The Hunt for a Breast Cancer Gene

When breast cancer struck her family last year, Susan M., as she wants to be called, was no stranger to the disease. Since 1978, she had watched it strike her mother at 46 and kill her oldest sister at 38. Two cousins had also died in their 30s. So when she learned that her surviving sister Janet (also a pseudonym) had been stricken at 41, Susan did what any rational person would do. She panicked. "I knew I wanted to have my breasts off," the blond, athletic 36-year-old recalls. "I was so afraid I was going to get cancer before I could get them off." Though the benefits of such surgery are unproven, Janet supported the decision. "I felt like to save our family, all girls had to get their breasts off," she says. "The feeling is, 'You can cut anything off. I just want to live'."

Though she had no way of knowing it at the time, Susan was in no such mortal danger. Researchers at the University of Michigan suspected that a deadly gene was worming its way through the M. family tree, and they had devised an indirect method of identifying the carriers through blood analysis. Susan had already scheduled a double mastectomy in August 1992 when she learned her test results. "It never occurred to me that I wouldn't have the marker," she says. But the findings were clear. Dr. Barbara Weber, an oncologist and molecular biologist, assured her that she had almost certainly escaped the family curse, and Susan canceled the operation. Today she's perfectly healthy.

Breast cancer is notoriously difficult to predict, but family history is one of the few clear risk factors. Of the 180,000 cases diagnosed in the United States each year, as many as 10 percent stem from hereditary defects. Experts now believe that at least half of these inherited cases involve flaws in a single gene, which they've dubbed BRCA1 (for Breast Cancer 1). Though a laboratory can sometimes concoct a "marker" test to chart the gene's movement through a family like the M.'s, BRCA1 itself has not been isolated. As a result, few of the million or so Americans with dangerous mutations have a good way of knowing who they are. If researchers can zero in on BRCA1 within the next year, as expected, simple blood tests may soon enable any woman with stricken relatives to gauge her own risk. And by illuminating the mechanics of familial breast cancer, the discovery could expand our general understanding of the disease.

Scientists are thrilled by the prospect; James Watson, one of the founders of modern genetics, has called BRCA1 the most exciting quarry in medical science. It heralds an era in which gene tests will help predict a range of chronic ailments, from high blood pressure to Alzheimer's disease. But even enthusiasts worry about the coming boom in genetic fortunetelling. Genetic tests are imperfect predictors of health, and there's no guarantee that widespread testing will improve people's lives. Knowing where the gene is won't lead immediately to better treatments, and women deemed vulnerable will have more than the illness to fear. Under the current system of health insurance, they could find themselves ineligible for coverage. As the epidemiologist Hoda Anton-Culver observes, "Finding the gene isn't the end of the challenge. it's the beginning."

THE search for BRCA1 is now a fiercely competitive marathon involving some of the world's sharpest gene hunters. But for nearly two decades, it was the quiet pursuit of a California researcher named Mary-Claire King. King, a 47-year-old geneticist at the University of California, Berkeley, set off the race three years ago by showing for the first time that familial breast cancer could be tied to a single gene. Thanks to that discovery, she now finds herself balancing the rigors of the lab with those of popular stardom. Unlike most molecular scientists, she has joined Jodie Foster and Anita Hill as one of Glamour magazine's "Women of the Year." "Sometimes," she muses, "I think I'll wake up and find it's just the five of us in this lab again."

King spent the early 1970s in Chile, married to a zoologist and planning to stay indefinitely. After the Allende regime fell, she found herself in San Francisco "with the unformed idea that I wanted to apply basic science to big [practical] questions." By the time she started studying breast cancer, doctors had long noticed that some families suffer more than their share of the illness. Studies suggested that any woman with an affected mother or sister was at high risk herself. The reasons were unclear, but King thought heredity offered the likeliest explanation. "Familial patterns can result from people in the same family having the same occupational exposure," she says, "but for breast cancer there was nothing like that." Indeed, the disease seemed to follow a "dominant" pattern of inheritance, in which anyone born to a carrier has a 50 percent chance of inheriting the trait.

The challenge was to find the relevant gene. It was well known that heredity is governed by the DNA housed in the nucleus of each cell. An earlier generation of scientists had learned that human cells contain 3 billion units (base pairs) of DNA arranged in linear sequence along 23 pairs of chromosomes. These long chains were known to contain the 50,000 to 100,000 functional segments, or genes, required to construct and operate the body. Unfortunately, no one knew the first thing about locating an unknown gene within this vast apparatus. So King focused initially on patterns of illness in the population. Her studies yielded small insights, confirming for instance that both men and women can pass the tendency on to their offspring, even though women are the only ones affected. But the evidence for a breast-cancer gene was still circumstantial.

Everything changed in the early '80s, after David Botstein, a biologist then at MIT, conceived a new technique for analyzing genetic material. In test-tube procedures, genetic engineers had begun using chemical scissors known as restriction enzymes to snip particular chromosomes at designated points. Applied to identical chromosomes, a restriction enzyme produces identical fragments. But if the chromosomes of two family members differ within the relevant region, the enzyme creates fragments of different lengths. Botstein suspected that these variations, known as RFLPs ("riflips"), might sometimes correspond to medically important traits. If researchers could link a hereditary illness to a RFLP pattern, he reasoned, they would learn roughly where the responsible gene lay, and they'd gain a too] for identifying carriers.

Armed with restriction enzymes, the Berkeley group collected blood from breast-cancer families and started searching for patterns that would distinguish the stricken members. Over several years, the team analyzed 182 RFLP markers and got nowhere. But in the summer of 1990, a 183d marker produced tantalizing results. In some of the 23 families under study, the breast-cancer sufferers seemed to lack a small part of one copy of chromosome 17, yet in other families no such pattern emerged. The researchers were stymied. Finally one morning, King's colleague Beth Newman proposed organizing all the families by age of onset. At that moment, everything fell into place. For each of seven families in which women had been stricken before 50, the abnormal marker was a powerful predictor of risk.

Geneticists had never shown much interest in familial breast cancer. Sorting truly hereditary cases from the myriad cases that occur by chance seemed an impossible task. And even if a hereditary syndrome existed, the experts said, it would probably involve complex interactions among numerous unknown genes. "My colleagues were very skeptical," King recalls, "and you know how skeptical boys can be. Scorn! Scorn! Scorn!" Everything changed when she laid out her data at a conference in Cincinnati that fall. She spoke at the end of an evening session. with a World Series game roaring at nearby Riverfront Stadium. Yet within 20 minutes she had a roomful of colleagues poring excitedly over her seven pedigree charts. In these families, premenopausal breast cancer was linked to a region of chromosome 17 just 50 million base pairs in length. Within weeks a team led by Dr. Gilbert Lenoir of the French-based International Agency for Research in Cancer started applying King's RFLP markers to other high-risk families to see if her results held up. They did. In January of 1991, Lenoir reported that his group had linked the new markers to early breast cancer in three families, and to ovarian cancer as well. The race for the gene was on.

A dozen labs in the United States, Canada and Europe have since leapt in, and they're charting remarkable progress. In studies of more than 200 high-risk families, researchers have found abnormal BRCA1 markers in more than 80 percent of the families prone to both breast and ovarian tumors and in half of those plagued by breast cancer alone. At the same time, they've narrowed the window around BRCA1 from 50 million base pairs to 300,000 base pairs, just 12 genes' worth of DNA. At first their main tactic was to study "recombination events," in which someone inherits only part of a known marker but still exhibits the related trait. But the suspected region is now so small that researchers are analyzing tiny segments one base pair at a time. "Your whole genome [genetic endowment] can be thought of as 23 sets of encyclopedias," says Dr. Francis Collins, the renowned gene hunter who codirects the University of Michigan's BRCA1 effort and heads the federal government's National Center for Human Genome Research. "We're down to part of one volume. What we're looking for is a misspelling."

Even without the gene itself in hand, researchers have learned a lot about its character. Because the affected members of a high-risk family lack a small, unknown portion of chromosome 17, logic suggests that the missing material belongs to a gene whose job is to keep tumors from developing. According to this model, most people start life with two copies of BRCA1 in every cell in the body, one from each parent. If a breast cell loses one copy, through wear and tear or the effect of some carcinogen, the second copy can still prevent uncontrolled cell growth. But not everyone gets that margin of safety. If a woman inherits a faulty copy of BRCA1 from one parent, damage to a cell's normal copy can have a devastating impact. In the families studied, the women with an abnormal BRCA1 marker have a 60 percent chance of developing breast cancer by the time they turn 50 (the usual rate is less than 2 percent), and an 80 percent chance of getting it by 65.

Unfortunately, these insights have yet to affect many lives. Though the family-linkage studies have helped scores of women assess their risk, linkage analysis is basically a research tool. To find a marker for one family, researchers have to analyze chromosomes from a number of members, sick and healthy, and seek out a pattern. As a result, says University of Utah geneticist Mark Skolnick, only one affected woman in 50 could learn her status from a linkage study. Consider the Cunninghams of Philadelphia. Three years ago Charlene Cunningham, then 25, found a lump in her left breast. Tests revealed a large tumor that had already seeded a nearby lymph node. Her three older sisters were terrified, but the family was too small to support a linkage study. ideally, researchers need five cancers, spread over more than one generation, to establish a pattern.

Charlene's sister Julie Maravich (pictured on the cover) was 26 at the time and healthy. But when a mammogram revealed that her mother also had the beginnings of a breast tumor, and a doctor guessed that her own odds were 50-50, she quickly settled on a double mastectomy. "I felt I had to address this immediately," she says. "I just couldn't deal with those odds." Neither could the two other sisters. Thirty-year-old Katie Mullins had her own breasts removed last year (both she and Maravich have had reconstructive surgery). And 38-year-old Dorothy McKenna is planning an operation next year. As they care for Charlene, whose cancer has invaded her hips and skull, the women can only wonder what's in store for their own children. "I would like to know," says Maravich, referring to her 5-year-old daughter, Alexandra. "I'd like to be able to plan."

Thirty-three-year-old Kris Glavaris, a Baltimore medical worker, has similar feelings. Breast cancer has already killed her mother (at 44) and one of her sisters (at 30). In 1989, she and a second sister were stricken. She's now hoping that a blood test will arrive before her 2-year-old daughter grows up. Most geneticists oppose testing children, but with BRCA1 in hand, the task would be simple. Instead of comparing family members' chromosomes, doctors could use standardized tests to check for common mutations in the gene. King predicts that if the gene is discovered next year, such tests will appear by 1995.

Besides making carrier testing easier, the discovery of BRCA1 might improve breast cancer screening in general. If doctors could periodically inject women with probes designed to bind and illuminate BRCA1 mutations within cells, a standard breast X-ray might detect trouble before tumors start to form. "It's what I call a molecular mammogram," King says. "It could be useful for any woman, whether she had an inherited susceptibility or not."

Scientists aren't optimistic about learning to fix or replace defective copies of the gene any time soon. You would have to remake millions of cells in the breast and ovaries to give a high-risk woman any protection. But the isolation of BRCA1 could help advance other kinds of therapy. If it turned out that the gene produced some anti-tumor protein, for instance, researchers might learn to synthesize it for use as a drug. "Understanding the gene may give us an important hand," says Weber. "But first we need to know what it does."

Until then, the capacity to test people for BRCA1 defects will be a very mixed blessing. Women who test negative will be spared some terror (though they'll still face the threat of nonhereditary breast cancer). But for those who learn they are carriers, the alternatives will be no less grim than they are today. As King wrote recently, "We do not have a good option...for women with inherited risk." Intensive mammography can help carriers spot tumors at early stages. Mammograms aren't normally recommended for women under 40, and experts differ strongly on regular screening for women under 50. But known carriers have much more to gain: mammography has already helped some detect tumors in their 20s. Unfortunately, the technique isn't always sufficient. If a young woman has particularly dense, fibrous breasts, it can produce false alarms while missing tumors.

The estrogen-blocking drug tamoxifen may someday provide a second weapon. Researchers have noticed that when tamoxifen is used to treat tumors in one breast, it seems to keep them from occurring in the other. Last year the government launched a huge study to see whether the drug could help prevent first tumors as well. British researchers are conducting a similar trial. Unfortunately the drug's side effects can range from hot flashes to depression, and until the results are in there is no reason to assume it will work.

Given such meager choices, some highrisk women are opting to have their healthy breasts removed. Some doctors now advocate mastectomy for any woman who sees two immediate family members develop breast cancer before menopause. No one knows how widespread the practice is, but in a survey conducted last year, Dr. Kathy Helzlsouer of the Johns Hopkins School of Hygiene and Public Health found that 90 out of 700 Maryland surgeons had performed at least one preventive mastectomy. BRCA1 testing could help confine the trend to those at greatest risk, but many experts fear that surgery will become standard treatment for anyone found to harbor a mutation. That would be inappropriate, they say, for no one knows how much good it does. Even a total mastectomy is likely to leave some breast cells behind.

Medical questions aside, the new tools for breast-cancer prediction could foster countless psychic dilemmas, for people don't face familial diseases by themselves. Kay Dickersin, an epidemiologist at the University of Maryland, was just 34 when she developed breast cancer seven years ago. She had no sense that she belonged to a high-risk family. Today there's little doubt. Two of her three younger sisters have been hit in their 30s (one of them twice), and a third is waiting for the King lab to determine whether she shares their predisposition. Dickersin has accepted her own lot, but she worries about the legacy her teenage sons may have to live with. Several years ago, as she tried to explain the hereditary aspect of the illness, one of them asked, "Does this mean my daughter will get breast cancer?" If he carries the gene, it may. But he doesn't know his status. "Should he be tested before starting a family?" she asks. "And if he does [carry the gene], what are the implications for his own relationships?"

Janet M., the Michigan woman who saw her sister spared a mastectomy, is now watching her own teenage daughter confront those issues. At 19, Ann (a pseudonym) knows she's a carrier. Because BRCA1 mutations can trigger both breast and ovarian cancer, she feels she should have both organs removed by the time she's 25. But she doesn't want to miss out on having, and breast-feeding, children. Her mother has a more mundane concern. "We feel we have to be very, very careful," she says. "I know that when Ann is 25, she's going to have a $50,000 to $75,000 operation. If that becomes public, who's going to want to insure her? Who's going to want to hire her?" Ethicists have worried for years that genetic knowledge would breed new forms of discrimination. The imminent discovery of BRCA1 makes the hazards more vivid than ever. Fortunately, the possibilities are just as vivid.