Tracing Breast Cancer’s History

Photo of Angelina Jolie by mfrissen via Creative Commons
Photo of Angelina Jolie by mfrissen via Creative Commons

In today’s New York Times, the actress Angelina Jolie published a remarkably forthcoming op-ed about getting a double mastectomy. Jolie carries a variant of a gene called BRCA1 that makes women highly likely to develop breast and ovarian cancer. Her mother, who also carried the variant, died of ovarian cancer at age 56. Like a number of other women with her condition, Jolie decided to get a mastectomy as a preventative measure.

A number of studies have shown that bilateral risk-reducing mastectomies (the official term) do indeed reduce the risk of breast cancer in women with the BRCA1 mutation. In a study published last month in Annals of Oncology, a team of Dutch medical researchers tracked 570 women with BRCA1 (or a mutation in the related gene BRCA2). At the start of the study, all the women were healthy; 212 of them later chose to get risk-reducing mastectomies. Over the next few years, the researchers followed their progress. Sixteen percent of the women who didn’t get risk-reducing mastectomies developed breast cancer; of the women who went Jolie’s route, none did.

The initials in BRCA1 stand for breast cancer. Its name reflects how it was discovered: scientists found it as they were searching for the cause of the disease. But such names are really misnomers. After all, genes don’t simply sit in our DNA so that they can mutate in some people and make them sick. Normally, they have a job to do. In the case of BRCA1, there are many jobs. For one thing, it protects DNA from harmful mutations that can arise as it’s getting replicated. And if DNA does get damaged, the BRCA1 protein helps fix it. It joins together with several teams of other proteins, and each team carries out a different part of the complex task of DNA repair.

BRCA1, in other words, normally keeps our cells in good shape. If it mutates, though, it can’t do its jobs properly. Cells with a mutant copy of BRCA1 let mistakes slip by. Mutations in other genes can accumulate in a line of dividing cells. Some of those mutations will cause cells to die, but sometimes they have the opposite effect: the mutant cells grow and divide rapidly. As they proliferate, they accumulate even more mutations, eventually becoming full-blown cancer. As a result, women who carry BRCA mutations have a 40 to 85 percent risk of developing breast cancer during their lifetime. (They also run a 16 to 64 percent risk of ovarian cancer.)

Last year, a team of scientists at the University of Utah discovered an unexpected side effect of BRCA mutations. They looked at medical records of women who carried BRCA mutations and compared them to women with a normal version of the genes. The scientists found that women with the mutations weren’t just more likely to develop cancer. They also had more children. The effect was particularly strong among women born before 1930: they had, on average, two additional children (6.22 compared to 4.19).

The Utah scientists couldn’t say from their study how the mutations could lead to more children. But they offered one suggestion. A woman’s fertility depends on the viability of her eggs. Like other cells, eggs have caps called telomeres on the ends of their chromosomes that keep them from getting damaged. The longer the telomeres, the better shape an egg is in. Among its many jobs, BRCA1 helps control the length of telomeres. The Utah scientists suggest that mutant BRCA1 proteins may lengthen the telomeres in eggs, keeping them more viable.

BRCA mutations are so good for fertility that Jack da Silva, a biologist at the University of Adelaide,  has pointed out that they should be a lot more common. Only a few percent of women carry them, but they enable women to have so many more children, you’d expect the mutation to become more common with each generation. Within a few centuries of the mutation first appearing, everyone should have it.

Da Silva proposes that the mutations hang in an evolutionary balance. Before the age of 40, a woman with a BRCA mutation only has a 20% chance of developing breast cancer. That risk rises to 37 percent by the age of 50 and continues going up; by the age of 70, it’s 70 percent. In other words, these women have good odds of surviving to the point at which they can give birth to children, but they’re less likely to be alive to see their own children become parents.

A number of biologists have argued that grandmothers have played an important part in the survival of their grandchildren. In fact, some maintain that their help is so valuable that menopause evolved as a result. Women who stop raising their own children can better channel their efforts into helping to raise their grandchildren. Women with BRCA mutations are less likely to be able to provide that help. As a result, Da Silva suggests, the odds of their grandchildren surviving may have been somewhat lower than for grandmothers without the mutations.

This balance could account for the puzzling nature of BRCA mutations. They’re more common than other potentially fatal disorders like cystic fibrosis, and they’re unable to spread to more than a few percent of the population. It is this ancient legacy of BRCA’s many effects that women like Jolie are grappling with today.

[Update 5/14: I corrected the cause of Angelina Jolie’s mother’s death from breast to ovarian cancer.]

10 thoughts on “Tracing Breast Cancer’s History

  1. I was raised somewhat by a grandmother. We did a lot of things together. She did not have breast cancer and neither does my mother.
    I feel in pretty good shape I am 63. I am sorry for Angeline Jolie and think as others she is a very brave person. Blessings to her.!

  2. Ms Angelina jolie is a woman on with special powers, showing the world how to live with strong heart. Love.

  3. “you’d expect the mutation to become more common with each generation. Within a few centuries of the mutation first appearing, everyone should have it.”

    So my understanding is that homozygotes for BRCA1 mutations are thought to have low fitness (e.g. BRCA1 homozygotes are embryonic lethal in mice ). That would explain why a BRCA1 mutant allele couldn’t become fixed in the population. Or am I missing something?

    At best a mutation with fertility advantage for heterozygous individuals, but serious problems for homozygotes (as potentially here) could be a heterozygous advantage balanced polymorphism. However, I guess I’m also a little skeptical of the fertility boost in heterozygotes. Larger studies have not found any effect:
    Although that could be because they are in the modern populations, with access to contraceptives etc. This first study is looking at women who had their children before 1970 to side step that problem. However, in doing so the method of choosing cases and controls is (necessarily) in this study is somewhat complex due to the fact that you need individuals with and without BRCA1 but you have the assay this by their descendants. It would be good to see this replicated in a larger study in a contemporary population.

  4. Graham Coop raises two good points. If BRCA1 mutations are homozygous lethal in humans, as they are in mice, then the mutations will not fix but may exist at some intermediate frequency at equilibrium. If BRCA1/2 mutations in heterozygotes increased fitness by 50% then these mutations would have a frequency of 25% at equilibrium. However, this is still two orders of magnitude higher than the observed frequency of these mutations of about 0.23%.

    The other point is that the method used to estimate the nearly 50% higher fertility associated with BRCA1/2 mutations depends on sampling from pedigrees. This is bound to bias estimates of fertility upward because the ancestors of sampled descendants will have higher fertilities on average than individuals that did not leave descendants in the generation being sampled. Ignoring the estimate from this method and using only data from descendants themselves still gives an 18% higher fertility associated with BRCA1/2 mutations. In this case, BRCA1/2 mutations should have a frequency of 13% at equilibrium – still much higher than the observed frequency. This will be an important result if it can be replicated.

  5. Sorry to hear about encroachment of cancer to Jolie. But life is life. My wife also got breast cancer in 2008 and had masectomy of right breast after 8 course of Chemotherapy. Then operation an dget it removed. 12 application of Radiotherapy an dsecond time suspicious operation again. But by the good grace of god and modern achievement of science and honest an dwonderful treatment of British Doctors team saved her life and she is still with me alive . Claims some time weakness and heaviness of the body. She is normal thesedays simple complaints. But those horrible days and bodily pains after medication and falls of hairs and very difficult to swallow the foods were painful.
    Jolie what she decided beforehand is praise worthy. A stich in time saves nine. She is brave I share this message to my wife, she will be happy.
    One most important thing which I would like to addhere that courage is very ointment treatment in cancer. The faith and belief of the patient helps to get care soon.

    Best regards,

    Deepak Chalise

  6. Hi Jack,
    Thanks for your response. I’ll add one more comment, which I think is clear to you, but flesh out on the population genetics side of things.

    I’d done rough the same balanced polymorphism calculations and come to similar conclusions. It seems likely that the BCRA1 mutation homozygote has low fitness in humans because BCRA1 is a key DNA repair gene [even ignoring the mouse results].

    If BRCA1 mutant alleles had no fitness consequences in heterozygotes, but was fatal in a homozygous state, then the allele would be maintained at 0.4% simply by the constant influx of mutations (mutation-selection balance of sqrt(1.65e-5/1)). That’s not far off its observed frequency, and I’m sure you’ve done that calculation. So putting aside the recent report of a fitness advantage to heterozygotes the data and theory are reasonably in accord.
    To explain the frequency of the allele under mutation selection balance we only need heterozygotes to have ~2% less children than the homozygote with two working copies. Before the increased fertility result this cost seemed quite plausible [even for a usually late onset disease, as there are some early incidences].

    In light of the higher fertility result, if it is true, obviously there is a need for a more complex explanation. As I understand your paper you are arguing that even if there is an initial fertility benefit, the grandmother effect might drop the relative inclusive fitness of the heterozygote below that of the homozygote with two working copies of the gene. This then allows the allele to be at mutation-selection balance, where mutation is keeping the allele in the population even though it has lower inclusive fitness. Is that right?

    [I note that I’m still a little skeptical about the BCRA1 increased fertility result. The sample sizes are not very large, so it would be good to see it replicated. ]

  7. PS my interest in this, apart from my general interest in human popgen, is that I use BCRA1 as an example of mutation-selection balance for my undergraduate class. So I want to make sure I understand the details here.

  8. Hi Graham,

    I suggested that if fertilities were limited in hunter-gatherer populations, then a carrier (heterozygote) would not have experienced the maximal increase in fertility possible without help due to the mutation, and would still have fewer additional children due to help than a non-carrier (non-mutated homozygote). In other words, help from a grandmother would give non-carriers higher inclusive fitness only if fertilities were limited.

    So, yes, with limited fertilities carriers would have lower inclusive fitness than non-carriers, and the mutation would be maintained at mutation-selection balance.

  9. As I have noticed that the D’Auria Family Tree on my paternal side, and the De Moss Family Tree on my maternal side don’t have the family history of any of cancer like breast cancer and ovarian cancer, I am 63 years old then I am the Second-Generation Italian American who is Jewish in my D’Auria Family Tree on my strong paternal side. I know that Magna Greica is covered on Central and Southern Italy since the Greek Civilization started to rule before the Roman Empire ruled on most of Europe, Northern Africa, and Middle East in the history in a fact. My ancestors came from Italy as One-Counting European Ancestry as I finally found out so recently.

Leave a Reply

Your email address will not be published. Required fields are marked *