Hi, I’m Charlie Craddock from Birmingham, UK, and I’m with my great colleague and friend Chris Hourigan from NIH in the United States. And we’re at the ISAL AML meeting. So Chris, a lot of the talk at this meeting is about the emerging and increasingly important role of transplant in the management of fit adults with AML. And there’s a lot of thought about how we improve transplant outcomes. And your group has just had a really important paper published in JAMA. I think it was last week, around pre-transplant, MRD and transplant. And I wondered if you could discuss the findings with us.

Yes, absolutely. Thank you. So, the study we published almost two weeks ago now in JAMA was the pre-MEASURE study. And the pre-MEASURE study was really focusing on this idea of, can we actually use MRD as a test in clinical practice? What are we currently doing in terms of assessments prior to transplant and what’s the opportunity to do even better in risk stratifying patients? My hope is this becomes something that we just don’t talk about at meetings like this, but something in your routine practice. When you’re seeing patients in clinic, it’s an actionable test. You can order and you can decide what’s the most appropriate transplant for the person in front of you rather than for wide populations of patients.

So, Chris just kind of summarize some of those findings and perhaps in particular molecular subgroups.

Yeah, absolutely. So we’ve known about MRD as you know, for 40 years now. Um and one of the, one of the particular opportunities that seemed to me was to look at now, we have such great profiling at diagnosis for molecular subgroups to use those same markers to track the depth of response prior to a transplant. And so we focused initially on the two most common mutations in AML: FLT3 and NPM1, and actually showed for six patients who are presenting for a transplant in first remission, for a first transplant, perhaps one in six of those still has detectable variants which are associated with a very highly increased risk of relapse and worse survival. And it’s a simple blood test, a single blood test prior to transplant can really help you separate what on average is a 30% risk into five people with a 20% risk and that one person with a 70% risk. And we think that kind of stratification could be useful in the kind of trials you’re doing.

So, just a bit on the technology there, Chris, because we’ve been quite comfortable and we’re using well established qPCR technology for NPM1 quantitation. But your lab and I think a German lab have really pioneered a much more sensitive approach for FLT3 patients. Just discuss the technology with us a bit.

Absolutely. So there’s been great work in doing individual molecular tests using PCR and the ELN guidance, as you know says for core binding factor leukemia or for NPM1, you should use a validated test, which at the moment is PCR. We’re really looking forward to getting these kind of tests in the hands of everyone, though. I think these tests are only useful if they can be run at every place that’s looking after these patients and next generation sequencing gives this huge opportunity where many pathologists are now comfortable running these kinds of tests. And so can we translate some of that, the findings from those specific favorable risk subgroups into the kind of patients who get a transplant? And so that’s why we focused on FLT3-ITD. As you said, there’s actually three papers concurrently, the British and Australian collaboration, the HOVON-SAKK and ourselves in the US. So we had 600 patients in our study with FLT3-ITD. There’s about 100 in each of those other cohorts all using different methodologies using next generation sequencing, all showing the same effect, that there is a clear risk stratification of those who have variants detectible at a 1 in 10,000 level to those who don’t.

So that’s actually quite a breakthrough in patients with FLT3 mutations. So, what would you be advising then, what should be the technology going forward in terms of MRD quantitation, pre-transplant in patients going to an allograft, some of whom don’t have such easily detectable markers?

Yeah, absolutely. So we don’t want to do a top down approach. And the approach my groups used is generate the evidence, so that it convinces clinicians at the bedside that this has value for them. If this is marginal changes, and you can’t really confidently say to the person in front of you, this is a good and important test for you to run, then it’s not convincing enough evidence. So we’ve really started with NPM1 and FLT3-ITD. We think we’ve made the case for both those detection prior to transplant. We’ll continue to iterate as we go through and add other mutations. There may be a point where we have a wide panel and a transplant panel you can send prior to someone going to transplant, but we didn’t want to start from that point. We wanted to start with very robust tests that we felt very confident about and then add to that as we go along.

So I guess there’s a bunch of patients, aren’t there, who are coming to transplant now, who are NPM1-positive, FLT3-positive. And our group has, you know, Sylvie Freeman really has pioneered with great skill, I think, the use of imminent phenotypic strategies for detection of MRD. Do you feel those still have a role?

Absolutely. And if every center had a Sylvie Freeman, this would be a moot point, right? And I think that’s part of the issue, just the heterogeneity of the testing. The ELN’s been working hard to try and harmonize those tests between centers. But there’s, there’s a real skill to doing flow cytometry, what we do with molecular is quite straightforward. I could train a college graduate within a few months to run the test we do. And that’s the hope, is the pathologists who are very comfortable with molecular testing now could adopt this to a portfolio test they do. They don’t need a Sylvie Freeman next door, as you have, to run these.

So that’s important, Chris. But also the sensitivity that you’re achieving, I think is great. I think even Sylvie would admit it’s a bit greater than you get with flow.

Yeah. And I think there’s certainly going to be, we found in the particular case of NPM-1 and FLT3-ITD, there probably wasn’t a need to do both those tests. I think for other subtypes and other treatment modalities, we don’t have any data in our cohort, AZA-VEN treated patients, you know, we don’t have complex karyotype. There may be other groups of patients where actually flow is going to be the best test and we really want to come in open minded rather than being advocates for one particular technology.

I think you’re saying that the prognostic impact of MRD detection is to some degree context-dependent, but it’s also in the pre-transplant setting dependent on disease biology. Is that right?

Absolutely. So are there some subgroups, molecular subgroups, Chris where a degree of MRD-positivity pre-transplant may not matter quite so much?

Yeah. So I think about the p53 mutated group where actually I’m not sure, we’ll test that formally, but I’m not sure the MRD status is going to be that helpful in really, which is typically a group where we need better therapies. And so, I think it is going to mean different things in different situations, different context of use. We were very specific in the pre-MEASURE study to narrow down or what could be seen as a narrow warring question rather than the universal truth, blood, first remission, first transplant for those particular molecular groups. Because I think it is going to change over time and it’s going to mean something different if you have that after transplant, if you have that in an older person versus a younger person, different therapy and would use it to build up that evidence base every time.

So I guess we’re going to come on to the challenging question about what are we doing with these folks? But I suppose related to the other question I have around your JAMA paper, is you talk about the prognostic significance in these defined molecular subgroups in terms of cumulative incidence of relapse. But does it impact kinetics of relapse, Chris?

Yeah, that’s a good question, it’s something we’ve looked at. What we would have loved to have done is looked at the degree of disease burden measured by MRD, and then how fast or late you relapse, right? Just even with a study like this, which is several-fold bigger than other studies that have been done of this kind, we just don’t have the data to answer that question. We certainly saw cases where it was an MRD, it was D going into transplant, very high level MRD where there was relapse in the first three months. And so those cases, which were probably misdiagnosed remissions. But, you know, I think transplant is such a individualized and such a heterogeneous therapy, it’s very hard to isolate the effect of MRD with such different treatment, you know, transplant works and it works to different ways as you know, with different kinds of transplant. So we could we capture that real-world heterogeneity in this study.

So, I mean, I think we’ve captured enough content already around this really important day that you published. But just in the last couple of minutes, I suppose you and I spend a lot of time, don’t we, thinking about, okay, there is a high risk of relapse. Well, what are we going to do about it?

Absolutely. And I would say, you know, the value of all this is predicated on people collecting detailed clinical annotation, on carefully conducted studies. And so the value we saw, one of the values we saw in pre-MEASURE was how great it was to have previously been able to analyze a randomized controlled trial. And I think it’s so easy to fool ourselves. That’s why I think the work you’re doing is so important with these randomized questions around the peritransplant period because it’s so easy to fool ourselves with the wide range of variables we have going in here to get a signal or an anecdote and to run with that. Our patients deserve randomized evidence. You know, we really want to continue to build on that.

We’re fortunate because we’re working quite closely together, certainly sharing thoughts. And I think one of the fundamental points, Chris before we come on about what you do about. It is just it is important for our scientific colleagues to remember that clinical trials networks are very important to generate the robust, carefully curated matching sample and data sets for them to do pivotal results. And you and I are aware, we won’t name them, but there are a number of papers in the leukemia setting and in the transplant setting that have just been retrospective have changed practice, but actually it turned out to be highly misleading.

Yeah, so I think as you say, it’s really important that we have data sets that reflect the practice of all of us practicing. And I think we want the evidence to be generalizable. And I think the kind of networks of cooperative groups of clinical trials really help answer those questions. There’s a bias obviously of who gets access to a clinical trial and the availability of network, but it takes us a step closer to that patient you see in front of us. It’s not precision medicine if you can’t apply it to the person you’ve seen in clinic. And so I think that’s…

So my thinking about what we do in patients who are pre-transplant MRD-positive, has been very heavily influenced by another of your papers, which is in JCO and that came out of the US CTN 0901, where to cut the long story short, if I’m thinking about it with a patient sitting there in clinic and they’re MRD positive pre-transplant, I think, well, could I do something about that or could I do something about the condition regimen or could I try and optimize the graft-versus-leukemia effect?

Where I’ve landed is that at the moment, it’s actually quite difficult to do anything in terms of pre-transplant MRD-positivity. There’s no data that doing something will necessarily improve outcome. It could be dangerous and prevent people going to transplant. We don’t go there, we go straight to transplant, your data in 0901 to my mind, you can correct me, but essentially it said, if you’re fit enough to have a myeloablative transplant, your MRD pre-transplant, you should obviously do that. And at the moment, actually, if the predicted TRM of the myeloablative transplant is acceptable, then why not do that anyway? But of course, there’s a bunch of people who we have to do a RIC on and it’s about 60% or 70% of patients. And so, that’s where I think the issue is. And I suppose we’ve got some options, Chris, do you want to just kind of…

You’re absolutely right. And you know, that study was for younger patients who could tolerate either intensity. So, it’s an artificial setting. It doesn’t reflect all the practice. But I think it answers that question. I think since then, there’s been a registry study where the signal’s less clean. But again, it’s younger patients, as we know, average age of AML is 68 years of age at first diagnosis. And so that’s the key unmet challenge is what we do for patients who can tolerate a non-myeloablative or reduced intensity transplant, have MRD-positivity, have had incredibly high risk of relapse. We need adjunct to that therapy. And what that is, whether that’s augmenting graft-versus-leukemia effect, whether that’s donor selection, whatever that’s going to be, that to me is the next key challenge.

And those are the patients with unmet need.

So my sense of it is in this older population, which is many of our patients, I, although we’re doing the randomized studies, I think it’s unlikely that changing the conditioning regimen much is going to do a lot. Okay. I think you tweak a bit, you may get a bit more… a bit less relapse. But I don’t see the breakthrough. We had some data in the FIGARO study, which was a large randomized study in this older population which interestingly showed no difference between the intensified FLAMSA-Bu regimen and a standard RIC, which showed that patients who acquired full donor T-cell chimerism at day 100 – subset analysis, not enormous sums of patients, but those patients, they abrogated the adverse impact of pre-transplant MRD.

So that’s where I’m focusing in the first three months, actually, Chris. Can we do something about tapering immunosuppression? Can we think about introduction of early cellular therapies, DLI, possibly that’s challenging because of GvHD, but alternative cellular therapies, or could we think about the introduction of drug therapies early? So, we’ll have to see what plays out with the MORPHO data. I still think that maybe a subgroup who actually benefit from gilteritinib maintenance.

And of course, we have two studies with sorafenib that are positive. I think there is room, isn’t there, for more maintenance studies? I think that’s an important area but also other strategies that switch patients, to full donor chimerism. Is that fair?

Absolutely. It’s the reason we do an allo-transplant, is to have an allo effect. And the fact we don’t have predictive biomarkers and we don’t have strategies to augment a suboptimal alloresponse is what we need to go, where we need to go next. You know, the MORPHO study is a great example. Randomized trials are so important, even if they’re negative, they teach us a lot. And so, the value of doing randomized trials in transplant, I think just can’t be overstated.

So we started in agreement. We finished in agreement.

Unusual. You caught it on camera.