Immunotherapy for Metastatic Castration Resistant Prostate Cancer Presentation - Charles Drake
September 17, 2019
Charles G. Drake, MD, PhD, Department of Medicine, Division of Hematology/Oncology, Co-Director, Cancer Immunotherapy Programs, Columbia University Medical Center, New York, New York, USA.
Charles Drake: Good afternoon. I'd like to echo the previous of speakers by thanking Silke Gillessen, Silke and Aurelius for organizing a meeting that's really seminal and really lets us to get up to date on many subjects. This is Columbia actually and if you look across your left there is William Oh sitting on the far side of the park actually. I can see the river, he can see the park and we do switch patients. These are my conflicts of interest, which are probably not as relevant as my general bias towards immunotherapy because I have a PhD and I love immunology. Maybe this'll go forward.
We were charged in this meeting to discuss first things that were standard of care, things that are of practical use to the everyday practicing clinician physician. And so that's going to be fast actually. We'll talk about Sipuleucel-T and the other place where immunotherapy is approved for prostate cancer is in patients who have microsatellite instability. I'll show you some provocative data suggesting that there are other potential patient selection markers including CDK12 and a homologous repair or DNA damage repair defects. Like every other cancer, immunotherapy will not be used as monotherapy. This is true for kidney cancer, melanoma, bladder cancer, and this is what's going to happen in prostate cancer. I'll talk to you about combinations with next-generation anti-androgens and also briefly talk about chemotherapy. And then at the last second, if I don't get belled actually, I'll talk about my personal bias of that. Immunotherapy for castrate-resistant disease has a number of challenges inherently against it actually that might not be as predominant if we were able to move to the castrate-sensitive setting.
This is the trial that led to the approval of Sipuleucel-T. You can see the date on the bottom. This trial had the clever acronym of IMPACT. You'll see that it's very important for immunotherapy trials to have clever acronyms. This was IMPACT actually, and this had a four-month overall survival published by Phil Kantoff. One challenge with the use of Sipuleucel-T, among many is that these data are not so new. Right now it's 2019, this trial and these patients were accrued before the era of apalutamide, abiraterone enzalutamide, and whether this would look better now or worse, we really don't know actually at this point.
The other drug that is approved for metastatic castrate-resistant prostate cancer, the name of the session, in fact as Anti-PD-1, pembrolizumab in particular, this is based on a very comprehensive trial that enrolled one, uno, prostate cancer patient. Actually, there they are. I couldn't even figure out what color they are if you look at that slide, but apparently they did have a response actually. Dr. de Bono did a nice job talking about how to measure MSI in the clinic and particularly men mentioned that the PCR analysis is not particularly good on FFPE and really suggested, and I do agree that we rely more on the IHC analysis and also on sequencing.
How do these patients do? How does pembrolizumab work in metastatic castrate-resistant prostate cancer? There are not a lot of data. The best data are from Wassim Abida, actually Memorial. At Memorial they have a wonderful program where they sequence everybody with the MSK impact assay. And he was able to accrue 1,300 patients with prostate cancer that were tested and as expected, about 2% are MSI positive, 10 of these were treated. And out of those about half responded. And that's probably real actually.
Johann, with Eli Allen published a nice paper in JCI showing that not all MSI tumors are infiltrated for some reason actually. And so, we don't expect everyone to respond. This is different, by the way, from colorectal cancer where the response rate is 70% upwards and some of those patients look like they're cured. Their response is three and five years out. But this is probably what to expect.
That was the end of what's approved. Then we're done. But these are the things that I think are in development. There's a beautiful paper from the Michigan group showing that CDK12 biallelic mutated tumors were heavily infiltrated, might respond to PD-1. And there's a trial testing that directly, that trial is called IMPACT. You can see actually. This is being done out of University of Michigan with some support from PCF of course.
This is ongoing, and this has a been open for about a year. Hopefully, that if you have a patient with biallelic CDK12 mutation, you should send them to Michigan or one of the other sites. Honestly, in real life, that's probably not what's going to happen. Probably in real life you're going to get sequencing, you're going to see this and frankly there's a lot of off label use of these drugs, which is going to complicate the interpretation of trials going forward.
If immunotherapy doesn't work as monotherapy, the idea is to combine an Anti-PD-1 mostly works on the CD8, the killer cells in the tumor. CTLA-4 it works differently. CTLA-4 works on the regulatory T cells. These are a population of evil cells, they're CD4 cells that turn off an anti-tumor immune response. What would happen if you combine Anti-PD-1 plus Anti-CTLA-4? If you take care of other cancers, this schema is exactly, that's now FDA approved in first-line kidney cancer where the response rate is about 40% and the rate of complete response is around 10%.
How does this work in prostate cancer? These are data from a very small trial that we did at Hopkins. My friend Emmanuel Antonarakis was the PI and the answer is not as good. Basically, in kidney cancer I told you response rate 40%, maybe even a little higher, complete responses in the small number of prostate cancer patients, only a handful are valuable by resistant, some of these responded. Perhaps not coincidentally both of the responders did have defects and homologous repair. One was BRCA, one I believe was ATM actually. This is data going forward.
This same experiment is being repeated in a larger form in the form of a trial called CheckMate 650 by BMS. And we saw some of the early data from that at GU ASCO. It was about the same frankly. This is not going to be a home run in prostate cancer.
What about another strategy of adding? Julie Graff and Tom Beer at Oregon had a really seminal observation. They had a patient who was progressing on enzalutamide. They added Sipuleucel-T, the person had a complete response, they're still in complete response. Based on that, Julie started a really nice small trial taking patients who responded to enzalutamide and were progressing and then adding. This is different than other screen. This is adding FIPD-1, pembrolizumab, to progressors on enzalutamide. Julie published a little part of this. We did the molecular analysis of patients. The first 10 patients, three of these look like they had a response. This is a very small numbers. The update will be a published very, very soon, but these are probably real data actually. Adding Anti-PD-1 to enzalutamide is probably not the same as doing them at the same time.
This is the trial that's testing, doing at the same time. This is a large randomized phase three trial. Tom Powles who's here as the PI and Chris Sweeney, who I forgot to put his name on, sorry. Chris, is the co-PI. This is a trial that takes patients who are progressing on abiraterone and doesn't add enzalutamide. It's a different, it starts enzalutamide plus Anti-PD-1 at the same time. And we learned yesterday actually from Mack Roach that there is some magic to timing in an immunotherapy. This timing might be even more critical than in radiation therapy. This is fully accrued, accrued very beautifully and quickly and we'll have results at some point.
Finally, I'll tell you why I think that that the idea of treating patients with metastatic castrate-resistant disease is hard. ADT has actually, everything we do to prostate cancer patients has immunologic effects, surgery, radiation, chemo, ADT. ADT in particular is balanced by these negative immunologic effects. And you can see on your left, you can look at the FoxP3. What happens is after you castrate mice with prostate cancer, they have a dramatic increase in these regulatory T cells that turn an immune response off. And when they're castrate-resistant, they still stay there. On the right is what happens in patients. This is real. These are patients treated with degarelix for two weeks prior to surgery. You can see that after degarelix, you doubled the T regs in the tumor. We have the same data from radiation in animal models. Dr. Reiter did a brilliant trial looking at radiation to humans. We'll see if this holds as well.
But the bottom line is when you push the tumor, the tumor pushes back with negative immunological effects. One way to see if you could do anything about this is with a trial. We're starting a trial where we combine a drug that depletes regulatory T cells. There's a drug that's modified off ipulimumab. It's modified to increase ADCC, so not only does it block CTLA-4, it depletes cells that express CTLA-4. It's called the son of ipi actually. And this is a trial where we're going to randomize men to either degarelix alone or degarelix plus the T reg depleter mostly to see if we really can deplete the T regs in the tumor actually. There will also be clinical and biological outcomes.
But other things happen too. When you castrate either mice or humans, we've done this in multiple cell lines. We have this for multiple clinical datasets. What happens is a cytokine called IL-8, interleukin-8, goes up. Interleukin-8 is bad because what it does is it attracts a population of cells that are also evil suppressor cells. They're known as myeloid-derived suppressor cells. Dr. De Bono in Elimonte, published a brilliant paper in Nature in the last year or so showing that these cells make IL-23 and that might contribute to castration resistance. Our data suggest that the way they get in there is through IL-8. IL-8 comes up with castration. You can see these are human cells. We've done it with mice and with antibodies and drugs. The bottom line is if the mice tumors don't have IL-8, then the MDSCs don't get in actually and this leads to an improved mouse clinical outcome.
What about humans? Well, to test this in humans we have a trial with another clever acronym of course (MAGIC-8). This is a trial it takes men with castrate-sensitive prostate cancer. We actually don't really care in this trial whether they're metastatic or non-metastatic because you've learned through the course of this meeting that it's kind of hard to tell. Is PSMA positive metastatic or not? These are men with a PSA between two and 50 and rapid-ish doubling time. They have a doubling time in the range of 12 months or earlier. And what we do is we try to prime the immune system with an Anti-PD-1 and then they get a short course of ADT based on a paper by Laurie Klotz showing that, oh good, I'm almost done. Showing that four months of ADT is very similar to actually is probably slightly superior to 10 months. Patients get primed with Anti-PD-1, then they get Anti-PD-1 plus immunotherapy. The other arm, it's randomized, the other arm is much better. The other arm gets Anti-IL-8. And that arm we expect not to see a an influx of these evil myeloid suppressor cells. This trial has recently been honored actually with a PCF award that allow us to fully support the correlative studies for which we're deeply appreciative.
In the end, there's ongoing trials. Those are all sort of anecdotal-ish, but I can tell you that KEYNOTE-199 from Merck, which was monotherapy, showed the same thing. CheckMate 650 of Ipi/Nivo showed the same thing. These are kind of data that I think are going to come forward. I think what's going to happen is there's combinations and phase three, which you've all heard of, Anti-PD-1 plus chemo, which seems to work in lung and potentially in bladder. Anti-PD-1 plus ADT, again, there timing might matter. Anti-PD-1 plus PARP inhibitors, which the next speaker, not coincidentally, Dr. Chi will talk about a little bit. And if you're anxiously awaiting data on these phase threes, it's not going to be soon.
With that, I'd like to finish. I'd like to thank the folks in my lab who did this, Areli Lopez-Bujanda did the IL-8 work, which is under review in a challenging review process. The clinical trials were written by Matt Dallos and especially the Prostate Cancer Foundation, among our funding sources. Thank you.