Dr. Amer Zeidan:

Hi, everyone. Good morning, good afternoon, good evening, wherever you are. Welcome to another episode of VJHemOnc MDS Sessions. We are going to be talking today about a very interesting subject, but very difficult subject at the same time, which is TP53-mutated MDS. This is an entity that has been quite challenging, both in all myeloid malignancies, in fact including MDS and AML.

Today, our discussion is going to focus more on the MDS side of TP53. And we will review some of the controversies that connect to the prognostication and classification as it relates to TP53, and then cover some of the therapeutics, with a particular focus on ASH 2023 presentations. And to do this today, we have a very exciting panel that’s joining me.

So I have Dr. Naval Daver from the MD Anderson group, as well as Dr. Matteo Della Porta from the Humanitas University in Rome, Italy. And I’m joined by one of our brighter fellows at Yale University, Dr. Tariq Kewan, who had the most challenging task of putting all the data together for us in a presentation. So Tariq, we’ll do this over 30 minutes and then we will have the panel discussion. Again, this is Amer Zeidan from Yale University, and thank you so much for joining us. Tariq, you can proceed.

Dr. Tariq Kewan:

Yeah. Thank you very much, Dr. Zeidan, and thank you for having me today. So I will discuss the updates in myelodysplastic syndrome neoplasms with TP53 mutation from ASH 2023 updates. So as you know, mutation in TP53 is very common and present in approximately 10% of patients with MDS, and represent a unique subtype with poor outcomes and prognosis. From the data presented by Grob et al, adverse risk ELN 2017 with TP53 mutation, had worse outcome compared to wild type TP53. However, no differences was noticed in overall survival observed between acute myeloid leukemia and MDS excess blast with TP53 mutation.

So when we’re looking at TP53 mutations, the characteristics of these mutations are very important to demonstrate outcomes in MDS patients. Several factors including, the type of TP53 mutation, locations and hotspot involvement, variable allele frequency and the estimated allelic burden or the allelic status can significantly affect the prognosis among patients with TP53 mutations. In the criteria published by Bernard et al, monoallelic TP53 mutation was defined by a single TP53 mutation with a VAF less than 50%, and biallelic mutation was based on multiple TP53 mutations or a single mutation with Del-17p or copy-neutral loss of heterozygosity.

And as you can see in the Kaplan-Meier panels in different settings, whether at the de novo setting or treatment-related, patients with multi-hit TP53 mutation had worse overall survival compared to single-hit TP53 mutation. Important to mention that VAF percentage of less than 50% alone is not necessarily indicating monoallelic inactivation and may reflect multiple heterogeneous subclones with heterogeneity at different allelic status among those patients.

Based on that and given the importance of TP53 mutation impact on outcome, both the ICC 2022 and the WHO fifth edition included TP53 mutation in their classification. According to the ICC, patients with a single TP53 mutation with more than 50% VAF or with Del-17p or copy-neutral loss of heterozygosity were assigned to have biallelic mutation. And also patients with multiple TP53 mutation with a VAF more than 10% at least were considered to have biallelic inactivation.

However, multiple differences between both classification schemes are present, and I would like to highlight three important points. The first point is that TP53 mutation with complex karyotype is considered biallelic equivalent per the ICC but not the WHO classification. The second point is that ICC exclude TP53 mutation with VAF less 10%. And lastly, ICC stratified TP53 mutation with 10% to 19% of blast to be considered as MDS/acute myeloid leukemia.

Based on that, I’ll start with data on ICC 2022 versus WHO fifth edition from the ASH 2023. This abstract looked at the comparison of the WHO and the ICC classification of the myeloid neoplasm harboring TP53 in a cohort of 536 patients. 71% of these patients had MDS. Importantly, the discordance between the two classification schemes was highlighted in three subcategories. The first one is in patients with VAF less than 10%, where the allelic status of TP53 was defined according to WHO, not ICC, and those patients with biallelic state, according to the Kaplan-Meier on the right, had worse adverse events or overall survival compared to the patient with single-hit.

Second, based on the bone marrow blast cut-off, the median overall survival of TP53-mutated patients with 10% to 19% of blasts was worse compared to the patients with 0% to 9% of blasts. However, the allelic status did not provide any significant finding among patient with 10% to 19% of blasts. And finally, based on the definition of multi-hit equivalent per the ICC, patients with TP53 mutation with complex karyotype had significantly poorer survival outcomes compared to the patient without complex karyotype. Overall, the study provided evidence that poor prognosis with multi-hit TP53 and VAF less than 10% is underestimated by the ICC. However, on the other hand, WHO is not providing much estimation on a blast percentage 10% to 19%, and did not account for the presence of a multi-hit equivalent.

In this study, the mutational status of myelodysplastic syndromes and AML patients according to the WHO 2022 ICC criteria in a Korean multicenter cohort was evaluated. 1,264 patients were diagnosed with MDS and TP53 mutation were observed in 9% of the patients, and 90% of them were single TP53 mutation. Among the patients with low blasts, WHO identified 56% of the patients to have biallelic TP53 compared to 75% per ICC. Among the patients with MDS high blasts, WHO identified 63.3% to be biallelic versus 87% in the ICC. And as you can see in the table, patients with biallelic mutation were more enriched with 17p deletion and complex karyotype. This study showed that ICC classification has a tendency for over-diagnosis of biallelic TP53 mutations compared to the WHO criteria.

In this study, the clinical impact of the new bone marrow blasts per the ICC on the patients with a mutated TP53 mutation was evaluated. Only patients with available TP53 mutation were evaluated and 107 patients were actually assessed in this study. And as you can see from the Kaplan-Meier, patients with MDS TP53 and MDS/AML TP53 had comparable survival outcomes.

However, patients with acute myeloid leukemia with blasts more than 20% and TP53 mutation have a worse overall survival outcome compared to the MDS patient. This study highlighted that the new proposed 10% blast cutoff is not adding much clinical value among patients with MDS according to the ICC guidelines. However, the cohort sample size was low with 107 patients.

So showing this discrepancy between both classification schemes, this study by Dr. Della Porta’s team tried to use an adopted data-driven machine learning algorithm to provide a harmonization roadmap between the ICC and the WHO classification. As you can see, 7,000 patients with MDS were included, and they identified different clusters according to the machine learning algorithm. The cluster of the highest hierarchical importance was characterized by biallelic inactivation of TP53, and the biallelic TP53 mutation were defined as the presence of two or more TP53 mutations or one mutation with copy-neutral loss of heterozygosity, as you can see in the first panel. Most patients assigned to biallelic TP53 cluster had a TP53 VAF more than 10%, consistent with the ICC, in about 78% of them and complex karyotype and 70% of the patient. Assignment to biallelic TP53 cluster was irrespective to the blast count, however. Patients with monoallelic TP53 mutation were actually segregated into other clusters.

Now we’ll talk about the abstracts talking about the impact of TP53 allelic status on outcomes. First, I’ll start with the abstract reported by our group from the VALIDATE database talking about the clinical implications of TP53 mutation and allelic status among high-risk MDS patients treated with hypomethylating agent. The VALIDATE database included high-risk MDS patients treated with frontline HMA from 14 centers, and the study included 816 patients. Only patients with high-risk disease were included and the cohort was enriched with complex karyotype in 76% of the patient. The response criteria for the HMA were assessed according to the IWG 2023 criteria. CR rate for patients with biallelic TP53 mutation was 18% and was similar to patients with monoallelic TP53 mutation. The median overall survival for the patients with TP53 mutations was much worse compared to the patients with wild type- 12 months compared to 28 months. When we define the allelic status based on the Grobi criteria published in Blood 2022, patients with biallelic TP53 had comparable median overall survival compared to monoallelic TP53. However, when defining the allelic status based on the ICC’22, more patients were assigned to biallelic TP53, 92% of the cohort, and those patients had worse overall survival compared to the patients with monoallelic state. This was similar when we identified biallelic mutation based on a VAF cutoff of 23%.

In another study, the TP53 gene allelic status was actually evaluated among patients with isolated del5q, and in this study, the TP53 allelic state was defined according to the Bernard criteria and the study included 682 patients with isolated del5q. Overall, 18.7% presented with TP53 mutation and most of these patients, 73% of them, had monoallelic TP53. The multihit group presented with worse prognosis compared to monoallelic, with a median overall survival of 55 months compared to 73 months. AML evolution at 60 months was also much higher at 40% compared to 25.5%. When they looked at the TP53 monoallelic MDS with a VAF cut-off of more than 20%, those patients had similar prognosis to the patients with multi-hit TP53 mutation. This study showed that multi-hit TP53 status is uncommon in isolated del5q. However, if it presents, it can present with prognostic significance, with worse outcomes and higher rates of AML evolution.

The singlehit TP53 mutation was also evaluated in this cohort of therapy-related MDS. This is a multicenter international cohort of 544 patients with TP53 mutation, and only patients with a VAF of more than 10% were included. And those patients were stratified based on therapy-related myeloid neoplasms and de novo myeloid neoplasms, and then after that, the allelic status was identified according to the ICC. As you can see, the patients with therapy myeloid neoplasms were enriched with multi-hit and multi-hit equivalent TP53 mutations at about 90% compared to 72%, and this was statistically significant. When the survival outcome has been compared, patients with therapy-related MDS with either single-hit or multihit TP53 and patients with de novo MDS with multi-hit TP53 had comparable overall survival. Patients with de novo MDS with single-hit TP53 had a better overall survival of 25 months compared to the other population.

In another study looking at the TP53 variant allelic frequency in patients with isolated del5q, this study included patients with isolated del5q. However, multi-hit TP53 mutations were included to allow unbiased analysis of impact on prognosis. A total of 210 patients were evaluated, 172 with a primary MDS and 38 with therapy-related MDS. And as you can see from the Kaplan-Meier figure that the presence of therapy-related characteristics and the presence of a VAF cutoff of more than 22% were associated with a hazard risk of survival. Therapy-related setting was associated with a hazard risk of 2.3 and TP53 with a VAF more than 22% was associated with a hazard risk of 2.4. At the median follow-up of 3.9 years, death was reported in 45% of the patients, and the study suggested to use the VAF cutoff and the therapy-related setting to provide those patients with a transplant in the front settings.

The impact of TP53 allelic stat on the outcome of patients undergoing allogeneic hematopoietic stem cell transplantation for myelodysplastic neoplasms was evaluated in this multicenter study of 347 patients. 45 of these patients had TP53 mutation, and 21 of these patients were monoallelic and 24 were multiple-hit. And as you can see from the Kaplan-Meier figures that patients with wild type had better prognosis compared to the patients with TP53 mutation, and patients with multi-hit had a worse prognosis compared to the single-hit. When we looked at the overall cumulative incidence of relapse of patients with myelodysplastic neoplasms after stem cell transplantation, patients who had multi-hit had a higher cumulative incidence of relapse at 47.3% compared to 10.8%. A multivariable cox proportional hazard model from the study demonstrated that monoallelic and multi-hit TP53 were an independent risk factor for survival and relapse after transplantation.

Now, I’ll shift gears a little bit and talk about the role of TP53 immunohistochemical staining as a sensitive and specific global tool for rapid identification of TP53 mutation among patients with AML and MDS. As you can see in panel A, patients with wild type TP53 will have a dim staining for TP53 protein. However, patients with missense mutation will have a bright staining in about 20% of the cells. In panel C, patients with in-frame deletion or nonsense mutation will have non-staining, helping to identify different patterns of mutation among TP53 newly diagnosed patients. This abstract looked at the early identification of TP53 mutation and MDS neoplasm and AML via point-of-care TP53 immunohistochemical staining. The concordance between the IHC and the mutational testing as a gold standard was used. The study included 42 patients with MDS and showed that the sensitivity of IHC can reach up to 68.1% with a specificity of 93.3%. This study suggested that the high specificity of the IHC allows it to be used as a roll-in test and additional workup could be done and performed to assess the link between TP53 staining pattern and the allelic status mutation.

Now, I will talk about the management of high-risk MDS with TP53 mutation. So we know that according to the current standard of care, patients with high-risk MDS who are eligible for bone marrow transplantation can proceed directly or undergo disease optimization with HMA or AML-like therapy to provide allotransplantation as the only cure options for these patients. However, patients who are ineligible for a transplant can be treated with hypomethylating agent or HMA-based combinations.

Now, I will talk about the allogeneic hematopoietic stem cell transplantation in high-risk MDS patients with TP53 mutation. Now we know that from recent data, that allogeneic transplantation can improve outcomes in TP53 mutation MDS. In this recent study, 309 patients with high-risk MDS were included. 29% of them had TP53 mutation, and 55% were multi-hit. As you can see from the survival curve, patients with TP53 mutation had the worst overall survival compared to the patients with wild type. Similarly, the cumulative incidence of relapse or progression to acute myeloid leukemia was higher in TP53 multi-hit compared to single-hit and significantly higher than the wild type. However, when the hematopoietic stem cell transplantation was considered as a time-dependent covariate and a multivariate regression model, patients with TP53 mutation who underwent the transplantation had improved overall survival compared with non-transplanted patients. The overall survival at three years was 23% among the transplanted patients versus 11% in the non-transplanted patients.

This abstract looked at the impact of the pre-transplant molecular and cytogenetic remission on outcomes of allogeneic transplantation in MDS patients. The measurable residual disease in this study was measured by a commercial myeloid NGS panel, and FISH cytogenetics performed on peripheral blood or bone marrow aspirates within three months prior to the transplant were included in this analysis. As you can see, 69 patients had a residual disease, MRD-positive, at the pre-transplant level, and at the time of diagnosis, 16 of these patients were TP53-positive. TP53 mutation in this study was the most commonly associated with relapse and eight out of the 24 patients, almost one-third of them, with a median overall survival of about 10 months. The median overall survival was not reached in the MRD- negative patients as shown in the Kaplan-Meier on the right, however, it was 14.1 months in MRD-positive patients.

This study showed favorable outcomes for high-risk MDS and oligoblastic AML with MDS-related changes with reduced intensity allogeneic transplantation followed by post-transplant cyclophosphamide, but not in the TP53 and the RAS-mutated patients. Patients with 5% to 30% of blasts and IPSS of more than or equal to three who underwent bone marrow transplantation with PTCy were included in this study. Overall, 94 patients were included and two-thirds of them underwent haploidentical transplantation, and almost all of them received reduced-intensity conditioning. The median time from diagnosis to transplant was around eight months, and as you can see from the Kaplan-Meier, patients with TP53 mutations and RAS mutations were associated with worse overall survival outcomes.

This study looked at the final results from a Phase I study of briquilimab, which is an anti-CD117 monoclonal antibody, in combination with non-myeloablative conditioning, that show a durable remission in older adult patients with acute myeloid leukemia in complete remission and MDS but not with TP53. Briquilimab is a first-in-class monoclonal antibody that inhibits stem cell factor by binding to CD117 and depleting hematopoietic stem cells in combination with the standard non-myeloablative conditioning of low-dose irradiation and the fludarabine. It has been evaluated in older adult patients with AML and myelodysplastic syndrome.

This study included 32 subjects, among of them 16 were diagnosed with MDS. There were no infusion toxicities and no briquilimab treatment-related adverse events reported. However, if you can look at the figure on the right, five of the 16 MDS subjects had TP53 mutation, three of these patients had persistent disease or progressed/relapsed prior to one year post hematopoietic stem cell transplantation, and the other two patients demonstrated molecular relapse at one year after transplant. The study showed a durable response with the briquilimab among patients without TP53 mutations since all these patients relapsed at one year after transplantation.

Now, I will talk about the hypomethylating agents and combinations and novel therapeutics in high-risk MDS patient with TP53. I’ll start with the impact of the type of the hypomethylating agent used on the outcomes of high-risk MDS patients from the VALIDATE database. The aim of this study was to assess the clinical outcomes of the high-risk MDS patients from the VALIDATE database who were treated with a hypomethylating agent regimen in terms of overall survival and the treatment response. The cohort was really enriched with adverse risk cytogenetics, with 38% of the patients having complex karyotype and 36% with TP53. Overall, 76% of these patients were treated with HMA monotherapy, 56% with azacitidine and 20% with decitabine, and 24% of them received an HMA-based combination. Importantly, there was no significant differences in overall survival or overall response rate according to the IWG 2023 criteria between the azacitidine and the decitabine-treated patients in the adjusted analysis.

This trial looked at the efficacy and the safety of venetoclax in combination with azacitidine for the treatment of patients with treatment-naive high-risk MDS. Adult patients with de novo treatment-naive high-risk MDS defined according to IPSS/IPSS-R more than or equal to 1.5 and more than three is respectively included. All patients should have bone marrow blasts less than 20% and no prior therapies were allowed. And the primary endpoint was CR per IWG 2006. Important to mention that the study did not report specific outcomes among patients with TP53 mutation. However, the CR was achieved in 30% of the patients and the modified ORR was achieved in 80% of the patients and 37% of the patients achieved molecular marrow CR with hematologic improvement. The important side effects included infection in 57% of the patients and neutropenia in around 49% of the patients.

This study looked at the azacitidine combination with lenalidomide in treating high-risk myelodysplastic syndrome. It’s a single-center Phase II trial, looking at the azacitidine, 75 milligrams per square meter per day on days one to seven, administered in combination with lenalidomide, 10 milligrams per day on days one to 21, every 28 days for patients with high-risk MDS. The primary endpoint was overall response rate, and secondary endpoints included overall survival, progression-free survival and adverse events. A total of 41 MDS patients were enrolled and 32 patients were evaluable for efficacy. And as you can see in the table, 72% of the patients achieved overall response rate and 22% of the patients had CR. The median overall survival was not reached and the median progression-free survival was 12 months. Importantly, 27 patients had at least one mutation and seven of these patients had TP53. And as you can see in panel P, all these patients with TP53 mutation achieved CR or molecular CR, however the duration of the response is not known. Another important fact from that study is that there were VAF changes of the TP53 mutations when compared between before treatment and after treatment.

This trial looked at the preliminary results of a Phase II study of the IMM01 combined with azacitidine as a first-line treatment in adult patients with high-risk MDS. IMM01 is a component signal-regulatory protein alpha, SIRPalpha, monoclonal antibody that activates the “eat me” signal to induce a strong antibody-dependent phagocytosis. This study was an open-label multicenter Phase II study that evaluated safety and efficacy of this medication in combination with AZA as the first-line treatment in patients with untreated high-risk MDS. Only high-risk patients with IPSS-R more than 3.5 who are not eligible for transplantation or intensive chemotherapy were actually enrolled. Efficacy was assessed according to the IWG 2006 criteria, and 54 patients were enrolled. Among the 22 efficacy evaluable patients who received the initial treatment more than four months, the overall response rate was 82% and about 36% of these patients achieved CR. Among the 17 efficacy evaluable patients who received initial treatment more than or equal to six months, the overall response rate was 88% and the median duration of response was not reached. Biomarker data from the trial showed a mutation burden of several markers including TP53, however detailed data is not presented. And the most common adverse event included leukopenia in about 82% of the patients, thrombocytopenia in 70% of the patients, and neutropenia in 66.7% of the patients.

Lastly, the safety and the pharmacodynamics and anti-tumor activity of SL-172154 as a monotherapy and in combination with azacitidine in relapsed/refractory acute myeloid leukemia and high-risk myelodysplastic syndrome was evaluated in this study, The SIRPalpha CD40 ligand protein is a bi-functional fusion protein consisting of SIRPalpha domains linked to CD40 ligand domains, and demonstrated improved anti-tumor activity in comparison to the naked CD47 blocking antibody in a preclinical study. This was a Phase I dose-escalation cohort that compared SL-monotherapy to SL-AZA combination in patients with high-risk MDS or acute myeloid leukemia. Only patients more than or equal to 18 years old with relapsed/refractory MDS or AML were eligible. However, the cohort included untreated TP53 MDS patients and were eligible to be enrolled in the SL-AZA cohort. 37 patients were enrolled, 19 in the monotherapy cohort and 18 in the SL-AZA cohort, five had relapsed/refractory MDS and five were untreated TP53-mutated MDS. Infusion-related reactions was the most common related side effect in both SL-monotherapy and the combination therapy reported in 68% and 44% respectively. SL-monotherapy induced elevation in serum interleukin levels and cytokines, and there was a greater response at the dose of three milligrams per kilogram compared to one milligram per kilogram. However, the responses were similar among patients receiving the three and six mg per kg. In four evaluable patients with untreated TP53 mutation MDS, the response rates were complete response in one patient, one patient had a marrow CR and two patients had stable disease. Two of these patients proceeded to allogeneic hematopoietic stem cell transplantation.

So, as a conclusion and the future perspective, myeloid neoplasms with TP53 mutations continues to have dismal prognosis and better novel therapies are needed. Determining the most effective timing, combinations, and order of therapeutic agents remains a critical factor for the treatment of patients with TP53 mutations. IHC testing for TP53 is beneficial for informing clinical treatment options and decisions, and studies are needed to evaluate the possible role of this model with other models in allelic status determination. Finally, defining the TP53 allelic state is very important for the prognosis and very challenging at the same time, and applying multiple modalities is needed for further characterization of this important, I would say, option. Thank you.

Dr. Amer Zeidan:

Thank you so much, Tariq. This was a truly outstanding tour de force and all the new data from ASH. And I want to start the discussion, Naval and Matteo, I think to divide it into three major sections, each one around 10 minutes. So I think the first section is going to be on the impact to the clinical practice. What do physicians do today when they see these patients? The second one in issues related to clinical trial design and how should we think about trials when thinking about TP53-mutated myeloid neoplasm. And the last thing is about the new therapies and what do we think are the most promising.

So starting with clinical practice, maybe I start with you, Matteo, since you have done a lot of the work that connects the classification and the prognosis in terms of those patients. And I think one of the better words to use is dizzying because when you go through all these presentations that Tariq has mentioned, you feel like the data is all over the place in terms of does the blast count matter? Is it five to nine, is it 10 to 19? What is the multiallelic versus monoallelic cutoff? Is it 50%, 22%, 40%? And how does that affect the decision? First, how to treat the patient. Would you add venetoclax as standard of care or just go and give HMA? And the issue of the transplant, we have gone through this period of nobody wanted to transplant these patients because all of them were, I guess, sought to be not cured, and then we know that some of them are, but we cannot figure out who are those ones. So maybe you can give us your sense about how do you actually use the issue of TP53 when you treat patients in your clinic today?

Dr. Matteo Della Porta:

Yes. Very important questions, Amer. So I think that, at least in my mind, there are three major topics to be faced in terms of clinical implementation of the recognition of patients with TP53 dysfunction with MDS. The first one refers to the need, to the urgent need, to harmonize the classification of the patient between WHO and ICC because we observed in a large cooperative international cohort that a disagreement in the assignment of patients with TP53 mutation is around 25% in these two already available systems. So, my first point would be, in order to do a more reliable implementation of the recognition of the patient we need for harmonizing the classification.

The second issue is about, at least in my mind, the definition of the prognostic relevance of monoallelic TP53 dysfunction, because according to the first study provided by the IWG committee, apparently the dismal outcome was most associated with biallelic dysfunction. But in a more recent report, also including the post-transplantation outcomes of MDS patient, also monoallelic TP53 is associated with reduced prognosis and increased risk of disease relapse. So, we have to refine and to provide more clear evidence about this patient group.

The third topic, in my mind is what is the, let’s say, precision in which we are defining TP53 dysfunction in MDS by genomics. In other terms, are we currently able to capture all the patients with MDS and TP53 dysfunction? My answer is potentially no. And maybe we have to move to more functional-driven studies in order to check if other, let’s say, forms of TP53 dysfunction, not necessarily correlated with the mutation of the TP53 gene, are, let’s say, providing a clinical role in myeloid neoplasms.

Dr. Amer Zeidan:

Thank you, Matteo. I’ll probably get back to you again in terms of how do you apply the knowledge. I think you raised the questions very well, but how do you apply it when you are seeing a patient in the clinic. But maybe I’ll go to you, Naval, first. MD Anderson has led the field on the use of decitabine in myeloid malignancies in general, but particularly in MDS and AML. And there has been all this literature, initially starting with the 10-day of decitabine and then subsequently some data with the oral decitabine/cedazuridine, that there might be some differences in activity for those patients with TP53. I never understood from a pathophysiologic point of view why would that be the case. And several of the studies, including the one that Tariq mentioned from the VALIDATE consortium, doesn’t seem to suggest there’s a difference between the two HMAs. So the same question I posed to Matteo, outside of a clinical trial, when you see those patients, how do you approach them in terms of who is considered really TP53 versus not in terms of your choice of the HMA and taking them to transplant?

Dr. Naval Daver:

Yeah. I think what Matteo said is the key. I think we need to harmonize what we consider as functionally real TP53, or you can call it real high-risk TP53, because you see all the data that Tariq very nicely presented, and if you’re not, forget about being an alter expert, even if you’re just an expert in MDS, I don’t think you’ll get the same answer if you ask many people. So we’ve been looking at this, and I think in AML for example, it’s much more clear what is considered biallelic is very well-defined now, so you have to have either one mutation above 40%, Moffitt and our group showed that, or you have multiple mutations that add up to above 40% or you have a mutation with a loss of heterozygosity. So we don’t really just consider any two hits as automatically being a biallelic equivalent.

Now in MDS, I think it may or may not be different, and I do think that the ICC is over-capturing patients because if you just say 10% cutoff regardless of multi-hit, you’re going to get some patients who actually may have a not so bad outcome, none of them are favorable but not so bad outcome. And I think that is really, clinically, the key message because you don’t want to miss those patients who may have a lower VAF and may actually have a good survival with traditional therapy HMA or if it’s above 10%, one could consider intensive chemo, whatever the practice may be, and go to transplant and you could have a tail to the curve in those groups.

So that’s really our focus. So if I see a new TP53 in the MDS setting, and let’s say they have a low VAF, less than 20%, we’re actually doing this right now in our dataset and we’ll be publishing soon. So less than 20% seems to be predicting for patients who have a better outcome and they don’t have an associated chromosome 17p or complex karyotype. Now you’ll see that this is going to get to a small group, this is probably 15% to 20% of patients, but those 15% to 20%, we do not want to deprive them of a potential curative outcome and we also don’t want our transplanters, because this is the big battle, many of our transplanters, they just see TP53 and they say no transplant. And at least in Anderson we’re still doing it, but other centers, I’m getting patients from many other centers, where they haven’t even looked at the VAF, the associated cytogenetics, single hit, multi-hit, just say TP53, no transplant. So today I think we should look at the VAF-associated cytogenetics and those that have a low VAF, whether you take 20% or 23%, other groups have looked at those and can be curative, I would give them HMA-VEN potentially if they are older, if they’re 10% and higher and we’re considering intensive chemo, that’s another option. But for everybody else, we’re predominantly HMA. So those who have a high VAF above 20%/25%, those who have associated complex cytogenics or chromosome 17p, HMA is what we’re doing. We’re not doing intensive chemo anymore in those patients. And there, more and more the data, and we’ll be publishing soon our data, the venetoclax does not seem to help much. In fact, if anything numerically, it’s showing the detriment because of early mortality inflammation, more severe pancytopenia. So there I would just go for HMA alone that seems to be the most reasonable.

And I agree with you. I don’t think there’s cumulative data that suggests decitabine overall is better than azacitidine. I think the oral decitabine abstract you mentioned is interesting, the one from Savona and Guillermo Garcia-Manero I think last year at EHA or the ASH before, where they showed this 20 plus month survival. But again, it’s the same problem: how many of those patients were monoallelic, less than 20%? How many of them had higher VAF? We don’t know. So I think it’s fine to try oral decitabine, but I don’t necessarily think it’s better per se yet than IV AZA or IV decitabine.

Dr. Amer Zeidan:

Yeah. Thank you so much. I think there’s a clear agreement that the issue is I think very complex, and I have the same concerns like you, is that there has been somewhat the message in the community or especially for some of our transplant colleagues has been that TP53, those patients are not potentially curable, and I think that has been a disservice to some of the patients because I do think there’s a clear tail at the end of that curve. The problem is that we cannot seem to identify very well those 20% of patients who seem to benefit from transplant, and I think we really need to provide better guidance, hopefully, using some of the data sets that you just mentioned.

So taking this to the second, I guess, focus for today, which is the clinical trial design, and I think this has been a painful issue, I would say in particular in high-risk MDS because of the repeated failure of Phase III trials. Now, with sabatolimab and magrolimab, we still don’t have the data publicly released yet. The press release has indicated negative outcome in OS, but of course we don’t know the specifics within the TP53 patient population. But there has been a general concern in MDS that every single Phase III trial seems to have 20% to 30% at least of those patients and whether they could be affecting the overall outcome of the entire trials. But on a bigger picture, the issue of do you stratify by them, do you study them separately, do you stratify by them and cap how many patients who have TP53? And if you want to do all of this, how do you do it when we cannot seem to agree on what is actually a TP53? Do you use the local assessment? Do you use central assessment and have to wait probably at least a week to 10 days depending on where it gets done centrally? So maybe I’ll start with you, Naval, and then get to you, Matteo, about your thoughts about this.

Dr. Naval Daver:

Yeah. I mean, I think it’s a very complicated scenario, which I think now we all agree it is very complicated, especially four years later, as you me, Matteo all know, with many drugs that we all personally were very involved with and showing signals that were very good in Phase I and II studies. Now, whether that’s because of the centers, the investigators, or because of the drugs themselves, it’s always hard to tease out, you know the management of toxicities, et cetera. But I think you have to look at AML, I guess, as an example. And even in AML, if you look at most of the approvals, these were for particular targeted subsets. The FLT3 inhibitors, IDH inhibitors, now the menin inhibitors, CPX, except for venetoclax, these were not across the board. And even with venetoclax, we did have 24% of patients who are TP53 in the VIALE-A, so it’s not that they were stratified for excluded, but we also had a very favorable and venetoclax-sensitive population, which included NPM1, IDH1 and 2, RUNX1. And so I think in the balance, the benefit that was derived from those favorable overcame the negative impact of TP53, because if you look at the TP53 outcomes, they’re exactly the same, 5.6 months. There’s no difference. I think in MDS, the problem is you have the TP53, it’s probably similar or a little bit more, and we don’t have those de novo AML-like populations that can respond really well to venetoclax. And I think that is the challenge, and this is probably the challenge for other drugs as well.

So I think going forward, what we probably want to do is, yes, you want to stratify, but I think we need to get a better understanding early on what the signal really is in TP53. Because you see these ORR rates, and we have reported them as well, everybody does it because it’s a competitive field, but really what does it mean to have just a marrow CR or some HI. So we probably need to really focus on durability, molecular, and then outcomes with and without transplant. And probably have either a preplanned stratification, which is complicated because then that means you need to have the molecular data before randomizing those patients if you’re going to have it really preplanned, or at least some way to have a discussion with the regulators as to how we could protect the study even if we did not get a good signal or even a negative signal in TP53 for the remaining population. So we’re all hopeful with the venetoclax that it will be overall positive, but for some reason we’re all nervous as well, and if it’s not, I still think we need to look at each population and we shouldn’t say venetoclax is not good, because there will be populations, the IDH, the RUNX1, the ASXL1, the splice where it benefits, and there may be others where it doesn’t, and I think we need to move with MDS, like in AML, to more focused molecular outcomes than the whole group. But, yeah.

Dr. Amer Zeidan:

Matteo, I give you the same question. How do you advise sponsors who are looking at high-risk MDS, how to tackle the issue of TP53 in terms of their trial design?

Dr. Matteo Della Porta:

Yes. So a very, let’s say, difficult question. Basically, I do agree with what Naval discussed before, but I would like to, let’s say, add some comments from my side. The first one is that the increasing scientific and clinical evidence may suggest that TP53-mutated patients within all myeloid neoplasms can be maybe considered as a separate category. Also considering that blast count is not so effective in stratifying the prognosis across different patients with TP53 mutation. So maybe this clinical evidence suggests that we should move to either modifying the design of clinical trials by providing a separate analysis for TP53 patients since the beginning of the study design or maybe to cumulate all myeloid neoplasm with the TP53 mutation as a specific category in which we have to provide specific and innovative clinical trials.

In this sense, one of the, let’s say, points of contact in terms of biological relevance of TP53 mutation between MDS and AML is that in both cases, several studies reported an immune-specific dysregulation for these patients that is consistent with a leukemic escape phenotype of immune niche. So maybe this is both operational in order to consider these two populations as the same category from a clinical perspective and to provide the rationale for new biological targets for TP53-mutated patients.

Dr. Amer Zeidan:

Yeah. No, thank you so much, Matteo. I think both of those are great concepts. I think in terms of dealing with TP53 as one entity that is molecularly defined, I guess, maybe you can do that within the context of aggressive myeloid neoplasms, meaning that if you have more than 5% blasts, whether it’s AML, whether it’s MDS, whether it’s MPN or accelerated MPN, I think the regulatory challenges in terms of the standard of care on those therapies, but the FDA actually has set up a similar approach using mutation agnostic… or sorry, disease agnostic, basically approvals with some of the immune checkpoint inhibitors that you mentioned.

And I think, actually, I’ll pivot to our last segment of the day using something you just mentioned, which is the fact that TP53-mutated MDS, and AML for that matter, has been associated with an immune escape signal, and this is something that Naval and I and others have tried to use therapeutically for years now, but unfortunately, we still do not have very clear therapeutic success. I think discussions later this year after some of the data of those trials comes out are going to be very interesting. But based on what we currently, I guess, can discuss and know, Naval, in terms of seeing all of these PD1, PDL1, CTLA4, potentially some of the CD47s and, I guess, some of the TIM-3 apparently not working, at least again across the board, not necessarily only in TP53, we’ll have to see the TP53-specific data and looking at the data that are represented on some of the novel agents, some of which are we are also involved in with some of those novel CD47 agents or other immune-targeting approaches, do you view there are major differences that could still be the therapeutic window? You mentioned the issue of is it the disease, is it the drug, is it the center, is it the investigators? It’s many, I think, variables that could affect the outcomes of individual studies.

Dr. Naval Daver:

Yeah. Yeah. I mean, as a very simple answer level, I would say yes. I still think that immunotherapeutics are probably going to be the most likely thing that’ll work in these very high-risk, whether it’s TP53 inversion three, chromosome 7, but I do think we probably need to get maybe more myeloid-specific immunotherapeutic agents. We’re borrowing a lot of things from solid tumor, PD1, PDL1, CTLA-4, and sometimes these things work, but actually if you look at it, most of the times they have not, even in AML, ALL, we have to develop specific drugs.

So I think the CD47 SIRPalpha story, if you look at the preclinical data, the science behind it with Irving Weissman, Majeti, Mark Chao, it’s amazing. I mean, very few drugs have had this level of science before going into clinic. So, I still think that this pathway is very relevant, I would be very surprised if the entire pathway and all the preclinical work was not relevant, but I think it may be about the drug. And as you said, we cannot discuss, but at EHA and upcoming meetings, there will be data. And I think the safety, for example, anemia with magrolimab, of course did play a much bigger role than we thought and expected compared to the single arm in big academic centers. So I think if you have a drug that targets SIRPalpha, for example, that does not have expression on the RBC cytoskeleton, does not have the anemia baggage signal with it, and potentially maybe can bring in some innate immune activity from T-cell engagement, like one of the bispecifics we’re working with the SL-17 from [inaudible 00:50:02]. There’s others, DSP107 from a company in Israel. I think this could be still something interesting.

I always like to give the example of FLT3 inhibitors. The first FLT3 inhibitor, lestaurtinib, Phase III randomized study was completely negative, and everybody at that time started feeling, oh, FLT3 is not going to work. 10 years later, of course, many people don’t know that history and four FLT3 inhibitors have made it. The same is true with JAK inhibitors. Four of them went to development, only ruxolitinib initially got approved, then 10 years later we have others. And even with BCL2 inhibitors, there were three of them before venetoclax, oblimersen, navitoclax, obatoclax, but then it took the right drug with the right combination of safety and efficacy. So I think I would not probably give up on the field yet. I think we have better compounds, better targeting, less toxicity, potentially innate immune engagement, and we should look at that.

But I personally feel, in the future it’s probably going to be allogeneic or off-the-shelf approaches that really work because as you said, Amer, the T-cells themselves have the mutation, they’re impacted, they’re functionally not as good. So I think if we’re really going to succeed, it’s probably going to be some form of external immune replacement, whether it’s with T-cells, CAR T-cells, NK-cells, combinations. I think that is what, five to eight years down the line, I feel may be really the success with these TP53-type disease.

Dr. Amer Zeidan:

Yeah. Thank you, Naval. Clearly, we’ll make sure to bring the whole group after later in the year to discuss, I think, once we have more specific data to review. Matteo, what’s your perspective on, I guess, the difficult broad of the high-risk MDS drug development?

Dr. Matteo Della Porta:

Yes. Basically I do agree with Naval’s comments. So the immunotherapy is really promising in this specific subset of patients, because of the results of many translational evidence for specific immune dysfunction. But again, as underlined by Naval, so we have to study more specifically the possible efficacy target for this type of treatment in the setting of a liquid tumor, because we know very well that in terms of niche and in terms of relationship between immune cells and tumor cells, there is a big difference in a solid tumor such as, let’s say, breast cancer or even in lymphomas with respect to the relationship between tumor cell and immune system in a liquid tumor. So I think that, based on this very simple concept, we have to move to more deep and specific immunological studies for MDS with the TP53 mutation in order to check for what is the most promising target for this specific clinical scenario, not just to move evidence from solid tumor to leukemia. That is not so effective in my mind.

Dr. Amer Zeidan:

Yeah. Thank you so much. I think this is a fantastic summary for the direction we should be taking forward. Tariq, I want to give you the chance to give any final thoughts before we conclude today. I think you have done an outstanding job in overviewing all these drugs and where they are, but any concluding remarks you’d like to get?

Dr. Tariq Kewan:

Yeah. I think I agree with what all have mentioned, mainly about the identification of the allelic status and identifying those patients with multi-hit, is still not clear. Harmonization of the WHO and ICC is really needed. And regarding the treatment options, I do agree we’re using a kind of diverse treatment, not personalized or presaging treatment among those patients, and definitely targeting specific markers that can predict response is going to be more beneficial for these patients.

Dr. Amer Zeidan:

Yeah, thank you. And clearly, I think for the issue of understanding the specific relevance of different subsets and where do you draw the cutoff between monoallelic and biallelic, et cetera, I do think large collaborations internationally are very important and this group here on the call have been very collaborative. Many of our colleagues in a rare entity like MDS, where it’s very difficult to get a lot of knowledge just based on limited number of centers or single-center data. So I think this is also very important to continue to push the field forward. So thank you so much for joining us. Thank you so much for the audience, and look forward to seeing you on a subsequent episode of MDS sessions.

Dr. Naval Daver:

Thank you.

Amer Zeidan: Consultancy: Schrödinger, Notable, Chiesi, Mendus, Otsuka, Foran, Syros, ALX Oncology, Kura, Seattle Genetics, Servier, Boehringer-Ingelheim, Jazz, Orum, Geron, Syndax, Gilead, Zentalis, BeyondSpring, Incyte, Pfizer, Celgene/BMS, Janssen, Amgen, Agios, Daiichi Sankyo, Tyme, Taiho, AbbVie, Ionis, Takeda, Genentech, Epizyme, Lox Oncology, Novartis, Astellas, BioCryst, Regeneron; Honoraria: Schrödinger, Notable, Chiesi, Mendus, Otsuka, Foran, Syros, ALX Oncology, Kura, Seattle Genetics, Servier, Boehringer-Ingelheim, Jazz, Orum, Geron, Syndax, Gilead, Zentalis, BeyondSpring, Incyte, Pfizer, Celgene/BMS, Janssen, Amgen, Agios, Daiichi Sankyo, Tyme, Taiho, AbbVie, Ionis, Takeda, Genentech, Epizyme, Lox Oncology, Novartis, Astellas, BioCryst, Regeneron; Research Funding: Foran, Shattuck Labs, Astex.

Naval Daver: Consultancy: Daiichi Sankyo, Bristol-Meyers Squibb, Pfizer, Gilead, Servier, Genentech, Astellas, AbbVie, ImmunoGen, Amgen, Trillium, Arog, Novartis, Jazz, Celgene, Syndax, Shattuck Labs, Agios, Kite, Stemline/Menarini; Research Funding: Daiichi Sankyo, Bristol-Meyers Squibb, Pfizer, Gilead, Servier, Genentech, Astellas, AbbVie, ImmunoGen, Amgen, Trillium, Hanmi, Trovagene, FATE Therapeutics, Novimmune, GlycoMimetics, Kite.