Biological Pathways in Urothelial Carcinoma - Carissa Chu

April 12, 2022

Carissa Chu joins Sam Chang to discuss data on genetic changes and alterations in urothelial carcinoma. The purpose of the study is to characterize human epidermal growth factor receptor 2 (HER2) (ERBB2) and fibroblast growth factor receptor 3 (FGFR3) mutations in a prospectively collected cohort of urothelial cancers. Dr. Chu highlights the key findings from this study.


Carissa Chu, MD, Urologic Oncology Fellow, Memorial Sloan Kettering Cancer Center, New York, NY

Sam S. Chang, M.D., M.B.A. Patricia and Rodes Hart Endowed Chair of Urologic Surgery Professor Department of Urology at Vanderbilt University Medical Center

Read the Full Video Transcript

Sam Chang: Hello, everyone. I am fortunate to have Carissa Chu, today with me. She is actually an urologic oncology fellow at Memorial Sloan Kettering Cancer Center. We're going to focus on the abstract that she's presenting at this year's GU ASCO 2022 meeting. It's actually an evaluation of some genetic changes and alterations looking at urothelial carcinoma. So I'm going to have her actually give a short synopsis of what her abstract is about, and then I'm going to ask her some probing questions. So Dr. Chu, Carissa, thank you again for being here with us.

Carissa Chu: Thank you so much for having me, Dr. Chang.

Sam Chang: So tell us some key highlights of your abstract.

Carissa Chu: So this was an initial foray into our institutional database of over 1200 patients who'd undergone targeted exome sequencing, focusing specifically on two hot drug targets right now. So FGFR3 mutated tumors, as well as HER2 altered tumors as well. We drew those tumors out, altogether there were about 360 of them. We just did a little bit of characterization looking at the genomic profile of what co-alteration patterns occurred with these mutations, as well as just some general clinical outcome data on them, how they presented, the stage and grade of their tumor, and then things like recurrence free survival, long-term survival outcomes.

Sam Chang: What were the key findings that you found between those two modification? Were there differences, very similar? Did they coexist, co-align? What did you guys find?

Carissa Chu: I think the main key finding there was that the FGFR3 altered tumors were mutually exclusive from the HER2 altered, suggesting that there are essentially divergent molecular pathways that may give rise to these tumors.

Sam Chang: So that's the first I've heard, that there really is kind of a disparate or a separated pathway in terms of mechanism. Was there any alignment with those modifications or changes, any alignment that you also saw in terms of worst disease, stage, grade, et cetera, or were there not enough samples that you could actually tease that out?

Carissa Chu: Yeah, I think there are some interesting findings there. Some of this has actually been described in the non muscle-invasive space with a prior study from our group, essentially showing that there seems to be very little overlap between the FGFR3 altered and the HER2 altered tumors. We did find, similar to another abstract that's presented at this meeting, that HER2 altered tumors tend to be more high grade, and to be higher stage, so T2 or higher, and present with metastatic disease. That being said, I think the caveat there is that these are both kind of very selected cohorts, depending on when the tumor was sent, at what time point of the patient's clinical course it was sent. So I would take that data with a grain of salt. But I think it is generally accepted that HER2 altered tumors tend to be associated with high grade, non-muscle invasive disease, more locally advanced disease, and that's why it's such an important drug target for us going forward.

Sam Chang: So looking at that HER2, did you see any correlation with some of the molecular subtypes that have been described? More luminal, more basal. Were there some overlap in terms of those types of classifications?

Carissa Chu: We didn't overlay the molecular consensus, molecular subtype, because the actual gene expression data that we have to work with currently for the entire cohort is limited, but we did find that there were co-alterations with other important mutations that we know, or that are driving, some of the oncogenesis of these tumors. So p53, RB1 alterations, ARID1A is associated with HER2. On the flip side, FGFR3 alterations are associated with [inaudible] and N2A, N2B, and stag two.

Sam Chang: So in looking at kind of individualized and personalized medicine forward, do you think that we're going to be really, I mean, obviously there are studies going on now in terms of predictive biomarkers and it's role in terms of choosing in terms of therapies and how eventually people do. How important do you think FGFR3 and HER2 are going to play. If you go looking at kind of your future in glass five years, 10 years, are we going to be trying to classify all tumors with these types of modifications? Or do you think it's only going to be the ones that are higher risk or metastatic to do areas in terms of drug targeting? What do you think?

Carissa Chu: I do think there's an explosion of interest in biomarker are right now and a lot of them are built as correlative studies and ongoing clinical trials, whether you're looking at BCG unresponsive disease, all the way out to your heavily pretreated metastatic disease. I do think that there is a role for, for a biomarkers very soon in helping clinicians decide which treatment am I going to pull for in which disease setting. Once it helps, it needs to be a predictive biomarker that can actually help clinicians stratify patients to the appropriate treatment. I think that's where the key is and I think that's a big challenge for our field. I think those studies are born out of what we can see right now in terms of molecular heterogeneity of these different tumors.

So for example, I think one very small subset of patients in our cohort were co altered, FGFR3 and HER2. It was like 16 patients out of the 367 and a very small minority of patients, but I think if we dig a little deeper into those tumors and we see are these the same cells that are co-altered or are there sub-clonal populations and if you were to treat one with the FGFR inhibitor, if you would then select for the tumor to be then just predominantly HER2 altered, can you go one and then the other? Do you do both at the same time? I think there's a lot of potential interesting translational science there.

Sam Chang: So that leads me to my last two questions. One is there's always been a belief in talking to pathologists as they look at variations within the same tumor, the heterogeneity that you expressed. This is just on histopathologic evaluation. How often does a single tumor have strong expression of one of these modifications versus very little. So sampling wise, how big a role does that play?

Carissa Chu: I think it's the key question, is sort of what part of the tumor did you decide to sequence? I think if you look at, for example, IHC stain slides, you can see, particularly for example, variant histology, right. It's rarely just pure one histology that lives on that slide. Usually, it's a mix it's sort of a mosaic and you can see, for example, that there's islands of, for example, areas enriched in HER2 expression in micropapillary variants.

Sam Chang: Okay.

Carissa Chu: I think it is kind of more mixed than, it's a little more complicated than we think, not so simple. I think looking in particular at the expression of these different targets in variant histology, for example, is a really important way forward because those are our patients that sometimes we don't really necessarily know what to do with in either the neoadjuvant space, adjuvant. If you can say, for example, micropapillary tumors are high in HER2 expression, then one of these trastuzumab ADCs coming down the pipeline may be the perfect treatment for that patient.

Sam Chang: Be the way to go. So along those lines, because that difficulty of trying to determine what's actually going on likely, it leads me to my last question is how easy is it to consistently do the test for HER2 change or an FGFR3 because you mentioned the IHC, the immunohistochemistry and how that can play a role in what shows up, not only from the sampling standpoint, but say you have the same specimen and you do it in, and there's some studies out there, and you do it at different labs.

Carissa Chu: Yeah.

Sam Chang: And read by different pathologists. Are they understood enough that there's some consensus regarding, Hey, this is a HER2 change, this is a FGFR3 change or is there some variability like we have with PD1 and PDL1?

Carissa Chu: Yeah.

Sam Chang: What do you think of the current state is in terms of that's pretty reliably, that's a change. So what do you think about that?

Carissa Chu: I feel like that's almost a philosophical question about biomarker driven treatment in general is what assay are you using?

Sam Chang: I see.

Carissa Chu: And what's the consensus for your cutoff? I think her IHC is something that the pathologists that we work with are very comfortable doing as if the first starting point is nice to know, is this tumor enriched in the actual protein expression of HER2 An FGFR3 alteration is something that you can check for and if you have access to a foundation or some of those next generation sequencing, but that may not necessarily be available in every institution. So I think that's a potential limiting factor there.

Sam Chang: Yeah. In thinking about it, five, maybe not five years ago, but 10 years ago, we'd never be having this conversation or there'd be thoughts about it, but now with just as you said, the idea of having a target, but then actually having a drug that actually can affect that target, incredibly exciting as move on. So the work that you've done and the folks at Sloan Kettering have you've done have really kind of enriched our understanding what's going on, on a molecular basis. I look forward to kind of your future work. Where are you guys going to go next in terms of the next studies?

Carissa Chu: Yeah. I think I want to delve a little deeper into these co-altered patients actually, because I think if we can show that they're potentially sub-clonal populations that are sort of lying in wait as treatment comes and gives rise to the evolution of the tumor to drive treatment resistance, for example. I think that'll be very, very interesting to look at those qualtric tumors to see how resistance is being built in these tumors.

Sam Chang: Yeah. I think that, to me, there's always two exciting populations. Those that were over treating that we need to back down and those that we're under treating and better understanding those that we are under treating, perhaps not treating in a sequential way, that would work the best.

Carissa Chu: Yeah.

Sam Chang: If we can figure that out, then that really would find out the bad actors more quickly and be able to treat them appropriately. So well, thank you Dr. Chu.

Carissa Chu: Thank you very much.

Sam Chang: I look forward to your very bright and promising future.

Carissa Chu: Thank you so much.