KC: Hello and welcome to this ecancer expert to expert discussion on PARP inhibitors and testing for homologous recombination repair defects in patients with advanced prostate cancer. I am Dr Kim Chi, I’m a medical oncologist and professor of medicine at University of British Columbia and I work out of the BC Cancer. Joining me today is Alex Wyatt who is from the Vancouver Prostate Center and the University of British Columbia. So, first off, we’re going to discuss a little bit about the background on the DNA repair landscape in prostate cancer. Alex, do you want to start?

AW: Yes, thanks Dr Chi. We now know that particularly in advanced prostate cancer it’s very common to find DNA repair defects in both the germline of the individual and also in the tumour itself. In fact, up to about a quarter of all metastatic prostate cancer cases harbour DNA repair defects. This is significantly higher than in the local or locally advanced setting. The most common gene that’s altered is BRCA2 but other genes linked to the homologous recombination repair pathway are also recurrently altered. One other pathway that can be altered in metastatic prostate cancer is the mismatch repair pathway, so genes such as MSH2 and MSH6 but these are not as common as BRCA2, ATM, CDK12 and other HRR genes.

KC: This has very important clinical relevance around the recent FDA approvals of the PARP inhibitors olaparib and rucaparib for patients with metastatic castration resistant prostate cancer and a pathogenic alteration in a homologous recombination repair gene. The olaparib FDA approval was based on the PROfound study which showed an improved radiographic progression free survival, particularly in the patients with BRCA1/2 and ATM group. Then also the rucaparib study, TRITON-2, received accelerated approval based on patients with a somatic or germline BRCA alteration and, as you said, most of these patients had BRCA2. These patients all had previous AR targeted therapy and taxane based therapy and showed a very high objective response rate in the 40-50% range, many patients having responses that were lasting more than six months. The challenge, though, will be identifying these patients and there are many hurdles to overcome. For example, in the PROfound study I think approximately 4,400 patients were enrolled for screening, of those approximately 400 did not have tissue available for testing and of those there was a proportion that couldn’t be sequenced successfully. Then it got down and down until the point that about 28% of patients, or ultimately less than 10% of patients, got on the study from the original 4,400 patients screened. So there are a lot of challenges to some of the testing that needs to be done. So let’s start with that, how do we test for these patients? So who and when should we test? What do you think, Alex?

AW: Typically at the moment people are considering testing when a patient is metastatic castration resistant. The trouble in this setting is that tissue is not always easy to access. So ideally one would have access to a fresh metastatic tissue biopsy but I’d maybe comment on how straightforward or not that is in the advanced setting.

KC: Yes, in terms of who and when to test, mostly because we’re looking at the PARP inhibitors in that late stage mCRPC setting, that’s where we’re thinking about testing, but I really think we should be starting to think about testing patients even before that. Patients when they’re developing metastatic castration sensitive disease. In fact, the NCCN guidelines from the US suggest testing even in patients with high risk and very high risk localised prostate cancer. Of course the tissues available for testing can be a challenge, as you mentioned and, indeed, in many patients it can be a challenge to get those archival tissues. Some patients with metastatic disease may not even have an archival tissue specimen, that they were clinically diagnosed and started on treatment without a tissue diagnosis. But, for the most part, 70% or probably closer to 80-90% of patients will have some archival tissue for testing but even if they have tissue there may be a challenge with that, is that right Alex?

AW: Yes, that’s correct. It can be challenging to get sufficient DNA of sufficient quality from archival tissue. Fortunately in recent years there has been big improvements in the technology to do this and, as you mentioned, the success rates are getting better and better. So once those technical issues are overcome there can actually be some biological issues because you may be looking at the primary tumour from a patient whose primary was treated ten or more years ago. So there are sometimes questions about the representativeness of that archival tissue so that’s something that you have to bear in mind when you’re considering a result.

KC: What about fresh biopsies? What is the success rate with fresh biopsies? There can be both technical challenges with that but it’s also quite invasive for our patients.

AW: That’s right. I think for fresh biopsies the success rate is quite dependent on the site. So a soft tissue biopsy if a patient has a liver metastasis or an accessible lymph node, those often have quite a lot of tumour in them and the yield of DNA and the quality of that DNA is often reasonably high. Most patients with advanced prostate cancer, as you know, have bone metastases and they can be harder to biopsy. Consequently, the amount of tumour material that you obtain is low and sometimes of lower quality.

KC: Now, we’re also hearing about circulating tumour DNA or cell-free DNA and that’s across many, many cancers, including in prostate cancer. This is a promising area as well to test for patients. What do you think about the ctDNA test? You’re quite a leader in this area.

AW: Yes, we’re quite fortunate in a sense in the advanced prostate cancer setting, in the metastatic setting, that many patients have high amounts of circulating tumour DNA in their bloodstream. So this means that a plasma cell free DNA test is often reasonably informative in the sense that you have enough material to actually say positive versus negative. There are challenges, just like tissue there are patients where you can’t get enough material and so you can’t actually judge whether or not a somatic alteration is present or not. I would say that for now tissue remains the gold standard for testing but plasma ctDNA is definitely emerging on the scene as a valuable component to HRR testing.

KC: It may be a good alternative to a fresh biopsy which is much more invasive than taking a blood test for circulating tumour DNA.

AW: That’s right and, of course… sorry to interrupt but of course considerably cheaper, both from an actual cost point of view but also from a patient morbidity point of view. I think the other thing to remember is that because close to half of all the alterations we’re detecting in the advanced setting are at the germline level, a simple blood test is often enough to identify a germline alteration, so even a cheek swab could do this. So even if tissue isn’t available it’s still possible to identify potentially relevant defects just by screening a germline surrogate.

KC: And now there are a number of commercial as well as institutional assays that are validated to identify these both somatic and germline alterations. So, what does a negative test mean? If send off a patient and it comes back negative, that could have many implications – is it truly negative?

AW: Yes, the interpretation of a negative test has been a challenge for years. It’s going to depend upon the source of tissue, so a fresh metastatic tissue biopsy that’s probably relatively representative of the patient’s disease as a whole perhaps can be interpreted with a little more confidence than, let’s say, a very old piece of tissue from the primary site. So we have to consider even in the setting of a negative test about testing again potentially in the future, particularly if you’re looking at that patient in the castrate sensitive setting.

KC: Then also a positive test, what does a positive test mean? It’s not just a simple yes/no answer, there are many different kinds of genes and there can be monoallelic versus biallelic losses. So how do we interpret that? How would the average investigator and average clinician, what should they be cautious about when they’re looking at a test result that comes back to them?

AW: That’s correct. We have to realise the limit of our knowledge at the moment and we currently know enough to realise that there are a lot of variables here, that not all DNA repair defects are the same. So we like to group things together as a field but the reality is that there are many different genes involved, they don’t all have the same impact on the HRR pathway. But even within a gene the location of a mutation, as you mentioned whether it’s mono- or biallelic, these things matter. We just don’t know quite the extent they matter for PARP inhibitor response yet. So that’s likely to be elucidated over the coming years.

KC: Yes, I think we see most of the benefit coming from patients with a BRCA2 alteration but we’re also seeing activity in other HRR associated genes, including PALB2, for example. So one issue, as you mentioned at the beginning, is that many of these patients have germline alterations in one of these genes, meaning that they inherited it from their mother or father and they could pass it down to their children. So an important issue is understanding if a patient does have a germline alteration because that has broad implications, not for that patient in terms of their own treatment but also implications for their brothers, their sisters and their children as they are at risk for developing prostate cancer as well as breast cancer, ovarian cancer, pancreatic cancer and other cancers. So certainly if there is a germline alteration found then patients and their families should be referred for hereditary cancer counselling.

So, we’re going to end it there with this discussion. Thank you very much, Alex, that was very interesting. And thank you for listening in to this expert to expert ecancer discussion.