Drake Highlights PARP Inhibitors, Immunotherapy in Prostate Cancer

Charles Drake, MD, PhD, discusses the current status of both PARP inhibitors and immunotherapy in prostate cancer, emphasizing that immunotherapy is still very much a player in the prostate cancer field, with ongoing trials exploring checkpoint inhibitors.

Charles Drake, MD, PhD

Following in the steps of ovarian cancer, PARP inhibitors have begun to show promise in metastatic castration-resistant prostate cancer. Olaparib (Lynparza), in particular, has demonstrated high response rates in patients with DNA-repair defects.

Immunotherapy in prostate cancer, however, has experienced a fair share of successes and disappointments. With a bevy of initial excitement surrounding this treatment option, enthusiasm has waned in much of the community. Yet, a future for immunotherapy may very well exist in combinations, explained Charles Drake, MD, PhD.

“It’s funny; we began with unbridled enthusiasm, came down to total depression, and now I think we are back to cautious optimism for immunotherapy for prostate cancer,” said Drake, director of Genitourinary Oncology at NewYork-Presbyterian/Columbia University Medical Center.

OncLive: What is the excitement with PARP inhibitors in prostate cancer?

In an interview with OncLive, Drake shared insight on the current status of both PARP inhibitors and immunotherapy in prostate cancer, emphasizing that immunotherapy is still very much a player in the prostate cancer field, with ongoing trials exploring checkpoint inhibitors.Drake: The paper that Dr Johann de Bono's group published in the New England Journal of Medicine a couple of years ago suggested that patients with mutations and DNA damage-repair genes demonstrated high responses to PARP inhibitors. The responses were nice and very meaningful, and some of them persisted for a very long time. That is a very important clinical finding.

What is your advice for community oncologists and urologists approaching PARP inhibitors in this patient population?

There was also a paper by the Seattle Cancer Care Alliance by Dr Colin C. Pritchard and Dr Peter Nelson's group where they showed that about 12% of men with advanced metastatic prostate cancer had germline defects in these DNA mismatch-repair genes. These were things like BRCA1/2, and that was fascinating because that meant with a simple genetic test—a germline genetic test—you could identify potential patients. If you add up the germline defects and the new or acquired defects, it is probably high—in the 20% to 30% range. That is important because that means there is a very meaningful treatment option for about one-quarter of men with prostate cancer.Right now, it is in flux. There are large randomized trials—I believe there are 3 of them—that are testing PARP inhibitors in selected patients. There are a lot of unknowns right now. I always ask the experts, “What is the best way to screen patients for PARP inhibitors?” It turns out that the clinical trials are using different screening methods. To be honest, right now, we do not know the best way to screen these patients. The other thing is that not all mutations are created equally. Some of the mutations render the tumors sensitive to PARP inhibition, and others do not. These are data that will come out from the randomized phase III trials.

It is a very interesting time right now. You could easily do genomic testing to predict for PARP inhibitor response, and you could also test the tumor by sequencing. Do you need to do both? Who are the best patients to do it in? What are the chances that the drug works? Here is what I can tell you that is interesting: if you talk to oncologists in academia as well as those in private practice, you'll find out that many of them have tried PARP inhibitors.

What are your thoughts on this shift toward precision medicine?

The simplest test to do is basically a germline DNA test. It is available from a few different companies but one that is used frequently is from a company called Color. The patient swabs their cheek, the test performs germline sequencing on 30 genes, and we get the data back about 2 or 3 weeks later. That catches about half of the patients. It is fascinating clinically, because if the patient cannot get on a trial or does not have access to a trial, how do you know when to use this drug? [Do we give it] before or after chemotherapy? Certainly, [we should give it] after second-line hormonal therapy, such as abiraterone acetate (Zytiga) or enzalutamide (Xtandi). We don't know if we should do chemotherapy first; it is a very evolving field in terms of where and when these drugs are going to be used. There will be a lot of clarity when these trials come out, but right now it is very much a field in flux.We have to be a little careful with the term "precision medicine" because what is happening now is very similar to what happened with targeted therapy and immunotherapy. In the early days, people would present beautiful, anecdotal responses to mutations that they found. For immunotherapy, many early trials of vaccines showed that when you did careful randomized trials, the clinical benefit of these approaches was maybe not as great as we thought.

For example, in prostate cancer, you will hear people say that there are targetable mutations in 80% of cases. Then, they count the androgen receptor (AR). We knew [that] was there; we did not have to sequence the tumor to know that many, if not all, prostate cancers were driven by AR signaling, and sometimes driven forever by it.

What is the status of immunotherapy in prostate cancer?

It is okay to be optimistic, but the optimism needs to be balanced by caution and by the conducting of clinical trials. There was a paper published by a French group looking at a precision medicine clinic and, when you looked at the bottom line numbers, the overall response rate per screened patient was just not as high as everyone was hoping it to be. Again, these are early days and early trials but, in the words of Larry David, it is important to “curb our enthusiasm” to some degree.In prostate cancer, we have tested the immune checkpoint blocker ipilimumab (Yervoy), an anti—CTLA-4 inhibitor, in a randomized phase III trial that we did after chemotherapy. The hazard ratio was around 0.08 and the P value was .053. Apparently, you don't get to round up, so the drug did not get approved. There was clearly some activity of anti–CTLA-4 and there have been some case reports of patients achieving long-term benefit. However, based on that and the experience of nivolumab in the phase Ib trial—where I was privileged to treat some of these patients—the objective response rate was nowhere near what we have seen in other cancers, including cancers of the lung, kidney, and bladder. There was a little bit of a loss of enthusiasm.

Recently though, it has changed. A woman named Dr Julie Graff did a very clever trial. She took patients who were progressing on enzalutamide and added anti—PD-1 therapy with pembrolizumab (Keytruda) and has documented, both in a paper and in some subsequent abstracts, that there is a reasonable response rate—between 10% and 20%.

In another trial, Johns Hopkins Medicine researchers are treating patients with a combination of anti—PD-1 and anti–CTLA-4 inhibitors; this is from my colleague Dr Emmanuel Antonarakis. There are responses to that, as well.

More recent data suggest that prostate tumors can respond to immunotherapy and there are 2 ways to think about that. One is so-called “precision medicine” where they are like, "Let’s pick out the patients who are going to respond." The work that Dr Antonarakis presented was really quite beautiful because he showed that DNA damage-repair mutations may predict for response to immunotherapy, as well. That was interesting, but it was a small number series of patients. Those are some encouraging data.

Meanwhile, Merck is doing 2 trials; one is called KEYNOTE-199. Patients receive an anti—PD-1 agent after chemotherapy. They have recently expanded that trial, so it is probably a positive sign. Then, they are also doing a series of signal-seeking combination trials combining variant PD-1 pembrolizumab with other drugs. There is some enthusiasm and there are definitely some objective responses. Julie Graff’s paper published in Oncotarget last summer was quite nice and provocative; it has really gotten the field moving in the right direction. The data from the Merck KEYNOTE-199 trial, which has several hundred patients, will give us a better picture of patients that respond and the frequency of response.

The other approach is to say that anti—PD-1 is a reasonable target, but could there be another way to target prostate tumors? There is some enthusiasm for these drugs that are being retargeted; the prototype is blinatumomab (Blincyto). There is a drug that binds to B7-H3, which is highly expressed in many prostate cancers, and this is in a phase I study. There is some enthusiasm about other ways to get T cells into tumors in prostate cancer.

Genentech has launched a phase III trial that I am on the steering committee for. This trial is for patients who have progressed on abiraterone, and then they get randomized to either enzalutamide, which has some response rate, or enzalutamide plus the anti—PD-L1 agent atezolizumab (Tecentriq). There is some feeling that perhaps cancer can respond to this.

Is it fair to say that the future of immunotherapy in prostate cancer is in combinations?

If you look at the data with pembrolizumab and the upcoming data with ipilimumab and nivolumab, those data suggest that there is going to be a role at some point in the future. It is certainly fair, and it is what a lot of people are saying. In animal models, we have beautiful data. There are good data and, at least in animal models, combinations can be more than additive. The combination of anti—PD–1 and anti–CTLA-4 was already approved in melanoma, and it will certainly be approved in kidney cancer. Whether you need that combination to have an effect in prostate cancer remains to be seen. There is an ongoing 90-patient industry-sponsored trial run by Bristol-Myers Squibb that is testing that combination. It is a very reasonable idea and combinations might be required.

The question though, honestly, is whether the same combinations that will show effect in prostate cancer are the ones that showed effect in melanoma and kidney cancer. The answer is maybe not. The thing is that the tumor microenvironment of prostate cancer is certainly not the same as melanoma and, to be fair, kidney cancer is different than melanoma, too. It might be that there are different combinations. One idea might be using T-cell retargeting drugs and getting some T cells to the tumor; that would be great. However, when the T cells get activated, then they are going to upregulate PD-1; then, the tumors are going to upregulate PD-L1 and that is going to be bad. Perhaps you are going to need to use one of these retargeting agents plus anti—PD-1 therapy.

Combinations are definitely a great idea and most certainly will be required. However, the combinations might not be the obvious ones. For example, we have data from animal models that we just published suggesting that hormonal therapy is quite a beautiful recruiter of T cells to the tumor. Perhaps initial hormonal therapy plus immunotherapy might be a good combination. That statement is right—the suggestion that, to make an impact in prostate cancer, combinations will be required.

Is there anything that you are working on related to the therapies we have spoken about?

The metabolic microenvironment is very different than other tumors, too. It might be that you need to target some of the metabolic effects. The challenge is also that a lot of people have done these trials in lung cancer, kidney cancer, and bladder cancer. Screening all of these combinations intelligently in prostate cancer is going to start up again, but we are a little bit behind other cancers, regarding thinking about and testing different combinations, in my personal opinion. The one thing that we are trying to do is to move some of these newer agents into the neoadjuvant space. When I was at Johns Hopkins Medicine, we completed a nice trial where we gave a vaccine prior to surgery and it allowed us to learn a lot about the prostate microenvironment. That gave us some ideas for newer combinations.

What we are trying to do at Columbia University Medical Center is to initiate neoadjuvant trials of some of these DART molecules, as well as potential combination trials with anti—PD-1 in the neoadjuvant space. The neoadjuvant space is so beautiful. The great part is that patients are healthier; they are going to get surgery anyhow. To be honest, no one needs surgery tomorrow; prostate cancer has been there for 10 to 15 years, so they can wait a few months. The other things is that, when you resect that tumor, you have plenty of tissue to do any kind of careful bioinformatics or molecular analysis. Through these neoadjuvant trials, we will hopefully rapidly gain some insights into the rational combinations.

We are working very hard to start one of these neoadjuvant trials soon. My goal at NewYork-Presbyterian/Columbia University Medical Center is to have a situation where most of the men who go for surgery don't just get surgery, they get something before their surgery that helps us hopefully help them. However, it also helps us learn how to treat prostate cancer immunologically in the future.