Transcript: NCCN Guidelines for Treating Metastatic Castration-Resistant Prostate Cancer: Biomarker-Selected Patients

Dr Atish Choudhury: Welcome back to our 2-part series on NCCN Guidelines for treating patients with metastatic castration-resistant prostate cancer with the Journal of Clinical Pathways.

My name is Dr Atish Choudhury, Senior Physician at Dana-Farber Cancer Institute and Co-Director of the Prostate Cancer Center at the Dana-Farber/Brigham and Women's Cancer Center. I'm also assistant professor of medicine at Harvard Medical School.

I'll be moderating today's discussion. I'm joined today by Dr Alicia Morgans. Dr Morgans, can you please introduce yourself and tell us a bit about your background?

Dr Alicia Morgans: Sure. Thank you, Dr Choudhury. As you heard, my name is Alicia Morgans. I'm a junior medical oncologist and the medical director of the Survivorship Program at Dana-Farber Cancer Institute.

Dr Choudhury: Wonderful. In Part 1, we discussed treatment modalities that have demonstrated survival benefit in mCRPC for genetically unselected patients, and now in Part 2, we'll discuss treatment modalities for biomarker-selected patients.

We'll start with a very basic question. What does that mean, Dr Morgans? What is biomarker selection, and what does that mean for mCRPC patients?

Dr Morgans: That's an interesting question, because I do think that there have been changes and updates over the last few years that can make it a little bit more challenging.

I think of it as 2 separate ways of understanding how we might be able to target different abnormalities in patients, the first being germline genetic testing, and the second being somatic testing.

Both of these testing modalities are important, because they have different aspects that allow us to identify patient populations, but then have other implications as well. The germline genetic testing that I think is so important is used to identify whether patients may be eligible for PARP inhibitors because they can have DNA repair defect alterations, for example.

This also, potentially, has implications for our patient's family, because identifying those alterations that are in cancer syndromes may be important to family members who also have inherited these particular DNA repair defect mutations, and then may, of course, have implications for those family members, in terms of their screening for different cancers.

When we think about somatic testing, we're thinking about identifying mutations that are occurring in the cancer itself. We have to think about not only that we're doing somatic testing, because that's, of course, critical in itself, but how we're doing somatic testing.

Are we testing a metastatic site? Are we testing the primary tumor? Are we testing circulating free DNA or circulating tumor DNA? What exactly are we testing? What's the assay? That can lead to differences in outcomes that may, in some cases, miss an alteration that we're looking for, or perhaps identify an alteration that's not there.

All of this is important. Certainly, you're an expert in this type of testing. I would love to hear you dig into that, when it comes to somatic testing in particular.

Dr Choudhury: Somatic testing is a very complicated issue. Again, I have a particular perspective, which I do think that tumor biopsies of a metastatic site at the time that you're planning treatment for a patient is the favored approach.

Certainly, it can identify a lot of these truncal alterations that we're discussing, particularly alterations in genes involved in the homologous recombination pathway. There might be certain new abnormalities that could be detected there that might make patients candidates for clinical trials.

Thinking about primary tumor versus metastatic site versus circulating cell-free DNA, it depends on what sites are available for biopsy, what the primary was, and when was it acquired, and is there still tumor tissue available from there?

Oftentimes, the primary tissue is biopsies from many years before and there might not be adequate tissue for this sort of testing. When to profile the primary tumor compared to a circulating tumor DNA test in patients who do not have a metastatic site that's amenable for a biopsy and analysis right now is fairly complicated.

There have been recent studies that do demonstrate that most homologous recombination repair deficiencies that are present at the time of metastasis were present as a truncal mutation at the time of the original biopsy.

There have also been certain studies that demonstrated that sometimes, especially when you're doing just a 12-core biopsy, that the areas that you're doing the biopsy of might not represent the pre-metastatic clones.

I do think that there are advantages to profiling at the time that you're going to treat a patient, but circulating cell-free DNA is also a somewhat limited assay. In order to get great quality readouts, you do need a reasonable amount of tumor DNA that's represented in the circulation.

It's not a test that you would do while a patient is responding to treatment, because while patients are responding, their circulating tumor DNA content is quite low. It also profiles the set of genes that are in whatever panel, so it wouldn't identify alterations in other genes that might make patients candidates for other trials.

The main limitation for circulating tumor DNA testing is that it's very difficult to identify copy number losses in circulating tumor DNA. There are some patients who have germline point mutations or somatic point mutations that are easy to find.

If patients have actual two-copy loss of particular homologous recombination genes, like BRCA2 or ATM, which I've certainly seen, those are able to be detected from tumor biopsies and it's very difficult to identify those particular events in circulating cell-free DNA, unless the tumor content is high enough to be able to make those calls.

What's your perspective then on the timing of when you do germline testing versus somatic testing? Let's start with germline. When do you think that we should do that?

Dr Morgans: I try to do germline testing on the initial visit or within the first few months of meeting someone with metastatic disease. We always talk about it early on.

As I said, usually, I mention it at least in the first visit, even if we don't do the testing at that visit, because the NCCN Guidelines stipulate that any patient with metastatic disease is eligible for germline genetic testing.

They also mention that patients who have high-risk or very high-risk localized disease, even unfavorable intermediate-risk disease can consider germline genetic testing. If we're not seeing those patients, our colleagues are often seeing those patients and are probably and hopefully referring those patients for germline testing as well. That's a very early event.

When I'm seeing a patient with metastatic disease, I often talk to them, at least, about sending somatic testing early on, but not necessarily in that first visit. If the patient hasn't had any treatment before, that somatic testing result is not typically going to make any difference in my treatment decision conversations with the patient at that time.

Although, of course, if I have a clinical trial that is, a frontline trial for patients that requires some testing, then we can discuss it. Usually, it's in the first-line mCRPC setting or in the context of a trial when I would do it earlier than that.

Often, the germline testing is happening first, and then the somatic testing is happening a little bit later. I would say that this absolutely varies by practice patterns and by operating procedures within a particular practice, because these are things that we've had to develop to work into our routines over the last number of years, and every clinic is doing this differently.

How are you doing it? When are you incorporating these tests?

Dr Choudhury: I agree, very early for the germline profiling, but I do think that the sweet spot for somatic profiling is after progression on abiraterone or enzalutamide as far as that's where the actual approvals are for the agents that we'll be discussing.

However, if a patient has a biopsy, particularly for diagnostic reasons before abiraterone or enzalutamide, and especially if they have a fairly short interval from the biopsy that they had before to their particular decision on next line of treatment, I will use the original diagnostic biopsy for that particular purpose.

There are patients who have primarily bony disease, and I do think that it is difficult to get great tissue for genomic profiling from bone metastases unless there's a prominent soft tissue component. Those are the appropriate patients for circulating tumor DNA kind of tests.

If the circulating tumor DNA test is nondiagnostic or doesn't identify a particular targetable biomarker, then there remain opportunities to do more tissue testing, either later on or going back to the primary tissue.

Let's say you have a patient and they were identified to have a germline alteration in a homologous recombination repair gene, or they were identified somatically to have one of these alterations, how do you interpret that test to really understand if this is a patient that's likely to benefit from, let's say, a PARP inhibitor?

Dr Morgans: The mutation itself is going to be important. Which gene is this alteration in? Certainly, if this is a complete loss of a gene, of a BRCA2, for example, especially if it's homologous, I do think that that's going to be a patient who has a higher chance of responding.

There is some information sometimes on the report that you'll get that will suggest that a certain type of mutation may be a loss of function mutation. Again, if that is in a gene like BRCA2, or PALB2 is another, then the likelihood that this is going to be something that will ultimately render the disease sensitive to PARP inhibition is higher.

When I think about other alterations, though, I do have some thoughts about a given alteration and its likelihood of response to treatment. All of the gene assessments that were done in studies like PROfound or in TOPARP-B were exploratory and meant to get us started to think about whether certain alterations may be more or less responsive to treatment. They were not definitive by any stretch.

However, looking at those, there are certain genes that I wouldn't necessarily think are going to lead to very profound or rapid responses to PARP inhibition. ATM is one of those. I'm never extremely confident that that's going to lead to a very rapid reduction in tumor burden, for example, for a patient with mCRPC.

Though, that's not to say that I wouldn't use olaparib in a patient with an ATM alteration. It just might not be the first line of therapy that I would want to use after an AR-targeted agent, especially if the patient had visceral metastases that were rapidly progressing or the patient was highly symptomatic.

From your perspective, though, because this is also something that differs a little bit from physician to physician, are there differences in these genes or in the reports that you're seeing? Are there alterations that you would say, "Well, I don't think that that's really going to lead to loss of function, so I'm not actually going to use a PARP inhibitor on this patient."?

Dr Choudhury: These are all great questions. We should probably start with what the actual FDA approvals are. Rucaparib is a PARP inhibitor that's approved in patients who have deleterious germline or somatic alterations in specifically BRCA1 and BRCA2 and they've received a prior taxane.

Whereas, the approval for olaparib is across a large list of homologous recombination genes and does not require a prior taxane. The one exclusion from that list is PPP2R2A, which is probably not a gene involved in homologous recombination at all.

This goes back to what we were discussing in part one with, what is the patient presentation and what are the alternative treatments that we're considering?

In a patient with a BRCA2 alteration and probably PALB2, we can be relatively comfortable that this is at least the population that's likely to respond in benefit. In a patient even with symptomatic progression, you might feel like that's an appropriate treatment option compared to something like a taxane.

However, for something like an ATM alteration, if they have high-volume symptomatic progression, I don't think that they're very likely to have a profound response to a PARP inhibitor, and I would certainly prefer a taxane in that particular setting, and even consider a taxane in combination with a platinum agent.

However, if you have a patient who has, again, very slowly progressing disease, not very symptomatic, and they need something to maybe tide them over, I do think that it's very reasonable to consider a PARP inhibitor in a patient with an ATM alteration to see if that might prolong their progression-free survival.

Similarly, if somebody was on a taxane chemotherapy, like docetaxel or cabazitaxel, and had nice cytoreduction, but their PSA may be starting to creep back up, that might be an opportune time to use a PARP inhibitor in something like an ATM alteration.

It depends a lot on the clinical context, and then also if they have no other treatment options. Either they're not a taxane candidate or their disease has progressed on taxane, but they're still fit for further treatment, but not a candidate for a clinical trial, then certainly, it makes sense to consider a PARP inhibitor, because there are no other options in that particular setting.

As far as the genomics are concerned, certainly, there's evidence that patients who have loss of function of both copies of one of these DNA damage repair genes seem to be more likely to benefit, but certainly, it's not required, in terms of the FDA label.

If a patient has a germline alteration in BRCA2, you don't need to prove that they have loss of heterozygosity of the other allele to prescribe them a PARP inhibitor.

However, if you do somatic genetic testing and you see a one-copy loss of one of these genes, that almost certainly is not functionally relevant and that patient is almost certainly not going to respond to a PARP inhibitor. There are some subtleties, in terms of interpreting these genetic tests to understand who's likely to respond and who's not.

Now, let's switch over to immunotherapies. What test do you use to see if somebody is a candidate for pembrolizumab?

Dr Morgans: We would need to test for microsatellite instability. I often usually test for tumor mutational burden as well. This is all done through somatic NGS testing that I have sent out in the past to different companies that can do this.

Of course, some institutions have in-house testing, where this can also be done. To emphasize, this is a somatic test. That's either done on the tissue itself, which is the way that I like to do it. Just like you, I prefer to have a biopsy that's relatively recent, hopefully within the last year, from a metastatic site.

This is another opportunity if you do not have that. If you have bone-only metastases, if you have a patient who doesn't have circulating tumor DNA or doesn't have a large amount of circulating cell-free DNA, we can use the primary prostatectomy specimen or even biopsies if you have sufficient material to do that testing.

For pembrolizumab, which is the only immunotherapy that's—outside of sipuleucel-T, of course—that is approved in this setting, we need to have microsatellite instability-high or tumor mutational burden that's greater than 10.

It's interesting that I've not had any patients by TMB respond to treatment. Certainly, I have had several patients who have had MSI-high status who have had a nice response. That is only somewhere around 3% to 3.5% of patients in most populations, though, at least in the US, from what I can tell.

It's not especially common, but again, if you don't look for these kinds of things, you won't find it. If you do look and you find it, this can be a highly effective and relatively well-tolerated treatment for most patients.

Dr Choudhury: As you know, IHC testing for mismatch repair proteins is very common in colon cancer. Do you ever do that for your prostate cancer patients?

Dr Morgans: I do not. I have had that turn up a few times on my testing at my previous institution where it was done. I don't always know the reasons why that was done, but it has been done.

It is something that can be informative, but I typically have done the NGS testing anyway, because I want to know the entire mutational profile for a particular patient. Is it something that you normally send off or that you would use?

Dr Choudhury: I agree that the genetic testing is preferred for a few different reasons. One is that you can find the actual mutations, and then see what the tumor mutational burden is and see the mutational pattern and if it's consistent with a microsatellite unstable phenotype.

Sometimes, there is limited tissue available that there's not a specimen that's amenable for genetic testing, but you might have a few slides available for IHC.

If you do see loss of one of the mismatch repair proteins on IHC, you can consider that patient to be in that category of mismatch repair-deficient and still prescribe pembrolizumab on-label. It does allow us to use it in those particular circumstances if there's not tissue that's amenable for actual genetic profiling.

The last thing that I wanted to talk about before we moved off this topic is I mentioned briefly some other genetic alterations that might suggest candidacy for other types of biomarker-based treatments that are more in trials. Are there any that you think of off the top of your head or any trials that you have in mind?

Dr Morgans: PTEN alterations for use with ipatasertib is a potential. This was something that was described in the IPATential150 trial in combination with abiraterone in an mCRPC population. For those PTEN alterations, it did seem to have somewhat of a signal.

Although, the trial overall did not seem to lead to a survival benefit, at least at this point in the analysis. It's not something that has changed standard of care, but it's definitely of interest.

With that particular drug, I would say that the toxicity profile, at least reported in the trial, was a little bit concerning for me, in terms of the degree of diarrhea, perhaps, and some other maybe difficult-to-control adverse events.

I say that not having used the drug, so I don't have experience with this particular agent. Of course, that should be considered as someone simply reading the manuscript or reading the data that was presented. What are your thoughts on the potential and this ipatasertib approach?

Dr Choudhury: I think AKT inhibition is a very interesting pathway. Capivasertib, which is the AstraZeneca version, is in trials in hormone-sensitive prostate cancer. Certainly, I do think that patients with PTEN alterations very well might have a benefit, in terms of prolongation of progression-free survival using it earlier in the space.

As you know, many, many patients have a very prolonged response to abiraterone on its own, so adding the additional toxicities upfront for patients who are going to be on treatment for a very long time, it is going to be interesting how we do choose to incorporate this in our practice.

Certainly, there are other pathways that are relevant in prostate cancer, Wnt signaling pathways, MAP kinase signaling pathways, cell cycle pathways. It's very early days, in terms of testing inhibitors of those particular pathways and in terms of translating it to potential clinical benefit. AKT is probably the next target that is likely to show benefit, at least in some populations.

To conclude this particular section, I think that Dr Morgans and I are fairly aligned, in terms of the importance of biomarker testing and the importance of doing the genetic profiling, the germline early on so that patients have time to discuss this with their family and their relatives.

Certainly, we want to make sure that this is done in a way that everyone, all of our patients have access to this sort of testing and it doesn't promulgate any disparities. Do you have any closing thoughts, Dr Morgans?

Dr Morgans: To emphasize what you said and to echo that by encouraging everyone to do the testing, because if you don't do the testing, certainly, patients won't be able to inform their families if there are germline alterations that can be identified, and we can't use the data for therapeutic approaches. Doing the testing is, of course, step number one.

Dr Choudhury: Absolutely. Thank you so much, Dr. Morgans. Thank you for joining and thank you to the Journal of Clinical Pathways for the opportunity to have this discussion. If you haven't already watched Part 1, it will be linked in the description below.

For the latest updates on topics related to the development, implementation, and evaluation of clinical pathways, please check out the Journal of Clinical Pathways website at www.journalofclinicalpathways.com. Thank you so much.