Too good to be true? A blood test ‘to predict breast cancer’

Can ‘a simple blood test’ predict ‘a woman’s risk of breast cancer’? You might think so if you’d read the news this morning.

But as the old aphorism goes, if it sounds too good to be true, it probably is – and this story is no exception. The research that led to these stories isn’t ready for the clinic yet, and it’s far too early to be talking about “saving millions of lives”, or incorporating it into the NHS screening programmes “within five years”.

But it is a fascinating and intriguing finding.  So, given the widespread coverage, we thought it would be useful to have a quick look at what the researchers – who were led by an EU-funded research team at University College London – actually found (you can read their original paper here if you’re interested).

What did they do?

The researchers looked at blood and cheek-swab samples from women who’d taken part in several previous large-scale cancer studies (including the CRUK-funded UKCTOCS ovarian screening study).

Some of these women had gone on to develop breast cancer, so the researchers wanted to look back and see if there were any signs of the disease in samples taken before the disease developed.

They were particularly interested in something called ‘epigenetic’ changes in DNA from the samples. These changes are little chemical tags attached to our genes, which turn them on and off – and there’s a substantial body of evidence suggesting that changes in our epigenetic tags play a role in the development and growth of cancer, including breast cancer.

To find out if there was an epigenetic ‘signature’ in the blood of women who developed breast cancer, they looked at blood from women who carried faults in the BRCA1 gene, which means they were at unusually high risk of developing breast and ovarian cancers. Some of these women had developed breast cancer, but others hadn’t. (It’s worth noting that only about five per cent of women with breast cancer carry this gene fault).

They found a promising-looking signature in the women’s white blood cells that was linked to their chance of developing cancer. But what about women who didn’t carry a faulty BRCA1 gene?

When they looked at larger studies for this signature in women who didn’t carry a BRCA1 fault, they found that the signature appeared in the blood of women who went on to develop the disease.

This strongly suggests there’s something in these women’s blood that could be used to help work out their chance of breast cancer in later life. But – crucially – it also throws up several questions:

• What is this a signature of?
• Why is it in white blood cells?
• How long before breast cancer develops does it appear?
• Why does it appear?
• Is it an early sign of a cancer itself, or of an underlying process that precedes the disease?
• How is it affected by other things that affect breast cancer risk, like bodyweight and age?

All of these questions – and more – need to be answered before we can start talking about ‘tests’ available on the NHS.

Let’s clear a few things up

First, this isn’t ‘a simple blood test’ – a claim initially made in the press release put out by the journal which published the research, and then echoed in all the subsequent coverage (although hat-tip for the Independent for an otherwise excellent article). The analysis involved is extremely complex, requiring specialist technology and expertise. So it’s a long way from being able to provide the sort of reliable, evidence-based information a doctor would need to rely on to make clinical decisions about a woman’s care.

Secondly, the test didn’t ‘predict’ breast cancer, in the everyday sense of the word – it was present in the blood of some women who didn’t go on to get cancer, and also absent in some women who did get cancer. But it was more likely to appear in women who went on to get the disease.

Thirdly, many of the stories mentioned timescales that, we think, are wildly overoptimistic, given the unanswered questions above. There’s a very long way to go, and a lot to prove, before this could be routinely used. Almost certainly more than five years.

Finally, using words like ‘breakthrough’ and ‘revolutionise’ is really over-egging the pudding of what’s a relatively early-stage finding. (And please, can we not call BRCA1 ‘the Jolie gene’)?

So why’s it exciting?

The real interest to us is what this says about the biology of breast cancer. The changes, as we mentioned, were spotted in the woman’s white blood cells. This could mean several things.

There could be something inherent in these women’s immune systems that makes cancer more likely to develop. That would be a really exciting finding, especially in the light of the excitement about new drugs that reawaken the immune system to target cancer.

Or, it could be the hallmark of something in these women’s lives that makes cancer more likely, written into their immune genes. The researchers think they’ve ruled out hormones (there wasn’t a link between hormone levels of the women on the study, and the presence of the signature), but there could be other things, and this needs more investigation.

On top of this, the actual genes that were modified to make up the signature, known as ‘polycomb group target genes’, were all known to be involved in a process called stem cell differentiation – this is really interesting, and we need to know more about how these genes are involved in breast cancer development.

Being able to predict a woman’s breast cancer risk would be an incredibly valuable tool, and is an area of intense investigation (and of huge interest to the media and public). Our researchers have been working hard on it too, and have identified many of the genetic elements involved in breast cancer risk.

So we’re sure that, one day, researchers will develop a reliable way to look at a woman’s DNA and predict whether she’s at low, medium or high risk of cancer. This could allow her to make changes to her lifestyle, go for extra (or, indeed, less) screening, or even take drugs to prevent the disease. We hope this day arrives soon. But it’s still a way off for now.

Henry

Reference

• Anjum S et al (2014) A BRCA1-mutation associated DNA methylation signature in blood cells predicts sporadic breast cancer incidence and survival. Genome Med DOI:10.1186/gm567

Image via Flickr