Scott Tykodi, MD, PhD, Discusses Biomarker Detection, Analysis in RCC

Transcription:

Our center participated in the 427 study. And there were 2 abstracts relating to the data set. One of them was looking at biomarker detection and analysis. So that performed a couple of different types of analysis. There's an RNA sequence assessment, and then clustering genes by coherent gene pathways that have been described, looking at 11 different gene pathways and looking for an association with overall response or progression-free survival. And there's only a single clear association statistically in the data set. And that was a T-cell inflammation, T-cell effector pathway signal that showed a higher overall response rate in the pathway positive patients, so that biologically makes sense that tumors that have prestanding to inflammation would appear to be more likely to respond to PD-1 blockade. And that sort of association has been made in other data sets with atezolizumab (Tecentriq) with avelumab (Bavencio), although those are in combination with the IGF inhibitor.

Of course, PDL-1 is an on target, biological marker that's been carried through the clinical

development of these drugs. And although it's not a companion diagnostic for renal cell, it does enrich for overall response rate as well. So it's pretty clear that the PDL-1-positive tumors have a higher chance of response, but it doesn't seem to impact the overall survival gain of the drugs. Both PDL-1-positive and -negative tumors beat the competitor, whether it was the second-line study with everolimus or the frontline data with sunitinib (Sutent).

So it doesn't split the population and change your therapy recommendation, but it does modulate the response to these drugs. More complex gene signatures also seem to be able to do the same thing and modulate the response outcome. But again, not to the degree that the signature-negative population fare so poorly that you'd like me to not offer them checkpoint blockade. So I think it's providing some level of insight about what's happening. I think it's not a surprising outcome. It's what people would have expected. But it's not rising to a robust marker that that is going to change your therapy plan, that marker-negative patients would be offered some other type of treatment. So I think that's true across the board for similar marker studies in some of the larger clinical data sets. There really aren't strong signals that are emerging that split the population in such a dramatic fashion that would actually impact your clinical management for the patients.