Improving Genetic Testing Rates in Advanced Prostate Cancer - Pedro Barata
March 14, 2023
Pedro Barata, MD, MSc, Leader of the Clinical GU Medical Oncology Research Program, University Hospitals Seidman Cancer Center, Associate Professor of Medicine, Case Western University, Cleveland, OH
Alicia Morgans, MD, MPH, Genitourinary Medical Oncologist, Medical Director of Survivorship Program at Dana-Farber Cancer Institute, Boston, Massachusetts
Alicia Morgans: Hi. I'm so excited to be here with Dr. Pedro Barata, who's the director of the Clinical GU Medical Oncology Research Program at the University Hospitals Seidman Cancer Center, as well as being an associate professor of medicine at Case Western Reserve University School of Medicine. Thank you so much for being here with me today, Dr. Barata.
Pedro Barata: Thank you so much, Dr. Morgans. It's always a pleasure to chat with you and highlight some of our work at ASCO GU.
Alicia Morgans: Well, it's always a pleasure to talk to you too, and I think one of the things I'd love to hear from you about are three presentations that you gave at GU ASCO 2023. You really dug into the whole idea of genetic testing, germline testing in patients with advanced prostate cancer, and you tried to think about how these testing practices may actually impact our outcomes for patients with advanced disease. Let's start with the metastatic hormone-sensitive presentation that you gave, and this was a lovely poster at GU ASCO. Tell me a little bit about it, please.
Pedro Barata: Sure. Thank you for the opportunity once again. Basically, as you know, there's a subgroup of patients with advanced prostate cancer who harbor these homologous recombination repair, HRR, mutations, and they can be either germline or somatic. Actually, we do have ongoing trials asking the question about the role of PARP inhibition in this setting, the hormone-sensitive setting, usually combined with novel hormonal therapies. I'm thinking of T-3, I'm thinking AMPLITUDE, for example. However, I don't think we know a lot about the prevalence of HRR mutation testing rates in the real-world setting beyond trials for this patient population. So that was really the thought process we had to conduct this study.
We selected eligible physicians that were treating prostate cancer out there and they were able to abstract medical records from their own patients. Each physician could get up to 8 patients, if you will, and basically, patients, as long as they had hormone-sensitive disease and not being part of a clinical trial, they were eligible. We consider HRR status or HRR genomic alterations in one of the genes of interest, common usual suspects, ATM, HER, BRCAs, TDK, CHEK2, FANCA, RAD51, PALB2. Those are some of the 12 HRR genes of interest. And then we look at what happened as far as genomic testing patterns. So we basically included a little bit over 200 patients. What was interesting to us is actually more than two-thirds of that patient population has not been tested for HRR by germline or somatic. So, actually, no test for them, compared to about 28% of patients, a little bit over a quarter of those patients, who end up having testing.
Though, we also did a multivariate analysis, an MVA model, to identify, perhaps, characteristics associated with a chance to get tested for HRR genes. And among all the factors we took into consideration, age, family history, treating the treatment by urology or medical oncology community versus academia, race, visceral disease, Gleason score, metastatic disease, I think we identified two factors that are being the strongest associated with lack of genetic testing. There were the lack of visceral metastasis and no family history. So in other words, in general, the rate of testing has been in a little bit over a quarter of the patients, and of the patients, if you didn't have visceral disease nor family history, that put the patient at a low or very low chance of getting testing for HRR gene alterations.
That was basically the take home point from this study one, if you will, in the hormone-sensitive space, and so I think there's a little bit of learning lessons that we can take out of this. Of course, this needs to be validated in other cohorts, but I actually think that, based on us talking to the community colleagues out there, I was actually surprised to see 28% of patients are being tested in this setting, because in some of the situations, I would predict that number to be even lower. And as we know, Dr. Morgans, it does help us when we get genetic testing upfront, both somatic and germline, not just because we can initiate cascade testing for the family members in case you have a germline-positive alteration, but also you can start planning ahead and start thinking about optimal sequencing, whether you're going to use immunotherapy or a PARP inhibition, et cetera, for those patients. So I think there's an evolving story, but I think it's interesting to see or get the findings of this particular study.
Alicia Morgans: Absolutely, and I agree with you. I actually am happy that we at least have that proportion of patients getting tested, but we do have a long way to go and the guidelines suggest that all of these patients should be tested. And so, our aim should be that high and we'll have to see how we continue to do. I'm sure you'll update this work over time and let us know, because without looking in the mirror and really thinking about what we're doing, measuring it, we don't really have the benchmark or the opportunity to say, "We need to do better. What are strategies we can put in place to try to improve this particular aspect of our care?"
Now, I know you also looked at the metastatic castration-resistant prostate cancer population and would love to hear, are things similar? Are they different? What does it look like in mCRPC?
Pedro Barata: Yeah, that's a great question. We really wanted to look not only to the hormone-sensitive space, but also to the CRPC space, and for feasibility reasons, we were able to dig into the Flatiron Health database, which is basically a longitudinal database capturing data from patients that were cared for in the community as well as in the academic setting. So we got a much larger cohort that started early 2014, all the way to the end of '21, so it's a relatively contemporary dataset, and that included the time around COVID, which I think is important and I'll mention in a second. But, bottom line, we had a much larger cohort here. We were talking about over 8,100 patients included who were diagnosed with mCRPC after 2013.
What is interesting among the results is we actually were able to show the breakdown of the HRR mutation testing rates by year among men with mCRPC who were actually eligible for mutation tests. It's interesting to see the bars. In 2014, 1%, 2015, 2% and 3% and 5% and 7% 2019, we got the top 10% in 2020, and then it dropped a little bit to 7% in '21. So, very small numbers in this particular setting. Again, it's a much larger cohort, again, over 8,000 men here, 8,166, to be precise. We saw a little dip in the 2021, which might be COVID-related, this is completely hypothesis generated. We have not confirmed that, but that was the findings.
Now, we look specifically about the Black population versus White population, and in our numbers we showed a breakdown per year. It is true that, at least in this cohort, the lower proportion of Black patients received HRR mutation testing compared with White patients. There were a few exceptions in 2015 and 2021, the numbers were very, very similar for both groups. But in general, Black men were destined to have less genetic testing.
But what we did in addition to that is actually we ran a logistic model to see what factors were, indeed, associated with HRR mutation testing. It's interesting to see that we did find a number of factors, and some of them are kind of predictable, others not so much. So, patients without prior surgery or radiation with a higher performance status, newly diagnosed metastatic disease, a lower socioeconomic status, treatment in the community, African Americans, as I mentioned previously, and elderly patients, this is defined as over 65, were actually, all of this, associated with lower odds of HRR mutation testing in these men with mCRPC.
Of course, this is a retrospective analysis, there's obviously inherent limitations to these kinds of analyses, but I think it's a significant cohort for us to look at those factors and think, what can we do to improve the testing rates in men with mCRPC that, as you know, Dr. Morgans, we can actually do something about it. It does have treatment implications. And by the way, this testing included both germline and/or somatic testing. So, to me, it's telling that the rates are around a little bit below the 10%, if you will, and I think that's a lot. There's some learning lessons we can take from these analyses.
Alicia Morgans: Absolutely. And also to recognize which populations are being undertested, and these are some of the populations that certainly deserve the opportunity to have access to these targeted therapies that may be identified as potential treatment strategies if we do the testing and understand the results. And in some cases, like the older adult population, these are patients who really need, potentially, a treatment option that may have a lower burden of side effects. Sometimes our therapeutic options when we do this testing and can use things like pembrolizumab, for example, will be for most patients, relatively low burden, if not minimal burden. And that's really a value to everybody, but particularly to older, frailer patients who may not even realize that they have that opportunity if you're not doing the testing.
I know you took this all a step further and you didn't just shine a light on our practices, but also try to understand patient outcomes in this context. What can you tell us about patient outcomes? And I believe you looked at the mCRPC setting.
Pedro Barata: That's absolutely correct. We looked at this data and we said, "Boy, it would be interesting to actually understand the impact of the HRR mutation status in clinical outcomes." So what we did is we conducted an observational retrospective study, and so we got, again, data abstracted from medical charts from patients treated by oncologists and/or urologists in the US who had mCRPC, who had known HRR, and started any first-line systemic therapy, whether it was chemo or hormonal therapy or any other therapy, really, radium, immunotherapy, or even PARPs. This cohort went through September 2016, all the way to September 2019. So solid 3 years here, if you will. Again, this is completely off clinical trials, so it's completely real-world data. We were able to include 130 patients in our full cohort, and about 51 patients had mutated HRR compared to 79 patients with wild-type HRR.
What's interesting about the testing timing, about a third or so of the patients were actually tested prior to first line CRPC, whereas two-thirds of those patients were tested after starting first-line CRPC within the first line or beyond, second, third line, et cetera. So, we're really interested in reporting treatment patterns and also outcomes with those treatments in front-line CRPC. So, interesting for us to analyze. And we got a breakdown by systemic therapy, if you will.
It's interesting to see that actually novel hormonal therapies, both abiraterone and enzalutamide here, or even a combo in a few patients, we got a subset of other novel hormonal therapies used in this setting. Actually, NHTs were less used among patients with homologous recombination repair-mutated patients versus those with wild-type. And that was basically double 29% versus 58%, if you will. It was also interesting because we thought, "Okay, so those patients probably got more chemo in that setting." Actually, the chemotherapy rate was very similar in both these group populations. So this was one part of our findings.
The other pieces we really wanted to know how long were these patients on treatment? So in the real world, this real-world progression-free survival seems to correlate quite nicely with the PFS the way we know it. And so looking at the real-world progression-free survival, we analyzed the real-world PFS by HRR status for the frontline CRPC. What's interesting is, when we compare the group of patients with wild-type versus patients with mutated HRR, we found that the median real-world PFS was 30.9 months versus 19 months favoring the wild-type. And the same was true when we looked at the group of patients with HRR wild-type with BRCA1 and 2 mutations. And very similar numbers as well. We're talking 19, again, versus 30 months or so.
Now we're living in a time that we're talking about the outcomes and the poor prognostic factor that HRR represents to this man. And at least in this real-world dataset, that seems to be the case. In other words, patients with mutated genes have poor outcomes, poorly on front-line systemic therapy, compared with patients which HRR wild-type, which seems to be concordant with what we've been seeing from different trials. That's also important, and I think obviously there's more work to be done, but this was the piece of the story that brought the treatment implications and what happened to these patients as far as outcomes based on these HRR pattern testing.
Alicia Morgans: I think that's so helpful, so important, and certainly, as you said, it is concordant with what we've seen. Having these HRR alterations, particularly BRCA2, BRCA1, even in the localized disease setting, has been associated with shorter time to progression, poorer prognosis, and your real-world data from the US population here really bears that out.
It's also interesting to see differences in utilization and perhaps we're recognizing with these poorer prognosis patients that we shouldn't be sequencing AR-targeted agents. But then the question is, what should we do? An unmet need here where we certainly have the opportunity to use PARP inhibitors and potentially if we're talking first-line mCRPC, we may ultimately even have the combination as an option with a PARP inhibitor and AR signaling inhibitor. But we have to see how that all plays out.
Thank you so much. I think that these three analyses really provide so much context for us. I want to give you the opportunity if you have anything else you want to add as just a final word on this topic.
Pedro Barata: Sure. No, I couldn't agree with you more. Really well summarized, Dr. Morgans. This is fantastic. I just want to obviously highlight the great job that the Pfizer team did on this. Actually, this work was funded by Pfizer, I should acknowledge that. They're really looking at the real-world outcomes. So, completely separate from, obviously, the motivations around the PARP inhibition. Pfizer is one of the companies that has skin in the game in trying to understand who are the patients who benefit the most from PARP inhibition. But these real-world efforts are really relevant, in my opinion. Kudos to Alex and the other group and the group from Pfizer who helped us putting this, I think, to me, important data, real-world data to be presented. And hopefully we'll be able to publish this soon. So, thank you to them for allowing us to be able to conduct this study and there has been a great effort they put into it. So I just wanted to acknowledge that.
Alicia Morgans: Well, thank you for acknowledging that and I do think it's important always to consider sponsors and to recognize how much of a team sport this has become, not just in our clinical practices where we really use a multidisciplinary team to support our patients holistically and throughout their disease course, but also, of course, we have collaborations with pharma companies, with other academic centers, with community partners, and with real-world database vendors who help us get this work done. It is really taking a village, and we have a fantastic one. Thank you so much for your time and for your expertise today.
Pedro Barata: Thank you so much.