AACR 102nd Annual Meeting, 2—6 April 2011, Orlando, Florida

Heavy beer drinking and a gene variant increases gastric cancer risk

Dr Eric Duell – Catalan Institute of Oncology, Barcelona, Spain

ecancer.tv now is looking at the causes of cancer. Eric Duell, you are from Barcelona, you have just been talking about beer and causing gastric cancer. The whole nutritional involvement with human cancers is fascinating and large and also vexing, because although we are sure that nutritional factors of one kind and another are involved in causing cancer, it is really hard to pin them down. What have you done with the EPIC study?

In EPIC we have basically conducted a perspective cohort study, so essentially we’ve measured dietary factors, nutritional factors, including alcohol consumption; about 8 or 9 years before diagnosis of cancer. And of course as time goes on that time window will increase but in terms of the analysis that we are talking about here, we assessed alcohol and nutritional factors which we adjusted for our analysis about 8.9 years before diagnosis of cancer. And the advantage of that is it is essentially a prospective study so you don’t have, when we ask people questions, they had not yet developed disease.

So it’s not case controlled?

Actually it is not a case controlled study.

The ideal case study to have?

We avoid this bias; some of these biases with case controlled studies they can exist.

Give me some idea of the scope of the EPIC study; how big is it and how many subjects are there?

So EPIC was initiated in the early ‘90s and from ’92 to ’98 they enrolled over 500,000 men and women in 10 European countries, so there are 23 research centres. And Spain, of course, is one of the countries and our particular institute is sort of the coordinating centre for EPIC Spain.

So epidemiological surveys have shown a link between some alcoholic drinks and cancer but now, thanks to this prospective study, you’ve got hard data. What are the data that have come out?

Basically what we have found is we looked at total alcohol or grams per day of pure ethanol from different types of alcoholic beverages and we basically had a variable of grams per day of drinking. So we looked at total alcohol from all types of beverages and then we looked at beverage specific variables as well, so we looked at beer consumption, we looked at wine consumption and we looked at liquor consumption.

And I know beer came to the top didn’t it?

Right; and we did not see any associations between wine drinking and gastric cancer or liquor consumption. We did see an association with very heavy beer drinking in gastric cancer risk, and those associations were adjusted for the other types of alcohol. So we were able to actually, in a sense, isolate or say that the association seems to be stronger in the beer drinkers.

Right, now heavy beer drinkers have quite a few litres of beer per day?

Right, and so we basically have in the top category, we looked at different categories of drinking and the top category was 30 grams or more per day and that is about equivalent to three or more drinks of beer per day.

So what some people would not regard as particularly heavy, even normal in places in Great Britain.

No, and as I mentioned before, there are quite a few issues with measuring alcohol consumption in any epidemiologic study and there is always going to be some amount of measurement error. And we need to keep in mind that this group of heavy drinkers, the lower end is about three drinks, it is three beers per day and of course then you have a lot of people falling into the higher levels. So I don’t think we necessarily can focus on, say, it’s right at three that we see this increase risk; it’s sort of a group of people and three is the lower end, and then we have much higher drinkers. So it is a bit difficult though to separate it out and say exactly where and when increased risk occurs.

I seem to have pulled out of your figures that in chronic heavy beer drinkers you’ve got something like a 70% increase of risk but in chronic heavy beer drinkers who have a gene variant it’s 700% increase.

Right.

Tell me about that.

Right, so for this particular set of analyses we looked at sixteen genetic variants across the ADH1 low side, so ADH1 Class 1 includes ADH1A, B and C. So we looked at sixteen different SNPs across this region and we found one in particular that was associated with about a 30% increase in gastric cancer risk just by itself, ignoring alcohol drinking. But when we combined these data and when we looked at people who had the genetic variant and who were heavy drinkers of beer, we found this over 700 increased risk, 700 fold increased risk which is ...

That sounds like a serious situation.

It’s an interaction.

A call for action.

Right, it’s what we call a gene environment interaction. So you have the genetic component by itself; you have the environmental component with the alcohol drinking. When you put them together it appears that the risk multiplies even greater than what you would expect from just putting the two together.

What does this imply about action that needs to be taken to prevent gastric cancer?

Well I think we first need to state that this is the first study to find an association between this particular SNP (single nucleotide polymorphism) in gastric cancer risk. So I think the first thing we need to say is we need other studies; other populations need to look at this variant and replicate this finding. And in terms of the message taking back to heavy alcohol drinking or heavy beer drinking, I think there are other health reasons why it might be a good reason to try to decrease our alcohol consumption. But back to the genetic effect and the interaction, I think the first thing to do is to call upon other studies, other cohort studies that are out there, even case controlled studies, to look at this variant and see if they see in it the same association or similar association that we see in our study.

So you could go public, then?

Then I think we could maybe start talking about what are the implications of this. In our study we intend to look at other variants that are near this variance; as I said, we only looked at sixteen. We are covering about 45% of the variation in this particular genetic region. So our responsibility is to go back and look at other SNPs and try to evaluate more other  SNP variants that may be close to this variant. So we need to really pinpoint exactly where the variant is that may be causal in this pathway, because in fact this variant is between, is not in the gene, it is outside of the ADH1 gene. But this is something that has been discussed at this meeting, in fact, that a lot of genetic variants that are being found to be associated with cancer are not in genes but outside of genes and maybe somehow related to the expression of the gene.

Scientifically this is all absolutely fascinating. In the practical day-to-day issues though, are there are any lessons that we can pull out of it now for ordinary people everywhere who may or may not drink, and for their doctors who may want to advise them?

Well I think in the US we should be aware that gastric cancer rates and mortality have been decreasing and this has been going on for quite some time, and this is probably tied back to H. pylori infection rates which have been decreasing as well as diets, improving the salt intake in our diets are going down. We are not eating as much preserved meat because of refrigeration. So basically this is one of the reasons why gastric cancer rates have been going down, it is because we are no longer eating preserved meats. So I think, taking it back to the result that we found, we see this association with heavy beer drinking and I think that is something we could focus on at this point, and whether we should say anything about the genetics, I think at this point we need to replicate this finding in other populations and I want to emphasise that. I think that is important for any study, and this is normally what is done is initially a study finds a result and then they say, OK, let’s look at this in other populations. So I think that’s the first step before we can say anything about what we should do in terms of screening.

Would you advise everybody to follow the French example and drink wine?

Well I am not sure if I would suggest drinking a lot of wine, but yes.

That’s a possibility?

Why not?

Eric Duell, thank you very much for being with us here on ecancer.tv.

Thanks.