Shifting Endpoints for JAK Inhibitors in the Treatment of Myelofibrosis

Andrew Kuykendall

Can you walk me through the original U.S. Food and Drug Administration (FDA) approvals of the JAK inhibitors ruxolitinib and fedratinib? What data were the decisions based on?

Natasha L. Johnson

Both the COMFORT¹ and JAKARTA² studies showed benefits in spleen volume reduction, measured as a greater than 35% and a greater than 50% total symptom score reduction for ruxolitinib and fedratinib, respectively, in patients treated with those therapies.

Andrew Kuykendall

These are interesting endpoints, right? When you’re comparing these to other endpoints for drug approvals, they’re certainly quite unique. We don’t see true objective, clinical, complete, or partial responses with JAK inhibitors. These are largely anti-inflammatory agents, and early on, we knew that they were doing something for the disease, that they were rapidly and massively improving the size of people’s splenomegaly while also improving their symptoms. Ultimately, this was something that the FDA was able to agree on, to suggest that these were meaningful endpoints for patients. I think symptom response certainly goes along with quality of life; obviously if patients are feeling better, that’s a benefit. Spleen response, however, is a little bit more nuanced. Obviously, it is associated with symptom benefit, but we’ve also been able to correlate responses in spleen volume with overall survival. When you look at even the early studies of ruxolitinib in phase I or II trials, as well as phase III trials, we do see that achieving spleen response tends to correlate with improved overall survival—patients who had a spleen response have better overall survival than those patients who lack a spleen response. However, it does slightly cloud development of further agents, because now our endpoints in myelofibrosis are spleen response and symptoms. That makes me think: is that really all we care about when we’re treating patients with myelofibrosis? I think that’s great for drug approvals, but in the clinic, what are we thinking about?

Natasha L. Johnson

Yes, not at all. And from what I’m hearing and learning, is that with myeloproliferative neoplasm trials in general, there’s a real push to go after disease modification and improvement of overall survival vs spleen and symptom reduction.

Andrew Kuykendall

We always want patients to live better or live longer. Certainly, I think we’ve done a pretty good job in having them live better—living longer is a bit debatable, right? There is some overall survival benefit from ruxolitinib, but it’s quite modest and it doesn’t seem to be associated with true modification of the underlying disease. The disease is still there. We’re just improving patient’s functional status and therein, they’re able to have a longer lifespan because of that.

I don’t want to minimize the impact that ruxolitinib and fedratinib have had on the treatment of this disease, but I think there’s certainly room to do better as well. To that end, one of the things that I wanted to discuss with you is the unique dosing of ruxolitinib. This is different than a lot of other drugs we may give. How do we determine the dose? I know that certainly we have available doses that start at 5 mg twice a day, all the way up to 25 mg twice a day. But how do you determine which dose you’re going to give a patient?

Natasha L. Johnson

Dosing of ruxolitinib is determined by a patient’s platelet counts, but it’s important to note that after patients start therapy, they're more than likely going to have a drop in platelets and hemoglobin as well.

Andrew Kuykendall

Absolutely. If you look back at dosing of ruxlotinib in the COMFORT study, they enrolled patients who only had to have at least 100,000 platelets.¹ If their platelets were between 100,000 and 200,000, they started at 15 mg twice a day; if their platelets were over 200,000, they started at 20 mg twice a day, relatively high doses. In practice, that's not necessarily how we dose this. In your practice, what's the importance of getting to a high dose of ruxolitinib, and what have we seen with these higher doses?

Natasha L. Johnson

Studies have demonstrated that optimal dosing of ruxolitinib provides the greatest benefit for spleen volume reduction and symptom improvement. So we know that patients who are receiving inadequate or suboptimal dosing don't get the greatest response. When starting at the recommended and optimal dose, we unfortunately often see cytopenias develop, and it often requires dose modification.

Andrew Kuykendall

Right—if you look at the breakdown of responses based on dose in the COMFORT trials, you see that patients who received 5 mg twice a day as their last dose did not get too many responses at all in terms of spleen or symptoms.¹ Once you get to 10 mg twice a day, you start to see spleen and symptom responses. Maybe for symptoms, you do just as good at 10 mg, which is probably good enough if you’re just targeting symptoms, as you might do with 25 mg twice a day—it's pretty stable there.

But what we do see is a clear dose-response relationship with splenomegaly. If you're thinking that achieving these spleen responses is the one surrogate we have for overall survival, it goes to show that we probably should be trying to get to as high a dose of ruxolitinib as possible to achieve this endpoint, which we know has some clinical significance for survival. The problem, as you mentioned, is cytopenias. We run into problems where we're not able to fully dose ruxolitinib and are not able to achieve maximum spleen response or, by extension, survival benefit that we saw on clinical trials.

That brings up a problem: is suboptimal dosing of ruxolitinib, done slightly different than dosing in the confirmatory clinical trials, going to be as good as fully dosing another agent that might not cause the same degree of cytopenias? We do have some data from the REALISE study, which was largely conducted in Europe, that suggested starting ruxolitinib at 10 mg twice a day and then escalating it at prespecified time points, as long as the platelets would allow, to get to a maximum dose.³ What was seen there is that you can escalate these patients that start off anemic and actually achieve clinically meaningful responses. I believe they were looking at spleen length reduction on physical exam, but we saw that in more than 50% of patients on the trial. I think still the take-home point remains the same, which is that it's important to get these patients to as high of a dose as possible. You can start high, as they did in the trials, or you can start low, but if you do, you still have to get there and think about how to do that. Is that something you’re discussing in clinic with patients when you're seeing them, or are you constantly thinking about the dose of ruxolitinib that they're on?

Natasha L. Johnson

When I see the patient, it really depends on how they feel, how their spleen is, and how negatively impactful their symptoms are. So like you said, if they're doing well on a 10-mg or 15-mg twice daily dose, then we’ll leave them there and let them continue to feel good. But when they're coming in complaining of more and more symptoms, or more and more feelings of being uncomfortable because of their spleen, then we try to get them to the maximum tolerated dose.