Magrolimab and the tale of indigestible AML Free

In this issue of Blood, Zeidner et al and Daver et al report the results of the randomized ENHANCE-2 and ENHANCE-3 trials, showing that the CD47 antibody magrolimab did not improve outcomes in patients with newly diagnosed acute myeloid leukemia (AML).^1,2 

By blocking the inhibitory transmembrane protein CD47 on leukemic cells with magrolimab, the appetite of phagocytic cells, such as macrophages, should be unleashed, resulting in an unstoppable feeding frenzy of macrophages engulfing malignant myeloid blasts.³ This mechanism was first described in AML, subsequently confirmed in cell culture and animal models and supported by clinical results from early nonrandomized studies.⁴ Based on these data and the hope for a new effective immune mechanism, clinical development took the final step to randomized comparisons.

In 3 randomized ENHANCE trials, magrolimab was used in combination with standard of care (SOC) for 3 subgroups of patients with particularly high unmet medical need. The results of the 2 trials in AML, ENHANCE-2 and ENHANCE-3, are discussed in this issue of the journal.

Based on encouraging data from early clinical trials with a median overall survival of 19 months in TP53 wild-type patients and 10 months in patients with TP53 mutations,⁴ the likelihood of a meaningful clinical benefit appeared strong, raising optimism throughout the hematology community. The innovatively designed ENHANCE-2 trial had the ambitious goal of improving survival in the notoriously difficult-to-treat TP53 mutation population and compared magrolimab-azacitidine with the azacitidine-venetoclax combination in patients ineligible for intensive treatment or to the intensive standard 7+3 in fit patients.

ENHANCE-3 targeted patients ineligible for intensive therapy, representing approximately half of all newly diagnosed AML cases.⁵ Although the introduction of venetoclax in combination with hypomethylating agents has marked a significant leap forward in this difficult-to-treat patient population, the median overall survival is still limited to 15 months and long-term remissions are rare.⁶ In ENHANCE-3, patients across all genetic subgroups were treated with the standard dose of azacitidine plus venetoclax, with half of the patients receiving magrolimab and the other half receiving a placebo.

The recently published results of both trials show that the absolute efficacy figures from previous studies were not reproduced and that magrolimab did not improve clinical outcomes compared with the respective control arms. In the TP53-mutated population of ENHANCE-2, the composite complete remission rate was 12.9% and the median overall survival was 4.4 months with magrolimab-azacitidine compared with 43.3% and 6.6 months with venetoclax-azacitidine in the less fit patient group. In patients eligible for intensive treatment, the corresponding figures were 22.2% and 7.3 months with magrolimab and azacitidine as compared with 44.0% and 11.1 months with 7+3.

Magrolimab also failed to meet expectations in the ENHANCE-3 all-comer population. Its addition to the azacitidine-venetoclax combination resulted in a numerically but not significantly shorter median overall survival of 10.7 months vs 14.1 months in the placebo arm. With similar response rates in both arms, this difference was mainly due to an increase in infectious and respiratory early deaths in the magrolimab arm, highlighting the challenge of adding novel agents to the standard combination in the frail patient population ineligible for intensive treatment. In these patients, hematopoiesis is already compromised by the doublet regimen and is further strained by the addition of a third, potentially hematotoxic agent. Unlike the increased toxicity of the triplet combination in ENHANCE-3, replacing venetoclax with magrolimab did not affect the tolerability in unfit patients in ENHANCE-2.

What is the takeaway from the ENHANCE trials, and why, despite the unsatisfactory results, are they still milestones in the evolution of AML treatment?

First, both trials are among the largest prospective studies in AML and provide valuable reference data. The ENHANCE-2 results provide robust response and survival benchmarks for TP53-muated AML from a large, international, concurrently treated patient population. These data can also help guide prognostication in routine clinical practice, as well as serve as a reference point for new and hopefully continued improvements in clinical drug development. ENHANCE-3 is one of the largest prospective data sets of first-line AML treatment in patients ineligible for intensive treatment.

Second, the results clearly show that targeting of CD47 alone was not powerful enough to significantly change the course of disease, as evidenced by the response rates and survival not only in the 2 AML ENHANCE trials 2 and 3, but also by the results of the first ENHANCE randomized controlled trial in high-risk myelodysplastic syndrome.⁷ In addition to the general challenges of immunotherapy in AML, including antigen sink from CD47 expression on erythrocytes, a pro-inflammatory tumor microenvironment, and immune evasion through CD47 upregulation,⁸ macrophage activation on its own may not be enough to generate the durable adaptive immune response needed for sustained tumor control. Potential tolerability issues may also have shortened the exposure time for magrolimab, which may have further reduced its antileukemic effect and may have outweighed its potential benefits.

Third, the results of the studies underscore the preliminary and uncertain nature of results from small, oligocentric early clinical trials and the fact that beneficial effects, although biologically and mechanistically plausible, may still not be able to profoundly alter the course of disease. They highlight the value of randomized studies and caution against designating new SOCs based on nonrandomized trial results.

Finally, ENHANCE-3 is, to our knowledge, the first completed and published randomized trial to evaluate a triplet regimen. It highlights the importance of careful dose-finding when introducing a third drug to a SOC doublet in patients who are ineligible for intensive chemotherapy, as the risk of toxicity potentially outweighing the benefits is considerable.

As AML research progresses, it appears that macrophages’ teeth have not been sharp enough to chew up AML. The ENHANCE trials may guide us toward exploring alternative or combined targets and immune strategies, while emphasizing the importance of evidence-based medicine, randomization, and careful dose assessment when integrating novel agents with SOC therapies.

By directing our efforts toward one or multiple novel antigens and boosting the antigen-specific armamentarium, we can bridge the therapeutic gaps left by conventional and other targeted treatments, bare the teeth to AML, and ultimately improve our patients’ overall prognosis.

Conflict-of-interest disclosure: C.R. has received honoraria from AbbVie, Astellas, Bristol Myers Squibb, Daiichi Sankyo, Jazz, Janssen, Novartis, Otsuka, Pfizer, and Servier, as well as institutional research funding from AbbVie, Astellas, Novartis, and Pfizer.