KEYNOTE-010 and KEYNOTE-042 were clinical trials that looked at pembrolizumab as a single agent, KEYNOTE-010 in the refractory setting, meaning second or more line therapy, and KEYNOTE-042 in the front line setting, both versus the appropriate chemotherapy. Both were positive trials showing a superiority for pembrolizumab, the PD-1 immunotherapy, over docetaxel or over platinum based chemotherapy.

What we’ve done here is now that the trials have matured we’ve looked at the data with tumour mutational burden as a potential predictive marker. We already know that tumour proportion score, PD-L1, could predict activity but now what we’ve done is we’ve looked at the number of mutations by doing whole exome sequencing. We looked at this in two ways – as a continuous variable of the number of mutations and also using a cut-off of 175 mutations per megabase which had come from some of our prior work in other datasets.

Actually the data were quite interesting. Granted, it’s an exploratory analysis that was pre-specified and the numbers are not the complete number of patients in the trial because there weren’t enough samples or tissue to do every patient. We found that TMB was a predictive marker, both in trial 010 and 042, both for overall survival, progression free survival and response rate. Also using the cut-off of 175, both in KEYNOTE-010 and KEYNOTE-042, we were able to show that the group of patients that had the higher number of mutations, regardless of PD-L1 status, seemed to have improved overall survival, progression free survival and response rate.

So what does this all mean? It means that this needs to be validated prospectively, no doubt. But it also means that as we try to personalise immunotherapy, and we must – immunotherapy is historic in how it’s helping patients with lung cancer, but it really only in the long-run helps 15-20% and we’ve got to figure out how to get more involved and do better. We’ll start to pick patients for single agent immunotherapy using PD-L1 and now perhaps we can add in tumour mutational burden. Then, of course, we need other markers regarding other resistance pathways that we’ll discover as they emerge.

But it’s a very exciting trial and one that I think will lead to other studies, both retrospective and prospective in the not too distant future.