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

Breast milk analysis may offer early cancer test

Professor Kathleen Arcaro – University of Massachusetts Amherst, USA

Kathleen, you’ve just been talking about detecting breast cancer risk. We’ve been at a session about detection of cancer and, of course, there’s a hope of getting cancer at very early stages where, hopefully, it can be treated more effectively. What were you doing and why were you doing it?

We think that you can use breast milk to detect increased risk and to then identify women who are at increased risk and separate them from ones who are not. And then the women who are at increased risk could have more closer follow-up with the traditional methods.

So if you have a woman who is having a baby then, get some breast milk and then you can assess whether or not that woman is at risk at any time in her life.

Well we don’t know for certain how the signals that we’re trying to detect are epigenetic signals and so many other researchers have shown that these epigenetic signals, this methylation of the DNA occurs early in disease. It precedes disease but we don’t know how early. So we don’t know can we, if we get milk from a woman at 25, can we say she’s going to be at risk at 50? We don’t know the answer.

You’re looking then for methylation which is an indicator of a kind of potential for breast cancer?

That’s right.

What exactly did you look for?

When methyl groups, a carbon with three hydrogens, get attached in certain regions of the gene, they turn the gene on.

Which gene?

We looked at three genes here, we reported on three genes: RASSF1, SFRP and GSGP. The first two, we call them tumour suppressor genes and we call them tumour suppressor genes because when they’re not functioning, when the protein is not made, they tend to push the cell into promoting a tumour in that there is uncontrolled cell growth or lack of DNA repair.

And if there is methylation then, they’re more likely to go towards cancer?

That’s right. If there’s methylation the proteins don’t get made so the gene gets shut off so they would be suppressing the growth but now they’re not and so you’re more likely to end up with a cancer-like profile.

Now the good thing about breast milk is that there’s lots of it but is it actually containing the sorts of genes that you want? Couldn’t you do, say, aspiration or biopsy, fine needle biopsy?

If you want to screen and you think about maybe putting a catheter through a woman’s nipple to go into a duct, the ductal lavage, that’s another way of doing it, it really is quite invasive. Then you’ve also only surveyed one small area of the breast and you usually get a limited number of cells. Whereas breast milk it’s easy to do both breasts; if you were unsure of your results women could give you a second sample easily and the milk that comes out does contain these cells that naturally slough off, these epithelial cells, it contains millions of them and they come from the whole breast.

So they are the tumour suppressor genes that you’re looking for and you can check to see if they’re methylated?

Right, they contain the tumour suppressor genes that we’re looking for and you can test, yes.

Could you give me the numbers then? How sure are you from the data that you have so far, the tests you’ve made?

We can’t talk about specificity; the two things that you really want to know about – you want to know about the specificity and the sensitivity and we can’t talk about that yet. We know that we’re seeing increased methylation in a small subset and that’s what we might expect, that only a small subset are going to be at risk, but we haven’t followed them long enough to see, even though we are looking at this population where we ended up with thirteen cancers because they’re higher at risk. We need long term follow-up and we need to look at more genes.

In fact you’ve looked at 182 women, that’s the latest figures and you’ve found thirteen that actually had breast cancer, so you have got some relatively hard data. Have you got anything for clinicians to latch on to at the moment?

Every clinician, I’m not sure yet that we’re giving them information that they don’t know already. Most clinicians would agree that methylation in breast cells is going to put a woman at increased risk of breast cancer, most would agree with that. If anything I think maybe clinicians can help me, if they have a nursing mother who needs a biopsy, please give me a call and I’ll take her milk. We need to examine more samples before we can say anything more definitive.

In the scheme of things, what would you like to see happening? Because breast cancer is a massive risk to the population at large, how do you see this particular sort of test helping towards reducing that burden?

Right now the recommendation is that women only start mammographic screening at age 50, except if you’re at high risk for developing breast cancer. But most breast cancers are sporadic so women don’t know if they’re at high risk. I think a screen like this early on, could tell women, you’re at high risk or you’re not at high risk. I think that would be beneficial.

Every woman who has a baby then potentially has the chance to have this method of determining her risk?

That’s right. Every woman who gives birth and as the population of women give birth later and later, as you know, they’re at greater risk for developing breast cancer when they give birth later. They might benefit significantly more from this screen.

Now, as you say, it’s early days yet, you don’t have enough data to say everything about this but what would you tell clinicians at this point about how they should view these sorts of developments?

You could summarise that it really is clear that the cells, the epithelial cells in the milk, will give us an indicator of risk if we can look at a larger panel of genes. So that might be the take home message – there may be women who clinicians are particularly concerned about, BRCA mutations and all, and something assessing their risk would be of considerable importance, especially when you think that women have prophylactic mastectomies. If we could give them more accurate information they could make a more informed decision.

So yet more information that will help us view the whole patient and individualise the approach to that patient.

I would hope to see this assay move into something like that, yes.

Kathleen Arcaro, thank you very much for joining us here on ecancer.tv.

Thank you.