Expansion of AML Treatment Landscape Underscores Importance of Molecular Testing

Naval Daver, MD

The treatment landscape for acute myeloid leukemia (AML) has evolved significantly over the last decade, and with the approval of several targeted therapies, conducting genomic testing on all new patients has become essential.

While many patients receive intensive induction chemotherapy today, physicians can select a number of different therapies or combinations to improve outcomes in patients on top of the chemotherapy. For example, patiens with FLT3 mutations now have the opportunity to receive a FLT3 inhibitor in conjunction with chemotherapy. Likewise, IDH inhibitors are also available for patients with IDH-mutant AML.

In order to appropriately sequence treatment in patients and prolong their survival outcomes, it is important to evaluate the potential cytogenetic abnormalities or mutations that may be present in each patient with AML, said Naval Daver, MD. Once molecular and cytogenetic profiles are determined, physicians can personalize their treatment approach to each individual patient.

In an interview with Targeted Oncology, Daver, associate professor in the Department of Leukemia at the University of Texas MD Anderson Cancer Center. discussed the current treatment landscape for AML and how he goes about sequencing the available agents when he treats patients in clinic.

TARGETED ONCOLOGY: What are some of the latest updates we have seen in the treatment landscape for patients with AML?

Daver: There has been major progress in the treatment of acute leukemias, both AML and acute lymphocytic leukemia (ALL). Some of the important updates that were presented at the 25th Congress of the European Hematology Association and 2020 American Society of Clinical Oncology Virtual Scientific Meeting were updates in AML, specifically with the VIALE-A study, which was the combination study of azacitidine and venetoclax (Venclexta) versus azacitidine. This was a multinational phase 3 randomized study and showed that the combination of azacitidine with venetoclax significantly improved the overall survival (OS) as well as dramatically improved the response rates compared to azacitidine alone.

Historically, we used azacitidine alone for many years. This gave us a complete response/complete response with incomplete hematologic recovery (CR/CRi) rate of about 20% to 25% with a median OS of 8 to 10 months. Based on encouraging phase 2 single-arm data for azacitidine and venetoclax, we started seeing CR/CRi rates of about 70% to 75% and a median OS of 17 to 18 months. This led to the initial FDA approval of the combination of azacitidine or decitabine with venetoclax in 2018. This required a confirmatory phase 3 study, which showed that the CR/CRi rates were about 70% with azacitidine/venetoclax as compared to about 20% with azacitidine, both meeting the expected historical outcomes. The median OS was significantly improved at 15 months versus 9 months for azacitidine/venetoclax versus azacitidine. This is important in clinical practice because this now confirms what most of us in the United States have already been doing, which is using azacitidine in combination with venetoclax in older patients who are not considered good candidates for intensive chemotherapy. Also, this is very important clinically because these patients tend to have very dismal outcomes historically. Now for the first time with the combination, we're seeing both high response rates as well as 2- to 3-year survival rates of 40% to 50%, so in fact, we think that some of these patients may have durable and potentially even long-term survival.

Additionally, there has been a lot of updates of new combinations. There are combinations of azacitidine and venetoclax with targeted therapies, specifically IDH inhibitors that are showing that adding a third drug, specifically a targeted therapy, to the backbone of azacitidine/venetoclax can further improve the depth of response and hopefully over time, you will see the duration of response and survival [improve].

There are some new immunotherapies that are looking encouraging in acute myeloid leukemia. These include a CD47 antibody, which is an immune checkpoint on the surface of macrophages. By blocking CD47, we're able to enhance macrophage function to allow phagocytosis to acute leukemia cells. Some of the early data from the ongoing phase 1B study both in frontline and older patients with AML, including high-risk and TP53-mutant AML, as well as in the higher risk MDS patients, showed very encouraging early responses, as well as very good safety and durable survival.

The other immune therapy that looks like it may have some activity in AML similar to what has been shown in ALL is blinatumomab (Blincyto). This is a bispecific antibody that has been used in ALL and now in AML, and we have a number of similar bispecific antibodies. AMG 330 was the 1 that was presented at the ASCO and EHA meetings this year. It's shown signs of activity in multiply relapsed patients with AML. I think, ideally, bispecific antibodies will be best tested in the low minimal residual disease (MRD) burden setting, and these trials are now being amended to focus on the MRD, early relapse population, but we are seeing activity with these agents. There is hope that there will be a future for these to be developed in AML.

TARGETED ONCOLOGY: When you see a patient diagnosed with AML, how do you sequence therapy?

Daver: For AML, the frontline treatment now has become much more diversified, and this is a good thing because we are seeing improvement in outcomes by using mutation-directed personalized therapy. When I see an AML patient in clinic, the first thing that I decide is whether I think this patient will be a suitable candidate for intensive induction chemotherapy or not. That decision will heavily base what we think the early mortality with induction therapy will be for that patient, so for a patient who is above 70, and definitely above 75 years, who has comorbidities or other issues, such as recent hospitalizations or infections, they often tend to have a high early mortality. There are calculators that can be used to predict the early mortality from intensive chemotherapy, and we think that the early mortality will be 15% or 20%, or more if we are looking more towards what we call lower intensity therapies.

Let's say I do decide the patient is a good candidate for intensive chemotherapy, then the next 4 steps are trying to identify what intensive chemotherapy or what combination of drug I can add to the intensive chemotherapy, so we first check for cytogenetics looking for core binding factor. Core binding factory AML is an AML that is associated with cytogenetic abnormalities. If a patient has 1 of these, we will go for induction chemotherapy such as fludarabine, cytarabine and filgrastim with the addition of gemtuzumab ozogamicin (Mylotarg). There's a lot of data showing that gemtuzumab significantly improves OS, especially in patients with core binding factor AML receiving induction chemotherapy. That makes up about 15% of all newly diagnosed AML, and the survival for that population can be up to 80% at 5 years, so very curable.

Second, I will look for other cytogenetic changes that are associated with myelodysplastic syndrome (MDS). These are called myelodysplasia-related changes. These include deletions 5, 7, 17, INV(3) or INV(1). If we find these, we will consider the use of a drug called CPX-351, which has shown specifically improved survival in secondary or therapy-related AML, including AML with MDS-related cytogenetics.

The third thing I will look for is a FLT3 mutation. This usually comes on molecular sequencing. If a patient has a FLT3 and I'm giving induction therapy, I will add a FLT3 inhibitor to the standard, which today is midostaurin (Rydapt). However, the next-generation FLT3 inhibitors, such as gilteritinib (Xospata), or quizartinib, will probably give us better outcomes when added to chemotherapy, and we're doing trials combining these agents now.

The fourth thing is that if the patient has none of these mutations or cytogenetic abnormalities, then we will consider the use of induction chemotherapy. It could be done alone as is the standard, but we have a trial using induction chemotherapy with the addition of venetoclax. With this approach, we're seeing much higher MRD-negativity, and we will see over time if that translates to prolonged survival.

These are the 4 specific subsets based on molecular and cytogenetic profiling at baseline. Putting the patient in the right treatment can really improve the survival by up to 20% to 30%.

What would you recommend for a patient who may not be a good candidate for intensive chemotherapy?

Daver: If the patient is considered to be not suitable for intensive induction, then we are usually going for a hypomethylating agent (HMA) with a venetoclax-backbone. Now, the standard is HMA/venetoclax based on the BELIA study in the FDA approval, but we're trying to build on this, so if somebody has a FLT3 or IDHmutation, we're doing HMA/venetoclax with the FLT3 or IDH inhibitor added to it. If somebody has no mutations that are targetable, we may do HMA with venetoclax plus the addition of an antibody or an immune checkpoint inhibitor, so this is the current approach. As you can see, it's very personalized. It's no longer 1 treatment fits all, so we do need molecular cytogenetic information to make the decision.

TARGETED ONCOLOGY: Do you find that maintenance chemotherapy is effective in this space?

Daver: In AML for many years, there's been debate about whether maintenance will be effective or not. A number of phase 2 studies have shown that there was potentially improved outcomes with maintenance, but it was always hard to tease these out because then there were also some phase 2 studies that showed that maybe there was not a pure benefit. I think the large study that has kind of put this debate to rest, at least in my mind, is the oral CC-486 study. This was a maintenance study using oral CC-486, which is an epigenetic agent, as a maintenance in patients who had completed their planned induction/consolidation and were not going to stem cell transplant (SCT). It's important to know that these are patients who are not going to SCT, and the reason they were not going to transplant was either because of age, no donor, financial restrictions, or patient preference. If a patient was able to go to transplant and 1 thought that he should go to transplant, then the patient should still proceed to SCT. The study did not take patients off transplant, so if for some reason the patient eventually could not go to transplant, which is common in 30 to 50% of patients, then they were randomized to receive either CC-486 or placebo, which is a standard where we would do observation.

This study showed that with the addition of the CC-486 as maintenance after planned induction/consolidation in patients who could not make your transplant, there was a killer survival advantage with a median survival of 26 versus 15 months, and this was maintained at 2- and 3-year follow-up. Now I think will lead to the approval, hopefully, of CC-486 as the maintenance, post-induction/consolidation for patients with AML who cannot go to transplant. It also establishes the transplant is definitely beneficial. Now, this is the first step.

TARGETED ONCOLOGY: Where do you see the future headed for this treatment landscape?

Daver: The next questions we'll look at are, what about maintenance in patients who can go to transplant? Could you potentially reduce the need for transplant by doing maintenance? That's a big question and a completely separate study that will have to be done. Also, what about maintenance with retargeted agents? There is data to suggest that post-transplant FLT3 inhibitor maintenance can improve survival, and we routinely use FLT3 inhibitors as maintenance in FLT3-mutated patients going to transplant. Other trials will look at IDH inhibitors, both post-transplant and post-induction/consolidation.

The most recent and exciting trials are going to look at oral CC-486 or azacitidine with venetoclax as maintenance, both post-transplant, and post-induction/consolidation. I think the maintenance field in the next 5 to 6 years in AML is going to develop quickly.