Challenges of Applying MRD Decision-Making to Determine Transplant Among Patients With AML

Naval Daver, MD, MD Anderson Cancer Center, Houston, Texas, participated in a debate at the 2025 Great Debates in Hematologic Malignancies meeting in New York, New York, in which he argued against minimal residual disease (MRD) negativity being the primary tool to recommend patients with acute myeloid leukemia (AML) to transplantation, in part due to the heterogeneity of AML across patients and the impact of underlying mutations. 

Transcript:

My name is Naval Daver. I'm a faculty in the Department of Leukemia at the MD Anderson Cancer Center in Houston, Texas. Today I was having a debate presentation with Dr Andrew Artz, my colleague and friend. The topic of debate was whether MRD should be used as a key decision maker for transplant or no transplant in patients with acute myeloid leukemia.

My assigned side was the “No” side, meaning that we should not make the transplant decision solely on MRD, whereas Dr Artz’s side was that MRD should be the key or the sole factor to make that decision. Of course, this is all in fun and we all believe that MRD will and is already playing a role in acute myeloid leukemia.

The key learning points that we both wanted to highlight is that there are challenges in applying MRD when it comes to acute myeloid leukemia compared to acute lymphoblastic leukemia or CML. The big challenge there is the heterogeneity of the disease. In AML, we see that AML is really often a constellation of multiple different clones rather than one homogeneous entity. Also, between different AMLs, the type of underlying mutation has very different impact, both on their response to therapy, their long-term outcome, their post-transplant outcome, and also whether or not they will be likely to achieve MRD negativity.

With that aspect, how do we now build on that? That is the challenge, and I think we have already made a lot of progress in the last 7 to 8 years now. We have good NGS [next-generation sequencing]  or PCR [polymerase chain reaction], high-sensory MRD assays for at least 2 of the key mutations in AML, NPM1 and FLT3. These 2 make up about 40% of AML when you take both of those together and there are new MRD assays coming for KMT2 rearranged, which is another 10%.

We think that in the very near future there will be, and already for many are available commercial assays that for at least about 50 to 55% of AML could give us a good trackable MRD marker with known biological and prognostic significance. That will be good because now there are data sets that are emerging. 

The British group has actually championed a lot of these efforts and published recently 2 or 3 high impact papers where they have shown that with both intensive chemo and low intensity therapies like azacitidine venetoclax in NPM1-mutated patients, the most critical prognostic factor was actually achievement of NPM1 PCR eradication using a 10 risk to -5 assay. This actually overcame baseline risk factors such as cytogenetics, de novo or secondary AML co-mutations. For the first time they're showing that if you had to choose 1 key decision maker, it's actually achievement of MRD after 2 cycles of intensive chemo, 3 to 4 cycles of azacitidine venetoclax.

Similar data is also now coming out in FLT3 where from the US, Chris Hourigan and Laura Dillon have now shown that achievement of FLT3 NGS-MRD negativity, again enter to -5 sensitivity was clearly associated with better survival post-transplant and without transplant. What they also showed is that we have to be careful. It is not that every mutation has that same prognostic MRD value. They showed NPM1, FLT3-ITD, FLT3-TKD have the prognostic value. If you can eradicate them, you will have a better survival with or without transplant.

But for IDH1 and 2, which is quite common, 20% of the AML have an IDH1 or 2. They showed there was no difference in survival with or without transplant if you achieve clearance of IDH. This is a much more nuanced discussion than ALL where any MRD is bad, any B lymphoblasts residual is bad, here it is dependent on the specific mutations where MRD is clearly tracking with the disease and outcome and others where it may not.

We biologically now understand that this may be because about 20%, 30% of AML [cases] arise from a preexisting underlying hematological disorder like MDS or what we call CHIP. Some of these mutations may have been those precursor existent CHIP mutations for months or years and they will not go away because they have just been there as stagnant mutation. We are learning more. As we use more and more the NGS-MRD, at least for those driver mutations with NPM1, FLT3, KMT2A, that will help us decide which patients may not need a transplant. 

Then importantly also data now showing, especially for FLT3, that even if you transplant the patients, we today transplant most sure FLT3 -ITD, MRD may help you know which patients are done. Once you're done transplant, their outcome is great, you don't need to do anything further. Just monitor them post-transplant versus those where even with transplant they have a higher risk of relapse because they were MRD positive at any low level before after transplant, then you don't want to just wait and watch you add a post-transplant drug to mitigate or eradicate that risk of MRD persistence.

We are already starting to see clinical impact of the MRD, and I think in the next 5 to 7 years, this will become wider and wider. Of course, the key thing this will help is number 1, patients who absolutely need transplant should be decided early, go to transplant, early, use maintenance, everything possible to bring that MRD negative impact down. Those who don't need transplant, I think that's equally important, maybe 40, 50% of AML we are overt transplanting because we don't have good technology and maybe we will be able to avoid transplant, which is expensive, has a lot of comorbidities, quality of life expense to the entire system in nation. I do think that we are now getting more and more delineated and intelligent, and this will continue into the future.

Source:

Daver N. Debate- Should the Recommendation for HSCT in the Treatment of Acute Myeloid Leukemia be Determined Predominantly by MRD? Yes or No. Presented at the Great Debates in Hematologic Malignancies meeting. June 28-29, New York, New York.