Identification and management of inherited cancer susceptibility.

Identification of inherited cancer-predisposing genes offers opportunities for cancer prevention. Inherited susceptibility genes have been identified, primarily through studies of unusual cancer cases and families but also through general population studies. Examples include the RB1 gene for retinoblastoma; the WT1 gene for Wilms' tumor; germline p53 mutations in families with the Li-Fraumeni syndrome; the NF1 and NF2 genes for neuroblastomatosis, types 1 and 2; the VHL gene for renal cancer and other tumors associated with Von Hippel-Lindau disease; the APC gene for adenomatous polyposis coli; the BRCA1 gene for hereditary breast and ovarian cancer; and the mismatch repair genes for colon and other common cancers. For some cancers, identification of gene carriers might be beneficial for targeting screening and chemopreventive interventions. On the other hand, predisposition testing for cancer has the potential for harm from loss of insurability and employability, psychological distress, social stigmatization and other adverse effects. Research is needed to identify predisposition testing procedures that maximize benefits while minimizing harm to subjects. Chemoprevention trials in genetically susceptible populations offer the prospect of finding effective methods of reducing future cancer risk.


Introduction
Speakers at this meeting have summarized current knowledge of cancer etiology and considered opportunities for prevention. Epidemiological studies of human populations have demonstrated the carcinogenic effects of ionizing and ultraviolet radiation, tobacco, alcoholic beverages, viral and bacterial agents, and chemical compounds (1,2). In the 1994 Annual Report on Carcinogens (3), nearly two dozen substances, occupational exposures, and medical treatments were classified as carcinogenic in humans. Many more substances were said to be "reasonably anticipated to be carcinogens," but uncertainties remain. Presently, known carcinogens account for perhaps half the incident cancers in the United States. For many common forms of cancer, such as cancers of the breast, prostate, and gastrointestinal tract, lifestyle factors have been implicated, although specific causes have been difficult to pinpoint.
A constant challenge to cancer epidemiologists is the relatively small excess risk associated with exposures to many suspected carcinogens (4). To overcome the problem, researchers are utilizing laboratory tools to refine dose-response relationships and to reinforce the biological plausibility of hypotheses (5). Molecular biomarkers also can be applied to classify patients and strengthen associations in study subgroups (6). In this emerging era of molecular epidemiology, substantial progress has been made in the identification of inherited single-gene traits that markedly increase cancer risk (7). In the future, the knowledge of the more complex genetic systems involved in the activation and detoxification of carcinogens should further enhance the power of cancer epidemiology, as described in Rothman's paper (8). This overview will explore opportunities for cancer prevention based on recent discoveries of inherited cancer-predisposing genes.

Susceptibility Genes
Historically, the effects of inherited cancer susceptibility genes were discerned through studies of cancer occurrence in families. More recent advances in cytogenetics provided the first glimpses of inherited genetic changes that confer susceptibility to cancer. In 1986, the first inherited cancer susceptibility gene, the RB1 gene for retinoblastoma, was cloned (9). Since then, a dozen such genes have been identified, mostly through studies of unusual cases and families.
Epidemiological studies have consistently shown that a family history of cancer is a risk factor for virtually all forms of neoplasia in humans (10). The relative risk is typically 1.5to 3-fold for the forms of cancer that have occurred in close relatives. In some instances, the elevated risk extends to multiple organ sites, as in the multiple endocrine neoplasia syndromes (11). Familial aggregation of cancers is usually attributed to inherited susceptibility, although shared exposures to environmental hazard among close relatives often cannot be excluded. For the common forms of cancers, chance aggregation among blood relatives is a third explanation. Historically, inherited susceptibility factors seldom were the focus of epidemiological research, in part because laboratory tools for genetic analyses were rudimentary. Inherited predisposition was also viewed as immutable, whereas environmental carcinogens should be avoidable. The attributable risk of individual genes seems trivial compared with the effects of tobacco and sunlight exposure. However, the relative risk is exceptionally high among carriers of many inherited cancer susceptibility genes. The lifetime risk of specific cancers is 80 to 90% among carriers of certain susceptibility genes. These persons often are prone to early onset disease and multiple primary cancers.
Cancer in certain families appears to be transmitted as an autosomal dominant trait with high penetrance. Clinical observations at the bedside have helped to identify inherited susceptibility genes, and affected families. Epidemiological studies of these kindreds have quantitated their excess risk. Laboratory studies have helped clarify the biological basis of susceptibility, Environmental Health Perspectives F.P. Li including identification of inherited cancer-Major discoveries in the last year predisposing genes. Genes found to date include identification of the BRCA1 gene include tumor suppressor genes, an onco-for hereditary breast and ovarian cancer and gene (RET), and more recently, mismatch the mismatch repair genes (MSH2, MLH1, repair genes (11,12). In these studies, and PMSI and PMS2) for colon and other clinicians and epidemiologists provided common cancers (16)(17)(18). Discovery of the tissue specimens essential to the mole-these genes has vastly increased the numcular discoveries by laboratory scientists. bers of cancer susceptibility gene carriers Identification of inherited susceptibility who can be identified. In contrast to the genes has, in turn created new clinical rarity of carriers of RBI, WTI, p53, APC, opportunities to reduce cancer morbidity and RET, approximately 5% of breast or and mortality. colon cancer patients might carry an inher-Retinoblastoma is a prototypic herediited susceptibility gene. Carriers of susceptary cancer in humans (13). Mutations in tibility genes may account for more than the retinoblastoma gene, even single-10,000 cases of breast cancer and 10,000 nucleotide alterations, can confer a 90% cases of colon cancer diagnosed annually in likelihood of development of cance-r" the United States. Up to 1 to 2 million (9). Based on studies of retinoblastoma, Americans are estimated to carry a breast Knudson developed a 2-mutation model or colon cancer susceptibility gene. Because that, provided the conceptual framework survival from these cancers depends largely for studying rare family aggregates of can-on disease stage, identification of carriers cer to gain new understanding of human for targeted interventions might be benecarcinogenesis (14). The model postulates ficial. However, seeking gene carriers in that at least two mutations are required to cancer-free populations (genetic predispositransform a normal cell into a cancer cell. tion testing or predictive testing) is new At the molecular level, the same mutant (19). Benefits of predisposition testing are genes are involved in both familial and determined largely by the availability of sporadic (nonfamilial) forms of a cancer. prevention and early detection measures. Patients with hereditary retinoblastoma On the other hand, predisposition testing have inherited a germinal mutation and for cancer or other hereditary disease runs subsequently acquired another in the sec-the risk of harm from loss of insurability ond RBI allele, whereas those with spo-and employability, psychological distress, radic cancers had to acquire both RBI social stigmatization, and other adverse mutations within one retinal cell. Despite affects. Research is needed to identify prethe rarity of retinoblastoma as a clinical disposition testing procedures that maxidiagnosis, recent studies have shown that mize benefits while minimizing harm to acquired mutations in RBI can occur in subjects (20). Investigators, clinicians, govmost forms of cancer (15). Thus, studies ernmental agencies, and institutional of a rare hereditary cancer led to the review boards are working to set guidelines identification of a gene that is commonly to protect subjects who volunteer for testmutated during neoplastic transformation. ing (21). The social and ethical issues of Knowledge of the structure and functions offering commercial testing to the general of RBI might yield novel approaches to population are drawing headlines in both therapy of cancer in the future. the lay and professional press.
Most known tumor suppressor genes Genetic research on human cancers is have been discovered through studies of not new. Until inherited cancer susceptibilfamilies with hereditary cancers: the WTi ity genes were cloned, however, genetic gene for Wilms' tumor; germline p53 studies were limited to analyses of somatic mutations in families with diverse childmutations acquired during neoplastic hood cancers, and early onset breast cancer transformation. The results had no rele-(Li-Fraumeni syndrome); the NFl and vance to future cancer risks among relatives NF2 genes for neurofibromatosis, types 1 of the patient. In marked contrast, cancer and 2; the VHL gene for renal cancer and predisposition testing is conducted in canother tumors associated with von Hippelcer-free individuals with the goal of disclos-Lindau disease; and the APC gene for ing results that might sharply elevate adenomatous polyposis coli (12). Mapping estimates of future risk. and cloning of these genes have been The potential for adverse effects of disfacilitated by finding large affected families closure of inherited cancer susceptibility and rare cases with constitutional chromorequires that adequate safeguards be prosome markers, particularly translocations vided to test subjects (22,23). There must and deletions. be freedom from coercion to be tested.
Informed consent is imperative, with provisions for freedom to withdraw or postpone disclosure. Before testing begins, discussions are necessary regarding disclosure of results to persons other than the subject, including physicians, relatives, and third parties. Providers of predisposition testing should assess the mental and emotional competence of test subjects. Counseling and education should include interpretation and limitations of test results. The investigators should be prepared to manage complications of testing and to follow up to assess long-term outcomes.
After new inherited susceptibility genes such as BRCA1 and the colon cancer genes are cloned, population surveys often are conducted to assess population frequency of inherited mutations. A dilemma is whether to disclose individual or aggregate results from these surveys to the participants. Surveys for mutations in large numbers of specimens in research laboratories may not be as accurate as those conducted by clinical laboratories where government agencies monitor quality control. Considering the profound adverse effects of reporting erroneous results, survey findings probably should be considered preliminary and not disclosed to individuals without confirmation. In these survey studies, most subjects do not show inherited mutations, but the clinical significance of a negative finding is problematical. Multiple inherited susceptibility genes, some still unknown, can predispose a person to breast, colon, and other cancers. A negative test may simply mean that the mutation is in another susceptibility gene. Thus, the absence of a germline mutation is meaningless unless a germline mutation was previously identified in an affected relative.
A common misconception is that a positive genetic test predicts that cancer will develop, whereas a negative test is associated with the baseline population frequency for that cancer (23). This view disregards data on risk conveyed by the family and personal medical history. Cancer predisposition testing modifies clinical risk estimates but does not predict with certainty when, where (organ site), and if cancer will develop. To convey these concepts, counseling about cancer risk should begin even before specimens are collected for laboratory testing. Individuals should be counseled on the basis of family history of cancer, personal history of antecedent diseases, and known risk factors such as hormonal and reproductive factors for breast cancer. Risk estimates associated Environmental Health Perspectives with family history can be refined further on the basis of number of affected relatives, ages at diagnosis, relationship to the test subject, and cases of multiple primary neoplasms. In this context, subsequent DNA test results further refine risk estimates.
Identification of an inherited susceptibility gene per se provides no benefit to the subject. Benefit from cancer predisposition testing accrues when primary prevention, targeted screening, and more effective treatments lead to reduced morbidity and mortality (24). Finding carriers of inherited susceptibility genes creates an obligation to develop strategies for the care of these cancer-prone individuals. Increased medical surveillance might detect early cancers in at-risk organs and tissues. For example, identification of an inherited RBJ mutation in an at-risk infant can lead to surveillance and curative treatments with minimal loss of eyesight. Benefits of targeted screening might also accrue to carriers of colon and breast cancer susceptibility genes (25). Targeted colonoscopies can lead to excision of pre-cancerous polyps, but the procedure is neither inexpensive nor free of morbidity. An alternative is primary prevention through reducing exposures to carcinogens among those with inherited susceptibility to cancer. Avoidance of sunlight exposure in patients with xeroderma pigmentosum or hereditary melanoma can reduce the risk of skin neoplasms. Dietary modification might reduce the risk of colon and other cancers in gene carriers, but sustained change in dietary practices is difficult to achieve. There are a few indications at present to perform prophylactic surgery to remove the target organs in genetically susceptible individuals. Prophylactic total colectomy is standard treatment for patients with adenomatous polyposis coli due to the APC gene. However, morbidity of the procedure is high, and deaths from dismoid tumors arising at the surgical site are frequent. A few patients in breast-ovarian cancer families have undergone prophylactic mastectomy, oophorectomy, or both. Unfortunately, abdominal carcinomatosis have developed subsequent to prophylactic oophorectomy in rare cases, and the extent of risk reduction after prophylactic mastectomy is not known A new and exciting area of research is chemoprevention (26). The goal of chemoprevention is to prevent or delay cancer occurrence through the use of chemical agents. Chemopreventive agents under study include natural and synthetic products including vitamin micronutrients, natural compounds, and pharmaceutical agents. A growing number of clinical chemoprevention trials are in progress, and early results from several studies have been encouraging. Target populations in these studies include cancer patients at high risk of second cancers, those with precancerous lesions, and carriers of cancer susceptibility genes. The latter are an attractive group for chemoprevention trials, in part because their risk of future cancer development is exceptionally high. In these groups, chemopreventive agents with some toxicity might be acceptable. Chemoprevention trials require large numbers of subjects. When cancer prevention is the end point, studies should be extended over many years; their cost is high. Intermediate biomarkers of efficacy could give early indications of benefit. Unfortunately, validated biomarkers have been difficult to identify and their predictive power is difficult to measure. Problems notwithstanding, chemoprevention trials in genetically susceptible populations offer the prospect of finding effective methods of reducing future cancer risk.