iwAL 2023 | TP53 mutation in AML

Hello, my name is Andrew Wei from the Peter MacCallum Cancer Centre in Melbourne, Australia. And today I have with me Dr Sanam Loghavi from MD Anderson Cancer Center. And she gave a really interesting talk today on, P53 in AML. Sanam, did you want to tell us more about what you discussed today?

Sure. Very good to be here and happy to discuss this, this perplexing topic. You know, we spoke about the significance of TP53 mutations in AML and specifically with regard to classification and prognostication. As you know, the presence of TP53 mutations in AML right now, at least in the ICC and ELN classifications is a disease-defining event. And so, you know, we know from therapies that we have currently available that this class of disease is incredibly difficult to treat. And there are a lot of subtleties and nuances in the way we identify TP53 mutations and characterize TP53 mutations and loss that could potentially impact prognostication. And I think we’re still trying to learn, what is really the best way of prognosticating and maybe pulling out the subset of disease that is not necessarily as bad as the other ones.

You know, there are studies that show that, you know, obviously having a complex karyotype in addition to a TP53 mutation that leads to loss of both copies of TP53 is just excessively detrimental. But I think, again, there’s a lot of information that we’re starting to acquire as we’re looking more into the details, including variant allele frequency, the allelic state, other chromosomal abnormalities, chromothripsis, telomere length. So these are things that I think we should be considering in our future studies when we try and prognosticate, you know, develop prognostication models for these patients.

Thanks. And Sanam, can you tell us a bit about this concept of double-hit P53. In practice, how do we define that?

So, you know, this is very interesting because I think if you think about this logically, it makes perfect sense that loss of both copies of TP53 would be much, much more detrimental because, you know, maybe going back to high school or college biology, TP53 is a tumor suppressor. When you think about malignancies, typically the way tumor suppressors, you know, invoke malignancy or are involved in the pathogenesis, is you usually have loss of both functional copies. Whereas oncogenes, like, let’s say for example, MYC, you can have one activated copy of MYC and, you know, it’ll do the magic, it’ll do the job. But for TP53, you know, for the longest time we were just looking at TP53 mutations, the presence or absence of TP53 mutations in myeloid malignancies until the landmark study that was published, by Elsa Bernard and Elli Papaemmanuil in 2021, I believe,  where they showed that in myelodysplastic syndromes, patients that had biallelic loss of TP53, and I’ll explain that in detail, were the ones that had very poor outcome. And at least in their study, patients that had just one mutation did almost similarly in terms of, you know, outcome and disease behavior to patients that had wild type TP53.

And so how do we define that in practice? You know, I’m a hematopathologist, how do I define a biallelic loss of TP53? There’s really three major scenarios. One is if you have two or more mutations. And again, realizing that, you know, the way we detect mutations right now is by bulk NGS, So we’re really not looking at individual alleles or clones, but again, by way of biology, we are imparting that when there’s two mutations, they’re probably in two alleles.

The other way is if you have mutation of one copy and deletion of the wild type copy, so that all you’re left with is the dysfunctional mutated copy.

And then the third scenario, which is a little bit more tricky, and you know, some labs don’t actually do this routinely in practice, is if you have a mutation and copy neutral loss of heterozygosity, and what that means is that the wild type allele is deleted and the mutant allele is actually duplicated. So if you FISH for TP53, you’re gonna see that there’s two copies. You’re not gonna see a loss of TP53, but functionally you don’t have a wild type TP53 copy anymore. So that is also, you know, the third scenario, and again, because I said we don’t look for that and, you know, practice at least many labs, don’t we use the variant allele frequency as a surrogate to imply that there’s copy neutral loss of heterozygosity, and, you know, a 50% variant allele frequency cut-off more or less seems to correlate with copy neutral loss of heterozygosity. So if you have a mutation with a variant allele frequency of more than 50%, you can assume that there’s probably copy neutral loss of heterozygosity.

And you mentioned that P53 mutation is a really important and poor risk entity, but that it can also overlap with other, you know, commonly occurring mutations in AML. And so if we have a patient with a P53 mutation and an, say, an NPM1 mutation  God forbid, doesn’t happen very often – how do we, you know, how do we designate that patient? Is that patient favorable or unfavorable?

So, you know, I think this is still, you know, as you said, fortunately, this does not happen very common, but again, because it doesn’t happen very common, I think our information is really limited in this scenario. But the way the risk stratification systems are designed now, at least ELN 2022, if you have an NPM1 mutation without a FLT3-ITD, obviously, that disease is considered favorable risk. Now, if you have an NPM1 mutation with an adverse-risk cytogenetics, the adverse-risk cytogenetics trump NPM1, and that patient is considered to have adverse risk disease. And we know that the majority of TP53-mutated AMLs at least for the most part, tend to have complex karyotypes. So those would be considered high-risk disease. But in terms of designation of the, you know, your top line on the report that is still by ICC and by WHO fifth edition, that is still considered an NPM1-mutated AML, because NPM1 is a class-defining alteration that is above TP53 in classification.

Conversely, I think are the myelodysplasia-related mutations, where if they occur with NPM1, at least in our current stratification- risk stratification systems, they do not alter the risk of disease. So if you have an NPM1 mutation together with an SRSF2 mutation without a FLT3-ITD, that patient is still considered favorable risk.

And briefly, when I did my talk today, I was focusing on what are the possible therapeutic options for patients with adverse-risk disease, such as P53-mutant AML? And one of the areas that I discussed was the possibility of combining BH3 mimetics, targeting BCL2 and MCL1. The difficulty, of course, is that MCL1 is an important pro-survival molecule for critical organs, such as particularly the heart. And that future therapeutic modalities might need to determine ways to reduce this risk by perhaps directing the MCL1 inhibitor more directly to its target, perhaps by either PROTAC-relevant options or antibody dependent conjugates, that can direct MCL1 inhibitors to leukemic cells.

And the second possibility as a new area of potentially leveraging STING agonists, which are currently used in solid cancer circles, to activate the immune system and utilize, I guess, a less recognized possibility that activating STING with these agonists can actually drive apoptosis and drive apoptosis in a P53 independent manner. And that by combination with venetoclax, we might have a very effective combination, that can perhaps make some inroads into a really difficult disease.

So, today we’ve discussed, obviously, the therapeutic possibilities, but also the diagnostic aspects and the complex nature of P53. And hopefully all of these together will provide some diagnostic, but also some therapeutic benefits for patients in the future. Thank you for joining us today, and we’ve hope you found this session and discussion interesting and educational.