Myriad Genetic Laboratories, Salt Lake City, Utah 84108
Address correspondence and reprint requests to: Dr. Thomas S. Frank, Medical Director, Myriad Genetic Laboratories, 320 Wakara Way, Salt Lake City, UT 84108.
Consider a test result that indicates an increased risk of cancer. What if the amount of risk were not known? What if it were unclear which intervention would most effectively reduce the risk of subsequent invasive cancer? What if the result itself were subject to a great deal of interlaboratory variation? Should such a test result be used to make medical management decisions?
Such tests have been in routine use in medical care for decades. For instance, the diagnosis of ductal carcinoma in situ (DCIS) of the breast is essentially the identification of a microscopic marker of increased risk of breast cancer. Even though only a minority of incompletely or untreated intraductal carcinomas progress to invasive carcinoma (1, 2), and even though the diagnosis itself is subjective and often unreliable (3), DCIS constitutes a well-established medical call to action for nearly 40,000 women each year in the U.S.
The identification of cancer risk through tissue diagnosis is well-established and familiar to physicians. Through routine exposure to glass slides and photomicrographs of neoplastic cells it becomes tempting to conceptualize cancer as a disease of abnormal cellular appearance. Cancer is actually the direct result of sequential accumulation of mutations in genes required for normal growth and differentiation. Although malignancy is truly a genetic disease, the epiphenomenon of altered microscopic appearance remains to many physicians a far more comfortable basis of medical decision-making than the more abstract concept of underlying genetic mutations.
Although genetic tests for hereditary cancer risk have provided valuable information to many patients and physicians, they have also been the source of controversy and consternation. A frequently stated concern is that such tests indicate an increased likelihood but not a certainty of developing cancer, and that not enough is known about the actual cancer risk conferred by mutations. It is pointed out, for instance, that while one study of women in high-risk families showed that mutations in the genes BRCA1 and BRCA2 confer a 50% risk of breast cancer by age 50 (4), another population-based study concluded that this risk was "only" 33% (5). For the patient who has a 3350% chance of breast cancer before age 50, this range of risk would likely produce similar choices regarding health care options. Nonetheless, citing such differences in risk estimates, Dr. Bernadine Healy (6) commented that people who use genetic tests to identify hereditary susceptibility to breast and ovarian cancer are "fortunetellers ... reading a pretty cloudy crystal ball."
In comparison, few physicians could cite the percentage of women with DCIS who develop invasive breast cancer [which in most studies is less than 50% (1)]. Similar or greater uncertainties exist for histological indicators of increased cancer risk such as carcinoma in situ, dysplasia, and atypical hyperplasia of various tissues. Every day in the practice of medicine, nonetheless, thousands of surgical interventions are prompted by such diagnoses.
Physicians therefore commonly use test results that indicate an increased risk of cancer, but most such tests do so indirectly by examining cells rather than genes. Genetic tests are now available to definitively identify many hereditary cancer syndromes. The risks of cancer associated with positive results of such tests are in fact better characterized than those associated with many of the tissue-based abnormalities that physicians routinely act upon. It does not seem, therefore, that uncertainty regarding risk estimates could or should be the true basis of the distrust held by some physicians for genetic tests that identify hereditary cancer risk.
Another concern expressed by some physicians is that testing for hereditary cancer susceptibility might jeopardize an asymptomatic patients access to health insurance. Despite (or because of) extensive discussion of this issue, it has become apparent that the perceptions of discrimination are greatly out of proportion to the reality (7, 8). The Health Insurance Portability and Accountability Act of 1996 (HIPAA) guarantees access to health insurance coverage regardless of health status and pre-existing conditions for individuals in or entering group insurance plans and specifically precludes the use of genetic information to demonstrate a pre-existing condition in otherwise healthy individuals. Although such explicit protections have not been in place for individuals undergoing other medical tests, few physicians have advised women to pay personally for Pap smears, dilatation and curettage, or breast biopsies to prevent an insurance company from learning of these procedures. There is at present no evidence that genetic testing for cancer risk in healthy individuals is accompanied by discrimination in health insurance or employment. Although there are no specific protections regarding other forms of insurance such as life and disability, neither are such protections in place for any medical test that indicates an increased risk of subsequent disease.
Concerns about insurance discrimination are often based upon reports of adverse insurance consequences of other genetic diseases such as Huntington disease. Such reports, although anecdotal, have documented that some patients with serious genetic illness have concerns about obtaining or changing insurance and employment (9). However, the experiences of individuals with Huntington disease (which is incurable and invariably fatal) cannot be generalized to individuals with hereditary cancer risk simply because both are genetic. As the Human Genome Project continues to demonstrate the genetic contribution to most diseases, it is increasingly impossible to characterize genetic conditions as a specific class of illness. Generalizations about the ramifications of genetic disease are as invalid as comparable statements about the consequences of having an infection or even a neoplasm. It is far more appropriate to consider the consequences of tests for cancer risk as a whole (whether hereditary or sporadic) rather than to consider all diseases with a genetic contribution as a single entity.
Some physicians may dispute the characterization of a diagnosis such as DCIS as an indicator of cancer risk rather than a biological precancer. If DCIS were to represent an actual early form of cancer [which is speculation, not fact (10)], then the 40,000 patients diagnosed annually with DCIS would have more reason to be counseled about the potential impact on insurance and employment than patients with a demonstrated risk of future cancer. Such discussions in fact do not routinely occur.
Upon scrutiny, many concerns about genetic tests for cancer risk have less to do with identifying cancer risk and more to do with genetics. In fact, developments in genetics are frequently met with concern by society as a whole, of which physicians are, of course, a part. In the mid-1970s, for example, a 5-yr moratorium was placed on recombinant DNA research, primarily at the instigation of the scientific community itself. The concerns were based on conjectural rather than actual risks. The fears of recombinant DNA-induced diseases were soon recognized as greatly exaggerated, and such prospects were even likened by some scientists to concern over "UFOs or witches" (11). The same genetic tools that were banned as unacceptably hazardous fewer than 20 yr ago are now routinely used in student laboratories. A more recent example of public reaction to a genetic advance occurred in 1998 after the announcement of Dolly, the cloned sheep. The federal government deferred acting on numerous health issues affecting millions in order to swiftly draft legislation to address the hypothetical threat of human cloning.
Advances in genetics are thus often received with societal uneasiness. In an article in Newsweek, science writer Sharon Begley recently observed that "genetics is not just a science, let alone a technology or a business. Genetics is a profound idea." (December 28 1998, vol. 133, No.26, page 68). Shortly thereafter, in a special issue of Time magazine devoted to genetics, Walter Isaacson predicted that we are entering a "century of biotechnology," in which "well be able to alter our DNA radically, encoding our visions and vanities while concocting new life-forms." (January 11 1999, vol. 153, No.1, page 43) One of many polls presented in that issue revealed that 58% of respondents would not buy food labeled as "genetically engineered," even though such techniques offer distinct advantages over traditional (and inefficient) genetic manipulation through selective breeding. This reflects an attitude towards genetics not as a useful tool for mankind but rather as an unwelcome interference with nature.
The emerging field of cancer genetics has thus been described by some detractors as a "Pandoras box," referring of course to the ancient Greek myth of the evil-filled vessel that was presented as a gift by Zeus to Pandoras husband. It should be noted that this punishment was actually intended for Prometheus, who had committed an unpardonable offense against heaven. His crime was to share with man the secret of life, namely fire. The myth of Prometheus is one of many stories that reflect our discomfort with unlocking long-concealed secrets of nature, and such uneasiness may be the basis of much of the controversy over genetic tests of hereditary cancer risk. As human beings, we fear that we will somehow be punished if we explore certain aspects of ourselves. Sharon Begley wrote that, because cancer is genetic, it occurs "because of who you are ... cancer becomes an expression of our essential nature." In this view, genes are not merely chemical blueprints for proteins but rather the "unconsciousness" of our biological essence, a part of our true selves over which we have no control. Certainly societal discomfort with genetic evaluation has been exacerbated by proponents of the pseudo-science of "eugenics" who have made false claims to the practice of genetics in order to justify their bigotry. Given these attitudes towards genetics by society as a whole, it is not surprising that many physicians and patients tend to place genetic tests in a category separate from other medical tests even when the information they provide is comparable.
The scientific practice of medicine, however, was founded on the principle that patients with phenotypically expressed differences respond differently to medical interventions. The basis of such differences is in fact genetic, and now we have Prometheus giftwe can examine genes to identify hereditary cancer risk biologically, not just statistically. Genetic testing for hereditary cancer risk in a clinical setting is justified by the ability to identify appropriate patients for testing, the accuracy of laboratory analysis of these genes, and the medical utility of the information (12).
This does not suggest that such tests should be ordered indiscriminately for all persons with a family history of cancer. As with other medical tests, the benefit of the information for the care of the individual patient must be evaluated. It is equally important, however, that health care professionals provide an accurate and objective assessment of the benefits and limitations of genetic tests so that patients are not, in effect, denied access to potentially important information.
The benefit of hereditary risk assessment is inextricably linked to its value in guiding medical management for an individual patient. Although (as in all aspects of medicine) these management options now available will undoubtedly be improved by future research, there are currently several recommendations for individuals with hereditary cancer risk. For women who carry mutations in BRCA1 and BRCA2, for example, a national task force organized by the National Human Genome Research Institute has specifically recommended annual or semiannual clinician breast examinations as well as annual mammography to commence before age 35 (13). Breast cancer risk in high-risk women can be reduced through selective estrogen receptor modulators (SERMs) such as tamoxifen (14), and prophylactic mastectomy has also been shown to reduce the risk of breast cancer in high-risk women by 90% (15).
Given the efficacy of surveillance for breast cancer, prophylactic surgery is understandably chosen by only a minority of women with mutations in BRCA1 and BRCA2 (16). Those women who do consider prophylactic mastectomy because of their family history alone are ill-served by an absence of consideration of genetic testing. This is because mutations in BRCA1 or BRCA2 confer increased cancer susceptibility as an autosomal dominant disorder, so each individual in the family has only a 50-50 chance of having inherited the predisposing mutation and thus the accompanying cancer risk. Without knowing which individuals inherited the mutation, every member of the family must be assumed to be at increased risk. By identifying the familial mutation through genetic analysis it is then possible to identify those women who did not inherit this predisposing mutation and thus are not at increased risk themselves (17). Thus, genetic testing will likely decrease the incidence of prophylactic mastectomy, as it is most commonly performed on women based solely upon their family history, many of whom are truly only at population risk.
Aside from identifying hereditary risk of breast cancer, mutations in BRCA1 and BRCA2 also confer a greatly increased risk of ovarian cancer. Unlike breast cancer, which can be usually be detected at an early stage by widely-employed screening methods, ovarian cancer is usually not diagnosed until it is advanced and relatively incurable. In fact, screening for ovarian cancer is difficult and often ineffective. By identifying hereditary risk of this disease, physicians may offer specific interventions that would not be appropriate for women in the general population. For example, an NIH consensus development panel on ovarian cancer has concluded that "the risk of ovarian cancer from families with hereditary ovarian cancer syndromes ... is sufficiently high to recommend prophylactic oophorectomy in these women at age 35 yr of age or after child-bearing is completed" (18). In addition, oral-contraceptive use has recently been shown to reduce the risk of ovarian cancer specifically in women with mutations in BRCA1 and BRCA2 (19). These illustrate the availability of medical management options for individuals with hereditary cancer risk. Genetic testing for other cancer syndromes, including hereditary nonpolyposis colon cancer and multiple endocrine neoplasia type 2, can similarly guide medical management.
All medical tests, including genetic tests, have limitations as well as benefits. Hereditary risk assessment can be expensive, but often no more so than tissue examination, and both are usually covered by insurance. In our experience with over two hundred submissions for insurance coverage of BRCA1 and BRCA2 analysis, over 94% of claims have been reimbursed by insurers (20).
Patients should be informed that the results of genetic cancer risk assessment may not be definitive. Although a positive test (i.e. identification of a gene mutation that impairs protein function) indicates an increased risk of cancer, the clinical significance of some gene alterations is not yet known. In addition, the meaning of a negative test result may vary depending upon the circumstances. Such results are not unique to genetic tests, however; there are many examples of comparable or greater potential ambiguities among other medical tests. This demonstrates the need for a physician to be able to adequately interpret cancer predisposition tests before offering them (12).
Whereas other diagnostic tests related to cancer may impact the statistical risk for family members, the identification of a specific mutation responsible for hereditary susceptibility to malignancy has direct implications for identifying individual risks of the relatives of the patient being tested. For this reason, and because of medical and psychosocial issues that may accompany identification of hereditary cancer risk, individuals being assessed for hereditary cancer syndromes should be offered an unhurried and thorough discussion of the relevant issues. Genetic counselors and other health care professionals are specifically trained to assist in this process, although genetic counseling may represent an added expense not covered by some insurance plans.
Genetic evaluation of hereditary cancer risk is not a population screen but rather a test for a clinically significant condition that should be used selectively and appropriately. Because published information about hereditary cancer susceptibility has outpaced the ability of many physicians to keep current, some effort is required to catch up to the numerous recent advances in the field. Such challenges are no different for physicians in other areas of medicine that have recently undergone a rapid expansion of knowledge. Those physicians who do not have the expertise or interest in the use of such tests should be prepared to refer appropriate patients to specialists who do.
Genetic testing for hereditary cancer syndromes can assist physicians in identifying and managing at-risk individuals. Identification of hereditary risk of cancer can facilitate the medical care of mutation carriers, including those already diagnosed with cancer who are at risk of a second malignancy. The ability also exists to identify individuals who do not carry a mutation previously characterized in their family and who therefore have no elevated risk of cancer despite their family history. The availability of clinical genetic testing for many hereditary cancer syndromes is already having an impact on the identification and management of individuals with inherited susceptibility to cancer. The use of genes rather than slides to identify individuals with increased cancer risk before the development of disease is a fundamental extension of skills already possessed and used by many physicians. Like Prometheus, we must have the courage to acquire and use knowledge of life. The progress of medicine depends upon it.
Received February 5, 1999.
Revised March 18, 1999.
Accepted March 19, 1999.
References