We’re trying to use the information that we gained from sequencing primary pancreas cancer and our idea was that we could use that sequence data to inform the treatment for those patients if and when they recurred later.

Unfortunately that didn’t really work because getting the trial up and running took such a long time that most of those patients were, in fact, deceased by the time we opened the study.

So we had to modify our study.

Now, looking at the genome, what sorts of things could you have found or did you find?

We found and chose three targets which you could say were the low-hanging fruit.

We chose HER2 amplified pancreas cancer, which in our data represents about 2% of pancreas cancer overall.

We chose KRAS wild-type pancreas cancer because we know that about 93% of pancreas cancer harbours a mutation in the KRAS gene so we thought that would be another subset to go for, using an EGFR inhibitor in combination with gemcitabine.

Then the third group we chose were those cancers that had DNA damage repair pathway issues like BRCA1 and 2 mutations or mutations in the PALB2 or ATM genes.

So the idea would be to identify potential targets?

Exactly, yes.

And what might you have been able to do if you had identified the right targets at the right time?

We were hoping that we would have been able to institute targeted treatment for those patients who developed metastatic disease.

For example, if they had HER2 amplified disease to give them gemcitabine in combination with trastuzumab which would be a novel treatment for pancreas cancer.

So what went wrong?

What went wrong was in that design unfortunately we weren’t able to get all of the results by the time the trial opened.

It took such a long time to open the trial that those patients who had had primary pancreatectomies with curative intent who then recurred were mostly already deceased by the time the study started and opened.

So what we’ve done now is we’ve changed the study and we’ve made it applicable to people who present de novo with metastatic pancreas cancer.

The issue we now face is that people who present with metastatic pancreas cancer have only a few months to live and we have to act very quickly.

So we’re actually taking tissue samples in real time from people who present with metastatic disease and doing tissue analysis for these three targets on them and then trying to inform their treatment straight away.

Could you summarise for me, then, in terms of day to day clinical practice what the possibilities might be?

We’re hoping that if we can do this fast enough that we can give people that targeted treatment as I’ve outlined – platinum based therapy for the DNA damage repair pathway gene issues; erlotinib with gemcitabine for KRAS wild-type disease and trastuzumab with gemcitabine for HER2 amplified disease.

But to date we haven’t got a single patient on the study because of all the issues that have come about with trying to get this tissue analysis done quick enough.

We’ve modified the study now so that at least patients can begin standard therapy while they’re waiting for us to do the analysis and that has been a really useful amendment.

Now they can begin… well we had gemcitabine as standard therapy a little while back, now that’s changed to gemcitabine with nab-paclitaxel.

So you’re hoping to generate more informed standard therapies, depending on individualising the treatment based on the molecular results or the genome results.

That’s right.

How much of a difference could this make? Because you’re talking about metastatic disease, a very aggressive disease at that, so what are your hopes?

We’re hoping it will make a big difference to a small number of people.

We acknowledge that even with all of our targets we’re still only targeting 10-15% of the pancreas cancer population with our study as it currently stands but we’re hoping that we can make big differences to those people.

If you establish the principle, presumably you could take it back in time and apply it in an earlier setting of the disease?

Absolutely.

You’ve hit the nail on the head. If we can show that in principle doing this, that is dividing pancreas cancer into a series of molecular subsets, we might actually be able to start treating it better.

But how does the mutation of tumour cells, the evolution of tumour cells, affect this and perhaps throw a spanner in the works, make it difficult?

Yes, it does make it difficult but we haven’t even got to the first base, as it were, with pancreas cancer and I think we’re a little behind other cancers.

We haven’t even tried this first step yet.

But you’re right, there’s probably heterogeneity, both within the tumour and between different metastatic tumours and resistance will probably develop, yes.

So what should ordinary cancer doctors think about this sort of highly sophisticated development right now?

I’m hoping that people who treat pancreas cancer will be looking for clinical trials to put their patients on because, even with the best standard treatment, and now chemotherapy has improved, the best we can do for someone with this disease is eleven months.

So hopefully everyone’s looking for a trial that does something a bit different.