Cancer Genetics

A CLINICIAN’S GUIDE TO: BREAST CANCER RISK ASSESSMENT

How best to apply the various tools available to help assess a woman’s risk for breast cancer.

Professor of Medicine, University of Vermont, College of Medicine, Burlington, Vermont

BREAST CANCER RISK ASSESSMENT is a routine aspect of women’s healthcare, and 3 models are available to help clinicians with this task: Gail, Claus, and Tyrer-Cuzick. A fourth—and increasingly common—tool is genetic testing, which can be an important piece of genetic risk assessment for some individuals. Recommendations for the use of genetic testing for hereditary breast and ovarian cancers were published by the American College of Obstetricians and Gynecologists (ACOG) in 2009.¹ These recommendations are reviewed here, with a particular focus on the components of genetic counseling, identifying individuals for testing, and interpreting test results. Additionally, we review current risk models and apply these models to hypothetical cases that illustrate their strengths and limitations.

Models for risk assessment

Lifetime risk of breast cancer for the average woman is approximately 12%.² Risk factors for breast cancer are well described.³ The use of exogenous hormones (estrogen and progesterone) are known to increase risk, and endogenous hormonal factors such as ages at menarche, menopause, and first childbirth are associated with a slight, but significant, increase in breast cancer risk. Environmental factors, such as diet and alcohol use, although not well understood, also are associated with breast cancer risk. Having had a breast biopsy—and especially a biopsy that demonstrated atypical hyperplasia or lobular neoplasia—is associated with a significant risk of breast cancer.⁴ Increased breast density is associated with a higher risk of breast cancer independent of the increased difficulty of reading a mammogram.⁵ (See “Mammographic breast density, endocrine therapies, and breast cancer risk”). A positive family history of cancer also suggests a genetic component to a person’s risk. Many of these factors have been incorporated into the models discussed.

The Gail model incorporates age, history of breast cancer in first-degree relatives (mother or sisters), hormonal factors (age at menarche and first child), biopsy history, and biopsy results to calculate 5-year and lifetime risks of developing breast cancer. The model was validated in the Breast Cancer Prevention Trial for women older than 35 years who undergo screening.⁶ The National Institutes of Health has made this model accessible online (www.cancer.gov/bcrisktool). This model does not account for family history in relatives other than the mother and sister. Importantly, the Gail model fails to account for paternal family history of breast cancer.

The Claus model factors in breast cancer history in up to 2 first- and/or second-degree maternal or paternal relatives and the age at which these persons were diagnosed with breast cancer⁷ (TABLE 1). This information is used to calculate risk based on present age and lifetime risk. The Claus model does not incorporate information regarding breast pathology or menstrual and reproductive history and is limited to information regarding only 2 relatives affected by breast cancer (as opposed to other potentially hereditary cancers).

The Tyrer-Cuzick model, presently a research tool, attempts to address the limitations of the previous models by including a variety of risk factors.⁸ The model uses age, body mass index (BMI), hormonal and reproductive factors, breast disease, and an extensive family history to calculate personal breast cancer risk. One study showed that this model produced a higher ratio of expected to observed cases of breast cancer than did the Gail and Claus models, indicating that it may produce a more accurate assessment of risk.⁹ This model has not yet been validated and should, therefore, be used cautiously in a clinical setting.

Although each of these models provides a short- and long-term risk, the use of lifetime risk allows comparison of risk estimates generated by each model. An understanding of each model’s limitations will allow the clinician to use the most appropriate model for assessing risk and counseling a woman in a given situation. The cases that follow compare the use of each model in a common situation.

Lifetime risk of breast cancer based on family history: Claus model

Two case studies: lifetime risk estimates

PATIENT 1. Lisa is a 39-year-old, premenopausal woman, with a BMI of 22 kg/m². Menarche was at age 11 years, and she had the first of 2 children at age 19. A recent breast biopsy demonstrates atypical hyperplasia.

Using the 3 models, Lisa’s lifetime risk of developing breast cancer is estimated to be as follows:

Gail: 18.3%;
Claus: Cannot be used;
Tyrer-Cuzick: 27.6%

Use of the Claus model would be inappropriate, given Lisa’s lack of family history. The Tyrer-Cuzick model, which places Lisa at high lifetime risk, has not been validated and may overestimate her risk. The Gail model, which places Lisa at moderate risk, is the most appropriate model to use in this situation. Based on her moderate risk, she is a candidate for slightly increased screening (twice-yearly clinical breast exams and annual mammography beginning at age 40). She is not a candidate for screening breast MRI but may be a candidate for tamoxifen.

PATIENT 2. Ann is a 35-year-old, premenopausal, nulliparous woman with a BMI of 20 kg/m². She reached menarche at age 11 years and is currently using an oral contraceptive. One paternal aunt was given a diagnosis of breast cancer at age 40, and her other paternal aunt was given a diagnosis of ovarian cancer at age 55.

The models estimate Ann’s lifetime risk of breast cancer as follows:

Gail: 12.4%;
Claus: 13.2%;
Tyrer-Cuzick: 31.7%

The Gail model seriously underestimates Ann’s risk because it does not take into account her paternal family history. Similarly, the Claus model may underestimate her risk, as the family history of ovarian cancer is not accounted for. The Tyrer-Cuzick model may be the most accurate model to use in this situation. In such cases, genetic testing can often assist in obtaining a more accurate risk assessment. For example, if Ann’s paternal aunt is found to have a mutation in BRCA1, then Ann is a candidate for genetic testing. If Ann carries the same mutation in BRCA1 as did her aunt, she has a greatly increased lifetime risk of both breast and ovarian cancer. Current estimates suggest a 65% lifetime risk for breast cancer and 45% lifetime risk for ovarian cancer.¹⁰ A negative test result would indicate that Ann’s risk was average. In Ann’s case, however, family history of breast cancer can be ignored, because the family history is related to the presence of a genetic mutation that she has not inherited.

Hereditary breast and ovarian cancer: a cancer family syndrome

Approximately 5% to 10% of breast cancer cases are hereditary,^11,12 and 15% to 20% of breast cancer cases likely have a familial component and result from gene-environment interactions.¹³ Although BRCA1 and BRCA2 are the genes most commonly screened for, they are probably not the only genes implicated in familial breast cancer: in as many as 70% of cases, testing the person in a family most likely to have hereditary cancer yields a negative result.¹⁴

BRCA1 and BRCA2 mutations are associated with a 45% to 85% lifetime risk of breast cancer, with average age of onset of 43 years for BRCA1 and 45 years for BRCA2.¹⁰ Both BRCA1 and BRCA2 are associated with a >50% risk of a second primary breast cancer. BRCA1 is associated with a 45% lifetime risk of ovarian cancer, whereas BRCA2 has half that risk.¹⁵ Researchers have identified regions within BRCA2 that are associated with higher rates of ovarian cancer; in clinical practice, however, all deleterious mutations within each gene are given similar rates of cancer risk.¹⁶ Families with BRCA1 or BRCA2 mutations have an increased risk of prostate cancer.¹⁷ Families with BRCA2 mutations are at risk for male breast cancers as well as pancreatic, laryngeal, and gallbladder/bile duct cancers, and melanoma.^17-20 The prevalence of BRCA1 and BRCA2 mutations is estimated to be between 1 in 300 to 800 persons of European heritage and 1 in 40 persons of Ashkenazi Jewish heritage.^21,22

Genetic testing for BRCA1 and BRCA2, which has been clinically available for years, involves evaluation of the entire coding region of each gene in search of large deletions. Interpretation of genetic testing can be complicated, especially if the significance of the genetic alteration identified is uncertain. In such cases the mutation is labeled a “variant of uncertain significance” and cannot be used for testing of other family members.

In 2009, ACOG published a recommendation that women with a 20% to 25% chance of having a BRCA1 or BRCA2 mutation undergo genetic risk assessment and be considered for genetic testing (TABLE 2).¹ ACOG further identified women who have a 5% to 10% risk of BRCA1 or BRCA2 mutation as potential candidates for genetic risk assessment.

Candidates for genetic risk assessment: 2009 ACOG recommendations

Risk assessment is more than genetic testing

Cancer risk assessment is a comprehensive review of, and discussion about, a person’s risk for cancer. This includes obtaining information regarding factors that may mitigate cancer risk (eg, use of hormone therapy, use of oral contraceptives, age at menarche and at first childbirth). Other key components of the risk assessment include discussion of the genetic basis of cancer, family history of cancer, likelihood of a genetic syndrome, and options for screening and prevention.

Genetic testing is a tool that can refine risk assessment. Before any genetic test is performed, the possible results—and their implications for the individual and her family—should be discussed: this expensive test has minimal value to a patient who is unwilling to share the results with her relatives. Patients should be aware of supportive resources available to them after they receive test results.

The cancer risk assessment should be conducted by a healthcare provider with expertise in cancer genetics. Clinicians can search for qualified professionals using the website of the National Society of Genetic Counselors (www.nsgc.org), which identifies providers with a variety of specialties, including cancer, who are located within a specified radius of a zip code.

If genetic testing is appropriate, the genetic counselor or other healthcare provider may facilitate genetic testing by investigating insurance coverage, providing test results and follow-up care, and helping the patient inform at-risk relatives.

The value of a full pedigree analysis

Open-ended and far-reaching questions are critical to obtaining a cancer family history. Given that BRCA1 and BRCA2 are associated with a variety of cancers, patients should be asked about all known cancers and age at diagnosis among relatives. Patients should also be asked specifically about cancers present in second- and third-degree relatives (grandparents, aunts, uncles, and first cousins), as this information may be helpful in identifying an autosomal dominant pattern. It is important to capture paternal family history and to ask about adoption.

Some situations may obscure a family history of hereditary breast and ovarian cancer. Relatives who have had a hysterectomy or oophorectomy have significantly reduced their chances of ever having breast or ovarian cancer. Family size and gender distribution will also affect the prevalence of hereditary cancer within a family, eg, a very small family with few women will have very few members who might express the phenotype.²³

Lastly, it is often helpful to review medical records of family members to develop a family history of hereditary cancer, since not all gynecologic cancers are equal in this regard.

Interpreting genetic results

It is often the case that a patient who asks about genetic testing does not have breast cancer herself. A key principle in genetic testing for hereditary cancer is to test the person in the family who is most likely to have that cancer. In some families, that person is a relative who has been given a diagnosis of breast cancer. Testing that person can determine whether a mutation exists in the family, establishing a background against which to interpret the genetic test results of relatives.

If a family is known to have a genetic mutation, then interpreting the test of another relative is straightforward: A positive test result for the same mutation is associated with increased cancer risk. A negative test result means that the patient is at average cancer risk; her family history is related to the presence of a mutation that she has not inherited.

Interpreting results in the first person in a family is rarely simple. A result that reveals a deleterious mutation suggests that the person may be at very high risk of developing breast cancer. More often is the case that results are negative and therefore inconclusive or uninformative. If the presence of a mutation has not been established in a family, negative results in an individual could mean several things:

• There may be no mutation in the family

• A mutation may exist that the person did not inherit

• The test may not have been performed correctly

• A mutation could exist in a gene that was not tested (Genetic causes of breast cancer appear in TABLE 3.)

In the case of an inconclusive or uninformative test result, screening and prevention recommendations should be based on family history information. For example, a woman whose mother and sister died of early breast cancer, who is tested for BRCA1 and BRCA2 and receives a negative test result, is at very high risk for breast cancer herself. (See “Patient management after inconclusive results”.) Because patients may respond differently to notification of inconclusive results, a thorough discussion prior to the test and support even for patients who do not receive a positive test is important.²⁴

Causes of hereditary breast cancer

BOX 1

Summary

The models available to clinicians to assess a person’s risk of breast cancer can produce widely varying estimates, and clinicians must understand the limitations of each model to choose the most appropriate model for their patient. Genetic testing can be a powerful risk assessment tool. New recommendations from ACOG suggest a genetic risk assessment for individuals whose risk of carrying a BRCA mutation is 20% to 25%. Women at lower risk (5% to 10%) may also benefit from this type of assessment. Genetic risk assessment includes a comprehensive review of family cancer history and discussion of options for screening, prevention, genetic testing, and the possible results of a genetic test. Inconclusive or uninformative results from genetic testing are common, and in such cases patients should be managed according to family history.

Sexuality, Reproduction & Menopause ©2010 Lebhar-Friedman, Inc.