Consensus on Clinical Trial End Points Is Essential for Improved Drug Development Across MPNs

Efforts to redefine traditional clinical trial end points such as spleen volume reduction (SVR) and symptom scores in myeloproliferative neoplasms (MPNs) will likely remain limited until constrained until truly biologically targeted and disease-modifying therapies capable of producing substantive molecular responses in addition to symptom control are developed, according to Andrew Kuykendall, MD.

Right now, there’s a desire to create new end points, [such as] molecular response, but there are challenges with each of those,” Kuykendall said in an interview with OncLive® regarding a State of the Science Summit™ on myeloproliferative neoplasms, which he chaired. “What’s really going to drive the needle [forward] is the development of more comprehensive disease-modifying and more mutation-specific [agents]. There are agents in the pipeline designed with this intent, and if they prove successful, that could allow us to shift toward end points that are more reflective of true disease biology.”

In the interview, Kuykendall expanded on difficulties measuring traditional clinical trial end points, and challenges with the identification of novel end points, for myelofibrosis and polycythemia vera; the lack of disease-modifying therapies in MPN; and the potential of non-covalent BTK inhibitors to improve toxicity profiles.

Kuykendall also touched on emerging treatments, such as BCL2 inhibitors and CD19 CAR T-cell therapies, on treatment sequencing in chronic lymphocytic leukemia in another article.

Kuykendall is an assistant member of the Department of Malignant Hematology at Moffitt Cancer Center in Tampa, Florida.

OncLive: What have been some of the challenges with developing novel clinical trial end points in MPN?

Kuykendall: The biggest issue is that we do not have any home-run therapies that eliminate the disease. Myelofibrosis, polycythemia vera, and essential thrombocythemia are chronic malignancies that, unfortunately, affect the very cells we rely on most: red blood cells, neutrophils, and others. [As a result,] we cannot simply eradicate those malignant clones without compromising essential cell populations and, ultimately, the patient’s viability.

We also don’t have truly disease-modifying therapies that directly target the mutated clone. We do have effective therapies; when we first leverage those therapies, they [more so help] to improve patients’ quality of life and alleviate symptoms like splenomegaly. When we first began using these therapies, we developed endpoints that reflected their observable effects. For better or worse, we have [remained reliant] on those same end points to a large extent. We would love to identify new end points, but it's hard to do that when we don't necessarily have existing therapies or therapies in development that have shown they can do something new or beyond that.

What are some of the limitations of current quality of life or symptom-related end points?

For all the criticism these end points receive, they are still reasonable, as they reflect clinically relevant aspects of the disease. Patients with myelofibrosis often present with inflammatory symptoms and splenomegaly, and these features tend to respond when therapies that impact disease biology are administered. Logically, a therapy with true anticlonal activity or disease-modifying potential would be expected to reduce spleen size and improve symptoms, even if the timelines for these effects differ.

The challenge arises in how these end points are applied in clinical trial design, potentially setting too high a bar for new therapies. For example, in the [phase 3] MANIFEST-2 trial [NCT04603495], patients with higher-risk myelofibrosis were randomized to receive ruxolitinib monotherapy or ruxolitinib [Jakafi] plus pelabresib [CPI-0610]​​. The combination significantly improved SVR compared with ruxolitinib alone. There was also a nominal improvement in symptom scores, as measured by a 50% [or greater] reduction in total symptom burden. However, the trial did not meet the statistical threshold for superiority on this symptom end point.

This outcome reflects the complexity of demonstrating additive benefit when comparing a combination regimen against an already effective monotherapy, especially when increased toxicity from dual therapy may counteract symptomatic gains. In this case, the high expectations likely influenced both the regulatory outlook and the pace of development for pelabresib.

It remains critical to better understand how these end points are weighted and validated. Demonstrating that greater SVR or longer duration of spleen response correlates with improved survival or long-term outcomes would strengthen their utility in therapeutic development. However, many trials to date have limited follow-up—often around 24 weeks—and lack robust data on the durability or depth of response. It is therefore incumbent upon us as both researchers and industry [sponsors] to generate these data.

How has the chronic nature of these conditions contributed to difficulties leveraging overall survival (OS) or progression-free survival (PFS) end points?

OS is an aspirational end point. We know that a lot of these diseases dictate survival, and what better way to show value than someone living longer than they would otherwise live if they didn’t have therapy. The problem is, if you start to deal with chronic diseases, there are competing risks; other things can impact survival beyond just the disease. In myelofibrosis, especially higher-risk myelofibrosis, survival is estimated between 3 and 5 years. [However], a lot of our patients are older and have other comorbidities, cardiovascular issues, so there are competing risks.

If [a patient does] have a disease that has a longer survival, we need to watch these patients for a long time. In this day and age, clinical trials and monitoring patients for years are not something that people necessarily want to invest in. If we’re going to do that, we need to enroll a ton of patients. That’s hard to do if you have a rare disease like myelofibrosis. We're dealing with a rare disease state with a modest OS [benefit]. If we're going to use OS [as an end point], we need to enrich for high-risk patients or need to have an effective intervention that's going to meaningfully change OS.

Right now, it's unclear whether we have all those factors to use that end point. [Using] PFS [comes with] the same risk, because our current definition of progression just relies on myelofibrosis turning into acute leukemia/blast phase, or the spleen growing.

That’s a bit challenging, because even though our currently approved JAK inhibitors don’t necessarily make the disease go away, they invariably reduce spleen size, so the spleen is not often growing or progressing. PFS is quite challenging when we have other agents that control that. Amending our definition of progression might be helpful.

How can the MPN field come together to develop a consensus on the development and optimal use of these clinical trial end points?

We can’t just change where the end zone is and then call it a touchdown. At the end of the day, if we develop better drugs and we see promising activity beyond what we’ve seen with our existing agents in early-phase trials, there will be a lot of motivation to redraw what these end points are. Regulatory bodies play a role there, because we have to ask [them whether something is a] reasonable end point as we’re designing a trial.

If we’re able to show why [molecular response] should be [a viable end point], and make a reasonable argument, then [it could] be accepted. If [we see] a molecular response or the fibrosis goes away, but the person’s spleen is still big, they feel terrible, and they’re getting transfusions, that’s still not a win. We have to listen to the biology and design intelligent agents and then see what they do in early-phase trials. If we start to see something different, maybe [we can] adjust at that point in time.

We also have to understand that myelofibrosis is probably not one disease. It’s probably a lot of different diseases that we’ve grouped into a single basket. As we develop trials, we need to consider how to enrich populations with more specific, more homogeneous subtypes of disease that are biologically similar, and say: “Okay, let’s study this drug in this group of patients, because that’s where it makes sense.”