With Novel Combos on the Horizon, End Points Beyond OS May Hold Promise in HR-MDS

Mikkael A. Sekeres, MD

Novel combinations with agents, such as pevonedistat, eprenetapopt (APR-246), venetoclax (Venclexta) or magrolimab, plus hypomethylating agents (HMAs) like azacitidine are being evaluated to determine whether they can improve upon the lackluster median survival for patients with higher-risk myelodysplastic syndrome (HR-MDS), said Mikkael A. Sekeres, MD, who added that although overall survival (OS) remains the gold-standard end point, trials are evolving to incorporate event-free survival (EFS) and other relevant end points for this patient population.

“We've got to do better for patients with HR-MDS, many of whom have a median survival that is certainly less than 2 years, and, for many, less than 1.5 years,” Sekeres said. “As a primary goal of trials in HR-MDS, we are looking for an improvement in OS, but an interim marker may emerge in EFS for something that is clinically meaningful.”

In an interview with OncLive®, Sekeres, a professor of medicine and chief of the Division of Hematology at the Sylvester Comprehensive Cancer Center of the University of Miami Health System, discussed the unmet needs of single-agent HMAs in HR-MDS, novel combinations in development, and the incorporation of other end points beyond OS in clinical trials. 

OncLive®: What is the current role of HMAs in the treatment of patients with HR-MDS?

Sekeres: When I first see a patient with HR-MDS, the first drug I turn to is an HMA. We have 3 to choose from: azacitidine, decitabine, and oral decitabine plus cedazuridine. [Azacitidine and decitabine] have been around for around 50 years. It has taken us a long time to figure out the right dose and schedule to give to patients, as well as the right disease to choose. We finally were able to settle on MDS. Both azacitidine and decitabine were approved by the FDA back in 2004 and 2005 for all subtypes of MDS, but there are a couple of studies that led us to land on treating these patients with HR-MDS first with these drugs.

The first [trial] was the AZA-001 study [NCT00071799] that was conducted in Europe. This was a randomized trial in which patients with HR-MDS received azacitidine or conventional care regimens. Those conventional care regimens could have consisted of best supportive care, which 60% of patients received, or low-dose cytarabine, or acute myeloid leukemia (AML)–type 7+3 chemotherapy.

In that study—the only prospective, randomized trial to ever show a survival advantage in HR-MDS—patients who received azacitidine lived a median of 24.5 months vs 15 months for those patients who received conventional care regimens.

For decitabine, a similar study was conducted in Europe, but had quite different results. Those patients with HR-MDS who were randomized to receive decitabine, lived a median of 10.1 months compared with only 8.5 months for patients randomized to receive best supportive care. Those differences were not significant as opposed to [those observed in] the AZA-001 study.

How do we tie all of this together? We believe azacitidine and decitabine are very similar drugs and should have similar outcomes, so these studies just simply enrolled different patient populations. My conclusion from this is that when I am meeting a patient with HR-MDS for the first time, I recommend azacitidine as my first choice, but I don’t criticize [clinicians] who choose decitabine instead because it should work just as well as azacitidine.

We have a new drug that is oral decitabine in combination with cedazuridine that was approved by the FDA. This regimen was approved based on pharmacokinetic and pharmacodynamic data showing that it was similar to intravenous [IV] decitabine but without survival data. Although we can technically use this drug for patients with HR-MDS, I reserve it for patients with lower-risk MDS because we don’t have comparative data to show that it is as good as azacitidine or IV decitabine.

What benefits does transfusion yield for patients with HR-MDS? Should transfusion independence be a goal of treatment?

Patients with HR-MDS almost invariably present with profound cytopenias and may already require transfusions before I even meet them. Those could be blood or platelet transfusions. About 80% of patients with MDS have some degree of anemia at presentation or early in their disease course. These folks continue to receive transfusions, sometimes as we are starting therapy or well into therapy. We measure benefit in patients with HR-MDS by whether they achieve a complete response [CR], but also by whether they achieve transfusion independence or a reduction in the frequency of transfusions. That definition is called hematologic improvement. Patients achieve a hematologic improvement if they start a study already dependent on blood transfusions and then go an 8-week period on that study becoming transfusion independent. Someone may take a step back from this definition and ask, “Is it really worth my patient getting a shot in the stomach—which is how we give azacitidine—or a shot in the arm for 7 days out of every month just so they can go an 8-week period without requiring a blood transfusion?” A lot of us in this field think that is not worth it.

A couple of years ago, a bunch of us internationally proposed a revision to those criteria where a patient would have to go 16 weeks without requiring a transfusion to achieve transfusion independence. We do think [transfusion independence] represents an improvement in quality of life for patients.

What unmet needs regarding survival are research efforts aimed at addressing?

I described a drug that was shown in a randomized, prospective trial to provide a survival advantage compared with conventional care regimens. In the AZA-001 study, the median survival advantage was 24.5 months [with azacitidine] vs 15 months [with conventional care regimens], so a 9-month survival advantage. I would love to know what “magic dust was sprinkled” on these patients in this AZA-001 trial because no randomized trial subsequent to this has ever shown a median survival for azacitidine-based therapy that has reached 24.5 months. To be realistic, much more typically, a patient will have a response to azacitidine that lasts about 17 months if that patient is among the 35% to 40% who achieve a response.

Classically, we’ve looked at CRs as an interim marker of improved OS, but, unfortunately, several studies in HR-MDS and in older adults with AML who have a disease that is biologically very similar to HR-MDS have shown that patients can achieve upward of double the CR rate without improving their OS at all.

Trials that are more recent are focusing on [end points] like EFS. How long does a patient live or live without transformation from HR-MDS to AML, which can serve as an interim marker of improved OS?

What investigational combination regimens do you find most compelling?

Several trials are combining drugs with azacitidine. A potential future in therapy for HR-MDS is to combine drugs with an HMA-based backbone. One combination of drugs involves the NEDD8–activating enzyme inhibitor pevonedistat plus azacitidine. In a trial in HR-MDS, patients were randomized to receive pevonedistat and azacitidine vs azacitidine alone. I was fortunate to be involved with this study.

In this study, among patients with HR-MDS, there was a significant improvement in EFS, at a median of [21] months for patients who received pevonedistat plus azacitidine vs [16.6] months for patients who received azacitidine. This study was not powered to show an improvement in OS, yet numerically, [OS] was also improved in patients who received the combination, at 24 months compared with 19 months for those patients who received azacitidine.

Another attempt at a combination of therapy is the drug eprenetapopt that was combined with azacitidine, specifically to treat patients who have a TP53 abnormality. Eprenetapopt restores wild-type TP53 confirmation and activity. In a single-arm study, which I was also privileged to participate in, the overall response rate [ORR] approached 85% to 90% in patients who received the combination.

Unfortunately, in a randomized trial, preliminary data that have been released in a news release but have not been peer-reviewed, have shown that although the response rates were improved for patients who received the combination at 33% compared with 22% for patients who received azacitidine alone, the OS did not appear to be improved. There’s more to come on that as the phase 3 trial data are released.

Another combination approach that has become a standard in older adults with AML is to combine azacitidine plus venetoclax, which is a BCL-2 inhibitor. There was a phase 1b study of the combination of azacitidine and venetoclax that was presented during the 2020 ASH Annual Meeting that showed a CR rate of 42% with an additional 14% of patients who had a hematologic improvement. These are single-arm data with impressive response rates so far. We are waiting to see what a randomized trial will yield comparing azacitidine and venetoclax to azacitidine alone.

A final combination is that of magrolimab, which is an anti-CD47 monoclonal antibody, combined with azacitidine. Magrolimab is a major macrophage immune checkpoint and [releases a] “don't-eat-me” signal in myeloid malignancies.

In this trial, patients with HR-MDS and AML were treated with the combination [of magrolimab and azacitidine]. Once again, [the data showed] a very high response rate that approached 90% for the combination vs 35% to 40% with azacitidine alone, which is what we would expect. A randomized trial has been initiated comparing azacitidine and magrolimab vs azacitidine alone. We are eagerly awaiting results of that trial to be reported.

How is pevonedistat in combination with azacitidine performing in patients with HR-MDS?

[Regarding] the trial that looked at pevonedistat plus azacitidine vs azacitidine alone, we presented some data at the 2021 European Hematology Association Virtual Congress. We demonstrated that response was associated with a lower likelihood of molecular evolution, or acquiring additional molecular abnormalities, as well as a shrinking of the variant allelic frequency of known molecular mutations. That was among responders vs non-responders to pevonedistat plus azacitidine vs azacitidine alone.

What clinical end points are important to evaluate beyond OS for emerging therapies?

For patients with HR-MDS, OS is the gold standard. We would love for our patients to be able to live longer. We would also love for them to live better, or in other words, to have improved quality of life [QOL]. These are the 2 standards that the FDA uses for approving drugs for any cancer indication.

Unfortunately, QOL studies and instruments measuring QOL in MDS haven’t demonstrated that routinely. There are a couple of reasons why. First, a lot of studies have been investigated using the EORTC QLQ-C30 QOL instrument. This simply isn’t very sensitive to changes in patients with MDS who are undergoing treatment. A better instrument would be the QUALMS [Quality of Life in Myelodysplasia Scale], which was developed by Gregory A. Abel, MD, MPH, of the Dana-Farber Cancer Institute.

Secondly, QOL instruments and measurements are tricky in patients with HR-MDS because in one arm we may have a patient who has been given a drug and responds. Therefore, her blood counts improve where she doesn’t need transfusions as often. In the arm in which patients aren’t responding, they are still getting blood transfusions, which are in themselves improving the patient’s QOL. That equilibrates the differences in QOL that we may see between the arms.

OS is still the goal standard. I would love to see, for example, the use of wearable devices to measure QOL so we can see how patients are performing at home rather than when they are putting on their best face in the clinical setting.

How is pevonedistat tolerated among patients? What are the potential clinical implications of the activity demonstrated with pevonedistat?

One of the reasons we’ve seen good results from the combination of pevonedistat and azacitidine so far is that it is so well tolerated. I led a North American intergroup trial in HR-MDS that randomized patients to receive azacitidine monotherapy, azacitidine combined with lenalidomide [Revlimid], or azacitidine combined with the histone deacetylase inhibitor vorinostat [Zolinza]. In that study, although we did show an improvement in ORR for azacitidine plus lenalidomide compared with azacitidine alone of 11%, it wasn’t significantly different because the study was powered to show an improvement in ORR of 20% [or more]. However, from that study we learned that patients who have HR-MDS and are receiving combinations of therapy are at risk of having that therapy dosing lowered or therapy stopped because of toxicity. Often, the adverse effects [AEs] that are justified to lower or stop therapy are not serious, at least not as reported in the clinical trial. Those dose reductions or stopping treatment occurs in ways that are not specified by a clinical trial. In other words, a clinician decided to [move forward with treatment modification or discontinuation].

In the pevonedistat plus azacitidine trial, we had much stricter language in the protocol itself so that providers had to check with a medical monitor before they lowered or stopped dosing of the combination. We also found that AEs to the combination were about the same as what we saw with azacitidine monotherapy. Because the combination was better tolerated, patients remained on it longer, and that is probably why we saw the emergence of better responses even with azacitidine monotherapy.

There are a couple of clinical implications to studies with combination therapy. First, we have to give something that is tolerable. Secondly, we have to stick with [treatment] in older adults with AML so that they are on therapy long enough to see improvements in EFS and OS.

Are there other areas of ongoing research you are keeping an eye on?

One other area that is evolving is the use of targeted or genetically directed therapy. We have data on 17 patients who were treated with enasidenib [Idhifa]. Those patients all had IDH2 positivity. This is relatively infrequent in MDS compared with AML and occurs in about 5% of patients. When we can identify those patients with MDS, we have an opportunity to treat them after HMAs have failed them or to combine those drugs with HMAs to see if we can get a more robust response that is more durable up-front with those patients.

We are starting to see this more with other types of drugs that are targeted; however, not all of them have been successful. For example, there was a recent drug made by H3 Biomedicine that targeted patients who had spliceosome mutations [and failed]. There are second-generation drugs that are coming out focusing on spliceosome mutations, as well as those that are trying to take advantage of the immunologic mediated aspects of MDS, such as sabatolimab, which is now in a phase 3 trial combined with azacitidine.