ESH AL 2026 | MRD as a surrogate endpoint in AML: challenges, ongoing studies, and the need to standardize assays

The debate is on measurable residual disease as a proper surrogate endpoint in AML. And I’ve been assigned the affirmative end of the argument. It’s clearly a very important position to be able to take but also to prove that MRD is a proper surrogate endpoint. And the reason for that is currently our approvals of new drugs are depending on showing an overall survival benefit in people with acute myeloid leukemia. And those years, those phase three trials take years to complete, to design and complete. And by the time they’re completed, new questions have come up. And so we need to move quicker, especially in the era where we have novel therapies being developed and being brought into frontline therapy for patients with acute myeloid leukemia in order to try to improve the survival. So we need a powerful surrogate that correlates with overall survival. Now the challenge in doing this is what MRD assay are we going to use? Flow cytometry, PCR, RT-PCR, FISH, there are a variety of them. What tissue, peripheral blood or bone marrow, at what time point immediately after induction? Is that one or two cycles of induction after consolidation? And so there are a number of questions that we don’t have answered yet. And of course, these assays have not been validated and standardized internationally, including what is the most appropriate cutoff that is informative at predicting the surrogate endpoint, in this case, overall survival. What complicates the matter in large phase 3 studies is that patients are able to undergo allogeneic stem cell transplantation after they’ve achieved remission. An allotransplant through the graft-versus-leukemia effect has the greatest impact on measurable residual disease. And so it can end up negating any benefit that may have been seen by the intervention to induce an MRD-negative remission, complicating the interpretation of clinical trials. Nonetheless, I think there are situations where we can consider MRD a proper surrogate. One endpoint for my patients with acute myeloid leukemia is avoiding the toxicities, the morbidity, the mortality associated with allogeneic stem cell transplantation and first remission. And I do think we now have enough data using an MRD assay, the RT quantitative PCR assay for the NPM1 mutation, that help us decide which patients may benefit from an allotransplant in first remission and those who may not need an allotransplant in first remission. Now, of course, no assay is perfect, and patients who are MRD negative by that assay and don’t undergo a transplant may end up relapsing. However, in NPM1 mutated disease, we have a number of options. If the patient just received chemotherapy, they could receive gemtuzumab, azacitidine plus chemotherapy. They can receive venetoclax-based regimens, either with intensive or non-intensive chemotherapy. And then, of course, we have the menin inhibitors. And all of these are going to be combined in different combinations for patients with relapsed disease. So I think we’d be able to salvage those patients and then direct them to allotransplant. There’s another place where the reverse transcriptase quantitative PCR assay for NPM1 can be actually quite helpful and informative for patients. In those patients receiving less intensive therapy with a hypomethylating agent or venetoclax who have NPM1 mutated disease, we know that some of these patients may very quickly, within a year of treatment with a Ven-HMA combination, achieve a very deep remission that’s MRD negative by this assay. Some of those patients may actually be able to go on to treatment discontinuation and remain in a treatment-free remission. I have such patients in my own clinic who’ve been off therapy for three years, having achieved an MRD negative remission after just venetoclax and azacitidine. And these have been reported in case series. However, as a proper surrogate endpoint for regulatory approval, I know of only one study that has prospectively evaluated a specific MRD assay at specific times along the course of the treatment protocol, and that’s the QuANTUM-First study. In that study, patients who were up to and including the age of 75 with a FLT3-ITD mutation received initial therapy with standard 7 and 3, followed by high-dose cytarabine, in combination with either quizartinib or placebo during the induction, consolidation, and then as maintenance for up to 144 weeks. We know that quizartinib was associated with a survival benefit. But what we learned from that study, where we prospectively captured these samples from bone marrow specimens, is that patients who received quizartinib were more likely to become MRD negative by a very sensitive next-generation sequencing FLT3-ITD-specific assay compared to those who got placebo. And the median variant allelic frequency was also lower following induction and consolidation and prior to maintenance in patients receiving quizartinib. To me, what’s remarkable is in patients who became completely MRD negative, not only were there more of those patients in the quizartinib arm, however, in patients who were MRD negative, who did not undergo an allogeneic stem cell transplant in first remission, there was a survival benefit with quizartinib compared to placebo. Now, this is very important to me because although many of my patients should be considered for allogeneic transplant at first remission, we know that it’s only the minority in the phase three QuANTUM-First trial and in clinical practice who are actually able to undergo allogeneic transplant. You may not know this on the first day that you’re meeting the patient. And so in my opinion, this supports the use of quizartinib in the FLT3-ITD mutated patients, and the MRD data supports the general benefit of quizartinib in that patient population. So I think going forward, we need to be mindful of the need to actually develop, to validate, and to standardize these assays, which may be different across different subsets of patients with acute myeloid leukemia, and incorporate those assays into well-designed clinical trials in order to try to prove that MRD is a proper surrogate endpoint, not only in FLT3-ITD-mutated AML, but in other subpopulations of people with acute myeloid leukemia. So we’re not waiting another 10 years for the next phase three survival study to read out.

This transcript is AI-generated. While we strive for accuracy, please verify this copy with the video.