Impact of p53 in Acute Lymphoblastic Leukemia (ALL)

Podcast Transcript

Dale Shepard, MD, PhD: Cancer Advances: a Cleveland Clinic podcast for medical professionals exploring the latest innovative research and clinical advances in the field of oncology.

Thank you for joining us for another episode of Cancer Advances. I'm your host, Dr. Dale Shepherd, a Medical Oncologist, Director of International Programs for the Cancer Institute and Co-Director of the Sarcoma Program at Cleveland Clinic. Today I'm happy to be joined by Dr. Moaath Mustafa Ali, a Medical Oncologist and Hematologist here at Cleveland Clinic. He was previously on this podcast to discuss real-world data for the treatment of small-cell lung cancer with atezolizumab, and that episode is still available for you to listen to. He's here today to discuss the impact of p53 mutations in patients with ALL. So welcome back.

Moaath Mustafa Ali, MD, MPH: Thank you. Thank you for having me.

Dale Shepard, MD, PhD: So remind us again about what kind of things you do here at Cleveland Clinic.

Moaath Mustafa Ali, MD, MPH: So, I specialize in malignant hematology, but I'm very interested in real-world data. Part of my training in master's in public health, I focused a lot on epidemiology, biostatistics, and causal inference.

And I work with large research groups to gather and generate data, observational data, on how the outcomes in different kinds of cancers. And we usually look to different diseases, not necessarily only malignant hematology, but also sometimes even solid tumors. And since I started at Cleveland Clinic for three years, we've generated massive databases for acute lymphoblastic leukemia, acute myeloid leukemia, chronic myeloid leukemia, and even solid tumors, one of the largest database for immunotherapy in the world.

Dale Shepard, MD, PhD: So you've been collecting lots of data and now we're going to be looking for lots of results coming out. So if something tells me, we'll have you back before we get into this specifically about p53 and ALL, let's... You talked about collecting databases, maybe give some insight about kind of things that you need to think about in terms of having comprehensive databases and avoiding biases and things like that.

Moaath Mustafa Ali, MD, MPH: Yes. So, collecting data for observational studies is not necessarily an easy task. It needs a lot of planning and using the right tools to collect the data as well as the right team members. It is always best to be done in a way we include all patients sequentially, so we avoid selection bias. And the best thing is to, for example, start in a specific year. Say, okay, I want to start in 2015 and move forward and not to skip years because that can introduce selection bias and affects any prognostic studies you're interested in.

It does need significant amount of effort. However, if it's done properly, it can generate significant amount of research in a short period of time. The most important thing is that the data is collected comprehensively, in terms of if you need to design another study, it can be easier to do by just reclassifying the comparison groups, which is something we've done for the last ASH meeting as well as ASCO meeting, collect the data comprehensively, analyze different groups. And as long as it's done appropriately, you can generate significant amount of research that is very useful and also can help you confirm the outcomes of clinical trials.

Dale Shepard, MD, PhD: When you think about the data collection, how do you ensure that you don't miss cases in terms of ICD codes? And then, is this just brute force effort to get the data or are you using sort of language models to do searches or how does that work?

Moaath Mustafa Ali, MD, MPH: Yes. So the process is that at any hospital a physician works at, he or she needs to contact the IT department, or the cancer registry, and figure out how to obtain the list of the patients. And you simply start with getting the IRB approval obviously, and requesting patients starting, for example, from 2015 and onwards. And then you start collecting data.

Now, my suggestion or recommendation is to work with the IT department closely because they can help you extract certain data much more accurately. For example, like the age, the gender, ethnicity, when was the last date to follow up and their vitality state, their survival outcomes.

And then subsequently you have to generate a team. And then this team has to be trained very well in terms of to understand oncologic outcomes, how to collect data, how to avoid biases. And then to also train the team very well on to avoid missing data, because missing data actually is very toxic to research and can lead to wrong conclusions. So try to maximize data collection.

And lastly, it needs to be updated frequently and undergo frequent monitoring for the quality of the data. Now using artificial intelligence to collect data, that is something that is gaining more and more momentum. We have not used it in our research yet. That is something I'm interested in and I'm looking forward to the future, but that has not been integrated yet. And that's why one of the reasons creating databases does need extensive amount of energy and effort.

Dale Shepard, MD, PhD: And I guess, one last thing from the sort of how we collect data before we delve into some results and what you've done here in ALL, but you mentioned missing data, and that's something that as you go back and retrospectively it seems like that would be a big problem sometimes because everyone, even if they have the data, they may not document it the same way or document it at all. Are there cases where you kind of find fields of things you'd like to collect to answer questions, but there's just enough missing data you have to scrap the entire, sort of field and category?

Moaath Mustafa Ali, MD, MPH: Yeah, sometimes there are certain variables that there are missing data because that test was not done. For example, like you want to study the impact of a mutation on outcomes in a certain cancer. Now if you go back to 2010 or 2012, the test was not done simply.

And that's something we call, "missing by random," because missing values can be missing by random or a missing due to a certain reason. I'll give you another example. So let's assume you have a physician, Dr. X, he treats patients for some reason, Dr. X does not document ECOG, and most of his patients do not have ECOG.

And then you want to look to survival outcomes. This missing data is not by random, and that can potentially actually bias or confound the results of the study. And this is how missing data can be very toxic.

Now, as a general rule, you need to have less than 10% missing data and anything more than that, it becomes problematic. And as an advice, if you have missing data, try to do multiple imputation. There are several packages in our statistical program, Python, that can do multiple imputation. That's something we always do for our studies to decrease the risk for bias.

Dale Shepard, MD, PhD: Excellent. All right. Well, we're going to shift over to some results from data collected. We're going to talk about p53 mutations in ALL, so maybe as a background, a lot of different people might be listening. What are these mutations and kind of what are the implications for cancers in general and ALL specifically?

Moaath Mustafa Ali, MD, MPH: Yes. So first of all, as everybody know, these cancers with time tend to accumulate mutations. Now mutations can be classified into different categories. Some of them are proto-oncogenes, which a mutation can result in an increase in the proliferative activity of the cancer. It can be also a mutation in a tumor suppressor, where a loss of function can also promote cancer, but not necessarily by increased proliferation, just by loss of the antitumor activity of this gene.

Different genes obviously have different mutations, and the genes can have different prognostic values as well as therapeutic values. And one of the more important genes, the tp53, which we call the guardian of the genome, can be mutated in several of the malignant disorders including solid tumors as well as hematologic neoplasms.

Dale Shepard, MD, PhD: So ALL, specifically these p53 mutations, kind of what was the basis for looking at this particular group and setting upon this study?

Moaath Mustafa Ali, MD, MPH: Yes. So acute lymphoblastic leukemia is a somewhat rare acute leukemia. It accounts for almost one fourth of acute leukemias in adults. And traditionally, we've known for at least 20 years that recurrent cytogenetic abnormalities have strong prognostic implications in acute lymphoblastic leukemia.

The last 10 years, there was a revolution in the prognostication and diagnosis of different cancers, solid tumors, and malignant hematologic neoplasms. And we started using next-generation sequencing to detect mutations. In acute myeloid leukemia that was more used, more prevalent. It's standard. Everybody used to get NGS, at least since 2015 till 2017. It became used almost routinely.

Now in acute lymphoblastic leukemia, that has lagged a little bit. One of the reasons is that acute lymphoblastic leukemia is less common. Like I said, almost 5,000 cases of acute lymphoblastic leukemia in adults occur in the United States compared to 20,000 of acute myeloid leukemia.

And hence, most hospitals have not necessarily adopted using NGS, or next-generation sequencing, in acute lymphoblastic leukemia, for example, in Cleveland Clinic, we started doing that routinely somewhere around 2019. And hence, our understanding for these mutations and how they affect prognosis is still somewhat limited and it's growing with time. And hence, these observational studies become very important to understand what is the significance of these mutations on long-term survival.

Dale Shepard, MD, PhD: In this particular set of data, how many patients did you look at and what did you find? From a sequencing standpoint?

Moaath Mustafa Ali, MD, MPH: Again, because the acute lymphoblastic leukemia is rare, only we in Cleveland Clinic, we had 30, we almost see almost 30 new cases of acute lymphoblastic leukemia a year, which is generally a much higher number than other centers. Other centers, even sometimes tertiary, they may see one or two or three. So 30 is actually a quite big number.

Among the patients that we have acute lymphoblastic leukemia who had NGS at baseline, because we use strict criteria, we had 72 patients had NGS at baselines, and among them 11 patients had mutation in tp53. In these subsets of patients, we compared their clinical outcomes, their cytogenetic abnormalities, mutations in other genes, as well as their long-term outcomes, including response to chemotherapy, MRD response, and the overall survival and event-free survival.

Dale Shepard, MD, PhD: So in the group that you looked at, and as I recall it was over about six years, you had looked at patients that had been seen here. Was that about the level of mutation that you'd expect?

Moaath Mustafa Ali, MD, MPH: Yeah, tp53 occurs in adults almost around 10 to 15%. Now, what's interesting, we found that tp53 mutation, it's more commonly encountered in B-cell acute lymphoblastic leukemia, not very commonly in T-cell acute lymphoblastic leukemia. We also found that it's mutually exclusive with BCR-ABL1, also known as Philadelphia chromosome. So patients with Philadelphia chromosome, they very rarely have a tp53 mutation, which tells you that both of them work on different molecular pathways.

We also found that there's a strong association with complex cytogenetics. That's something we see frequently in acute myeloid leukemia, but it's less known in acute lymphoblastic leukemia. And we saw that there's a statistically significant association with complex cytogenetics.

Dale Shepard, MD, PhD: What did you find about the presence of the mutation in response to treatments?

Moaath Mustafa Ali, MD, MPH: So people who have a tp53 mutation, tp53 mutated acute lymphoblastic leukemia, the odds of achieving MRD, or measurable residual disease negative response, was almost 0.1 of that of people who have tp53 wild type. So that's almost like 10%, like one-tenth of the patients will achieve an MRD negative.

So what does that translate into, that if patients have an MRD positive response and tp53 mutation and tp53 mutated ALL, that's a poor prognostic sign, and it's very rare actually to achieve MRD negativity.

Now, another thing that we found that people with tp53 mutated ALL, they have almost 60% 12-month overall survival, compared to 90% 12-month overall survival in the TP 53 wild type. So there's almost 30% difference at 12 month mark, which is a quite big drop in survival in just one year.

Dale Shepard, MD, PhD: And so ,can this data, is this data being used to modify first-line therapy? Is it being used to sort of start another therapy sooner? How are we using this data?

Moaath Mustafa Ali, MD, MPH: So the research on tp53 mutated ALL is somewhat limited. There are few ,publications in the literature. Now, our findings, what we suggested based on our findings is that introduction of immune therapy should be considered, hopefully in clinical trials in the future in tp53 mutated ALL, if the use of immunotherapy can be used upfront or introduced at an early stage, as well as considering doing allogeneic transplant at an early stage. The reason why, although these patients sometimes may achieve complete response or complete remission, however, they tend to have frequent relapses and poor overall survival.

Dale Shepard, MD, PhD: So much like these cytogenetics are being used more routinely in AML, then perhaps we can start shifting and doing this in ALL as well.

Moaath Mustafa Ali, MD, MPH: Yeah, definitely in ALL, the use of NGS should become routinely done in all institutions in the United States and worldwide where it's available. The NGS should not be only for DNA, but also should look into RNA fusion. Because fusion proteins are very common in acute lymphoblastic leukemia and they have significant prognostic implications.

Dale Shepard, MD, PhD: What do you think is the next best step, in terms of utilizing the mutational status in a meaningful way? What other research do you think is going to be required to make it more standard?

Moaath Mustafa Ali, MD, MPH: Community of hematologists/oncologists learning about this, having this standardized in the guidelines, which it became standard nowadays in the guidelines to do these diagnostics, but it's just a matter of time until it becomes more adopted. Definitely there are advances in molecular techniques, understanding the mutations, the different kind of chromosomal abnormalities, the different kind of fusions, partial tandem duplications, all of that stuff. The more we understand the molecular biology, the better we will understand the prognosis and the more possible targetable therapy we can hopefully introduce for these blood cancers.

Dale Shepard, MD, PhD: Are there other mutations that you're looking at within ALL that might be promising as well?

Moaath Mustafa Ali, MD, MPH: Yeah, so part of our acute lymphoblastic leukemia database, as I mentioned, because we've done a comprehensive database, we actually have a couple of studies under review. One of the studies we're looking into Philadelphia chromosome, and we found some interesting results. I don't want to share them now, but since it's under review. So maybe in a future podcast.

Dale Shepard, MD, PhD: We'll have you back.

Moaath Mustafa Ali, MD, MPH: Yes, thank you. But we're also looking to the differential mutational profile in B-cell versus T-cell acute lymphoblastic leukemia. We're looking into the mutation profile impact on first-line treatment in acute lymphoblastic leukemia, including pediatric-inspired regimen CLGB 10403, also hyperCVAD. So it's very interesting research, and again, I want to emphasize the role of observational studies. It's very important for prognostication, and as well as to confirm the outcomes of clinical trials. And lastly, for hypothesis, to generate hypothesis.

Dale Shepard, MD, PhD: That's great. Well, appreciate you being with us today and sharing your insights.

Moaath Mustafa Ali, MD, MPH: Thank you. Thank you for having me.

Dale Shepard, MD, PhD: To make a direct online referral to our Cancer Institute, complete our online cancer patient referral form by visiting clevelandclinic.org/cancerpatientreferrals. You will receive confirmation once the appointment is scheduled.

This concludes this episode of Cancer Advances. For more podcast episodes, visit our website, clevelandclinic.org/canceradvancespodcast. Subscribe on Apple Podcasts, Spotify, or wherever you listen to podcasts.

Thank you for listening. Please join us again soon.