Genes and Cancer
Leafing through Molly Fitzpatrick's bulky family albums, one is struck immediately by the strong family resemblance. From grandmother to mother to daughters, from summer picnics to Christmas mornings, photo after photo has captured sparkling eyes, ivory skin, auburn hair, mischievous Irish smiles, and natural grace. The hundreds of photos, some almost a century old, record the rise of a family started by penniless immigrants whose descendants now enjoy the comforts of middle class life in the United States. But albums tend to record the good times, and candids cannot tell us what is going on inside. Only when Molly paused and touched a photo here and there and wistfully reminisced did I begin to understand the burden that her family has carried.
Over the last four generations, seven women on Molly's mother's side of the family have developed breast cancer and two have died of ovarian cancer. Of these nine women, only one has yet lived past 60 and most have died in their 50s. There are only two survivors, Molly's mother, Elizabeth, and her sister, Kathleen. After watching her mother and an older sister die, Kathleen, now 70, had her breasts surgically removed. At the time, more than 20 years ago, prophylactic mastectomy was an extremely unusual operation, and Kathleen had fought hard to convince the surgeon to operate. Elizabeth found her lump when it was small; she had her mastectomy 15 years ago.
Each year in the United States more than 150,000 women are diagnosed with breast cancer, and about 20,000 learn that they have ovarian cancer. During the year, 40,000 women will die with breast cancer and more than 10,000 will die from ovarian cancer. The lifetime risk of developing breast cancer is about 1 in 9, but it is important to remember that most cases occur in older women. In the general population, the risk of developing breast cancer by age 50 is "only" about 2%. The lifetime risk of developing ovarian cancer is about 1 in 65, and most cases arise in women over age 60. About 5-10% of the women who develop either of these cancers have a strong family history for one or both of them.
Physicians began writing about families with unusually high numbers of cancers more than a century ago. For decades such reports were treated as little more than medical curiosities, but as life expectancy soared (since 1900 the average life expectancy for white women born in the United States has climbed from 45 to 78, and it continues to rise) and we learned more about the diseases of old age, and our knowledge of pathology and genetics grew, the evidence that a significant fraction of breast cancer (and many other cancers) occurred in people who were born with a strong genetic predisposition mounted.
At the molecular level, every cancer is a genetic disease. We know this thanks to the work of many scientists over the last 25 years who have discovered and elucidated two classes of genes known as oncogenes and tumor suppressor genes. The several dozen oncogenes code for proteins that interact at key positions on the surfaces of cells and instruct them to grow and divide. They trigger a cascade of molecular messengers (a sort of biochemical pony express) that travel to the cell's nucleus and order certain genes to make the cell replicate. When they are normally functioning, oncogenes act to orchestrate the diverse aspects of human growth, the remodeling of injured tissue, and the replacement of dying cells. When oncogenes behave abnormally, the growth factors that they produce can cause massive overgrowth of cells, the essential feature of cancer. Signals to overproduce cells are, in most cases, not enough to cause cancer. To become the founder cell in a cancer, a cell must also overcome the forces of the tumor suppressor genes, proteins that normally act as restraints on growth.
Cancer arises when a single cell accumulates a sufficient number of mutations in several different genes that are each in some way responsible for regulating the cell cycle—the timing and pace of cell division. When these mutations significantly cripple the system that regulates that cycle, the cell may begin to replicate wildly. As further genetic changes occur, some progeny cells become so misprogrammed that they metastasize (break away and migrate to other parts of the body). This puts the individual at high risk of dying from metastatic cancer.
In most cases, the mutations in the oncogenes and the tumor suppressor genes (several dozen of which have now been identified) that lead to cancer were not present in the DNA with which the individual started life. They are somatic mutations that occur in cells in the various tissues over the course of one's life. Many cells in the body, especially those that are exposed to the outside world, such as those in lung and colon, must regularly divide to replace those that die off in the normal course of events. When they divide they pass on a copy of their DNA—including any newly acquired mutation—to the two daughter cells, which in due course may pass them on to their progeny. A few of these billions of daughter cells (there are on the order of 10 trillion cells in the human body) will in turn acquire additional mutations to other genes. Some of them can then pass on a bigger mutational load to their progeny. If a cell emerges that has inherited by chance just the right set of mutations, it has the potential to be the founder of a population of tumor cells. For some cancers we now know in exquisite detail the pathway to cancer. Among the most elegant work is that on the origin of colon cancer. During the early 1990s, a team led by Dr. Bert Vogelstein at Johns Hopkins University School of Medicine discovered which genes must mutate for colon cancer to arise.
No one understands exactly why most somatic mutations occur, but we have some good hypotheses. Certainly, as in lung cancer, there is overwhelming evidence that direct, repeated exposure to chemicals in cigarette smoke eventually cripples the key genes. There is also moderately impressive epidemiological evidence that a high-fat diet is a risk factor for cancer of the colon and, to a lesser extent, the breast. In fact, some of the nation's leading cancer epidemiologists think that smoking and a high-fat diet account for nearly two-thirds of all (non-skin) cancers in the United States. Given that percentage, none of the other major suspected causes actually constitutes a very large slice of the pie. For example, despite all the attention it has received, the low-level radiation to which we are all chronically exposed is thought to cause only about 2% of all cancer deaths.
Despite much concern about exposure to chemicals in the workplace, the epidemiological evidence suggests that it is a source of relatively few cancers, again probably only about 2% of the total. This is almost certainly much lower than was the case during the 19th and early 20th centuries. In fact, among the strongest evidence that chemical exposures in the workplace cause cancer is an observation made nearly two centuries ago. In the early 19th century, Sir Percival Potts, a British physician, noted that young men who worked as chimney sweeps often developed cancer of the scrotum. The cause was prolonged exposure of the skin to coal tars. Another well documented example of occupational cancer is the hugely increased rate of thyroid cancer among workers who were exposed to radium as they manufactured luminescent dials for watch faces (a practice now long abandoned). Today, thanks to better containment systems, the hundreds of dangerous chemicals that can be found in the workplaces of modern industrial society probably account for many fewer cancers than they did just 50 years ago. In poorer nations, which have much more dangerous workplaces, the risk of occupational cancer is probably much higher.
Most hereditary cancers, like those that have occurred in Molly Fitz-patrick's family, arise when several somatic mutations (changes in the DNA in cells in a particular organ) accumulate in the cells of an individual who was born with a germ-line mutation in a tumor suppressor gene. Because the germ-line change was present from the moment of conception, it is in the DNA of all cells in the body. In essence, people with germline mutations start life one big step closer to cancer than do those not so burdened. That is why one of the classic hallmarks of hereditary cancers is that they tend to strike at a younger age. Indeed, two of the cancers through the study of which the concept of tumor suppressor genes was in part elucidated, retinoblastoma (a tumor of the eye) and Wilm's tumor of the kidney, are childhood cancers.
In 1989 a team led by Mary Claire King, a scientist then at Berkeley, California, after painstakingly studying many families like the Fitzpatrick family, found convincing evidence that there is a gene on the long arm of chromosome 17 which, if mutated, greatly increases the lifetime risk for breast and ovarian cancer. The news set off a spirited race to find and clone the gene, a race that was won in 1994 by a team led by Mark Skolnick at Myriad Genetics, at the time a fledgling biotech company in Salt Lake City. Since then, the same research strategy has led to the discovery of a second gene predisposing to breast cancer, and in coming years it is likely that one or two more will be uncovered. Together, the two genes that have already been cloned, known as BRCA1 and BRCA2, account for much more than half of all hereditary breast and ovarian cancer. One of the curious facts about molecular genetics is that scientists now routinely find culprit genes before they have any idea about the gene's role in the life of a cell. We do not yet know much about what the BRCA1 and BRCA2 proteins do, although those secrets will soon be decoded.
Although the data are still being compiled, it appears clear that carriers of a harmful BRCA1 mutation have a 50-70% lifetime risk of developing breast cancer and a 30-50% risk of doing so by age 50. That is, their risk for early breast cancer is roughly 15 to 20 times higher than that of the average woman. Perhaps even more significant, but less discussed, is that their risk of ovarian cancer is also much higher than that of the general population. The lifetime risk for ovarian cancer in BRCA1 carriers is probably as high as 1 in 6, about 10-15 times higher than the background risk. It is crucial to inform women who do carry one of the harmful mutations about the risk of ovarian cancer. Indeed, a good case can be made that it is an even more important risk to discuss than is the risk of breast cancer. Most women are already highly aware that breast cancer is common, that there are simple methods (self-examination and periodic mammography) to detect most cases early, and that it is a highly treatable disease with steadily improving survival times. In general, women know far less about their risk of ovarian cancer. Furthermore, it is highly likely, although not yet absolutely proven, that in women known to carry a BRCA1 mutation, the removal of the ovaries greatly reduces the future risk of ovarian cancer. The surgery does not completely eliminate the risk because 5-10% of ovarian cancer arises from biologically similar cells that line the nearby pelvic wall which cannot be removed with the ovaries.
In 1996 scientists at Myriad Genetics launched the world's first laboratory test that reads the precise DNA sequence of a particular gene. Using highly automated techniques, the scientists at Myriad extract DNA from blood samples of at-risk individuals and examine about 16,500 DNA letters in the coding sequences of these two genes, searching for changes in even a single letter. This astounding new test (which on a single sample analyzes more DNA letters than all the laboratories in the world were analyzing in the mid-1970s) has created new hope and profound uncertainty for people like Molly Fitzpatrick. It also greatly exacerbated an often bitter policy debate about whether such a test was in the patients' best interests. Molly is among the first to face a dilemma that millions of us will experience in coming years. Should she find out if she carries a mutation that predisposes her to cancer?
Molly's mother, Elizabeth, developed breast cancer in 1982 at age 48.
Yet, Elizabeth considers herself lucky on four counts. First, her mother (Molly's grandmother) was diagnosed with breast cancer at 38 and died when she was 39, so Elizabeth thinks of herself as having eluded the disease for 11 years longer than did her mother. Second, because she regularly and meticulously examined her breasts, Elizabeth found her lump when it was small and she is doing well more than 15 years after her breast surgery. Third, she has not developed a second breast cancer or ovarian cancer, for both of which she is at high risk. She recently had her ovaries removed.
The fourth reason that Elizabeth considers herself lucky is that she knows she has a mutation in her BRCA1 gene. By becoming one of the first people to undergo the new DNA test, she converted a deeply held family suspicion to a fact. Elizabeth told me that she underwent the test for two reasons. For her the less important is that if she tested positive, it would provide a powerful rationale for having her ovaries removed. The more important is that she wanted to provide better information to her two daughters and her son about the chances that they could have inherited a gene that confers a significant cancer risk on them. Even though the absolute risk to her son remains low, men with BRCA1 mutations are at much higher risk for breast cancer than are other men. Early studies suggest that they may be at higher risk for prostate cancer as well. Because Elizabeth has a BRCA1 mutation there is a 1 in 2 chance that she passed it on to each of her children.
The fact that men are just as likely to carry BRCA1 and BRCA2 mutations and transmit them to their children as are women creates a clinically deceptive situation. In more than a few instances, women who turn out to have hereditary breast or ovarian cancer will be the first in their families to be afflicted. In such cases, the mutation may have passed silently through the male line. It could have arisen four generations ago in the sperm cell that started the life of a male fetus. In time the young man who was the product of that conception married and passed the mutation to both his sons. During the next generation perhaps each son, neither of whom developed cancer, had three children—two sons and a daughter. It would be quite plausible that only one of the two granddaughters inherited the mutation, which caused her to develop breast cancer when she was, say, 52. When that woman recounted her family history for her doctor, she would disclose nothing that would suggest that she was born with a hereditary predisposition to breast cancer.
Scientists were able to find the BRCA1 and BRCA2 genes by studying unusually large families in which many people had been diagnosed with breast cancer. These represent the tip of an iceberg of unknown dimensions. As we learn more, it will probably turn out that the genetically based risk for breast and other cancers is somewhat higher than we now think, but that the risk is spread over many different genes, some of which contribute only a small component to the overall predisposition. The role of still poorly understood environmental factors in breast cancer is and always will be high.
Her mother's decision to be among the first people to undergo a DNA test for cancer risk has had a huge impact on Molly. The positive result means that she has a 1 in 2 chance of having inherited a gene that (almost certainly) caused her mother's breast cancer. She could find out at any time if at conception she won or lost that particular toss of the genetic coin. Molly has agonized over whether or not to be tested. On the one hand, the possibility of learning that she won the toss and did not inherit the chromosome with BRCA1 mutation has tremendous allure. If she is not a carrier, many of the fears that trouble her would vanish. She would no longer be terrified about dying young of ovarian cancer as did one of her aunts, she would stop wondering if she should have her healthy breasts removed to reduce her risk of cancer, and she would stop worrying about whether she will transmit a predisposing cancer gene to her children when she has them.
What if she took the test and the results showed that she did inherit the mutation from her mother? All those worries that she is usually able to keep at bay may mushroom. Old questions may take on new urgency. Should she consider mastectomy? There is some fairly persuasive evidence that it saves the lives of BRCA1 or 2 carriers. In 1996 Lynn Hartmann, a researcher at the Mayo Clinic, published her findings from a study of women who, in addition to having a breast removed to treat an existing cancer, also had the other removed to avoid a future cancer. In comparing them to similar women who did not have the second surgery, Dr. Hartmann found fewer deaths from breast cancer in the group that had both breasts removed. This was the first evidence that prophylactic surgery saves lives, but as it was a retrospective study, it left some scientists unconvinced.
Imagine the difficulty of deciding whether or not to undergo prophylactic surgery. Currently, few women request this option and few surgeons recommend it, largely because they believe that by carefully following women at high risk they can find a breast cancer so early that it will be highly curable. On the other hand, a growing number of medical centers are offering surgery to remove the ovaries of women who carry a deleterious BRCA1 mutation because it is not yet possible to monitor such women to catch ovarian cancer early enough to cure it. Molly, who is 30 and soon to marry, hopes to have two children over the next four years. Assuming no major breakthroughs in chemoprevention that would allow her to avoid surgery, she currently plans to have her ovaries removed before she is 40.
The hunt for genes that may predispose to breast and ovarian cancer has explained an old and puzzling observation. Ashkenazi Jewish women (about 90% of the Jews in the United States are of Ashkenazi descent) have an above average risk for breast cancer. For years this increased risk was attributed to unknown factors in diet or lifestyle. We now know that the reason is largely genetic. Two mutations in the BRCA1 and one mutation in the BRCA2 gene are unusually common among the Ashkenazim. Overall, an Ashkenazi Jew has about a 1 in 40 chance of being born with one of these mutations. The high (from a geneticist's point of view) prevalence of these mutations in this population is probably due to a combination of three factors. First, because the Ashkenazim constitute a relatively small population group whose members have through most of history tended to marry other members (a pattern that geneticists call endogamy), it is likely that some mutations that appeared by chance were propagated. Second, since these mutations, which usually cause illness well after the individual has had children, do not have any negative effect on reproductive fitness (the likelihood that each person has of passing on his or her genes), they will not tend to drain out of the gene pool. Third, because the three mutations are so common, there is even a chance that they may confer some (as yet completely unknown) benefit during the first four decades of life.
Because only three mutations explain most of the hereditary risk for breast cancer in the Ashkenazi population, it has been technically easy to launch a low-cost DNA-based test for Ashkenazi Jewish women. So far, however, relatively few people have made use of it. Because there is no simple medical intervention, such as taking a medication with few side effects, that will reduce the risk faced by carriers, and because few women want to undergo bilateral mastectomy, many Jewish women are debating whether learning that they do carry one of the three mutations will be helpful. Since breast cancer is a risk faced by all women and since established surveillance techniques will, if carefully adhered to, catch most breast cancers at a relatively early stage, why not just take monitoring seriously? This might well work, but it does not solve the problem that being a BRCA1 carrier also confers a significantly increased risk for ovarian cancer for which there is not a good surveillance system. Ovarian cancer is an insidious disease that in the vast majority of cases is only diagnosed in an advanced stage, and is usually fatal in less than five years. If, like Molly, one has a significant family history of breast cancer and one can learn whether her risk for that disorder and for ovarian cancer is or is not elevated, the information is potentially of great value.
Another major worry articulated by many women with a family history of breast cancer (and a major concern of the National Action Plan on Breast Cancer) is that when a healthy individual learns that she carries a mutation that predisposes to breast cancer, she and her relatives will have trouble obtaining health and life insurance. Worries about genetic discrimination are understandable. Genetic ideas were misused in the United States for much of the first half of this century, leading to the involuntary sterilization of about 60,000 persons with mental retardation and to national immigration quotas keyed to notions of relative racial superiority (see Chapter 24). Socioeconomic discrimination on the basis of race and gender (both genetically determined) remains widespread today. People with cancer have traditionally had great difficulty in finding affordable health insurance and life insurance.
What evidence is there that healthy people in the United States are being denied access to health insurance or charged higher premiums because of a genetic test that indicates an increased risk of a serious disease or of bearing a child with a genetic disorder? In a word—almost none. In 1992 and again in 1996, a Boston-based group of scientists published papers based on surveys of families at risk for various genetic disorders which purported to show that a significant fraction had been denied health insurance. But, because the authors did not contact the insurers to study their side of the story, it is impossible to verify the claims of the consumers. The 1992 paper generated a lot of media attention and public interest. Since then I have frequently sought firm evidence of genetic discrimination in health insurance underwriting and have repeatedly come up empty. The insurance executives with whom I have spoken deny any such practice. They point out that (1) almost all health insurance is group-written so that the health of any single person or family is largely irrelevant, (2) the cost of asking people to take genetic tests as part of their application is far too expensive (for now) to justify its use, and (3) it would be legal and political suicide to be caught engaging in genetic discrimination.
The era of DNA-based testing to find out whether one is at special risk for breast or other cancers is just beginning. Thus far, there is little evidence to suggest that persons who test positive for BRCA1 or BRCA2 mutations or for other genes that cause risk for other cancers have had trouble obtaining health care coverage. Indeed, in the only study conducted of a large cohort of women who had undergone the BRCA1/2 test offered by Myriad, the vast majority indicated that they were glad that they had taken the test and had not suffered any economic discrimination as a consequence of so doing. This work was done by researchers who were not associated with the company. Most Americans now obtain health care coverage either through federally funded programs such as Medicare or Medicaid or through employer-based group health insurance. On August 1, 1997, the federal Health Insurance Portability and Accountability Act (HIPAA) went into effect. That law expressly forbids group health insurers from defining predictive genetic information as a "preexisting condition" (which would give the insurers a right to deny coverage for a particular condition for up to 18 months). In the summer of 1997, President Clinton put genetic discrimination on the national agenda when he came out strongly in favor of laws to prevent it. Several bills that would, if enacted, broaden the protections provided by HIPAA are now before the Congress. The protections offered by HIPAA as well as a steadily growing number of state laws (at the close of 1999 there were at least 30) that forbid health insurers from using predictive genetic information in underwriting, suggest to me that our society has done a pretty good job of heading off a possible new form of discrimination.
Nevertheless, a new problem has emerged. Advocacy groups, bioethi-cists, lawyers, and some leading scientists have been so effective in pointing out the potential threat of genetic discrimination that many people who are members of families with a history suggesting that they might carry mutations like BRCA1 or BRCA2 are telling their physicians that fear of discrimination is the main reason they will not take the predictive test. Even though there is little direct evidence to justify that fear, if the perception cannot be changed, it defines reality. Hopefully, over the next two or three years people will be gradually reassured by the enactment of laws to ameliorate this fear. If not, we may face the tragic situation of people refusing to undergo DNA tests that could eventually lead to life-saving strategies in the management of cancer risk.
(Left to right) Alois Alzheimer, Emil Kraepelin, Robert Gaupp, and Franz Nissl.
(Reprinted from Pollen 1993 [photo from Dr. Paul Hoff, Psychiatric Hospital of the Ludwig-Maximilians-University, Munich].)
Was this article helpful?