Read the Full Video Transcript

Alicia Morgans: Hi, my name is Alicia Morgans, and I am an associate professor of medicine in medical oncology at Northwestern University, and I am delighted to have here with me today Dr. Petros Grivas who is an associate professor of medicine at the University of Washington, where he practices medical oncology at The Fred Hutchinson Center and the Seattle Cancer Care Alliance.

I'm so happy to have you here. Thank you for your time today, Petros.

Petros Grivas: I'm very very delightful to be here and thank you for inviting me.

Alicia Morgans: Of course. So, you know, we started this new thing, these quick takes, and I really want to kind of hone in on a fast snippet, high-yield information so that we can educate people as rapidly and efficiently as possible. And today I wanted to speak with you about how we think about biomarker testing in patients with urothelial cancer with a real emphasis on muscle-invasive disease and metastatic disease. So, this landscape has become increasingly complicated. It has really affected our day to day treatment algorithms.

Who are you testing, when are you testing them, and how are we using these results? Let's begin with that.

Petros Grivas: Thank you, Alicia. Great question in a very rapidly evolving landscape. I would probably focus my answer in the metastatic urothelial cancer setting where we have the most data in terms of clinical utilization of putative biomarkers in urothelial cancer. There is a lot of clinical trial landscape that is evolving in earlier disease settings. Neoadjuvant, adjuvant, even non-muscle-invasive disease, but the role of biomarkers there, I think it's not that clear. The data is emerging.

If we focus our discussion in the metastatic urothelial cancer setting, I think most of us utilize Next Generation Sequencing, and it's most commonly applied in the tumor tissue that can be available usually through our type of specimens, TURBTs or cystectomy specimens or nephroureterectomy upper tract disease. And we use Next Generation Sequencing to analyze the tumor profile and what we call fingerprinting of the cancer and identify potential genomic alterations and mutations, copy number variations, fuse into arrangements and else that can potentially be targets for targeted therapies in the context of clinical trials or approved agents, and also putative biomarkers of response.

There are multiple commercial vendors out there. I don't need to mention particular names, but they're a number of vendors who offer this opportunity for genomic sequencing of the tumor tissue, and the results usually come back in the form of a report that outlines particular known genomic alterations that have known significance, but also what we called VUS, which is Variants of Uncertain Significance. The later has to do with finding alterations, mutations, etc., that do not have a known impact on the function of the protein. So when the provider receives the report, has to go through all those pages, figure out what these mutations mean, and that's the tricky part, and I think education's needed in that regard.

In terms of what to do with the results, there is what we call actionable items. In urothelial cancer specifically, I think what comes to mind in terms of actionability is for providers to look for genomic mutations or fusions in the five or plus growth factor receptor, FGF receptor, specifically isoform two and isoform three because now there is an FDA approved agent called Erdafitinib for tumors of metastatic bladder cancer that exhibits FGF instead of two of three mutations from fusions. Now the tricky part is, which mutations and which fusions to focus on, and I think that's a very confusing and challenging part. What needs to happen is for those mutations and fusions to be activating, meaning that activate the receptor to drive signaling pathways of single aid tumors in proliferation and progression. And in those reports, usually there is some explanation of the makings of those mutations and whether or not they act as a function of the protein if they drive signaling through their check receptor.

There is some data that have been published and also if I recall correctly, might accompany the approval of this compound that call out specific mutations, but I think it's important for the provider to review its mutation or fusion and to evaluate whether this indeed acts as a function of the protein. And they can do that by either contacting the vendor who does the sequencing or looking at publicly available databases, because not all mutations are created equal, so have to look into specific genomic alterations and see what the impact is.

In addition to that actionable item, there are multiple other chemical trials that could potentially use mutations or other alterations as screening eligibility for targeted therapies, so that's another potential opportunity. For example, there's a trial across tumor types, a basket trial looking at FGF receptor fusions that can be, you know, just finding something useful in that trial and other therapies as well.

The last point to make is that the VUS findings, Variants of Uncertain Significance, those, I would say are not known, but they made it some review, ideally by some expert because more knowledge is being gained all the time. And the last point is those somatic testing of platforms, they look for particular alterations in the tumor tissue for the cancer, and they do not address the germline question, which means hereditary cancer predisposition. However, I think it's ideal even if it isn't available to another counselor there to review those reports to see if there are any hidden hints for germline mutations that can trigger a dedicated separate germline testing that can probably benefit the patient and broader family in terms of cancer predisposition, which is a separate question. And of course, somatic testing of the tumor cannot replace germline testing, but can potentially give some hints before chemical decision making for that part.

So I think that applies to the tumor tissue. There is some other discussion can be done on the separate question about circulating tumor DNA, which has less data at this point, but emerging data, but that's kind of a quick guide, you know, of how someone can review the genomic sequencing reports using the metastatic disease setting, and most of us do it at the time of diagnosis of metastatic disease. Someone can argue that it might use full-duty [inaudible 00:07:08] at the time of progression of particular therapies. However, logistical factors, you know, accessibility to the tissue, you know, cost, inconvenience, come to play and that's where liquid biopsies may be having a role in the future.

Alicia Morgans: Absolutely. And we're going to get into the FGF alterations a little bit more in a separate recording, but just to expand a little bit before we wrap this up, the other place where these NGS tests can be useful is if you have a patient who is not cisplatin-eligible but is chemotherapy-eligible, and you're in this in-between where you're thinking about, should I give this patient with new metastatic disease a carboplatin-based combination or single-agent chemotherapy, or should I think about, you know, one of the checkpoint inhibitors. You can, in many of these reports, actually get PD-L1 or PD-1 testing within that report, or at least I do in the reports that I get and you have to remember that you're looking for, if you're using pembrolizumab, you're looking for a particular assay reporting a CPS score where this is going to be different if you're looking for atezolizumab or you're looking for a different, you know, a different type of a report and a different threshold.

So for pembrolizumab, it's really looking for a CPS score greater than 10, and Petros, I don't know if you use atezolizumab in this setting or not, but I believe that's greater than five on that assay.

Petros Grivas: It's a great question, Alicia. So also very, very corrective point about there is a... We [inaudible 00:08:50] genomic alterations that other biomarkers that are relevant clinical practice in metastatic urothelial cancer, and you mentioned one of the most commonly used nowadays, which is PD-L1, by immunohistochemistry that quantifies a protein expression in a tumor tissue and as you mentioned, the only setting where there is an indication to test PD-L1 is in the first-line setting of patients who never received chemotherapy for advanced disease, and they are not fit for cisplatin, as you mentioned, but they're still fit for carboplatin chemotherapy, and as you mentioned, the FDA and the EMA have made the label change last year that said that those patients in the first-line setting who are not fit for cisplatin... But they are considered to get checkpoint inhibitor and there are two of them approved in that setting.

You mentioned pembrolizumab and atezolizumab. In order to use those, either of those checkpoint inhibitors, the tumor tissue has to have what we can PD-L1 high package or high expression based on a companion diagnostic assay. For pembrolizumab, use the Agilent Dako 22C3 acid with a CPS score 10 or more, as you mentioned. For atezolizumab is the VENTANA SP142 assay that measures PD-L1 expression only in tumor-infiltrating cells but not in tumor cells. And then that assay, 5% or more, as you called it, is considered positive for PD-L1s.

So based on either assay, it has to be the right assay for the right agent, not mix and match. But based on the right assay, for the right agent, if there's a PD-L1 high status, atezolizumab or pembrolizumab can be considered in those patients. So there is some practicality in testing PD-L1 in that first-line setting because of the safety and legal things. The tale here is that the US only... The FDA allowed the use of pembrolizumab or atezolizumab in this first-line setting of cisplatin-unfit patients, also for those who are unfit also for carboplatin. They are too frail to seek, let's say, to use any of the carboplatin. In those patients, there's no need to check PD-L1 expression in the US. They're again too frail to receive any of the carbo, what we call platin overall, cis and carbo unfit patients. But this provision is not present in the EMA report, so I think in Europe, the first-line study on cisplatin-unfit patients, with people still have to test PD-L1 based on either of those assays, depending on which agent you might consider.

Having said that, PD-L1 high patients we still do not know whether chemotherapy, for example, carboplatin gemcitabine, or checkpoint inhibitor is better. There are phase three trials that are ongoing and we try to answer the question. We don't have the results yet, but what we know is, for patients who are fit for carboplatin, not for cisplatin, but they have PD-L1 low, in those patients, they may get chemotherapy or a clinical trial, but no checkpoint inhibition if they have PD-L1 low expression and they are fit for carboplatin chemotherapy. So that's another, I would say, relevant biomarker, and of course, a few patients beyond that, they may have...

I would say there probably is less than 10% of patients, they may have microsatellite instability in the tumor, what we call MSI-high and there is an approval of checkpoint inhibition across tumor types. I think pembrolizumab in tumors across tumor types, regardless of the origin of the tumor, if they have MSI-high status, and that's another relevant biomarker, which is being reported in those reports.

The other biomarker being reported, which is still of questionable clinical value, is the TMB, tumor mutational burden, but we're still trying to define what the role of TMB is in urothelial cancer and other tumor types, so this is being reported, but we do not know yet how to utilize it, and this is being tested in clinical trials.

Alicia Morgans: Great. Well, I appreciate your time in this discussion today. Very, very useful, and as always, an enjoyable time talking with you. Thank you so much, Petros.

Petros Grivas: Thank you, Alicia, for your great questions.