Thanks, John and Claire. Thanks, Charlie and the whole VJHemOnc team. Really excited to be opening this particular workshop. I mean, again, we’ve not really had anything quite of its kind. So my task is really more of a setting-the-stage type of talk, but I think we’re all looking forward to the discussions that will ensue.
So this is kind of a review of the endpoints of the approved drugs. So as you recall, in myelofibrosis, the four JAK inhibitors were primarily built around spleen, right? So I just listed the primary endpoints in the column to the right: ruxolitinib, fedratinib, pacritinib. Now, pacritinib had co-primary endpoints, remember that; that was spleen and symptoms. And momelotinib, at least in the SIMPLIFY trials, was again built around spleen, although MOMENTUM was a little different.
So MOMENTUM, obviously building on the strengths of this drug, it’s fairly obvious what led to this design. Looking at the lessons learned from SIMPLIFY-2, this is a good drug for symptoms. Perhaps the ability of this drug in the second-line setting to improve symptoms, which was seen in SIMPLIFY-2, led to symptoms being the primary endpoint of the MOMENTUM trial. And then we’ve always known that this is a drug whose particular strength is anemia, right? So danazol was picked as an active control, as an anemia comparator. And symptoms here were the primary endpoint: the first Phase III with symptoms as the primary endpoint, not spleen. And all endpoints were met. With danazol being the active control, the comparison for transfusion independence was a non-inferiority one, but everything else was superiority, and they were all met.
Now, obviously, we are in the combinations era. Pelabresib was among the first. So in MANIFEST-2, again, the primary endpoint was spleen, and that was met. Obviously, the symptoms have been a struggle, and that’s emerged as a common theme. But the primary endpoint was spleen, and that was clearly met. There were some combinations that were discontinued. So navitoclax was discontinued, and parsaclisib was discontinued, for slightly different reasons. Navitoclax easily met the spleen primary endpoint, so again, the primary was spleen there, but not symptoms. And then parsaclisib really halted early because it didn’t look like that was going to be a positive trial.
Now, SENTRY, the most recent, we published this earlier this year: co-primary endpoints this time. So spleen, again, easily met, 50 versus 28, and also the absolute change in TSS, or the modified absolute change in TSS, at week 24. So co-primary this time. So remember, pacritinib: spleen and symptoms co-primary. And here we have another one: spleen and symptoms co-primary. Now, in SENTRY, we did see something we had not expected. At an early time point, we saw an overall survival difference, with a significant p-value and an impressive hazard ratio. An early time point, of course.
Now, navtemadlin, in the POIESIS trial, this is a bit of a unique design, right? There’s nothing quite like it in the field. And the attempt here was to try and isolate the effect of prior ruxolitinib and make it sort of homogeneous across patients, so that you don’t have greatly varying durations of prior ruxolitinib. And so here, there’s sort of a control for that, in the sense that everybody starts ruxolitinib, and then after 18 to 24 weeks, if they are suboptimal, that is when they randomize to navtemadlin or placebo. So it’s pretty much a unique design, and although it’s a frontline trial, so to speak, it’s really a suboptimal setting that is being looked at and tested here. And importantly, they have to be suboptimal for both spleen and symptoms for that randomization to occur.
Now, imetelstat, again, building on its strengths: in the IMbark study, the Phase II, there was what appeared to be a survival signal. Again, compared to historical controls, you’d expect a 13- to 14-month median in the post-JAK inhibitor failure setting, and they showed around 29 months in the Phase II, and that led to the Phase III, which is here. That’s ongoing; we don’t have results. It’s completely accrued, which is imetelstat versus BAT in the second-line setting, with an overall survival endpoint. Again, it’s often said that this is the only trial that has an overall survival primary endpoint, which is true, but obviously it’s a very different trial from everything else I showed you. This is a JAK inhibitor failure trial, and that’s probably not an argument that can be made in the frontline setting.
Now, there have been efforts, obviously, to revamp endpoints, since we’ve had several examples now of not being able to overcome the symptom barrier when we are comparing combinations to ruxolitinib. So Claire has called for, as you see on the upper left, all stakeholders, regulators and pharma, to essentially pool the data and really look at what matters, right? Clearly, spleen and symptom endpoints were designed with JAK inhibitors in mind, but now that we have non-JAK inhibitors, perhaps it’s time to look beyond those. And then some of us were also involved in this effort with the ELN to try and define a composite endpoint, to sort of rank the outcomes by their desirability, or lack thereof, and just sort of see: if we had a wish list, what would we want our patients to not have? And so it would be freedom from certain things, and you would rank those according to the setting; the treatment setting would be different for frontline versus failure versus suboptimal.
Now, in PV, ruxolitinib was approved based on a composite of hematocrit and spleen, and the RESPONSE trials were clearly positive. RESPONSE-2 did not require spleen, as you recall, but RESPONSE did. And so essentially, again, spleen, symptoms and hematocrit control in that setting.
Now, Claire led the landmark MAJIC study, which taught us a lot and really took PV in a different direction, right? So event-free survival emerged as something important in PV, disease modification for the first time, moving away from just reducing clots or just reducing counts. And so here, showing that ruxolitinib improved event-free survival, which was defined as being alive without a clot, MF or AML. And also that molecular response mattered: molecular response correlated with event-free survival. Again, a trial taking the field, I think, in a different direction, which was continued by ropeg. These are the PROUD-PV and CONTINUATION-PV studies, which led to ropeg’s approval in Europe; in the US it was PEGINVERA, so that’s different. But here, again, they were able to show… this trial again was designed kind of like RESPONSE, around blood counts. CHR was the main thing, also spleen, and looking at VAF separately. But this trial also showed event-free survival improvements. So with two drugs now, disease-modifying ability, we believe, has been shown in PV. And importantly, for both, there was a correlation between molecular response and event-free survival.
Now, obviously, the latest drug to be approved for PV is very different. So rusfertide is a hepcidin mimetic, and obviously, it requires different endpoints, right? So, as you recall from Andrew’s presentation at ASCO last year, this is the VERIFY study, where you’re looking at freedom from phlebotomy eligibility. Again, it’s a totally different type of drug, not felt to be a disease-modifying drug, but a useful drug nevertheless in terms of reducing phlebotomies and making patients feel better. So clearly, freedom from phlebotomy is emerging as the primary endpoint for this type of drug.
And finally, I’ll end with SURPASS-ET, which led to the most recent MPN approval, on August 31st. So this is ropeg for ET. And here again, this is important because other drugs are being looked at in Phase III, like the mutant CALR antibody from Incyte, rolling out now in Phase III. So again, I’m not sure what their exact endpoints are, but in SURPASS-ET, this was the endpoint that was used: essentially a durable hematologic response rate and lack of progression as well. So controlled counts, without any vascular events or spleen progression, and that was really the composite endpoint there. So I think there is an appetite for composite endpoints, and we’re starting to see that in PV and ET already. So again, I look forward to the discussion, and thank you for this opportunity.
AI-generated transcript. Please refer to the video as the authoritative source.