On April 30 at 7:30 PM, I’ll be part of a panel on Health Link with Benita Zahn, WMHT TV, to discuss the genetics behind the “Angelina Jolie effect” that has catalyzed testing for the BRCA mutations that increase risk for breast and ovarian cancer. It’s a tough assignment.
A letter in the current People magazine referring to Jolie’s recent announcement of the removal of her ovaries, following a double mastectomy last year, illustrates how at least one person is misconstruing the genetics of Jolie’s situation:
“How did she go about getting these types of tests and elective surgeries? It would be good to know if the same options are available for all women or if these procedures are something only afforded to the rich and famous.”
RISK MUST BE INDIVIDUALIZED
Angelina Jolie was tested and treated because of her family history, not her fame and fortune. Any genetic counselor or primary care provider would have been alert to her background: Her mother, grandmother, greatgrandmother, aunt, and a cousin had BRCA1-associated cancers.
That heritage places Jolie at far greater risk than most women. Mutations in the two BRCA genes account for only 5-10% of breast cancer cases and about 15% of ovarian cancers, according to the National Cancer Institute. Based on her family history, Angelina faced an 87% chance of developing breast cancer — about five times the general population risk — and a 50% risk of developing ovarian cancer.
Two charts from the upcoming Health Link show will use the NCI data to compare lifetime risks for people with BRCA mutations to those of the general public. (The “lifetime” is important. An article in the Boston Globe recently made the common error of stating that 1 in 8 women has breast cancer.)
BRCA mutations are “germline.” That means a person is born with one mutated copy in each cell. With a second mutation in the gene, years later, in a breast or ovary cell, cancer develops. This “2-hit hypothesis” is from Alfred Knudson’s work in 1971, reflecting even earlier ideas and still fueling the delineation of multiple mutations unfolding as cancer progresses. However, just the presence of one BRCA mutation disrupts messenger RNA production enough to increase cancer risk — a “one hit” effect.
But BRCA isn’t all there is to inherited cancer risk.
MANY GENES CONTRIBUTE
We each have two copies of the two BRCA genes (unless a very rare mutation deletes the entire gene). But as is true for all genes, our personal versions vary, because DNA is an informational molecule, a sequence of four nucleotide base types. These two giant genes can differ in many ways, only some of which increase susceptibility to cancer.
Three mutations – the founder or Ashkenazi mutations – are “pathogenic” and elevate risk considerably. Yet some other versions of either gene are “variants of uncertain significance” — a patient could learn her gene has an unusual sequence, but be told that it isn’t known whether or not it increases cancer risk.
The BRCA genes encode proteins that repair DNA broken in both strands of the double helix. That means DNA anywhere, not just in the gene itself, so the normal versions of these genes are part of the cell’s arsenal to limit DNA damage. Other proteins do this too, and the repertoire of repair proteins interact to protect the cell.
We can tier the risks. BRCA1 and BRCA2 mutations are the most dangerous, while mutations in other genes, such as ATM, TP53, CHEK2, and PTEN, confer a more moderate increase and are behind family cancer syndromes. But that’s not all. Variants of dozens of other genes contribute in still-little-understood ways that slightly elevate cancer risk.
I’ve recently written about all of these, so I thought I’d summarize the findings here. Skip to the end for a look at online cancer risk calculators.
LOCATION, LOCATION, LOCATION
A report in the April 7 Journal of the American Medical Association shows that how likely a BRCA mutation is to lead to cancer depends on where it is in the gene.
Timothy Rebbeck, an epidemiologist at Penn Medicine’s Abramson Cancer Center and colleagues, displayed the BRCA genes like maps of cities scattered among vast underpopulated areas, highlighting hot spots where meaningful mutations lie.
Trouble arises from the middle and the ends of BRCA1. The gene has 24 exons (the parts that encode protein). Mutations in exon 11, right in the middle, raise risk of ovarian cancer, while mutations at the tips raise risk of breast cancer. BRCA2 harbors three regions where mutations correspond to breast cancer and a different three that lead to ovarian cancer. Plus, mutations in different parts of the genes may influence age at onset. That’s important when weighing a decision to take a preventive measure that would affect fertility, or a treatment that has long-term adverse effects.
We don’t yet know how women will use this type of information. “If before testing a woman has a 50% risk and once she knows a particular mutation it changes to 60% risk, is that enough to change her pattern of behavior to prevent cancer?” asked Dr. Rebbeck.
MORE THAN A DOZEN GENES
Several companies, such as GeneDx and Myriad Genetics (known for the Supreme Court Decision about patenting the genes), have offered tests for the genes behind familial breast and ovarian cancer for years. They sequence entire genes as well as the most common mutations, including duplications and deletions.
A newer player is Invitae. “BRCA is just the beginning,” reads a banner on their website. For $1500, a physician can order pre-curated collections of up to 17 genes that confer risk to developing breast cancer, add another 17 cancer predisposition genes, or create a panel for a particular patient.
Removing total cost from the equation frees physicians to order tests based on a patient’s history and clinical situation. Plus, because all samples are tested for all the genes that the company covers, the physician can dip back into the data as a patient’s cancer progresses, seeking clues that might inform prognosis and treatment. A study from the company showed that doctors choose both pre-curated and customized test panels.
THE POLYGENIC COMPONENT
Genome-wide association studies (GWAS) take 2 populations differing by one factor, such as a disease, compare the DNA bases at millions of sites in the genome (single nucleotide polymorphisms, or SNPs), and see what the sick people have that the healthy ones don’t. GWAS results are associations, not causes. Another recent paper, in the Journal of the National Cancer Institute, reveals how analyzing SNPs can refine cancer risk prediction. (I wrote about it for Medscape.)
Celine Vachon, an epidemiologist at the Mayo Clinic and colleagues, used a “polygenic risk score” based on 76 SNPs identified in various GWAS, plus breast density, to predict risk among 334 healthy women. Women with the highest SNP scores and dense breasts had a 2.7-fold increased risk of developing breast cancer, and 11% of women previously classified as low risk were actually of higher risk, making them candidates for chemopreventive drugs, MRIs, and prophylactic mastectomy.
LET’S NOT FORGET THE ENVIRONMENT
Most cancers are sporadic (non-familial), beginning with two mutations in both copies of a gene in the same cell, sometime after birth. The cause is an accident of DNA replication or a response to an environmental insult. The much rarer familial cancers tend to appear at younger ages because one cancer mutation is inherited, the other a consequence of later mutation in that first errant cell. The 2-hit hypothesis.
The idea that the environment can influence even the strongest genetic predisposition comes from a 2008 study. It found that average cumulative breast cancer risk to age 70 among women with BRCA1 mutations was 50% for women born between 1920 and 1929, but 58% among those born after 1950. Genetic change isn’t that fast.
Genetic counselors look for clues that a BRCA gene mutation may be at play in a family. They include:
• breast cancer in a man
• more than one type of primary cancer in one relative
• several generations with breast and/or ovarian cancer
• breast cancer in a person under 50
• cancer in each breast
• Ashkenazi Jewish, French Canadian (including Cajun), or Icelandic background.
Online Mendelian Inheritance in Man indicates other populations with elevated risk.
A genetic counselor would also consider other types of cancers in a family. BRCA1 mutations predispose to cancers of the cervix, colon, and pancreas, and BRCA2 to cancers of the stomach, gallbladder, pancreas and bile ducts, as well as to melanoma.
Even with lists of what to look for, BRCA mutations can go unnoticed for many generations, not causing cancer. That’s the case for my friend’s family. The 21-year-old son found out he had a BRCA1 mutation using 23andMe’s tests (when that was kosher), and we all assumed his mom, of Ashkenazi heritage, had passed it to him. But further testing revealed that his Catholic father had the mutation. Fortunately, no one in the family has cancer, although several relatives have refused to believe the test results.
EXPLORING RISK CALCULATORS
I looked up my own risk using a few breast and ovarian cancer risk calculators. Everyone in my extended family has had cancer of some form, including me, except my sister. But the cancers were all late onset or due to some obvious environmental exposure, like my aunt the chain smoker and sun worshipper, my uncle the long-time radiologist, and me exposed to years of x-rays for orthodontia without appropriate shielding. My comparison is unscientific and apples-and-oranges, and some omissions, like Susan B. Komen, adapt their tests from the NCI tool.
The Myriad calculator for having a BRCA1 or BRCA2 mutation gave me a risk of 8.2%.
The Penn risk calculator gave me a risk of 6% of having a BRCA mutation. It suggests anyone over 5% risk take a BRCA test.
The NCI’s Breast Cancer Risk Assessment Tool isn’t specific for the BRCA genes, includes some questions about reproductive history (important risk factors), but oddly, once I put in that I was white, did not give me the option to add Ashkenazi background. Still, it gave me a 14.3% risk of developing cancer by age 90 compared to the 8.6% of the general population. Glad to see lifespan extended!
The National Foundation for Cancer Research quiz looked only at age, and gave me a 284/100,000 risk for breast cancer, which they deemed “very high.” Huh?
Most helpful was the Bright Pink general breast cancer risk tool, because it asked everything. BMI. Exercise. Alcohol consumption. Breast density. Age at first period, first kid, breastfeeding. Family history.
Bright Pink provides a “report card” of sorts that tells you good things (“working in your favor”) so you don’t panic or feel helpless, then lists modifiable risk factors – things you can do to lower your cancer risk. I came out at “potentially high risk” and was advised to “see a doctor or genetic counselor to confirm that your baseline risk truly is only increased, and not actually high.” I appreciate the qualifiers.
NOT A SIMPLE SITUATION
Genetics is an imprecise science, and that’s why I try to use precise language. People develop cancer, they don’t contract it. The BRCA genes do not cause cancer, they elevate risk. I don’t even call people who have one BRCA mutation “carriers,” as is common, because as an old-school geneticist, carrier to me means recessive. The risk that inheriting a mutation imparts is dominant. It’s not hidden.
Some people with family histories riddled with cancer never develop it. And a person can be the only one in her family with a BRCA-associated cancer, never coming to the attention of a health care provider asking about family history — there is none. That’s what led Mary-Claire King, who discovered the mutations, to write a controversial viewpoint last fall advising population-wide screening.
Angelina Jolie has done a world of good by opening up about her actions to prevent the cancers that might have been her genetic destiny. But at the same time, her message that her family history is unusual may not have gotten through. According to Reuters, insurers are indeed balking at the increased demand for BRCA testing in the wake of Angelina’s messages. Genetic counselors are the experts who know the most about the complex issue of hereditary cancer, and can help to distinguish it from sporadic cases. See the National Society of Genetic Counselors to find help — and likely reassurance.