The Uromigos Episode 126: APCCC—Germ Line Testing and Molecular Testing in Prostate Cancer

Ros Eeles gives a clinically focused overview of this topic.

Episode Transcript

Tom:
Welcome everybody to our third in the most recent series of the APCCC podcasts. Silke Gillessen and Chris Sweeney are going to do the steering. We’re really, really pleased to be joined by Ros Eeles. Ros, do you want to introduce yourself and then Silke, do you want to ask the questions?

Silke Gillessen:
Yes.

Ros Eeles:
Yes. Thank you so much for asking me to be involved in one of these exciting podcasts. So, I’m a professor of oncogenetics at the Institute of Cancer Research and an honorary consultant at the Royal Marsden Hospital in London. I’m a clinical oncologist by training, with a special interest in cancer genetics in adults.

And I’ve got a prostate cancer treatment practice and then specialize in cancer genetic testing, particularly in the germline in adults and obviously, for this podcast, with its relevance to prostate cancer patients and their families.

Silke Gillessen:
Great. So, Ros, I think I just thought of with the first question that I’m also getting asked a lot. So, what do you think, and you said for prostate cancer, which patients need germline testing and counseling, of course?

Ros Eeles:
So, unfortunately, I think it depends which country you work in. So, if we take the ideal scenario. For example, if we take the recommendations in the United States, then we would want to test in men who have prostate cancer at young age of onset, by which we would mean really individuals at least under the age of 65.

In fact, in our research studies, we see under the age of 70 who, if they do not have any spread of their prostate cancer, might have between maybe a 4 and 7% chance of having a germline mutation in a DNA repair gene. And then those with metastatic disease, maybe at least one in six of those, 13%, might have a germline mutation in a DNA repair gene.

And unfortunately, in the United Kingdom at the moment, those individuals can only have testing if they have a very strong family history of breast or ovarian cancer on the testing directory, or if we found a mutation in their tumors [inaudible 00:02:43] will actually change in the near future. So that then we’re all testing young onset cases or cases with metastatic disease.

Chris Sweeney:
That’s brilliant. And maybe we can just expand that a little bit to us. The question is why should we test these patients? What is the clinical implication of finding and what examples of genes are you specifically referring to with DNA damage repair?

Ros Eeles:
Absolutely, Chris. So, the sorts of genes that were that we would recommend testing would be germline mutations in the BRCA1, BRCA2 genes, ATM, and then mismatch repair. So, the genes that predispose to Lynch Syndrome and the reason for testing these is that first of all, for the individual themselves, it could have implications for their management.

So, if you put a germline mutation in the BRCA1 or BRCA2 gene or ATM, there’s some early data, particularly from Johns Hopkins that suggest that they might have a threefold higher risk of progression if you manage them with active surveillance. So, it’s not yet known whether you definitely therefore should manage such an individual with surgery if they have local disease rather than active surveillance.

I don’t think that’s really proven definitely at the moment, but many units are now veering towards doing that. And we have a very large research study going on to see if we can validate those data, but many surgeons now, where if you have a germline mutation in particularly BRCA1, BRCA2, or ATM would prefer to offer surgery rather than active surveillance.

And then for metastatic disease, that you can now offer targeted treatments using PARP inhibitors, if somebody has relapsed, particularly if they’ve relapsed on first-line hormone therapy, following the profound study results.

Silke Gillessen:
Okay. Well, one thing, because that’s really something that interests me, what you think about. So, there was also in the NCCN guidelines, also at least the recommendation to also test patients who have intraductal, ductal, or cribriform components in their primary tumor. So, what do you think about that?

Ros Eeles:
Yes. And we know we now know that those features are associated with the worst prognosis. And I have to say my personal opinion is that this is very forward-thinking, and we would like to be able to do that. Unfortunately, in the UK, we can’t on the current guidelines, but I think following the NCCN guidelines in the US would definitely be the right way forward in my opinion, but we definitely need more data to prove that’s the correct thing.

Chris Sweeney:
Can I pick up on two aspects of that it’s really high-risk disease patients, patients with high-risk disease who actually are recommended. And I think it’s very hard to get intraductal without actually being high risk.

And that, that NCCN recommendation passing in those out with potential thought that patients with a BRCA [inaudible 00:05:58]

Silke Gillessen:
Chris, I guess we lost you?

Chris Sweeney:
I got back, I just someone just tried to call me. And so sorry about that. So, Ros, I think half that NCCN guideline was there was some early notion that maybe patients with BRCA2 mutation were more likely to have this intraductal feature and Elena Castro’s work is maybe counter that I know Elena and you have worked closely together. What’s your comment on that?

Ros Eeles:
I, yes, I think that’s a very good point. Is that introductory disease was thought to be associated with a higher risk of being aggressive. And certainly, BRCA2 was also thought to be associated with a higher risk of aggressive disease, but they weren’t necessarily totally correlated.

And certainly, the BRCA1 mutation carriers, we’ve now got early data suggesting, and so does Elena Castro, that they maybe don’t behave in such an aggressive way as germline mutations. It’s those with germline mutations in BRCA2, and it’s very interesting when you think about the other aspect of how you might manage individuals, for example, their relatives.

In the impact study, we did targeted screening of individuals with BRCA1 and BRCA2 germline mutations, and that was done with annual PSA starting at 40. And in the impact study results published in European Urology, we found that there was a significant benefit in doing screening and BRCA2 mutation, but not in those with BRCA1, that didn’t yet reach statistical significance and that fit with this prominent and now the new guideline that we should do targeted screening and BRCA2 mutation carriers, but not yet the BRCA1.

Chris Sweeney:
Can I put a patient scenario in front of you, Ros, and see how Silke and you would react on this? And so, a 50-year-old man, let’s call it 52. My age. Presents with a PSA of 3.2 and had a family history actually of prostate cancer and he’s young and he gets a biopsy and he’s found to have Gleason three plus three in 5% of one core.

And he got sent off for germline testing for whatever reason. And it did come back with a presence of a BRCA2 variant. Pathogenic, germline variant. How strong do you think that person would need a prostatectomy? And is [inaudible 00:08:38] and would you say you definitely need a prostatectomy because you’ve got Gleason six and peri-BRCA2 mutation.

Ros Eeles:
This is, this is such a classic dilemma that we see in the clinic. So, very small amounts of Gleason three plus three. You know, if you didn’t have a BRCA2 mutant, then there’s no question.

Most people would offer him active surveillance unless he was incredibly anxious about surgery. And it’s based on the progression data. Yes, the threefold higher risk of progression that some surgeons would offer surgery. And to be honest with you, nobody really knows the right answer.

We need more follow-up data, but we tell people about the results, watch them very close if they decide to have active surveillance. It’s up to you to decide to go back to surveillance. I have to say, we wouldn’t go beyond a year to repeat their MRI. So, we would do, we will do MRI. And so, any suggestion of progression we’d re-biopsy.

Chris Sweeney:
So, to take that a step further, by progression we mean a person going from a Gleason six small volume, low volume to someone who may have a touch of Gleason four not talking about progression to metastatic disease. It’s just less likely to stay on surveillance longer.

Ros Eeles:
Absolutely. Or maybe more cores becoming involved because we know the molecular data, they do have a field effect with quite derangement has shown on a race gauge and the informal areas around where the tumors are.

Silke Gillessen:
Ros, can I ask you something? I, so when we speak about family history, positive family history, I think this is probably one of the easiest things at least to do a family history, when you think about it. But what would you ask in that family history?

Ros Eeles:
Yes. And, and if one thinks that it takes a long time to take a family history in the clinic, the ask is, is there any cancer in your first-degree relatives? So, you’re talking about children, siblings, and your parents. And you want to know type of cancer and the age of onset. And what we’re really looking for is breast or ovarian cancer.

That’s your BRC1 or 2 indication that there may be a higher chance or colon cancer. Uterine cancer, maybe see even other types of kidney urothelial track, which might indicate a Lynch Syndrome or mismatch repair mutation being in the family. And then all you need to then ask is, are there any other relatives? And so, so that’s going to capture over 90% of the significant family histories that we see.

Silke Gillessen:
So pancreatic cancer is not on your list.

Ros Eeles:
It is younger onset for BRCA2, but it’s quite interesting with BRCA2 families, not all BRCA2 families have pancreatic cancer within them. When we have a BRCA2 carrier, we don’t do pancreatic screening, for example, unless we see another case in the family. It is an indicator, and so is melanoma of the skin, and in some BRCA2 families, we also see thyroid cancers, lymphoma, head and neck cancer. In those that smoke it’s much less common.

Silke Gillessen:
Okay. Interesting. And do you ask for Ashkenazi ancestry or other ancestries specifically?

Ros Eeles:
Yes, we do actually. It’s also worth making the question worded so you say where people from because of the history of migration and tragedy that causes migration, due to disruption in countries, sometimes people change their names.

So, it’s often worth saying what your actual origins as well. So, it isn’t just Ashkenazi origins that might be important. So, for example, we do see certain mutations, for example, the Icelandic population with BRCA2 mutation carriers be more common in that population.

Chris Sweeney:
Can we maybe just also ask about Lynch Syndrome and prostate cancer. I think That doesn’t get enough attention sometimes. What are your thoughts about patients with a family history of colon cancer?

Ros Eeles:
That’s a really good question. So, after the impact data that we’ve just talked about a minute ago, for BRCA1 and BRCA2, after we finished recruiting those mutation carriers, we then have been recruiting people with germline mutations that predispose to Lynch Syndrome.

So MLH1, MSH2, and MSH6, we haven’t been recruiting PMS2 carriers cause they’re much rarer. And we’re just analyzing the data at the moment and preparing analysis for submission for publication. So, from the published data so far, the indication is they may have approximately a threefold increased risk of prostate cancer.

We don’t know if it applies to all of those genes that I mentioned. It may be some rather than others, but this one paper from Manchester, from Barrow, that suggests that the tumors may actually also be more aggressive in some of those gene mutation carriers. They are, of course, as you know, with the colon cancer literature often very responsive to immunotherapy.

So of course, there’s the possibility if you have prostate cancer and you had a mutation in a mismatch repair gene, would you be more likely to respond to the immunotherapy, particularly for metastatic disease? And we’ve certainly seen reports of, of dramatic responses.

Silke Gillessen:
Ros, can I ask you a very practical question? So, if someone had, for some reason, a somatic testing in his tumor first, and there wasn’t any BRCA1 and 2 mutation, let’s say. Do you do a germline testing because there’s always written that you could still find one in the germline testing. So, what is your recommendation for that?

Ros Eeles:
Yeah, so that’s a really good question. We’ve certainly seen instances where we have not found a mutation in the tumor and it has been in the germline and what’s happened in the tumor, is that it has lost both copies because it’s so deranged, so that wouldn’t stop me testing somebody.

Silke Gillessen:
Okay

Chris Sweeney:
That’s fascinating, Ros. I never thought of it that way. Yeah. Oh, that’s fascinating. So, the report would show BRCA2 loss on the foundation one and you therefore can’t detect the mutation. That’s actually gold information. Okay.

Silke Gillessen:
No, it’s the other way round, in reality, because it would say negative, if I didn’t misunderstand, it would be saying negative in the foundation one and you should still do a germline. Oh, sorry.

Chris Sweeney:
That’s exactly what I’m saying. Yeah. It’s because the copy number loss of the BRCA gene, so they don’t see a mutation. That’s fascinating. So, this is all gold information, Ros, it’s spectacular.

So we’ve got implications for patients and their family members, patients with regards to close watching them if they’re under active surveillance and potential therapeutic options if they develop metastatic disease and also cascade testing, can we just quickly turn to the world of work that you’ve been leading, and fantastic and how close it is to the reality of risk testing, looking at SNPs and how SNPs different to exomes and recognizing, you talk about that maybe 10% of patients with prostate cancer, we can account for by exome mutations.

There’s so much more in the SNP literature, but we haven’t implemented that in screening tests yet. And how close we are there? That’s a very big question, but what can you say in the next couple of minutes to enlighten us about that?

Ros Eeles:
It’s a potentially very exciting area. So, there are what we call common variants. So, they’re in general present in greater than 5% of the population and they are widespread across the genome they’re so-called SNPs or single nucleotide polymorphisms, and they are numerous.

So, so far, we have what we call a SNP profile or a series of genetic variants that you can type in an individual that, at the moment in Europeans, it would be about 170 of these and in people of African ancestry, about 86 unique ones, and they’re common, they are distributed across the genome and each of them has a small increased risk, but together it’s like having a hand of cards in a bridge game, and I’m taking the total points of those, each person has an individual risk profile.

So, they’re not like the BRCA1 or 2 where there’s a one in two chance that your child will have this high risk, one single alteration.

Here, it’s like you shuffle your pack of cards and you give your child another set of points. It could be totally different from yours. So, each one is very individual to that person. And that raises the possibility that you can do testing of large numbers of individuals in the population and then stratify them. And that therefore raises the possibility, could we use that for population stratification to target our screening?

So particularly in that applies in most countries at the moment is where there’s no set screening of the population because of the finding that you might find just disease that’s never going to bother anyone. And the question is, could you use these to actually identify who you should offer a screening to, not screen other individuals? And then would you enrich the population for the number of cases?

At the moment, the jury is out as to whether this will work, but we’re doing studies to see whether this is a possibility. And each SNP is only about seven p to test. So, the total cost can be quite small.

Chris Sweeney:
That’s fascinating. And this has the potential, I think, from reading your work, Ros, that you’ll be able to find a, the terminology I’m hearing you use is, polygenic risk score using the SNPs to identify patients at risk of having high risk disease, clinically significant prostate cancer, rather than Gleason six and 5% core, which has a lot of men may have

Ros Eeles:
That’s right. That’s what we don’t know at the moment is will it, because you see more cases and people with a higher polygenic risk score. In other words, the higher number of sort of card counting points in your hands, as it were, in the bridge game.

So, the people at the higher polygenic risk score, because they are more likely to have disease, that group might be more enriched for those with aggressive disease.

Also, what we don’t know is how many men you would have to screen in that group to find those that have the aggressive disease, but the modeling suggests that that should enrich for people who have disease that should be treated.

Chris Sweeney:
Fantastic. Fantastic work Ros.

Tom:
Well, we’re coming to the end. One of the questions I’ve got is what are the next big steps in this area? And what is the next big breakthrough for patients in this field?

Ros Eeles:
I think one of the next six steps is getting a multi-ethnic profile for common variants because we want one profile. Now, if we are going to do mass population testing and it is going to work, because populations, for example, in the United Kingdom, in the US, and across Europe, many populations are now very diverse and we’re seeing a lot of admixture.

So, we want one profile. We don’t want to be doing one profile for one population, one for another, because actually it doesn’t work if you’ve got a very admixed population. And then the next challenge is, is the DNA repair gene profile big enough? And so, we’re trying to find other genetic mutations in other DNA repair genes that might also enable you to do further targeted treatment.

And the early research data suggests that the current gene testing profile, even with the NCCN guidelines, is probably not wide enough.

Tom:
Silke, Chris any last questions or should we call it?

Chris Sweeney:
I think we should call it and we should give Ros a big hand of applause for such an eloquent and elegant discussion.

Tom:
Tour de force, Ros. Thank you very much. Fantastic. It’s pleasure. I’ll see you soon. Bye-bye.