HRR Mutation Testing in mPC
Transcript: Navigating the genetic landscape of mCRPC
Gerhardt Attard, MD, PhD, FRCP, and Alicia Morgans, MD, MPH
Interview recorded August 2025. All transcripts are created from interview footage and directly reflect the content of the interview at the time. The content is that of the speaker and is not adjusted by Medthority.
- [Alicia] Hi, and welcome. My name is Alicia Morgans. I'm a GU medical oncologist at Dana-Farber in Boston, and I'm really, really pleased to be here today with Professor Gerhardt Attard, who is joining me from London in the UK to talk about how we think about BRCA and beyond really navigating the genetic landscape of metastatic castration resistant prostate cancer. Thank you so much for being here with me today, Professor Attard.
- [Gerhardt] Thanks for inviting me. - [Alicia] Wonderful, well, Gerhardt, let's dive right in. We have some really important things to consider and to talk about to really help share your experience and really the way that we're thinking about these alterations and a path forward in prostate cancer. So let's start at the beginning, you know, how do you really think about and approach discussions about these HR mutations with your patients? And of course, we're really thinking about both BRCA mutations and non-BRCA mutations when we're seeing patients in clinic.
- [Gerhardt] Yeah, so I guess this complexity here, not just in discussions with patients, but also as physicians, how we think through this group of gene alterations. And I think with BRCA it's very clear, cancers with an alteration in BRCA2 are sensitized to PARP inhibition, with BRCA1 probably similarly so, although whether it is equally so remains I guess on the study, and there's probably a number of other genes that sensitize cancers to PARP inhibition. But that list includes small subgroups and has a heterogeneous effect and there's double hit of small numbers. And heterogeneous effect really creates a challenge in answering the question we need in clinic, which is yes or no, treat or not treat, and we don't want to handle them as one large population, the non-HRR mutant group, we want to make a decision based on that gene alteration. But for many of these genes in our registration trials, we've had a small handful of patients, 10 to 30, you know, it's events in a smaller number of those.
One of the useful outputs that has come from at least the line one MCRPC phase three trials is an independent patient meta-analysis published by the FDA, which allowed pooling of data across the three trials for individual gene subtypes. And I found that enormously informative. And what that study showed, suggested or showed was sensitization in the CDK 12 mutant group and then the PALB2 mutant group. Now PALB2 is promoting and localizing BRCA. So the strong biology to support that PALB2 would have the same sensitization effect as BRCA, but the numbers are really small. So the uncertainty remains there. With CDK12, that's a bit more common. There is emerging biological evidence to support that sensitization. And increasingly I would like to have the option of using a PARP inhibitor for a MCRPC patient with a CDK12 alteration. And of course, we know from a number of studies now that that group of patients have a worse outcome and do not appear to have greater benefit from immune checkpoint inhibition. So we need a targeted therapy for that group.
The third large group, I should mention in this non HRR population is ATM. And in this there have been signals for effect in ATM, there's interesting biology to support ATM as involved in homologous recombination repair and sensitization to PARP inhibition. But I think alterations in ATM are heterogeneous within patients. And we published a small study a couple of years ago showing monoallelic loss of ATM. There was loss of one allele and because there was duplication of the remaining wild type allele, that was called as LOH and the patient was, I wouldn't say mistakenly, but he was labeled as ATM loss, there clearly wasn't loss, neither genomic level nor at the protein level, and he did not respond to PARP inhibition. And how extensive that challenge is in terms of ATM loss is hard to say because we're dealing with challenging tissue, FFP, which doesn't allow us to dissect that granularity.
But I suspect the lack of efficacy that has been reported in the ATM subgroup is at least in part explained by ATM monoallelic loss being much more common than biallelic loss. and the biallelic loss being required for sensitization. - [Alicia] I think that's so interesting and so important because there have been multiple clinical trials and certainly there are sort of real world studies, studies taken from claims data that are very conflicting in terms of certain populations having responses or lack of response. And a lot of that might be related to the method of testing in one allele versus both alleles being out and with ATM of course, there's this concern around CHIP really confounding a ctDNA interpretation of these alterations in addition to just losing one versus losing two. So how do you think through all of the clinical data, you mentioned the FDA meta-analysis, but so many of these exploratory analyses end up giving us insights, but then those are potentially refuted in the next exploratory analysis.
How do you put it all together? - [Gerhardt] And you raise the issue of testing Alicia, I mean the devil's always in the detail. So we have a range of different testing methodologies that have been used, broadly germline testing, tissue testing and ctDNA and the last two are subject to significant variability both across assays and within the false negative rate. The tissue we are dealing with is several years old and of very small quantity. And in most studies it has been unable to differentiate between biallelic and monoallelic. And in fact, copy number calling has been challenging. So most of the patients we recruit and identify have a mutation in one of the genes. And we don't know whether the other allele is lost. And I think we are missing many patients who have solely loss, loss of both alleles, and that's in tissue and ctDNA becomes even more complicated. 'Cause although the DNA is very good quality, the tumor fraction is low. And this is becoming an increasing issue as we use PARP inhibitors early in the disease landscape where tumor burden is lower. And I have a strong belief that that's where we should be using PARP inhibitors when tumor burden is lower. But of course then circulating tumor DNA fractions are lower and at the very lower limits of detection, calling somatic mutations at 1% to 2%, as you said, that can be confounded by mutations in white blood cells as a result of clonal hematopoiesis and the patient being falsely labeled as having a somatic mutation in ATM or check one or a range of other genes.
And to avoid that it requires germline testing, which many of the commercial ctDNA assays do not do. And if one sets higher thresholds for detection, they were really excluding a lot of our patients. The, the other consideration, so there's two parts to this. One is relevant to clinical practice and I think it's really important that we do not label patients as HRR negative based on a ctDNA test if we do not know the tumor fraction. So if the tumor fraction's high in ctDNA, we're probably comfortable that there is no somatic alteration in the gene. But if one does not have sight of the tumor fraction and most ctDNA reports do not provide information on circulating tumor fraction, then do not label a patient as negative because that just could be a false mess. And of course, we're missing the chance, you know, if that goes in the letterhead for that patient's in the diagnostic box, HRR negative, then they may not be tested in future and miss out on that chance.
I think the other important consideration is some of our trials have selected patients based on ctDNA, and that in itself selects a group of patients with a worse prognosis 'cause the only ones who are identified as positive are cases where there is a sufficiently high ctDNA fraction, which we know is poorly prognostic. And that shift in the case mix can affect results in a range of ways. Some expected, others potentially not expected, such as, you know, of course time to progression is shorter and there could be a number of influences there. So as you said, we have some discordance in the literature, which could relate to different PARP inhibitor mechanisms of action. It could relate to different populations, but I think a lot of it is driven by different tests. - [Alicia] Yeah, thank you for that. And I think that that's especially important, at least in my practice for these non BRCA HRR mutations, which I end up picking up on some of these somatic tests more commonly than my germline tests, though though of course we can see these in germline testing as well. I so appreciate the points that you raised around the complications or challenges of testing. I wonder if you could give just maybe two to three very practical points, just even summaries of what you just said, but just very pointed and practical for folks who are trying to find both BRCA and these non-BRCA HRR mutations. Just what, what should we look out for? One, two, three.
- [Gerhardt] So if you can only do one test, test tissue, that will capture both somatic and germline. For BRCA, for example, half of alterations are in germline, but for the other HRR genes, it's a lower proportion. So germline testing alone will miss a high proportion of somatic alterations. However, tumor testing is challenging both in terms of finding the tissue and then obtaining a successful result. So when that fails, germline test, so that's either saliva or blood could identify a smaller population of patients with an alteration. And from the evidence we have, when there's a germline alteration, at least for BRCA, but looks similar for other genes than the other allele is lost in tumor tissue. So that is a clinically relevant result. The third option is ctDNA and that of course is accessible. Do a ctDNA test when a patient is progressing, not when they are responding to treatment, because then ctDNA fractions would be too low and be cautious about ctDNA fractions. So don't label a patient as negative without being certain that the circulating tumor DNA fraction is high enough to be certain of the somatic HRR gene status. - [Alicia] Wonderful, thank you for that.
You know, I think these non BRCA HRR mutations are also really, as we talked about, kind of different from each other. We mentioned a few of these that may have better responses or be more interesting in terms of their potential response. I wonder if you could give an example, have you seen any patients with a PALB2 alteration or with a CDK 12 alteration who on review, was found to have one of these alterations and actually had a nice response or a response either to a single agent PARP inhibitor or to a PARP and AR pathway inhibitor combination treatment? - [Gerhardt] Yeah, so for CDK12, yes, and in fact a patient who was on abiraterone, and then we were able to access, you know, abiraterone plus par and the patient was progressing and then responded, had a CDK12 mutant, one always has to treat anecdotal cases with caution. I must say, you know, I haven't seen long responses, long durations of responses with PARP inhibition in metastatic patients with a CDK 12 mutation. So I don't think I've had a PALB2 mutation, no one that comes to mind. As I've said, they're rare. The ATMs, the other subgroup I've treated a large number of patients in, it's challenging because ATM mutant cancers tend to have a more indolent history. So we have several patients we tracked over time. We collected ctDNA, we've performed a lot of research on them.
They tended to have quite unusual distribution of disease, but they also had a relatively indolent course. And so it's hard to say what the contribution of treatment to that subgroup is. - [Alicia] Great, well thank you for sharing that though. I think it's interesting, if you as a person who's so dedicated and interested in this field and certainly doing a lot of germline and somatic testing, finds that there's still variability, but still opportunities to see some nice responses in some patients, then the rest of us should not feel so on the spot to have to know everything about everything, we're all learning together and I appreciate those insights. You know, one of the things that's really important and interesting too is that some of our testing, particularly somatic testing, when we might be testing at different points of disease with, you know, new metastatic sites or ctDNA, we may see multiple mutations.
And I wonder how you interpret that. If you see multiple mutations, perhaps multiple HRR mutations that are non-BRCA or BRCA and additional, how do you interpret those? - [Gerhardt] That's, in my experience, not that common. Although it does happen, as you say, and my hypothesis is having alterations in two HRR genes increases the sensitization, whether it's additive or synergistic, I don't know. But I think the effect is greater if more than one gene is altered or lost. So I think, you know, BRCA is mutated, that opens access to PARP inhibitors in most of our jurisdictions in terms of funding. And that's the one we call when we have two alterations in non-BRCA HRR genes. I think that increases enthusiasm to use a PARP inhibitor. But because this is so uncommon, we don't have sufficient evidence in our trials in randomized patients.
- Yeah, I agree. I think the times that I see it most frequently, there's a very low tumor fraction in a circulating tumor DNA assessment. And so it almost looks like noise in these 0.1% of this or that. And so I think that's an opportunity too, to think about your sample, think about retesting, think about ensuring that you know the truth because there can be a lot of noise in some of this, you know, Gerhardt, as we finish up, would you mind sharing your thoughts on where we're going from here and how you're going to be integrating these mutations in BRCA, non BRCA HR mutations as you move forward managing patients with metastatic CRPC? - [Gerhardt] So I think the evidence is very clear that every patient, it's a tumor with a BRCA mutation should receive a PARP inhibitor at some point. And my bias is the earlier, the more effective, and we see this with the line one MCRPC trials that have recently reported on survival and long-term RPFS, the benefit in the BRCA mutants is so large that using a PARP inhibitor later is unlikely to, I guess, catch up, make up for that difference.
So that's the first thing, for the non-BRCA mutants, there are patients who will respond. I think it's going to continue to be a case by case basis making an assessment of other treatment options for that patient. And the evidence at the time and access, I think for patients without an HRR mutation detected by any of the tests, we don't have enough evidence to support using PARP inhibition when there are other effective treatments. As you know, there's now emerging data on MHSPC, so I think a proportion of patients will receive treatment a PARP inhibitor for MHSPC rather than MCRPC. So an MCRPC will start to treat patients who have RP resistant PARP inhibitor resistant disease. And going forward, we're now going turn into the sort of new paradigm of how do we reverse resistance to PARP inhibition in a tumor with an HRR mutation.
- [Alicia] Well, I so appreciate your thoughts both in terms of what you're doing now and, you know, overcoming some of the real world challenges that we see in testing and also as you think about the future and where patients may have access to these types of targeted therapies and benefit from them as we move forward in treating advanced prostate cancer. Thank you so much for your time and for sharing your expertise today. - [Gerhardt] Thank you.
Developed by EPG Health. This content has been developed independently of the sponsor, Pfizer, which has had no editorial input into the content. EPG Health received funding from the sponsor to help provide healthcare professional members with access to the highest quality medical and scientific information, education and associated relevant content. This content is intended for healthcare professionals only.