Unlocking the Future of mCRPC Treatment: Exploring Immune Checkpoint Inhibitors for MMR Loss and MSI-High - Sumit Subudhi

November 8, 2023

Program: Beyond Androgen Blockade – New Pathways and Novel Treatments in mHSPC and mCRPC

Part of an Independent Medical Education Initiative Supported by  LOXO@Lilly


Sumit K. Subudhi, Associate Professor, Department of Genitourinary Medical Oncology, University of Texas, MD Anderson Cancer Center, Houston, TX

Neal D. Shore, MD, FACS, Medical Director, Carolina Urologic Research Center, Atlantic Urology Clinics, Myrtle Beach, SC

Read the Full Video Transcript

Neal Shore: Well, hi everybody. And welcome to UroToday, and our online medical education program entitled Beyond Androgen Blockade: To New Pathways and Novel Treatments in MHSPC and MCRPC. I'm Neal Shore, and I have the real privilege and honor to moderate today's discussion following a presentation by a really good friend and colleague, Dr. Sumit Subudhi from MD Anderson. His topic, which he is going to review today, is IC inhibitors targeting immune checkpoints, MMR loss and MSI-high. So thanks very much, Dr. Subudhi.

Sumit Subudhi: Thank you, Neal, for the kind introduction as well as the opportunity to talk about this topic. So here's a presentation outline. I'll first talk about immune checkpoints, then mismatch repair loss, followed by tumor microenvironment subtypes, and then steps forward. So this is a complicated slide that shows the immune tumor microenvironment, but I just want to get two main points out of here. So in this slide, you see cancer cell in the center surrounded by a lot of immune cells. The first point is that not all immune cells are good. The ones on the right of the cancer cell help the cancer grow, the ones on the left help kill the cancer. So that's the first point. And one of the key cells that kill the cancer are CD8 T-cells. The second point I want to make is that the goal of immunotherapies is to shift the balance of this immune tumor microenvironment so you have less of these bad immune cells and more of the good ones.

So when we talk about immune checkpoint inhibitors, I think it's important to understand basic T-cell biology. So what we have here is a T-cell with a T-cell receptor. That T-cell receptor recognizes the antigen in the form of a MHC peptide cognate complex. And when this interaction occurs, this does not lead to T-cell activation. The analogy I like to give is this is like putting a key in the car ignition, and turning that ignition, you hear the motor or the car engine running, but the car does not move. So what was discovered in the mid-1980s is that there's a second signal required that has to be given concurrently with signal one. And this second signal's called a co-stimulatory signal. And it's best characterized by CD28, which is expressed on T-cells and its interaction with the B71, B72 ligands expressed on professional antigen presenting cells. B71 and B72 are also known as CD80 and CD86.

When you get these two interactions occurring at the same time, you get co-stimulation, and proliferation and activation of these T-cells. The analogy I'd like to go back to is, this would be the equivalent of signal one putting the car key in the ignition, and then signal two would be putting your gas on the accelerator and then the car moves. The problem is, that if this was all we had, then we'd probably have a lot of car accidents because you need a brake. And that's something that both Jeffrey Bluestone and Jim Allison independently discovered in the early 1990s. They discovered the break on T-cell is CTLA-4, and that became known as an immune checkpoint. But then what Jim Allison did was that he ended up developing a drug that blocks the break CTLA-4. And in mice that had tumors in it, he was able to cure mice. And then this drug eventually went into clinic, and is known as ipilimumab.

I just want to point out that this break is so important that when you knock it out or when mice are deficient of CTLA-4, they actually die of lymphoproliferative disorder within a few weeks of being born. So CTLA-4 was the first immune checkpoint discovered and identified, and then the next two were PD-1 and its ligand PD-L1. These are the three most commonly targeted in clinic, but I do want to point out that LAG is another immune checkpoint that has been recently FDA approved as a target. Because of the work of CTLA-4, PD-1 and PD-L1, Jim Allison and Tasuku Honjo were recognized for the 2018 Nobel Prize in physiology and medicine. So who responds to treatment with anti PD-1 and PD-L1? Well, this here I'm going to show you an example of a melanoma patient, and this you can see is melanoma before getting treatment with the immune checkpoint inhibitor.

And what you're seeing is, is tumor microenvironment and in particular CD8 T-cells that are stained brown within the melanoma tumor, and you can see in red is a tumor margin. So the tumor is right here with very few T-cells inside of it. And a lot of the CD8 T-cells, which are the good T-cells that can kill the tumor, they're hanging right outside of the tumor at the margin. When you give anti-PD-1 treatment to this patient, you can see after a few weeks, the patient's tumor starts disappearing. And when you look at the tumor itself, the immune microenvironment, you can see a lot of T-cells have now come in to kill the cancer cells. So this is a patient who responded clearly to the treatment. Now, we have a patient who did not respond to the treatment. And you can see prior to the treatment, there are very few T-cells located in the margin or even inside the tumor.

And then when you give anti-PD-1 or PD-L1 treatment, again, there are few T-cells present. This is what we call an immunological desert, just very few T-cells. And you can see a few weeks later, the patient's melanoma has actually gotten worse. So what this tells us is that these drugs, these anti-PD-1, PD-L1 inhibitors, they actually work best in patients where there's CD8 T-cells present near the tumor microenvironment. So let's talk about prostate cancer and what's the response rate with anti-PD-1 therapy. So in prostate cancer, the response rate is somewhere between 3 and 6%, 3% in the PD-L1 negative tumors and 6% in the PD-L1 positive tumors. So it's very low. And just in contrast, the response rate with these types of therapies in metastatic melanoma is approximately 30%.

So why is that? So this actually helps explain it. So if you look at the prostate tumor microenvironment, there's very few CD8 T-cells present in almost all prostate cancer patients. So that's why it's less likely to respond to immune checkpoint inhibitors. However, in patients with mismatch repair defects or loss in prostate cancer, you can see there are a lot more CD8 T-cells present. And so these are patients that are more likely to benefit from immune checkpoint inhibitors. So mismatch repair defects are not unique to prostate cancer. In fact, they're most commonly seen in colorectal cancer. But as you can see, there are many other cancer types that are represented, including prostate cancer, here in purple. What we're seeing here in the bottom is a waterfall plot looking at change in baseline tumor regression. And so these patients who have minus 100% have a complete response. And you can see many of them in red are colorectal patients, but there are a few prostate cancer patients also included.

And you can see that these are patients with mismatch repair defects, and the majority of the patients actually have shrinkage of their tumor with a smaller percentage that have tumor that are resistant to this approach and continue to grow. So Wassim Abida at Memorial Sloan Kettering, he was a co-fellow with me, and he actually has described mismatch repair loss in prostate cancer, and these patients were treated with either anti-PD-1 or PD-L1 treatment. And what he found is that the durable response rate is approximately 25% in prostate cancer. That's in stark contrast to what we see in colorectal cancer where a majority of patients actually respond to this treatment. Why is that? We believe it has to do with the tumor microenvironment. So it's not just the immune cells but the stroma cells that contribute to that.

So this is a paper that was published in 2021 performed by the BostonGene group. And what they did was they did a transcriptional analysis, bulk RNA-seq analysis on multiple tumor types including melanomas and they were able to segregate the tumors into four different TME subtypes. There's immune-enriched fibrotic, immune-enriched non-fibrotic, fibrotic, and desert. And then after doing this, they looked at, "How did these patients, when subgrouped like this, respond to immune checkpoint inhibitors?" And they found that the immune-enriched non-fibrotic were the ones that responded the best. So let's take a step back and think about how's prostate cancer, where does the cancer go when it metastasizes? This is work done by Susan Halabi, an analysis of approximately 10,000 patients with metastatic castration-resistant prostate cancer that were treated on phase three docetaxel-based chemotherapy trials.

And what you see here is that approximately 43% of patients have bone-only disease and approximately 30% of patients have bone plus lymph node metastasis. So if you think about how many patients with metastatic castration-resistant prostate cancer have bone metastasis, it's somewhere between 70 and 80%. So when we look at where bone metastasis fits into these immune subtypes, we find that a majority of the bone metastasis actually are part of this fibrotic subtype. And remember the fibrotic subtype also was a poor responder to immune checkpoint inhibitors. It turns out these fibrotic subtypes, whether it's immune-enriched fibrotic or fibrotic itself, have high levels of TGF-beta transcriptional expression.

Our group actually looked in patients with bone metastases versus those without bone metastases and healthy donors. Again, these are metastatic castration-resistant prostate cancer patients and healthy donors here. These are healthy donors, meaning they don't have any cancer. And we looked at different cytokine levels and found that there were elevated TGF-beta-1 levels in the patients with bone metastasis. We then turn to a mouse model. And what's really neat about this mouse model is you can inject a prostate tumor into one bone and the other side is tumor free, and you can actually measure TGF-beta levels in the tumor free and the tumor bearing. And we've found is that in the tumor-free bone, there's significantly less TGF-beta-1 present compared to the tumor-bearing bone, similar to what we see in our patients. So when we treat these mice that have tumor-bearing prostate cancer in their bone, we find that the immune checkpoint inhibitor plus targeting the TGF-beta pathway actually improves survival compared to anti-CTLA-4 alone versus anti-TGF-beta alone. We are now trying to confirm these in ongoing clinical trials.

So with mismatch pair loss, we want to point out that they're not always associated with a high intratumoral T-cell density. So here's one example of a patient who has an immune-enriched non-fibrotic TME subtype. And I want to point out that this patient has a lot of CD8 T-cells. Why do we know that? Because this bubble here is dark blue, which means there's a lot of it and it's a large blue. This is in contrast to another mismatch repair loss that's an immune desert, okay? So in this patient, you can see that there's very few CD8 T-cells. Why? Because it's light blue and a small blue bubble for the CD8 T-cells. And then you have another patient who is fibrotic where there's less CD8 T-cells compared to immune-enriched, but more than an immune desert, but it has a lot of fibrosis in it.

And what I'm trying to say here is that this patient here is the one that would be predicted to respond to monotherapy with anti-PD-1 or anti-PD-L1 treatment, but these other patients would probably benefit from different immunotherapy strategies. For example, the immune desert would probably benefit from a CAR T-cell plus anti-PD-1 or a T-cell bispecific plus anti-PD-1. Why those options? Because there's very few T-cells present, and CAR T-cells and T-cell bispecifics can drive T-cells into the tumor microenvironment. The fibrotic, as I mentioned before, is affected by the TGF-beta pathway. So drugs targeted in TGF-beta pathway plus in anti-PD-1 would be probably more beneficial for these patients. And what I'm talking about now needs to be confirmed in prospective clinical trials.

So conclusions, one, MMR loss is the best currently available predictive biomarker for responses to immune checkpoint inhibitors. That doesn't just include for prostate cancer, but for other cancers. The TME likely influences anti-tumor responses to MMR loss tumors. I believe I showed that to you with the TME subtypes developed by BostonGene. And we believe that better predictive biomarkers will likely incorporate combinations of tumor intrinsic, which includes the MMR loss, plus the immune component, and also taking account the stromal components of tumor microenvironment, and this will likely inform immunotherapy treatment strategies in the future.

Neal Shore: Wow. Sumit, that was fantastic. That was really good. And I think a lot of our audience will want to go back and forth and review your presentation. The science on the immunobiology, the way you presented it was really, really clear. And I couldn't help but thinking throughout, what you're doing in conjunction with your colleagues, professor Sharma and Allison, is the holy grail of trying to personalized our care decision making, picking the patients who will benefit from I-O therapy, a checkpoint inhibitor, but understanding that not everybody will, even if they have MSI-high or an MMR component. But breaking it down into these subtypes is brilliant because our patients with MCRCP, the clock is ticking and we want to make sure they get a therapy that is effective as quickly as possible. So brilliant work, and I think the folks at BostonGene and the work that you're doing with them, this is great stuff. Even though prostate has been, as you described, somewhat of a cold tumor as opposed to melanoma and bladder and kidney cancer, we still have this subset of patients who can have remarkable responses.

Sumit Subudhi: Absolutely.

Neal Shore: So my question for you, and I know we're running up on time, for our busy community oncologists and urologists, what is your recommendation for someone who comes into clinic, they're MCRPC and they haven't had any genetic testing, no next generation sequencing? What do you do at Anderson? Someone comes in, they've got extensive bone disease, some nodal disease, let's assume they may or may not have visceral and they want to be treated, and maybe they've just had a novel hormonal agent, they've not had genetic testing, maybe they've had a quasi-successful family history, how do you approach that?

Sumit Subudhi: Yeah, so we try our best, even in the localized setting, to try to start getting both germline genetic testing as well as molecular tumor testing. And the reason for that is we want to anticipate the patient's future. So even in a localized setting, if they have high grade disease, meaning at least in grade four and five disease, then these are patients that have an approximately 50% chance of recurrence, even if they have definitive treatment with surgery or radiation or both. And so these are patients that includes my father, who was actually treated at a Memorial Sloan Kettering Cancer Center, and he got the genetic testing. And luckily, he's been cured for over two years now. But it turns out he did not have any germline-positive mutations, but his somatic testing revealed BRCA2 mutation. So what that gives us, it gives us, I guess, more opportunities.

So in case my father's cancer comes back, we have PARP inhibitors that are already FDA approved. And even more exciting is the second generation of PARP inhibitors are even supposed to be more effective, and they're going through clinical trials now. So it opens another avenue of treatments that my father has despite the standard ones that are available. So we try to start early, is my point. But in the case that you were talking about, the metastatic CRPC, if the patient... Because sometimes it takes anywhere from two to four weeks to get these genetic results back. So it all depends how fast the patient's disease is progressing. If they're symptomatic, then we probably start chemotherapy immediately, but at the same time, we would get the genetic, both germline and somatic testing so that it could help us think about what the next treatment would be for the patient.

Neal Shore: Yeah, that was great. I love your message. I love your personal experience, your professional experience. Yeah, genetic testing is here. I completely agree with you, being more proactive in the high-risk localized and certainly in the sensitive metastatic. There's a lot we can be doing better. We still, across the board, under-utilize this testing. And the great work that you're doing and looking at other blood-based tests such as TGF-beta to better understand these subtypes, these four subtypes you talked about that BostonGene is interrogating with your research, really wonderful. So thank you very, very much.

Sumit Subudhi: Thank you.