Dr Peter Black

Shomik Sengupta: Peter, that was a nice talk this morning about the subtyping of bladder cancer. You have started off by saying that bladder cancer was the sort of poor cousin of urological cancers. Why do you think that is and do you think that’s about to change?

Peter Black: Yes, it’s a good question. I think we always compare to kidney and prostate cancer. In prostate cancer we’ve had the androgen receptor and kidney, we’ve had remarkable molecular understanding with the tyrosine kinase inhibitors and VEGF signaling. In bladder cancer, we just haven’t had anything to pin our progress on. But I think this molecular subtyping is something that really helps. As I said in my talk, it really helps give us a framework to study disease a little bit a better.

Shomik: Yes. You showed us a number of different groups that have, I guess, subdivided into kind of different categories what we understand to be bladder cancer and there are some overlaps and there are some differences. How do we pick which one?

Peter: Well, that’s a big question. Fur different groups internationally used especially RNA expression to define the subtypes and there were commonalities obvious right from the beginning and then in the interim, a lot of work has been done to sort of sort out how to explain the differences. I think we can explain a lot of the differences and when you can explain them, then you can see even better how to align the different and essentially a lot of it is just how detailed you want to go, how many levels of classification do you want to have.

Shomik: So of the different subtypes schemes there are different categorizations, so as there is some synthesis and all of that. In your mind, what are the kind of key subtypes that are emerging that might be in clinical use?

Peter: I think at the highest level, I think we can agree that we can divide muscle invasive bladder cancer into luminal and basal. Those are groupings that are established and also in other cancers like breast cancer and then the big question is within basal, should we subdivide further into— there two different types of basal and for luminar there are potentially two or three different types of luminal, I personally and some people in the field believe that we should divide into four so we would have the usual luminals  but then also a luminal subtype that has a lot of immune infiltration and that might be particularly relevant for immunotherapy. Same thing for basal, we will have a usual basal which we think responds really well to neoadjuvant chemotherapy, but then an immune infiltrated basal and that doesn’t respond as well. That’s one model, but there are other models out there.

Shomik: And related to that, do you think there will be a continuing evolution of the classification? I mean, the parallels with, say, grading in prostate cancer what you do today compared to what was done in the past can sometimes lead to a little bit of evolution of how the prognostication perhaps works and things like that.

Peter: I think for sure. We’ve seen just at ASCO this year the TCGA The Cancer Genome Atlas, they have modified their classification based on just more data. What’s also going to change is if we do a better job of into of integrating the DNA level data, some mutations and copy number changes and then also the proteomics data, then I’m sure we’ll fill in some of the gaps so. I think it is a field in evolution and it will change in the future.

Shomik: I guess this from clinician’s perspective, the million-dollar question is, is this ready yet to try and classify patients as their utility to how we manage patients?

Peter: I don’t think there is, right now, in July 2017 but I think there will be with additional validation. I mean, the most obvious thing I see is is we know that DNA repair gene alterations and ERCC2 mutations looks like they predict response to Cisplatin-based chemotherapy and then with the subtypes it looks like the basal subtype really does well with neoadjuvant chemo. If we could put that together, so mutations and subtypes, we would probably be able to pick out a population that’s going to do well with Cisplatin-based chemo and then we can focus in on doing something else for the other subtypes. It could help development of new treatments.

Shomik: Yes, I mean, we’ve always grappled with the difficulties of enrolling patients into neoadjuvant chemotherapy. I suppose if there was a more strong case to be made into subgroup, then that might be one way that might work and obviously immunotherapies and other arena where there’s sort of a subgroup that responds whether we can identify those. Are these areas where you do see the sort of immediate implications for this?

Peter: Absolutely. I mean, in the neoadjuvant setting, we worry about the patients who are going to progress and for them, cystectomy is truly delayed. And so, if we can pick out patients who are probably not going to respond and then move on and do surgery or do something else and really give chemo to those who have a higher probability of responding.

Shomik: Excellent. Well, we look forward to more updates on all of this and thank you very much for enlightening us this morning.

Peter: Thank you.