Stanford Center for
Biomedical Ethics–

Genetic Testing for Breast Cancer Susceptibility

Stanford University Program in Genomics, Ethics, and Society

Draft Executive Summary

Report of the Working Group on Genetic Testing for Breast Cancer Susceptibility

November 23, 1996 draft for discussion - not for citation. Please note:

Introduction

Every year, about 185,000 Americans are diagnosed with breast cancer; about 45,000 die of the disease. The news that mutations in first one, then a second, gene had been linked to hereditary breast cancer excited many and led people to hope that a new, genetic, era of prediction, prevention, and treatment for breast cancer was opening.

Two years after the successful cloning and sequencing of BRCA1, the hopes remain, but are tempered, and they are joined by an increasing realization that the discoveries bring costs as well as benefits. In many respects, those costs are currently more numerous, and more significant, than the benefits.

In the fall of 1995, the Stanford Program in Genomics, Ethics, and Society assembled a multi-disciplinary Working Group on Genetic Testing for Breast Cancer Susceptibility. Over sixty people have participated in the group, which averaged two meetings per month for the past year. They have included physicians (surgeons, psychiatrists, medical geneticists, oncologists, and internists), basic scientists, breast cancer activists, social scientists, undergraduate, graduate and professional students, genetic counselors, philosophers, journalists, law professors, and others. Working group members came primarily from within Stanford University but also from the rich pools of talent in the surrounding San Francisco Bay area, including, notably, a rich infusion of expertise from the Kaiser Permanente Medical System.

The Working Group has spent the last year educating itself about the issues raised by genetic testing for breast and ovarian cancer susceptibility and debating the consequences of testing for individuals, families, and society. It has turned what it has learned and concluded into a book-length report that covers the vast sweep of issues raised by our ability to identify a genetic susceptibility for breast and ovarian cancer with a blood test. We concluded early that the consequences of genetic testing would necessarily vary from disease to disease: because the genetics will vary, the disease will vary, and, most importantly, the social and cultural meaning of the disease will vary. Thus, although we hope that our report will prove of value in the growing public debate about the ethical, legal, and social implications of advances in molecular genetics, our recommendations are focused quite specifically on BRCA1/2 testing.

This Executive Summary includes two main sections. An overview of the scientific background is included in Part I of the Executive Summary, including a review of the identification of BRCA1/2, and a discussion of our current knowledge of the clinical benefits and burdens of genetic testing. The full report contains a more comprehensive background to BRCA1/2 testing. The volume opens with the stories of six women, potential candidates for DNA testing. The dilemmas experienced by these women and their families are developed and elaborated throughout the report. Members of the Working Group also contributed chapters elaborating the social and cultural context of genetic testing for BRCA1/2, including the particular fear breast cancer evokes, the changed meaning of cancer when viewed as a genetic disease, and how unexamined features of the American cultural background may constitute a bias in favor of testing.

Part II summarizes the Working Group's recommendations and conclusions. They focus exclusively on the consequences of testing for mutations in the BRCA1/2 genes and are specific to the American context. The report includes ten general recommendations as well as more detailed suggestions. Although the ten general recommendations were endorsed by the full group, our deliberations were not always unanimous. Where relevant, strong minority views are highlighted. We believe that the depth, detail, and focus of the full report adds weight to our conclusions, but we recognize that reasonable people may, and undoubtedly will, disagree with us on particular points. We cannot claim that our findings are "right" or that our process represented every viewpoint. We do believe that they resulted from the dedicated and prolonged efforts of a diverse group of people, who, unlike many of those who have issued recommendations on these issues, are not closely connected to one profession, specialty, or perspective. We hope the recommendations and the report from which they are drawn will help us move forward wisely as our society confronts what seems likely to be an inevitable increase in our ability to predict genetically-linked risk.

Part I - Genetic Testing for Breast and Ovarian Cancer Susceptibility: Scientific Background

Breast cancer is the most common non-skin cancer found in American women and their second-leading cause of death from cancer. Although about 1,400 men each year will be diagnosed with breast cancer, ninety-nine percent of those affected are female. Over an average lifetime, an American woman has about a 10 percent chance of being diagnosed with breast cancer. Although breast cancer is a heterogeneous disease, about 80 percent of the cases are "infiltrating ductal carcinoma," a tumor of the cells that line the milk ducts of the breast.

The risk of being diagnosed with breast cancer appears to increase steadily with age, at least until the late 70s. The median age of diagnosis with breast cancer is in the mid-60s. Only a very small fraction of women diagnosed with breast cancer are below forty; fewer than a quarter are below fifty. Of those diagnosed with breast cancer, approximately one-fourth will die from it.

Overall, five to ten percent of cases of breast cancer, are believed to be associated with an inherited genetic change, such as a mutation in BRCA1/2. The definition of a significant family history of breast and/or ovarian cancer requires multiple affected family members; the disease is so common that two breast cancer cases in a family often occur randomly, unrelated to genetic predisposition. Researchers' definitions of familial breast cancer vary, but all require several cases of breast or ovarian cancer among near relatives in the same line of descent. An early onset of the disease somewhat increases the significance of any one case.

In 1990, after years of work, a research team led by Dr. Mary-Claire King was able to demonstrate that some affected women in high-risk families shared genetic markers in common in a region of the long end of chromosome 17. Dr. King named the gene she had identified BRCA (for breast cancer) 1. In 1994, a group from the United Kingdom, identified another set of women in high risk families who shared genetic markers on chromosome 13 and named this gene BRCA2.

Localizing a marker associated with a disease to a particular region of a chromosome is a crucial step, but much work remains to be done before the actual gene is precisely located and sequenced, or "cloned." The effort to clone BRCA1 took on the aspects of a race, won in 1994 by Dr. Mark Skolnick and his team from the University of Utah, Myriad Genetics, Inc., and collaborators at other institutions. BRCA2 was cloned by Michael Stratton and his team in 1995.

BRCA1 proved to be a fairly large gene, with 5,592 base pairs in 22 coding exons, spread out over about 100,000 base pairs of genome. BRCA2 is another relatively large gene, with 27 exons distributed among about 70,000 base pairs of DNA. For both genes, multiple mutations have been found throughout the gene's coding regions and in some of the surrounding, non-coding DNA. There are a few locations where mutations are particularly common, at least in some populations. The most publicized of these mutations is a two base pair deletion of an A and a G at position 185 in the BRCA1 gene; the so-called 185delAG mutation. This mutation has been found at the relatively high rate of about one percent in women of Eastern European Jewish (Ashkenazic) ancestry. Similar specific mutations with high frequencies in particular populations have been found in both BRCA1 and BRCA2 in Sweden, Iceland, and Finland, among other places.

Both BRCA1 and BRCA2 are thought to be tumor suppresser genes. When they function properly, they inhibit the uncontrolled growth of cells. When they are not working properly, cells have lost one source of control of their growth and thus have taken a step toward becoming tumors. Both genes confer breast cancer risk in an autosomal dominant manner; a person apparently needs only to inherit one mutated copy of the gene, from either parent, to be at higher risk for cancer. Thus a child of an affected parent has a 50 percent risk of inheriting a mutated copy of a BRCA1/2 gene. Presumably, the higher risk of developing cancer comes because each cell then only has one properly functioning copy of the gene, which, if mutated in a cell during the person's lifetime -- by radiation, chemical carcinogens, diet, or other factors, genetic or environmental -- will no longer provide protection against the cell becoming malignant.

The extent of the increased risk conferred by mutated versions of these genes is not entirely clear. Research to date has been done in families selected because of their known high risk for breast and/or ovarian cancer; it is possible that other characteristics, genetic or environmental, in those families may make their risks from mutated BRCA genes different from those of the general population. It is also unclear whether all mutations in the BRCA genes put a person at higher risk for cancer or whether the extent of the heightened risk varies. These issues continue to be explored.

Nonetheless, the work done thus far suggests that both BRCA genes, when present in mutated form, confer a risk of breast (and in some cases ovarian) cancer substantially higher than most women face. A woman with a mutated BRCA1 gene has a lifetime risk of contracting breast cancer that has been estimated at approximately 70 to 85 percent, compared with a risk of roughly 10 percent in the general population. BRCA2 mutations confer a lifetime risk of breast cancer comparable to BRCA1 mutations. BRCA1 mutations also confer a heightened risk of ovarian cancer. Patients in high risk families who have mutated BRCA1 genes have an estimated 28 to 40 percent lifetime risk of being diagnosed with ovarian cancer, compared with the roughly two percent lifetime risk of this disease in the general population. BRCA2 mutations have also been linked to ovarian cancer, although at a somewhat lower rate. BRCA1 may also be linked to somewhat higher-than-average risk of colon cancer in both sexes and of prostate cancer in men. BRCA2, but not BRCA1, is linked to a somewhat higher risk of male breast cancer.

BRCA1 and BRCA2 mutations are found in most, but not all, of the families with strong histories of breast or breast and ovarian cancer, with BRCA1 mutations about twice as common as BRCA2 mutations. As many as twenty percent of those families, do not have inherited BRCA1 or BRCA2 mutations. Other factors, genetic or environmental, may be at work in those families. It is not yet clear how common BRCA1 and BRCA2 mutations are in the general population. Calculations based on the extent of familial breast cancer and the estimated penetrance of the mutations in high risk families lead to estimates that about one of eight hundred women carry an inherited BRCA1 mutation. Somewhat surprisingly, mutations in BRCA1 or BRCA2 have not been found in sporadic cases of breast cancer -- cancers where the patient was born with two functional copies of the gene but lost both of them.

Limited genetic testing for risk from the mutations was possible as soon as the markers were localized. Individuals could be checked through so-called linkage studies to see if the relevant piece of their chromosomes was the same as that carried by family members with breast or ovarian cancer, based on the presence or absence of markers known to be near the gene. This kind of test required samples from numerous affected and unaffected family members and necessarily involved a risk that the marker used had become separated from -- unlinked -- from the mutated gene.

With the cloning and sequencing of the genes, tests have become simpler. The BRCA1 or BRCA2 genes from a patient can be isolated, in whole or in part, and studied for mutations. If the test looks for a particular, known mutation, such as 185delAG, testing is relatively cheap, with direct charges of approximately $350. If, however, the test needs to examine the full length of the gene for mutations, the patient's entire gene must be sequenced. With current technology, this is more complicated and more expensive. OncorMed charges $1995 for full BRCA1 testing and $2100 for full BRCA1/2 testing. Myriad charges $2400 for full sequencing of BRCA1/2. (Technology under development is likely to reduce substantially the time and cost of such tests in the near future.) Testing a family member for a mutation that has already been identified in a relative with cancer is significantly cheaper; laboratory charges vary from $260 at the University of Pennsylvania to approximately $350 at OncorMed.

Patent applications have been filed for "inventions" covering both BRCA1 and BRCA2. Myriad Genetics has filed applications covering both genes; it faces a competing application for the BRCA2 patent. If those applications are granted, the inventors, or their assignees, will be able to limit or control testing using those genes, and theoretically could limit research use as well. Unless or until the patents are granted, however, the technology can be used freely, within the context of applicable regulations.

There are, however, no substantial regulations that necessarily apply to genetic tests. Genetic tests are not "drugs" or "biologicals" and hence do not have to be approved by the United States Food and Drug Administration after proof that they are safe and effective. If a DNA analysis is offered as a test kit, to be sold to individuals, physicians, or laboratories, it then would be a "medical device" and would have to be approved by the FDA. If, however, the test is offered by clinical laboratories, using readily available or locally developed materials and reagents (sometimes called "home brews"), it is not currently classified as a medical device and is not subject to FDA regulation. The FDA has proposed regulations that would expand its coverage of certain materials used by clinical laboratories, in a manner that could have important implications for genetic testing. Whether those regulations will be adopted and how stringent they might be remains unclear. The clinical laboratories that perform such tests are regulated by the federal government under the Clinical Laboratories Improvements Act and by state regulation. These regulations seek to ensure the competence of the laboratories and their staffs, but do not attempt to assess the clinical efficacy of genetic tests that are offered.

Currently, testing for BRCA1 and BRCA2 mutations is available through a number of sources. Most research teams have offered the testing -- using linkage analysis or direct gene testing -- based on protocols reviewed and approved by institutional review boards that are charged with protecting research subjects. BRCA1 tests have been available commercially or some time through a clinical laboratory firm called OncorMed and through a laboratory at the University of Pennsylvania. Myriad Genetics announced in late October 1996 that it was about to begin offering tests for both genes through its subsidiary, Myriad Genetic Laboratories, Inc. At least one other laboratory, The Genetics and IVF Institute in Fairfax, Virginia, has offered testing for the 185delAG deletion for $295.

Questions remain, however, about the benefits and the drawbacks of genetic testing for BRCA1 and BRCA2 mutations. There are no known methods for preventing breast or ovarian cancer that would be particularly important to women with mutated versions of these genes. Although many activities have been linked to higher risk for breast cancer, including age of menarche and late childbearing, whether avoiding high risk choices will affect women with BRCA1 or BRCA2 mutations is not known and will not be known for years. And many epidemiological risk factors, such as age at first pregnancy, are not subject to easy modification. Prophylactic surgeries -- either mastectomy or oophorectomy -- are another, drastic option, but their value in preventing breast and ovarian cancer in high risk women remains unclear. They do not perfectly prevent disease; at least some women have developed breast or ovarian cancer after such operations. Similarly, chemoprevention is of unproved efficacy.

It is also unclear whether genetic test results would help advance early detection of breast or ovarian cancer. All women should have regular breast examinations, including breast self exam and examinations by professionals; all women after an appropriate age should have regular mammograms. Women with significant family histories of breast cancer are generally advised to take those early detection steps whether or not they have BRCA1 or BRCA2 mutations. Although no reliable method for early detection of ovarian cancer exists, women may avail themselves of currently accessible early detection methods, such as pelvic examination, transvaginal ultrasound, and blood tests for CA125.

Finally, at this point, for women without a significant family history, treatment for breast cancer is no different whether or not a woman has a mutated BRCA1 or BRCA2 gene. Recent research suggests that breast and ovarian cancer found in mutation-positive women may be less aggressive than sporadic cancer; survival may be longer. The implications of these preliminary findings for treatment decisions are currently unknown. A woman with the mutation might decide that the risk of a second tumor justifies broader surgery, including perhaps a prophylactic mastectomy of the other breast, but that course, at least with current data, is by no means compelling.

On the other hand, it is clear that genetic testing has costs beyond its price. Women who undergo genetic testing may face psychological problems in adjusting to the results, whether positive or negative. Their test results may cause tension and discord within their families, immediate and extended. Genetic tests are unique, yielding results that extend beyond the individual, affecting other family members whether that is intended or not. If they do have a mutated BRCA1 or BRCA2 gene, knowledge of that mutation may affect their ability to be insured, or employed, in the future. Additionally, women who test negative for the mutation still face the roughly 10 percent lifetime risk of breast cancer of the general population. If a negative genetic test result led them to stop breast examinations or mammography, the results could be tragic.

The health care system in which women face these issues remains undeveloped. There are few qualified genetic counselors and most have little experience with genetic testing for cancer susceptibility, or with the specifics of BRCA1/2 testing. Physicians typically have very little knowledge of human clinical genetics, including the BRCA1 and BRCA2 genes. Providers of health coverage need to develop policies about whether and when genetic testing will be covered and about which other genetics-related services they will include.

Part II - Working Group Recommendations

Against that background, the Working Group developed the following sets of recommendations. These recommendations arise from the discussions in the body of the report and are discussed more fully in those chapters. The recommendations are aimed at different audiences, from positions about needed legislation to suggestions for health care providers and individual people. Some are expressed as general principles, some could be implemented by health care providers or organizations at the level of policy, and others address needed legislative change. The recommendations are divided into two sections: 10 broad recommendations which provide an overview of the Working Group's positions, followed by a set of more detailed recommendations with respect to implementation. They are explicitly based on the state of scientific knowledge -- and legal and cultural constraints -- as of late 1996. As both science and society change, the recommendations will need to be modified .

General Recommendations

1. The potential value of genetic testing must be judged in the context of the meaning of the disease for those affected or threatened, taking into consideration the social, cultural, political, and economic environments of testing, as well as its medical implications.

Genetic testing for breast cancer susceptibility is not just about medicine or science; it is about people's lives and the families and societies in which they live. In weighing the benefits and costs of genetic testing, the focus must be on the people involved. The meaning of any genetic disease, condition, or predisposition is not uniform. Breast and ovarian cancer have medical, social, and personal consequences and meanings different from those of colon cancer or heart disease. Even when talking about a single disease, condition, or predisposition, the significance of the genetic test result will vary among people. It will depend upon personality, social class, gender, and ethnicity, as well as complex family relationships. The meaning will also change over time for the person, depending on age, experience with the disease, and self-assessment of risk.

In addition, genetic testing for cancer susceptibility is growing out of particular social, cultural, economic, and political contexts. Genetic testing will be implemented in those contexts and the consequences of such testing will feed back into those same contexts. Genetic testing cannot be looked upon in isolation; it represents a very particular approach to health, illness, and death, and makes a strong cultural statement that we must try to understand.

2. The effects of genetic testing for BRCA1/2 mutations are complicated. The tests may have both positive and negative consequences for individuals and families. At our present level of knowledge, the tests should be offered, and taken, only with great care.

Medical "tests" are sometimes viewed as inherently benign procedures, where there is little or no risk of harm to the patient. In light of the risks and consequences of false positives and negatives, this is rarely true with any medical test. It is definitely not true with genetic testing for breast and ovarian cancer susceptibility, but there the risks are primarily non-medical. Women who are tested for BRCA1/2 mutations risk unforeseen psychological consequences, family disruption, and adverse social consequences, such as loss of insurability or employability. Most women, and many health care providers, will not necessarily know or consider those kinds of risks when contemplating genetic testing for breast cancer susceptibility.

These risks do not mean that BRCA1/2 testing is never beneficial or should never been offered or taken. They do mean, however, that testing should be approached, by providers and individuals, cautiously.

Although no one strongly objected to this recommendation, some members of the Working Group believe that genetic testing for BRCA1/2 should be framed in a more positive light.

3. For most people, BRCA1/2 mutation testing is not appropriate. For people at high risk for carrying a mutation, either as a result of their family history or their own early onset of disease, testing is an option that should be discussed and that could reasonably be accepted -- or declined. Even for those not at high risk, testing, though not encouraged, should not be prohibited.

After careful weighing of the advantages and disadvantages of genetic testing, we believe most people both should and will decide not to be tested for BRCA1/2 mutations. For people who do not have a strong family history of breast or ovarian cancer, the test will offer almost no benefit. For a properly informed person from a lower risk family who remained anxious about her BRCA1/2 genes, a negative test would not necessarily allay that anxiety -- and a positive or ambiguous test would not offer any attractive medical options, but could lead to psychological, familial, and social problems. A negative test might even have the adverse effect of making a woman less concerned about her general risk of contracting breast cancer. The small benefit and potentially high cost for lower risk people necessarily imply that screening programs for large populations are not appropriate, at least with our current knowledge of prevention, early detection, and treatment for breast and ovarian cancer.

We do not believe that many women from lower risk families who have already been diagnosed with breast or ovarian cancer are likely to find a genetic test useful. Except perhaps among those with a very early age at diagnosis, only a few will test positive. In light of the increased risk of a new primary cancer in any woman diagnosed with breast cancer, it is not clear that many women would change their treatment options as a result of genetic testing. On the other hand, the reality of their cancers will already have subjected these women to many of the psychological, familial, and social discrimination problems of genetic testing, so the disadvantages of the tests may be of less concern to them.

We expect that some people from high risk families would reasonably opt for the test. For them, the chance that they carry a BRCA1/2 mutation is substantial -- if a parent carried the mutation, the chance is 50 percent. They may be so anxious about their genetic status, or that of their children, that a negative test would improve their lives. Some may be so concerned that they would consider prophylactic mastectomies or oophorectomies. For many of these women, however, the possible psychological, familial, and social consequences of a positive result may persuade them not to be tested (even though their being in a high risk family already confers some of those costs). The possibility of testing, and its benefits and burdens, should be discussed with people from high risk families. For them, a decision either to be tested or not to be tested seems reasonable.

Although we do not believe testing is appropriate for people at lower risk, we are not prepared to recommend that it be prohibited to them. To prevent a competent, fully informed adult from obtaining information about his or her genetic status seems too paternalistic to us. This reluctance to ban such testing, however, does not mean that providers should be obliged to offer or provide testing in cases where they believe it is not appropriate. Nor does it mean that others should bear through insurance the cost of a lower risk individual's test.

This recommendation about the appropriate scope of testing is particularly sensitive to advances in medicine. Should scientists discover ways in which knowledge of BRCA1/2 mutations could significantly aid people in preventing, detecting, or treating breast or ovarian cancer, far more people may reasonably seek testing.

4. Whenever undertaken, testing should be done by competent professionals who have received specialized training. Testing programs should encompass adequate informed consent, genetic counseling, and multi-disciplinary follow-up care, including psychological counseling for the individual and family when appropriate.

In spite of the limited value and substantial costs we see in testing for mutations in BRCA1/2, we believe that a properly implemented testing program should be open to almost anyone who wants to participate in it. (The specific groups we believe should not be tested, fetuses and children under age 18, are discussed in detail in the specific recommendations.) Although further research is essential, it is now known that many mutations in these genes do give women substantially higher risks of disease. Although little can be done with the information from the tests, they can have enough beneficial value that we believe properly informed adults should be allowed to give informed consent to be tested in an appropriate program.

But the issues involved in making a decision to be tested are complex and personally sensitive. An appropriate program must include competent professionals, who have the knowledge and experience necessary to assist people in their deliberations about testing. We believe that access to the tests should be allowed only through a relationship with a trained professional and that more evidence of competence in this area should be required than a general physician or genetic counselor license. Governments, health plans, and clinics should require special training of those who would offer genetic testing for BRCA1/2 mutations, whatever their medical background.

An appropriate program of genetic testing for these mutations must include substantial genetic counseling so that the patient knows whether he or she is in a high risk family, and understands in advance of her decision the likely consequences of a negative or a positive test result. An appropriate program also must include follow-up after the test results have been given. People should be counseled about the implications of the test and should be evaluated to determine whether they are experiencing significant psychological consequences from the test. If so, they should be referred for psychological counseling or treatment.

Restricting the tests to those who go through qualified professional intermediates and requiring that approved programs include follow-up and psychological help where needed are good practice. They could be legislated by the states or the federal government or might be imposed through health plans, medical clinics, professional organizations, or testing firms. They may effectively be imposed as part of the standard of care for practitioners.

5. Additional research on the full implications of BRCA1/2 mutations and testing is crucial. Testing need not be limited solely to research protocols, but all testing programs -- whether conducted by research groups or commercial laboratories -- should meet the standards set out in General Recommendation 4 and should include a confidential data collection component.

The unanswered questions about BRCA1/2 mutations are many and vital. We do not know enough about the risks conferred by the mutations, the nature of the cancers they generate, or the consequences of testing on those who receive it. A national registry -- with appropriate strong safeguards for patient privacy -- should be established to track information about detection, prevention and treatment in people who receive BRCA1/2 testing and made available to qualified researchers. Patients who go through informed consent for testing -- whether they decide to be tested or not and whether they test positive, negative, or indeterminate -- should be encouraged to participate in that registry. In effect, this would give all genetic testing for BRCA1/2 mutations a research component.

There has been much discussion about whether testing should be limited exclusively to research settings. We believe that such a restriction is not now justified. We know thus far little about the mutations and the consequences of testing, but we know enough that the tests cannot fairly be considered purely "experimental." It does not seem justified to require people to participate through a research protocol if they do not want to, particularly if geographical or other factors make research participation difficult for them.

Both researchers and commercial laboratories can have interests in tension with those of their patients. Although research protocols typically provide substantial protections for participants as a result of federal regulations and the activities of Institutional Review Boards (IRBs) to govern human subjects research, it is possible for research programs to fall short of the standards we would set for all testing. This is even more true of commercial testing, which does not have to comply with IRB requirements.

But we believe that those kinds of protections are essential to a good testing program. Research protocols should be held to high standards, perhaps higher than those imposed by IRBs, and commercial testing should be held to the same standards. The focus needs to be on the adequacy of the protections built into the testing program, not on whether the testing program is called "research".

This recommendation was problematic for a minority of Working Group members. These members are in agreement that a research context for genetic testing does not necessarily assure a high-quality environment for offering tests. However, the minority believes that restricting testing to research settings is one way to limit the rapid diffusion of BRCA1/2 tests into medical practice in advance of clear knowledge of the tests' value.

6. Improved knowledge about BRCA1/2 mutations is essential to appropriate testing programs. Education programs need to be undertaken for people at risk, for the general population, and for health care providers. Education programs need to reflect the currently understood advantages and disadvantages of testing and to make clear that not being tested can be a wise decision.

Few people understand very much about BRCA1/2 mutations or testing for those mutations. For society to use these tests wisely, that number must grow, among several different groups.

People at higher risk for breast or ovarian cancer, who might appropriately choose to be tested, obviously will need good information before deciding whether to approach a testing program. Existing breast cancer awareness outreach and education programs for patients, their families, and other potential consumers should be expanded to include information on genetic testing for hereditary susceptibility to breast and ovarian cancer.

The general population also needs to know more about BRCA1/2 mutations. Few people understand dominant inheritance, penetrance statistics, the kinds of effects testing might bring, or even the general risks of breast and ovarian cancer. Increased public education would both help people decide whether testing held any appeal for them and improve their understanding of the meaning of test results for others who did choose testing. Special efforts should be made to educate policy-makers, legislators, and legal professionals about genetic testing for BRCA1/2 mutations.

Health care providers need to learn about the test. Those who plan to offer testing must be competently educated in not only the medicine and the science of BRCA1/2 mutation testing, but in the possible effects of testing on individuals. Practitioners who do not plan to offer testing should still be able to provide their patients with some useful information about testing that may help them decide whether to seek out a testing program. Existing breast cancer outreach and education programs for health practitioners should be expanded, and new, well-coordinated programs initiated to enable clinicians, health care providers, and other health professionals to learn about and discuss the ethical, legal, social and medical issues around genetic testing for breast and ovarian cancer susceptibility with their patients.

The issues raised by testing are sufficiently complex that educational presentations could easily be consciously or unconsciously biased in any number of ways. For this reason, it is important that educational programs not be provided primarily by groups with financial interests in testing.

7. Marketing of genetic tests for BRCA1/2 mutations should be carefully limited.

The Food and Drug Administration limits the ways in which pharmaceuticals can be marketed. Drugs can only be marketed for their labeled uses. Full marketing must include substantial labeling information, including lengthy discussions of contraindications and possible risks. Marketing for genetic tests, although subject to general Federal Trade Administration oversight along with all other advertising, is not specifically regulated by the Food and Drug Administration. It should be.

Many women are terrified of breast cancer. Marketing campaigns might easily prey on this fear, by first increasing women's anxiety and then offering testing as a solution to this heightened concern. Genetic testing for BRCA1/2 mutations holds too many dangers to allow unrestricted advertising. Marketing of genetic tests should be brought under FDA regulation in the same manner as prescription drugs.

In recent years, some pharmaceutical firms have begun "indirect" marketing of prescription drugs to consumers. These advertisements generally talk about a use for a pharmaceutical, suggest that the consumer see his or her doctor, and mention the firm's name. In light of the complex issues involved in BRCA1/2 mutation testing and the strong emotions that breast cancer can evoke, genetic testing programs and clinical laboratories should voluntarily avoid such advertisements.

8. Health plans should cover genetic testing for BRCA1/2 mutations for appropriate plan members; when genetic testing is covered, the full scope of an appropriate program must be covered.

Health coverage in the United States typically includes treatments that are non-experimental and are "generally accepted." Although much remains to be learned about testing for mutations in BRCA1/2 even for people at higher risk, enough is known that the tests of such people should no longer be considered experimental for purposes of health coverage.

For a small group of people, we believe BRCA1/2 testing has reached the point where it should be termed "generally accepted." This group includes adults from high risk families. For most other people, based on our current knowledge, testing offers little if any benefit but imposes some costs, both on them and on the overall health care system. At the present time, the financial costs for these tests are substantially more than $2,000 (not including counseling and follow-up care). Health plans should not normally cover BRCA1/2 testing for people outside of groups that are clearly and openly defined as appropriate. Whatever definition of "high risk family" an insurance company adopts, it should be readily available to plan members.

When health plans do provide coverage of genetic testing for breast cancer susceptibility, they need to cover all the services necessary to providing a proper testing program. These include genetic counseling and follow-up and, when appropriate, psychological counseling after the test. We recognize that by requiring these services, we will necessarily increase the price of the test. But without such services, a testing program is inadequate and may be dangerous; our society has not generally allowed substandard care just because it is cheaper.

9. Federal and state laws should clearly prevent the use of information about genetic testing -- such as BRCA1/2 mutation testing -- in health coverage or employment decisions. The fact that a genetic test has been administered should be protected in the same way as the information gained from the test. Governments should study carefully limiting uses of genetic information in other contexts.

Among the most feared consequences of genetic testing in the United States are its effects on health coverage and on employment. Currently, the law in the United States is variable and unclear. Employment discrimination based on some genetic characteristics may be illegal under the federal Americans with Disabilities Act . . . but it may not be. State laws on the subject are few and variable. With some limitations imposed by the new Kassebaum-Kennedy legislation, federal law does not stop health plans from considering genetic information in deciding whether to cover someone, how much to charge them, and what conditions to cover. About a dozen states have passed statutes imposing some limitations, but the limits are variable, untested, and cannot help the many people who are covered by most self-insuring employers.

In employment, there is no justification for discrimination on the basis of genetic susceptibility to breast cancer, particularly as discrimination based on a cancer diagnosis, past or present, is clearly banned by the Americans with Disabilities Act. Congress should equally clearly ban discrimination based on a heightened risk of a future diagnosis; if Congress fails to act, the states should.

In health coverage, insurers in many contexts are allowed to consider present or past cancer diagnoses in making underwriting decisions. Although the majority of covered Americans are not subject to this medical underwriting, we believe that its use is wrong. Health coverage is essential to leading a good and secure life, a fact recognized by all other developed countries, which guarantee health coverage to effectively all their residents. The United States should do the same. But as the realization of that broader goal no longer seems to be in the immediate offing, we believe that legislation limited to genetic risks is appropriate in order to prevent the extension of the current flawed system of medical underwriting to more people.

Currently, other insurers, such as providers of disability or life insurance coverage, are not generally forbidden to use genetic information. With respect to those forms of insurance, which have been the subjects of much less attention and study, further work is necessary. Similarly, genetic information might be used to try to determine whether a person should have custody of a child, become an adoptive parent, or be an adopted child. Although it seems unlikely that information about BRCA1/2 mutations would be appropriate as part of such decisions, these issues also require further study.

The passage of legislation against the use of genetic information in employment and health coverage will not make those uses automatically disappear. Nor will it do much to ameliorate the potential psychological harms tested people might face. Passage of such legislation would not, therefore, fundamentally change our assessment of the balance of advantages and disadvantages in genetic testing for breast cancer susceptibility. But it would be a good thing and should be done.

A minority of the Working Group believe that disability and life insurance, which may be essential for a woman to provide a reasonable quality of life to her family, should be included in legislation barring the discriminatory use of genetic information. Substantial efforts should be made to understand the impact of the potential loss of these forms of insurance on high risk families.

10. The federal government should regulate the introduction and use of genetic tests for conditions, disease susceptibility, or carrier status.

Currently, genetic tests are almost totally unregulated. As long as the test does not involve a marketed "test kit," the FDA does not now assert jurisdiction over it as a drug, biological, or medical device. The Clinical Laboratory Improvements Act provides potentially useful regulation of how laboratories do clinical tests, including genetic ones, but does not assess or regulate what tests are appropriate for widespread use. This state of affairs should be changed.

We believe that genetic tests should be subject to FDA regulation. Our preference is for legislation creating a new regulatory status for such tests, slightly different from current drug or device regulation.

Human subjects research on such tests should require FDA approval, akin to an Investigational New Drug approval. When sufficient information is established that the test does in fact accurately provide information about genetic conditions, disease susceptibility, or carrier status, it should receive a preliminary approval for marketing. That marketing would be limited to programs involving competent professionals and including all necessary safeguards. During the period of this preliminary approval, however, those administering the tests would be required to collect additional data about the safety and efficacy of the test, based on the results with those who take it. Only after sufficient information has been obtained to prove that the test is, as used, adequately safe and effective for patients -- not just medically, but psychologically and socially -- should final approval be granted.

We believe that Congress should pass legislation establishing such a process, for currently considered tests and for the expected flood of future tests. If Congress fails to do so, the regulatory approach currently being considered by the FDA to increase regulation of clinical laboratory "home brews" would be a useful step. State legislation setting out requirements for the introduction of genetic tests could also be valuable.

More Specific Recommendations

1. Appropriateness of BRCA1/2 Testing for Particular Groups Defined by Risk, Ethnicity, or Cancer History

a. Testing may be particularly appropriate for someone considering prophylactic mastectomy or prophylactic oophorectomy to limit cancer risks.

This group includes women from high risk families who are considering seriously such surgery. It may also include women diagnosed with breast or ovarian cancer who are considering additional, prophylactic surgery.

b. Testing may be appropriate for women and men from high risk families, especially those who have access to genetic test results from a relative with breast or ovarian cancer.

Different definitions of "strong family history" have been used in research on these issues. Further research may provide a better way of choosing one definition; we believe that the high risk group might reasonably be defined in several ways. Whatever definition is used, we believe it would be appropriate for such women and men to be offered and to undertake testing if they so choose, after receiving full information about the test and its possible consequences. However, we also believe it is appropriate for women and men in those categories to decide not to be tested, and it is imperative to preserve the option of declining.

c. We do not recommend testing for women and men without a strong family history of breast or ovarian cancer.

We believe the value of the information to be gained is substantially outweighed by the costs, financial and personal, of genetic testing for women and men from lower risk families. A properly informed woman or man from these groups should be able to choose to be tested, but the tests should not be recommended for them or covered by their health plans.

For men, such testing could be justified on three grounds. First, although the possible increased risks of colon cancer and prostate cancer for men with BRCA1 mutations are probably too low to justify testing, male breast cancer is associated with BRCA2 mutations. Although the absolute risk of male breast cancer even for a man with a BRCA2 mutation is quite low, the disease is so rare in men that heightened awareness of risk might lead to earlier detection and treatment. Second, men may wish to be tested to help make reproductive decisions. Finally, they may want to be tested in order to avoid the need to test their daughters.

Some members of the group dissented from this recommendation, believing that the information gained would not be sufficiently useful to justify testing either for male breast cancer risk or for reproductive decision making and that, rather than have a parent tested solely to determine the possible status of his children, any interested adult children should test themselves.

d. Population screening, for the general population or any racially or ethnically-defined subpopulation, is not appropriate.

It follows from our discussion of testing lower risk women and men that population screening is not appropriate. A mandatory screening program is quite clearly inappropriate; no current benefits from testing come near to justifying that kind of invasion of personal autonomy. Even a program of encouraged, voluntary screening is clearly unjustified. The costs of such screening would be substantial. The number of women not from high risk families who were identified as having mutations would be relatively small. The existence of a screening program would tend to encourage people to believe that the test results have more value than they actually do.

There has been some discussion of screening subgroups within the population, particularly women of Ashkenazic descent. Research has demonstrated that particular mutations may have an incidence of one percent in this group. However, the clinical significance of detecting a mutation -- outside the context of a known high-risk family -- remains uncertain. Screening for a particular mutation, such as 185delAG, would cut the financial costs of screening substantially, but it would increase the "false negatives" of women who did not have the tested-for mutation, but did have other harmful mutations. Finally, screening populations defined by race or ethnicity, even when medically justified, must be done with great care for fear of exacerbating ethnic tensions or stereotypes. With such small expected benefits, this concern alone argues sufficiently against a screening program.

2. Appropriateness of Testing for BRCA1/2 Mutations for Vulnerable Groups

a. BRCA1/2 testing should not be available to children, with or without their parents' consent.

We believe that testing for mutations in BRCA1/2 carries substantial risks for the psychological, familial, and social well-being of the tested individual. The benefits, though at present usually small, are sufficient that adults should be allowed to choose to be tested. The potential burden to children -- in altered self-concept and differential treatment within families -- is enormous. For a condition that rarely, if ever, manifests itself before the third decade of life, we can see no advantage that a child gains from being tested before adulthood that would not also apply after she reaches the age of majority and can make her own decision. Although the child's parents might derive a psychological benefit from knowing her status, that cannot justify putting the child's interests at risk. Before adulthood, children do not often have the capacity to make a truly informed and autonomous decision. Although we recognize that eighteen is not a magic birthday, conferring maturity, it is a reasonable place to draw the line. We believe that the decision to test should be made only by competent individuals, for themselves. There may be exceptions due to legal considerations or very unusual situations, where older adolescents may appropriately decide about testing. For example, "emancipated minors," who are otherwise able under state law to make their own medical decisions, should be allowed to make this decision.

b. Testing should not be available to individuals of any age who lack the capacity to make informed medical decisions.

Those lacking decisional capacity, such as the seriously mentally ill or individuals with severe cognitive impairment, should not be tested. No substantial clinical or preventive benefit will accrue to them from the tests, at least with our current knowledge. They are also unlikely to gain any psychological benefit from the relief of anxiety about breast or ovarian cancer, which they may not have or which the test may not dispel. Individuals lacking decisional capacity may face disadvantages as a result of testing and, by definition, cannot make a competent decision whether to pursue it. Prophylactic surgery, perhaps the most plausible preventive step that could be taken after a positive test result, seems entirely inappropriate for these individuals.

c. Prenatal testing of embryos or fetuses should be discouraged.

We strongly discourage prenatal testing of embryos or fetuses for BRCA1/2 mutations. There remains uncertainty about the importance of particular mutations. In addition, the penetrance of the gene is not 100%, the age of onset is late, current treatments are often effective, and the efficacy of breast and ovarian cancer treatments in 20 or 30 years is difficult to predict. Children born today with a BRCA1/2 mutation may not face the same risk of disease as their mothers or grandmothers.

The Working Group is split over whether such testing should be allowed at all. Some believe that its value is so low and the consequences, in abortions (probably almost always of female fetuses), are so grave that it should not be allowed. Others favor a broad right of women to choose to carry or not carry pregnancies to term and to get as much information as possible about the fetus in making that decision. We do agree, however, that such testing should not be encouraged.

There are some situations where such testing seems less troubling, even though the same issues apply. Some forms of assisted reproduction allow for, or include, pre-implantation genetic testing of pre-embryos, although these tests are not yet sufficiently accurate to make testing feasible for BRCA1/2 mutations. If it becomes feasible, we do not believe it would be inappropriate to test pre-embryos for BRCA1/2 mutations or to choose to implant only pre-embryos that did not carry mutations.

d. Testing should not be encouraged for women who have been newly diagnosed with breast cancer.

It has been suggested that testing should be recommended for women who have been newly diagnosed with breast cancer. We do not agree. First, it remains unclear what percentage of women diagnosed with breast cancer, even at an early age, will carry a BRCA1/2 mutation. The data collected thus far has not been consistent; the percentage of breast cancer cases with germ line BRCA1/2 mutations may be on the order of 5 to 10 percent, even for young women. Second, the value of a genetic test to a woman diagnosed with breast cancer remains unclear. It is known that one primary breast cancer already increases substantially the risk of a second primary breast cancer; although it is plausible, it is not known whether women with mutated BRCA1/2 genes are at even higher risk for a second primary cancer than women with sporadic tumors. (As our knowledge on these points changes, the groups for whom testing would be appropriate may also change.)

In addition, we recognize that women who have just received a diagnosis of breast cancer are preoccupied with treatment decisions and may be particularly vulnerable. This may not be the ideal time for clinicians to offer, or for women to consider, the option of genetic testing. Given the same choice months or years later, they might make a different decision. Once taken and the results revealed, the test can never be undone; it can always be done later. With our current knowledge, there is no particular reason to think that it is important for a women to receive information on BRCA1/2 testing at all, let alone near in time to her diagnosis of breast cancer.

The exception to our stated recommendation is that testing for BRCA1/2 mutations would be appropriate for women from a high risk family diagnosed with breast and ovarian cancer and may be appropriate for women seriously considering prophylactic surgery. Physicians should discuss such testing with those women. Otherwise, given the current lack of well-understood preventive or treatment options for women with breast cancer who test positive, it is not appropriate to recommend the test for all women diagnosed with breast or ovarian cancer.

3. The Crucial Role Of Genetic Counseling

a. We recommend that high risk families (or individuals at risk) should receive care in a setting where comprehensive services, including genetic counseling, preventive care, and psychological follow-up, can be provided.

Testing for BRCA1/2 mutations is not a simple matter. Adequate care for women considering testing demands a range of skills and knowledge that will probably be wider than any one professional possesses. As a result, we believe that, when possible, individuals considering testing should receive care in a setting that integrates a variety of services. Genetic counseling is the linchpin to such care, but it is not, by itself, sufficient.

b. Testing should not be offered in settings (such as submitting blood tests by mail) where adequate genetic counseling cannot be assured.

The interpretation of information about testing for BRCA1/2 mutations is complicated for several reasons. Individuals may seek testing based on an inaccurate understanding of the significance of their family history. This is particularly true in light of the limited accuracy of information people often have about the nature of the cancers their relatives had. Making the determination of who is "high risk" is a complex task. The risks faced by patients who are tested go beyond the medical risks normally considered in the health system and expected by patients. People need to be warned, by knowledgeable professionals, about the possible psychological, familial, and social consequences of testing. Similarly, they need to be told quite carefully what the tests can, and cannot, reveal about their risks of breast and ovarian cancer and the limited medical interventions possible after a positive test.

After the test, whether the results are positive or negative, the patient needs to be told -- and to understand -- what the results mean. Women who test negative must understand that they still bear the same substantial risk of breast cancer faced by those in the general population. Women who test positive need to understand fully their heightened risks and the options available to them to deal with those risks. In addition, follow-up counseling for individuals testing both negative and positive is essential to determine whether they are having any strong psychological reactions to their test results.

For all these reasons, testing without counseling is inappropriate and may be dangerous. Testing programs should not be allowed unless they include counseling as a mandatory part of the process. Although face-to-face counseling is preferred, in certain situations, such as people living in remote areas, it may be appropriate to provide counseling by telephone.

c. Counseling may be provided by a wide variety of trained staff, including genetic counselors, oncology clinical nurse specialists, oncologists, geneticists, primary case physicians, and others. Whoever provides counseling should have received special training regarding genetic testing for breast and ovarian cancer risk.

Each group of health professionals has unique skills, knowledge and experience to offer individuals and families undergoing genetic testing for BRCA1/2 mutations. We do not believe that any of the specialties involved has a monopoly on the information and skills necessary to provide counseling and hence we do not recommend that counseling be limited to a particular profession or specialty.

We do, however, believe that particular information and skills are essential to counsel people about BRCA1/2 mutation testing. We therefore recommend that such counseling only be performed by professionals who, by virtue of their past training or by their completion of special training courses, have sufficient knowledge and skill to provide good counseling. We urge the creation of continuing education training courses for oncologists, geneticists, OB/GYNs, primary care physicians, genetic counselors, and others to provide them with the background they will need to provide such counseling. Counseling provided by professionals without such specialized training will often fall below the standard of care and must be avoided.

4. The Content of Genetic Counseling

Good genetic counseling for BRCA1/2 mutations must cover a wide range of topics, in a sensitive way. Careful counseling is necessary before the test is offered, both to ascertain whether the patient is in a high risk category and to provide the patient with enough information about the benefits and risks of the testing to allow her to make a truly informed decision. Careful counseling after the test results is necessary both to help the patient react to the information in an informed and useful way and to look for adverse reactions to the results. The body of our report contains detailed recommendations for the specific components counseling should cover (see the Table 8-1 in Chapter 8).

5. Informed Consent

As noted above, one key goal of genetic counseling is to provide the information patients need to make an informed choice about testing. This is particularly critical in testing for BRCA1/2 mutations, where testing does not have any strong clinical benefits and does carry non-clinical risks. People are likely to accept or reject testing based on personal considerations, which will, in turn, be based on information about the possible consequences of testing that may not be obvious to them prior to counseling.

In many states, informed consent requires that a person be given the information customarily provided by reasonable professionals under the circumstances. Other states require that patients be informed of whatever information a reasonable patient would want to know under the circumstances. Under either test, the crucial factors for patients are likely to be personal.

a. Informed consent for BRCA1/2 mutation testing must involve information about non-medical benefits and risks.

The process of informed consent usually focuses on the medical risks and implications of patients' decisions. The medical risks (and benefits) of testing for BRCA1/2 mutations, based on our current knowledge, are very limited. The important implications for people considering testing deal with the possible psychological, familial, and social effects of their test results. It is absolutely essential that, before a patient decides whether to be tested, the counselor review these possible consequences. Pre-test counseling that ignored those risks would be grossly inadequate and should be considered a violation of the standard of care under either version of state laws governing informed consent.

b. The waiver exception to informed consent should be applied very narrowly in the context of testing for BRCA1/2 mutations.

The law of informed consent does contain some exceptions, including medical emergencies, public health emergencies, incompetent patients, and situations where information about the risk would harm the patient. The only exception potentially relevant to genetic testing for BRCA1/2 mutations is waiver. Normally, patients can waive informed consent, by making a conscious, voluntary decision not to be informed. We believe that waiver should be construed very narrowly in this context. The tests have no compelling clinical value and the risks to the tests may well not be obvious to patients. Waiver should be allowed only in extraordinary circumstances, where the provider knows that the patient has already been informed about the risks and benefits of the procedure (such as when the patient is a researcher who has studied BRCA1/2 genetic testing).

c. The informed consent process should remain individual in nature but should encourage the patient to consider the implications of testing for other family members and to discuss the question with them.

Genetic testing of an individual necessarily provides some probabilistic information about that individual's relatives. It might be argued that, given the inherently familial nature of the information, informed consent should come broadly from the affected family, not from the individual.

This view is plausible and may be appropriate in some cultural contexts. It seems to us, however, to conflict too strongly with the individual emphasis of American culture (and American law). Familial considerations, as noted above, must be one part of the process of informed consent. Patients need to understand some of the kinds of tensions that genetic testing for BRCA1/2 mutations might cause within their family. We believe that counselors should, at the same time, inform patients that test results also will confer some information about their relatives. It may show that a parent carries a mutation; it may demonstrate that siblings or children have a fifty percent chance of carrying a mutation. Patients should then be advised that they may wish to discuss their decision whether to be tested with their families. Counselors should be willing to help explain the issues to family members, if the patient requests it.

A minority of the Working Group felt that the unique aspects of genetic information may warrant developing new models of informed consent to allow for family, as opposed to individual, decisions to initiate genetic testing.

6. Any Program of Genetic Testing for BRCA1/2 Mutations Must Include Follow-Up Care

Testing for BRCA1/2 mutations is not an end, but a point in an on-going process of patient care. It is essential that those who consider such testing have access to good follow-up.

Even women who consider testing but decide not to be tested should receive some follow-up. Decisions about testing should be influenced heavily by the current knowledge about the mutations and about prevention, early detection, and treatment of breast or ovarian cancer in women with the mutations. As those facts change, as they almost certainly will, women who in the past began the testing process should be informed of the changes and their possible implications.

Women who decided to be tested and had negative results also need follow-up services. They need to be informed of the meaning of the negative results and to be reminded of the continuing general risks of breast and ovarian cancer. An appropriate follow-up regime would include reminders about breast examinations and mammography as recommended for women of their age at the general population risk for breast cancer. Women who test negative should also be warned about possible negative psychological consequences.

Some women may receive indeterminate test results, for example where they had a mutated BRCA1/2 gene, but the connection of the mutation to higher cancer risk was unclear. As information improves about the meaning of various mutations, those women should be informed of the changed information and its possible consequences for them.

Finally, women who test positive definitely need access to care beyond counseling about their test results. Services aimed at early detection and, if developed, prevention of disease in women with mutations should be available to women who test positive. Psychological counseling and treatment for patients and family members should be available when appropriate.

7. Protections for Patient Privacy Should Be Maintained and Improved

a. The privacy of genetic information, and all medical information, should be better protected.

Genetic information is currently protected like other medical information -- not very well. Stronger protections for patient information, either genetic information or, preferably, medical information in general, should be adopted by the federal or state governments. In addition, all those involved in genetic testing need to be sensitive to heightened patient concerns about confidentiality.

b. Privacy legislation must be strengthened as one part of a strategy to limit discrimination in employment and insurance against people with mutated BRCA1/2 genes.

As discussed below, people considering genetic testing have great concern about their possible loss of employment and insurance opportunities. One way to prevent employers or insurers from making decisions based on genetic information is to prevent them from obtaining that information, in part by putting physicians, researchers, genetic counselors, and others with the information under an obligation not to disclose it.

Legislation to achieve that goal would have to limit not only the ability of employers and insurers to acquire such information without a patient's knowledge, but would need to prevent them from asking patients for the information or asking patients to authorize the release of that information. Such legislation could prove useful and should be given serious consideration. In light of other ways in which some genetic information may reach employers and insurers, however, it would need to be coupled with substantive prohibitions on such discrimination in order to be maximally useful.

c. Researchers must design their research and their embodiment of its results in ways that protect patient privacy.

All agree that further research is essential on the effects of BRCA1/2 mutations on cancer risks. That research needs to collect information on the mutation status of many people and follow their health over many years. Ideally, the resulting data would be widely available to researchers in order to allow many different analyses of the results. These goals imply an accessible computerized database of information on tested people.

The existence of such a data base automatically raises concerns about the confidentiality of information within it. That confidentiality obviously would be compromised if the data base contained personally identifying information, such as names or social security numbers. It would also be threatened, perhaps less obviously, by full background information on research participants. An anonymous entry on, for example, a white woman, aged 49, born in Illinois and living in a particular section of the District of Columbia, with a professional education and one teenage child, might effectively identify Hillary Rodham Clinton, even without a name or street address. The same issues arise in more limited research.

All researchers need to be extremely sensitive to the confidentiality of the information they gather. The research needs to be designed to minimize the risks that confidentiality will be breached. When appropriate, this should include measures that provide some protection even against court orders, through Public Health Service "Certificates of Confidentiality." Patients should be fully informed of the protections for confidentiality -- and the risks that it may be breached before they decide whether to participate in the research.

d. Even within the family, physicians and counselors should maintain privacy rights.

An unusual aspect of genetic information is the extent to which information about one person conveys information about another. If one person carries the mutation, her siblings and children have a fifty percent chance of carrying it. If a parent from a high risk family does not have the mutation, her children are very unlikely to carry it. If a child in such a family carries the mutation, one of her parents almost certainly carries it.

It can be argued that, just as a psychotherapist with a violent patient may have an obligation to warn a targeted person of her danger, a physician for a family member who receives genetic testing should have an obligation to tell other family members of important information about them revealed by the test. With the current state of knowledge about prevention or early detection of breast or ovarian cancer, we see no good justification for such disclosure. The risks to the other family members are not immediate and not subject to substantial change as a result of genetic test information. This recommendation applies not only to the results of genetic tests, but whether an individual has taken a genetic test.

Physicians may want to encourage their patients to tell affected family members something about the tests or to suggest that they receive genetic counseling, but physicians must not take that step themselves without the patients' consent. This recommendation may change if options for prevention, early detection, or treatment of breast or ovarian cancer in women with BRCA1/2 mutations improve.

The tested person is not the only person whose privacy should be upheld. Family members should have a right to refuse to receive information that is relevant to their BRCA1/2 status, even if the tested person consents. Physicians should not provide that information to a family member without the informed consent of the person receiving the information. Patients should be warned that providing information that informs their family members that they have a mutated gene or have a fifty percent chance of having a mutated gene may have serious consequences for the family member.

8. Liability Issues In Testing For BRCA1/2 Mutations

a. In general, with current knowledge of BRCA1/2 mutations and of breast cancer prevention and treatment, malpractice liability for failing to offer genetic testing should not be used as a reason for doctors to offer such testing broadly.

A physician is liable for malpractice only if she has fallen below the standard of care in her treatment of a patient. The standard of care is generally a question of fact: what would other reasonable physicians do under similar circumstances? This issue is usually decided by a jury, based on expert testimony and other evidence of what physicians normally do in similar circumstances.

At the present time, there is no justification for a reasonable physician to encourage testing for BRCA1/2 mutations in anyone. Based on the risks and the possible benefits, physicians might be expected to discuss BRCA1/2 mutation testing with women from high risk families or women giving serious consideration to prophylactic surgeries. Physicians should not be expected even to offer or to raise the issue of such testing for individuals without demonstrated family risk. This standard of care, based on current medical knowledge, should be applied by physicians and should govern malpractice claims.

Even when a plaintiff is able to show that the standard of care has been breached, she must still demonstrate that the breach of the standard of care harmed her. Often, this question of causation is not difficult; the negligent removal of the wrong limb clearly causes damage. The issue would be much more complicated with BRCA1/2 mutation testing. Women who claimed that they should have been offered the test would have to show that a) if offered the test they would have taken it, and b) if they had taken the test, they would have taken other steps that would have prevented them from being harmed. Based on our current knowledge of prevention, early detection, and treatment for breast and ovarian cancer in women with mutated BRCA1/2 genes, it will be rare that a woman could prove such a claim. Therefore, for both reasons of the standard of care and proof of causation, fear of malpractice litigation does not justify encouraging or broadly offering BRCA1/2 mutation testing. Of course, if prevention, detection, or treatment of breast or ovarian cancer turn out to be substantially improved by knowledge of a BRCA1/2 mutation, this standard of care may well change.

b. Professional organizations and others should continue to draft and refine guidelines about whether and when such testing is appropriate.

Statements by professional organizations and other authorities are useful as evidence of the standard of care, although they usually do not legally determine that standard. Such statements provide some additional guidance for juries and assurance to physicians that a proper standard of care will be applied in court. For purposes of preventing malpractice litigation, as well as for purposes of informing and guiding physicians and patients, authoritative guidelines for BRCA1/2 mutation testing should be developed by appropriate bodies. As information about the mutated genes and cancer risks changes, those guidelines should be updated.

c. When testing for BRCA1/2 mutations is offered, the use of genetic counseling in connection with the testing should become part of the "standard of care."

For the reasons noted in Specific Recommendation #3, genetic counseling is absolutely vital to testing for BRCA1/2 mutations, both to help the patient in the process of informed consent and to help the patient usefully deal with the test results. This counseling is so important that testing offered without counseling should be viewed as falling below the standard of care and inherently negligent. Should any harm come to a patient as a result of the lack of adequate counseling, the patient may well have a valid malpractice claim.

d. When such testing is offered, practitioners without substantial knowledge or experience in genetics may well risk violating the standard of care.

Similarly, as noted above, counseling about the risks of BRCA1/2 mutations (and mutation testing) involves special skills and knowledge. We have recommended that the tests be allowed to be offered only in conjunction with counseling from properly trained professionals. Even if that recommendation is not adopted by legislatures, regulators, clinics, or health plans, it may be effectively enforced through the liability system. People not competent to provide genetic counseling should not offer genetic testing. A person harmed as a result of inadequate counseling from an unqualified counselor may well have an excellent malpractice claim.

e. When such testing is offered, in light of some evidence from other genetic tests of substantial psychological effects from testing, access to psychological counseling for those who receive genetic testing should be considered part of the standard of care.

People sometimes have strong and occasionally unpredictable psychological reactions to the results of genetic tests. Those who provide such tests need to be on guard for the possibility of such reactions; those who take such tests should have access, where appropriate, to psychological counseling or treatment. It should be negligent to offer testing without a process that looks for such reactions in patients and family members, and provides access to counseling or treatment when necessary. A patient injured as a result of such a negligent process may well have a strong malpractice claim.

9. Ownership Issues In Genetic Testing For BRCA1/2 Mutations.

a. Congress should consider the possible implications of a patent monopoly resulting from a patent on an important human genetic sequence.

The issue of the patenting of genetic material, information, or "inventions" is controversial. Patents are praised as making the discovery and development of useful pharmaceuticals possible. They are attacked on several grounds, including the breadth of the conferred monopoly, the question whether gene discovery is really an "invention," and the hubris involved in patenting aspects of life. The present system of genetics-related patenting in the United States followed from earlier practice concerning organic chemistry without special legislative action. We do not have a position as a group on the proper stance for patent laws to take with respect to genetic "inventions," but we do believe that the competing interests involved in gene patenting should be considered on their own merits. Thus, we think that Congress should examine the issues involved in genetics-related patents.

b. A patent should not be allowed to be a significant barrier to research in genetic links to human disease.

The commercial monopoly provided by a patent raises one set of issues; its potential for controlling research raises different and, we believe, clearer issues. Research into human genetics, including particularly into BRCA1/2 mutations, is too important to be controlled by those who control patents. Patent owners should not limit substantially research on these genes. If the patent owners were so shortsighted as to try to do so, their efforts should be resisted. One possible legal course would be through the exploration of some discussions in American patent cases of a "research exemption" from patent infringement. Other possibilities would be through commercial or social sanctions.

c. When the federal government is an assignee on a patent application, its agreements with other assignees or co-inventors should be public and should protect the public interests in the uses of the invention.

In many cases, the federal government provides substantial funding for research that leads to patents. Since 1980, Congressionally-mandated federal policy has made it easy for universities and other non-profit organizations to patent inventions made, in part, with federal funds. When, however, the federal government and its employees play an important and direct role in the invention, the federal government has the right to be one of the assignees of the patent. The rights and duties of multiple assignees to patents are often defined in agreements among them.

Federal employees were listed as co-inventors on the patent applications for BRCA1 and BRCA2 and the federal government is thus an assignee on the patent application. The federal government's agreements with other assignees concerning their respective rights in patents should be public and should provide for the protection of the public's interests. In the case of BRCA1 and BRCA2, the federal government should endeavor to protect the research uses of the gene and to protect the public from the possible abuse of the patent by the other assignees.

10. Regulation Of Testing For BRCA1/2 Mutations

a. The federal government should regulate the commercial use of genetic tests based on the safety and efficacy of the testing.

b. This regulation probably needs to take a form somewhat different from FDA regulation of drugs or medical devices.

c. Such regulation should impose substantial data collection and research requirements after interim approval of a test.

d. Such regulation should limit the use of such tests to qualified practitioners.

e. The federal government should regulate the research use of genetic tests along the lines of regulation of research on experimental drugs.

f. The regulation of clinical laboratories, by federal and state governments, needs to test the abilities of those laboratories to perform genetic testing adequately.

g. Marketing of testing should not prey on women's anxieties about breast cancer.

11. Financing Testing And Follow-Up For BRCA1/2 Mutations

a. Where genetic testing for breast cancer predisposition is medically appropriate, it should be reimbursed by health coverage in situations where it is medically appropriate, just as should any other accepted medical service.

b. With current knowledge, such testing may be medically appropriate for some individuals in high risk families; there is no reason to believe that it is medically appropriate for people who do not have high risk genetic backgrounds.

c. Health insurers, HMOs, and self-insured employer plans ("payers") should cover genetic counseling for all people for whom they cover testing.

d. Payors should also cover at least some genetic counseling for all women concerned about whether genetic testing is appropriate for them.

e. Payors should cover psychological counseling when made medically appropriate by a patient's or family member's reaction to genetic testing.

f. Payors should participate in properly planned research on genetic testing for breast or ovarian cancer susceptibility.

12. Employment And Insurance Discrimination Issues

a. The federal government should clearly ban employment discrimination on the basis of genetic information that is not manifest in the employee's condition or behavior.

b. Discrimination on the basis of conditions or behaviors linked to genetic information shall be prohibited or allowed as provided under existing laws, such as the Americans with Disabilities Act.

c. Before federal legislation is effective, or where federal legislation does not reach, states should ban such employment discrimination.

d. The federal government should require that the availability or price of health coverage should not be conditioned, limited, or affected by genetic information.

e. Preferably, such a requirement would be in the context of guaranteed health coverage for all Americans.

f. Before federal legislation is effective, or where federal legislation does not reach, states should ban such activity.

g. With respect to life and disability insurance, the federal and state governments should assess each type of insurance individually, considering both the importance of the coverage, the context in which it is usually provided, and the significance of genetic test results for the insurers.

h. The states or the federal government may want to provide for guaranteed availability of some minimum level or type of life and disability insurance without regard to genetic information.

i. Genetic information might also be used in making determinations about child custody or adoption. Given the uncertainties associated with BRCA1/2 mutation testing, the use of such information in these contexts seems very likely to be inappropriate, but the Working Group has not considered these issues in detail. Further study would be appropriate.

13. Social Education And Outreach

a. Existing breast cancer awareness outreach and education programs for patients, their families, and other potential consumers should be expanded to include information on genetic testing for hereditary susceptibility to breast and ovarian cancer.

b. Existing breast cancer outreach and education programs for health practitioners should be expanded, and new, well-coordinated programs initiated to enable clinicians, health care providers, and other health professionals to learn about and discuss the ethical, legal, social and medical issues around genetic testing for breast and ovarian cancer susceptibility with their patients.

c. Programs that educate policy makers, legislators, and other legal professionals about the ethical, legal, and social implications of genetic testing should be encouraged.

d. Efforts to heighten the level of scientific literacy and educate the public about the work of the Human Genome Project, its possible applications and consequences, should continue.

e. Educational partnerships and collaborations between agencies, and between professions, should be sought and nurtured.

f. Resources providing technical assistance in setting up and maintaining outreach and education programs should be established and publicized.

g. All the education programs around BRCA1/2 mutation testing need to address, actively and visibly, the special needs and perspectives of women.

 

Stanford Medicine Resources:

Footer Links: