iwAL 2025 | Refining the diagnosis and classification of CMML

Sanam Loghavi:

Hi, everybody. I’m Sanam Loghavi. I’m a hematopathologist and a molecular pathologist at MD Anderson Cancer Center. 

Guillermo Montalban-Bravo:

And I’m Guillermo Montalban-Bravo. I’m a physician and associate professor at the leukemia department of MD Anderson. 

Sanam Loghavi:

And we’re here today to discuss the topic of oligomonocytic CMML. This is a relatively newly defined entity. So CMML is chronic myelomonocytic leukemia. And in the most recent iteration of the hematolymphoid neoplasm classifications, both the WHO fifth edition and the ICC, they have proposed to lower the absolute monocyte threshold for CMML in order to allow us pathologists to make an earlier diagnosis of CMML. And they did this by saying that in order to be able to make this diagnosis, you need to have the persistent 10% relative monocytosis, but they lowered the absolute monocyte threshold count to 500, but then to kind of account for lowering the count, they said you require clonality. So you have to have clonality in order to prevent diagnosing reactive monocytosis as CMML. But they really didn’t specify what type of clonality. So with this in mind, we tried to look at a series of cases with Guillermo here, to see if we could refine the diagnostic criteria for CMML. So, Guillermo, tell me, what did we find? 

Guillermo Montalban-Bravo:

Thank you, Sanam. So, I mean, I think, you know, to follow up on what you were saying, one of the key things was that we wanted to try and understand, is there any specific information that we can use that could help refine it so that we ensure that we’re not inappropriately calling CMML patients who may actually not have CMML, right? Because, just to clarify, one of the major caveats that this reduction has, even if we account for clonality, is that there’s patients with other myeloid neoplasms, like MDS, for example, who might have a certain degree of monocytosis, but that may for sure not behave like CMML. So then what we wanted to do is we wanted to look at a large cohort of patients and try to characterize, based on their genetic features, can we predict who might be actually CMML? So what we did is that we looked at around 900 patients, including MDS patients, CMML patients, and oligomonocytic CMML patients, and we clustered them based on their genetic composition. And what we found is that there’s a group of patients, like approximately over a third, that have a specific signature when we merge all the diagnoses together, which is presence of bi-allelic TET2 mutations and SRSF2 comutations, majority of which are CMMLs. And then when we tried and understand how does this correlate with the disease behavior, we saw that that bi-allelic TET2, SRSF2 mutation predicts for monocytic bias, and that that is not because patients have more RAS mutations, and that these are the patients that if we looked only at the ones with oligomonocytic CMML, they actually evolved to CMML. While if we account for other patients without those features, it’s much less likely that it happens unless that there’s a… we know, factually, there’s an increased monocyte population in the bone marrow. So if we account for the genetic of the patients and then how many monocytes they may have in bone marrow, either looked by morphology or flow cytometry, we can quite reliably predict who is going to become CMML. So we feel that, you know, and I feel that, incorporating this into the kind of diagnosis framework that pathologists use probably could help us. But I’m interested to know what your opinion is on this as well. 

Sanam Loghavi:

Yeah, for sure. Because I think, you know, again, the refined current diagnostic criteria, I think, was done with good intention because there were studies that had showed that true oligomonocytic CMML behaves like real CMML, right? And we know that. But the problem, you know, we kind of found this out in practice when the diagnostic criteria came out. You know, I was looking at cases that are clearly MDS, like let’s say an MDS with biallelic TP53 loss, but it has borderline monocytosis. So, you know, you know that this is MDS, of course, but it kind of fits the criteria for the other one too. And nowhere in the diagnostic criteria does it say that one trumps the other, right? So, and I think, you know, TP53 is an exaggerated example of this, but you could see this in other low-grade MDS where they have borderline monocytosis, but that’s not really CMML. So I think, you know, I encounter this not infrequently in my practice, and I really want to be able to make an accurate diagnosis of CMML. But you tell me, so as the treating physician, does it matter if I call something MDS versus CMML? 

Guillermo Montalban-Bravo:

I think it matters at different levels. So first, it matters because, I think one of the most important is, because I assume as a patient you want to know accurately what you have, right? So I feel that the understanding that, oh, well, they’re like similar disorders is probably not the most accurate way to think about what patients deserve. But also from a manner of perspectives in terms of like clinical trial accessibility that may be disease-specific or obtaining access to certain drugs that may be current or maybe future for a selective diagnosis on another. Reimbursement for certain testing or transplant accessibility. So there’s a whole array of downstream implications that having an accurate diagnosis carries that I think is sufficient to want to be very careful in assigning an appropriate diagnosis to a patient. 

Sanam Loghavi:

I could not agree more with you. I think I asked the question knowing the answer. But of course, no, really. And I think even for us, we want to try it be accurate. And I think I’ve used this analogy with you before, you know, back in the day when we didn’t have rituximab, it didn’t matter if the patients had B-cell lymphoma or T-cell lymphoma, right? Everybody got CHOP. So, and, you know, people argue and say, oh, what does it matter? You know, they’re going to get HMA anyway. But I think just because we don’t know, you know, or we don’t have specific therapies right now, it doesn’t mean that we’re not going to develop specific therapies in the future. And also, if we’re lumping everything together, we’re never going to develop good therapies because, you know, we’re not dissecting the disease based on its biology. 

Guillermo Montalban-Bravo:

Totally agree. 

Sanam Loghavi:

So I really think it does matter that we make an accurate diagnosis. 

Guillermo Montalban-Bravo:

Fully agree. Yeah. 

Sanam Loghavi:

Yeah. All right.

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