The Forecast for Genetic Fortune-Telling
Cloudy but Gradually Clearing

by M. Stephen Meyn

(Posted July 25, 1997 · Issue 13; archived August 15, 1997)

Epidemiologic studies have predicted that certain cancers, including some breast cancers, have an inherited genetic basis. Molecular genetic tools are now becoming available to screen for specific mutations in individuals in high-risk families. At the same time, these tools help to improve the estimation of risk. Sometimes better estimations indicate lower risks than previously believed. But confusion is created when medical science reports changes in estimated risk. Many clinicians, as well as much of the general public, now wonder what to do with the information derived from molecular genetic screening for breast cancer risk.

The current article outlines some of the problems created by the emerging studies. The second article will deal with challenges of designing and executing epidemiologic studies to address questions of risk and potential prophylactic treatment. Finally, we will present the opinions of a genetic counselor who must deal directly with the potential recipients of this new knowledge.

Men at some times are masters of their fates.
The fault, dear Brutus, is not in our stars,
But in ourselves. . . .

- Julius Caesar, Act I, Scene II.

At the dawn of a new millennium, we think that we know where the fault lies: in our genes. In the past, genetics had little to offer mainstream medicine. Clinical genetics was an obscure pediatrics subspecialty that tried to make sense of tragic, but rare, accidents of nature. Now, armed with advanced biotechnology and the fruits of the Human Genome Project, geneticists promise to help diagnose and treat common adult-onset diseases. One major attempt to apply the new genetics to clinical practice has focused on familial breast cancer. Recent developments in the translation of genetic analysis to epidemiology offer a glimpse at how far we have come and how far we must go to realize the promise of molecular medicine.

The cloning of the BRCA1 and BRCA2 genes, and the demonstration that in families with multiple affected women, mutations in these genes carry respective lifetime risks of ~85% and ~50% for breast and ovarian cancers, started a revolution in the counseling of women with family histories of these malignancies. By demonstrating that they have not inherited their family's BRCA mutation, we can now tell many women that they have escaped the genetic curse that killed their relatives. But the dark side of BRCA testing illustrates some of the problems of applying our new knowledge of human molecular genetics to clinical medicine. BRCA1 and BRCA2 are large genes and germ-line BRCA mutations are private, complicating the task of identifying clinically relevant mutations and distinguishing them from benign polymorphisms. In addition, linkage studies suggest that current methods of clinical screening, primarily RT-PCR followed by SSCP, direct sequencing and/or protein truncation assays, may miss more than 30% of germ-line BRCA mutations. As a result, BRCA testing is frequently uninformative, even in high-risk families. When we do identify women who are positive for disease-related BRCA mutations, we offer them only difficult choices. They can undergo increased screening for breast and ovarian cancer, knowing that such screening is of uncertain merit. Alternatively, there is prophylactic mastectomy and oophorectomy, but these surgeries only reduce rather than eliminate the risk of breast and ovarian cancer. Population screening is another thorny issue. Four BRCA mutations occur at a combined frequency of over 2% in Ashkenazi Jews, making cancer predisposition from BRCA mutations potentially the most common genetic disease of the Ashkenazim. This observation incited a heated debate as to whether Ashkenazi women without positive family histories should undergo BRCA screening, given that the risk carried by these mutations in unselected individuals was unknown. [1] Finally, the information gained from presymptomatic genetic testing can be used to deny a BRCA carrier insurance or employment.

Four recent articles in the New England Journal of Medicine have helped to clarify the situation. In a study of more than 5300 Ashkenazi Jews, Struewing et al. (1997) [2] found that the Ashkenazi BRCA mutations gave only a 56% risk of breast cancer and a 16% risk of ovarian cancer by age 70, considerably less than the 85% breast cancer and 50% ovarian cancer risk predicted by the Breast Cancer Linkage Consortium (BCLC) analysis of BRCA mutations in families with multiple women affected by these cancers [6]. In companion pieces, Couch et al. (1997) [3] reported that only 16% of women with a positive family history of breast cancer who were seen in their cancer genetics clinic had BRCA1 mutations, rather than the 45% predicted from linkage studies, and Krainer et al. (1997) [4] found that BRCA2 mutations make relatively little contribution to early-onset breast cancer. Finally, Schrag et al. (1997) [5] provided the first comprehensive estimates of the effects of prophylactic surgery on the life span of BRCA carriers.

These reports represent real progress in the application of molecular medicine to cancer. They put us in a much better position to counsel BRCA carriers as to their true risk of developing cancer and their options for decreasing that risk. However, the popular press viewed the studies as seriously weakening the case for the clinical application of BRCA mutational analysis. Headlines in American newspapers included "Research questions role of breast cancer genes," "Breast cancer test may be overrated," and "Genetic marker assailed." The general medical community was quoted as reacting with skepticism toward BRCA testing. An editorial by former NIH head Bernadine Healy set the standard [7], concluding that "it is too early to use BRCA gene testing in everyday clinical practice."

Much was made of the reported differences between the Struewing study and previous work by the BCLC. But despite the media spin, the new data do not contradict the old. The Struewing study estimates the risk conferred by BRCA mutations in the general population, while the BCLC study risk estimates are for those BRCA carriers who are members of families with a high incidence of breast and/or ovarian cancer. Not surprisingly, the risk conferred by BRCA mutations in the general population was found to be lower than that in high-risk families. Even this difference may not be real, as the overall risks for breast cancer reported in the two studies are not statistically different and the age-specific risks are quite similar until age 65. The Struewing study did provide crucial information for evaluating the efficacy of population-based screening and, rather than invalidating the role of BRCA mutations in breast and ovarian cancer, it demonstrates that even in a relatively unselected population, BRCA1 and BRCA2 mutations remain two of the strongest known risk factors for breast and ovarian cancers.

This last point is emphasized by Schrag et al., who calculated that middle-aged BRCA carriers who undergo prophylactic surgeries will gain several years of life expectancy. For example, using a set of conservative assumptions, they found that a 30-year-old BRCA carrier from a high-risk cancer family would gain, on average, 5.3 years from a prophylactic mastectomy alone and 7.2 years from mastectomy at age 30 and oophorectomy at age 40. As pointed out by the authors, this effect on life span is greater than that achieved by preventing heart disease through smoking cessation or reducing serum cholesterol from >300 mg/dl to 200 mg/dl.

While the lay press may have been off base in concluding that the aforementioned studies invalidated BRCA testing, caution still is warranted regarding the widespread use at this time of genetic testing for cancer predisposition. Our current ability to provide clinically useful information is constrained by a variety of basic scientific, clinical, and political issues. The small number of cancer susceptibility genes that have been isolated to date and the difficulties faced in identifying the multitude of disease-causing mutations in these genes limit the clinical utility of presymptomatic testing, but these are problems that will eventually be overcome as human genetics and biotechnology advance. In addition, insights into these genes' normal function, such as the recent demonstration that the BRCA genes are involved in DNA repair [8, 9] may lead to new approaches for prevention and treatment. However, without the kind of translational studies discussed above, our expanding basic science knowledge will be of little practical value.

At the same time, the scientific community needs to educate not only the general public but also medical practitioners on the proper clinical application of molecular genetics and the true nature of presymptomatic genetic testing. This problem's depth is illustrated by a recent study of genetic screening for familial adenomatous polyposis, a condition in which colorectal cancer is caused by germ-line mutations in the APC gene. The study found that less than 20% of the patients received genetic counseling before the test and that physicians misinterpreted test results in almost a third of the cases [10]. Finally, there needs to be appropriate legal protection for individuals who seek presymptomatic testing. President Clinton's recent call to outlaw discrimination against people on the basis of their genetic inheritance is a move in the right direction.

In one way, the new genetics is a victim of its own success. We have promised the moon when it comes to our ability to look into someone's genetic future. As a result, both the public and the medical profession yearn for absolute answers from presymptomatic genetic testing. Yet we offer only genetic predisposition, not genetic determinism. We stand accused of being genetic bookmakers who deal in odds, not certainties. We should face the truth, plead guilty, and work on improving our fortune-telling abilities. Even a cloudy crystal ball is better than blind luck for mastering one's fate.

M. Stephen Meyn is Associate Professor of Genetics and Pediatrics, and Codirector of the Cancer Genetics Program, at Yale University School of Medicine.

Send us your comments and ideas for future articles.

Endlinks

Oncolink - multimedia oncology information resource maintained by the University of Pennsylvania Cancer Center. It offers a wide range of cancer information, from research article listings to support groups to art. The BRCA 1 Information Center maintains links all over the Web.

New England Journal of Medicine - offers full-text online collections of articles on given topics on a rotating basis free of charge. Breast cancer is one of the current topics and most of the articles discussed in this opinion are available. NEJM also offers a search of previous issues that will deliver abstracts. Full text-reprints must be ordered from NEJM.

The Gene Letter - an online, bimonthly newsletter focusing on genetics and public policy. It discusses science, medicine, ethics, and law. It has resource links and a chat room. The article "How Many People Seek Genetic Testing for Cystic Fibrosis, BRCA1, and Huntington Disease?", published in May 1997, is one of the various articles now available.

The American Cancer Society - a vast Web site. Information on breast cancer can be found in the Breast Cancer Network, which may be found on this site.

OncorMed - recently received exclusive rights to the BRCA 2 gene for diagnostic purposes. They are in the process of developing diagnostic tests for the gene. Their Web site provides information about clinical uses.

An example of the lay press's reporting of the NEJM studies can be found in the May 26 issue of Time magazine, in an article entitled "Cancer Genes Revisited".