The Role of the PD-L1 Checkpoint in BCG Resistance - Max Kates
November 15, 2020
PD-L1 was found to be a mediator of bladder-cancer stage progression in a 2007 Mayo Clinic study by Inman et al. A primary observation was that as local tumor stage advanced, levels of PD-L1 expressed by tumor cells increased. Another key discovery was that the majority of patients in whom Bacillus Calmette-Guerin (BCG) immunotherapy failed had “extremely intense” PD-L1 expression within the BCG granulomas. Inspired by this early work, Dr. Kates and colleagues sought to answer the question: Does the PD-L1 checkpoint regulate BCG therapy response?
This talk, first given at BCAN’s Think Tank 2020, is freshly offered here for visitors to the UroToday Bladder Cancer Center of Excellence, along with a discussion about the evolving story of adaptive immune resistance and earlier treatment in bladder cancer.
The study by Kates et al included 63 patients with non-muscle-invasive bladder cancer (NMIBC): 31 BCG responders and 32 BCG nonresponders. In addition to demonstrating that a significantly greater proportion of nonresponders were PD-L1 positive, important discoveries were made regarding the roles of CD4 and CD8. Further, Dr. Kates and fellow investigators expanded this cohort to explore gene expression relative to BCG response. They also examined the potential for PD-L1 combination therapy because, while the anti PD-1 monotherapy pembrolizumab was approved for patients with BCG-nonresponsive, high-risk, CIS-containing NMIBC in January 2020, PD-L1 expression was not associated with a complete response.
Dr. Ashish M. Kamat, Professor of Urology and Cancer Research and Wayne B. Duddleston Professor of Cancer Research at MD Anderson Cancer Center in Houston, Texas, USA, introduces and engages Dr. Max Kates, Assistant Professor of Urology and Oncology, Brady Urological Institute, Johns Hopkins University School of Medicine, and Director of the Bladder Cancer Program, Baltimore, Maryland, USA, in a post-talk discussion.
Max Kates, MD, Director, Bladder Cancer Program Assistant Professor of Urology, Department of Urology and Oncology, Brady Urological Institute, Johns Hopkins Hospital
Ashish Kamat, MD, MBBS, President, International Bladder Cancer Group (IBCG), Professor of Urology & Cancer Research, MD Anderson Cancer Center, Houston, Texas
Ashish Kamat: Hello, everybody, and welcome to UroToday's Bladder Cancer Center of Excellence. I'm Ashish Kamat from MD Anderson Cancer Center. It's a pleasure to introduce you all today to Max Kates, who really needs no introduction. Even though he's been in the field relatively for a short duration of time, he's made a huge impact both locally and internationally in his work with bladder cancer. Max is an assistant professor at the Brady Urological Institute, John Hopkins Medicine Hospital in Baltimore.
Max, it's a privilege and pleasure to welcome you today.
Max Kates: It's a pleasure to be here. Thank you so much for having me.
Ashish Kamat: So, Max, you gave many such talks, but you gave this talk at the BCAN Think Tank, which was virtual this year, on BCG unresponsive disease and the role of PD-L1 checkpoint inhibition. That garnered a lot of attention for obvious reasons with the explosion of activity in this space. So the field is all yours right now.
Max Kates: Wonderful. Thank you, Ashish, for having me. It's really a pleasure to talk to you about what is a really exciting topic in our field. So I'll be talking about the role of the PD-L1 checkpoint specifically in BCG resistance. I have no disclosures.
So the BCG mechanism of action is really multifactorial. Ever since Ratliff in the 1980s demonstrated in animal models that you needed a thymus for BCG to work, we've known the importance of T cells for BCG activity. But more recently, we've come to appreciate the role of the innate immune system and direct cytotoxicity as mediators of BCG response.
In this seminal paper by Claire Biot and Matt Albert's lab, BCG was shown to increase all T cell populations, and these populations increased further with multiple doses of BCG. Similarly, when T cells are depleted, innate cell populations decreased. So this suggests that T cells are important for innate immune response as well to BCG.
Some of the early work on the role of PD-L1 in BCG response was done by Brant Inman in Dr. Kwon's lab at the Mayo. They demonstrated that PD-L1 expression via IHC is highest in CIS and in areas of diffuse lymphocytic infiltration. An interesting finding in that paper, which was back in 2007, was that strong, positive PD-L1 staining of 11 of 12 patients with BCG granulomatous who had not responded to BCG therapy.
So, basically, the hypothesis here is that the immune checkpoint may be highly involved in regulating intravesical therapy response. So we really wanted to further this hypothesis in a next-generation version of that study. So we assessed immune cell expression on 31 responders and 32 BCG non-responders. PD-L1 expression via two different antibodies, SP-142 and 22C3, were found to be associated with recurrence and progression. We then looked specifically at this 25% of BCG non-responders that were PD-L1 positive and found that, really across the board, CD8 cells, which are here in pink, colocalized in a tumor microenvironment with PD-L1 positive cells, which are in brown here. Interestingly, in all cases, the PD-L1 tumors, the PD-L1 positive tumors, had high expression of CD8 positive cells but were completely devoid of CD4-expressing cells.
So what's going on here? So we attribute these findings to adaptive immune resistance whereby CD4 T cells recognize tumor antigen. This then leads to an antigen-specific T cell expansion, including CD8 cells, which are then able to kill the tumor and further propagate a memory response through CD4 cells. However, when cancers then hijack the PD-1/PD-L1 checkpoint, they silence these CD8 cells, which essentially make them still present in the tumor microenvironment, but make them ineffectual. So this limits their ability to both kill tumor and to propagate this memory response through the CD4 feedback.
We should note that more recently we have greatly expanded this cohort and performed RNA-Seq on BCG responders and non-responders, and that cohort, PD-1, and PD-L1, gene expression does not appear to be associated with BCG response. So, if one mechanism of BCG resistance is adaptive immune resistance mediated through the PD-L1 access, what is the role of anti-PD-1/PD-L1 monotherapy and combination therapy in BCG unresponsive disease? This question is obviously crucial as we have seen multiple drug approvals in the last several years in the advanced bladder cancer setting. Anti PD-1 monotherapy is now FDA-approved in the United States for BCG unresponsive, CIS-containing non-muscle invasive bladder cancer, with multiple other studies on the way.
Now, one question is the role of PD-L1 expression to predict response to these therapies in BCG unresponsive disease. In keynote 057, PD-L1 expression was not associated with a complete response to pembrolizumab. So we wanted to further interrogate the potential for PD-L1 combination therapy. So we compiled an RNA-Seq tissue cohort of paired BCG non-responder samples before and after BCG and, in each column, look at the ratio of gene expression in the post-BCG tissue compared to the pre-BCG tissue.
We specifically looked at genes associated with therapies currently FDA-approved for advanced bladder cancer, so FGFR3 and/or Gefitinib or Nectin 4 and enfortumab vedotin or PD-L1 and PD-1 or, in the case of Trastuzumab, a Trop 2 antibody-drug conjugate that is in late-state clinical trials for advanced disease. What we see is that clearly, among BCG unresponsive patients, there's increased expression of at least one, if not multiple, relevant genes and that there are distinct subgroups with differential gene expression. This in our mind provides some rationale for combination systemic therapy with PD-L1 inhibitors in the BCG unresponsive disease space.
Similarly, there's also a preclinical rationale for anti-PD-L1 combination therapy with local therapy, in this case with interferon gene therapy. In this BBN model of bladder cancer, we see that the lentiviral interferon recruits CD8 cells to the tumor microenvironment. This is very similar to how BCG works in terms of its cytotoxic T cell recruitment of CD8s. Similarly, lenti interferon gamma upregulates PD-L1 expression.
The two concepts that the lentivirus interferon gamma both brings CD8 into the tumor microenvironment and upregulates PD-L1 likely in those same areas leads to the idea of combination treatment. In preclinical animal models, this combination was associated with increased antitumor activity. Similarly, among patients treated with Adstiladrin therapy, five of eight patients with tumor recurrences had increased immune checkpoint expression.
So, in conclusion, a cohort, and this is around 25% of patients, with non-muscle invasive bladder cancer undergoing BCG may have evidence of adaptive immune resistance mediated through the PD-L1 access. PD-L1 positive IHC expression may be associated with BCG response, though this has not yet been validated with PD-L1 gene expression. There is a rationale for exploring anti-PD-1/PD-L1 therapy in combination with other systemic antibody-drug conjugates and targeted therapies, as well as with other local intravesical therapies.
I just want to thank you for listening and to particularly thank my lab and my close collaborator on much of this work, Dr. Woonyoung Choi, here at Hopkins. Thank you.
Ashish Kamat: Oh, that was excellent, Max. I mean, you presented a lot of heavy data in a short period of time, but in a very concise and succinct manner. I have to compliment you for doing that not just today but on most of your presentations that I have had the pleasure of listening to. For our listeners and viewers, if you had to summarize a bottom line as to where the research is pointing us when it comes to mechanism of BCG response, but also salvage mechanisms after BCG has not worked, what would you say?
Max Kates: Yeah. To answer the second question first, we know not all resistance to BCG is the same, and there are probably several pathways that are relevant. In this situation, I talked about adaptive immune resistance that could be overcome, for example, with PD-L1 inhibitor. But there are likely several other pathways that may not be overcome with a PD-L1 inhibitor and may be overcome with a different therapeutic, and so I think it's essential that we begin to try to rationally understand these different cohorts in order to understand what the second-line therapies should be and understand that they may not be the same for all patients.
Ashish Kamat: Yeah. Now, you raised a good point. One of the things that we and others have worked on is the trained adaptive immunity in response to BCG. That's actually the basis of the clinical trial that we have that's looking at BCG as a vaccine against COVID, a nonspecific booster of the epigenetic changes that occur. Do you think there is any role that that pathway plays when it comes to bladder cancer activity of BCG?
Max Kates: Yeah, I do. When you look at some of these patients who go on to recur after BCG and you look at their tumors, BCG floods CD8 cells really for all comers. So the cytotoxic T cells are in these bladder tumors with BCG, but yet a lot of those, you know, a lot of those cytotoxic T cells are not doing anything because the tumors are recurring, nevertheless. So this is what we try to tease out, is why are these CD8s being silenced? Is it always through the classic PD-L1 checkpoint or are there other reasons? I think that it's still an evolving story.
Ashish Kamat: I agree with you there absolutely. With the explosion of I-O therapy in all cancers and especially bladder cancer, and, of course, we were able to get pembro approved. When I say, "We," I mean, Mark and the community, and, full disclosure, I was there at the FDA with them, but we got them approved for BCG unresponsive bladder cancer. As you showed, there's a movement of these I-Os in the earlier disease state of bladder cancer, and you showed elegant work that suggests that the BCG unresponsiveness is linked to this whole pathway. What's your sense? Mechanistically based or... but also just styled from having been in the field like you are of moving these agents earlier when there has not been an exposure to BCG yet?
Max Kates: Yeah, so I think that there is a very cogent rationale based on our preclinical work and some of our translational work for moving some of these I-Os, and even there's some rationale for consideration of moving some of the ADCs and many... and other systemic therapies earlier in terms of treating the bladder cancer.
Now, the question is the toxicity. So there's the rationale that it can be an efficacious therapy. But when you start moving drugs earlier and earlier, toxicity becomes more and more important, and so I think that's really the balance. I think there's no doubt that there's a lot of potential for its treatment effects.
Ashish Kamat: Yeah. That's one of the advantages of having someone like you that's doing clinical work, but also doing lab work looking at the picture because, that's exactly right, a lot of these agents have activity. They're mechanistically based. We are all excited about it. But we have to remind ourselves and obviously the community that the patients ultimately, for them, it's a trade-off between not only efficacy but toxicity. So you have to factor all of that in. Just because you have a beautiful tool that works really well, it doesn't mean that we can ignore the toxicity that goes hand in hand with it.
So great points. Max, we could chat on this topic for a long, long time. But in closing, since we do have to close, let me ask you what are some of your one or two main points you want to share with the audience to bring this home?
Max Kates: Yeah. What I would share with the audience is that there is a clear rationale, a translational rationale, for I-O and specifically anti PD-1/PD-L1 therapy to overcome BCG resistance, and there is a clear rationale for combination therapy with both intravesical agents and other systemic agents. It's really going to just come down to, as we mentioned, whether the toxicity will accept these combinations.
Ashish Kamat: Great points. Once again, Max, thank you for taking the time from your busy schedule to join us today. Stay safe and stay well.
Max Kates: It was a pleasure. Thank you.