussion around this. I’d say, unequivocally, with the clinical experience that I’ve had with these patients, there is 0 doubt that there is a clear improvement in survival as a group. It’s tougher to say on an individual-patient level. That has become very clear.

How can we say that? Well, 1) we look at the data. Without question, even accounting for the crossover, there’s clearly an improvement in survival. Now, there is not a plateau. It is not a cure, but it clearly improves survival. We know there are individuals that remain on ruxolitinib who even started on the phase I studies that we started at Mayo Clinic and The University of Texas MD Anderson Cancer Center in 2007. There are patients who are now on the drug for 12 years who had a life expectancy of under 3 years when the trial started. I’ve had patients who have been on it since the drug became commercially available, who long ago should have passed away from the disease. The sheer rate of patients who are passing away from myelofibrosis has decreased, and it’s clearly from ruxolitinib.

Why is that the case? It’s tough for us to know for certain, but there are certainly several theories. One, there is clearly less debilitation. It must be even more than that, given the length, but that certainly is a factor. Two, there is potentially less progression toward acute leukemia. Again, that is very difficult to prove without a bulletproof control arm, but the rates certainly seem to be lower than they had been in the pre-ruxolitinib world.

Is there a benefit by a decrease in the inflammatory milieu in the bone marrow? In other words, it’s improving cytokines, but maybe that makes a big difference. If you have a toxic bone marrow microenvironment, maybe that leads to additional genetic mutations and a greater likelihood of disease progression. I think it’s been a big impact. It clearly still leaves us room to grow, because as 1 looks at those curves, there’s not a plateau, so there’s not a cure that is involved. However, it’s been highly impactful, and the data have really run true in that every parallel, open-label study that’s been done afterward has found that those data are accurate.

Harry P. Erba, MD, PhD: Even if you’re not convinced by the symptom control and controlled splenomegaly, which were the primary endpoints of the study, what I find fascinating is that survival is easy to measure, right? You’re either dead or alive. What was interesting about the survival curves in both COMFORT-I and COMFORT-II is they actually didn’t start to diverge until after the median from when patients actually crossed over. For me, that suggests that it supports what Mary Frances said. This really should be in a symptomatic patient—an intermediate or high-risk patient—and you need to start with your best agent to see that benefit. Would you agree with that?

Ruben A. Mesa, MD, FACP: I think so. There is no scientific rationale for using it as a salvage therapy. What we have learned is that, in a salvage therapy—type approach in MF—whether it’s waiting for people to get really sick to start ruxolitinib—they’re likely to respond, but they’re going to respond less. Likewise, with transplant, if you wait for patients to get really sick, their outcomes with transplant are worse. Myelofibrosis is not a disease to wait to initiate therapy. Starting therapy earlier—whatever that is—is highly impactful.

Transcript Edited for Clarity