Novel Therapeutic Approaches and Personalized Strategies for Managing Anemia in LR-MDS         

We will start with a patient case, and then we will get a discussion and talk about this. I know that we all like to talk about MDS. This is a 64-year-old lady with increasing fatigue over the last 9 months. Isolated anemia with hemoglobin of 8.6. Gets a complete workup done that reveals, eventually, with a bone marrow erythroblastic dysplasia, 8% ring sideroblasts, SF3B1 mutation, and non-del 5(q). When we calculate the risk, the IPSS score is 2. Endogenous serum EPO level is 390. The patient is becoming red blood cell transfusion dependent, getting 3 units every 8 weeks.

[00:33:35]

Pretest 1

The question for our audience. In your practice, how would you manage this patient?

1. Growth factor support
2. Hypomethylating agents
3. Imetelstat
4. Lenalidomide, or;
5. Luspatercept.

We got a split here between growth factors, luspatercept and hypomethylating agents. Hopefully, we will highlight this in the discussion and come back to this question.

[00:34:19]

Discussion Points

Let's talk about this case a little bit. Obviously, this is lower-risk MDS RS-positive, and this is what we call MDS SF3B1. What would be your first approach to this patient?

Dr Stahl: I think, Rami, you highlighted the key factors. That this is a lower-risk MDS, it is RS positive, SS3B1 mutated. I would say for this case, it is actually pretty clear-cut. That is where we probably have the best data. Based on the COMMANDS data, which I know we will talk a little bit about, this, I think, would be a pretty clear case for using luspatercept, in my mind.

Dr Komrokji: Absolutely. Amer?

Professor Amer Zeidan (Yale University School of Medicine): Yes, I agree. I think several factors here in my mind would favor luspatercept over ESA, which I think is the most kind of important selection point in this juncture, which is the fact that the patient has a ring sideroblast of 8%. But once the patient has SF3B1, that automatically confirms a sideroblastic anemia diagnosis. And then the EPO level above 200 generally predicts a lower chance of response to ESA. And then the patient is needing, I would say, a decent transfusion amount, which is 3 units per 8 weeks. Generally, those patients have a very predicted low chance of responding to ESAs. In my mind, luspatercept would certainly be better than ESA. Now, in terms of choosing vs HMA or imetelstat or lenalidomide, I would not say these are wrong options, but they are not options I would use as frontline because we tend to think of these patients as patients who live with anemia for the rest of their lives, and you're going to need multiple drugs. I think the question is how do you sequence them, rather than are you going to use them? You're probably going to use most of them at 1 point.

Dr Komrokji: To your point, probably the adverse events or the toxicities are going to be a little bit more with the other options, and particularly hypomethylating agents, we would like to save later on. Yes, I totally agree. I think, obviously, we were all part of the COMMANDS study. Historically, we got luspatercept approved in patients after ESA failure, particularly in the ring sideroblasts, based on the MEDALIST study. And then the COMMANDS study had moved, looking at luspatercept vs erythroid stimulating agents in the upfront setting. And the study took patients similar to this: lower-risk, transfusion-dependent, randomized between ESA and luspatercept. And indeed, the primary endpoint was looking at 12 weeks not needing blood with a 1.5 gram hemoglobin increase, which is a little bit even more robust than the classical endpoints we used in trials. It showed almost a doubling of the response and the durability, particularly in the RS positive. To your point, this is a clear-cut that basically those patients will be treated with luspatercept. I think we can get a little bit into the discussion of the RS negative because the COMMANDS study allowed both RS positive, RS negative. How do you use luspatercept in the RS negative? Have you moved up front? Are you convinced?

Dr Stahl: That is a great question. I think that is a key question where probably most of the debate still is, what do you use first line? I think some of the factors that Amer actually outlined can help make that decision in patients who are SF3B1 biotype. We know from an actually very old paper, from Eva Hellstrom, I think it is a 1990s paper, to look at who really benefits from ESAs. Armer outlined the main points and those were pertinent in this case as well: patients who have a higher EPO level already, and this patient was somebody who had an EPO level of more than 200, and patients who have a higher transfusion burden. This was a patient who had 3 units in 8 weeks. Even if this patient would be SF3B1 biotype, would probably benefit more from luspatercept. The COMMANDS study looked at all those different subgroups, and the higher EPO level benefited more from luspatercept, as well as the patients who had higher transfusion burden. That's the situation where I would use luspatercept even in SF3B1 biotype patients.

Dr Komrokji: Absolutely. Amer, how would you approach that?

Professor Zeidan: I think the same approach. My approach when we saw the results of the COMMANDS trial was more of a purist approach, that anybody who fits within the eligibility as a big randomized Phase 3 trial is a reasonable candidate for luspatercept. Of course, we ended up, unfortunately, with under-enrollment of patients who were sideroblast-negative. There were around 25% of patients. That, I think, created the confusion about how those patients do with the drug. Clearly, there are more RS negative patients than RS positive patients in everyday practice, but they were overrepresented for RS positive patients in the trial. This is where we landed with the trial itself.

The main question that lingers in my mind is, how well do patients do with ESAs after luspatercept? This would be the only downside in my mind of starting with luspatercept, is that if we know that giving ESA after luspatercept does not work very well, but I don't think, based on the mechanism of action – let me backtrack for a second here. We know that once you give something in the second line and third line, the response rate is going to be lower, regardless what the agent is. The question is, what is the best maximum time? Do you start with luspatercept and then give ESA? Or do you start with ESA and then give luspatercept? My opinion, based on the current data, is that you can start with luspatercept. Some of that is just tangential from the fact that on the MEDALIST trial, which in the second line, we know that luspatercept works median duration of transfusion dependence of around 38 weeks. You are losing a lot of time from 2 and a half years to 38 weeks once you go from the frontline to the second line. I imagine no matter how much you are going to lose with ESA from the frontline, which is 1 and a half years, it cannot be that short that it justifies giving ESA ahead of luspatercept. But ideally, we should have a randomized trial where you sequence with other drugs.

Dr Komrokji: Absolutely, yes. I think obviously Max made a great point, is the way we've used ESAs, even in practice, had been typically when patients are not transfusion-dependent. Once you need transfusions on a monthly basis or your serum EPO level, historically in the model, more than 500, then the chances of response really drop below 10%. And we actually saw that in the COMMANDS study. If you look at the subset analysis with patients that had EPO level – actually, I was surprised even above 200, not even 500, the response rate to ESAs was 10%. I think, to your point, you look at other factors, the transfusions, the EPO level. The EPO data were interesting. If you had less than 200, you were more likely to respond to ESA, but then also the responses were not as durable as with luspatercept. Patients that had less than 200 EPO with luspatercept, Dr Garcia‑Manero presented the updates in the past few meetings, looking at that subset of patients that had almost 2 years durability of response, and the patients responded. I think both are becoming an option. I don't think sometimes it's completely wrong to use ESAs on somebody that is not heavily transfusion-dependent, low serum EPO. But definitely, I think we have data. We have a label that allows the use of luspatercept. Again, sometimes some of the factors to keep are also the ease of use. Luspatercept is an injection every 3 weeks vs ESAs on a weekly basis. That also makes a difference.

Now, in terms of the adverse events and the tolerability of the treatment, anything you want to highlight about luspatercept? What do you counsel the patients on adverse events? What do you look for?

Dr Stahl: Maybe I can start. I think overall it is a very well-tolerated drug. I think we'll talk about some other drugs where you need to watch blood counts much more closely, but luspatercept is a very well-tolerated drug. Things that I tell my patients are that it can aggravate hypertension in patients who already have hypertension or if it's poorly controlled. I think that's important. And then some patients develop fatigue that might not be entirely related to the underlying hemoglobin. Doesn't correlate perfectly with it. I think we all had some patients who had actually very nice improvement in the hemoglobin but can have this fatigue, and that sometimes requires a hold of the drug or a dose reduction, and sometimes even stopping the drug. But this is a minority of patients. I would say, overall, this is a very well-tolerated drug.

Professor Zeidan: I agree. It's certainly 1 of the drugs where I necessarily don't need to see the patient every time they come in, compared to some of the, I guess, more involved drugs. I think the issue of fatigue, in particular, can be a little bit tricky with the MDS. There is not a very linear relation between hemoglobin and the degree of fatigue in patients. And when you throw in luspatercept in the mix, it gets even more confusing because certainly there are some patients in whom the fatigue worsens, but for the most part, it's not significant to a degree where you have to stop the drug. In my experience, the fatigue tends to improve within the first few cycles as long as you communicate well with the patient that this is expected and you should wait, because especially once the hemoglobin improves and especially once it's above 10, most patients, the fatigue will get better. I think giving the patient advance notice and discussing that with them is a good idea because the patient might think, okay, I am going to be stuck with fatigue the whole time, and what is the idea of being on the drug? I think good communication about that is important.

Hypertension in my experience is not a big issue, but you certainly need to watch for it. The only point I'd like to highlight is that some patients are anxious from coming to see us all the time. I tend not to be very aggressive in changing drugs. I usually ask them to check their blood pressures at home and log it so that I have a better sense rather than making adjustments just based on their visits to the office.

Dr Komrokji: Absolutely. I agree. I think data also from clinical trials, the rate of discontinuation because of adverse events was really very low. To your point, we always ensure that the blood pressure is controlled, and so we've not had issues once, if patients are known to have hypertension, their blood pressure is controlled.

I think fatigue is an interesting phenomenon. To your point, I don't know if it's well understood; it can be discordant from the blood counts typically in the first couple of rounds. When you look at the data from the MEDALIST vs the COMMANDS, it is actually very interesting to me that the rate of fatigue was way less in the COMMANDS than in the MEDALIST. One of the differences that came immediately in the MEDALIST was post-ESA failure, more transfusion burden. Maybe there's also something related to that. To your point, it is so much confounded by other things, and it usually goes away after a couple of rounds.

There has been no risk of thrombosis because anytime we are increasing the blood counts, people are worried about that. That has not been an issue with this. There are actually sometimes differences in the toxicity between this and thalassemia, where this drug is approved. In thalassemia literature, I think there is a little bit more bone pain and ache. I've rarely if ever had patients complain of that. In general, the rate of discontinuation because of adverse events is very low. I think 1 of the advantages, those drugs are not myelosuppressive. Even if patients have borderline platelets, neutrophils, you can use it. In fact, there is some data from, again, the MEDALIST, that you can see trilineage responses. Every now and then, you'll see some platelet and neutrophil improvement. Not that's the goal of the treatment, but it could be helpful if somebody has a borderline platelet or neutrophil count.

[00:46:55]

COMMANDS Long-term Outcomes: RBC-TI ≥1.5 Yr Across Subgroups

Again, I think we get more and more data from the COMMANDS now. As the data matures, we really have seen a doubling of the response rates and the durability. When we look at the subset of patients that had enjoyed transfusion dependency for more than a year and a half, as we discussed, it tends to be those patients who actually even they had lower serum EPO, but there was nothing differentiating much in RS-positive, RS-negative. We saw more durability of the response in patients that are low or high transfusion burdens. We saw the same.

[00:47:33]

Real-World Outcomes of Patients With LR-MDS Treated With 1L Luspatercept or 1L ESA

There is obviously a lot of data coming out on the real-world experience with this, and it seems that the real-world data are consistent with what's seen in the MEDALIST and the COMMANDS. We looked at our real-world data also in collaboration with our colleagues in Italy, and they're more like the MEDALIST style post ESA failure. Indeed, we saw very similar responses to what was reported on the MEDALIST.

Interestingly again, people always don't like this responder vs non-responder analysis, but those patients, even in the real world, that became transfusion independent with a hemoglobin increase, we're actually seeing a survival advantage. This is something we all have discussed, that we believe treatment of lower-risk MDS is not merely just rendering patients transfusion independent or supportive care, that we probably could impact patient outcome with treatment.

At ASCO this year, there will be a presentation at real-world data, more COMMANDS population upfront, looking at luspatercept vs ESA. The data or the abstract results at least show the real-world experience is very consistent with the COMMANDS.

[0:49:00]

Patient Case

Let's go back to our patient. Again, this is a 64-year-old, presented with anemia. Bone marrow findings consistent with MDS, 8% ring sideroblast, SF3B1 mutation, lower-risk disease. The EPO level is 390. Patient is receiving blood almost 3 units every other month or 8 weeks.

[0:49:26]

Posttest 1: In your clinical practice, how would you manage this patient’s anemia?

The question was, in the practice now, hearing our discussion, how would you manage this patient's anemia?

1. Growth factors,
2. Hypomethylating agents,
3. Imetelstat,
4. Lenalidomide, or
5. Luspatercept?

I think we shifted some more people to luspatercept. Again, as mentioned earlier, I think there are pros and cons for all of those options. I think by the guidelines now, by the NCCN guidelines, by the data from the study, it truly supports luspatercept use. I think ESA will have a low chance of response in this. Like Dr Stahl highlighted nicely based on the models, and as Dr Zeidan mentioned, hypomethylating agents, imetelstat, or lenalidomide could be an option, but probably not the first-line option. All of those have a rate of response, have a role, but the toxicity is more. Probably in the totality of therapy, we would start with an erythroid-maturating agent and move to something else.

[0:50:53]

Let’s Answer Another Question

I think I'll have Dr Stahl take over and lead us through the next part of the discussion here.

[00:51:01]

Pretest 2

Dr Stahl: This is a pretest for the next part. Your patient with RS-positive lower-risk MDS has been receiving luspatercept 1.33 mg per kg every 3 weeks, but is experiencing fatigue and is transfusion-dependent, 4 units of red cells per 8 weeks after 8 weeks of treatment. Which of the following would be the next best step for this patient?

1. Continue luspatercept at the same dose for an additional 8 weeks.
2. Continue luspatercept every 3 weeks at the same dose with an ESA.
3. Increase the luspatercept dose to 1.75 mg per kg every 3 weeks, or;
4. Discontinue luspatercept and switch to imetelstat.

Please vote. Here we go. Most respondents increased the luspatercept dose to 1.75 mg per kg every 3 weeks, which I think is the right answer here, and we'll go through a couple of slides to give the rationale for that.

[00:52:11]

Luspatercept Response-Adapted Dosing

I think it's very important to keep in mind that luspatercept is dosed in a response-adapted way. There's a starting dose of 1 mg per kg every 3 weeks, but this dose is increased if patients remain transfusion-dependent despite receiving twice the same dose. There are 2 other dose levels that can be increased to. One is 1.33 mg per kg. And then, if this is not sufficient, increase the dose further to 1.75 mg per kg. This is what this case was about. There are also dose decreases if patients have too rapid an increase. This was, again, with the safety, I think, kept in mind that Rami outlined that some of the other agents had thrombosis risk. Although I have to say this in real practice doesn't happen very often, but if the hemoglobin is more than 11.5, treatment would be interrupted and would be restarted once the hemoglobin is less than 11.

[00:53:13]

Real-world Experience of Luspatercept Dose Escalation in LR-MDS

Real-world experience in luspatercept shows that this is important to keep in mind. I think this is very important for practice because in this real world data analysis was shown 85% had a dose escalation or required a dose escalation, and then 54 of those maintained escalated doses. Among 66 patients who were transfusion dependent before starting luspatercept, 47% achieved transfusion independence using this dose-escalated approach. Of 61 patients who were transfusion independent before starting luspatercept, 82% remained transfusion independent.

Maybe some discussion points that I would love to hear the opinion of my colleagues. We will start with Rami here to my right. What is your experience with response-adapted dosing in the real-world setting? I'm sure you see a lot of patients. Either they come to you first, or maybe they come to you for a referral from the outside. What do you see, and what do you advise your colleagues to do?

Dr Komrokji: I think before calling treatment failure, I think patients should have the highest dose in general. Obviously, we do see responders on the 1 mg per kg dosing, the starting dose, but if not seen, one should escalate the dosing. Even from the data on the MEDALIST study originally when we looked at that and we looked at particularly if the patients are a little bit more on the high transfusion burden, almost all the responses are seen on the higher dose, that sometimes you could argue in practice, can I just go to the higher dose from the get going? Definitely the message is do not stop in the middle. Escalate the dosing up to the max. If the patient has had 2 doses at the maximum dose, then basically, the treatment is not working.

In my experience, I have not seen any adverse events that are dose-dependent or dose-related. Going up on the higher dose will not increase any of the toxicities we worry about. Amer, and I know his thinking about this because we are aligned on this, we think that you should actually be able to normalize the hemoglobin for those patients. And the package insert and even data with ESA limits you, that when you hit the normal hemoglobin, you have to hold. And that's probably not physiological, but it goes to your point of the fear about pushing thromboembolic events that comes from all data on the ESAs and actually solid tumor literature, but we are always cornered in that position. If the patient's hemoglobin goes higher than 11.5, you have to hold, adjust the dosing.

But I think the message really is, especially if the patients are needing blood transfusions, there is more likelihood that they will need higher dosing, and there is no increase in the adverse events. It's not dose-dependent, but we do not stop the treatment before we have tried the higher dosing. Again, in my experience, there is very consistent with the real data that 70, 80% of the patients will need dose escalations.

Dr Stahl: I think those are great points, and I think this will also be the subject of a clinical trial, to explore whether we could start right away, as you mentioned, with a higher dose. I think we'll talk about the MAXILUS trial a little bit later on.

Dr Komrokji: I always think that you should be the PI for that study. It's named after you.

Dr Stahl: Great point. Maybe I will hand it over to Amer for the next question. If you reach now this maximum level, Rami has titrated up to the maximum level, but the patient still remains somewhat transfusion dependent, what is your strategy for those patients? What do you do with those patients?

Professor Zeidan: I think 2 points I'd like to highlight. I think there is some confusion because the same drug is approved in the second line and the first line setting, and how we handled the dosing was actually a little bit different between MEDALIST and COMMANDS. In COMMANDS, we were not only trying to get to transfusion independence but also trying to get to a hemoglobin of above 10.

Actually, we have a number of analyses that showed that those patients generally have better patient-reported outcomes. Less fatigue. I think some of the confusion comes that the patient becomes transfusion-dependent because, again, depending on where you practice, some people will transfuse only if you're less than 6.5 or 7. The patient is hovering on a hemoglobin of 8, 8.5, and they're not getting transfusions, but the patient does not feel great, and the doctor stays at the 1 mg. I think it's very important in this setting, even technically the patient's transfusion dependent, is you want to try to get to the dose that gets you the highest hemoglobin.

I actually often see it the other way around, where you cannot get to very high hemoglobins, rather than overshooting the hemoglobin. All of us, I think, agree on this point, but I think we tend to be very conservative in our approach in general to treating anemia, but also in dosing this drug. I think the approach of dosing the highest from the get-go, is I think a very valid approach in a drug that is very generally safe to use.

To your direct question about the issue of the maximum dose of the luspatercept, what would you do at that point, I have 2 thoughts. I think one is that it is not always a clean win in these settings. I will give you an example. One of my patients, who was actually in his mid-80s and came to me from another state, has a medication list that's longer than a page, like 25 medications and a million comorbidities. He had low-risk MDS with ring sideroblasts, and he was not responding to ESAs. He had already started transfusions almost every other week.

I put him on luspatercept, and he went down to almost 1 unit every 8 weeks, 10 weeks. He still needed a unit every now and then. For me, a significant reduction in the transfusion burden and the frequency of transfusions, even at the maximum dose, is still a win. What we do at this point, I think you have approaches. You could potentially switch him, but this is a patient I probably would not give HMA to. He just was too frail. This was before we had imetelstat, but I think there are some patients for who you could take this as a win, that they have a significant reduction in their transfusion needs. They are needing intermittent transfusions.

We know this concept of cumulative transfusion dependence. This is what happened with this patient. He would get a UTI and get hospitalized every 3, 4 months and get a unit here. Each time he's in the hospital, they give him blood. I don't think it means that he stopped responding. I think you have to apply some judgment in those situations.

Now, there are situations where I think you might consider switching, and that depends on what are the symptoms, how the patient's level of energy, what is the maximum hemoglobin you get to. In that situation, I think you could consider switching to imetelstat or potentially HMAs or even lenalidomide.

Dr Stahl: Great. That's a wonderful answer. I think we all recognize that clinical trials and the real world can be different.

[1:00:45]

MAXILUS: Maximum Dose Luspatercept in Transfusion Dependent LR-MDS

Let me just, in the sake of time, go to what we already mentioned. The MAXILUS trial, which is going to look at the maximum dose of luspatercept in transfusion-dependent lower-risk MDS, right to start from the beginning at a dose of 1.75 mg per kg, maybe to get patients to a benefit quicker now that we know this is a very safe drug. How do you feel MAXILUS would potentially impact clinical practice?

Dr Komrokji: Obviously, I think the main rationale behind the MAXILUS study is really also to demonstrate the safety that you can start with 1.75. Regardless of even what we get the data, but I think it will establish the safety that those erythroid-maturating agents, you could go to the highest dose from the beginning with no worries about the adverse events. That we intuitively know, and we have done it, but you need the data to support it.

As I mentioned, I think in patients that are a higher transfusion burden, they will need that dosing. Maybe in that group of patients it will be easier in practice to go with the higher dose, because sometimes I feel the frustration for the patients, telling them we're going to do the 1, 1.33 and you're not going to respond because we already know you are going to need the 1.75. Hopefully, that will help with that. But I think the main issue that it is going to be establishing that it is safe. There are no adverse events, no discontinuation. That is probably the goal. It is also looking at this RS negative. We will get some indirect evidence about other things, questions in the field from the study.

Dr Stahl: That is a great point. Wonderful.

[1:02:23]

Let’s Revisit Our Question

Let's revisit our question. This is a patient who has not responded to a 1.33 mg per kg dose, and you see the options here. What would you do?

[1:02:31]

Posttest 2: Your patient with RS+ LR-MDS has been receiving luspatercept 1.33 mg/kg Q3W but is experiencing fatigue and is transfusion dependent (4 units RBC/8 wk) after 8 wk of treatment. Which of the following would be the next best step for this patient?

93%, wow! That was an effective discussion, I would say. That's great. Wonderful. Here's the rationale again that we outlined that the dose should be increased based on the package label. I think I will hand it over to Amer for the next session.

Professor Zeidan: Thank you so much.

[01:03:09]

Second-line Management of Anemia in LR-MDS

I think here we will talk about what you do after your first line of treatment for patients with lower-risk MDS, and we'll talk first about the patient case.

[01:03:20]

Patient Case

This is a case of a 70-year-old male with anemia, undergoes a workup for the anemia and turns out to have a lower-risk MDS. He has a couple of mutations, DNMT3A and SF3B1. The patient does not have deletion (5q). As often has been the case, the patient gets ESA in the frontline setting, short-acting ESA for 18 months. The patient, after the initial response, starts complaining of fatigue, and the anemia starts to worsen. The hemoglobin is 7, but the platelets and the white cell count are preserved. A bone marrow is repeated at that point because of loss of response and does not show progression in the sense of increasing blood count or chromosomal abnormality, but the patient clearly still has MDS with ring sideroblasts. And cytogenetics are still normal. The patient now needs a lot of transfusions. He needs 6 units of blood every 8 weeks.

[01:04:22]

Pretest 3

In your practice, how would you manage this patient's anemia?

1. Azacitidine or decitabine
2. Lenalidomide
3. Luspatercept or imetelstat, or;
4. Switch to another form of ESA, which would be the longer-acting version, like darbepoetin.

Please vote. The majority of people chose luspatercept or imetelstat, and we will discuss our thoughts on this case as well.

[01:05:00]

Pretest 4

The second question, working with your colleagues in the healthcare system, how would you first manage the first occurrence of Grade 3 neutropenia in patients with lower-risk MDS who are receiving imetelstat?

1. Do you continue imetelstat at the same dose that you started with, but just do more frequent monitoring?
2. Do you hold until resolution to ANC above 1? So, this is again Grade 3. You wait until ANC is above 1 and you resume at the same dose.
3. Do you hold until improvement of ANC to 1, but resume at the lower dose.
4. Do you discontinue imetelstat altogether?

Please vote. So, 60% chose to hold until ANC recovery and then go to the next lower dose.

Maybe I will start with you, Rami. This is a patient I think that's very typical, especially in the pre-COMMANDS era where the patients were getting started on ESA. The patient gets the typical response, which is generally around 1 to 1 and a half years of transfusion independence. And then, as I think, your group has outlined very nicely, that most patients when they progress within lower risk MDS, they still continue to be lower risk, but they are just needing transfusions again, and they need clearly a different strategy. How would you approach this patient?

Dr Komrokji: Absolutely. To your point, this is a lower-risk MDS patient. Again, MDS with SF3B1, that's a unique category and exactly to your point, majority of the lower risk MDS actually when we look will not progress to a higher risk disease. But inevitably almost everybody would lose the response to the ESA or to any line of therapy in that regard. And the unmet need always remains this ineffective erythropoiesis, patients becoming transfusion dependent.

Again, this patient just from the molecular profile having SF3B1, DNMT3A, no bad risk features, I would definitely have predicted that the bone marrow would not show progression. Now we are in a patient that is probably a relatively heavily transfusion dependent ESA failure and the options, obviously by data we have luspatercept from the MEDALIST study. This was the population that it looked at, lower risk, ESA failure, transfusion dependent and there was almost 40% responses at the primary endpoint. But even when you look into the data a little bit more there were higher.

Now, the subset that were high transfusion dependent, the response rates dropped down. And to the point you were discussing in some patients, you actually were reducing the transfusion burden then rendering the patients completely independent. But I think it is an option. I think the appeal about the luspatercept is the ease of use, the lack of myelosuppression. So if somebody says, I want to start with luspatercept, this is the population treated on the MEDALIST study and it's very well tolerable, and if it's not working then I'll move to the other option. I don't think somebody can fault that. The appeal is the ease of use for the patients.

Now the imetelstat I think is also another good option. And this is from the IMerge data that again looked at the same population here. ESA failure, transfusion dependent, and the study selectively had patients with higher transfusion burden, similar to here. And the responses were almost around 40%. And if patients responded in that 40%, they were durable with hemoglobin increase. The downside anyway is we do see more myelosuppression and grade 3 and 4 thrombocytopenia, neutropenia were in the range of 50 to 60%. But reminding everyone that the clinical events were really very low; febrile neutropenia, less than 1%, any clinical meaningful bleeding was low. But there is definitely the myelosuppression that will be seen, that you'll sometimes need to hold the dose, delay, reduce the dose later on based on the guidance.

So I think the argument for the imetelstat in this case, this is a high transfusion burden patient, 6 units every 8 weeks, the chances of response even to luspatercept are less. The best chance of response for this patient is probably imetelstat. The downside for it that we're going to see a bit more myelosuppression.

Sometimes in cases like that, you can actually discuss the 2 options with the patient, say this is the pros and cons of those 2, and sometimes the patients will weigh on it. I think from a response point, imetelstat has better chance. From tolerability, ease of use, luspatercept is better choice.

Professor Zeidan: I think this is an outstanding summary, but Rami I'm going to pin you down. We made it easy for the audience and put luspatercept and imetelstat together as reasonable options because this case really fits both the MEDALIST and the IMerge data, but my good mentor Dr Go told me you always discuss options but you owe to the patient to give them a specific recommendation what you think is the best. Between the 2 drugs, based on this specific case, which one would you – or you have no, absolutely no preference? You just leave it up to the patient?

Dr Komrokji: No, I think it depends on the goal. I'll ask the patient what do they want to achieve, and if they really will say, my goal is I want my hemoglobin up and become completely transfusion independent, I would say imetelstat would be my choice. If the patient doesn't want to be coming every week for the coming 2 months to check the blood counts, the dose delays and reductions, and wants to give it a try and he's okay with still getting blood for 2 months or so till we figure out luspatercept is working or not. My gut feeling or based on that, the chances of becoming transfusion independent are definitely higher with imetelstat and that's what I'll tell the patient. But the patient can talk me out of that and say, I don't want to come every week to check blood count for the coming 2 months, do that. Why not try luspatercept first? If I'm driving response, we do. If not, after 2 months we shift to the other one. I can be talked easily into that.

Professor Zeidan: Yes. How about you, Max? Do you just leave it up to the patient after the discussion, or do you have a stronger feeling towards 1 of the 2 drugs?

Dr Stahl: Yes, so just to entertain your point, I'm going to take a strong view, but I want to say I completely appreciate Rami's more nuanced view. That's probably what I would do in clinic too, honestly. But just to entertain for entertainment’s sake, I think the 1 thing to talk about is probably sequencing of the approaches. We have pretty good data from secondary analysis from the IMerge data that imetelstat can still work quite well post-luspatercept. We have much less data for luspatercept post-imetelstat. For this patient, just to put my nickel down, I would probably use luspatercept first because I know imetelstat might still work in the second-line setting. But of course, I agree with Rami, if a patient would say, I don't want to travel, I don't want to have blood counts all the time, it's probably more luspatercept, and the other way around you could make an argument as well. But my vote would be luspatercept first.

Professor Zeidan: Yes, and I agree with both of you. I think this is a very good example of why it's important to understand the clinical trial data and what are the patient goals and what is it that they value the most. Is it transfusion independence, or is it freedom from coming to do blood counts and having to go through, for example, those interruptions and growth factor support, et cetera, which is more common with imetelstat? But I also agree with your point that I think, which weighs in my mind, is that we know luspatercept, or we know that imetelstat works after luspatercept.

While I think it does work, we still don't have data with luspatercept after imetelstat, and I always have this concern about burning bridges down the road if somehow a drug does not work very well. Actually, 1 of the first papers I worked, I remember with Rami actually, was about the same issue with lenalidomide and azacitidine, and this was actually quite clear that you should give lenalidomide before azacitidine in that sequence.

I think now we have to do more of this, as we have a number of active agents that you can do after ESAs in terms of how you choose your sequencing. I think we would love to see some data about activity of luspatercept after imetelstat, because I think this certainly would make I think the discussion even better with the patients. But maybe we can cover some of the, I think, highlights. I think we covered this quite well.

[01:14:02]

MEDALIST: Luspatercept in ESA-Refractory Lower-Risk MDS

The MEDALIST and the IMerge are the 2 trials that led to the approval of those drugs in the second-line setting after ESA failure. The main differences between IMerge and MEDALIST, and both of them compared against placebo, the main difference is that MEDALIST only included RS-positive patients. This patient we just discussed was RS-positive, which I think made the patient potentially eligible for both approaches.

But the second difference is that IMerge had a little bit more transfusion dependency. The average transfusion or the median transfusion was around 6 units per 8 weeks, while it was I believe, around 4 in the MEDALIST. I think for patients who are more transfusion dependent, as I think Rami mentioned, you might lean more towards imetelstat.

[01:14:49]

METALIST: Reduction in RBC Transformation Burden and Improvement in Hl-E

Similar between the COMMANDS and the MEDALIST. Sometimes you don't get these clean wins. I think the same thing applies for imetelstat, although in my experience, imetelstat tends to be a little bit more black or white. You either have a great response or you have no response. This is my own sense of the data, but I think with luspatercept you can see these middle zones where you have significant reduction in transfusion, but actually, most patients with luspatercept will get some degree of reduction in their transfusion burden.

[01:15:25]

METALIST: Cumulative Duration of RBC-TI Among Patients Achieving ≥8-Wk RBC-TI

I think an important point to highlight with luspatercept in particular is the idea of the cumulative transfusion dependence. We covered this in the frontline setting, but it's particularly important in the second-line setting in my opinion, because many people underestimate these drugs and say, well, the median transfusion independence with luspatercept is 38 weeks. Why would you want to go through all these infusions every 3 weeks just to get the patient median transfusion independence period of 38 weeks?

However, when you take into account what we discussed earlier about those patients being older, often hospitalized, they need transfusions here and there, when you look at the cumulative transfusion independence with luspatercept, it's close to 2 years. It's not really 38 weeks. This is the number I tend to highlight for patients. And Rami covered this a little bit, but the issue of – and I’ve never been a big fan of these responder vs non-responder analyses, but it's clear that patients who respond tend to derive the most benefit from these drugs.

[01:16:29]

MEDALIST: Overall Survival by Response

I think whether there is an overall survival advantage is a very important question. The way we design our current trials clearly does not, I think, give us the option to answer these questions. There was a trend for improvement in the frontline setting. In the MEDALIST there was not a clear trend for improvement when you take the entire patient population. But again, remember these studies that include a couple of hundred patients, it's going to be very difficult to show overall survival improvement.

[01:16:57]

IMerge: Imetelstat vs Placebo in ESA-Refractory Lower-Risk MDS

With imetelstat, again, very effective drug. The rate of transfusion independence in a very patient or transfusion-dependent patient population was 40%, and this trial, the IMerge, included both RS-positive and RS-negative patients. If the patient we just talked about was RS-negative, I think that would sway me more towards recommending imetelstat over luspatercept.

[01:17:22]

IMerge: Rates of Durable RBC-TI Over Time

But the durability also with imetelstat is very nice. The median is around 52 weeks. One interesting phenomenon with imetelstat compared to luspatercept is that you tend to see more of this 1 single episode of transfusion independence rather than these cumulative transfusion episodes that we see with luspatercept.

[01:17:43]

IMerge: Most Common AEs

The side effects, I think Rami covered this quite well. In terms of the cytopenias, people for some reasons that are not fully clear to me, people freak out over neutropenia and thrombocytopenia with imetelstat, although they dose HME all the time. It's really very similar phenomenon. It's that you have part of the mechanism of action of that drug is that the neutrophils and the platelets will drop, but the way to do that, you have to monitor the patient. The first 2 cycles, you actually have to bring the patient every week for the first 2 cycles to make sure the platelets and the neutrophils are on good range. Sometimes you have to hold the drug, you have to reduce the dose, you have to give growth factor support. Rarely, you might even have to give platelet transfusions. It's not common, but it can happen. But generally these don't lead to severe infections or bleeding as long as you monitor and manage. If you just give the drug and don't see the patient for a month, you are going to run into trouble. It's very important that those patients should be monitored carefully.

[01:18:38]

Managing Severe Neutropenia and Cytopenia With Imetelstat

And I think Rami mentioned the point very well is that if you want to use imetelstat, you have to have a patient who's willing to do this, who's compliant, who's not going to skip doing the blood work and travel without telling you, et cetera. These are very important things to discuss with the patient.

Now, I think the dose reduction, and this was 1 of the questions that was asked to the audience, can get a little bit tricky. When you look at the label specifically, is that with the first episode, if it's grade 3 neutropenia, the first episode, you can just hold until improvement of ANC and resume at the same dose rather than at the lower dose. With grade 4, if you get to that, you basically resume until count improvement, but then reduce the dose.

Now in all honesty, again, this is what's in the label, but we probably, again, each one of us does this a little bit differently depending on the level of the neutrophil drop, how long does it take to recover, do you have to give growth factors, does the patient have a history of infections, et cetera. I think all of these factor in, but technically if you are grade 3, you can resume at the same dose level after recovery.

We covered, I think, a lot of points, but I want to get your thoughts. Rami and Max, both of you talked a little bit about this, but 1 of the main points that were presented in this ASCO by Dr Garcia-Manero on Friday actually is about the trend for overall survival benefit in the COMMANDS in the frontline setting with luspatercept compared to ESA. Now, this data comes with a lot of caveats, but you are going to the clinic on Monday, and when your next patient comes and you tell them we have this exciting update from ASCO, how do you frame the overall survival advantage? How do you view it, and how do you frame it to your patients? Maybe I'll start with you, Max.

Dr Stahl: Sure, yes, I think we're all looking for it. We all are hoping that we actually, instead of just improving blood counts, that we actually impact disease biology. And I think for luspatercept, compared to ESA, I think there's at least some rationale that you might change the microenvironment of the bone marrow that makes it more conducive to not only blood production, but might also lead to maybe an effect on the clone itself.

It's all what we're looking for, and I think the data is very early. It's definitely exciting. I would not say to a patient yet that we have clear benefit of survival advantage. I think that would be fantastic. But when we looked at the curves that are presented at ASCO, the curves separated quite late, and at that time, very few patients were actually at risk. Even small differences between a couple of patients can make a difference. I think longer follow-up is needed, but it's certainly an exciting data point. I don't know how you feel about it, Rami.

Dr Komrokji: Yes, I totally agree. I think first obviously this was not a planned primary analysis, this is secondary looking at endpoint, but it's interesting and sometimes for me those are more the principle that we've been for many years trying to advocate say, if you have good treatments for lower risk MDS, restoring hematopoiesis may impact the outcome.

Sometimes, for me, those COMMANDS, or now even when we are moving into the ELEMENT to start challenging this dogma that we should just treat patients when they're transfusion dependent. I think for me, obviously, it's very interesting and exciting because we've not seen any survival data. This is probably 1 of the first few studies where we see survival data. Again, it's not statistically significant, et cetera, secondary endpoint with all those caveats to your point. But it also raises the question, could there be more with those class of drugs in MDS rather than just anemia response.

Those are drugs that target the TGF beta pathway. There's inflammatory milieu. We are talking about cardiac mortality, about important factor in MDS. We're talking about even that in pre-MDS, now that patients can have increased cardiac mortality, could those drugs be doing something even more? It's reassuring, definitely.

I would not go and tell patients you should be on this treatment because you're going to improve your survival, but I would say that this is encouraging in patients that needed this treatment. Like we're starting to see some trend which is very promising, and I think we should really, in the next trial, start focusing more on this, look at things like cardiac events and other things beyond just the anemia response.

But it's very exciting to see a proof of principle that now we start to see a glimpse on a survival advantage in lower risk MDS patients, which reminding our audience, we don’t take lower risk MDS very seriously many times, but a patient who is a lower risk MDS that became transfusion dependent, those patients’ survival is impacted by the disease. We are talking about 4 or 5 years maximum survival in the best situation. Those patients are not going to be doing well for 10 years, needing transfusions.

Professor Zeidan: I agree with I think both of your approaches. In my mind I think it's another piece of the discussion with the patient, rather than saying, okay, your survival is going to be improved. Which I don't think you can say based on this data, with all the statistical caveats. But I think it's another piece you add to the higher rate of durable transfusion dependence with luspatercept, et cetera. I'd love to see more of this data, especially in RS-negative patient population, but I think it's trending at least in the right way. I think that's a good way to tell patients.

[01:24:16]

2025 Guideline Recommended Management of Non-del(5q) LR-MDS With Symptomatic Anemia

I think for the guidelines, people I think rely a lot on NCCN and other guidelines. I think the guidelines cover what we generally have discussed. The issue of the second line management continues to be, I think, a little bit tricky right now because we are in a good situation. We have more and more drugs, but we don't fully understand the sequencing, but I think we have a number of good options between imetelstat, luspatercept, lenalidomide and HMAs. All of us seem to agree that HMA probably should be the last line. Lenalidomide in non-del(5Q) probably should be the drug before the last, and the real question is, do you give luspatercept or imetelstat after ESA, assuming you didn't start from luspatercept in the frontline. But if you started with luspatercept in the frontline, I think there's another good discussion is do you try ESA at that point or do you just go directly to imetelstat?

[01:25:16]

Patient Case

Let's revisit our questions with the audience. This is the case we presented at the beginning. RS-positive patient who responded initially to ESAs for lower-risk MDS and then subsequently lost the response. So, second-line RS-positive patient.

[01:25:29]

Posttest 3

What would you give at this point?

1. HMA,
2. Lenalidomide
3. Luspatercept
4. Imetelstat or;
5. Switch to another ESA?

Okay. I think most people will agree that they would use either luspatercept or imetelstat before lenalidomide or HMA, or switching to another ESA would not do at all, because if you have clear failure of I think short acting, it doesn't make much sense to try to go with the longer acting drug.

[01:26:10]

Posttest 4

In terms of the dose modification, this was, I think, a little bit of a trickier question. In this patient, if they got a grade 3 neutropenia, how would you adjust?

1. Do you continue at the same dose and just monitor the patient more frequently?
2. Do you hold until resolution and resume at the same dose?
3. Do you hold until resolution, but continue, also resume at the next lower dose?
4. Or do you discontinue altogether?

I think more people got the right answer about resuming at the same dose level for grade 3, while with grade 4, you have to go to the next lower dose. I would highlight here that with imetelstat it’s the opposite case of luspatercept. Where most patients with luspatercept you have to escalate, with imetelstat it’s the opposite actually, where most patients you have to reduce. And that brings the question, is the opposite, I guess, concept of MAXILUS. Do you start with a lower dose and escalate? Of course, this is not how the trials were designed, but I think I certainly get asked about this often, and I think people with frail, older patients, they might start lower and then go up.

For future strategies, I think this is…

[1:27:40]

Future Treatment Strategies in LR-MDS

Dr Stahl: That's me actually. Yes. I think maybe we have a minimus trial, not a MAXILUS trial for imetelstat. I guess 1 discussion point is, and I think actually Rami mentioned this already a little bit, that the general thought process for lower risk MDS seems to change a little bit. That we're not just waiting until they’re heavily transfusion dependent, but we actually try to change the disease biology a little bit early. I think the first really big evidence that came into this was the Sintra-Rev trial. How do you view the data of the Sintra-Rev trial, and how does it impact your general thinking about lower-risk MDS treatment, Rami?

Dr Komrokji: Absolutely. I think you nicely summarized the thought process behind those studies because historically up to date, every trial that led to approval of a drug in lower-risk MDS was based on taking a group of patients that are already transfusion dependent, which in my mind is a definitive bone marrow failure biology, and rendering them transfusion independent. As we get active therapies the question, can we move those earlier? Would we get higher responses, more durable responses, would we make an impact?

The first study probably to test this concept is this Sintra-Rev study. We know that lenalidomide is pretty active in deletion (5q). It's been the standard of care for many years now for that group, and it's a treatment that can affect the disease biology in cases of deletion (5q). The Spanish group said, what if we think of starting lenalidomide earlier? We're going to go for patients that are anemic, not needing blood, lower risk, with the deletion (5q), and also we're going to give a lower dose, to Amer’s point now even about the imetelstat, and we're going to do it with a fixed duration.

A lot of advantages. Instead of 10 mg lenalidomide that we use, we're going to do 5, we're going to do it for 2 years fixed duration, not indefinite. Our endpoint or goal would be seeing the time it'll take those patients to become transfusion dependent. It's time to transfusion dependency, and indeed they showed that in the group. It's a small, randomized trial, but it's still randomization. In the group that got the fixed-duration lenalidomide for 2 years, it took 6, 7 years for those patients to become transfusion dependent, vs probably less than a year in the patients under observation, with obviously some other benefits. There were more cytogenetic responses seen, complete cytogenetic responses. Really, the first proof of principle in a way to challenge the dogma that we always have to treat patients once they're transfusion dependent.

Obviously, there could be some downsides for that approach, like you'd say, how about the toxicity? Why would I treat somebody earlier? In general, the toxicity was reasonable. There was some GI toxicity. People discussed clonal evolution with you. If you started treatment earlier, select bad clone, especially when you have treatments like lenalidomide. There were small subset that they looked at, there was no evidence of clonal evolution. On the contrary, actually in the P53 – again very small numbers, we're talking about few patients – stabilization of the clone.

But those are things to keep in mind. Anytime we're moving treatment earlier, it has to be well tolerable, not adverse event. It has to have probably not detrimental effect in selecting or clonal pressure for disease evolution.

The Sintra-Rev really is the proof of principle. Do we do it in practice? I actually discuss it with patients. Sometimes I bring it up and say, there is this now. You have the option of going on. It's very intriguing. And then, some patients will elect to do it because, especially when patients have symptomatic anemia, and nobody wants to become transfusion dependent on the long term.

I think the other trial now that's looking at the same principle is the ELEMENT. We're now moving luspatercept very similar to what we've been discussing all this morning in the MEDALIST: ESA failure, COMMANDS upfront, transfusion dependent. Now, we are moving luspatercept against ESA and non–transfusion dependent, and the first time looking at the rate of patients that will become transfusion dependent at 6 months or something like that. That also will be informative. A little bit different from the Sintra-Rev because luspatercept has less side effects, but it doesn't affect the clone in the same way. It'll be very interesting to see and I don't know what Amer also thinks.

Professor Zeidan: No, I agree. I think all of these trials go into the next wave of thinking about how should we manage MDS. We started with this, but probably ending, reminding people about this is important, anemia cannot be good for you. Anemia often gets ignored in lower-risk MDS. The prevailing practice has been, well, your hemoglobin is 9, but you are not short of breath all the time. Then we can monitor you until you are very short of breath and you're needing transfusions.

I always make the analogy to hypertension. Nobody would ignore a blood pressure that's persistently high just because someone has no symptoms, because we know in the long run this is going to cause cardiovascular problems, and this is what patients with MDS die from. They don't die generally from converting to acute myeloid leukemia. They die primarily from cardiovascular complications.

Now, this could be directly from anemia, or it could be due to other inflammation and other things that happen at the cardiovascular system. But I have a strong belief that treating patients earlier and aiming for higher hemoglobins will eventually translate into better outcomes, and I'm hoping the ELEMENT would be 1 of the first few trials to show that. In all honesty, what we saw with COMMANDS was quite encouraging earlier. I'm hoping we can see similar trends with ELEMENT.

Dr Stahl: Wonderful. That's a great summary.

[1:33:40]

ELEMENT-MDS: Luspatercept in Patients With ESA-Naive MDS and Transfusion-Independent Anemia

All right. Here is actually the element trial, which I think Rami already beautifully summarized. Okay, good.

Maybe just 1 more question to both Rami and Amer, and this is in the spirit of being a little bit of a contrarian here, although I totally agree with both of your thoughts about this. How do you think about the European data, who also try to give ESA early? Which was kind of a similar approach to Sintra-Rev, to give ESAs earlier, even before they became transfusion dependent, and they didn't show delay in the onset of transfusion dependence. Do you think ESA is just not impacting enough the disease biology and is not impacting? Or how do you interpret that data?

Dr Komrokji: Yes, so you're referring to the study that was presented at last ASH. I think there are a couple of things with that. First, the numbers were really very small to make a conclusion. The historical data from the European registry suggested the other way. And then even in the study, it was confusing because the hematological improvement was better, but they were not able to delay the transfusion. I walked out a bit confused from that, not knowing exactly, and to your point, it could be also you have to have a certain magnitude of response or delta to see really impact of moving the treatment earlier. You saw that clearly with lenalidomide because it's very effective. Again, I didn't walk away with a clear message from that data.

Dr Stahl: Perfect. I agree. Yes. Maybe last point too for Amer, what are you most excited about novel approaches in lower-risk MDS? What is on the horizon that excites you?

Professor Zeidan: Yes, so I think it's actually a very exciting part to be in MDS clinical research, in particular for lower-risk MDS, where we had more success, but also we had, I think, a number of interesting agents. Each time you have, I think, good drugs, you always get second in class agents, which could be better. I guess that's a question that remains to be seen.

We have elritercept, which is a second in class transforming growth factor agent, but also we have other, I think, interesting drugs. For example, some of the JAK inhibitors borrowing from the anemia improvement we see in myelofibrosis. There's this new class of drugs called pyruvate kinase agonist, which work for pyruvate kinase deficiency and have some activity in thalassemia and potentially in sickle cell, which is something by the way, we see with luspatercept as well, it works in thalassemia.

Sometimes, I guess could be borrowing from other diseases, and clearly for the small niche of patients who have IDH mutations, which is a small number probably around 5% or less, those are very good options, very well tolerated drugs, generally effective even in lower-risk MDS, significant transfusion independence. Which I think reminds me of a point we did not discuss much at the beginning, but it's very important to do Next Gen sequencing in these patients because if you have an IDH1 or IDH2 mutation it could open the door.

And then I think the last point I would make is that there's a number of drugs that are looking at the innate immunity. Unfortunately, we have not had so much luck with these so far, canakinumab and inflammasome modifiers, and these kind of drugs. So far, we have not seen significant success, but there's a number of these drugs, and there is some data with IRAK inhibitors. Again, I think relatively early data.

And then the last point I think is the combo, right? When you have all these good drugs, you are leading, I think, an interesting trial combining ESA with luspatercept in the frontline, and I think this is the direction we are going as earlier initiation of therapy combinations and how to understand sequencing. Similar to I guess what's happening in multiple myeloma. I like to take multiple myeloma as a good paradigm of what we're doing with lower-risk MDS, but we certainly need more agents.

Dr Stahl: That's great. Wonderful summary. Lots of excitement, I think, on the space, and this is what Amer already mentioned. And I'll hand it back over to Rami.

Dr Komrokji: All right, so I think we're getting to answer final few questions here. Those are again polling questions.

[01:37:59]

Poll 3

Do you plan to make any changes in your clinical practice based on what you learned today in program?

1. Yes
2. No
3. Uncertain

[01:38:17]

Poll 4

Please take a moment to enter one key change that you plan to make in your clinical practice based on what we discussed today. A good change in your practice, not a bad habit.

[01:38:33]

Q&A

All right, so we have questions. Let's see. Yes, so because of time, we'll probably try to take 1 or 2 questions or lump them together. There's questions about basically ties to the luspatercept. They're saying in a cohort of patients they've seen transfusion independence, but no increase in the hemoglobin. Describing that phenomenon, and also what's the mechanism of resistance of how do we lose the response to luspatercept. I'm trying to tie a couple of questions together, so if you guys want to…

Dr Stahl: Sure. Yes. Maybe the point of achieving transfusion dependence, but no increase in hemoglobin. I actually thought this was – and I think Amer mentioned this actually in the COMMANDS trial – I thought actually, the COMMANDS trial had the most robust endpoint of all of them because they demanded transfusion independence plus a hemoglobin increase. This was made because I think, very wise from the investigators, to ensure that it's not just like people are already almost transfusion independent, and then the placebo rate is very high. I think actually in most cases you do see a concurrent increase in the hemoglobin, but I think if somebody turns transfusion independent, that's already an achievement that I think is worthwhile on its own.

Professor Zeidan: No, I agree. I think 2 points that I think apply to clinical research in particular. Each time I'm approached about a drug for anemia in lower-risk MDS, and I'm being told that we had, I don't know, 3 out of 20 patients who had transfusion independence, I want to see the hemoglobin. Especially in the open-label setting where it's not like a blinded double placebo control. In open label trials I think you need to have some hemoglobin rise to feel confident that the drug is actually leading to a good change rather than a change in the doctor's practice, and that the patient now is in a trial, they stop transfusing. I think, again, this could happen with patients where they, because the doctor started them on something, they say, well, I feel better. Is that real biologic effect? I think it's tough to know, but my own approach is, if there's no real change in the hemoglobin and exactly where you are, where you were before you started the treatment, don't call that a response.

Dr Komrokji: Right. Last couple of questions I'll just answer quickly, just because the audience had asked them. Other mutations other than SF3B1 impact response to luspatercept?

In the PACE, which was the phase II of the luspatercept, actually other splicing mutations were favorable; that patients had higher response. We've presented data on the mutations and the mutation burden from the COMMANDS both in the RS-positive, RS-negative.

There were no negative predictors. Probably the exception was 7 mutation, which is very rare and typically seen in the MDS MPN. There were no negative predictors. The mutation burden sometimes could actually favor luspatercept. Once you get more than 1 mutation, ESA responses were less. Once you get to 3 or 4 mutations, both drugs did not perform well.

The last question was, we don't have imetelstat in our country. Can we use HMA instead?

Yes, of course. If there is no other option, as Amer mentioned more than 1 time, and I think it's a clear message, always try to push them as the last resort of treating those patients. ESA, lenalidomide, even in the non-del (5q), if the patients have anemia only, you can use lenalidomide. If you are in the corner, then obviously you can use HMAs.

If there are concomitant thrombocytopenia, neutropenia, they may force your choice for HMA as a second line sometimes. There are data on shorter courses, actually, you can get by 5 days or even 3 days.