'Breast cancer blood test' needs more work
“A genetic test could help predict breast cancer many years before it develops,” the Daily Mail has reported.
The newspaper says the test is based around identifying a type of DNA change called methylation, which is reportedly caused by “exposure to environmental factors such as hormones, radiation, alcohol, smoking and pollution”.
The research behind this news was a large study looking at how breast cancer risk might be linked to the levels of chemicals attached to certain sections of our DNA. Researchers analysed historic blood samples from over 1,300 women, some of whom had breast cancer and some who did not. They were interested in looking at a particular gene in white blood cells, comparing patterns of methylation between the two groups of women. They found that women with the highest levels of methylation had an 89% increase in the odds of developing breast cancer compared with women with the least modification. The researchers therefore concluded that methylation on the surface of the gene could potentially serve as a marker for breast cancer risk. They added that further research may identify similar markers.
Methylation has been in the news several times in recent months, with some studies linking it to disease risk and others looking at whether factors such as exercise could reverse the process. However, despite some news stories suggesting that blood tests looking at methylation may predict or detect early cancer, it is not yet known how this modification might influence risk, or how it interacts with other breast cancer risk factors. Importantly, a simple blood test based on this research is not available and is unlikely to be available for some time.
Where did the story come from?
The study was carried out by researchers from Imperial College, the Institute for Cancer Research and other institutions throughout the UK, Europe, the US and Australia. The research was funded by the Breast Cancer Campaign and Cancer Research UK.
The study was published in the peer-reviewed medical journal Cancer Research.
This study was covered appropriately in the media, with The Guardian pointing out that this research has only identified an association between DNA changes and breast cancer risk. It hasn’t identified a definitive link between the two nor the underlying mechanism that may be involved.
What kind of research was this?
Human DNA contains sections of code that perform a specific function, and these are known as genes. These genes contain instructions for making proteins, which then go on to perform a host of important functions in the body. This case-control study examined the association between a type of genetic modification called “methylation” within DNA and the development of breast cancer. DNA methylation occurs when a molecule binds to a gene. The addition of this molecule can “silence” (turn off) the gene and prevent it from producing the protein it normally would.
Case-control studies compare people with a particular disease or condition (the cases) with a group of comparable people without that condition (the controls). Case-control studies are a useful way to investigate risk factors for a relatively rare disease, as cases are identified on the basis that they already have a particular disease. This allows researchers to recruit a large enough number of subjects with a condition to produce a statistically meaningful analysis. This would be much harder to do if they followed a large group of volunteers and simply waited for a sufficient number to develop a particular disease.
In standard case-control studies, both cases and controls are asked about their previous exposure to risk factors, allowing researchers to analyse how their past exposure related to the risk of developing the condition being studied. This, however, does not always accurately measure risk factors, as the participants may not correctly recall their exposure, or information on the exposure may not be readily available. It is also difficult to guarantee that the exposure occurred before the development of the disease.
To get around these limitations, researchers may conduct what are called “nested case control studies”, in which the participants are drawn from existing “cohort studies” – where a large population is followed over time to see who went on to develop a particular disease. Sourcing participants from a cohort study means researchers can evaluate participants’ circumstances and exposures before they developed the disease, providing a better appraisal of participants’ past exposure than from simply asking about their histories, as would happen in a normal case-control study.
In this research, participants were drawn from three cohort studies that had collected blood samples from a large group of women who were judged to be free of breast cancer at the time they entered these studies. These women were then followed up over time. The researchers identified women from these cohorts who had gone on to develop breast cancer, and matched them to other cohort participants who had not developed the disease. Nesting the study in this way ensured that the analysed blood samples were drawn before the cancer was diagnosed, allowing researchers to compare pre-diagnoses methylation levels between the two groups of women.
What did the research involve?
The researchers used three prospective cohort studies to identify breast cancer cases and matched control participants. The first study involved women with a family history of breast cancer who were considered to be at high risk for developing the disease. The second and third studies were cohort studies conducted among the general population. All of the cohort participants had had a blood sample taken as part of the original study, before any cancer diagnosis.
All of the cohort studies collected blood samples from the participants. Samples were taken an average of 45 months before breast cancer was diagnosed in the first study, 18 months in the second and 55 months before diagnosis in the third study. In addition to blood samples, information was collected on other breast cancer risk factors, such as hormonal and reproductive factors, smoking status and alcohol drinking status.
The researchers analysed white blood cells in the blood samples to determine the degree of methylation they had within a specific gene called the ATM gene. The ATM gene is involved in many functions, including cellular division and the repair of damaged DNA. The researchers then compared the average level of methylation between cases and controls in each cohort study to determine whether there was a significant difference in the degree of modification to the ATM gene.
The researchers then divided the study participants into five groups based on their level of methylation. For each methylation group, the researchers assessed the odds of having breast cancer. They then compared the odds of developing the disease in the groups with the higher levels of methylation with the group with the lowest level. This analysis combined the data from the three cohort studies and controlled for a variety of confounders that could potentially account for the association between gene methylation and breast cancer diagnosis. This analysis was also stratified by participant age, family history of breast cancer and the length of time from blood test to diagnosis in order to assess whether or not these factors modified the relationship.
What were the basic results?
The exact number of women involved in the three studies is not featured in the study paper but the details mentioned suggest it was around 640,000 in total. Among these women, the researchers identified 640 breast cancer cases and 780 healthy control subjects. They found that, in two of the three studies, cases had significantly higher average levels of methylation at a specific point on the ATM gene than controls did.
When comparing the odds of developing breast cancer between the highest and lowest levels of methylation, the researchers found that:
- Participants in the fifth quintile (with the highest degree of methylation) had significantly higher odds of having breast cancer compared with the lowest methylation group (odds ratio [OR] 1.89, 95% confidence interval [CI] 1.36 to 2.64).
- Participants in the second, third and fourth quintiles (intermediate degrees of gene methylation) showed no significant difference in the odds of having breast cancer compared with the lowest methylation group.
When the results were stratified by participant age, the researchers found that this pattern was strongest among women under the age of 59, and not significant among women between the ages of 59 and 91.
How did the researchers interpret the results?
The researchers concluded that high levels of methylation (modification of the ATM gene) might be a marker of breast cancer risk.
This case-control study provides evidence that a type of molecular modification (methylation) at a particular genetic site may be associated with an increased risk of developing breast cancer.
The researchers said that the identification of a white blood cell DNA methylation marker for breast cancer is quite useful because it can be detected through assessing a simple blood sample, as opposed to the extraction of tissue samples that is often needed to identify cancer markers.
This study had several strengths, including:
- The case-control study was “nested” from three large, independent cohort studies. Nesting is a process were participants are taken from existing studies so that researchers can examine details of their histories that have been formally recorded at the time, rather than being simply recalled.
- Using blood samples taken before a cancer diagnosis allowed the researchers to be confident that the study results were not due to “reverse causality” (that is, the possibility that active cancer or treatment might cause DNA methylation).
There are some limitations to the study that should be considered:
- The selection of appropriate controls is important for case-control studies, as ideally subjects should be from the same study base. For the first study, cases consisted of women with a strong family history of breast cancer, while their friends with no family history were selected as controls. This is not an ideal method of identifying controls, as controls lacked the key risk factor of a family history of the disease.
- Across the three cohort studies, there were varying strengths in the association between white blood cell DNA methylation and breast cancer risk. The strongest association was seen in the cohort study that included women with a strong family history of the disease. Whether this association was due to genetic predisposition to the disease or weaknesses in the case-control design for this cohort is difficult to say at this point.
The researchers said that additional research is needed in order to investigate the effect of age on the association between methylation and risk of breast cancer. They also said that their results supported the further investigation of common variations in DNA methylation as risk factors for breast cancer as well as other cancers.
It’s important to note that a simple blood test based on this research is not available yet, and is unlikely to be available for some time. There are various known genetic, medical and lifestyle risk factors for breast cancer, and the extent to which any modification of this white blood cell gene influences risk, or interacts with other breast cancer risk factors, has not been established.
Although media reports suggest that these findings could lead to a simple blood test to screen women, or to detect the earliest stages of cancer, it is far too early to be sure of this. Before any screening test is introduced, extensive research and consideration is needed to determine in which groups of people the benefits of screening (such as reduced incidence of breast cancer and improved survival) would outweigh the risks (such as false positive or false negative results, further diagnostic tests and treatments or associated anxiety).
- Brennan K, Garcia-Closas M, Orr N et al. Intragenic ATM Methylation in Peripheral Blood DNA as a Biomarker of Breast Cancer Risk. Cancer Research, May 1 2012 72; 2304