Genetic Testing for Cancer Risk: Title and subTitle BreakHow to Reconcile the Conflicts

GENETIC TESTING for cancer susceptibility is moving from the research laboratory to the clinical domain and may become a useful tool for public health. However, clinicians and patients must balance potential gains in cancer prevention and control with the risks of testing and screening.

Recent policy directives, which conflict with one another, reflect the promise and the peril of this cancer genetics technology. In this article, using breast cancer predisposition testing as an example, we describe 3 such statements from organizations selected because they represent researchers, oncologists, and consumers. Because practicing clinicians will encounter patients who request testing, we conclude with suggestions to help clinicians understand these conflicting positions and respond to patients.

Breast cancer is the most common form of cancer in American women.¹ Approximately 5% to 10% of the 180000 new cases each year are related to 1 of several genes inherited in an autosomal-dominant fashion.² The breast-ovarian syndrome is a familial breast cancer syndrome caused by mutations in either of 2 susceptibility genes, the breast cancer 1 (BRCA1)³ and breast cancer 2 (BRCA2)⁴ genes. Because BRCA1 and BRCA2 were only recently identified, there is a paucity of population-based clinical data on known gene carriers. While initial studies showed that the cumulative lifetime breast cancer risk for putative female BRCA1 carriers was as high as 85% by age 85 years,⁵ more recent data suggest that the actual penetrance may be somewhat lower, at least in some populations.⁶

Genetic screening might contribute to cancer control by identifying a cohort of at-risk patients who could benefit from augmented screening or prophylactic measures. However, the nature and frequency of appropriate surveillance measures and the efficacy of interventions based on genetic information have not yet been determined. Prophylactic strategies, including medical options such as chemoprophylaxis, and preemptive surgical approaches such as bilateral mastectomy or oophorectomy may provide benefits to those with positive genetic test results.⁷ Unfortunately, these measures also carry their own risks; careful study and significant follow-up time will be required to demonstrate their utility.⁸

There are several genetic and epidemiologic issues that limit potential cancer control benefits. First, inherited forms of breast cancer may be caused by several distinct genes. This genetic heterogeneity, or many genes causing a disease process, is further complicated by allelic heterogeneity, where many mutations can occur at a single locus. For example, over 200 mutations have been identified in the coding regions of BRCA1 and BRCA2. Second, the predictive value of current testing for BRCA1 and BRCA2 has not been established. The lifetime risk for developing breast or ovarian cancer is extremely high in female members of BRCA-linked families who have multiple members with breast and ovarian cancer. However, the cancer risk for gene carriers who do not have a strong family history is less clear. Third, the natural history of malignancies caused by germline mutations in cancer susceptibility alleles is different from sporadic cancers,⁹ so that these patients may have an earlier average age of onset or a better prognosis than those with sporadic forms of the disease.¹⁰ Finally, because breast cancer susceptibility genes are associated with malignancies of other organs, a cancer control strategy based on predictive genetic testing must also provide surveillance measures for a multiplicity of primary sites. Despite these limitations, genetic screening holds promise because it may identify patients at risk prior to the actual development of cancer. If this identification can be coupled with interventions to reduce morbidity and mortality, genetic screening could contribute to improved cancer control. Together with other cancer susceptibility genes, which have been, or will soon be, identified, genetic screening may bring substantial public health benefits.¹¹

However, these benefits must be weighed against the risks that accompany genetic susceptibility testing for cancer. While there are no direct medical risks, the psychosocial and economic consequences of learning one's genetic status are significant. These risks may include psychological morbidity such as anxiety and depression,¹² risks of misinformation such as false reassurance or unwarranted prophylactic measures,¹³ the potential for breaches of confidentiality,¹⁴ and the risk of genetic discrimination.¹⁵ In 1 group of patients, those with health insurance were more likely than those without to request BRCA1 test results, suggesting the impact of economic considerations on patients' decisions.¹⁶ Genetic testing may also create tension within families, and even those who test negative may experience survivor guilt.¹⁷ These nonmedical risks may be important to patients, so they cannot be ignored.

The potential of DNA-based genetic tests to predict cancer susceptibility and the controversy surrounding possible risks associated with testing have prompted various organizations to issue statements about this emerging technology. These clinical and policy discussions address 2 major questions: (1) Who should undergo genetic testing or screening? (2) What are the clinical implications of discovering an inherited predisposition to cancer? A recent consensus statement provides guidance with regard to the latter issue.¹⁸ To examine the former question, we review positions published by the National Advisory Council for Human Genome Research (NACHGR) of the National Institutes of Health (NIH), the American Society of Clinical Oncology (ASCO), and the National Breast Cancer Coalition (NBCC).

In 1994, the year that the first major breast cancer susceptibility gene was discovered,³ a "Statement on the Use of DNA Testing for Presymptomatic Identification of Cancer Risk" was prepared by the NACHGR.¹⁹ This policy statement, endorsed by the NIH Department of Energy (NIH-DOE) Ethical, Legal, and Social Implications Working Group, takes a restrictive position on the question of clinical implementation. The NACHGR called attention to 5 issues and concluded that "until more information is available to address these critical issues, it is premature to offer DNA testing or screening for cancer predisposition outside a carefully monitored research environment."¹⁹ In 1996, Dr Francis Collins, lead author of the NACHGR policy statement and director of the National Human Genome Research Institute (NHGRI), reiterated the conclusion that genetic testing for cancer predisposition should "remain a research activity."²⁰

The American Society of Clinical Oncology (ASCO), which represents practicing oncologists, published a statement in 1996 that conflicts with the NACHGR position.²¹ While encouraging further research, ASCO recommends that "genetic testing should be made available to selected patients as part of the preventive oncologic care of families. . . . " The statement sets forth practice guidelines stating that the indications for testing should be (1) a strong family medical history of cancer or onset of disease at an early age, (2) a test that can be adequately interpreted, and (3) results that will influence the medical management of the patient or family member. The ASCO statement introduces a typology of cancer gene tests. Group 1 test results demonstrate the presence, or absence, of a genetic mutation and provide unequivocal clinical benefit. With Group 3 tests, clinical benefit has not been established. Group 2 tests, which include genetic testing for breast-ovarian syndrome, apply to those syndromes where the "medical benefit of the identification of a ‘carrier' is presumed but not proven."²¹ Yet, the ASCO document suggests that "oncologists should consider offering genetic testing only to the first 2 categories (ie, Groups 1 and 2), . . . and that testing for Group 3 is considered research with unknown clinical implications, and should not be offered in a clinical setting." In contrast to the NACHGR position, ASCO suggests that genetic testing for breast-ovarian syndrome should no longer be restricted to the research setting and is now appropriate for clinical use.

A third important perspective comes from the National Breast Cancer Coalition (NBCC), which is a self-described "grassroots advocacy organization dedicated to the eradication of the breast cancer epidemic." The NBCC, a consumer coalition of 350 organizational and 41000 individual members, issued a 1996 position paper calling for more research about genetic testing. Citing inadequate knowledge of sensitivity, specificity, reliability, and problems with genetic education and counseling, the NBCC suggests that "genetic testing should only be available within peer-reviewed research protocols." The statement goes on to urge patients to participate in such studies, asks the scientific community to make studies widely available, and requests that the Food and Drug Administration prohibit commercially available testing.

What accounts for the variation in these statements? Scientific, economic, and sociocultural pressures have each affected policy development in this area. Moreover, self-interest often plays an important role in policy development and genetic testing for cancer susceptibility is no exception. The NIH-sponsored NACHGR is a major research institute, while ASCO is a professional group concerned with clinical issues. While researchers may worry about the potential for misuse of the technology they develop, clinicians encounter patients actively seeking genetic testing. The NACHGR position is to withhold genetic testing from clinical practice, thereby denying it to patients except in carefully controlled research settings. ASCO wants to permit testing if the physician and patient agree on its potential medical benefit, and if the test meets the ASCO criteria. It is significant, then, that the consumer voice represented by the NBCC more closely resembles the restrictive NACHGR position than the permissive ASCO position. Supporters of ASCO's position may criticize the NACHGR's approach as paternalistic, but NBCC support of a restrictive policy constitutes a strong refutation of this critique; the consumers represented by NBCC share many of the NACHGR concerns. If the NBCC position paper leaves any uncertainty, a subsequent commentary on the ASCO statement makes it quite clear. "The NBCC . . . applauds ASCO for its attempt to limit the use of genetic testing by the medical community. But ASCO does not go far enough. . . . Under existing circumstances, genetic testing . . . outside of quality research protocols is harmful to your patients."²²

Genetic assessment for cancer predisposition may indeed confer significant psychosocial and economic risks, as noted above. Sound policy must balance those risks with the potential benefits to patients. The risks and benefits must be determined in particular cases by patients, with the help of knowledgeable clinicians. We suggest 4 basic principles to guide clinicians in understanding these conflicts and to facilatate their response to patients who request genetic testing for cancer predisposition.

First, clinicians must educate themselves. A recent review of methods for effective clinician education in genetics suggests that familiar strategies such as bedside consultation, case reports, and traditional continuing medical education can be adapted for this purpose.²³ One helpful resource entitled "Understanding Gene Testing" is available from the National Cancer Institute.²⁴ Clinicians should also recognize their own limitations in this rapidly changing field and make referrals to specialists in cancer genetics and genetic counseling when appropriate. This referral pattern may produce a learning curve, where primary care clinicians assist their patients and educate themselves through consultation, gradually decreasing the need for such referral as they become more familiar with cancer genetics. In this way, clinicians will be better able to help their patients understand the implications of genetic testing for cancer risk.

Second, clinicians and patients should engage in meaningful informed consent discussions about genetic testing. Whether the discussion is prompted by a patient's request for testing or is initiated by the clinician based on family history, a thorough consent process must take place prior to any decision about testing. When initiated by the patient, clinicians need to understand the reason for a patient's interest as they evaluate and discuss the request. When initiated by the clinician, patients need to understand the professional bias and perspective from which the recommendation is made. Guidance about specific elements of informed consent required in this context has recently been published.²⁵ When the current limitations, along with the potential risks and benefits, are clearly explained, patients may alter their initial decision about genetic testing.

Third, understanding the tensions that underlie conflicting positions will help clinicians to better serve their patients. While NBCC advocates an expanded research agenda and NACHGR calls for restricting genetic testing to research, ASCO encourages research but argues that access to testing should not be restricted. Further research in cancer genetics will benefit society, but restricting testing to the research context may harm individuals. Such restriction may also conflict with the fundamental notion of voluntary research participation. Should patients who do not wish to become research subjects have access to clinical genetic testing services if the clinician and patient agree that the benefits outweigh the risks? Is the ASCO position also too restrictive by requiring an effect on the medical management to result from the test? In some cases, simple knowledge of one's genetic risk, or its impact in the private realm of reproductive decision making, may be sufficient to justify an individual's decision to undergo testing. Alternatively, the combination of inherent risks to patients, limitations in proven benefits, and the need for better understanding of cancer genetics through research may support continued restriction. Clinicians who understand these tensions can evaluate the merits of the conflicting positions and better assist patients who are candidates for genetic susceptibility testing.

Fourth, because of the important distinction between testing and screening, clinicians should resist pressure to screen until data support the benefits of such a program. While the goal of clinical testing is improved individual health, screening programs aim to improve public health. By definition, a screening program targets a population to reduce the morbidity and mortality from a particular cancer among individuals screened.²⁶ The critical difference between genetic susceptibility testing and screening is that testing applies to those with cancer or a significant family history, while screening would assess many individuals without a personal, or family, history of cancer. With regard to screening, the NACHGR and NBCC positions seem appropriately restrictive: the current public health benefits of further research probably outweigh the potential benefits of screening, and the current risks of screening outside the research context are considerable. The robust informed consent necessary to maximize patient autonomy would be nearly impossible in widespread, population-based screening programs, and the risks of genetic discrimination would be amplified. Despite these concerns, and the considerable cost of the test, clinicians are now subject to aggressive marketing by laboratories, which encourage testing and screening of patients. Although questions about testing remain controversial, we suggest that the use of genetic screening for cancer is, at best, premature.

Position statements may influence the encounter, but clinical decisions about genetic testing will ultimately be negotiated between patient and physician. Legislators, industry leaders, researchers, clinicians, and patients must attend to the important distinctions between research and practice, and between diagnostic testing, and screening, as genetic technology advances make their way from the research laboratory to the clinical and public health arenas.