A paper published in the New England Journal of Medicine (NEJM) this week, entitled “Performance of Common Genetic Variants in Breast-Cancer Risk Models,” has led several media outlets to declare that common genetic variants have nothing to add when it comes to predicting breast cancer risk.  Here we’ll explain how the results of this study have been misinterpreted.

Researchers from the National Cancer Institute and other institutions looked at data from 5,590 women with breast cancer and 5,998 women without the disease.  These women, all between 50 and 79 years old, had participated in one of five studies, four in the United States and one in Poland.  Since it’s known already which women have or have had breast cancer, and which have not, the researchers were able to use their data, both genetic and non-genetic, to test the predictive power of different types of risk calculations.

The study tested five different models.  The “demographic model” considered only age, year of entry into the study and which study the woman was originally part of.  The “nongenetic model” added in several variables that are part of the so-called “Gail model,” which is the standard model used in clinical practice today to counsel women about their risk.  These variables were the number of first-degree relatives with a diagnosis of breast cancer, age at menarche, age at first live birth, and number of previous breast biopsies.  Two models used the demographic information plus genetic information for 10 SNPs (they differed in details of how the genetic risk score was calculated).  Finally, the “inclusive model” combined demographics, the Gail model and the genetic information.

The genetic models and the nongenetic model performed about the same, with genetics doing just a little bit better.  Perhaps not surprisingly, the best model was the one that used the most information.  With the inclusive model, which is based on genetic and non-genetic information, there was a net 12% improvement in risk classification over the nongenetic model for women with breast cancer.

So, SNPs did add something.  The improvement in prediction seen when SNPs are added to the nongenetic model is about the same as the improvement seen when the Gail model information is added to the simplistic demographic model.  Discounting the benefit of adding SNP information to the Gail model implicitly discounts the utility of the Gail model itself.  Peter Devilee and Matti A. Rookus made this very point in a NEJM editorial that accompanied the study.

It’s also important to remember that the study of common variations and their effect on common diseases like breast cancer is a relatively young science.  Many more variants are bound to be discovered, and these will only help further refine risk predictions.

It’s true that there is a cost associated with collecting a person’s genetic information, while collecting the information needed for the Gail model is essentially free.  (In fact, a risk calculator based on the Gail model is available at the National Cancer Institute website.) But for those people who do have their genetic information in hand, the current study shows that it can improve the estimation of their risk for breast cancer.  It should be remembered, too, that not everyone is fortunate enough to have access to their family medical history. Adoptees, for example, cannot get an accurate picture of their breast cancer risk from the Gail model.  For such women, the ability to use their own genetics to help assess their risk for breast cancer is an option that should not be dismissed.

Ultimately, it’s not about genetics vs. non-genetics — it’s about getting accurate estimates of risk to help doctors catch cancers early on, when they’re easiest to beat. Researchers should welcome any tool that can help them move closer to keeping more women healthy.  Giving up on genetics this early on in the game would be a real disservice.

Note: A similar study can be found in Nature Precedings.  The results are consistent with the NEJM study regarding the improvement in risk prediction when SNP information is added to the Gail model.  This work also shows that the improvement is larger for women at intermediate risk who are more likely to be reclassified.  Senior author David Hinds was at Perlegen Sciences when this paper was written and is currently an employee of 23andMe, Inc.

ShareThis

No matter when it happens, puberty is no fun.  But for girls, entering womanhood earlier or later than average can mean more than just the social awkwardness of being quick to mature or a late bloomer.  It can have important health consequences later in life.

In general, girls in the developed world are maturing faster now than ever before.  In the mid-nineteenth century the age at menarche (the first menstrual period, a milestone usually reached about two to three years after puberty begins) was about 16, but by the end of the twentieth century that number had fallen to about 13. Earlier age at menarche has been associated with an increased risk for breast, ovarian and endometrial cancer. On the other hand, later menarche is linked to increased risk for osteoporosis and cardiovascular disease.

Much of the change in the rate of maturation for girls has been attributed to better nutrition and rising rates of obesity, but new research shows that the age at menarche is also influenced by genetics.

Four papers (Ong et al., Sulem et al., Perry et al., and He et al.), published online yesterday in the journal Nature Genetics, analyzed data from a total of more than 60,000 women with European ancestry and identified two regions of the genome that contain variations associated with small differences in age at menarche.  One of these four studies, and a fifth paper (Stolk et al.), also found variations associated with the timing of menopause, the other end of the reproductive spectrum for women.

All four studies looking at age of menarche found a connection with a region of chromosome 6. Each paper identified slightly different variants, but they all appear to be zeroing in on a single, yet to be found, causative variation in the DNA.  As an example of the effect of the chromosome 6 SNPs, Ong et al. found that compared to those with two Cs, each T at rs314263 meant that woman got her first period about six weeks earlier.

(Ong et al. actually looked at rs314276, but we’re focusing on rs314263, a perfect proxy for the original SNP.  23andMe customers can check their data for rs314263 using the Browse Raw Data feature.)

Each T at rs314263 was also associated with 1.20 times greater odds of breast development at age 10, 1.26 times greater odds of advanced breast development between the ages of 9 and 16 and a faster rate of increase in height and weight.

Although girls who go through puberty earlier than their peers are tall for their age during childhood, their growth stops sooner and they are generally shorter as adults.  Ong et al. found that each T at rs314263 was associated with a reduction in adult height of about 0.15 inches.

Two studies also found an association between variations on chromosome 9 with age at menarche.  Each C at rs7861820 (the most significant SNP found in He et al.), for example, translated into about five weeks earlier menarche.

He et al. and Stolk et al. found variations in five regions of the genome associated with another big change in a woman’s life: menopause. The risks associated with the timing of menopause are the reverse of those associated with menarche: later menopause has been associated with higher risk for certain cancers and earlier menopause is linked to increased risk for osteoporosis and cardiovascular disease.

(23andMe customers can check their data for SNPs representative of these five regions using the table at the end of this post.)

Finally, all of you boys and men out there need not feel left out:  Ong et al. also looked at the effects of rs314263 on puberty in males.  They found that each copy of an T translated into 1.26 greater odds of advanced voice breaking at age 15, 1.19 times greater odds of advanced pubic hair at age 13, faster rate of growth at age 10 and about a tenth of an inch less in adult height.

SNPs Associated With Menopause

ShareThis

BRCA1 and BRCA2 (BRCA1/2) mutations account for most (though not all) cases of inherited breast cancer in women. These mutations are also associated with an increased risk for ovarian cancer. In men, BRCA1/2 mutations increase the risk for breast cancer and may also increase prostate cancer risk.  Research has indicated there may also be an increased risk of melanoma and pancreatic cancer in people with BRCA1/2 mutations.

Although BRCA1/2 mutations significantly increase the risk of cancer, having one of these mutations doesn’t mean a person actually has cancer or will necessarily ever get the disease.   But new research from scientists at the National Cancer Institute and the National Human Genome Research Institute, published online this week in the journal PLoS ONE, suggests that those BRCA1/2 carriers who do escape cancer may still be at risk of a shorter than average lifespan.

In a study of almost 5,300 people with Ashkenazi Jewish ancestry researchers found that, on average, both men and women who carry BRCA1/2 mutations die younger — even when cancer deaths are excluded from the analysis.  Among women free of melanoma, breast, ovarian and pancreatic cancer, BRCA1/2 mutation carriers died 5.8 years earlier (age 75.0 vs. 80.5) than those women without the mutations. In men, after excluding cases of melanoma, prostate and pancreatic cancer, BRCA1/2 mutation carriers died about 3.7 years earlier than those without the mutations (71.0 years old vs. 74.7).

“Theoretically, these mutations may either be associated with a small increase in risk of a variety of different diseases, or they may be associated with moderate increase in risk of a few major diseases,” the authors write.  They note that the current study is unable to make this distinction because it did not collect information on non-cancer related causes of death.

The researchers analyzed the effects of only the three BRCA mutations that are most common in Ashkenazi Jewish people: 185delAG in BRCA1, 5382insC in BRCA1, and 6174delT in BRCA2.  The mutations account for 80-90% of hereditary breast and ovarian cancer in this ethnic group.  But there are hundreds of other BRCA1/2 mutations that have been associated with cancer, and the authors caution that further studies taking these other BRCA1/2 mutations into account and using study subjects from diverse groups will be needed to confirm their results.

(23andMe provides data for the three BRCA1/2 mutations most commonly found in people with Ashkenazi Jewish ancestry in the BRCA Cancer Mutations (Selected) Carrier Status Clinical Report.)

The authors conclude that understanding the effects of BRCA1/2 mutations on non-cancer related deaths could eventually help scientists understand how the mutations affect cancer risk and may even facilitate efforts aimed at finding new ways of preventing disease in BRCA1/2 mutation carriers.

ShareThis

The cells in our bodies are subject to a constant onslaught of potentially damaging chemicals. Some of these come from the environment – the toxins in cigarette smoke, for an example – while others are natural byproducts of cellular life. Regardless of their origin, these chemicals have the potential to cause serious DNA damage that can eventually result in cancer.

Cells have evolved numerous mechanisms for defusing the danger posed by these chemicals. In the event that these efforts fail and DNA damage does occur, cells also have their own DNA repair systems. Not surprisingly, mutations in the genes that encode these DNA-fixers can leave cells vulnerable and increase the chances that a cancer will develop. Bloom’s syndrome, Fanconi anemia and hereditary breast cancers caused by BRCA mutations are all examples of cancer syndromes caused by relatively rare functional mutations in DNA repair genes.

Research from the University of Texas M.D. Anderson Cancer Center in Houston published online today in Clinical Cancer Research shows that common variations, not just rare mutations, in DNA repair genes can increase the risk for pancreatic cancer.

Pancreatic cancer is the fourth leading cause of cancer death for both men and women in the United States. Environmental factors such as smoking, obesity, diabetes, chronic pancreatitis and dietary factors are known to play a part in determining the risk of pancreatic cancer, but about 10% of patients have a family history of the disease, suggesting that there are also genetic risk factors.

Donghui Lie and colleagues examined variations in several DNA repair genes among 734 non-Hispanic white pancreatic cancer patients and 780 people without the disease. They found that having two copies of the A version of rs1801516 in the ATM gene increased the odds of pancreatic cancer 2.76 times compared to two Gs. Having just one copy of the A version did not significantly increase risk.

(23andMe customers can check their data using the Browse Raw Data feature.)

SNPs in the ATM gene have also been associated with an increased risk of developing breast cancer. The protein encoded by ATM is known to be crucial for the cellular response to DNA damage.

The researchers also found a significant association between having two copies of the A version of rs2074522 in the LIG3 gene (also involved in DNA repair) and risk for pancreatic cancer. 23andMe does not currently provide data on this SNP.

“Pancreatic cancer is a highly fatal disease because most of the cases are diagnosed at late stage and the tumors are resistant to most therapies. Early detection for pancreatic cancer is crucial to reduce the mortality,” the authors write.

“However, there is no screening method available to identify the high-risk individuals among those at risk,” they write — referring to people such as smokers, people with diabetes, and those with a family history of pancreatic cancer.

The authors suggest that in addition to aiding in the development of screening tests for pancreatic cancer, finding and understanding variants such as those identified in their study will help aid in developing new treatments.

ShareThis

Our SNPwatch posts here at The Spittoon are one of our most exciting features. They give our customers the opportunity to connect their genetic data to the newest discoveries, often within just hours of a study’s publication.

Looking ahead to 2009, we can only begin to imagine the exciting discoveries that will be made in genetics. In the meantime, here are a few of our favorite SNPwatches from 2008:

SNPs That Affect Drug Response
We reported on several studies this year that showed the importance of genetic variations in determining how different people react to certain medications.

• A report in Nature Genetics showed that some women with a particular version of a SNP in the NQO1 are less likely to survive breast cancer after treatment with the commonly used chemotherapeutic epirubicin.
• A study by the SEARCH Collaborative Group found that a version of one SNP is associated with an increased risk for myopathy (muscle pain and/or weakness) in people taking cholesterol-lowering drugs called statins.
• Mayo clinic researchers found that the weight loss drug sibutramine (Meridia) is effective only in people with specific versions of three different genes.
• And just this month we brought you news of three studies that showed that a genetic variant known to affect the metabolism of the anti-clotting drug clopidogrel (Plavix) also affects heart attack patients’ risk of a second major cardiovascular event.

Shared SNPs
Sometimes multiple conditions strike the same person or run in families. Several studies published this year showed that shared genetic risk factors may be part of the reason why.

• Obesity is a known risk factor for many cancers. Researchers found that a variant of adiponectin, a hormone released by fat cells, can increase the risk of developing colorectal cancer.
• Other researchers found variants that affect the risk of developing both type 1 diabetes and celiac disease, two autoimmune diseases that tend to cluster together. One of these shared variants is also associated with HIV resistance.
• Finally, a report published this month in the Proceedings of the National Academy of Sciences showed that a single genetic variant can make a person prone to greater indulgence in both smoking and drinking.

SNPs Associated with Risk Factors for Disease
Several studies this year looked beyond disease itself and instead found associations between SNPs and traits known to be risk factors for disease.

• One study found an association between several SNPs and fasting plasma glucose, a measure of how well a person’s body can control blood sugar levels – a process that goes awry in diabetes.
• Another research group reported SNPs associated with glycated hemoglobin levels, a measure of long-term blood sugar control and another factor associated with the risk of developing diabetes.
• The findings of three papers published in Nature Genetics roughly doubled the number of SNPs associated with blood levels of cholesterol and triglycerides, important risk factors for cardiovascular disease.
• And finally, in a study that looked at behavior instead of metabolic markers, researchers found that a variant in the FTO gene known to increase the risk for obesity affects food choices in children, pushing them towards foods denser in calories.

ShareThis

People sometimes think of ancestry as the “fun” part of the 23andMe Personal Genome Service™, an intriguing glimpse at the past that is completely unrelated to our more serious Health and Traits features.

But a new paper shows how ancestry can actually reveal quite a bit about the genetics of disease. In the December issue of Cancer Research, Laura Fejerman of the University of California San Francisco and several colleagues show that among Latinas, a woman’s breast cancer risk increases with her proportion of European ancestry.

Most Latinos are descended from mix of Native American ancestors and European immigrants. In countries such as Brazil, where slavery was practiced until the late 19th century, many people have substantial African ancestry as well.

Fejerman and her colleagues used 106 ancestry informative markers (AIMs) to estimate the relative European and Native American ancestry of 440 Latina women from the San Francisco Bay area with breast cancer and 597 control subjects without the disease. With every 25 percent increase in European ancestry, the researchers found, a woman’s chances of having breast cancer rose 79 percent.

The finding is no surprise — Latinas have lower breast cancer rates compared to European women, as do Native Americans. But the study does demonstrate that ancestry and population history can be valuable tools for understanding the genetics of disease.

And researchers still have a lot more to learn about population differences in rates of breast cancer. Though environmental factors are known to play a role, the new study strongly suggests that genetics is worth investigating as well.

There’s also an intriguing puzzle in breast cancer mortality differences among different populations. Though Latinas, Native Americans and African Americans are less likely to develop the disease, when they do their chances of dying from it are greater. Researchers would like to know whether those differences are due to genetics, differences in access to health care, or other factors.

Image: Peruvian “casta” painting, late 18th c.

ShareThis

New research suggests that a gene associated with breast cancer can influence a woman’s chances of surviving the disease, particularly if she receives a drug treatment commonly given after surgery, according to a study published online Friday in Nature Genetics.

Fagerholm et al studied the gene NQO1 in 1,005 breast cancer patients from Finland and found that a woman’s survival was associated with her genotype at a particular SNP in the gene – rs1800566. After 5 years, about 85% of patients with one or two copies of G at rs1800566 survived, whereas only 65% of those with two A versions (AA genotype) did. The researchers saw similar results were seen in a second set of 1,162 Finnish female breast cancer patients – 72% of GG and GA genotype patients survived after 10 years, while only 46% of AA genotype patients did.

These results could impact a large number of women – according to the study 4% of people of European descent and up to 20% of the Asian population have the AA genotype at rs1800566.

Epirubicin is part of a drug cocktail that is often given as a chemotherapy treatment after removal of a breast tumor. When Fagerholm et al studied a subset of patients who received the treatment, they found that those with the AA version of rs1800566 were much less likely to survive.

The survival rate for women with the GG or GA genotype at rs1800566 who received epirubicin-based chemotherapy was 75% after 5 years. But for those women receiving epirubicin who had the AA genotype at this SNP, the survival rate was only 17%. There was no statistical difference in this study in survival rate for women who received other types of treatment or no treatment at all, regardless of their genotype.

The A version of rs1800566 causes a change in the protein sequence of NQO1 that results in the protein being destroyed by cells. Having the AA genotype at rs1800566 leaves cells completely devoid of NQO1 protein. Laboratory tests found that cells with the AA genotype, as well as cells engineered to lack the NQO1 protein (mimicking the AA genotype), were not as efficiently killed by epirubicin as cells that do contain the protein. The study showed that the effectiveness of other types of cancer treatment, including radiation and other drugs, were not affected by rs1800566 genotype or the amount of NQO1 present.

Approximately 20% of breast cancer tumors have a mutated version of an important tumor suppressor protein called p53. When the researchers analyzed the original group of 1,005 patients according to whether or not their tumors had this mutation (p53 information wasn’t available for the second study group), the AA genotype at rs1800566 had an effect only in women whose tumors had mutated p53. Only 20% of these women survived 5 years, whereas 73% of women with the GG or GA genotype and mutated p53 survived. In women with normal p53 protein, rs1800566 did not affect survival.

Based on the clinical and experimental results in their report – and the fact that the NQO1 protein is known to interact with the p53 protein – the authors speculate that NQO1 and p53 not only work together to prevent cancer but are both needed for drugs like epirubicin to destroy cancer cells that do develop.

The researchers theorize that in cells with functional copies of both the p53 and NQO1 genes, epirubicin treatment causes cell death. In cells that are deficient in either one of these proteins the effects of epirubicin may be less pronounced, but the end result is still the desired cancer cell death. But in cells that lack functional copies of both p53 and NQO1, the researchers think that signaling pathways are disturbed in such a way that epirubicin treatment not only is ineffective at killing cancer cells, but it may even enhance cancer cell survival and/or promote progression and spread of the disease.

The authors say that although more studies will be needed to confirm their findings, “NQO1 genotype may provide a predictive factor for treatment, possibly in combination with p53 status, both for primary tumors and, crucially, in metastatic breast cancer, where response to chemotherapy is critical for the treatment outcome and life expectancy.”

ShareThis

It turns out that the riskier version of this SNP increases the amount of the FGFR2 protein that is made by cells, which can stimulate cells to grow – and increases the chance that they will proliferate out of control and lead to cancer.

Though the number of genome wide association studies is growing every day, very few of these studies have been able to explain why the SNPs they find are associated with a particular disease. Understanding how SNPs impact the biology of cancer and other common diseases will be critical for developing new treatments.

The FGFR2 gene encodes a protein that is inserted through the surface of cells where it acts as a receptor for certain growth factors. A growth factor is a chemical signal in the body that can tell cells to divide, move around, or differentiate into more specialized types of cells.

While several genetic mutations have been found that confer a high risk for breast cancer (most famously BRCA1 and BRCA2), these known genetic changes are actually relatively rare. There are probably many more common genetic changes that on their own don’t pose much of a risk, but in certain combinations can cause disease. The SNP in FGFR2 was one of the first of these common variants to be found.

On average, a woman’s chances of developing breast cancer some time during her life are one in eight. Having two copies of the G version of SNP rs1219648 (instead of the most common genotype: AG) increase her chances to about one in six. Two copies of the A version of this SNP, on the other hand, reduce a woman’s chances to about one in 10. (These numbers apply only to women with European ancestry. 23andMe customers can learn more in My Gene Journal (now called Health and Traits).)
Meyer et al showed that the cells of women who had two copies of the G version of rs1219648 expressed significantly more FGFR2 than women with two copies of the A version (the study actually looked at another SNP that is thought to be universally linked to rs1219648).

It’s known that in 5 to 10% of breast cancers, the FGFR2 gene is turned on at unusually high levels. And laboratory experiments have shown expressing extra FGFR2 in cells can cause them to take on cancer-like traits.

When they investigated further, the researchers found that the G version of rs1219648 was associated with increased expression of FGFR2 because the versions of two SNPs (located very near rs1219648) that are inherited along with the G version of rs1219648 have a greater affinity for DNA-binding regulatory proteins that help turn on the FGFR2 gene.

Meyer et al believe that theirs is the first study to address the function of the SNPs associated with breast cancer.

“Our study demonstrates that SNPs identified by whole-genome scans can be used as valid starting points for studying the underlying biology of cancer,” the authors write.

Like many of the SNPs found in genome wide association studies, rs1219648 and other SNPs in FGFR2 that regulate the binding of regulatory proteins are in a non-coding part of the gene called an “intron”. Many other SNPs found in these types of studies are in non-coding regions of the genome located between genes.

The authors speculate that their finding that SNPs in non-coding parts of the DNA can regulate gene expression will be a common theme.

ShareThis