Supplements and Featured Publications
Aaron Gerds, MD, outlines how cytopenias affect patients with myelofibrosis and how these can help drive JAK inhibitor selection.
With JAK inhibitors engrained as the primary treatment modality for patients with myelofibrosis, factors such as anemia status and platelet count can serve as treatment selection drivers among approved agents in this class, according to Aaron Gerds, MD.
“Some of these nuances [with JAK inhibitor selection] are pretty difficult to sort out,” Gerds said. “Most practicing doctors are quite savvy, and they’re up to speed on the commercially available agents for myelofibrosis, but some of those nuances of selection, sequencing, and dosing strategies often come up in discussion.”
In an interview with OncLive®, Gerds discussed different presentations of myelofibrosis and how factors associated with them could affect management strategies. He also talked about JAK inhibitor selection for patients with myelofibrosis and the role JAK inhibitor–based combinations could ultimately play in the treatment paradigm.
Gerds is a physician in the Department of Hematology and Medical Oncology at Cleveland Clinic, as well as an assistant professor in the Department of Medicine of the School of Medicine, deputy associate director for Clinical Research, and a member of the Developmental Therapeutics Program at Case Comprehensive Cancer Center at Case Western Reserve University in Ohio.
Gerds: There are no formal classification, criteria, or guidelines to separate out more proliferative myelofibrosis vs cytopenic myelofibrosis, but there are some intuitive things that we do recognize. First and foremost, [for] cytopenic myelofibrosis, these are patients typically with lower blood counts—anemia and thrombocytopenia, in particular. There are also some things that separate out [these classifications] beyond the simple blood counts. We often see that patients with thrombocytopenic myelofibrosis and anemia [with] myelofibrosis tend to have lower variant allele fractions [VAFs] of their driver mutation vs those who have more proliferative disease. In cytopenic myelofibrosis, we will see VAFs even in the teens or lower; in more proliferative [disease, these VAFs] will be much higher, maybe even beyond 50%.
Cytopenic myelofibrosis tends to be more primary myelofibrosis, and proliferative myelofibrosis tends to be more post–polycythemia vera and post–essential thrombocythemia myelofibrosis. There’s certainly a difference in the way we see the secondary and tertiary mutations aggregate out, too. Cytopenic myelofibrosis tends to have a higher number and higher VAFs of second, third, and fourth mutations, particularly adverse-risk mutations, where we tend not to see those as often with proliferative myelofibrosis.
In proliferative myelofibrosis, [patients] tend to have larger spleens vs cytopenic myelofibrosis, and there might be some differences in clot risk. There may be some differential effects of how [patients] respond to JAK inhibitors and which JAK inhibitors they respond to, so there may be some therapeutic implications as well; however, [we don't have] a unified, agreed-upon classification for cytopenic myelofibrosis, and it really hasn’t been studied in clinical trials. There are some retrospective data looking at the differences, but ultimately, we still pick the JAK inhibitor that best fits the individual patient, irrespective of whether they have cytopenic or proliferative myelofibrosis.
Anemia [management] is incredibly important, and it’s very common. Roughly 40% of patients will be anemic at the time of diagnosis, and virtually every patient will develop anemia at some point in time, whether it’s disease-related caused by [ineffective] erythropoiesis or treatment-related, [since] JAK inhibitors can make patients anemic. It [can also be] due to splenic sequestration or even inflammation, [where] cytokine levels, particularly interleukin-6, get high, then the hepcidin levels subsequently go high, and that can cause anemia. There’s a range for anemia or the reasons why anemia may happen in any given patient, and that actually makes anemia pretty difficult to manage.
[Anemia] is a very common problem and a very prognostic problem. Anemia predicts survival in patients with myelofibrosis, as well. It’s a very difficult problem to manage because it’s multifactorial.
We think about a lot of different things when selecting a JAK inhibitor for treatment. To paraphrase my good friend Gabriela S. Hobbs, MD, [of Mass General Brigham in Boston, Massachusetts], ruxolitinib [Jakafi] is still king. That is the first JAK inhibitor that was FDA-approved [for myelofibrosis], and it’s the one we often think about first. Many people will go through their minds and say, ‘Let's think about if ruxolitinib is a good idea,’ and if the answer is yes, we use it. If the answer is no, then we think about the other JAK inhibitors.
JAK Inhibitor Selection in Myelofibrosis
Another way of thinking about [JAK inhibitor selection] is the way we parse stuff out. If a patient has proliferative myelofibrosis, [where] they have a good platelet count and they don't have significant anemia [with] a hemoglobin [level of, for example,] less than 9.5 g/dL, ruxolitinib is the way to go. If a patient does have anemia—say, for example, their hemoglobin [level] is 9 g/dL or even lower—we often think about momelotinib [Ojjaara] in those patients who need spleen and symptom control [along with] that myelofibrosis-associated anemia.
For pacritinib [Vonjo], we think about patients who have platelet counts that are below 50 x 109/L, or even patients where their platelets could get to [below] 50 x109/L soon. Even patients with a platelet count of 65 x 109/L or 70 x 109/L were included in the [phase 3] PERSIST-2 trial [NCT02055781], so we’ll think about pacritinib for those thrombocytopenic patients.1
Then, we often think about fedratinib [Inrebic] for patients who have received prior ruxolitinib. That’s really where it shines. In the [phase 2] JAKARTA2 trial [NCT01523171], we saw high rates of response in terms of spleen volume response and symptom response in patients previously treated with ruxolitinib.2
Combination therapies have really been hindered by the fact that we are so hung up as a field and [in the] regulatory process on spleen and symptom [data]. Pelabresib [CPI-0610] is a great example of this, [where we saw almost] doubling of spleen volume [reductions] of at 35% [from baseline at week 24 with pelabresib plus ruxolitinib vs ruxolitinib plus placebo in the phase 3 MANIFEST-2 trial (NCT04603495)], but symptom burden [changes] didn’t quite hit statistical significance.3 Still, patients were, on average, feeling a little bit better [with pelabresib plus ruxolitinib]—again, not statistically significant, but definitely trending in that direction.
Additionally, we saw anemia improvements with pelabresib combinations vs a JAK inhibitor alone, and the totality of data suggests that this is a very powerful treatment. It’s certainly not adding a significant amounts of toxicity using the combination vs a JAK inhibitor. However, it didn’t hit [statistical significance] on that [symptom] end point, so things are kind of in disarray, which I think is really unfortunate because the trial design focused around these two end points, [spleen and symptom improvements], for regulatory approval, and that has failed this combination for our patients.
The [phase 3] SENTRY trial [NCT04562389] data are provocative. There is actually a measurable [overall] survival advantage; it’s very slim, and if you look at the curves, you’re not going to drive a truck through them, but there is a slight separation. If that continues to trend out, that could be an important signal. Along with that, we’ll have to link [the findings together]. Just showing survival is not good enough. We’ve shown that with ruxolitinib vs the control arms in the [phase 3] COMFORT-I [NCT00952289] and COMFORT-II [NCT00934544] studies, where ruxolitinib led to a survival advantage, but we didn’t call this disease modification just because there wasn’t all that complementary data. Patients’ symptoms also have to get better, and we have to have other markers of deepening responses.
In SENTRY, they did look at reductions in allele burden, which is important, and they arbitrarily picked a 20% reduction in the VAF [as a measure for the secondary end point]. Is that [20% reduction] enough? Is that not enough? [Either way], more patients had a reduction of at least 20% in their VAF with the combination of selinexor [Xpovio] and ruxolitinib vs ruxolitinib alone. There were associations [in improvements] with spleen volume response and survival in that study, and that could be helpful in terms of pulling all this together to see if this is a viable path forward.
To compare and contrast [selinexor plus ruxolitinib] against the pelabresib experience, the toxicity profile was much different. There’s always that plot where it almost looks like a funnel, where one way is the toxicity with the combination, and the other direction is the toxicity with the single agent alone. [When comparing selinexor plus ruxolitinib vs ruxolitinib alone in SENTRY, this funnel] wasn’t a perfectly symmetric figure. There was definitely much more toxicity with selinexor, and we know this from our experience in multiple myeloma. This is a pretty significant drug in terms of gastrointestinal toxicity in particular. That’s something we’ll have to keep an eye out for. [We need to see if] patients can take this medication and take it for a [long time], because this is a chronic illness. That’s the other thing we often get hung up on when we look at these adverse effects [AEs]. A grade 1 AE might be very significant because [we don’t need patients to only get through] 4 cycles of adjuvant chemotherapy, then they’re done. These are drugs that, presumably, patients will take for years and years. Even a grade 1 toxicity can significantly affect patient’s quality of life, and that needs to be a consideration.
There is a lot to unpack from SENTRY and with combination therapies, but this is the way forward. JAK inhibitors are great, and they’ve changed the field. Patients are better, and they’re living longer because of that, but [these agents] are not truly rooting out disease in the deep and fundamental way that we want. That is going to require combination therapies with additional therapies brought in that can be taken by patients for the long term to lead to these long-term benefits that we are all hoping for.
JAK inhibition is fundamental to what we’re doing, and no one will argue [against the fact] that JAK inhibitors shrink spleens, improve symptoms, and make patients’ lives better. A lot of these newer agents can do [these things], but not to the degree that I think JAK inhibitors can. A lot of times, these newer therapies take time to [affect a patient’s disease]. You may need 3, 6, or 12 months’ worth of treatment to really root out disease to get to that point. Because of the speed and intensity with which JAK inhibitors can improve symptoms and splenomegaly, I think they will continue to be a very important part of what we’re doing going forward, and it’s going to be with combination therapy. It’s just about finding the right partners that can affect the right things, work with JAK inhibitors, and get people feeling better.
The [CALR mutation–targeting] monoclonal antibody [INCA033989] is a great example [of a potential combination partner] because the AE profile for the fully humanized monoclonal antibody plays well with others. There’s not going to be a lot of overlapping toxicity [with a JAK inhibitor]. There was very little toxicity observed so far with [INCA033989 monotherapy], which we have the most data for [among CALR mutation–targeting agents under development], and you look at that naturally and say this will be easy to combine with other therapies. You can imagine combination therapies, triplet therapies, even quadruplet therapies, like in multiple myeloma, in the future [for myelofibrosis].
