Read the Full Video Transcript

Ashish Kamat: Hello, and welcome to UroToday's Bladder Cancer Center of Excellence. I'm Ashish Kamat, Professor of Urologic Oncology at MD Anderson Cancer Center in Houston. And it is my pleasure to welcome Dr. Bishoy Faltas, who is going to talk to us on novel IO approaches across the spectrum of urothelial carcinoma.

Professor Faltas needs no introduction. He is well-published, well thought, and it ia a pleasure to hear him talk on multiple different topics at multiple venues. Bishoy, I'll hand the stage over to you. Just briefly introduce yourself, and then launch into your discussion.

Bishoy Faltas: Well, thank you so much, Ashish. It's always a pleasure to talk to you and with UroToday. And my name is Bishoy Faltas, and I work at Weill-Cornell Medicine. I'm an Assistant Professor of Medicine, Cell and Developmental Biology, and the Director of Bladder Cancer Research for the Englander Institute for Precision Medicine. I'd like to thank you for this opportunity to talk to you today about novel immunotherapy approaches across the spectrum of urothelial cancer stages. These are my disclosures.

The development of immunotherapy has transformed, as you know, the treatment of patients with metastatic urothelial cancer. At the ASCO 2021 annual meeting this year, we heard three excellent presentations from investigators who are now working to extend these gains to earlier stages of the disease. The first two trials that I will talk about today were presented by Dr. Funt and Dr. Grivas, who tested immunotherapy combination approaches in the neoadjuvant setting.

So what is the biological rationale for early systemic immunotherapy? We have previously shown, by tracing the phylogenetic trees of MeSH primary and metastatic urothelial cancers from the same patient, that in many cases, micro-metastatic clones had already spread at the time of the initial presentation, and thus unlikely to be cured by surgery alone.

We know that neoadjuvant chemotherapy improves overall survival. We know that it is often better tolerated. We know that there is a potential for maximizing the impact on patient outcomes, by administering these systemic therapy drugs, at the earliest point in the natural history of the disease. Another advantage is the tissue availability from the TURBT and the radical cystectomy procedures, which offers opportunities to study biomarkers of response in clinical trials. And finally, surrogate endpoints of responsiveness to therapies, such as pathologic complete response, or PCR, enables early risk stratifications in select patients who could benefit from additional therapy.

At the ASCO 2021 meeting, we heard two actual presentations showing how immunotherapy can be incorporated into the neoadjuvant setting. Dr. Funt, and his colleagues at Memorial Sloan Kettering, tested a combination of atezolizumab with gemcitabine cisplatin chemotherapy, in cisplatin eligible patients. Dr. Grivas tested a combination of the PD-1 inhibitor nivolumab with lirilumab, which is a monoclonal antibody against killer cell immunoglobulin-like receptors, in cisplatin-ineligible patients.

The rationale for combining chemotherapy with immune checkpoint blockers is, that chemotherapy modulates tumor immunity by mechanisms that are distinct from immunogenic cell death. The rationale for combining lirilumab with immune checkpoint blockade is the potential for synergistic activation of both NK cells and cytotoxic T-cells.

With a pathologic complete response rate of 44%, the combination of atezolizumab with gemcitabine cisplatin appears favorably comparable with previously reported early phase immunotherapy/chemotherapy combination trials. Both regimens were well-tolerated, and they were comparable to previous single agents and combinations of immunotherapy that were tested previously.  Taken together, these data show that neoadjuvant immunotherapy combinations are well tolerated.

These are some examples of ongoing Phase III clinical trials. For cisplatin-ineligible patients, there is a trial of BEMPEG, which is a CD122-preferential IL-2 pathway agonist, in combination with nivolumab. There is another trial combining the antibody-drug conjugate, enfortumab vedotin, and pembrolizumab.

For cisplatin eligible patients, there is a combination of an oral IDO inhibitor, linrodostat, with standard chemotherapy, with, or without, the addition of nivolumab.

There are also combinations of durvalumab and pembrolizumab with GemCis chemotherapy that are currently being tested.

Several candidate biomarkers are being studied as predictors of response to immune checkpoint blockades such as single agents, or in combination, as with other immunotherapy or chemotherapy regimens, such as PD-L1 tumor mutational burden, immune gene signature, CDA density, mutations in DDR genes, have demonstrated some predictive potential in the neoadjuvant setting. TGF-beta signatures are also promising as biomarkers of resistance to immunotherapy. Overall, identifying a reliable biomarker of response to immunotherapy has remained elusive.

I'd like to recognize the tremendous progress that was made in bringing immunotherapy to the neoadjuvant space. However, in my mind, there are a few key unanswered questions that you can see here on this slide.  First, what is the optimal schedule, duration, and time to cystectomy, when using these regimens? What are the best surrogate clinical endpoints to predict overall survival? Are IO, in addition to chemo, IO plus IO combinations, IO plus ADC combinations, better than single agents in randomized controlled trials. Which biomarkers predict benefit from neoadjuvant immunotherapy trials? And finally, what is the best sequence of neoadjuvant, adjuvant, and maintenance lines of therapy.

Now, I'd like to switch gears to the third study by Dr. Powles and his colleagues, focusing on maintenance immunotherapy. The JAVELIN 100 study that was published, randomized 700 patients to maintenance avelumab, to the best standard of care, or best standard of care alone.

This was a positive trial, showing a significant improvement in overall survival for patients receiving maintenance avelumab, with a hazard ratio of 0.69 in the overall population. This resulted in the approval of maintenance avelumab for patients with locally advanced, or metastatic UC, who have not progressed on first-line platinum-containing chemotherapy.

We can see that the clinical benefit of maintenance avelumab is consistent across the majority of subgroups in this analysis that was presented by Dr. Powles. My interpretation of this data is that maintenance avelumab should be offered to all patients who would qualify. We see that patients with the luminal molecular subtype may not drive as much benefit, although the confidence intervals are wide, making this finding hard to interpret.

There are several existing classifiers of RNA expression-based molecular subtypes of urothelial carcinoma. These include the BASE47 UNC system, the MD Anderson classification system, the LUND classification system, two versions of the TCGA classification, and the final consensus classifier that was published recently. Several groups have examined whether these distinct subgroups derive more clinical benefit from immunotherapy.

Over the next few minutes, I will highlight the molecular subtypes deriving the highest clinical benefit, along with the relative position on the subtype grid below.

First, we will look at the IMvigor210 data with a K=4, according to the first version of the TCA sub-classification, which showed the highest benefits in the luminal infiltrated subtype.

In the CheckMate 275 data, there was some significant difference, showing higher responses in the basal 1 subtype. But there was also a significant percentage of partial responses seen in the luminal two groups.

In the maintenance setting, using a TCA K=5, we see that the basal subgroup is deriving the highest degree of benefits, with mixed benefits in the luminal subtypes.

And finally, there was a second look at the IMvigor210 data, including the newer endocrine subtype, which is quite rare, but exceptionally responsive to immunotherapy.

When we put all of this together, I think a pattern starts to emerge. And I suggest that the relationship between these molecular subtypes should be viewed more as a continuum. On the left, we see the neuroendocrine tumors, which again are rare, but exceptionally responsive. Moving to the right, we see basal tumors that seem to derive a high degree of clinical benefit as well. The luminal infiltrated subtype and its equivalents are probably also somewhat responsive to immunotherapy, whereas the luminol papillary subtype is probably less responsive, or even resistant in some cases.

In order to leverage this understanding to develop a clinical biomarker, we need a single sample classifier, such as the consensus clustering dual. However, one challenge for setting a biomarker threshold is that the classification of the luminal subtype has significant variability with these cases, falling in both sides of that threshold, as you can see on this slide.

Our group previously described that upper tract urothelial carcinoma has a luminal papillary and a T-cell depleted phenotype. This was subsequently validated by several groups, and this would support the notion that this subtype is less responsive to immunotherapy.

In Dr. Powles' trial, we see that UTUC tumors seem to derive less benefit from maintenance avelumab. However, the data on the relative distribution of the luminal papillary subtype between upper and lower tract tumors was not available.

Another challenge is that we are starting to understand that molecular subtype membership can actually switch over the natural history of the disease, and there will be more data coming up on this.

Another challenge that I would like to talk about here, is that the original model, and even the names of these subtypes, are based on the premise that they predominantly reflect the RNA expression patterns of cancer cells arising from the luminal or basal layers. However, it is important to understand that in practice, these subtype classifiers are applied to bulk RNA sequencing data of urothelial cancer tumors.

We are now starting to dissect the contributions of these different components at a single-cell level. On the left, you see a t-SNE plot of single-cell RNA sequencing data showing different cell types. In the middle and on the right, you can see tissue CyTOF data from our group, showing the spatial relationships between these different cancer cells, stroma, and immune components.

We're also starting to appreciate that even at the level of single cells, there is a certain degree of intrinsic plasticity in the expression patterns of basal luminal markers. This plasticity can be potentially altered by extrinsic factors, such as chemotherapy or immunotherapy.

So to conclude, immunotherapy is making inroads into the neoadjuvant and maintenance settings. Neoadjuvant immunotherapy approaches are promising for cisplatin-ineligible patients. Neoadjuvant IO combinations are safe, and the efficacy signals are promising. Most clinical and molecular subtypes derive clinical benefit from maintenance avelumab. Molecular subtyping offers a framework for understanding urothelial cancer biology. We still have challenges for biomarker development that require addressing some fundamental knowledge gaps in urothelial cancer biology.

Thank you, and I would be happy to discuss this with you and to take any questions.

Ashish Kamat: Thank you again, Bishoy. You covered a lot of very important information in a short time, in a very succinct fashion, so thank you for doing that. With all the work that you have been doing, have done, and will obviously, continue to do, when it comes to correlates between the clinic and the bench, if you could share with me some of your high-level thoughts on where are we today when it comes to,  outside of a clinical trial, where are we today when it comes to counseling patients on using molecular correlates to select for, or select against a particular systemic therapy? Are we there yet?

Bishoy Faltas: This is a fantastic question, Ashish. And I would say that the answer is, obviously, it depends. I would say, in certain, outside of specific gene alterations, so for example, FGFR3 activating mutations or fusions, which would obviously select patients who are likely to respond to FGFR3 inhibitors in the metastatic setting. We haven't really had a lot of success with essentially, multi-gene, or multi-protein, or multi-transcriptome, or genome-wide markers. And again, with the notable exception of MSI high tumors, which are also quite uncommon, aside from that, I still think we have a lot of the work that is cut out for us, because as I mentioned on my last slide, I think we still have some key knowledge gaps as to the significance of these markers, in terms of predicting response.

And also, which I think is not a question that can be, that is not necessarily a separate question, what are the essentially immunogenic mechanisms, for example, or the upstream mechanisms, that result in these signatures? And I think it's going to be hard to translate these signals that we see, even though we are starting to have a good characterization of them. But without sort of addressing these two important gaps, it's hard to translate them into clinical-grade biomarkers.

The other point is technical. It's not one point, it's a host of technical factors that are related to how the tests are done, which assay, is it CLIA certified or not, is it certified by the state? What is the regulatory environment for testing this? Is it a continuum, as I was mentioning, or is it more of a binary threshold? So these are all important factors that need to be considered.

Ashish Kamat: Yeah, I liked that slide and the point that you make is worth reiterating. That all of this is a continuum, and any arbitrary cutoff that we select is going to be probably arbitrary. Because as our audience knows, patients' tumors, the immune system, nothing is black or white, it is a continuum. So that slide of yours is very important, and I'm glad we have that for our listeners and audience to go back and forth and look at.

I do want to ask you one question. You have done work that suggests the FGFR3, the papillary altered tumors, that correlates with sort of a non-T-cell inflamed phenotype, and also there has been some suggestion that it is because of that, that these tumors are less responsive to IO therapy. Based on the work that you have done, I know I'm asking you something a little outside of what you presented, but based on your expertise, do you think it is FGFR driven, or do you think it is actually, T-cell driven, the relative lack of response of the papillary tumors, luminal, papillary, to IO therapies?

Bishoy Faltas: That's a great question. We think that the FGFR3 essentially coordinates the luminol papillary subtype and the T-cell depleted phenotype. There are some, we have some ideas about some pathways that could be implicated in this, and that is something that we are testing.

There are other aspects as well. So as you know, in our paper, we and others have shown that essentially, FGFR3 and PPAR-gamma are expressions correlated in these tumors, and it might be essential, mediated by PPAR-gamma, for example. So I guess, what I'm trying to say is that I think there is some degree of cancer cell-intrinsic signaling, in the form of FGFR3, and/or PPAR-gamma that is communicating with the T-cells in the microenvironment. It is something we do not exactly understand how, but we have some ideas that we are currently testing.

Ashish Kamat: Right. Bishoy, let me flip the stage back over to you, and have you maybe highlight and share some of your top closing thoughts with our listeners.

Bishoy Faltas: Thank you so much. I think this is exciting. I am very glad to see the work that was done by my colleagues and presented at ASCO 2021. We are continuing to try to do more work to understand the best predictors of response to immunotherapy, and hopefully, to try to extend these gains to more patients, and the neoadjuvant, and adjuvants, in other non-muscle invasive, in other settings, as well.

Ashish Kamat: Right. Thank you again for taking the time and sharing your knowledge with us. Hopefully, we will get a chance to see each other in person sometime soon.

Bishoy Faltas: I hope so. Always a pleasure.