The presentation really revolves around an update on data from the KEYNOTE-001 clinical trial. So this was the original trial looking at pembrolizumab in cancer, looked at multiple histologies, mainly melanoma and non-small cell lung cancer. What we’re updating is the non-small cell lung cancer cohort. There were 550 patients who were treated as part of that cohort, 101 of them were naïve to treatment for advanced non-small cell lung cancer, 449 of them had been previously treated for advanced non-small cell lung cancer. All patients received pembrolizumab and the results have been previously reported in The New England Journal of Medicine 2015 that there was about a 20% response rate in previously treated patients, a little bit higher in treatment naïve patients, and that the response, as well as progression free survival and overall survival, correlated with PD-L1 expression, particularly with better outcomes in patients who had PD-L1 in at least half of their cells.
What we update now is looking at long-term outcomes using what has typically been a benchmark in cancer research, of course, which is five year overall survival. The follow-up at this point is over five years median and what we see is that over 15% of the patients who participated in the study remain alive more than five years out. In particular, the patients who had high staining for PD-L1 that over a quarter of those patients remained alive at five years although patients with 1-49% staining still did have a real potential for response somewhere in the mid-teens in terms of their likelihood of survival at five years.
We also updated the toxicity, we have followed toxicity throughout on the study to ensure that we are not having problems with late, particularly immune-related, toxicities. Really we saw very little additional toxicity at our analysis at five years compared to the analysis that we conducted at three years.
How will those data change or impact clinical practice?
In some ways this data is data that clinicians are aware of, that drugs like pembrolizumab have really changed the management of patients with non-small cell lung cancer. But in many respects, as the discussant at the press release we just had indicated, this does change our whole outlook on the management of non-small cell lung cancer. We typically would go in to see a patient in the hospital with a new diagnosis of metastatic non-small cell lung cancer and we would give them a very uniformly poor likelihood of doing well, that it may be several months in terms of life expectancy and almost no likelihood of being alive five years out. We do now see that we have patients who are not only alive but thriving many years after initiating pembrolizumab for advanced non-small cell lung cancer. That, of course, is very encouraging but does change our approach throughout the treatment paradigm for patients.
Is there a potential to incorporate biomarkers in identifying the most effective patient population?
PD-L1 is one biomarker, of course, that is clinically used and actually predicted reasonably well in the previously treated cohort. Unfortunately the five year survival among PD-L1 negative patients was very similar to what is seen in historical controls whereas there was much higher likelihood of five year survival among patients who had high PD-L1 expression. I believe that over time we will be able to further refine biomarkers, have potentially biomarkers that would select perhaps along with PD-L1 so that we would be able to have a group that we could predict with even greater likelihood than 25% would be likely to be alive at five years.
What are the next steps for this research?
In many respects this is probably the end of our reporting, I would assume, on the KEYNOTE-001 study. It’s hard for me, this is a study that we invested quite tremendously, we treated nearly 100 patients at UCLA on this study. I continue to follow up on them, they are now, though, instead of being on this study there’s a rollover protocol. So this is probably the last time that we will be reporting data on this cohort. I think that there clearly have been advances in the field; adding chemotherapy in the patients with low or negative PD-L1 expression has looked very promising, whether we will be able to recapitulate these impressive five year results is something that we still do not know but it’s something that of course we will continue to assess.