Recent updates on the management of acute myeloid leukemia with targetable mutations

DR BRUNNER: Well thanks for having me today. My name is Andy Brunner, and I’m a medical oncologist here Massachusetts General Hospital, and I’m happy to go over some of the new developments in AML and MDS that were presented at the ASH Annual Meeting in December 2021.

And so first, one of the things I was hoping to discuss were how abstracts in AML reflect a really rapidly-changing practice environment and have created a much more targeted and nuanced approach to leukemia in general. And this is a slide. It’s meant to be complicated because the management of AML in 2022 has become a complicated affair. We increasingly have identified therapies that we may use depending on the patient’s mutational profile or depending on their fitness, or a combination of both. And I think that even now these therapies are evolving.

I guess it’s a blessing that we have a number of therapies that are very active in AML, and for any patient I meet in clinic I can now choose things both in the front-line setting and if they have relapsed or progressive disease. I think a challenge, though, is understanding what sequence we should use these in, when is the most appropriate time to use a targeted therapy, for instance, how can we optimize therapy to prolong life and quality of life for our patients.

So there are a number of mutations that increasingly play a role that will be highlighted through later abstracts, including mutations in p53, IDH1, IDH2, FLT3 among others, including cytogenetic risk. And so this is just to give a sense of how AML has really become a complicated space to some degree, but in many ways that means we’ve got a lot of new options for patients.

So one of the areas that AML has changed the most recently has been the discovery that combining azacitidine or decitabine, the so-called hypomethylating agents or DNMTi therapies, combining them with venetoclax, which is a BCL2 memetic that results in synergy when added to azacitidine alone. And this study from 2020 showed survival benefit compared to azacitidine monotherapy in AML.

At ASH 2022 we start asking the next questions. How does azacitidine/venetoclax compare to other regimens? Are all patients who were treated on this study, do they benefit the same, or should we think about other therapies for certain groups? And what about patients who weren’t included in this study? What about our younger patients or otherwise fit patients? How can we also improve upon this combination? Is there some triplet, for instance, that may give us even better results?

And one of the most challenging spaces in AML has been p53-mutated disease. It’s a challenging space across oncology, but particularly in AML patients with p53-mutated disease, they can have poor responses, and they often don’t respond to traditional cytotoxic therapies.

Here’s a study that was looking at the patients who were treated with azacitidine/venetoclax who had p53 mutations versus those who did not, according to their cytogenetic risk. So we know that patients with poor-risk cytogenetics have challenging outcomes. We know that patients with p53 mutations have challenging outcomes. Looking at this, they wanted to see well, how do the patients with the worst disease do with venetoclax and azacitidine compared to azacitidine. And in this study, they showed that even though we would like those patients to have better responses, the overall response rate was still 40% in this high-risk group of p53-mutated, poor-risk cytogenetic patients with venetoclax and azacitidine, which was higher than the azacitidine arm, suggesting that even in these poor-risk patients there is benefit to the combination therapy.

If we look at patient survival, we see that unfortunately these patients do more poorly than others, and those patients that do the worst are really the patients who have p53 mutations, unfortunately kind of regardless of venetoclax and azacitidine or azacitidine. So we can improve the response rate, but our ability to improve survival is still lacking in this group and makes it a key group for further study.

If we look at other venetoclax combinations, shown here is a study combining venetoclax with decitabine, another hypomethylating agent, and administering decitabine as per standard dosing and adding venetoclax. I think one of the challenges is what do you do when you have a patient who might be eligible for either induction or for decitabine/venetoclax, and this study tried to compare the decitabine/venetoclax combination, tried to compare it to historic parallel enrollment group of patients getting induction with standard “7 + 3” with idarubicin and cytarabine.

In that study they had 60 patients with 7 + 3 and 14 patients with venetoclax/decitabine, and they showed a similar complete remission rate, but more patients achieved CRh with the venetoclax/decitabine arm as a percentage of patients. Granted, it’s a small number, only 25 patients in the combination arm, but they also suggested that at least the survival and progression-free survival were not that much worse, and if anything, a trend towards being better in the venetoclax/decitabine arm compared to induction arm.

So I think this study suggests we should consider prospectively comparing HMA/venetoclax regimens to other standard regimens to see other patients who really would benefit more from this somewhat alternative regimen in the future, including in young adults.

Another question that’s being asked is well if venetoclax works in azacitidine or decitabine combinations, can we add it to standard induction? So here’s a study with FLAG-IDA as induction used. This is out of MD Anderson and showing that they could combine venetoclax with this regimen. It does add myelosuppression, and so there were strategies, including GCSF, to try to mitigate that. But the survival curve for patients treated on this, with relatively early follow up, was reasonably encouraging, and perhaps we should be exploring how venetoclax can be synergistic with other induction regimens as well.

Of course the number of agents that we use in the up-front setting has expanded quite a bit. And one of the indications that we think about for AML is for patients who have had prior malignancy or secondary AML, do they have an optimal regimen? And so many of those patients will be older. They will fit into the group of patients treated with hypomethylating agent venetoclax.

At the same time, recently liposomal daunorubicin/cytarabine, or CPX-351, was approved due to its superiority compared to 7 + 3 induction for patients with secondary AML who are 60 to 75 years of age. And so a question that has arose in the field is when I have a 67-year-old who has secondary leukemia, what is the preferred regimen of patients.

And so this was a retrospective study. It looked at 4 large academic centers, and it found patients treated over a similar period with either CPX-351 or with HMA/venetoclax. It found that in the CPX-351 arm there seemed to be a higher remission rate, so the CR rate was 98 patients out of 211, compared to 62 out of 226 in the HMA/ven arm. However, the CRi, so complete remission with incomplete count recovery, was higher in the HMA/ven arm compared to CPX-351. There did seem to be some differences, as well, in overall and relapse-free survival favoring CPX-351. And here is shown some Kaplan-Meier curves suggesting that there was more of an overall population benefit to the CPX-351.

Of course, this is a retrospective study, and so it’s hard to infer intent — treatment intent, also patient selection, things that just can’t be captured outside of a prospective analysis. But it does set the stage for what may be to come, a prospective analysis comparing HMA/venetoclax options to other standard induction regimens.

I think another way of looking at the same question was also presented; CPX-351 compared to venetoclax and azacitidine, again trying to get us to this question you’ve got a lot of patients who would meet criteria for either trial, which is the better route to take. And so in this study this was looking at data out of Penn, as well as out of Flatiron Health record, and it found 219 CPX-351-treated patients and 440 aza/ven-treated patients. And really in this group they did not find a significant survival benefit, 13 months in CPX-351, 11 months in azacitidine/venetoclax, with a higher rate of F & N in the CPX-351 arm. So again, setting the stage for what may be a potential head-to-head comparison of these 2 regimens, given the similarity of the patient populations that received them.

One of the other emerging spaces in AML is really how targeted therapies play a role. Gilteritinib is a targeted tyrosine kinase inhibitor that targets the FLT3-ITD and -TKD mutations that are present in AML. It is approved in the relapsed/refractory setting in the United States. And here was a study in Asia, similarly designed, comparing gilteritinib monotherapy head-to-head with either investigator’s choice of low-dose cytarabine, MEC, or FLAG as a treatment for relapsed/refractory FLT3-ITD or -TKD-mutated AML.

Notably, the comparison arm did not have a FLT3 inhibitor, and there is some availability issues around FLT3 inhibitors outside of the United States. But in this study, similar to the study that got approval of this drug in the United States, we see that there is a survival benefit to gilteritinib compared to other standard chemotherapies.

How it is used is — in the front-line setting is an increasingly important question, and I think that in contrast to the prior, where arguably FLT3 inhibitors are less available outside the United States, one of the updates here from a study that is highly anticipated, was the comparison of gilteritinib with azacitidine to azacitidine monotherapy. And in this study patients, in an open-label study, were randomized to either the combination of gilteritinib/azacitidine versus azacitidine alone for FLT3-mutated AML. The study found that there was no survival benefit to the combination. The survival curves were essentially overlapping.

I think there are some interesting takeaways from this study, however. First, we see that as far as how many of the patients on the gilteritinib/azacitidine arm went on to any subsequent therapy, only about 20% of those patients got a next line of therapy. Whereas patients who were on azacitidine alone, nearly half of them, of the 49 patients, went on to a second line of therapy. And of those patients, notably who did not get a FLT3 inhibitor front line, many of them did end up getting a FLT3 inhibitor as a second line of therapy.

Moreover, if we look at how the response rates were between the arms, we see that when we add gilteritinib to azacitidine, while the CR rates were not that different between arms, the rate of CRh was significantly higher. So you are getting some different responses that integrate both blast reduction and blood count improvement in the combination that may be meaningful for select patients.

So while we are still learning how we might use FLT3 inhibitors in combination with azacitidine, I think questioning really when is our sequence of FLT3 inhibitors, when does that play the main role. At the same time, another mutation in AML, IDH1 and IDH2, have been targetable with drugs ivosidenib for IDH1 and enasidenib for IDH2. The AGILE study shown here was presented. It was a Phase III study that combined ivosidenib and azacitidine compared to placebo and azacitidine in newly diagnosed IDH1-mutated AML.

And similar to the prior studies, it showed a higher remission rate of the combination, so a response rate when you add azacitidine to ivosidenib was higher, as well as the CR plus CRh rate. Interestingly, there were actually fewer infections in the azacitidine/ivosidenib arm, suggesting perhaps that does mitigate infectious risk. And quality of life also favored the combination.

And so here we see the survival curves, and really showing that by adding ivosidenib to azacitidine up front you have an improvement in survival compared to azacitidine monotherapy. The question unanswered is A) how does this compare, for instance, to azacitidine/venetoclax, which was not the comparator arm, but many would argue is the standard of care arm now for many patients with AML. And also, if these patients access to ivosidenib at progression, would that change some of the responses and survival that were seen here instead of just staying on azacitidine monotherapy?

All of this also is leading the way to questioning where will we see the use of triplet therapy in AML. So we know that 7 + 3, with or without gemtuzumab, there is potential benefit. We’ve also seen, we discussed with secondary AML CPX-351 compared to 7 + 3, a benefit. And now in “unfit” patients azacitidine versus azacitidine and venetoclax. So now that we have a number of new standards in this field, is there a way that we should be combining even additional agents to AML treatment?

So one study looked at the addition of azacitidine/venetoclax, so now our standard of care backbone, adding gilteritinib for patients who have FLT3-mutated AML. And this is a small study so far but suggested that you could look at this triplet. You do need to do a lot of monitoring. They did a bone marrow at day 14 to assess for aplasia. And there were a number of patients in the front-line setting who could achieve a remission, but also I think, excitingly, that a number of the patients with relapsed/refractory setting were able to achieve remission. Now numbers are very small at this point, but I think we’re increasingly seeing this explored.

Similarly, quizartinib, which is an investigational FLT3 inhibitor, was combined with decitabine and venetoclax, and it also showed a number of responses, both in the front line and relapsed/refractory setting. I will point out, however, that the rate of morphologic leukemia-free state, so meaning low blasts and low blood counts, was relatively high in the relapsed/refractory setting. So as we get further and further in treatment, using these triplets may become challenging on an already exhausted marrow.

Another agent in development is magrolimab. It’s an anti-CD47 antibody. And this study looked at combining azacitidine/venetoclax with magrolimab, and so again getting into this idea that will triplets be the way that we explore things further. Interestingly, this divided patients according to p53 mutation, as well as according to whether they received prior venetoclax.

So for p53-mutated disease we did see a number of responses. It’s a small cohort, and so we’ll need to see if those patterns continue to hold, and if we really have a more active agent in p53 disease.

I think this study was also important, though, because it has a prior venetoclax arm. And so really when we are thinking about a number of agents being approved in AML, we have to think about sequencing those agents. And in the prior venetoclax arm we did not see any CRs out of the 15 patients. We had a few CRis, but maybe this would suggest that magrolimab, if it’s got a role in disease, might need to go earlier in this space than venetoclax, especially with a triplet.

Shown here are some of the durations of response. And so again showing that post ven is much more challenging space, and that may become the next area that we’re really exploring in AML.

And so to conclude about AML in ASH 2022, we have a lot of active agents in the front line. We don’t have a lot of direct comparisons, so I don’t know which one to necessarily pick patient to patient, and I think that those are going to be the questions we need to ask in the next years.

Sequencing these agents may also be really important. If I have a potent targeted therapy, is it best to use it up front, or is it better to wait to use it once somebody has not responded to the first line?

And triplets are being explored. They can be administered. I think it’s a challenge to know whether triplet therapy is going to improve over just using therapy in a sequential nature.

Recent developments in the management of myelodysplastic syndromes (MDS)

DR BRUNNER: From MDS we also had a number of developments, and I’ll try to go over some of the bigger themes of the MDS abstracts today.

In MDS, I think one of the biggest challenges that we’ve had is how to define risk because you want to treat patients based on what the likely natural course of their disease is. Some patients are going to live with their MDS for a long time, but some patients have more active disease that may behave more like leukemia, and really we need to intervene sooner to change the natural course of disease.

And one of the biggest areas we’ve learned how to risk stratify, or understand disease, is based on the introduction of molecular classification into MDS. And so the long-awaited Molecular International Prognostic Scoring System was presented at ASH this year. It involves a number of genes, 156-gene panel, added to traditional cytogenetics, blood counts, and blast counts, to assess patients into a high-risk group and a low-risk group and really now 6 risk groups.

It was validated against a cohort from Japan, and it showed a number of findings that we’ve long suspected from other molecular studies, so p53 multi-hit mutations, those do the worst and really are in need of new therapies. And the mutations that are AML like, so MLL-PTD, or FLT3 mutations occurring in a small number of patients, but really having a poor prognosis and maybe we should be thinking about them more like AML. A more favorable clinical course confirmed that SF3B1 mutations, depending on their comutations, really can have a more favorable course.

So this is a bit more complicated metric. It does suggest that we need to ramp up molecular testing in MDS and make sure that each patient’s getting that at diagnosis, but this probably is the future of how we’re going to assess patients to really determine who is eligible for a higher-risk protocol and who can be monitored and treated with symptomatic therapy.

DR LOVE: In terms of the classification system, but when you move from the current system to this one, kind of what happens? Do people get usually more high risk, or what happens?

DR BRUNNER: Yeah. That’s a great question. I think that we’ll see in practice, in general many patients will stay, for instance at the very high risk and the very low risk, those patients maintain a very similar risk. I think the surprise was how many people in the middle were reclassified.

And I’m not sure how easy it is to interpret this little chart here, but you can see a sideways stacked bar chart showing that on the top is the IPSS-M, and the side is the traditional IPSS-R. And so the red group is the very high-risk group in the IPSS-M, and really purple, orange, and red are the new high-risk patients. And blue, green, and light green are the low-risk patients.

And what you see is, yeah, in the very low IPSS-R, largely you would have guessed that they’re either going to be very low or low, but there are a few patients who have higher-risk mutations that now would really move up to high risk. Perhaps those would be those patients in the TP53, MLL, or FLT3 group.

Similarly, in the very high risk, those patients largely were very high by IPPS-M and by IPSS-R. But in the middle, you see a large reshuffling. So especially for intermediate-risk patients, which are always challenging in the IPSS-R to know who’s going to have a more benign course and who’s going to have a more active or aggressive disease, here you can really see that those patients are divided significantly according to their mutation profile.

So I think that it will give us a tool in addition to the IPSS-R and IPSS, from which we still base all of our — azacitidine and decitabine data was enrollment based on IPSS. But I think that it will give us a little bit more nuance in being able to understand what a natural course of a patient is and how can we intervene. I think it will need to be — it was validated in the Japanese cohort, but we will — I anticipate people will be validating this risk similarly in other patient cohorts as well. And ideally it would be something that we’re prospectively looking at both risk scores when we are finishing clinical trials to assess how they differ.

DR LOVE: Is it available?

DR BRUNNER: Not yet. It’ll have to be electronic. It’s going to be such a complicated input system.

DR LOVE: And the — this assay, is this like a commercial assay that they developed for this?

DR BRUNNER: So the lab did this assay, but I believe it’s akin to most standard NGS sequencing panels. So 156 genes is a little on the large side. For instance many myeloid sequencing panels are closer to — commercially available are closer to 35 to 50 genes. But the main genes that are seen are typically covered, certainly p53, FLT3, SF3B1. MLL-PTD has historically been more of an AML gene, and it may be something that needs to be checked from assay to assay. It also sometimes requires —

DR LOVE: If an oncologist wants to use the system how do they do it?

DR BRUNNER: Yeah. So I think that you would be able to do most standard gene sequencing panels for the initial molecular testing.

DR LOVE: And then just go find the 159 genes?

DR BRUNNER: I don’t think you have to input them all. But my understanding is you would input the ones that came back. So say your sequencing came back with 3 or 5 gene mutations, you would mark yes/no, is what they ended up doing. Now there are different ways where you could look at point mutations or you could look at the variant allelic fraction of the mutation, and they ended up not using that. So I think it’s just a yes/no that they ended up using in the final analysis.

But you would go into your sequencing report, you’d see the 3 or 5 genes, you’d check yes on those.

DR LOVE: Right.

DR BRUNNER: And put in the cytogenetics. And so from a user standpoint it shouldn’t be too bad because they validated it kind of assuming that people would have missing data.

Current management of high-risk MDS

DR BRUNNER: Absolutely. And so yeah. Where are we in 2022 with managing high-risk MDS? This is basically the same discussion that could have been had 15 years ago that for the front line of treatment patients who have higher-risk MDS, there’s a survival benefit to azacitidine compared to conventional care. We’ve known for a long time that transplant has a role in MDS. We now have prospective data suggesting that yes, if you go to early transplant in patients with higher-risk disease, those patients also have a survival benefit compared to patients who are not treated with transplant early in their disease course.

And so by and large those 2 pillars, azacitidine-based therapy or HMA-based therapy, as well as consideration of early transplant for eligible patients, remain the same and kind of frame how we’re approaching MDS in this time. And I think that one of the main questions that has persisted in MDS is if azacitidine backbone is the therapy that is standard, can we improve upon either the frequency of response, the quality of responses, and the duration of those responses, and patient survival.

And the long-awaited study of pevonedistat with azacitidine versus azacitidine alone was presented at ASH this year. And I think it provides a lot of interesting questions for the field. It randomized patients 1:1 to receive the combination or azacitidine alone, and the primary endpoint was EFS with the secondary endpoint of overall survival. Overall the groups were pretty balanced, as you would expect.

I think the surprise to many of us was that really there were not a lot of differences between the 2 arms. The overall response rate, CR, CRi, PR were essentially the same. Similarly, just like in CR and CRi, there were numerically slightly higher survival metrics in the pevonedistat/azacitidine arm, but these were not significant, and they were on the order of a couple months. No differences were seen among subgroups of interest: high-risk MDS, high-risk CMML, low-blast AML.

I think the other main surprise was really just how well azacitidine did in this group. So the overall response rate was higher than we typically see, and some attribution, although we’re still getting data from that, suggested that perhaps that’s because azacitidine dose intensity was very high in each arm. So greater than 95% of patients received dosing as specified by days 1 through 7 and every 28 days.

And so I think what takeaway can we have in the field from this? It’s that azacitidine can be improved upon, even in regular clinical practice, by really improving your dose intensity. And so the other challenge is this is still not adequate for the majority of patients, how do we make this regimen better? And I think that the other main updates from ASH this year in MDS were really looking at that question.

And so, again, thinking about high-risk MDS, is this patient a transplant candidate, in which case do we want to optimize their disease prior to transplant? Is this patient not a transplant candidate? Maybe the duration of response or quality of that response over time is more important than getting a really fast up-front response. And then how do targeted therapies play a role in MDS? Can they play a role similar to that of AML?

And so just like in how AML has looked at CPX-351 for secondary AML, there were a few studies at ASH this year that looked at CPX-351 for the up-front treatment of high-risk MDS. And this is really — this is a more intensive therapy than azacitidine monotherapy, and the intent often is can we get this patient to a more definitive therapy. And so in this study, for instance, the goal — one of the goals of this study was to get to transplant, and they showed that the CR rate with this combination was 20%, but many patients had a reduction in blasts, and 75% of patients were able to then go on to transplant, including another patient pending transplant.

And so I think in general, yes, we can use AML-like regimens. This was also shown in a study out of France against CPX-351 as front-line treatment using the AML regimen, again seeing that there are about 20% CR, 23% in this arm, but then many patients will have a blast reduction. And so if you’re goal for a patient is to get onto allogeneic transplant, using an AML-like regimen including CPX-351, there is evidence that they can reduce blasts and potentially get them to that bridge.

Another regimen approved in AML that’s being explored in MDS is azacitidine combined with venetoclax. Of note, the dosing of venetoclax is different in this, so it’s 14 days instead of the continuous dosing used in AML. And here shown was preliminary data looking at response rates in front-line high-risk MDS. And I think the biggest takeaway is that it does seem to potentially be active. We need a head-to-head comparison, but historically a CR rate of 35% and a marrow CR, a blast reduction rate of another 50%, meaning that over 80% of patients had some blast improvement or response, seems higher than we’ve historically seen.

We also have some preliminary survival data from that abstract showing that there is some difference in survival according to the quality of response. But even patients who don’t meet CR or marrow CR do seem to have encouraging ability to be on this drug and potential response durations.

These responses don’t seem to really correlate with a specific mutation profile. And so when we’ve been — treatments according to a targeted therapy or really a therapy that we can give to anyone, azacitidine and venetoclax, if it is active in MDS and does improve upon azacitidine alone, seems like it’s something that will be able to give to — kind of across the board to most patients.

But again, if we’re not going straight into transplant, if we’re not trying to get a response right away for some reason, I think a big question that we’re going to have to ask in MDS, maybe even more so than in AML, is how do we sequence treatments. And in venetoclax I would bring up these 2 curves. In the front-line setting patients with azacitidine and venetoclax can seem to have long responses, and we’ll see if those are better than azacitidine alone. But at the same time, we can salvage a lot of people by taking patients who were on prior HMA and adding venetoclax. And if we’re getting good responses at the time of progression or at the time of lack or response, is it better to start with azacitidine alone and wait until patients progress and then add venetoclax? And I think that that’s an emerging question.

It’s one thing that the group from Moffitt tried to ask retrospectively. So they looked at patients who either had combination front line of azacitidine/venetoclax or had azacitidine or decitabine alone, HMA alone. And they also looked at a group where they started with HMA alone and then added venetoclax after progression. And they found that the group that really did best, although this is retrospective, seemed to be the one that started with monotherapy and added venetoclax later. So I think that this is really a space that is of ripe exploration.

If we’re looking at durations of response, really thinking about sequencing and thinking about agents that can prolong responses might become more relevant for MDS. So one of the agents we also heard about at ASH was this TIM3 inhibitor. TIM3 is an immune checkpoint that’s also expressed on early progenitors. And sabatolimab, or MBG453, is an antibody that targets TIM3. So this study was looking at whether it could be combined and looking preliminarily at some of the durations of responses. And so you’ll see still more than half of patients had some form of response, but perhaps the median duration of responses can be prolonged by using something that’s an immunologic agent.

Similarly, looking at mutation profile, there didn’t seem to be a specific mutation that did poorly. Even in the p53-mutated group, albeit small numbers, those patients had a duration of response that seemed on par with other patients, which really, if we’re trying to prolong responses, that’s one of the goals for that group.

We also got some updates in how to optimize responses. And so eprenetapopt, or APR-246, is an agent that binds to p53 and is supposed to reactivate its activity. And so in patients with p53-mutant MDS and oligoblastic AML, these patients were treated with this combination, and this study showed some longer-term outcomes of that combination. It showed in particular that if you get a response, so if you reduce the p53 burden, if you see that go below sequencing levels as a metric of measurable disease, those patients seem to do best.

And in particular, of all the patients, the patients who went to transplant, when they had a reduction in their p53 burden below detection, that cohort seemed to be the ones that had long-term survival. And so as we try to approach p53 MDS and AML, a challenging group of patients indeed, perhaps the way to get some long-term survivors is to try to reduce the disease burden as much as possible and then go to transplant. And it’ll be interesting to see if we can translate this into other patients in the future.

That said, we are really interested in making MDS a targetable disease. Although not as frequently seen as in AML, we did look at IDH1 and IDH2 mutations in MDS. IDH1-mutated disease can be targeted, both in high-risk and low-risk MDS. This was a study by the French group, again showing that there is activity with both monotherapy, as well as combining this with azacitidine in both the high-risk up-front setting, as well as in lower-risk patients post ESA.

Very similar data also from the GFM group with enasidenib targeting IDH2-mutated disease in MDS. And so although these patients are relatively infrequent, they can respond to monotherapy IDH inhibition and can have meaningful responses over time.

Oral agents also may provide flexibility. And so one of the recently approved agents, oral decitabine/cedazuridine, mimics the pharmacokinetics of IV decitabine. And this study was looking at it in lower-risk disease. Now there are some patients with low-risk for whom we wind up giving hypomethylating agent therapy and being able to give an oral agent may have some quality-of-life benefits, in particular for lower-risk disease. And shown here is just evidence that over long-term follow up patients could receive this therapy, looking out now several years we need to be able to compare this to what the natural course of disease is, but you can have responses with HMA, and these patients can be maintained on this for some time.

So to conclude from MDS updates in ASH 2022, I think in MDS characterizing disease with molecular characterization is increasingly important, both in the up-front setting, and we talked about how the molecular IPSS has updated, how we understand disease risk. When we meet someone, we really want to know are you going to be stable or is this going to progress on us. And so being able to use molecular data to identify high-risk patients or low-risk patients is important.

HMA combinations are ongoing. I think that for me it’s important to know what the goal of therapy is for somebody. Am I trying to transplant you? Do I want to get the longest duration of therapies out of your care? And maybe then I’m sequencing your therapies. And there are novel targeted therapies that have a role in MDS, but right now they apply to too few patients, so can we develop targeted therapies for broader populations of MDS. So hopefully I’ll be able to answer that in 2023.