BACR & ECMC: Therapeutic interventions for cancer prevention

Update on CaPP3 in high risk populations

Prof Sir John Burn – Institute of Genetic Medicine, Newcastle, UK

Could you please tell us about the CaPP studies?

It started just over 25 years ago when I began developing an interest in hereditary cancer and came across a family where a young boy had the gene on the basis of some physical features but hadn’t yet developed the polyps that would predispose him to cancer. It struck me that this was an ideal study population for interventions that might prevent cancer and particularly as we’ve learned more about first that condition called familial polyposis and then subsequently Lynch syndrome where there’s a breakdown in mismatch repair genes, basically these are model systems for the common forms of hereditary cancer, particularly bowel cancer. So if you lose an APC gene through inheritance it predisposes to polyp formation and cancer; if you could prevent that person developing cancer then you’re probably going to have a similar effect in people who develop sporadic cancer through that same mechanism.

In CaPP1 we focused on FAP, in CaPP2 our trial focused on Lynch syndrome and here you’ve got a breakdown in DNA repair systems. Again, that accounts for about one in six of sporadic cancers so if we can protect the Lynch syndrome population then we can potentially also have an impact on the general population with those more common cancers.

Could you please tell us more about Lynch syndrome?

Yes, so Lynch syndrome is actually a poor relation to the breast cancer genes - everyone’s heard of BRCA1 and BRCA2, there are a group of four key genes called the mismatch repair genes common throughout the animal and plant kingdom. Their job is to spellcheck your DNA as you copy it and if you get slippage, particularly bits of repetitive DNA. So if you’ve got, say, ten As in a row you might end up with nine or eleven and that completely wrecks the gene so there’s a very elaborate molecular system for spellchecking and correcting those slips. If you lose that system you start picking up spelling mistakes in key genes and obviously if you take out some of your tumour suppressor genes that can then lead to cancer.

So Lynch syndrome is effectively a familial form of cancer where you get from an early age, kicking off round about the age of 25 typically, colon cancer, endometrial cancer, ovarian, gastric, renal, a whole range of different tumours although the bowel and endometrium make up the bulk of the cancers.

What role does aspirin have in your study?

Back in the late ‘80s it was apparent that people who took non-steroidal anti-inflammatories seemed to get less cancer, particularly those who took aspirin. We were looking for something to try out in our hereditary cancer population. That was one of two interventions that we started out with, the other being resistant starch which has an impact on bowel development which it turned out not to work in our trial but it was probably a useful additional feature because it slowed us down and it meant that we ended up with a lot of long term follow-up data because it took us so long to do the trial. So we began designing CaPP1 in about 1990 and CaPP2 in 1993 because we were there at the beginning of the discovery of the mismatch repair genes. We actually got rolling with CaPP2, as it became known, in 1998 and finished recruiting in 2006. In fact this year the last recruits reached their tenth anniversary which is the planned end follow-up for the study.

We took a look at the data, or a look at the patient long term follow-up, when the first of them reached ten years and what we found was that the people who had taken two aspirins a day for a minimum of two years, as was the original plan, had a reduction in their cancer rates of about 50%. That was apparent not only in the colorectal cancers but in fact across all the cancers associated with the syndrome which came as a great surprise. Well, not entirely as a surprise because it fitted with the epidemiology but it was delightful to see a similar impact. What was surprising was we didn’t see a particularly big impact on polyp formation so at the end of the treatment phase we were a bit demoralised because we expected there to be a reduction in the number of polyps as well and that wasn’t apparent.

We now think what’s happening in Lynch syndrome is that you’re probably getting a progression down a slightly different pathway to that which we’ve become used to where people generally think of polyps forming, taking several years to eventually become a cancer. What probably happens in Lynch syndrome is you get loss of mismatch repair function in the crypts in the lining of the colon and then they accumulate mistakes in stem cells which then really pretty rapidly progress to become malignant cells and effectively emerge as cancers. They then take a long time to metastasise because one of the effects of this mechanism is you tend to lose key genes like B2M which are critical for your HLA expression. Now that’s quite useful in terms of local tissue expansion but if you try and metastasise you get picked up by the state police, the T killer cells, and they don’t like cells without HLA type so they’re taken out. So basically these cells become cancers then sit around for a long time as a cancer and it seems that the aspirin is having a fairly dramatic effect on that process.

But, again, the dramatic thing we’re seeing with the Lynch syndrome patients, as indeed is being seen in all the epidemiological studies of aspirin, is there’s some kind of delay from when you start taking the aspirin to the beneficial effect, about a 3-5 year gap. So what we think is probably happening is that the aspirin is having an impact on apoptotic mechanisms in terms of aberrant stem cells, that it’s helping you to take out the cells that will become cancer one day. It’s also probably having a direct effect on other mechanisms like new vessel formation and the direct immunological response to the tumour itself. So it’s a kind of dirty drug that’s hitting the system at multiple different levels. What’s really fascinating in a way is that we’ve found this effect and now we have to work out the explanation. It’s like the old Hitchhikers’ Guide to the Galaxy, the answer to life, the universe and everything is 42, now work out the question. So we know that aspirin prevents cancer but we still don’t actually know why and that’s really quite challenging because without knowing the mechanism we can’t think about interactions and so on with other drugs, we can’t think about whether we can change the dose.

So in CaPP3 we’re testing three different doses and that trial started two years ago almost and we’re up to 776 recruits of the minimum 1,500 that we need. What we’re doing in CaPP3 is randomising people to receive either the 600mg we used in CaPP2, 300mg, one standard aspirin a day, or 100mg of enteric coated which is the cardiovascular baby aspirin. The idea is to look at those three doses and look at enough people over a period of five years to see whether taking two aspirins a day gives you the same level of protection or a greater level of protection than taking a baby aspirin or the intermediate dose.

Clearly if we can demonstrate that the baby dose has as big an effect on cancer protection or prevention as the two aspirins a day we can move to the safer lower dose in terms of population treatment. So it will add to the case for general use of aspirin as a cancer preventive but particularly we can then go for a big push to get people with this hereditary predisposition, and other people with a high risk, to take aspirin as a routine standard protection.

At what point do you see aspirin being brought to a wider commission approval level?

I think if you look at the epidemiological and now clinical trial data, our CaPP2 trial and the Women’s Health Study which also used cancer as an endpoint and which, on review, showed a beneficial effect of low dose aspirin from ten years onwards. So we’ve got two randomised trials, one in a high risk population, one in a general population, using different doses, both of which show an effect. So, to be honest, the evidence is now dramatically stronger for using aspirin than it is for using colonoscopy which we all just accept is the right thing to do for people who are at increased risk of bowel cancer. So the answer is we should be using it, we need to repurpose aspirin. The problem is the legislation is such, has been such, that it’s very difficult to get a drug repurposed. It’s one thing to take it through for a licence for one application but if it’s already got a licence and it’s already off patent there’s no-one stimulated to actually get it a licence for another use. But we’re hoping to take advantage of the recent legislative changes in the UK for innovative drugs which will allow us to potentially get aspirin repurposed.

In the meantime we’ve already got a recommendation from our European Hereditary Tumour Group, of which I’m a member, that aspirin should be drawn to people’s attention with Lynch syndrome. Similarly the American Gastroenterological Association have now recommended that aspirin be offered to people with Lynch syndrome, accepting that the evidence is quite weak because we only have one randomised trial. But given the rarity of the condition that’s pretty good to have even one randomised trial. Actually it’s not that rare, we think probably about one in every thousand people will manifest Lynch syndrome so that means that in a population of our size, that’s 60-70,000 people, it’s not a trivial condition in its own right. So we should be pushing ahead with using aspirin now in that highest risk group. But one could take as a reasonable guide, given that the side effects of aspirin in a non-frail population are about the same as the side effects of colonoscopy, so in other words your risk of getting a serious bleed or a perforation or a death from a colonoscopy is small but real, is about the same as an ulcer or an intracranial bleed associated with aspirin. It’s about one in 10,000 people, which is not to be ignored but if you were at high risk of cancer it’s a risk you may well be prepared to take. And if someone is worthy of having colonoscopy regularly then there’s a very powerful argument for saying that the similar case could be made for giving them aspirin and certainly making them aware of the beneficial effects.

It’s still really quite hard to break through the old thing that it’s what we think we know already that prevents us from learning, as Claude Bernard the physiologist once said, so most doctors think they know everything about aspirin and so in fact studies have recently shown that a majority of them still aren’t recognising this highly protective effect of anything up to 50% reduction in bowel cancers. The other thing is that doctors generally over-estimate the side effects and that’s partly because we tend to see elderly people who’ve been left on aspirin forever presenting with life threatening bleeds in their 80s and 90s. That tends to skew our impression of the effects of aspirin which are otherwise quite mild in the vast majority of healthy people.

Any final thoughts?

The only one thing I would say is that if anyone is looking after patients with Lynch syndrome in the UK every genetic centre in Britain is now open for business and any adult with Lynch syndrome who doesn’t have a major illness or is already on anti-inflammatories is eligible to take part. So we need about a fifth of all the Lynch syndrome patients in the country that are known to us so we basically need as many as possible to be made aware of this. But given that there’s no placebo group and they’re learning something of relevance to their own future healthcare and that of their family, we’re finding a pretty strong response so far so I’m hoping that most will come forward.