At this meeting I’m going to be focussed on a fairly narrow application of immunotherapy. This is looking at a precision medicine enabled tissue agnostic approach to therapy. Biomarker enabled therapies have been around for a long time but they tend to have been tissue specific. So, for example, oestrogen receptor positive breast cancer, HER2 positive breast cancer, have been very specific precision mechanisms.

What we have shown, and this was worked on in collaboration with workers at Johns Hopkins in the United States, is that there’s a class of malignancy where there is a defect of DNA repair. This can occur as part of a syndrome that’s called [?? 0:48] Syndrome or it can be an acquired defect in cancer cells. For most of us our normal cells are being bombarded with insults all the time. These are either ultraviolet radiation, products of smoking and the like. What happens is when a cell is injured and the DNA is actually injured there’s a mechanism whereby that DNA gets repaired. If it’s not repaired we tend to accumulate mutations. It was argued on first principles that indeed if we had a circumstance where there was a very large number of mutations it was quite likely that those types of tumours would be quite immunogenic, the idea being that the more mutations you have the more likely you are to have mutated proteins which will lead to cancer-related antigens or neoantigens. And that’s exactly what happened. In fact, if you identify cancers which have a defect of DNA repair they are most likely to respond to immunotherapy and actually respond very well to monotherapy. Response rates have been incredibly high and they are very durable. We initially studied this in colorectal cancer but then it’s turned out to actually be generalizable across most cancers that have been studied, in fact all cancers that have been studied. On the basis of this the FDA recently approved Keytruda, or pembrolizumab, for use as monotherapy in circumstances where patients had malignancies which had this defective DNA repair. It’s identified by an assay called the MSI, or microsatellite instability assay. So essentially any cancer which has characteristics of being MSI high and that have failed their standard therapies actually are approved for use to be treated with Keytruda. So essentially this becomes a tissue agnostic molecularly defined indication.

Have you been working with combination therapies as well?

We’ve been working extensively with combinations but, indeed, the monotherapy rate in this particular circumstance is very high. If we actually look across the whole programme we have explored now over thirty different cancers, these are major cancers, and we’ve got evidence of efficacy now in more than 25 of those and that’s just as monotherapy. Using a series of biomarkers we are trying to identify those patients for whom monotherapy would be very sufficient. By inference, obviously, we’re identifying patients for whom more may be necessary and those patients we’re trying to put into clinical trials evaluating combinations. Combinations have been studied extensively. We have now more than 300 combination trials ongoing, the majority of those are the single detection nature but then we also have a number of registrational trials ongoing where we’ve already established a very good signal.

The combinations that we are exploring include combinations with standard therapies such as radiation and chemo. We’re looking at combinations with targeted therapies, combinations with other immunologic therapies and then combinations with vaccines or oncolytic viruses to try and improve antigenicity. To give you a few examples for each of these, standard of therapies we, for example, have done chemo combos with Keytruda in lung cancer. We explored a number of these, took one forward into a registrational trial. A small randomised phase II study called KEYNOTE-021, this was the G cohort. That showed remarkable improvement in progression free survival and at ASCO when we updated data there’s now the appearance early on of a divergence in the survival curves as well. That was the basis of a recent FDA approval for the chemo combo indication in front-line treatment in patients unselected by biomarkers.

We had some other rather nice data at ASCO looking at chemo combos. There was very nice investigator sponsored research looking at a chemo-radiation-Keytruda combo in head and neck cancer. This was in essentially locally advanced disease; it had a complete response rate in the order of 80% which is really quite remarkable. We had very nice data, for example, in breast cancer from the I-SPY study where Keytruda was added to standard of care, as is done in this platform trial, and to cut to the chase, basically in HER2 negative disease, in triple negative disease and in hormone receptor positive disease we saw almost a tripling in the pathologic complete response rate. PathCR has been used as a surrogate for outcomes in breast cancer, as is well recognised in the field.

Looking at targeted therapies we have shown very nice combinations with TKIs, or tyrosine kinase inhibitors, in the settings of diseases such as renal cancer and then most recently at ASCO we showed very nice data with that combination in endometrial cancer. Based on this we’ve initiated phase III trials. So, for example, in renal cancer we’re exploring Keytruda plus axitinib and then we have also now started to explore the combination of Eisai’s drug lenvatinib in combination with Keytruda. The endometrial data that were so exciting were lenvatinib plus Keytruda in endometrial.

Then when we look at the combinations of other immunologic drugs obviously the IOIO has been touted as an important platform. The real key here is to get to a combination which is effective but also has acceptable toxicity. Some of the early IOIO data where, for example, CTLA-4 was put together with a PD-1 antibody resulted in quite high rates of toxicity and seemed to have some enhanced response rates although it’s not clear yet how durable or important that’s going to be in terms of long-term outcomes.

We have been exploring, for example, a microenvironment inhibitory mechanism, the IDO1 mechanism. We’ve shown very nice initial data in melanoma, we took that forward in phase III and we now have phase III trials which are getting ramped up in lung, head and neck, bladder and renal cancers based upon encouraging signals seen in phase II. Then in terms of vaccines and oncolytic viruses furthest along here, a nice example, would be our combination with the Amgen molecule TVEC where we saw very nice response rates in melanoma. Based on that we’ve taken forward a combination phase III study. We’re also exploring other oncolytic viruses and vaccines.