Predictive Tools in Renal Cell Carcinoma

Transcript:Nizar M. Tannir, MD, FACP: So I think one thing also that we are doing more and more that maybe has not spread to the community yet to do it routinely is doing a biopsy and genomic analysis for sequencing of the tumors. But I think there is a cohort of patients that would be best treated with an mTOR inhibitor and that probably is like somewhere around 5% to 10% of those patients who have mutations in the mTOR gene, genes TSC1, TSC2, PTEN, PI3 kinase, etcetera. In those patients with tumors that harbor these mutations, if I have access to do genomic analysis on their tumor and patient agrees to have a biopsy right at the time of progression, that biopsy, that information that you gain from the sequencing would be very informative and will help in the selection of your subsequent therapy.

Daniel Heng, MD, MPH, FRCPC: Unfortunately, it’s not an all or none phenomenon though.

Nizar M. Tannir, MD, FACP: Sure, sure. But if they do have them it suggests that an mTOR inhibitor would be what I would do next.

Carlos H. Barrios, MD: That’s important, though, because the two trials that use the sequencing of sunitinib and . . .

Nizar M. Tannir, MD, FACP: RECORD-3.

Daniel Heng, MD, MPH, FRCPC: The RECORD-3 trials.

Carlos H. Barrios, MD: Yeah, the RECORD-3, essentially they used some of that information to select some of the patients that responded very well to everolimus. The group from Dana Farber did that, suggesting that these mutations you mentioned may actually be more prevalent in the patients with a very good response. But the results were in retrospective analysis, a small number of patients. We need to confirm and validate these data. And also the Memorial Sloan-Kettering group also showed, using a similar set of patients, some genes that are associated with better benefit for sunitinib and some that may be better, associated with more benefit from everolimus. That is important information that we need to address. And the issue of the biopsy at the time of progression is another important thing because for most of the clinicians it is not in our regular clinical behavior. So I think that tumor DNA, circulating DNA is probably one of the things that is being explored in other diseases and that should be explored in this situation in order for us to try to discover biomarkers more appropriately for resistance mechanisms.

Paul Nathan, MBBS, PhD, FRCP: The bottom line at the moment though is that we don’t have useful predictive tools to allow us to make these decisions.

Carlos H. Barrios, MD: Definitely.

Paul Nathan, MBBS, PhD, FRCP: And I think when we’re in a situation where we haven’t got tools that enable us to select a first-line treatment, asking, expecting, hoping to have tools that will allow us to make second-line decisions when one will inevitably have multiple resistance pathways is a really big ask and we’re a long way from doing that at the moment. The other point that I just want to make is that given the overall survival readout from the sequencing studies, I tend, more often than not, to go TKIïƒ TKI rather than going TKIïƒ mTOR in the absence of everything else, given the overall survival signals that we’ve got. And finally, you mentioned an interesting question about treatment withdrawal.

A colleague of mine, Janet Brown in Leeds, is running a large 1000-patient study looking at treatment withdrawal; so standard of care whether it’s with pazopanib or sunitinib versus treatment until nadir, and then you withdraw therapy, wait until progression, and re-challenge. You keep going around that loop. And it was powered for a health economic endpoint because there are obviously health economic implications, but also quality of life. There are major implications for patients. And certainly we’ve had some patients who I was expecting to progress once their disease had shrunk and they’ve withdrawn from treatment— and they haven’t. So if we can more wisely use the tools that we have, it may be that we improve patient outcomes and patient experience by doing that.

Carlos H. Barrios, MD: Right, we need to be creative in order to get around this kind of thing. But in clinical trials there are two things that I think, uniformly, we have not been able to collect appropriately and that have a significant impact on the results. One is material at the time of progression, to find out and address all this complexity and heterogeneity in order to inform our future research. And the second, which is very important in some of the trials, is data on postprogression therapy. We need to recognize that patients after progression are receiving more and more lines of therapy and that has an impact in survival. And if we do not collect that information, we’re not going to be able to appropriately analyze the differences that we have been seeing.

Daniel Heng, MD, MPH, FRCPC: So I think those are key messages. We don’t have any valuable biomarkers that are reproducible to help us choose the therapies today, but we do need a lot of research to find those biomarkers. It could be that these drugs, they’re all antiangiogenic agents, and so maybe it’s not very helpful. But now with newer drugs coming out, maybe there are significant substantial differences in mechanism of action where a biomarker might actually be helpful.

Transcript Edited for Clarity