We have a hypothesis-driven project that involves a collaboration between classic clinical geneticists and genetic counsellors, people such as myself who tend to study cancer signalling systems and some of our colleagues, also at Fox Chase Cancer Center, who are molecular modellers for protein structure.

The reason why we thought it would be important to bring this group together is that the whole area of cancer risk factors there have been some spectacular successes that have developed over the last number of years.

BRCA is the lead example that everyone always brings out where if you have certain classes of mutations in BRCA1 you know that you’re going to be at very high risk for breast.

And then that has gradually become a more complicated story because first it was a risk factor for breast, then it’s a risk factor for ovarian, now it’s a risk factor for prostate cancer as well.

In contrast to breast cancer there are very, very few genes that have been identified that are clear predictive markers for prostate cancer risk.

So what we have done is we have taken the hypothesis that many genes that are risk factors for all types of cancer, prostate cancer included, that have been identified are genes that impair DNA damage response or genes that affect androgen regulation, which are two very strongly linked elements to prostate cancer.

Then we’ve brought in the concept from systems biology and from precision medicine in treatment of sporadic tumours.

The concept that we’ve brought in is that genes that are affected are dispersed through a signalling network.

So it’s not just the primary gene but it’s all the partners with that gene.

So what we have done is we have gone data mining and we have identified a cluster of genes that all interact together in the areas of DNA damage response or androgen regulation and we’ve used this as a query set and then we have exome sequenced patients that have high risk families for prostate cancer.

So, just looking at the germline DNA of these patients, what we have identified is mutations that are very strongly predicted to be damaging with each patient carrying at least three or four or five.

So really large numbers of them, which is more than one would expect in random; we have been benchmarking this to the random population and it seems to be higher frequencies.

It seems that you’ve investigated and found some fascinating things. How might that help you either prevent prostate cancer or treat it better once it’s discovered?

Imagine the situation: somebody comes into the clinic and they want to know if they personally are at high risk for prostate cancer or whether their children are at high risk for prostate cancer.

So they do a basic panel test and none of the normal suspects are there.

What do you tell this person?

You can’t do a massive analysis of their family, they have a small family.

What this would allow you to do is say, ‘OK, these are classes of genes that are broadly linked with impaired DNA damage response.

You are more susceptible to pick up mutations from the environment and you have five separate mutations that have each separately the likelihood of contributing risk.

So we think that based on all of these together you are probably somebody who should be more actively surveilled.’ It might be PSA, it might be something else, it also offers some possibilities.

So if you have germline mutations in your DNA damage response, this is theoretically something where in the future you might be able to develop a test in your lymphocytes – just do a blood draw and treat your blood with DNA damaging agents and see do they respond differently from a normal benchmark population of blood.

In fact, preliminary work that we have ongoing is suggesting that there are differences that are detectable.

This is far from being a clinical test I would recommend at the moment but it is showing some very interesting patterns of difference in people at risk versus not at risk for cancer.

The clear view of the field for prostate and other hereditary cancers has moved from there are a small number of genes that are very high risk to the idea that there is quite a number of genes that have modest risk and that we now have enough information from exome sequencing, information about genes from many different sources, that we would be able to imagine that each individual family may be at risk because they have five different genes, none of which have a very strong risk, but they’re all linked in controlling DNA damage response.

One of the big issues in prostate cancer, though, is not so much whether you’re going to have prostate cancer but whether it’s going to be aggressive. Do you think you’ve got a handle on that important question yet?

I think that it would be premature to say whether we can make any claims about aggressiveness in the prostate cancer.

What we may have is information that can inform the use of DNA damaging drugs in those patients.

Because one of the large stories that is out there is that if you have, for instance, mutations in BRCA1 you respond better to specific classes of drugs like PARP inhibitors because you’ve already got a weakened DNA damage response.

So if you are tracking patients and you are profiling an underlying set of mutations that behave like a BRCA1 mutation that might similarly suggest that someone will respond better to those classes of drugs.

That is a hypothesis to be tested in the future and it has, of course, the risk it may be that if every cell in your body has these underlying mutations it may mean that you’re at risk of DNA damaging drugs and you need a lower dose.

But it does suggest that some people have specifically different vulnerabilities that are possible in the system.