I’m going to speak about two of the papers that were presented at the recent ASCO meeting in Chicago in gynaecological cancer. The first one was the results of the KEYNOTE-100 study which was a large single arm study using pembrolizumab in patients with recurrent ovarian cancer. There were two cohorts of patients differing in the interval from the previous line of treatment but essentially the treatment was the same – pembrolizumab given until progression.
Overall the results of the trial were perhaps a little disappointing. We’d hoped for a better response rate and a better progression free survival. But within the data there are some very interesting things that happened. For example, amongst the patients who had a response there were some patients that had quite a durable response and when we looked back at those individual patients they were really of a relatively poor prognostic group. So there are patients that clearly benefit from these immune checkpoint inhibitors, and here pembrolizumab, the difficulty is understanding who they are and trying to predict who might do well.

Now, there has been some data looking at PD-L1 expression and indeed that was done in this study both as a training set and then a validation set. A CPS score which has been developed by Merck MSD for a number of their other pembrolizumab studies was applied to this and there was a slight improvement in those patients who had a high CPS score but it’s small and there’s much more work being done. I hope that we’ll be able to present more data at the upcoming ESMO meeting, trying to identify which patients are going to respond. Because that’s the key, particularly in ovarian cancer but actually across the board, when you’ve got a drug such as pembrolizumab or any of the checkpoint inhibitors that’s really very expensive you really want to be able to target patients who are likely to respond and we have to have predictive markers of response. At the moment this is rather unclear in ovarian cancer.