PARP Inhibition in Castration-Resistant Prostate Cancer - Daniel Petrylak

May 24, 2020

Dan Petrylak joins Tom Keane to discuss PARP inhibition in castration-resistant prostate cancer.  In this lecture, Dr. Petrylak discusses the germline DNA repair mutations involved in prostate cancer.  This is a state of the art lecture establishing the rationale for the clinical trials evaluating PARP inhibitors in metastatic castration-resistant prostate cancer.  


Daniel P. Petrylak, MD, Professor of Medicine (Medical Oncology) and of Urology; Co-Leader, Cancer Signaling Networks, Yale Cancer Center, New Haven, Connecticut, USA.

Thomas E. Keane, MBBCh, FRCSI, FACS, Department of Urology, The Medical University of South Carolina, Charleston, South Carolina, USA.

Read the Full Video Transcript

Thomas Keane: Hello, everybody. This is Tom Keane coming to you from UroToday and from the Medical University of South Carolina in Charleston. Today, we have a special treat for you. Professor Dan Petrylak is the Professor of Medicine and Medical Oncology and Urology and Co-Leader of Cancer Signaling Networks at Yale Cancer Center. Dan received his MD from Case Western Reserve University School of Medicine and joined the Yale faculty in 2012. He's currently Co-Director of Signal Transduction, the research program at Yale Cancer Center and is PI on multiple trials in primarily prostate and bladder cancer. However, he is also one of the most up-to-date and educational speakers that I've had the opportunity to listen to. And today he's going to deal with PARP inhibition in castrate-resistant prostate cancer, which is a very attractive concept and it applies to a relatively small percentage of patients with castrate-resistant disease, but it does give us information on areas that can be exploited in the future. So without further ado, I'd like to turn it over to Professor Petrylak. Thank you, Dan.

Dan Petrylak: Thank you, Tom. It's always a great pleasure speaking to you. Tom has been one of my best friends in the business for a number of years, and it's a pleasure to give you this presentation. We're going to be talking about PARP inhibition in castrate-resistant prostate cancer.

So let's think a little bit about in our first slide what the rate of DNA repair mutations are in castrate-resistant prostate cancer. And this is the sum of the germline DNA repair mutations that have been seen in a number of different trials. Overall, we take a look at the 692 patients in seven studies. We found that about 11.8% of these patients have germline DNA repair mutations that are involved in these cancers.

So this basically breaks itself up into a number of different enzyme abnormalities. This includes BRCA2, ATM, CHEK2, BRCA1, and then there are smaller DNA repair mutations, such as PALB2, RAD51, ATR. The microsatellite enzymes, MSH2, MSH6, as well as RAD51C and other that are approximately 1%. But the predominant DNA repair mutation is BRCA2, and as we know, this is also present in breast and ovarian cancer. And if you identify a germline mutation in a patient with prostate cancer, I think it's important that you also start referring the patient members for counseling because this could predispose patients to breast cancer or ovarian cancer. And, again, it's important that the family members be notified.

So what are PARP inhibitors? These are drugs that have multiple functions in the repair of DNA. PARP inhibitors will affect double-stranded DNA repairs, we see in the left-hand portion of this particular slide. They also regulate transcription. They will modify chromatin and they will actually modify the length of the telomere. They're involved in mitotic spindle formation as well as mediating apoptosis. And when we have patients who have a damage repair as we see in the next slide, in double-stranded DNA, you can then, by giving a PARP inhibitor, knock out that single-stranded DNA repair pathway and this leads to something called synthetic lethality. So this can then cause apoptosis in patients with cancer who have these genetic abnormalities.

The first indication that PARP inhibition was important for prostate cancer came from a trial that was published in the New England Journal of Medicine called TOPARP, and Johann de Bono's group took patients who had been on previous docetaxel. These were 49 patients and Dr. Mateo was the person who led this particular trial. We looked at unselected patients. About a third of those patients responded to a PARP inhibitor. But when you started to analyze these patients based upon genomic markers, about a third of patients had mutations in DNA repair and that included BRCA2, ATM as well as the other tumors. Sixty-seven percent had no mutations. Fourteen of the 16 patients with the repair mutations responded to olaparib, and then two of these patients in the ones that did not have repair mutations responded. So there was a significant difference in this particular response pattern to olaparib based upon the DNA repair mutation pathways.

To further confirm this, a study called TOPARP-B was employed, and this took 711 patients who underwent molecular screening. Of those 711 patients, 161 had some form of a DNA repair mutation, and 98 patients were randomized to two different dose levels of olaparib. On the right-hand portion of this slide, you can see that BRCA1 and BRCA2 were the predominant mutations that also included ATM, CDK12, PALB2 as well as other DNA repair mutations.

So this is a heavily pretreated group of patients. All patients received prior docetaxel. About a third of patients receive prior cabazitaxel, and about 90% of patients received either abiraterone or enzalutamide. And these pretreatment characteristics were well distributed amongst both the 300-milligram group and the 400-milligram group. Radiographic progression was taken in three-quarters of patients. About a quarter of patients had metastases to the liver.

The primary endpoint was the objective response by RECIST, and there were also other response parameters such as CTC conversion and PSA responses. So we see from the 300-milligram group that there was a confirmed response rate of 39%, slightly higher in the 400-milligram group 54.3%. If we look at RECIST, it's 16.2 versus 24.2. I mentioned before, that's the composite rate like I said before. When we look at the other individual components, such as PSA response, about 30% CTC conversions, meaning going from five to less than five CTCs, about half of those patients converted. PSA response or RECIST response, about 28% in the 300 group versus 41% in the 400 group.

If you look at the particular gene subgroups, it does seem to be a very, very interesting pattern. The BRCA1, BRCA2s do best when you look at the composite response, 83.3%. ATMs are not quite as good, 36.8%; CDK12s, 25%; PALB2s, interestingly, 57%, but only seven patients in this particular group; and then the other DNA repair mutations is about 20%.

A randomized trial was performed to look at olaparib in patients who had BRCA mutations. This was PROfound and these were men with castrate-resistant prostate cancer who failed one next-generation hormonal agents, such as abiraterone or enzalutamide, and they had to have at least one DNA repair mutation. There are two different cohorts, Cohort A with a BRCA1, BRCA2s and ATMs, that comprised of 245 patients, and the B was the other alterations. Patients were randomized to receive olaparib versus physician's choice, and the lower dose of olaparib was used in this particular study, 300 milligrams BID.

The primary endpoint was radiographic progression-free survival. Secondary endpoints were overall survival, time to pain progression, and confirmed objective response rate. Patient characteristics, again, well balanced in terms of what we see as far as the sites of disease. About a quarter of patients having visceral disease, about 35% of patients having bone-only disease, and 30% of patients having other forms of combinations of disease states. About 65% of patients had prior taxane use, predominantly docetaxel. About a quarter of patients had both docetaxel, cabazitaxel.

So these are the most commonly reported adverse events either in the olaparib arm or the physician's choice arm: anemia, nausea, fatigue, decreased appetite. Those are all seen with olaparib, so you have to monitor these patients' blood counts while they're on treatment. It didn't seem to be terribly different between the physician's choice. Maybe the pattern's a little bit different, but the overall number of toxicities don't really seem to be much different between physician's choice as well as the data with the olaparib by itself.

This is the primary endpoint of the study which is radiographic progression-free survival. In those patients who are a BRCA1, BRCA2 or ATM, as we see there's about a four-month difference in the median PFS. The hazard ratio is 0.34 and that's really impressive and that's statistically significant. If we look at the objective response rate, 33% versus 2.3% in the physician's choice arm, very, very significantly different and very impressive, in my mind.

This is the slide for overall survival, both in the patients who are ATM, BRACA1 or BRCA2 versus everybody else in Cohort A and B or the combined Cohort A and B. We have some pretty impressive hazard ratios as well of 0.64 and 0.67, respectively. About three-quarters of patients in the BRCA/ATMĀ arm on the left are alive at a year and it's 56% at 18 months, a little bit lower, but an 18-month rate in the combination or for all-comers of the DNA repair mutations, but 66% at 12 months for everybody. So, this again is an impressive finding.

It's hard to determine, whether there's really any real difference in physician's choice versus activity in each of these particular groups because these numbers are fairly small. But there is a trend towards a better outcome with BRCA2s, CDKs. The interesting thing is the ATMs in terms of their PFS don't seem to be that much different in physician's choice versus BRCA, same thing in BRCA1. So there may be some difference in how these patients respond in as far as durability response goes to the individual DNA repair mutations.

There are other PARP inhibitors that are being evaluated. Rucaparib is being evaluated in two different studies, the TRITON2 trial and the TRITON3 trial. TRITON2 has been presented several times at the various meetings, and that's the slides up here. This is taking patients with DNA repair mutations, one prior line of taxane-based chemotherapy, no prior PARP inhibitors or mitoxantrone, cyclophosphamide or platinum therapy, and the primary endpoint was RECIST response. These patients were treated into radiographic progression or discontinuation for other reasons.

So this is the efficacy as of April of 2018, 85 patients, predominantly BRCA1, BRCA2s, also ATMs. As we see the clinical characteristics are very, very similar from these different groups. Again, predominantly bone is the metastatic site. About half of patients in the BRCA1, BRCA2s having visceral disease, but overall 28% having visceral disease, 14% having hepatic metastases. So this is the waterfall plot in those patients who had BRCA1, BRCA2 mutations. Response rates approximately 40%, which is what you see with the olaparib, very, very similar. On the next slide, the duration response with patients with BRCA1, BRCA2s where they're patients who have ongoing responses up to 36 months in this particular trial, and again, I think that this is an impressive finding.

So when we start looking at the various DNA repair mutations, we see that, again, this is predominantly BRCA1, BRCA2. The ATMs don't seem to have as robust a response rate as the BRCA1, BRCA2s, but there is one ATM responded by RECIST criteria. Again, this is very similar to the pattern we were seeing with olaparib. So these are not the only PARP inhibitors that are out there in clinical trials. The TOPARP-B, which we talked about before, GALAHAD is looking at niraparib in men with castrate-resistant prostate cancer. The TALAPRO is looking at talazoparib, again, in patients with castrate-resistant disease who've had prior taxanes and progressed on antigen-target therapy.

There is the TRITON3 trial, which also is looking at rucaparib, but this is a randomized Phase III and this is basically allowing for progression on one targeted therapy. Patients are randomized to receive rucaparib or enzalutamide. Rucaparib is the experimental arm, enzalutamide, abiraterone, prednisone and docetaxel as the control arms.

Then we also have some trials that are being done in hormones-tested disease. This makes logical sense, particularly if you'd like to try to avoid antigen blockade. This is the TRIUMPH trial is looking at olaparib in hormone-sensitive patients with biochemical relapse and PSA doubling times of less than six months. TRIUMPH is looking at hormone-sensitive disease, not on ADT with DNA repair mutations.

Now, how are we trying to move forward in the field and improve the response rates of PARP inhibitors? Because remember, only about 10 to 20% of patients have DNA repair mutations. So we want this to be more of a general treatment, of course, for the overall population, and it may be the combining PARP inhibitors with different agents may actually lead to synergistic activity. In the laboratory, next-generation antiandrogens, such as abiraterone, will synergize with olaparib, and the thought is that the synergy occurs because you're down-regulating the levels of DNA repair mutations in these patients.

There was a trial that looked at olaparib plus abiraterone and compared that to abiraterone alone in men with castration-resistant prostate cancer regardless of the DNA repair status. So this is the schema, 142 patients. They were pretreated with docetaxel, could not have any more than two lines of prior chemotherapy and no previous second-generation antihormonal agents, and they randomized to receive olaparib plus abi versus placebo plus abi. These patients were treated until progression. Median PSA 86 in the olaparib arm, 47 in the abiraterone arm. So small trials you can start seeing a little bit of a difference in distribution. Soft tissue disease in approximately 11 to 15% of patients, bone-only disease in about 46% of patients, bone and soft tissue, about 40% of patients overall.

This is the radiographic progression-free survival curve. There is about a six-month difference between the two different arms in terms of the median, and the hazard rate shows 0.65. When you start looking at the different DNA repair patterns, you see an interesting distribution. On the right is the wild type, and the combination of olaparib plus abiraterone does have a better radiographic PFS, than abiraterone alone. But look at the numbers, fairly small, only 35. You do see in the mutated or the partially characterized, you still see the same trends, but the real issue is, is this inducing BRCAness?

So the office concluded that there was reason to go further with further trials because of the activity that was noted. The one thing I didn't show is the toxicity slide, but one of the things that's very, very bothersome is the fact that the toxicities can be cardiovascular in nature, particularly in the combination. So this is something you've got to watch out for, and there's a Phase III trial that is ongoing right now.

This slide summarizes some of the combinations in advanced prostate cancer. These are basically going at different themes of DNA repair. The top, olaparib plus radium makes a lot of sense because it's COMRADE, this is a NCI study. This is potentially seeing if there's synergy between two different DNA repair agents. Of course, radium will cause DNA damage to double-stranded breaks, olaparib is on the single-strand of repair.

Joe Kim at my institution recently presented data at ASCO GU showing that olaparib plus cediranib, cediranib is an antiangiogenesis agent that potentially causes hypoxia and induces BRCAness by reducing DNA repair enzymes. He showed a six-month difference in radiographic progression-free survival in favor of the combination. And the interesting thing is we had a crossover, and there are patients who received olaparib frontline and then went on olaparib with cediranib and responded after they progressed on olaparib by itself.

Checkpoint inhibitors, olaparib plus durvalumab. There are similar studies being done with rucaparib and nivolumab as well. The thought is that by giving these PARP inhibitors you cause more mutations to occur and then they're recognized by the immune system. So those are now underway. I think that these are very, very interesting, and these are ways that we potentially could expand the use of PARP inhibitors in this disease.

In conclusion, PARP inhibition is effective in patients with DNA repair mutations and metastatic castration-resistant prostate cancer. We're going to try to increase the eligibility of patients for these treatments by looking at novel combinations that exploit the biology.

Thomas Keane: Thanks, Dan. That was a wonderful journey through what's being developed currently. I have a couple of questions that I would like to just ask. To your knowledge, so far the data doesn't appear to suggest that one of the PARP inhibitors is more effective than the others are. Which do you feel may be the most effective and also which do you feel might be the more toxic?

Dan Petrylak: It's hard to really say if there's really any difference. I think there's one that the traps the complex a little bit better, but right now I can't see any clinical difference in terms of toxicities or efficacy at this point. So we're just going to have to really rely on experience. We'll need comparative trials later on. The question will be is if you fail one PARP, then can you respond to another or vice versa?

They haven't been all that different in other solid tumor types. I wouldn't expect that they're going to be different in prostate cancer, but you never know. I mean, we may have some surprises.

Thomas Keane: Yeah. And then the second question I have is there are incentives now to do both hereditary and to do tumor-specific genetic testing and counseling, which will open the door for patients that could have clinical benefits for PARP. But when and whom would you recommend should undergo genetic testing? Because this is going to, we're going to see a lot of patients who are going to want to know once this becomes more commonly known. Patients are going to be wondering, well, should I undergo genetic testing? Should I undergo tumor-specific testing? What's your answer to those kinds of questions? Are these germline mutations or are they somatic mutations induced by therapy?

Dan Petrylak: Well, you have both. The germlines are going to require that patients do have genetic counseling. If you have a somatic mutation in the tissue specimen, then that does not require that the family be tested in that situation. So those can be distinguished. There are germline tests that we have. There a variety of different commercial tests that you can spit into a test tube and know if you have BRCA mutations, and then the appropriate people can be counseled in the family. Then we have next-generation sequencing, which is done directly on the tumor cells, and that can tell us whether there are somatic mutations or not.

Thomas Keane: And then at what point would just say that somebody should get genetic testing in terms of somatic mutations, I mean at what point in the disease? Say they get prostate cancer and they're now using it earlier and earlier. It used to be, as you said yourself, this started off in patients with metastatic disease. But if you had a patient today who you felt had a Gleason, say 4 and above, say 4+5, would you advise them to go straight to testing to see if PARP would be of help in the future that you could combine with one of the other novel therapies?

Dan Petrylak: Absolutely.

Thomas Keane: And at what stage?

Dan Petrylak: Well, again, I think you've got two different places to look. Number one, the guideline testing right now basically says that we really should be checking all people's germline right at the time of diagnosis, so that should be done. The issue as to when you should check the tissue, I like to do the most, obviously, when a patient becomes castrate-resistant, that's one time you should be checking the tissue for any sort of somatic mutations. The real question is how many times do you do it later on... on top of everything else?

Thomas Keane: And we don't know that?

Dan Petrylak: We don't know that. We don't know what the evolution is, how quickly you evolve somatic mutations or not. That's not clear, but I think that... I mean, for example, there was a study in, I believe it was in Clinical Cancer Research where they looked at microsatellite instability and found that in sequential specimens you did start seeing upregulation of them. So I think that that's just an interesting observation. When do you say no? When do you stop testing?

Thomas Keane: Right, right. This is data that we've never seen, well, we have seen it in recent years, but this is up-to-the-minute data, and it does point to us that there are certain selective prostate cancer patients who may well do very well from these treatments. And it is interesting to see that as you dial it back earlier in the disease that you're also starting to see responses. Okay, well, thank you once again, Dan. It was a pleasure having you on.

Dan Petrylak: Always a pleasure, likewise. Thank you.