Role of genetic testing in hereditary melanoma

One in 40 (2.5%) non-Hispanic, American Caucasians will develop melanoma during their lifetime [101]. The majority of these cases are sporadic, but 5–10% of cases occur in familial clusters [1]. Melanoma is a multifactorial condition with contributions from UV radiation exposure, phenotype and, in rare cases, inherited mutations in cancer predisposition genes. Mutations in CDKN2A are identified in 20–40% of families with multiple cases of melanoma but, for many families, the specific genetic etiology of the increased melanoma risk cannot be identified [2,3]. Despite remaining uncertainties regarding the basis of the cancer in all high-risk families, there is growing awareness among the general public about the contribution from inherited factors to cancer risk and a strong interest in genetic testing [4]. Dermatologists are on the front lines of addressing patient questions and concerns about genetic testing. With this article we will provide a concise review of the genetics of melanoma, criteria for identifying appropriate candidates for genetic testing, the genetic testing process and implications of test result reporting.

CDKN2A & cancer risk

In 1994, CDKN2A was the first melanoma-predisposition gene to be identified [5,6]. Other genes have been found to be altered in rare high-risk families [7]. However, CDKN2A is the most clinically relevant melanoma-predisposition gene identified to date. The CDKN2A gene encodes the p16 protein (also known as cyclin-dependent kinase inhibitor 2A), which functions as a tumor-suppressor gene [6]. Mutations in CDKN2A are associated with increased risks for both melanoma and pancreatic cancer. Estimates of melanoma risk have been quite variable. Studies based on data from families ascertained due to a strong family history of melanoma have found the lifetime risk of developing melanoma by 80 years of age to be 58% in Europe, 76% in the USA and 91% in Australia [8]. Lower risk estimates (28% by 80 years of age) have resulted from studies using population-based ascertainment [9]. Penetrance estimates (i.e., the likelihood of developing melanoma if one carries the CDKN2A mutation) tend to be higher in geographical areas and ethnic groups with high background rates of melanoma. This is probably due to the coinheritance of other predisposing genetic factors in these populations and geographical differences in UV intensity. An example of another genetic factor that can interact with CDKN2A is the melanocortin-1 receptor (MC1R). One study found that when CDKN2A mutations are coinherited with MC1R red-hair variants, the risk of melanoma increases from 50 to 84% [10]. Families ascertained on the basis of a strong family history probably have additional factors, such as MC1R, that contribute to melanoma risk. For clinical purposes, penetrance estimates determined from families with multiple cases of melanoma are likely to be more-appropriate estimates for counseling regarding risk.

Data on the risk for pancreatic cancer have also exhibited variability. Data from GenoMEL, an international melanoma genetics consortium, found that 28% of 178 families had at least one pancreatic cancer [4]. However, as with melanoma risk, geographical differences exist. An increased risk for pancreatic cancer was noted in European and North American families, but not in Australian families. Studies of families demonstrating an association with pancreatic cancer have reported a lifetime risk of up to 17% [11].

Identification of appropriate candidates for genetic testing

CDKN2A mutations are identified in only 1% of unselected melanoma cases, and genetic testing for all melanoma patients is not an appropriate or useful strategy [9]. Genetic testing is most useful and cost-effective when targeted toward families with a significant likelihood of harboring a mutation. Features in a family history that are suggestive of the presence of a CDKN2A mutation include multiple affected family members, individuals with multiple primary cancers and the presence of melanoma and pancreatic cancer in a family. As with penetrance estimates, geography and ethnicity play an important role in determining the likelihood of detecting a CDKN2A mutation and preclude the development of a worldwide standard for offering genetic testing for CDKN2A. The probability of detecting a CDKN2A mutation is greater in low melanoma incidence areas and conversely, the probability is lower in high melanoma incidence areas. This is probably due to other genetic or environmental factors contributing to the baseline melanoma risk; that is, when the baseline risk of melanoma is greater, a larger number of melanomas will arise in a familial context but will not be related to inheritance of a CDKN2A mutation. For example, the likelihood of detecting a mutation in a family with two relatives with melanoma is 5% in North America (high melanoma incidence region) and 11% in Spain (low melanoma incidence region) [9,12]. North American families require three or more affected family members to reach a comparable likelihood of mutation detection [9].

For non-Hispanic, Caucasian families in North America, we suggest the following the ‘rule of threes’ guidelines for genetic testing (Box 1):

• • Families with three or more diagnoses of melanoma and/or pancreatic cancer among close family members (first or second degree) on the same side of the family. At least one of the cancers should be a melanoma;

• • Individuals with three or more primary melanomas;

• • Individuals with both melanoma and pancreatic cancer (synchronous or metachronous).

The patient’s ethnicity, history of sun exposure, age of diagnosis and other risks factors should be taken into consideration when evaluating the appropriateness of genetic testing. Individuals from ethnic backgrounds or geographical regions associated with low melanoma incidence may have a higher likelihood of harboring a deleterious mutation and a lower threshold for testing should be considered.

Process of genetic risk assessment & testing

The first step in evaluating patients for hereditary melanoma is to review the family history. Ideally, a three-generation pedigree, including all diagnoses of cancer, age at diagnosis, phenotypic features, screening practices and removal of suspicious lesion, as well as environmental exposures (such as UV exposure and smoking) that may contribute to specific cancers, should be obtained. However, even though family history is an important predictor for determining the likelihood of detecting a CDKN2A gene mutation, its reliability depends on the accuracy of the reported information. Studies have found that up to 40% of reported melanoma diagnoses may be inaccurate [13]. Verification of reported diagnoses with medical records is important for accurate risk assessment. Incorporating a comprehensive review of family history into routine clinical practice can be difficult, and referral to a cancer genetics center (Table 1) may be necessary to ensure that patients’ family histories are evaluated appropriately before proceeding to testing.

Genetic testing should first be performed on an affected family member because they are most likely to harbor a deleterious mutation if one is present in the family. Several clinical laboratories offer genetic testing (Table 1). There are three possible outcomes from analysis of the CDKN2A gene. The first possible outcome is that a deleterious mutation is identified. This result can be used to plan a management strategy, as described later. This result also means that testing for this specific genetic mutation can be offered to other at-risk family members. Relatives who test positive will also be at an increased risk for melanoma and pancreatic cancer. The risk will be greatly reduced for any family member who tests negative for the mutation; however, it will not completely eliminate their risk. The lifetime risk of melanoma development in noncarrier family members within a family that carries a known CDKN2A mutation is slightly elevated to approximately 1.7-times that of the general population [14]. This is probably due to residual risk from shared phenotypic or environmental factors in the family that contributed to melanoma risk along with the CDKN2A mutation. Careful screening is still recommended for these relatives.

The second possible outcome from genetic testing is the identification of a variant of uncertain significance. This result means that a variation from the normal genetic sequence was identified, but it cannot be determined at this time whether this variation affects the function of the protein and contributes to cancer risk. In these cases, the family should still be considered at high risk, but predictive testing for at-risk family members cannot be offered until the significance of the variant is determined.

The final possible outcome from genetic testing is that no mutation is identified. This result does not rule out an inherited component to the risk of melanoma in a family and these families should still be considered at high risk. As with the finding of a variant of uncertain significance, predictive genetic testing cannot be offered to relatives. Families with strong histories of melanoma but no identifiable mutation in CDKN2A are excellent candidates for research to identify new predisposition genes and should be offered participation in appropriate studies (Table 1).

Prior to genetic testing, informed consent should be obtained [15]. The process of obtaining informed consent includes discussion of the purpose of the testing, possible results, implications for the patient and their family members, and the potential impact on insurability. There are often concerns that genetic testing will result in loss of insurance. Currently, there are several state and federal laws that offer protection against the use of genetic information by health-insurance companies. In May 2008, the Genetic Information Non-Discrimination Act (GINA) was passed. This federal act ensures consistent protection against the use of genetic information by health-insurance companies across the USA. When this law goes into affect in May 2009, it will prohibit the use of genetic information from genetic testing or inferred from family history, in determining eligibility for coverage or in setting premiums [16]. However, this law does not require that insurance companies provide coverage for recommended screening examinations for individuals who are at high risk, or prohibit the use of personal medical information, such as a previous or current cancer diagnosis, from being used as a factor in determining insurability.

Implications of genetic testing

Early melanoma detection leads to improved survival [17,18]. Both clinical and self-skin examinations (SSEs) have been shown to result in detection of thinner, earlier-stage melanomas [19,20]. At this time, there are limited data from evidence-based studies regarding the optimal approach for specifically following individuals with CDKN2A mutations, and current guidelines are based on expert opinion. Individuals known to carry a CDKN2A gene mutation should be followed with annual total body skin examination (TBSE) beginning at 10 years of age, and monthly SSEs [14,21,22]. Knowledge of mutation status may contribute to a lower threshold for carrying out biopsies of lesions. Photography may be incorporated into screening, and education about reducing sun exposure may also be helpful.

The optimal approaches for the screening for pancreatic cancer have also not been published. Ideally, pancreatic cancer screening should be pursued in the context of a research study. Current approaches under investigation include evaluation of tumor markers, imaging with CT scans or magnetic resonance cholangiopancratography and endoscopic ultrasound [23–25]. Approximately 40% of individuals diagnosed with pancreatic cancer develop diabetes within 2 years prior to their diagnosis [26,27]. Therefore, clinicians should be aware that the onset of diabetes in high-risk patients may be a harbinger of pancreatic cancer development. Since smoking is a major risk factor for pancreatic cancer, smoking avoidance or cessation should be encouraged strongly.

Concerns about patients not appropriately understanding the implications of genetic test results and being falsely reassured by negative results has been cited as a reason for not offering clinical CDKN2A testing [21,28]. Genetic testing for other hereditary cancer predisposition syndromes, such as hereditary breast and colon cancer, has been part of routine care for several years. Studies of these families have found little psychological sequela from genetic testing and have shown increased uptake of screening and risk-reducing strategies among mutation carriers [29–31]. There are specifically limited data regarding the medical and psychological impact of CDKN2A gene testing. However, preliminary data have not suggested adverse outcomes.

Studies of high-risk families in Australia have indicated a strong interest in genetic testing for familial melanoma [4]. Individuals from high-risk families also report anticipation that being found to carry a specific genetic mutation would motivate them to increase screening and precautionary behaviors [32]. A study by Aspinwall et al. looked at members of high-risk families who had been involved in gene-identification research [33]. Participants in this study had previously been counseled about their risk of melanoma based on family history and recommended to have annual TBSE and to practice monthly SSEs. However, adherence was low. Only 51% reported having had a TBSE within the last year and 51% were under screening with SSE. A total of 83% indicated that they intended to have TBSE within the next year, 1 month following disclosure of genetic test results, and 85% were practicing SSE at least monthly. Gains in intention to screen were seen both among individuals who tested positive and those who tested negative [33].

Future developments

Melanoma genetics presents unique challenges with multifactorial contributions to risk, regional and ethnic differences in penetrance and limited intervention methods. However, studies indicate that patients are interested in receiving this information in order to make informed healthcare and lifestyle decisions. Preliminary data from genetic test-reporting studies are encouraging that this information is helpful. However, continued research is needed to determine the long-term implications of genetic testing for familial melanoma. It is likely that genetic testing will be only one piece of a comprehensive program that includes ongoing education, counseling and dermatological care to ensure lifelong adherence to screening and precautionary measures to maximize early detection and risk reduction.

Table 1. Resources for finding genetic services and research opportunities.

Resource	Website	Information provided
Huntsman Cancer Institute Melanoma Research	www.hci.utah.edu/group/melanomaProgram/about.jsp	Genetic counseling and testing for familial melanoma and research opportunities
National Society of Genetic Counselors	www.nsgc.org	Genetic counseling and resources for identifying local cancer genetics resources
National Cancer Institute Cancer Genetics Service Directory	www.cancer.gov/search/geneticsservices	Genetic counseling and resources for identifying local cancer genetics resources
GenoMEL: international melanoma genetics research consortium	www.genomel.org	Familial melanoma and research opportunities
GeneTests	www.genetests.org	Laboratories offering clinical genetic testing services

Box 1. ‘Rule of Threes’ criteria for CDKN2A gene testing.

Genetic testing should be considered for individuals/families meeting the following ‘rule of threes’ criteria:

• • Families with three or more diagnoses of melanoma and/or pancreatic cancer among close family members on the same side of the family^*‡

• • An individual with three or more (synchronous or metachronous) primary melanomas

• • Individuals with (synchronous or metachronous) melanoma and pancreatic cancer

Financial & competing interests disclosure

S Leachman is on the speakers’ bureau for Myriad Genetic Laboratories. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.

Notes

^*Close family members include first-or second-degree relatives (parents, siblings, children, grandparents, aunts and uncles).

^‡At least one cancer should be a melanoma.