Specialty Article: Obstetrics and Gynecology

Symposium: Genetic Testing and Management of the Cancer Patient and Cancer Families

Moderator: Maurice J Webb, MD, FACS

• Cancer Family Syndromes
• Genetic Testing for Hereditary Cancer Syndromes
• Genetic Testing and Management of Patients with Colorectal Cancer
• Breast Cancer Risk: Screening and Prevention
• Hereditary Ovarian Cancer

Dr. Webb: It would be difficult as practicing clinicians or even as lay people to be unaware of the rapid expansion of the body of knowledge about the role inherited genetic alterations plays in the development of cancer. Numerous articles appear daily in the lay and professional press. Our clinical journals contain numerous review articles on the subject. Our textbooks are being revised with "updates" as new information becomes available. Cancer societies are producing material to help the professional and the patient answer their many questions. Committees of various colleges are producing opinions for our benefit, and commercial companies bombard us with advertising announcing their wares.

Because of the gradual fall in mortality from cardiovascular disease, it is predicted that before too long, cancer will be the leading cause of death. While environmental factors have long been known to contribute to the development of malignancy, we are beginning to understand that it is probably the result of a combination of factors, and particularly the interaction between genetic predisposition and the environment. We have also recognized that family history plays an important role in many of the most common cancers (eg, colon, breast, ovary, prostate), and the identification of some of the genes that contribute to an inherited predisposition to cancer makes us feel that we may be on the verge of discovering new ways to prevent, diagnose, and treat some cancers.

But even when the genetic basis of the disease is defined, it is obvious that a combination of factors is necessary to produce the disease process. Therefore, the predictive value of genetic testing will have to be assessed cautiously, and we must remember that an interpretation of a result today might differ from the interpretation in a few years' time.

Theoretically, genetic testing may be able to assist us with screening and early detection of cancers, and more effective therapy may result. It may also aid in assessing risk in a particular patient, allowing preventive measures (eg, prophylactic mastectomy, or prophylactic oophorectomy). Genetic testing may help to identify a particular type of malignancy, aiding in diagnosis, and there are hopes that genetic therapies might be effective treatment for the disease. Testing may also allow the doctor to counsel the patient about measures of a preventive nature.

Testing, however, certainly has a negative side. A positive test, though helpful in earlier intervention, may also cause considerable anxiety for the patient who worries about the loss of a job or inability to obtain life and health insurance.

Confidentiality issues abound with risks of release of information to insurers, employers, government departments, etc, with the inevitable consequences. In fact, in mid 1998, there are bills before Congress dealing with protection of genetic information in relation to insurability, insurance premium costs, disclosure of information to employers, etc. A major problem also is the misinterpretation of a test result; we are yet unsure about validity of testing in the longterm.

Genetic testing also brings about a change in the informed consent process because we now have to make sure the patient understands the various psychosocial implications of being tested (eg, discrimination in the workplace, the patient\'s self-image, effects of disclosure, and what this may do to the patient and family relationships). Lastly, genetic investigations are not inexpensive, and widespread use of genetic testing and its subsequent counseling, could have a significant impact on health economics.

It is because of the tremendous current interest in the topic of genetics and cancer, and the potential minefield of clinical, socioeconomic, psychosocial, and ethical dilemmas that this subject contains, that a seminar was presented on "Genetic Testing and Management of the Cancer Patient and Cancer Families" at the American College of Surgeons Annual Clinical Congress Meeting in October 1997. We, as surgeons, are still the front line in cancer diagnosis and treatment. We will increasingly have to answer questions from our patients about the role genetics plays in the development of their own cancer or in the cancer of a relative. The subject is very complex and we need to understand and keep abreast of the implications of this new science.

Cancer Family Syndromes

LC Hartmann, MD

Cancer family syndromes, or inherited cancer syndromes, appear to account for 1% to 5% of the overall cancer burden.1,2 Passing within these families are dominantly inherited mutations in key growth regulatory genes. Each individual within such a family has a 50:50 chance of inheriting the deleterious gene copy, which conveys a significantly increased likelihood of developing cancer. Approximately 30 inherited cancer syndromes have been identified to date, the most common of which are listed in Table 1. We will focus primarily on those syndromes with a preponderance of breast, ovarian, and colorectal cancer.

Table 1. Common Inherited Cancer

Syndrome	Primary Cancer(s)	Gene
Breast-ovarian	Breast; ovary	BRCA1
Breast	Breast; male breast; pancreas; others	BRCA2
Li-Fraumeni	Sarcomas; breast; brain; leukemia; others	p53
Familial adenomatous polyposis (FAP)	Colorectal	APC
Hereditary nonpolyposis colorectal cancer (HNPCC)	Colorectal; endometrial; ovary; urinary tract	MSH2; MLH1; PMS2; PMS1
Hereditary retinoblastoma	Retinoblastoma; osteosarcomas	RB1
von Hippel-Lindau (VHL)	Renal; pheochromocytomas	VHL
Multiple endocrine neoplasia (MEN1)	Pancreatic islet cell	MEN1
Multiple endocrine neoplasia (MEN2)	Medullary thyroid; pheochromocytomas	RET
Cowden disease	Breast; thyroid	PTEN
Familial melanoma	Melanoma; pancreas	p16
Neurofibromatosis type 1 (NF1)	Neurofibromas; neurofibrosarcomas; AML; brain	NF1
Neurofibromatosis type 2 (NF2)	Acoustic neuromas; meningiomas; gliomas	NF2

What are the clinical features that suggest the presence of an inherited cancer syndrome? Awareness of these features is important so that clinicians can identify members of true cancer families from the 5% to 10% of cancer patients who report some family history of cancer.

Characteristics of inherited cancer syndromes are early age onset, multiple primaries in the same individual, and autosomal dominant pattern of inheritance (ie, half at risk affected; maternal or paternal transmission).

A 33-year-old female was seen for risk assessment and management options. She had not had cancer. Her mother had bilateral breast cancer in her thirties and ovarian cancer at 52. Four of her mother's sisters had breast cancer, and multiple first cousins had developed the disease, including three daughters of her mother's brother. The average age of onset of breast cancer in this family was 32. These features are characteristic of an inherited breast-ovarian cancer syndrome. The pathogenesis of cancer in these families, as previously stated, is an inherited abnormality in an underlying cancer susceptibility gene.

By way of contrast, the more common clinical situation is where there is a family history of cancer but only one or two individuals within the family are affected, and at later ages. The pathogenesis of cancer in such familial situations is not well understood. It is not likely the result of an abnormality in a dominant-acting susceptibility gene. This may be a polygenic effect or may reflect multifactorial causes. Of course, with a common disease such as breast or colon cancer, having one or two affected relatives with cancer in a family could simply reflect sporadic cancer cases occurring by chance.

One of the earliest reports in the medical literature of an inherited cancer syndrome was described by Paul Broca, MD, a French surgeon.3 At that time, in the mid 1800s, there was widespread skepticism that a predisposition to cancer could be inherited. Dr. Broca, 100 years ahead of his time, concluded that heredity clearly could play a role in cancer predisposition.

Approximately 30 inherited cancer syndromes have been elucidated to date. To identify the underlying susceptibility genes responsible for these syndromes, scientists use an initial step of linkage analysis. By tracking DNA markers located in known chromosomal regions that cosegregate in family members affected with the cancer(s) of interest,
scientists can localize a putative susceptibility gene to a given chromosomal area. Once the region of interest has been defined, eg, 17q21 containing a breast cancer susceptibility locus, a variety of strategies can be used to identify the actual gene of interest within that region (BRCA1 in this example).2

What is the normal function of genes underlying inherited cancer syndromes? The proteins encoded by these genes normally control a variety of key regulatory pathways from proliferation to differentiation, apoptosis, and DNA repair. The most common type of gene mutated in these syndromes is a tumor suppressor gene. A normal individual possesses two functional copies of such a tumor suppressor gene in each cell. At-risk individuals in cancer families have inherited an inactive, nonfunctional copy of a tumor suppressor gene, leaving each cell with only one functional tumor suppressor gene copy. If that copy is then lost or mutated in a cell of an at-risk tissue (eg, breast), a cancer can develop.

Clearly, our understanding of inherited cancer syndromes has progressed dramatically since Dr. Broca worked to convince the medical community in the mid-1800s that a predisposition to cancer could be inherited within certain families. Although rare, these syndromes have enabled scientists to identify regulatory genes and processes that underlie malignant transformation in inherited and noninherited cancers.

References

1. Fearon ER. Human cancer syndromes: Clues to the origin and nature of cancer. Science 1997;278:1043-1050.

2. Knudson AG. Hereditary Cancer: Theme and variations. J Clin Oncol 1997;15:3280-3287.

3. Broca P. Etiologie des productions accidentelles. Traite Des Tumeurs. Paris, 1866:147-157.

Genetic Testing for Hereditary Cancer Syndromes

Noralane M Lindor, MD

Currently, more than 30 different cancer-predisposing syndromes are known. Some are profoundly rare, and others are sufficiently common that many clinicians will encounter them on a regular basis. If one were to create a "Top Ten" list of most studied or most common of these disorders, the list would likely include hereditary breast/ovarian/other cancer (due to BRCA1 or BRCA2); hereditary nonpolyposis colorectal cancer (Lynch syndrome; due to mutations in one of at least five genes whose products participate in DNA mismatch repair processes); familial adenomatous polyposis (APC gene); hereditary melanoma (p16 and other genes); Li-Fraumeni syndrome (p53 and other unknown genes); multiple endocrine neoplasias type I and II (mutations in the MEN1 gene and RET protooncogene, respectively); hereditary retinoblastoma (RB gene); von Hippel-Lindau syndrome (VHL gene); and neurofibromatosis type I (NF gene).

Genetic testing of some type is now available on a clinical basis for all of these disorders. It is likely that patients will expect their physicians to be familiar with genetic testing options. An informed discussion on genetic testing needs to address three discrete topics: scientific aspects, psychologic issues, and insurance/confidentiality issues.

Scientific issues in genetic testing include: test sensitivity, which varies from assay to assay, but is consistently below 85% for most tests; current limitations in ability to distinguish harmful versus DNA sequence changes; lack of good data on gene penetrance for various mutations; understanding how to interpret a negative test (which depends upon prior testing information on that family); discussing before testing how results might impact clinical decisions (eg, would recommendations for surgery or screening be different based upon test results).

The psychologic impact of genetic testing for any person is hard to overestimate, and can be difficult to predict; testing of an individual often leads to a ripple effect throughout the family. Very careful pretest evaluation and consultation with a mental health professional is advisable for most people seeking predictive genetic testing.

Much has been written about potential insurance discrimination based on genetic tests. Legal protection from such discrimination is in place in some states and is being drafted at the federal level. Clinicians must explain to patients how genetic test results are handled in their practice: are they incorporated in the general medical record or kept sequestered in a "shadow file?"

Having covered these three general topics, a few additional points deserve emphasis. First, the drive to have a genetic test should be patient initiated, not pushed by the doctor, spouse, or children. Second, genetic testing is never an emergency. Blood for DNA can be cryopreserved in any clinical molecular genetics laboratory while decisions are being made. Third, adequate time and education must be given to the patient about genetic testing. The time spent carefully considering all pros and cons will spare physician and patient anguish later. If this cannot be done in the office, arrange a referral to a geneticist or familial cancer program for the patient. Fourth, be sure to plan for alterations in clinical management before testing. If there is no way that this test would alter the plans, then why do it? Last, encourage patients to participate in an organized familial cancer research program or registry. There are many urgent questions to be answered in these high risk families, but without their active participation this process will not more forward.

Genetic Testing and Management of Patients with Colorectal Cancer

James M Church, MD, FACS, FRACS

Principles of genetic testing for colorectal cancer

Genetic testing can be performed in a variety of tissues for a variety of reasons. Genetic blood testing for presymptomatic diagnosis of an inherited disease is appropriate only when looking for a germline mutation in a dominantly inherited syndrome such as familial adenomatous polyposis (FAP) or hereditary nonpolyposis colorectal cancer (HNPCC). Here, the mutation is inherited at conception and affects every cell in the affected person\'s body. Leucocyte DNA can therefore be used. Patients with sporadic colorectal cancer do not have a germline mutation and are not suitable for blood testing. Because FAP (1%) and HNPCC (<5%) account for only a small minority of all colorectal cancers, most patients are not suitable for blood testing for colorectal cancer.

Testing for microsatellite instability (MIS) involves comparing a panel of 5 to 10 microsatellite markers in normal tissue and tumor. A tumor that is MIS positive (\g1 marker positive) is likely to be right-sided, mucinous, associated with a strong family history and young patient age, and have a prognosis that is better than expected. MIS testing may be a good screen to detect HNPCC patients who do not have an obvious family history.

Testing tumor tissue for loss of heterozygosity (LOH) of DCC or p53 may provide useful prognostic information. There is evidence that tumors without LOH have a better prognosis than do tumors with LOH, independent of pathologic stage. Testing for LOH or k ras activation in cells recovered from stool is a potential screening test for sporadic colorectal neoplasia or neoplasia in chronic ulcerative colitis. This is not yet fully refined.

Practicalities of genetic testing

Who to test?

Some patients present with a family history suggestive of HNPCC or FAP, while others may be obviously affected (eg, a colon full of polyps). In such patients genetic testing may be considered to confirm the diagnosis or identify the mutation. This starts with the proband (the first affected patient in a family to present). Once the mutation responsible for the disease in the proband is discovered, screening of unaffected at-risk relatives is simple. Genetic testing should always be arranged in the context of an institutional review board-approved genetic program including pre and posttest counseling, after informed consent has been obtained. Current criteria for genetic testing for inherited colorectal cancer are shown in Tables 2 and 3.

Table 2. Blood Tests for Inherited Colorectal Cancer Genes

Gene	Syndrome	Criteria
APC	Familiar adenomatous polyposis	< 100 adenomas in the large bowel or any adenoma in patients with family history of familial adenomatous polyposis
hMSH2	Hereditary nonpolyposis colorectal cancer	Amsterdam criteria; 3 relatives with colorectal cancer, 1 < 50 years at diagnosis, 2 first degree of the other, 2 consecutive generations
APC	FCC	Ashkenazi Jews with any family history, test for APC 1307 substitution

Table 3. Tests on Tumor Tissue

Syndrome	Criteria
Microsatellite instability	Amsterdam criteria positive families  Individuals with: hereditary nonpolyposis colorectal cancers (HNPCC) Colorectal cancer and a first degree relative with a colorectal cancer or an HNPCC-related cancer or an adenoma (< 40 y) Colorectal or endometrial cancer (< 45 y) Right-sided, undifferentiated colorectal cancer (< 45 y) Signet-ring colorectal cancer (< 45 y) Adenomas (< 40 y)
Loss of heterozygosity	Potentially any cancer; this is not yet in routine clinical practice

How to test?

DNA for mutational testing is usually obtained from white cells in the blood. The area of DNA of interest can be amplified by the polymerase chain reaction (PCR) and the sequence of the DNA determined by direct sequencing. If the actual mutation operating in a family is known, screening is quick and easy. If the mutation is unknown, searching for it by DNA sequencing is laborious and time-consuming and various "short cuts" are available. One such technique is single strand conformational polymorphism (SSCP).

Another is the protein truncation test. For this test patient RNA is used to produce complementary DNA (cDNA) that is amplified by PCR. The cDNA is used to make the protein normally produced by the gene in question. The protein products are then separated by gel electrophoresis. If proteins are shorter than normal, there must be a mutation affecting protein production. This test does not identify the mutation but will diagnose its presence without the need for information about family. Some mutations (missense) may not produce a truncated protein. Such mutations may or may not cause disease. Almost all APC mutations are nonsense mutations and cause protein truncation of the APC protein. Missense mutations are more commonly seen in HNPCC.

Linkage analysis is a test used when the location of the mutated gene is not known. It relies on finding a DNA marker close enough to the gene to be inherited with it. Then, if the marker is positive in a test patient, it is likely that the mutation is also present. This test relies on knowing the status (of both marker and disease) of enough family members to allow a confident prediction.

Implications of Genetic testing

A recent report by Giardello and coworkers showed that incorrect interpretation of genetic testing in FAP is common. A test that shows the presence of the mutation in an at-risk individual in an FAP or HNPCC family means that the patient will get the disease. Penetrance is close to 100% in FAP and more than 80% in HNPCC. Endoscopic surveillance should continue with prophylactic surgery when adenomas appear in FAP. The role of prophylactic surgery in HNPCC is not so well defined. A test showing the absence of the mutation in such a situation means that the patient will not get the disease, and endoscopic surveillance should be recommended as for average-risk people. If the family mutation has never been detected, a negative test does not mean there is not a mutation. At-risk relatives should continue to have intensive endoscopic surveillance.

Bibliography

1. Church JM, Williams BRG, Casey G. Molecular genetics and colorectal neoplasia: a primer for the clinician. Igaku Shoin, Tokyo, 1996.

2. Church JM. Familial adenomatous polyposis: a review. Perspectives in Colorectal Surg 1995;8:203-225.

3. Marra G, Boland CR. Hereditary non-polyposis colorectal cancer: the syndrome, the genes, and historical perspective. J Natl Cancer Inst 1996;87:1114-1125.

4. Burke W, Peterson G, Lynch P, et al. Recommendations for follow-up care of individuals with an inherited predisposition to cancer. 1. Hereditary Nonpolyposis Colon Cancer. JAMA 1997;227:915-919.

Breast Cancer Risk: Screening and Prevention

Joseph P Crowe, MD, FACS

Screening

Breast cancer screening in women >50 years old results in a 25% to 30% relative reduction in mortality. The benefit for women ages 40 to 49 years continues to be debated, but as additional data become available, the advantages for this group may be similar. In early 1997 the American Cancer Society advocated annual mammogram screening starting at age 40 years, and the National Cancer Institute suggested screening every 1 to 2 years from ages 40 to 49 years and yearly thereafter. Guidelines for screening of individuals with an inherited predisposition to breast cancer are in evolution. Breast self-exam should begin by ages 18 to 21 years, and clinical breast examination is recommended beginning at ages 25 to 35 years. There are no data on which to base mammography guidelines for this group. With a higher breast cancer incidence and earlier disease presentation among individuals with genetic predisposition compared to the average, annual mammographic screening starting between 25 to 35 years would seem reasonable. The recommendation is often made to begin screening 10 years earlier than the <zaq;7>youngest family member who has breast cancer. The benefit of early screening needs to be balanced against the potential increased cancer risk of early and repeated radiation exposure and against the possible increased sensitivity to radiation among individuals with genetic susceptibility to cancer. Currently the advantages of early screening in this group appear to outweigh any theoretical risks.1

Prevention

Chemoprevention. Chemoprevention is defined as the use of specific compounds to prevent, inhibit, or reverse carcinogenesis. Central to any consideration of chemoprevention are issues of appropriate population, disease prevalence, duration of the specific carcinogenic process, and overall risk/benefit of the chemopreventive agent. The two agents that are being studied in large clinical trails for prevention of breast cancer are tamoxifen and the retinoid, 4-HPR. The number of participants and the duration of followup required to demonstrate efficacy of chemopreventive agents in these studies makes the studies extraordinarily difficult to conduct. The tamoxifen prevention trial in the United States will require 13,000 women accrued over 5 years and followed for a minimum of 5 to 10 years before any results become available. It is estimated this trial will cost at least $60 million.

The possible advantages of including women at risk for inherited breast cancer in chemoprevention trials would appear obvious, however this population is not large. The relative infrequency of inherited breast cancer (<10% of all breast cancer) together with myriad issues surrounding the implication of a positive genetic test for any individual make it unlikely that genetic susceptibility testing will be used to identify individuals for such studies in the near future. The ability to measure and noninvasively monitor intermediate biomarkers of breast cancer risks for their response to chemopreventive agents would be ideal. Such biomarkers are not yet available.2

Risk Factor Modification. Both exogenous hormone use and dietary fat intake have received considerable attention as risk factors for breast cancer. Even though prolonged estrogen replacement therapy has been found to be associated with a higher risk of breast cancer (relative risk 1,3) in two metaanalyses, the reduction in cardiac disease and osteoporosis suggests that most women benefit from estrogen replacement therapy. Oral contraceptive use has also been found to be associated with a slightly higher but reversible risk of breast cancer. No relationship has been found between oral contraceptive use and family history in terms of breast cancer risk in these studies. Specific data are not available at this time to make clear recommendations concerning the exogenous hormones among individuals who are BRCA1 or BRCA2 carriers.

Dietary fat intake has been studied extensively as a risk factor for breast cancer. Even though there is considerable epidemiologic evidence indicating a positive association between increased fat consumption and breast cancer, longitudinal cohort studies such as the Nurses' Health Study in the United States have not found a relationship. Two possible explanations for this are that dietary factors may have an adverse impact during early phases of maturation and development not being measured in these studies, or that the range of fat consumption among the participants is still too high to demonstrate any association, if one exists.

Prophylactic Mastectomy. Historically, prophylactic mastectomy has been performed for various reasons such as fibrocystic disease, pain, cancer phobia, multiple breast biopsies, family history, and lobular carcinoma in situ. Without an appropriate high risk population, with the obvious cosmetic and psychologic implications of such procedures, and with the widespread use of more conservative approaches for management of invasive breast cancer, prophylactic mastectomy has not been advocated widely. At this point, the ability to identify individuals with an exceptionally high breast cancer risk based on genetic testing has led to a reconsideration of prophylactic mastectomy. In a recent report by Schrag and associates3 in which a hypothetical decision analysis model was developed, a 2.9-, 4.1-, and 5.3-year gain in life expectancy was predicted for a 30-year-old woman with either a BRCA1 or BRCA2 mutation assuming a 40%, 60%, or 85% risk of breast cancer development by 70 years of age. Perhaps the most provocative data comes from a report by Nelson,4 who reviewed 950 women who had prophylactic mastectomy at the Mayo Clinic between 1960 and 1993. With a mean followup of 17 years, 76 breast cancers were predicted among this group based upon estimates from the Gail model. Only seven cancers were observed, a 91% reduction in breast cancer risk. For individuals choosing prophylactic mastectomy, new surgical techniques that include skin sparing mastectomy and immediate autologous tissue reconstruction can give excellent cosmetic results. In spite of these technical advances, future breast cancer prevention undoubtedly lies in a better
understanding of individual risk factors and in modulation of this risk with nonablative measures.

References

1. Burke W, Daly M, Garber J, et al. Recommendations for follow-up care of individuals with an inherited predisposition to cancer. JAMA 1997;227:915-919.

2. Fabian CJ, Kamel S, Zalles C, Kimler BF. Identification of a chemoprevention cohort from a population of women at high risk for breast cancer. Breast J 1997;3:220-226.

3. Schrag D, Kuntz KM, Garber JE, Weeks JC. Decision analysis-effects of prophylactic mastectomy and oophorectomy of life expectancy among women with BRCA1 or BRCA2 mutations. N Engl J Med 1997;20:1465-1471.

4. Nelson NJ. Studies show prophylactic surgeries seem to reduce cancer risk. J Natl Cancer Inst 1997;89:762-763.

Hereditary Ovarian Cancer

Holly H Gallion, MD

In hereditary breast ovarian cancer families, disease is transmitted as an autosomal dominant trait with a high degree of penetrance. In these families, children of an affected individual have a 50:50 chance of inheriting the disease gene and a nearly 50% risk of developing breast or ovarian cancer by age 85 years. Rather than basing estimates of risk on family history alone, it is now possible to determine which gene is responsible for disease in the family, which of the children actually inherited the disease causing mutation, and which did not. Those who test negative can then be counseled that their cancer risk is similar to that of the general population. Conversely, those who have inherited a mutation can be advised that they have a 95% lifetime risk for developing breast cancer and a 40% lifetime risk for ovarian cancer. These mutation carriers can then be offered increased surveillance and prevention strategies.

There are currently a number of uncertainties that make testing for these disease genes problematic. First, can we accurately assign risk in mutation carriers? The original estimates of cancer risk in BRCA1 and BRCA2 carriers were based on data from the very large, multicase families used in the original linkage studies. In these families, lifetime risk estimates for ovarian cancer was approximately 60% in BRCA1 carriers. More recent studies of carriers of specific mutations in Ashkenazi Jewish women unselected for family history reveal that the lifetime ovarian cancer risk for carriers of the 185 del AG and 5382insC mutations in BRCA1 and the 617delT mutation in BRCA2 is closer to 16%, well below previous estimates based on high-risk families.1 The lower cancer incidence observed in Ashkenazi Jewish women may represent allelic heterogeneity in the expression of these genes, with different mutations conferring different cancer risks. In addition, pedigree analysis of large multicase families would support the theory of genetic heterogeneity, with some BRCA1 families containing multiple cases of both breast and ovarian cancer, while in others, the excess cancer cases are strictly breast or ovarian. Several studies have suggested that mutations in the 5` half of BRCA1 predispose to both breast and ovarian cancer, whereas mutations close to the 3` end of the gene are associated with breast cancer only.2 Such differences in gene penetrance would be important in counseling gene carriers, particularly those considering prophylactic oophorectomy.

Another area of uncertainty is whether traditional risk factors alter the expression of these highly penetrant genes. Very little is known regarding the influence of traditional reproductive and environmental risk factors, such as diet, in BRCA1 and BRCA2 carriers. For example, recent data from a study of 331 BRCA1 mutation carriers indicate that BRCA1 carriers decrease their risk of breast cancer somewhat with at least one full-term pregnancy (relative risk 0.85) but that ovarian cancer risk actually increases with increasing parity (relative risk [RR] 1.4).3 This increase in ovarian cancer risk with increasing parity is the opposite of what is seen in studies of the general population and suggests that oral contraceptive use, suggested as a possible prevention tool, may not decrease ovarian cancer risk in mutation carriers. Although current screening recommendations for BRCA1 carriers include annual pelvic examination, transvaginal sonography, and serum CA125, the effectiveness of these methods is unproved. For this reason, prophylactic oophorectomy at the completion of childbearing or at the time of menopause is recommended for women who are members of a documented hereditary ovarian cancer family. But as with prophylactic mastectomy, prophylactic oophorectomy is not completely protective. The occurrence of peritoneal carcinomatosis, clinically and histologically indistinguishable from ovarian cancer, has been reported to occur in as many as 2% to 25% of high risk women. Whether this is as a result of metastasis unrecognized at prophylactic surgery, or the later development of cancer arising from residual ovarian tissue or the peritoneum, is unknown.

Another concern with genetic testing is the implication of a negative test. Although the vast majority of BRCA1 mutations occur in the coding region of the gene, it should not be forgotten that mutations in the noncoding portions of the gene cannot be detected by current commercial tests. Although rare, these regulatory mutations can lead to false negative results. In addition, a small proportion of inherited breast cancer cases are from mutations in other genes, including p53, MSH2, and MLH1. Likewise, not all inherited ovarian cancer cases are from BRCA1 and BRCA2 mutations. Approximately 10% occur as part of the HNPCC syndrome (hMSH2, hMLH1, and hPMS2), and evidence suggests there are additional, yet-to-be-identified breast and ovarian cancer genes. Also, gentic testing has failed to find mutations in families predicted to contain mutations based on prior probabilities, suggesting that we are missing mutations or that previous estimates of the proportion of families with BRCA1 and BRCA2 mutations are overestimated.

Finally, although little is known about the function of the BRCA1 gene itself, wild-type BRCA1 gene has been shown to inhibit the growth of sporadic breast and ovarian cancer cell lines in vitro and in nude mice.4 These studies provide direct evidence that BRCA1 functions as a tumor suppressor gene and that this effect is not limited to hereditary disease, suggesting that it may be useful in the treatment of sporadic disease. In fact, delivery of wild-type BRCA1 using a retroviral vector has been shown to suppress tumor growth in nude mice, and gene therapy trials in ovarian cancer patients are underway.

Clearly, genetic epidemiologic studies aimed at determining factors that influence the penetrance of these genes and the effectiveness of surveillance and prevention strategies in mutation carriers are needed. But elucidation of the underlying molecular basis of disease in mutation carriers is equally important because this knowledge should provide novel prevention and treatment strategies for the inherited and the sporadic forms of ovarian cancer.

References

1. Struewing JP, Hartge P, Wacholder S, et al. The risk of cancer associated with specific mutations of BRCA1 and BRCA2 among Ashkenazi Jews. N Engl J Med 1997;336:1401-1408.

2. Gayther SA, Warren W, Mazoyer S, et al. Germline mutations of the BRCA1 gene in breast/ovarian cancer families provide evidence for a genotype/phenotype correlation. Nat Genet 1995;11:428-433.

3. Narod SA, Goldgar D, Cannon-Albright L, et al. Risk modifiers in carriers of BRCA1 mutations. Int J Cancer 1995;64:394-398.

4. Holt JT, Thompson ME, Szabo C, et al. Growth retardation and tumour inhibition by BRCA1. Nat Genet 1996;12:298-302.

Presented at the 83rd Annual Clinical Congress of the American College of Surgeons. Moderator: Maurice J. Webb, MD, FACS, Division of Gynecologic Surgery, Mayo Clinic. Presenters: Lynn C. Hartmann, MD, Department of Medical Oncology, Mayo Clinic; Noralane M. Lindor, MD, Division of Medical Genetics, Mayo Clinic; James M. Church, MD, FACS, FRACS, Department of Colorectal Surgery, Cleveland Clinic; Joseph P. Crowe, MD, FACS, Department of General Surgery, Cleveland Clinic; Holly H. Gallion, MD, Department of Obstetrics and Gynecology, University of Kentucky Medical Center.

Received April 13, 1998; Accepted April 30, 1998. Correspondence address: Maurice J. Webb, MD, Mayo Clinic, 200 First St. SW, Rochester, MN 55905.