Harry Erba, Md, PhD, Discusses Treatment Shift in AML with Maintenance Therapy

Harry Erba, MD, PhD, Professor of Medicine, Director of the Leukemia Program, Division of Hematologic Malignancies and Cellular Therapy, Duke University, Durham, North Carolina, discussed how the treatment paradigm for acute myeloid leukemia (AML) has changed with maintenance therapy at the virtual Great Debates & Updates in Hematologic Malignancies.

Currently, the treatment algorithm for AML starts at the time of diagnosis, when an assessment is made on whether a patient is fit or unfit for intensive chemotherapy, Dr Erba said. The decision also is based on features such as cytogenetics and mutational data.

After achieving remission, patients who were fit for intensive chemotherapy may undergo observation, consolidation therapy, or allogeneic stem cell transplant (HSCT), he said. The decision for allogeneic HSCT is based on risk stratification and minimal residual disease (MRD) assessments.

Dr Erba said the amount of appropriate consolidation therapy is a matter of some debate, as it varies between younger and older patients in terms of the number of cycles and the dose of cytarabine that might be used. There are even patients who may not undergo consolidation chemotherapy, such as those with complex karyotypes that received induction chemotherapy and achieved a remission but have such a low chance of prolonged survival that they don’t receive it.

During induction with either less intensive or intensive therapy, Dr Erba said, patients may get refractory disease. They may even relapse after achieving remission and then need to undergo salvage therapy. After induction therapy and achieving a remission or following consolidation, patients typically undergo observation, but Dr Erba said it’s important to consider maintenance therapy at this point as well.

There have been multiple attempts to improve the survival of AML patients following intensive induction therapy with or without consolidation, Dr Erba said.

For example, a recent trial by the HOVON Group studied patients with AML receiving 7x3 followed by an intermediate dose of cytarabine for 12 doses. Patients were then moved onto a study evaluating gemtuzumab ozogamicin, an anti-CD33 antibody drug conjugant, with 3 doses once every 4 weeks compared with observation.

“There was no difference, no improvement in disease free or overall survival (OS),” Dr Erba said.

Histamine dihydrochloride/IL-2 was also evaluated as a maintenance therapy after initial induction and consolidation therapy for patients with AML in remission, Dr Erba said. In this study, patients (n=320) received this regimen for 3 weeks followed by a 3-to-6-week rest for 18 months versus observation. The leukemia-free survival for those patients randomized to IL-2 with histamine was superior, however there was no difference in OS.

In terms of hypomethylating agents, Dr Erba referred to the ECOG-ACRIN E2906 randomized study, which evaluated a maintenance regimen of decitabine for older patients with AML who had achieved first remission with intensive chemotherapy.

There, patients were randomly assigned to either 7x3 followed by intermediate dose cytarabine for 2 cycles or clofarabine induction consolidation. The study stopped early, he said, because of a higher OS in patients receiving 7x3 followed by cytarabine. Patients who did not undergo allogeneic stem cell transplant after achieving remission were randomly assigned to receive decitabine 20mg/m² daily for 3 days or observation. According to the statistical plan of the study, there was superior OS (P=0.06) and improvements were seen in disease-free survival (DFS, P=0.12). However, there was significant OS impact on the FLT3-ITD-negative population.

Dr Erba also mentioned the phase 3 HOVON97 multicenter study, where patients with AML in remission who did not undergo allogeneic stem cell transplant and who had recovered from the toxicities were randomly assigned to either azacitidine as a maintenance therapy (n=56, 50mg/m², subcutaneously, days 1-5, every 4 weeks) until relapse, for a maximum of 12 cycles, or no maintenance (n=60).

“There was a clear improvement in DFS … however there was no significant improvement in overall survival unless you censored the patients the time of allogeneic transplant,” Dr Erba said.

That history, Dr Erba said, helped to created CC-486, an oral hypomethylating agent with a distinct PK/PD profile from injectable azacitidine.

CC-486, he said, demonstrates clinical activity in patients with hematologic malignancies. The oral dosing allows for extended drug exposure during each treatment cycle to prolong therapeutic activity. The mechanism of action appears to be through DNA hypomethylation inducing DNA damage, and the disruption of RNA.

The QUAZAR AML-001 study was designed to show an improvement in OS for patients who had received intensive therapy but had not undergone allogeneic transplant, Dr Erba said. This was an international, multicenter, placebo-controlled, double-blind phase 3 study from 148 sites in 23 countries. Patients enrolled were age ≥55, had an ECOG PS score of 0-3, were ineligible for HSCT, found to be fit for intensive chemotherapy, and received an intensive chemotherapy induction followed by consolidation or not. Overall, 472 patients were randomized 1:1 to receive either CC-486 300 mg daily for 14 days (n=238) or placebo (n=234) on the same schedule. Therapy was continued if there was no evidence of relapse or unacceptable toxicity.

“Patients who had over 15% blasts were removed from study, that was failure and relapse. If they had 5-15% bone marrow blast, they could receive a longer duration of placebo or CC-486. I will not show you the data from that cohort of patients, but about 25% of them could actually achieve again a complete remission or CRI with 21 days of azacitidine,” Dr Erba said.

In all, 79% (n=187) of patients were in complete remission, and 21% (n=51) had CRI. About 80% of patients had received some amount of consolidation, typically 1 or 2 cycles, and 20% did not receive consolidation. The primary end point of the QUAZAR study was met, Dr Erba said. Oral azacitidine given at 300mg daily for 14 days in a 28-day cycle significantly improved OS and relapse-free survival (RFS) compared with placebo. The median OS improved from 15 months with placebo to 25 months with oral azacitidine.

“Oral azacitidine may delay relapse but does not actually cure patients who are not already cured in a long-term disease-free remission with intensive therapy. That is supported by the RFS, showing again an improvement from 5 months to 10 months, but ultimately no difference in RFS after 24-36 months,” Dr Erba said.

In a variety of subset analysis, Dr Erba said, treatment with oral azacitidine was favored over placebo. There was an apparent benefit of oral azacitidine in each subset, including age, gender, WHO AML classification, performance status, history of MDS or CML, cytogenetic risk at induction, consolidation after induction, and whether they were in CR or CRI, or whether they were MRD positive or negative.

“The only hazard ratio (HR) to the left of that line is 3 cycles of consolidation, but that was a very underpowered group of patients,” Dr Erba.

Dr Erba pointed out that RFS following consolidation improved with oral azacitidine regardless of the number of consolidations. MRD was assessed using 22 cell surface markers by flow cytometry with an MRD positivity cutoff of ≥0.01%. The MRD assessments were done routinely every 3 cycles initially through the study and then every 6 cycles.

Gail Roboz, MD, presented during her ASH presentation that oral azacitidine showed a significant treatment benefit over placebo for OS (P=0.0067) and RFS (P=0.0001), regardless of MRD status, according to Dr Erba. The presence of MRD at study entry was significantly associated with shorter OS and RFS (both P=0.0001). For patients that were MRD negative vs. MRD positive at study entry, oral azacitidine resulted in improved OS from the time of randomization over placebo in both groups. The HR of 0.69 was statistically significant for patients who were MRD positive. This study also showed that 37% of patients became MRD responders with further oral azacitidine.

“If you look at the Kaplan Meyer curves for OS and RFS based on not only treatment assignment but MRD status, there was a benefit of oral azacitidine in both the MRD positive and MRD negative groups,” Dr Erba said.

He continued: “Now, what’s interesting about that data is 19% became MRD negative with placebo. Now this may be due to the fact that response following intensive chemotherapy could potentially be delayed. Personally, I think it’s more likely to reflect the vagaries and difficulties of MRD assessment on bone marrow aspirates done on clinical trials … However, those patients who are MRD negative at baseline or were MRD responders had a significantly prolonged median duration of MRD negativity with oral azacitidine versus placebo, 5 months vs 11 months.”

Dr Erba then discussed data from EHA on intermediate-risk versus poor-risk karyotype at baseline. He said this is the first study looking at data from the patient where their NPM1 status is known. Focusing on the NPM1 mutated group, Dr Erba said the median OS improved from 16 months with placebo to 46 months (HR 0.57) with azacitidine. This was also accompanied by an improvement in median RFS.

Turning his attention to gastrointestinal (GI) safety, Dr Erba focused on the 2 major toxicities that were seen in the QUAZAR study: GI and hematologic toxicity.

Dr Erba said GI adverse events (AEs) with CC-486 were the most common during the first 2 cycles, with more than half the patients having nausea (65%), vomiting (60%), and diarrhea (50%). There were very few grade 3 or 4, but neutropenia was more commonly seen with CC-486 (45%) than placebo (26%). The placebo group had 24% of patients develop grade 3-4 neutropenia, which may be due to relapse, Dr Erba said, not the placebo.

Serious AEs were more commonly reported in 34% and 25% of patients in the CC-486 arm and placebo arm, respectively. There were also no treatment-related deaths.

The median duration of exposure for CC-486 was 12 cycles and placebo was 6 cycles. Dose modifications were more commonly seen with CC-486, with 43% of patients requiring interruptions and 16% requiring reductions due to GI toxicity and cytopenia. Neutropenia was the most common AE requiring dose interruption/reduction in both arms. However, discontinuation due to AEs was infrequent (CC-486 13%; placebo 4%).

In summary, Dr Erba asked the question, “Has the availability of a maintenance therapy that now has been show to improve the OS survival of AML patients changed the treatment paradigm,” and then answered.

“I would say, ‘Yes absolutely.’ But, it has its place, and the impact in the entire treatment paradigm is marginal,” he said.

However, there are things to consider.

For example, Dr Erba said favorable risk cytogenetic patients were excluded from the QUAZAR study in general because they do well. However, there are patients with core binding factor leukemias that may still relapse. When relapse occurs in a patient with 821, they are often difficult to get back in remission. Furthermore, these patients often do not proceed to allo transplant in first remission.

“I think it is reasonable to consider oral azacitidine even for patients with favorable risk cytogenetics. However, there’s no data to support that statement at this time,” Dr Erba said.

Oral azacitidine is not indicated or appropriate as a substitute for potentially curative therapy, like allogeneic transplant, he said. However, oral azacitidine can be considered as a temporary maintenance as patients recover from the toxicities of intensive chemotherapy prior to a planned allo HSCT, especially if they are not fit to receive consolidation.

“I think what’s going to limit the use of oral azacitidine the most is the fact that we now have very effective regimens for older, unfit patients. When we look at the data from those trials, the initial responses are over 50% in poor-risk subsets of patients with hypomethylating agents with venetoclax. So, many patients with poor risk AML now, regardless of age or fitness, are not receiving intensive therapy,” Dr Erba said.

However, Dr Erba pointed out oral azacitidine should not be substituted for a parenteral azacitidine or decitabine in HMA/venetoclax treatment programs.

“Finally, managing toxicities with prolonged maintenance therapy is going to be critically important for the optimal treatment of AML patients following intensive therapy,” Dr Erba concluded.—Emily Bader