First Time User? Sign Up Now
First Time User? Enroll now.
Home > Health Library > Genetics of Skin Cancer (PDQ®): Genetics - Health Professional Information [NCI]
This information is produced and provided by the National Cancer Institute (NCI). The information in this topic may have changed since it was written. For the most current information, contact the National Cancer Institute via the Internet web site at http://cancer.gov or call 1-800-4-CANCER.
This executive summary reviews the topics covered in this PDQ summary on the genetics of skin cancer, with hyperlinks to detailed sections below that describe the evidence on each topic.
More than 100 types of tumors are clinically apparent on the skin; many are known to have familial and/or inherited components, either in isolation or as part of a syndrome with other features. Basal cell carcinoma (BCC) and squamous cell carcinoma (SCC) are two of the most common malignancies in the United States and are often caused by sun exposure, although several hereditary syndromes and genes are also associated with an increased risk of developing these cancers. Melanoma (which is sometimes referred to as cutaneous melanoma) is a less common type of skin cancer, but 5% to 10% of all melanomas arise in multiple-case families and may be inherited in an autosomal dominant fashion. It is the most lethal of the common skin cancers.
Several genes and hereditary syndromes are associated with the development of skin cancer. Basal cell nevus syndrome (BCNS, caused by pathogenic variants in PTCH1 and PTCH2) is associated with an increased risk of BCC. Syndromes such as oculocutaneous albinism, epidermolysis bullosa, and Fanconi anemia are associated with an increased risk of SCC. The autosomal recessive disease xeroderma pigmentosum (XP) is associated with an increased risk of BCC, SCC, and melanoma. The major germline tumor suppressor gene associated with melanoma is CDKN2A; pathogenic variants in CDKN2A have been estimated to account for 35% to 40% of all familial melanomas. Germline pathogenic variants in several other genes, including CDK4, MITF, BAP1, and BRCA2, have also been found to be associated with melanoma.
Genome-wide association studies show promise in identifying common, low-penetrance susceptibility alleles for many complex diseases, including melanoma, but the clinical utility of these findings remains uncertain.
Risk-reducing strategies for individuals with an increased hereditary predisposition to skin cancer are similar to recommendations for the general population, and include sun avoidance, use of sunscreen, use of sun-protective clothing, and avoidance of tanning beds. Chemopreventive agents such as isotretinoin and acitretin have been studied for the treatment of BCCs in patients with BCNS and XP and are associated with a significant decrease in the number of tumors per year. Vismodegib has also shown promise in reducing the per-patient annual rate of new BCCs requiring surgery among patients with BCNS. Isotretinoin has also been shown to reduce SCC incidence among patients with XP.
Treatment of hereditary skin cancers is similar to the treatment of sporadic skin cancers. One study in an XP population found therapeutic use of fluorouracil (5-FU) to be efficacious, particularly in the treatment of extensive lesions. In addition to its role as a therapeutic and potential chemopreventive agent, vismodegib is also being studied for potential palliative effects for keratocystic odontogenic tumors in patients with BCNS.
Most of the psychosocial literature about hereditary skin cancers has focused on patients with familial melanoma. In individuals at risk of familial melanoma, psychosocial factors influence decisions about genetic testing for inherited cancer risk and risk-management strategies. Interest in genetic testing for pathogenic variants in CDKN2A is generally high. Perceived benefits among individuals with a strong family history of melanoma include information about the risk of melanoma for themselves and their children and increased motivation for sun-protective behavior. A number of studies have examined risk-reducing and early-detection behaviors in individuals with a family history of melanoma. Overall, these studies indicate inconsistent adoption and maintenance of these behaviors. Intervention studies have targeted knowledge about melanoma, sun protection, and screening behaviors in family members of melanoma patients, with mixed results. Research is ongoing to better understand and address psychosocial and behavioral issues in high-risk families.
Many of the medical and scientific terms used in this summary are found in the NCI Dictionary of Genetics Terms. When a linked term is clicked, the definition will appear in a separate window.
A concerted effort is being made within the genetics community to shift terminology used to describe genetic variation. The shift is to use the term "variant" rather than the term "mutation" to describe a difference that exists between the person or group being studied and the reference sequence, particularly for differences that exist in the germline. Variants can then be further classified as benign (harmless), likely benign, of uncertain significance, likely pathogenic, or pathogenic (disease causing). Throughout this summary, we will use the term pathogenic variant to describe a disease-causing mutation. Refer to the Cancer Genetics Overview summary for more information about variant classification.
Many of the genes and conditions described in this summary are found in the Online Mendelian Inheritance in Man (OMIM) catalog. Refer to OMIM for more information.
Structure of the Skin
The genetics of skin cancer is an extremely broad topic. There are more than 100 types of tumors that are clinically apparent on the skin; many of these are known to have familial components, either in isolation or as part of a syndrome with other features. This is, in part, because the skin itself is a complex organ made up of multiple cell types. Furthermore, many of these cell types can undergo malignant transformation at various points in their differentiation, leading to tumors with distinct histology and dramatically different biological behaviors, such as squamous cell carcinoma (SCC) and basal cell cancer (BCC). These have been called nonmelanoma skin cancers or keratinocyte cancers.
Figure 1 is a simple diagram of normal skin structure. It also indicates the major cell types that are normally found in each compartment. Broadly speaking, there are two large compartments—the avascular epidermis and the vascular dermis—with many cell types distributed in a connective tissue matrix, largely created by fibroblasts.
Figure 1. Schematic representation of normal skin. The relatively avascular epidermis houses basal cell keratinocytes and squamous epithelial keratinocytes, the source cells for BCC and SCC, respectively. Melanocytes are also present in normal skin and serve as the source cell for melanoma. The separation between epidermis and dermis occurs at the basement membrane zone, located just inferior to the basal cell keratinocytes.
The outer layer or epidermis is made primarily of keratinocytes but has several other minor cell populations. The bottom layer is formed of basal keratinocytes abutting the basement membrane, along with interspersed melanocytes. The basement membrane is formed from products of keratinocytes and dermal fibroblasts, such as collagen and laminin, and is an important anatomical and functional structure. Basal keratinocytes lose contact with the basement membrane as they divide. As basal keratinocytes migrate toward the skin surface, they progressively differentiate, lose their nuclei and form the spinous cell layer; the granular cell layer; and the keratinized outer layer, or stratum corneum, which serves as a protective covering of the body.
The true cytologic origin of BCC remains in question. BCC and basal cell keratinocytes share many histologic similarities, as is reflected in the name. Alternatively, the outer root sheath cells of the hair follicle have also been proposed as the cell of origin for BCC. This is suggested by the fact that BCCs occur predominantly on hair-bearing skin. BCCs rarely metastasize but can invade tissue locally or regionally, sometimes following along nerves.
Some debate remains about the origin of SCC; however, these cancers are likely derived from epidermal stem cells associated with the hair follicle. A variety of tissues, such as lung and uterine cervix, can give rise to SCC, and this cancer has somewhat differing behavior depending on its source. Even in cancer derived from the skin, SCC from different anatomic locations can have moderately differing aggressiveness; for example, SCC from glabrous (smooth, hairless) sun-exposed skin has a lower metastatic rate than SCC arising from the vermillion border of the lip or from scars.
Additionally, in the epidermal compartment, melanocytes distribute singly along the basement membrane and can undergo malignant transformation into melanoma. Melanocytes are derived from neural crest cells and migrate to the epidermal compartment near the eighth week of gestational age. Melanocytes contain melanin, which is packaged into melanosomes and transported to nearby keratinocytes to induce pigmentation of the skin. Melanin provides a barrier for the nuclei of keratinocytes against ultraviolet radiation and also plays a role in the immune system.
Langerhans cells, or dendritic cells, are another cell type in the epidermis and have a primary function of antigen presentation. These cells reside in the skin for an extended time and respond to different stimuli, such as ultraviolet radiation or topical steroids, which cause them to migrate out of the skin.
The dermis is largely composed of an extracellular matrix. Prominent cell types and organelles in this compartment are fibroblasts, endothelial cells, smooth muscle cells, transient immune system cells, blood vessels, and nerves. When malignant transformation occurs, fibroblasts form fibrosarcomas and endothelial cells form angiosarcomas, Kaposi sarcoma, or other vascular tumors. There are a number of immune cell types that move in and out of the skin to blood vessels and lymphatics; these include mast cells, lymphocytes, mononuclear cells, histiocytes, and granulocytes. These cells can increase in number in inflammatory diseases and can form tumors within the skin. For example, urticaria pigmentosa is a condition that arises from mast cells and is occasionally associated with mast cell leukemia; cutaneous T-cell lymphoma is often confined to the skin throughout its course. Overall, 10% of leukemias and lymphomas have prominent expression in the skin.
Epidermal appendages are also found in the dermal compartment. These are derivatives of the epidermal keratinocytes, such as hair follicles, sweat glands, and the sebaceous glands associated with the hair follicles. These structures are generally formed in the first and second trimesters of fetal development. These can form a large variety of benign or malignant tumors with diverse biological behaviors. Several of these tumors are associated with familial syndromes. Overall, there are dozens of different histological subtypes of these tumors associated with individual components of the adnexal structures.
Finally, the subcutis is a layer that extends below the dermis with varying depth, depending on the anatomic location. This deeper boundary can include muscle, fascia, bone, or cartilage. The subcutis can be affected by inflammatory conditions such as panniculitis and malignancies such as liposarcoma.
These compartments give rise to their own malignancies but are also the region of immediate adjacent spread of localized skin cancers from other compartments. The boundaries of each skin compartment are used to define the staging of skin cancers. For example, an in situ melanoma is confined to the epidermis. Once the cancer crosses the basement membrane into the dermis, it is invasive. Internal malignancies also commonly metastasize to the skin. The dermis and subcutis are the most common locations, but the epidermis can also be involved in conditions such as Pagetoid breast cancer.
Function of the Skin
The skin has a wide variety of functions. First, the skin is an important barrier preventing extensive water and temperature loss and providing protection against minor abrasions. These functions can be aberrantly regulated in cancer. For example, in the erythroderma (extensive reddening of the skin) associated with severe sunburn, alterations in the regulations of body temperature can result in profound heat loss.
Second, the skin has important adaptive and innate immunity functions. In adaptive immunity, antigen-presenting cells engender T-cell responses consisting of increased levels of helper T cells (TH)1, TH 2, or TH 17. In innate immunity, the immune system produces numerous peptides with antibacterial and antifungal capacity. Even small breaks in the skin can potentially lead to infection. The skin-associated lymphoid tissue is one of the largest arms of the immune system and has a role in the prevention of infection. It may also be important in immune surveillance against cancer. Immunosuppression, such as when it is induced intentionally after solid-organ transplantation to reduce the risk of transplanted organ rejection, is a significant risk factor for skin cancer. The skin is significant for communication through facial expression and hand movements. Unfortunately, areas of specialized function, such as the area around the eyes and ears, are common places for cancer to occur. Even small cancers in these areas can lead to reconstructive challenges and have significant cosmetic and social ramifications.
Clinical Presentation of Skin Cancers
While the appearance of any one skin cancer can vary, there are general physical presentations that can be used in screening. BCCs most commonly have a pearly rim or can appear somewhat eczematous (refer to Figure 2 and Figure 3). They often ulcerate (refer to Figure 2). SCCs frequently have a thick keratin top layer (refer to Figure 4). Both BCCs and SCCs are associated with a history of sun-damaged skin. Melanomas are characterized by dark pigment with asymmetry, border irregularity, color variation, a diameter of more than 6 mm, and evolution (ABCDE criteria). (Refer to What Does Melanoma Look Like? on NCI's website for more information about the ABCDE criteria.) Photographs representing typical clinical presentations of these cancers are shown below.
Basal cell carcinomas
Figure 2. Ulcerated basal cell carcinoma (left panel) and ulcerated basal cell carcinoma with characteristic pearly rim (right panel).Figure 3. Superficial basal cell carcinoma (left panel) and nodular basal cell carcinoma (right panel).
Squamous cell carcinomas
Figure 4. Squamous cell carcinoma on the face with thick keratin top layer (left panel) and squamous cell carcinoma on the leg (right panel).
Figure 5. Melanomas with characteristic asymmetry, border irregularity, color variation, and large diameter.
Basal cell carcinoma (BCC) is the most common malignancy in people of European descent, with an associated lifetime risk of 30%. While exposure to ultraviolet (UV) radiation is the risk factor most closely linked to the development of BCC, other environmental factors (such as ionizing radiation, chronic arsenic ingestion, and immunosuppression) and genetic factors (such as family history, skin type, and genetic syndromes) also potentially contribute to carcinogenesis. In contrast to melanoma, metastatic spread of BCC is very rare and typically arises from large tumors that have evaded medical treatment for extended periods of time. BCCs can invade tissue locally or regionally, sometimes following along nerves. With early detection, the prognosis for BCC is excellent.
Risk Factors for Basal Cell Carcinoma
This section focuses on risk factors in individuals at increased hereditary risk of developing BCC. (Refer to the PDQ summary on Skin Cancer Prevention for information about risk factors for BCC in the general population.)
Sun exposure is the major known environmental factor associated with the development of skin cancer of all types. There are different patterns of sun exposure associated with each major type of skin cancer (BCC, squamous cell carcinoma [SCC], and melanoma). (Refer to the PDQ summary on Skin Cancer Prevention for more information about sun exposure as a risk factor for skin cancer in the general population.)
The high-risk phenotype consists of individuals with the following physical characteristics:
Specifically, people with more highly pigmented skin demonstrate lower incidence of BCC than do people with lighter pigmented skin. Individuals with Fitzpatrick type I or II skin (lighter skin) were shown to have a twofold increased risk of BCC in a small case-control study. (Refer to the Pigmentary characteristics section in the Melanoma section of this summary for a more detailed discussion of skin phenotypes based upon pigmentation.) Blond or red hair color was associated with increased risk of BCC in two large cohorts: the Nurses' Health Study and the Health Professionals' Follow-Up Study. In women from the Nurses' Health Study, there was an increased risk of BCC in women with red hair relative to those with light brown hair (adjusted relative risk [RR], 1.30; 95% confidence interval [CI], 1.20–1.40). In men from the Health Professionals Follow-Up Study, the risk of BCC associated with red hair was not as large (RR, 1.17; 95% CI, 1.02–1.34) and was not significant after adjustment for melanoma family history and sunburn history. Risk associated with blond hair was also increased for both men and women (RR, pooled analysis, 1.09; 95% CI, 1.02–1.18), and dark brown hair was protective against BCC (RR, pooled analysis, 0.89; 95% CI, 0.87–0.92).
Individuals with BCCs and/or SCCs report a higher frequency of these cancers in their family members than do controls. The importance of this finding is unclear. Apart from defined genetic disorders with an increased risk of BCC, a positive family history of any skin cancer is a strong predictor of the development of BCC. Data from the Nurses' Health Study and the Health Professionals Follow-Up Study indicate that the family history of melanoma in a first-degree relative (FDR) is associated with an increased risk of BCC in both men and women (RR, 1.31; 95% CI, 1.25–1.37; P < .0001). A family history of melanoma in the same cohorts, plus the Nurses' Health Study 2, showed a similar increased risk (hazard ratio [HR], 1.27; 95% CI, 1.12–1.44). A study of 376 early-onset BCC cases and 383 controls found that a family history of any type of skin cancer increased the risk of early-onset BCC (odds ratio [OR], 2.49; 95% CI, 1.80–3.45). This risk increased when an FDR was diagnosed with skin cancer before age 50 years (OR, 4.79; 95% CI, 2.90–7.90). Individuals who had a family history of both melanoma and nonmelanoma skin cancer (NMSC) had the highest risk (OR, 3.65; 95% CI, 1.79–7.47).
A study on the heritability of cancer among 80,309 monozygotic and 123,382 dizygotic twins showed that NMSCs have a heritability of 43% (95% CI, 26%–59%), suggesting that almost half of the risk of NMSC is caused by inherited factors. Additionally, the cumulative risk of NMSC was 1.9-fold higher for monozygotic than for dizygotic twins (95% CI, 1.8–2.0).
Previous personal history of BCC or SCC
A personal history of BCC or SCC is strongly associated with subsequent BCC or SCC. There is an approximate 20% increased risk of a subsequent lesion within the first year after a skin cancer has been diagnosed. The mean age of occurrence for these cancers is the mid-60s.[7,8,9,10,11,12] In addition, several studies have found that individuals with a history of skin cancer have an increased risk of a subsequent diagnosis of a noncutaneous cancer;[13,14,15,16] however, other studies have contradicted this finding.[17,18,19,20] In the absence of other risk factors or evidence of a defined cancer susceptibility syndrome, as discussed below, skin cancer patients are encouraged to follow screening recommendations for the general population for sites other than the skin.
Major Genes for Basal Cell Carcinoma
Inherited pathogenic variants in the gene coding for the transmembrane receptor protein PTCH1, or PTCH, are associated with basal cell nevus syndrome (BCNS), and somatic mutations are associated with sporadic cutaneous BCCs. (Refer to the BCNS section of this summary for more information.) PTCH1, the human homolog of the Drosophila segment polarity gene patched (ptc), is an integral component of the hedgehog signaling pathway, which serves many developmental (appendage development, embryonic segmentation, neural tube differentiation) and regulatory (maintenance of stem cells) roles.
In the resting state, the transmembrane receptor protein PTCH1 acts catalytically to suppress the seven-transmembrane protein Smoothened (Smo), preventing further downstream signal transduction. Binding of the hedgehog ligand to PTCH1 releases inhibition of Smo, with resultant activation of transcription factors (GLI1, GLI2), cell proliferation genes (cyclin D, cyclin E, myc), and regulators of angiogenesis.[22,23] Thus, the balance of PTCH1 (inhibition) and Smo (activation) manages the essential regulatory downstream hedgehog signal transduction pathway. Loss-of-function pathogenic variants of PTCH1 or gain-of-function variants of Smo tip this balance toward activation, a key event in potential neoplastic transformation.
Demonstration of allelic loss on chromosome 9q22 in both sporadic and familial BCCs suggested the potential presence of an associated tumor suppressor gene.[24,25] Further investigation identified a pathogenic variant in PTCH1 that localized to the area of allelic loss. Up to 30% of sporadic BCCs demonstrate PTCH1 pathogenic variants. In addition to BCC, medulloblastoma and rhabdomyosarcoma, along with other tumors, have been associated with PTCH1 pathogenic variants. All three malignancies are associated with BCNS, and most people with clinical features of BCNS demonstrate germline PTCH1 pathogenic variants, predominantly truncation in type.
Truncating pathogenic variants in PTCH2, a homolog of PTCH1 mapping to chromosome 1p32.1-32.3, have been demonstrated in both BCC and medulloblastoma.[29,30]PTCH2 displays 57% homology to PTCH1. While the exact role of PTCH2 remains unclear, there is evidence to support its involvement in the hedgehog signaling pathway.[29,32]
Putative Genes for Basal Cell Carcinoma
BRCA1-associated protein 1 (BAP1)
Pathogenic variants in the BAP1 gene are associated with an increased risk of a variety of cancers, including cutaneous melanoma and uveal melanoma. (Refer to the BAP1 section in the Melanoma section of this summary for more information.) Although the BCC penetrance in individuals with pathogenic variants in BAP1 is not known, there are several BAP1 families that report diagnoses of BCC.[33,34] In one study, pathogenic variant carriers from four families reported diagnoses of BCC. Tumor evaluation of BAP1 showed loss of BAP1 protein expression by immunohistochemistry in BCCs of two germline BAP1 pathogenic variant carriers but not in 53 sporadic BCCs. A second report noted that four individuals from families with BAP1 germline pathogenic variants were diagnosed with a total of 19 BCCs. Complete loss of BAP1 nuclear expression was observed in 17 of 19 BCCs from these individuals but none of 22 control BCC specimens. Loss of BAP1 nuclear expression was also reported in a series of 7 BCCs from individuals with loss of function BAP1 variants, but only in 1 of 31 sporadic BCCs.
Syndromes Associated With a Predisposition to Basal Cell Carcinoma
Basal cell nevus syndrome
BCNS, also known as Gorlin Syndrome, Gorlin-Goltz syndrome, and nevoid BCC syndrome, is an autosomal dominant disorder with an estimated prevalence of 1 in 57,000 individuals. The syndrome is notable for complete penetrance and high levels of variable expressivity, as evidenced by evaluation of individuals with identical genotypes but widely varying phenotypes.[28,38] The clinical features of BCNS differ more among families than within families. BCNS is primarily associated with germline pathogenic variants in PTCH1, but families with this phenotype have also been associated with alterations in PTCH2 and SUFU.[40,41,42]
As detailed above, PTCH1 provides both developmental and regulatory guidance; spontaneous or inherited germline pathogenic variants of PTCH1 in BCNS may result in a wide spectrum of potentially diagnostic physical findings. The BCNS pathogenic variant has been localized to chromosome 9q22.3-q31, with a maximum logarithm of the odd (LOD) score of 3.597 and 6.457 at markers D9S12 and D9S53. The resulting haploinsufficiency of PTCH1 in BCNS has been associated with structural anomalies such as odontogenic keratocysts, with evaluation of the cyst lining revealing loss of heterozygosity for PTCH1. The development of BCC and other BCNS-associated malignancies is thought to arise from the classic two-hit suppressor gene model: baseline heterozygosity secondary to germline PTCH1 pathogenic variant as the first hit, with the second hit due to mutagen exposure such as UV or ionizing radiation.[44,45,46,47,48] However, haploinsufficiency or dominant negative isoforms have also been implicated for the inactivation of PTCH1.
The diagnosis of BCNS is typically based upon characteristic clinical and radiologic examination findings. Several sets of clinical diagnostic criteria for BCNS are in use (refer to Table 1 for a comparison of these criteria).[50,51,52,53] Although each set of criteria has advantages and disadvantages, none of the sets have a clearly superior balance of sensitivity and specificity for identifying carriers of pathogenic variants. The BCNS Colloquium Group proposed criteria in 2011 that required 1 major criterion with molecular diagnosis, two major criteria without molecular diagnosis, or one major and two minor criteria without molecular diagnosis.PTCH1 pathogenic variants are found in 60% to 85% of patients who meet clinical criteria.[54,55] Most notably, BCNS is associated with the formation of both benign and malignant neoplasms. The strongest benign neoplasm association is with ovarian fibromas, diagnosed in 14% to 24% of females affected by BCNS.[47,51,56] BCNS-associated ovarian fibromas are more likely to be bilateral and calcified than sporadic ovarian fibromas. Ameloblastomas, aggressive tumors of the odontogenic epithelium in the jaw, have also been proposed as a diagnostic criterion for BCNS, but most groups do not include it at this time.
Other associated benign neoplasms include gastric hamartomatous polyps,congenital pulmonary cysts, cardiac fibromas, meningiomas,[62,63,64] craniopharyngiomas, fetal rhabdomyomas, leiomyomas, mesenchymomas, basaloid follicular hamartomas, and nasal dermoid tumors. Development of meningiomas and ependymomas occurring postradiation therapy has been documented in the general pediatric population; radiation therapy for syndrome-associated intracranial processes may be partially responsible for a subset of these benign tumors in individuals with BCNS.[70,71,72] In addition, radiation therapy of malignant medulloblastomas in the BCNS population may result in many cutaneous BCCs in the radiation ports. Similarly, treatment of BCC of the skin with radiation therapy may result in induction of large numbers of additional BCCs.[46,47,67]
The diagnostic criteria for BCNS are described in Table 1 below.
Of greatest concern with BCNS are associated malignant neoplasms, the most common of which is BCC. BCC in individuals with BCNS may appear during childhood as small acrochordon -like lesions, while larger lesions demonstrate more classic cutaneous features. Nonpigmented BCCs are more common than pigmented lesions. The age at first BCC diagnosis associated with BCNS ranges from 3 to 53 years, with a mean age of 21.4 years; the vast majority of individuals are diagnosed with their first BCC before age 20 years.[51,56] Most BCCs are located on sun-exposed sites, but individuals with greater than 100 BCCs have a more uniform distribution of BCCs over the body. Case series have suggested that up to 1 in 200 individuals with BCC demonstrate findings supportive of a diagnosis of BCNS. BCNS has rarely been reported in individuals with darker skin pigmentation; however, significantly fewer BCCs are found in individuals of African or Mediterranean ancestry.[51,75,76] Despite the rarity of BCC in this population, reported cases document full expression of the noncutaneous manifestations of BCNS. However, in individuals of African ancestry who have received radiation therapy, significant basal cell tumor burden has been reported within the radiation port distribution.[51,67] Thus, cutaneous pigmentation may protect against the mutagenic effects of UV but not against ionizing radiation.
Variants in other genes associated with an increased risk of BCC in the general population appear to modify the age of BCC onset in individuals with BCNS. A study of 125 individuals with BCNS found that a variant in MC1R (Arg151Cys) was associated with an early median age of onset of 27 years (95% CI, 20–34), compared with individuals who did not carry the risk allele and had a median age of BCC of 34 years (95% CI, 30–40) (HR, 1.64; 95% CI, 1.04–2.58, P = .034). A variant in the TERT-CLPTM1L gene showed a similar effect, with individuals with the risk allele having a median age of BCC of 31 years (95% CI, 28–37) relative to a median onset of 41 years (95% CI, 32–48) in individuals who did not carry a risk allele (HR, 1.44; 95% CI, 1.08–1.93, P = .014).
Many other malignancies have been associated with BCNS. Medulloblastoma carries the strongest association with BCNS and is diagnosed in 1% to 5% of BCNS cases. While BCNS-associated medulloblastoma is typically diagnosed between ages 2 and 3 years, sporadic medulloblastoma is usually diagnosed later in childhood, between the ages of 6 and 10 years.[47,51,56,78] A desmoplastic phenotype occurring around age 2 years is very strongly associated with BCNS and carries a more favorable prognosis than sporadic classic medulloblastoma.[79,80] Up to three times more males than females with BCNS are diagnosed with medulloblastoma. As with other malignancies, treatment of medulloblastoma with ionizing radiation has resulted in numerous BCCs within the radiation field.[47,62] Other reported malignancies include ovarian carcinoma, ovarian fibrosarcoma,[83,84] astrocytoma, melanoma, Hodgkin disease,[87,88] rhabdomyosarcoma, and undifferentiated sinonasal carcinoma.
Odontogenic keratocysts–or keratocystic odontogenic tumors (KCOTs), as renamed by the World Health Organization working group–are one of the major features of BCNS. Demonstration of clonal loss of heterozygosity (LOH) of common tumor suppressor genes, including PTCH1, supports the transition of terminology to reflect a neoplastic process. Less than one-half of KCOTs from individuals with BCNS show LOH of PTCH1.[49,92] The tumors are lined with a thin squamous epithelium and a thin corrugated layer of parakeratin. Increased mitotic activity in the tumor epithelium and potential budding of the basal layer with formation of daughter cysts within the tumor wall may be responsible for the high rates of recurrence post simple enucleation.[91,93] In a recent case series of 183 consecutively excised KCOTs, 6% of individuals demonstrated an association with BCNS. A study that analyzed the rate of PTCH1 pathogenic variants in BCNS-associated KCOTs found that 11 of 17 individuals carried a germline PTCH1 pathogenic variant and an additional 3 individuals had somatic mutations in this gene. Individuals with germline PTCH1 pathogenic variants had an early age of KCOT presentation. KCOTs occur in 65% to 100% of individuals with BCNS,[51,95] with higher rates of occurrence in young females.
Palmoplantar pits are another major finding in BCC and occur in 70% to 80% of individuals with BCNS. When these pits occur together with early-onset BCC and/or KCOTs, they are considered diagnostic for BCNS.
Several characteristic radiologic findings have been associated with BCNS, including lamellar calcification of falx cerebri in the brain;[98,99] fused, splayed or bifid ribs; and flame-shaped lucencies or pseudocystic bone lesions of the phalanges, carpal, tarsal, long bones, pelvis, and calvaria by diagnostic x-ray imaging. Imaging for rib abnormalities may be useful in establishing the diagnosis in younger children, who may have not yet fully manifested a diagnostic array on physical examination.
Table 2 summarizes the frequency and median age of onset of nonmalignant findings associated with BCNS.
Individuals with PTCH2 pathogenic variants may have a milder phenotype of BCNS than those with PTCH1 variants. Characteristic features such as palmar/plantar pits, macrocephaly, falx calcification, hypertelorism, and coarse face may be absent in these individuals.
A 9p22.3 microdeletion syndrome that includes the PTCH1locus has been described in ten children. All patients had facial features typical of BCNS, including a broad forehead, but they had other features variably including craniosynostosis, hydrocephalus, macrosomia, and developmental delay. At the time of the report, none had basal cell skin cancer. On the basis of their hemizygosity of the PTCH1 gene, these patients are presumably at an increased risk of basal cell skin cancer.
Germline pathogenic variants in SUFU, a major negative regulator of the hedgehog pathway, have been identified in a small number of individuals with a clinical phenotype resembling that of BCNS.[41,42,103] These pathogenic variants were first identified in individuals with childhood medulloblastoma, and the incidence of medulloblastoma appears to be much higher in individuals with BCNS associated with SUFU pathogenic variants than in those with PTCH1 variants.SUFU pathogenic variants may also be associated with an increased predisposition to meningioma.[64,103,105] Conversely, odontogenic jaw keratocysts appear less frequently in this population. Some clinical laboratories offer genetic testing for SUFU pathogenic variants for individuals with BCNS who do not have an identifiable PTCH1 variant.
DNA repair genes
In addition to pathogenic variants in genes primarily associated with BCC, other cancer-associated genes may confer an increased risk for BCC. A study of 61 individuals with a high number of BCCs (mean, 11 BCCs; range, 6–65) underwent genetic testing for 29 high-penetrance cancer susceptibility genes. Thirteen pathogenic variants were found in 12 of 61 individuals (19.7%). This was higher than expected compared with individuals in the Exome Aggregation Consortium (ExAC) database (3%). All of the genes with pathogenic variants were involved in DNA repair, suggesting that defects in DNA repair pathways may increase the risk of BCC. Of these 61 individuals, 21 (34.4%) had a previous diagnosis of another cancer including melanoma, breast, colon, and prostate cancers. To further evaluate the link between a BCC diagnosis and secondary cancers, non-BCC cancer diagnoses were assessed in more than 111,000 individuals from an insurance database with 1 BCC, 13,264 individuals with 6 or more BCCs, and 2,920 individuals with 12 or more BCCs. There was an increased risk of any additional cancer compared with a control group (OR, 1.61 [95% CI, 1.58–1.65], OR, 3.12 [95% CI, 2.98–3.26], and OR, 4.15 [95% CI, 3.79–4.53], respectively). These data should be interpreted with caution as the sequence of cancer diagnoses is unclear, and individuals with one or more BCCs may have been followed more closely by their health care providers or may be more likely to receive cancer screening.
Rombo syndrome is a very rare genodermatosis or genetic disorder associated with BCC. It is thought to have an autosomal dominant inheritance pattern, and it has been outlined in three case series in the literature.[107,108,109] The cutaneous examination is within normal limits until age 7 to 10 years, with the development of distinctive cyanotic erythema of the lips, hands, and feet and early atrophoderma vermiculatum of the cheeks, with variable involvement of the elbows and dorsal hands and feet. Development of BCC occurs in the fourth decade. A distinctive grainy texture to the skin, secondary to interspersed small, yellowish, follicular-based papules and follicular atrophy, has been described.[107,109] Missing, irregularly distributed, and/or misdirected eyelashes and eyebrows are another associated finding.[107,108] The genetic basis of Rombo syndrome is not known.
Bazex-Dupré-Christol syndrome, another rare genodermatosis associated with development of BCC, has more thorough documentation in the literature than Rombo syndrome. Inheritance is accomplished in an X-linked dominant fashion, with no reported male-to-male transmission.[110,111,112] Regional assignment of the locus of interest to chromosome Xq24-q27 is associated with a maximum LOD score of 5.26 with the DXS1192 locus. Further work has narrowed the potential location to an 11.4-Mb interval on chromosome Xq25-27; however, the causative gene remains unknown.
Characteristic physical findings include hypotrichosis, hypohidrosis, milia, follicular atrophoderma of the cheeks, and multiple BCC, which manifest in the late second decade to early third decade. Documented hair changes with Bazex-Dupré-Christol syndrome include reduced density of scalp and body hair, decreased melanization, a twisted/flattened appearance of the hair shaft on electron microscopy, and increased hair shaft diameter on polarizing light microscopy. The milia, which may be quite distinctive in childhood, have been reported to regress or diminish substantially at puberty. Other reported findings in association with this syndrome include trichoepitheliomas; hidradenitis suppurativa; hypoplastic alae; and a prominent columella, the fleshy terminal portion of the nasal septum.[117,118]
Epidermolysis bullosa simplex, Dowling-Meara
A rare subtype of epidermolysis bullosa simplex (EBS), Dowling-Meara (EBS-DM), is primarily inherited in an autosomal dominant fashion and is associated with pathogenic variants in either keratin-5 (KRT5) or keratin-14 (KRT14). EBS-DM is one of the most severe types of EBS and occasionally results in mortality in early childhood. It has an estimated prevalence of 0.02 per million individuals in the United States and an incidence of 1.16 per million live births. One report cites an incidence of BCC of 44% by age 55 years in this population. Individuals who inherit two EBS pathogenic variants may present with a more severe phenotype. Other less phenotypically severe subtypes of EBS can also be caused by pathogenic variants in either KRT5 or KRT14. Approximately 75% of individuals with a clinical diagnosis of EBS (regardless of subtype) have KRT5 or KRT14 pathogenic variants.
Characteristics of hereditary syndromes associated with a predisposition to BCC are described in Table 3 below.
(Refer to the Brooke-Spiegler Syndrome, Multiple Familial Trichoepithelioma, and Familial Cylindromatosis section in the Rare Skin Cancer Syndromes section of this summary for more information about Brooke-Spiegler syndrome.)
As detailed further below, the U.S. Preventive Services Task Force does not recommend regular screening for the early detection of any cutaneous malignancies, including BCC. However, once BCC is detected, the National Comprehensive Cancer Network recommends complete skin examinations every 6 to 12 months for the first 5 years, and then at least annually for life.
The BCNS Colloquium Group has proposed guidelines for the surveillance of individuals with BCNS (refer to Table 4).
Level of evidence: 5
Avoidance of excessive cumulative and sporadic sun exposure is important in reducing the risk of BCC, along with other cutaneous malignancies. Scheduling activities outside of the peak hours of UV radiation, utilizing sun-protective clothing and hats, using sunscreen liberally, and strictly avoiding tanning beds are all reasonable steps towards minimizing future risk of skin cancer. For patients with particular genetic susceptibility (such as BCNS), avoidance or minimization of ionizing radiation is essential to reducing future tumor burden.
Level of evidence: 2aii
The role of various systemic retinoids, including isotretinoin and acitretin, has been explored in the chemoprevention and treatment of multiple BCCs, particularly in BCNS patients. In one study of isotretinoin use in 12 patients with multiple BCCs, including 5 patients with BCNS, tumor regression was noted, with decreasing efficacy as the tumor diameter increased. However, the results were insufficient to recommend use of systemic retinoids for treatment of BCC. Three additional patients, including one with BCNS, were followed long-term for evaluation of chemoprevention with isotretinoin, demonstrating significant decrease in the number of tumors per year during treatment. Although the rate of tumor development tends to increase sharply upon discontinuation of systemic retinoid therapy, in some patients the rate remains lower than their pretreatment rate, allowing better management and control of their cutaneous malignancies.[137,138,139] In summary, the use of systemic retinoids for chemoprevention of BCC is reasonable in high-risk patients, including patients with xeroderma pigmentosum, as discussed in the Squamous Cell Carcinoma section of this summary.
A patient's cumulative and evolving tumor load should be evaluated carefully in light of the potential long-term use of a medication class with cumulative and idiosyncratic side effects. Given the possible side-effect profile, systemic retinoid use is best managed by a practitioner with particular expertise and comfort with the medication class. However, for all potentially childbearing women, strict avoidance of pregnancy during the systemic retinoid course—and for 1 month after completion of isotretinoin and 3 years after completion of acitretin—is essential to avoid potentially fatal and devastating fetal malformations.
Level of evidence (retinoids): 2aii
In a phase II study of 41 patients with BCNS, vismodegib (an inhibitor of the hedgehog pathway) has been shown to reduce the per-patient annual rate of new BCCs requiring surgery. Existing BCCs also regressed for these patients during daily treatment with 150 mg of oral vismodegib. While patients treated had visible regression of their tumors, biopsy demonstrated residual microscopic malignancies at the site, and tumors progressed after the discontinuation of the therapy. Adverse effects included taste disturbance, muscle cramps, hair loss, and weight loss and led to discontinuation of the medication in 54% of subjects. A subsequent, open-label, phase II study included 37 patients from the same cohort who continued vismodegib for up to a total of 36 months. Patients treated with vismodegib had a lower mean incidence of new, surgically eligible BCCs than did placebo-treated patients (P < .0001). However, only 17% of patients tolerated continuous vismodegib for the full 36 months. Tumors reappeared after treatment was stopped, but patients who resumed treatment again experienced tumor response. The duration of benefit after stopping vismodegib appeared to be proportional to the duration and compliance of taking the drug during treatment. Intermittent dosing schedules of vismodegib (8 weeks on/8 weeks off after an initial schedule of daily dosing for 24 weeks or 12 weeks on/8 weeks off) have also been shown to be effective in the reduction of BCCs in the BCNS population, although there has been no direct comparison between continuous dosing and intermittent dosing schedules. On the basis of the side-effect profile and rate of disease recurrence after discontinuation of the medication, additional study regarding optimal dosing of vismodegib is ongoing.
Level of evidence (vismodegib): 1aii
A phase III, double-blind, placebo-controlled clinical trial evaluated the effects of oral nicotinamide (vitamin B3) in 386 individuals with a history of at least two keratinocyte carcinomas (BCC or SCC) within 5 years before study enrollment. After 12 months of treatment, those taking nicotinamide 500 mg twice daily had a 20% reduction in the incidence of new BCCs (95% CI, 6%–39%; P = .12). The rate of new keratinocyte carcinomas was 23% lower in the nicotinamide group (95% CI, 4%–38%; P = .02) than in the placebo group. No clinically significant differences in adverse events were observed between the two groups, and there was no evidence of benefit after discontinuation of nicotinamide. Of note, this study was not conducted in a population with an identified genetic predisposition to BCC.
Level of evidence (nicotinamide): 1aii
Treatment of individual BCCs in BCNS is generally the same as for sporadic basal cell cancers. Due to the large number of lesions on some patients, this can present a surgical challenge. Field therapy with imiquimod or photodynamic therapy are attractive options, as they can treat multiple tumors simultaneously.[144,145] However, given the radiosensitivity of patients with BCNS, radiation as a therapeutic option for large tumors should be avoided. There are no randomized trials, but the isolated case reports suggest that field therapy has similar results as in sporadic basal cell cancer, with higher success rates for superficial cancers than for nodular cancers.[144,145]
Consensus guidelines for the use of methylaminolevulinate photodynamic therapy in BCNS recommend that this modality may best be used for superficial BCC of all sizes and for nodular BCC less than 2 mm thick. Monthly therapy with photodynamic therapy may be considered for these patients as clinically indicated.
Level of evidence (imiquimod and photodynamic therapy): 4
Topical treatment with LDE225, a Smoothened agonist, has also been investigated for the treatment of BCC in a small number of patients with BCNS with promising results; however, this medication is not approved in this formulation by the U.S. Food and Drug Administration.
Level of evidence (LDE225): 1
In addition to its effects on the prevention of BCCs in patients with BCNS, vismodegib may also have a palliative effect on KCOTs found in this population. An initial report indicated that the use of GDC-0449, the hedgehog pathway inhibitor now known as vismodegib, resulted in resolution of KCOTs in one patient with BCNS. Another small study found that four of six patients who took 150 mg of vismodegib daily had a reduction in the size of KCOTs. None of the six patients in this study had new KCOTs or an increase in the size of existing KCOTs while being treated, and one patient had a sustained response that lasted 9 months after treatment was discontinued.
Level of evidence (vismodegib): 3diii
Squamous cell carcinoma (SCC) is the second most common type of skin cancer and accounts for approximately 20% of cutaneous malignancies. Although most cancer registries do not include information on the incidence of keratinocyte carcinomas (BCC and SCC), annual incidence estimates range from 1 million to 5.4 million cases in the United States.[1,2,3] Multiple studies indicate an increased risk of SCC after a first keratinocyte carcinoma; a meta-analysis and review of 45 studies estimated that after a primary SCC diagnosis, 13.3% of individuals would develop a second SCC (95% confidence interval [CI], 7.4%–22.8%).
Mortality is rare from this cancer; however, the morbidity and costs associated with its treatment are considerable.
Risk Factors for Squamous Cell Carcinoma
Sun exposure and other risk factors
Sun exposure is the major known environmental factor associated with the development of skin cancer of all types; however, different patterns of sun exposure are associated with each major type of skin cancer.[5,6,7] Unlike basal cell carcinoma (BCC), SCC is associated with chronic exposure, rather than intermittent intense exposure to ultraviolet (UV) radiation. Occupational exposure is the characteristic pattern of sun exposure linked with SCC. (Refer to the PDQ summary on Skin Cancer Prevention for more information about sun and other environmental and therapeutic exposures as risk factors for skin cancer in the general population.) Other environmental agents associated with SCC risk include tanning beds, arsenic, therapeutic radiation (such as psoralen and ultraviolet A therapy for psoriasis), and immunosuppression.[8,9,10,11,12,13,14]
Characteristics of the skin
Like melanoma and BCC, SCC occurs more frequently in individuals with lighter skin than in those with darker skin.[5,15] A case-control study of 415 cases and 415 controls showed similar findings; relative to Fitzpatrick type I skin, individuals with increasingly darker skin had decreased risks of skin cancer (odds ratios [ORs], 0.6, 0.3, and 0.1, for Fitzpatrick types II, III, and IV, respectively). (Refer to the Pigmentary characteristics section in the Melanoma section of this summary for a more detailed discussion of skin phenotypes based upon pigmentation.) The same study found that blue eyes and blond/red hair were also associated with increased risks of SCC, with crude ORs of 1.7 (95% CI, 1.2–2.3) for blue eyes, 1.5 (95% CI, 1.1–2.1) for blond hair, and 2.2 (95% CI, 1.5–3.3) for red hair.
However, SCC can also occur in individuals with darker skin. An Asian registry based in Singapore reported an increase in skin cancer in that geographic area, with an incidence rate of 8.9 per 100,000 person-years. Incidence of SCC, however, was shown to be on the decline. SCC is the most common form of skin cancer in black individuals in the United States and in certain parts of Africa; the mortality rate for this disease is relatively high in these populations.[17,18] Epidemiologic characteristics of, and prevention strategies for, SCC in those individuals with darker skin remain areas of investigation.
Freckling of the skin and reaction of the skin to sun exposure have been identified as other risk factors for SCC. Individuals with heavy freckling on the forearm were found to have a 14-fold increase in SCC risk if freckling was present in adulthood, and an almost threefold risk if freckling was present in childhood.[19,20] The degree of SCC risk corresponded to the amount of freckling. In this study, the inability of the skin to tan and its propensity to burn were also significantly associated with risk of SCC (OR of 2.9 for severe burn and 3.5 for no tan).
The presence of scars on the skin can also increase the risk of SCC, although the process of carcinogenesis in this setting may take years or even decades. SCCs arising in chronic wounds are referred to as Marjolin's ulcers. The mean time for development of carcinoma in these wounds is estimated at 26 years. One case report documents the occurrence of cancer in a wound that was incurred 59 years earlier.
Immunosuppression also contributes to the formation of BCCs and SCCs. Among solid-organ transplant recipients, the risk of SCC is 65 to 250 times higher, and the risk of BCC is 10 times higher than that observed in the general population, although the risks vary with transplant type and with the immunosuppressive agent used.[23,24,25,26] BCCs and SCCs in high-risk patients (solid-organ transplant recipients and chronic lymphocytic leukemia patients) occur at a younger age, are more common and more aggressive, and have a higher risk of recurrence and metastatic spread than these cancers do in the general population.[27,28] Additionally, there is a high risk of second SCCs.[29,30] In one study, more than 65% of kidney transplant recipients developed subsequent SCCs after their first diagnosis. Among Medicare patients with an intact immune system, BCCs occur as frequently as SCCs; in transplant patients, SCCs outnumber BCCs by a 2:1 ratio.
Personal history of BCC, SCC, and melanoma skin cancers
A personal history of BCC or SCC is strongly associated with subsequent SCC. A study from Ireland showed that individuals with a history of BCC had a 14% higher incidence of subsequent SCC; for men with a history of BCC, the subsequent SCC risk was 27% higher. In the same report, individuals with melanoma were also 2.5 times more likely to report a subsequent SCC. There is an approximate 20% increased risk of a subsequent lesion within the first year after a skin cancer has been diagnosed. The mean age of occurrence for these these cancers is the middle of the sixth decade of life.[32,33,34,35,36,37]
A Swedish study of 224 melanoma index cases and 944 of their first-degree relatives (FDRs) from 154 CDKN2A wild-type families and 11,680 matched controls showed that personal and family histories of melanoma increased the risk of SCC, with relative risks (RRs) of 9.1 (95% CI, 6.0–13.7) for personal history and 3.4 (95% CI, 2.2–5.2) for family history. Another study of 216,115 individuals from the Nurses' Health Study, the Nurses' Health Study 2, and the Health Professionals Follow-Up Study found that individuals with a family history of melanoma showed a 22% increased risk of SCC (hazard ratio,1.22; 95% CI, 1.06–1.40).
Family history of squamous cell carcinoma or associated premalignant lesions
Although the literature is scant on this subject, a family history of SCC may increase the risk of SCC in FDRs. In an independent survey-based study of 415 SCC cases and 415 controls, SCC risk was increased in individuals with a family history of SCC (adjusted OR, 3.4; 95% CI, 1.0–11.6), even after adjustment for skin type, hair color, and eye color. This risk was elevated to an OR of 5.6 in those with a family history of melanoma (95% CI, 1.6–19.7), 9.8 in those with a family history of BCC (95% CI, 2.6–36.8), and 10.5 in those with a family history of multiple types of skin cancer (95% CI, 2.7–29.6). Review of the Swedish Family Center Database showed that individuals with at least one sibling or parent affected with SCC, in situ SCC (Bowen disease), or actinic keratosis had a twofold to threefold increased risk of invasive and in situ SCC relative to the general population.[40,41] Increased number of tumors in parents was associated with increased risk to the offspring. Of note, diagnosis of the proband at an earlier age was not consistently associated with a trend of increased incidence of SCC in the FDR, as would be expected in most hereditary syndromes because of germline pathogenic variants. Further analysis of the Swedish population-based data estimates genetic risk effects of 8% and familial shared-environmental effects of 18%. Thus, shared environmental and behavioral factors likely account for some of the observed familial clustering of SCC.
A study on the heritability of cancer among 80,309 monozygotic and 123,382 dizygotic twins showed that NMSCs have a heritability of 43% (95% CI, 26%–59%), suggesting that almost half of the risk of NMSC is caused by inherited factors. Additionally, the cumulative risk of NMSC was 1.9-fold higher for monozygotic than for dizygotic twins (95% CI, 1.8–2.0).
Syndromes and Genes Associated With a Predisposition for Squamous Cell Carcinoma
Major genes have been defined elsewhere in this summary as genes that are necessary and sufficient for disease, with important pathogenic variants of the gene as causal. The disorders resulting from single-gene pathogenic variants within families lead to a very high risk of disease and are relatively rare. The influence of the environment on the development of disease in individuals with these single-gene disorders is often very difficult to determine because of the rarity of the genetic variant.
Identification of a strong environmental risk factor—chronic exposure to UV radiation—makes it difficult to apply genetic causation for SCC of the skin. Although the risk of UV exposure is well known, quantifying its attributable risk to cancer development has proven challenging. In addition, ascertainment of cases of SCC of the skin is not always straightforward. Many registries and other epidemiologic studies do not fully assess the incidence of SCC of the skin owing to: (1) the common practice of treating lesions suspicious for SCC without a diagnostic biopsy, and (2) the relatively low potential for metastasis. Moreover, NMSC is routinely excluded from the major cancer registries such as the Surveillance, Epidemiology, and End Results registry.
With these considerations in mind, the discussion below will address genes associated with disorders that have an increased incidence of skin cancer.
Characteristics of the major hereditary syndromes associated with a predisposition to SCC are described in Table 5 below.
Xeroderma pigmentosum (XP) is a hereditary disorder of nucleotide excision repair that results in cutaneous malignancies in the first decade of life.Affected individuals have an increased sensitivity to sunlight, resulting in a markedly increased risk of SCCs, BCCs, and melanomas. One report found that keratinocyte carcinomas (BCC and SCC) were increased 150-fold in individuals with XP; for those younger than 20 years, the prevalence has been estimated to be 5,000 to 10,000 times what would be expected in the general population.[45,46]
The natural history of this disease begins in the first year of life, when sun sensitivity becomes apparent, and xerosis (dry skin) and pigmentary changes may occur in the sun-exposed skin. About one-half of XP patients have a history of severe burning on minimal sun exposure. Other XP patients do not have this reaction but develop freckle-like pigmentation before age 2 years on sun-exposed sites. These manifestations progress to skin atrophy and formation of telangiectasias. Approximately one-half of people with this disorder will develop BCC or SCC, and approximately one-quarter of these individuals will develop melanoma. In the absence of sun avoidance, the median age of diagnosis for any skin cancer is 8 to 9 years.[45,46,47] On average, BCCs and SCCs occur at a younger age than melanoma in the XP population.
Noncutaneous manifestations of XP include ophthalmologic and neurologic abnormalities. Disorders of the cornea and eyelids associated with this disorder are also linked to exposure to UV radiation and include keratitis, corneal opacification, ectropion or entropion, hyperpigmentation of the eyelids, loss of eyelashes, and cancer, including conjunctival and corneal cancers. About 25% of the XP patients examined at the National Institutes of Health (NIH) between 1971 and 2009 had progressive neurological degeneration. Features included microcephaly, progressive sensorineural hearing loss, diminished deep tendon reflexes, seizures, and cognitive impairment. Neurological degeneration, which is most commonly observed in individuals with complementation groups XPA and XPD, was associated with a shorter lifespan (median age of death was 29 years in individuals with neurological degeneration and 37 years in individuals without neurological degeneration). De Sanctis-Cacchione syndrome is found in a subgroup of XP patients, who exhibit severe neurologic manifestations, dwarfism, and delayed sexual development. A variety of noncutaneous neoplasms, most notably SCC of the tip of the tongue, central nervous system cancers, and lung cancer in smokers, have been reported in people who have XP.[45,49] The RR for these cancers is estimated to be about 50-fold higher than in the general population.
The inheritance for XP is autosomal recessive. Seven complementation groups have been associated with this disorder. About 40% of the XP cases seen at the NIH were XPC. ERCC2 (XPD) pathogenic variants were present in about 20%. Complementation group A, due to a pathogenic variant in XPA, accounts for approximately 10% of cases. Other variant genes in this disorder include ERCC3 (XPB), DDB2 (XPE), ERCC4 (XPF), and ERCC5 (XPG). An XPH group had been described but is now considered to be a subgroup of the XPD group.Heterozygotes for pathogenic variants in XP genes are generally asymptomatic. However, one study reported a threefold increase in BCC in Japanese individuals who were heterozygous for XPA pathogenic variants.Founder pathogenic variants in XPA (R228A) and XPC (V548A fs X572) have been identified in North African populations, and a founder pathogenic variant in XPC resulting in a splice alteration (IVS 12-1G>C) has been found in an East African (Mahori) population. It has been proposed that direct screening for these pathogenic variants would be appropriate in these populations.[53,54,55,56] A founder pathogenic variant at the 3' splice acceptor site of intron 3 in the XPA gene is present in approximately 1% of the population in Japan representing nearly 1 million people.
The function of the XP genes is to recognize and repair photoproducts from UV radiation. The main photoproducts are formed at adjacent pyrimidines and consist of cyclobutane dimers and pyrimidine-pyrimidone (6-4) photoproducts. The product of XPC is involved in the initial identification of DNA damage; it binds to the lesion to act as a marker for further repair. The DDB2 (XPE) protein is also part of this process and works with XPC. The XPA gene product maintains single-strand regions during repair and works with the TFIIH transcription factor complex. The TFIIH complex includes the gene products of both ERCC3 (XPB) and ERCC2 (XPD), which function as DNA helicases in the unwinding of the DNA. The ERCC4 (XPF) and ERCC5 (XPG) proteins act as DNA endonucleases to create single-strand nicks in the 5' and 3' sides of the damaged DNA with resulting excision of about 28 to 30 nucleotides, including the photoproduct. DNA polymerases replace the lesion with the correct sequence, and a DNA ligase completes the repair.[44,58]
An XP variant that is associated with pathogenic variants in POLH (XPV) is responsible for approximately 10% of reported cases. This gene encodes for the error-prone bypass polymerase (polymerase eta) which, unlike other genes associated with XP, is not involved in nucleotide excision repair. People with polymerase eta pathogenic variants have the same cutaneous and ocular findings as other XP patients but do not have progressive neurologic degeneration. A founder pathogenic variant resulting in a deletion of exon 10 was seen in 16 of 16 individuals from ten Tunisian consanguineous families.
Work on genotype -phenotype correlations among the XP complementation groups continues; however, evidence suggests that the specific pathogenic variant may have more influence on the phenotype than the complementation group.[44,62] The main distinguishing features appear to be the presence or absence of burning on minimal sun exposure, skin cancer, and progressive neurologic abnormalities. All complementation groups are characterized by the presence of cutaneous neoplasias, but skin cancers may be more common in XPC, XPE, and XPV groups.. There is additional clinical variation within each complementation group. Mild to severe neurologic impairment has been described in individuals with XPA pathogenic variants. Individuals with XPA pathogenic variants in the DNA binding region (amino acids 98–219) may have a more severe presentation that includes neurological findings. Individuals within the XPC complementation group have higher incidences of ocular damage. A very small number of people in the XPB, XPD, and XPG complementation groups have been identified as having xeroderma pigmentosum-Cockayne syndrome (XP-CS) complex. These individuals have characteristics of both disorders, including an increased predisposition to cutaneous neoplasms and developmental delay, visual and hearing impairment, and central and peripheral nervous system dysfunction. It should be noted that people with Cockayne syndrome without XP do not appear to have an increased cancer risk. Similarly, trichothiodystrophy (TTD) is another genetic disorder that can occur in combination with XP. Individuals affected solely with TTD do not appear to have an increased cancer incidence, but some affected with XP/TTD have an increased risk of cutaneous neoplasia. The complementation groups connected with XP/TTD (XPD and XPB) and XP-CS (XPB, XPD, and XPG) are associated with defects in both transcription-coupled nucleotide excision repair and global genomic nucleotide excision repair. In contrast, XP complementation groups C and E have defects only in global genomic nucleotide excision repair.[44,65] In addition, individuals in the XPA, XPD and XPG groups may exhibit severe neurologic abnormalities without symptoms of Cockayne syndrome or TTD. Cerebro-oculo-facio-skeletal syndrome, which has been described with some ERCC2 (XPD) or XP-CS pathogenic variants, does not appear to confer an increased risk of skin cancer.[66,67,68,69]
The diagnosis of XP is made on the basis of clinical findings and family history. Functional assays to assess DNA repair capabilities after exposure to radiation have been developed, but these tests are currently not clinically available in the United States. Clinical genetic testing using sequence analysis to identify pathogenic variants is available for multiple XP-associated genes; the list can be found at the NIH Genetic Testing Registry.
Multiple self-healing squamous epitheliomata (Ferguson-Smith syndrome)
Multiple self-healing squamous epitheliomata (MSSE), or Ferguson-Smith syndrome, first described in 1934, is characterized by invasive skin tumors that are histologically identical to sporadic cutaneous SCC, but they resolve spontaneously without intervention. Linkage analysis of affected families showed association with the long arm of chromosome 9, and haplotype analysis localized the gene to 9q22.3 between D9S197 and D9S1809. Transforming growth factor beta-receptor 1 (TGFBR1) was identified through next-generation sequencing as the gene responsible for MSSE. Loss-of-function pathogenic variants in TGFBR1 have been identified in 18 of 22 affected families. Gain-of-function variants in TGFBR1 are associated with unrelated Marfan-like syndromes, such as Loeys-Dietz syndrome, which have no described increase in skin cancer risk.
Somatic loss of heterozygosity in Ferguson-Smith–related SCC has been demonstrated at this genomic location, suggesting that TGFBR1 can act as a tumor suppressor gene. The long arm of chromosome 9 has also been a site of interest in sporadic SCC. Up to 65% of sporadic SCCs have been found to have loss of heterozygosity at 9q22.3 between D9S162 and D9S165.
Albinism is a major risk factor for skin cancer in individuals of African ancestry.[18,73] One report describing a cohort of 350 albinos in Tanzania found 104 cutaneous cancers; of these, 100 were SCCs, three were BCCs, and one was melanoma. The median age for this population was 10 years. Similar proportions of skin cancer diagnoses were observed in a Nigerian population, with 62% of dermatological malignancies diagnosed as SCC, 16% as melanoma, and 8% as BCC. Of note, some melanomas found in individuals with albinism do contain melanin.
SCC occurring at extremely early ages is a hallmark of oculocutaneous albinism. In a cohort of nearly 1,000 Nigerian patients with albinism, all had malignant or premalignant cutaneous lesions by age 20 years.
Two types of oculocutaneous albinism are known to be associated with increased risk of SCC of the skin. Oculocutaneous albinism type 1, or tyrosinase-related albinism, is caused by pathogenic variants in the tyrosinase gene, TYR, located on the long arm of chromosome 11. This type of albinism accounts for about one-half of cases in individuals of Caucasian ancestry. The OCA2 gene, also known as the P gene, is altered in oculocutaneous albinism type 2, or tyrosinase-positive albinism. Both disorders are autosomal recessive, with frequent compound heterozygosity.
Tyrosinase acts as the critical enzyme in the synthesis of melanin in melanocytes. A variant in this gene in oculocutaneous albinism type 1 produces proteins with minimal to no activity, corresponding to the OCA1B and OCA1A phenotypes, respectively. Individuals with OCA1B have light skin, hair, and eye coloring at birth but develop some pigment during their lifetimes, while the coloring of those with OCA1A does not darken with age.
The gene product of OCA2 is a protein found in the membrane of melanosomes. Its function is unknown, but it may play a role in maintaining the structure or pH of this environment. Murine models with variants in this gene had significantly decreased melanin production compared with normal controls. In one international study of individuals with albinism, biallelic variants in OCA2 were found in 17% of participants.
Genetic variants in SLC45A2 (MATP associated with OCA4), SLC24A5 (associated with OCA6), and TYRP1 (tyrosinase-related protein 1 associated with OCA3) are associated with less common types of oculocutaneous albinism. Reported incidences for these genes in an international population of patients with albinism are 7% for SLC45A2, 1% for TYRP1, and less than 0.5% for SLC24A5.SLC45A2 is found in 24% of oculocutaneous albinism cases in Japan, making it the most common type of albinism among Japanese individuals with identifiable variants. A study of 22 individuals of Italian ancestry without pathogenic variants in TYR, OCA2, or TYRP1 found 5 individuals with biallelic variants in SLC45A2, 4 of whom met clinical criteria for a diagnosis of oculocutaneous albinism. Collectively, more than 600 unique ocular albinism–related genetic variants have been identified. The increased risk of SCC of the skin in people with these variants has not been quantified. It is generally assumed to be similar to other types of albinism. Of note, a meta-analysis demonstrated that the SLC45A2 p.Phe374Leu variant was protective for melanoma, with an OR of 0.41 (95% CI, 0.33–0.50; P = 3.50 x 10-17). However, at this time, it should be noted that clinical testing is not routinely performed for protective variants.
Additional genes associated with oculocutaneous albinism have been found in small numbers of patients. OCA5, located on chromosome 4q24, has been identified in a Pakistani family, whereas OCA6 appears to be caused by pathogenic variants in SLC24A5 on chromosome 15q21.[85,86,87] Pathogenic variants in C10orf11 (LRMDA) cause OCA7, which has been found in patients from the Faroe Islands and Denmark. Small numbers of pathogenic variant carriers have been reported to date. One woman with OCA6 had actinic keratosis, but the incidence of skin cancers in these populations is unknown.
Although oculocutaneous albinism is inherited as an autosomal recessive disorder in most instances, one study has found that heterozygous variants in genes such as TYR, OCA2, TYRP1, and SLC45A2 are overrepresented in families with multiple cases of melanoma. Further investigation is warranted to determine if these genes may be moderate penetrance melanoma susceptibility genes in heterozygotes.
Other albinism syndromes
A subgroup of albinism includes people who exhibit a triad of albinism, prolonged bleeding time, and deposition of a ceroid substance in organs such as the lungs and gastrointestinal tract. This syndrome, known as Hermansky-Pudlak syndrome, is inherited in an autosomal recessive manner but may have a pseudodominant inheritance in Puerto Rican families, owing to the high prevalence in this population. The underlying cause is believed to be a defect in melanosome and lysosome transport. A number of pathogenic variants at disparate loci have been associated with this syndrome, including HPS1, HPS3, HPS4, HPS5, HPS6, HPS7 (DTNBP1), HPS8 (BLOC1S3), and HPS9 (PLDN).[106,107,108,109,110,111,112,113] Pigmentation characteristics can vary significantly in this disorder, particularly among those with HPS1 pathogenic variants, and patients report darkening of the skin and hair as they age. In a small cohort of individuals with HPS1 variants, 3 out of 40 developed cutaneous SCCs, and an additional 3 had BCCs. Hermansky-Pudlak syndrome type 2, which includes increased susceptibility to infection resulting from congenital neutropenia, has been attributed to defects in AP3B1.
Two additional syndromes are associated with decreased pigmentation of the skin and eyes. The autosomal recessive Chediak-Higashi syndrome is characterized by eosinophilic, peroxidase-positive inclusion bodies in early leukocyte precursors, hemophagocytosis, increased susceptibility to infection, and increased incidence of an accelerated phase lymphohistiocytosis. Pathogenic variants in the LYST gene underlie this syndrome, which is often fatal in the first decade of life.[116,117,118]
Griscelli syndrome, also inherited in an autosomal recessive manner, was originally described as decreased cutaneous pigmentation with hypomelanosis and neurologic deficits, but its clinical presentation is quite variable. This combination of symptoms is now designated Griscelli syndrome type 1 or Elejalde disease. It has been attributed to pathogenic variants in the MYO5A gene, which affects melanosome transport. Individuals with Griscelli syndrome type 2 have decreased cutaneous pigmentation and immunodeficiency but lack neurological deficits. They also may have hemophagocytosis or lymphohistiocytosis that is often fatal, like that seen in Chediak-Higashi syndrome. Griscelli syndrome type 2 is caused by pathogenic variants in RAB27A, which is part of the same melanosome transport pathway as MYO5A. Griscelli syndrome type 3 presents with hypomelanosis and does not include neurologic or immunologic disorders. Pathogenic variants in the melanophilin (MLPH) gene and MYO5A have been associated with this variant of Griscelli syndrome.
There are numerous forms of epidermolysis bullosa (EB), which is characterized by cleavage and blistering of the skin. Dystrophic EB and junctional EB are associated with an increased risk of skin cancer, particularly SCC. The type of EB can be difficult to determine clinically, although genetic testing may aid in the classification. In one study of 91 probands with features of EB, a next-generation sequencing panel of 21 genes associated with different forms of EB or skin fragility syndromes was able to predict the subtype in 76 of 87 probands of undetermined subtype (83.5%). Similar multigene panels are clinically available for EB. The types, pathogenic variants involved, and phenotypic characteristics are detailed in the following review.
Dystrophic epidermolysis bullosa
Approximately 95% of individuals with the heritable disorder dystrophic epidermolysis bullosa (DEB) have a detectable germline pathogenic variant in the gene COL7A1. This gene, which is located at 3p21.3, is expressed in the basal keratinocytes of the epidermis and encodes for type VII collagen. This collagen forms a part of the fibrils that anchor the basement membrane to the dermis, thereby providing structural stability and resistance to mild skin trauma. The lack of type VII collagen results in generalized blistering, often starting from birth, and is associated with skin atrophy and scarring. A registry of DEB pathogenic variants, The International DEB Patient Registry, is accessible on the Internet.
There are two recessively inherited subtypes of DEB: severe-generalized (RDEB-sev gen; previously named Hallopeau-Siemens type) and generalized-other or generalized-intermediate (RDEB-O; previously named non–Hallopeau-Siemens type); and a dominantly inherited form, dominant dystrophic epidermolysis bullosa (DDEB). These syndromes are rare. The prevalence per million individuals in the United States and incidence per million live births are 0.36 and 0.57 for RDEB-sev gen, 0.14 and 0.30 for RDEB-O, and 1.49 and 2.12 for DDEB, respectively. The clinical manifestations demonstrate a continuum of severity that complicates definitive diagnosis, especially early in life. The severe generalized subtype, associated with formation of pseudosyndactyly (a mitten-like deformity secondary to fusion of interdigital webbing) in early childhood, carries an SCC risk of up to 85% by age 45 years.[128,129] These cancers arise in nonhealing wounds and usually metastasize to cause death within 5 years of the diagnosis of SCC. In one case series, SCC was the leading cause of death for the 15 patients with the severe generalized subtype. The incidence of SCC appears to be highest in the RDEB subtype. In a review of 69 articles that included all types of EB, 117 individuals with SCC were identified; 81 of these cases (69.2%) were in individuals with RDEB. In this group, the median age of diagnosis was 36 years (range, 6–71 y). A cohort study found that 18 of 283 (6.4%) individuals with RDEB had a diagnosis of SCC with a median age of diagnosis of 22.6 years (interquartile range, 20–27.7 y). Early mortality also has been observed in this disorder, with a mortality rate of up to 40% by age 30 years. Extracutaneous manifestations of RDEB-sev gen include short stature, anemia, strictures of the gastrointestinal and genitourinary tracts, and corneal scarring that may result in blindness.
Diagnosis of EB may be accomplished by immunofluorescence or electron microscopy. A list of recommended diagnostic antibodies and their suppliers is available on the Dystrophic EB Research Association website. Pathogenic variant testing is generally used for prenatal diagnosis rather than for the primary diagnosis of EB.[134,135]
The rate of de novo pathogenic variants for DDEB is approximately 30%; maternal germline mosaicism has also been reported.[136,137] Glycine substitutions in exons 73 to 75 are the most common pathogenic variants in DDEB. G2034R and G2043R account for half of these variants. Less frequently, splice junction pathogenic variants and substitutions of glycine and other amino acids may cause the dominant form of DEB. In contrast, more than 400 pathogenic variants have been described for the two types of recessive EB. The recessive form of the disease is caused primarily by null variants, although amino acid substitutions, splice junction variants, and missense variants have also been reported. In-frame exon skipping may generate a partially functional protein in recessive disease. A founder pathogenic variant, c.6527insC (p.R525X), has been observed in 27 of 49 Spanish individuals with recessive DEB. A founder pathogenic variant in COL7A1, pVal769LeuFsXI, was identified in 11 of 15 families in Sfax, Southern Tunisia. Three of 12 individuals carrying at least one copy of this variant developed SCC, including two young-onset cases at ages 16 and 29 years. Genotype-phenotype correlations suggest an inverse correlation between the amount of functional protein and severity.
Pathogenic variants in COL7A1 result in abnormal triple helical coiling and decreased function, which causes increased skin fragility and blistering. In studies of Ras-driven carcinogenesis in RDEB-severe generalized keratinocytes, retention of the amino-terminal NC1, the first noncollagenous fragment of type VII collagen, is tumorigenic in mice. This retained sequence may mediate tumor-stroma interactions that promote carcinogenesis.
Junctional epidermolysis bullosa
Junctional epidermolysis bullosa (JEB) is an autosomal recessive type of EB with an estimated prevalence of 0.49 per million individuals in the United States and an estimated incidence of 2.68 per million live births. JEB results in considerable mortality, with approximately 50% of cases dying within the first year of life. Pathogenic variants in any of the genes encoding the three basic subunits of laminin 332, previously known as laminin 5 (LAMA3, LAMB3, LAMC2), or variants in COL17A1 can result in this syndrome.[142,143,144] Individuals with the Herlitz type (a severe clinical form) of JEB are at increased risk of SCC, with a cumulative risk of 18% by age 25 years. A study of COL17A1 in individuals with a milder subtype of JEB, called JEB-other, identified pathogenic variants in 85 of 86 alleles from 43 individuals. Total loss of COL17A1 protein staining correlated with a more severe phenotype.
Pathogenic variants in either of two adjacent genes on chromosome 17q25 can cause epidermodysplasia verruciformis, a rare heritable disorder associated with increased susceptibility to human papillomavirus (HPV). Infection with certain HPV subtypes can lead to development of generalized nonresolving verrucous lesions, which develop into in situ and invasive SCCs in 30% to 60% of patients. Malignant transformation is thought to occur in about half of these lesions. Approximately 90% of these lesions are attributed to HPV types 5 and 8, although types 14, 17, 20, and 47 have occasionally been implicated. The association between HPV infection and increased risk of SCC has also been demonstrated in people without epidermodysplasia verruciformis; one case-control study found that HPV antibodies were found more frequently in the plasma of individuals with SCC (OR, 1.6; 95% CI, 1.2–2.3) than in plasma from cancer-free individuals.
The genes associated with this disorder, EVER1 and EVER2, were identified in 2002. The inheritance pattern of these genes appears to be autosomal recessive; however, autosomal dominant inheritance has also been reported.[151,152,153] Both of these gene products are transmembrane proteins localized to the endoplasmic reticulum, and they likely function in signal transduction. This effect may be through regulation of zinc balance; it has been shown that these proteins form a complex with the zinc transporter 1 (ZnT-1), which is, in turn, blocked by certain HPV proteins.
A recent case-control study examined the effect of a specific EVER2polymorphism (rs7208422) on the risk of cutaneous SCC in 239 individuals with prior SCC and 432 controls. This polymorphism is a (A > T) coding single nucleotide polymorphism in exon 8, codon 306 of the EVER2 gene. The frequency of the T allele among controls was 0.45. Homozygosity for the polymorphism caused a modest increase in SCC risk, with an adjusted OR of 1.7 (95% CI, 1.1–2.7) relative to wild-type homozygotes. In this study, those with one or more of the T alleles were also found to have increased seropositivity for any HPV and for HPV types 5 and 8, as compared with the wild type.
Some evidence suggests nonallelic heterogeneity in epidermodysplasia verruciformis. An individual born to consanguineous parents with epidermodysplasia verruciformis and additional bacterial and fungal infections was found to have homozygous R115X pathogenic variants in the MST1 gene. Another susceptibility locus associated with this disorder has been identified at chromosome regions 2p21-p24 through linkage analysis of an affected consanguineous family. Unlike those with pathogenic variants in the EVER1 and EVER2 genes, affected individuals linked to this genomic region were infected with HPV 20 rather than the usual HPV subtypes associated with this disorder, and this family did not have a history of cutaneous SCC.
Fanconi anemia is a complex disorder that is characterized by increased incidence of hematologic and solid tumors, including SCC of the skin. Fanconi anemia is inherited as an autosomal recessive disease. It is a relatively rare syndrome with an estimated carrier frequency of one in 181 individuals in the United States (range: 1 in 156 to 1 in 209) and a carrier frequency of up to 1 in 100 individuals of Ashkenazi Jewish ancestry. Leukemia is the most commonly reported cancer in this population, but increased rates of gastrointestinal, head and neck, and gynecologic cancers have also been seen. By age 40 years, individuals affected with Fanconi anemia have an 8% risk per year of developing a solid tumor; the median age of diagnosis for solid tumors is 26 years. Multiple cases of cancers of the brain, breast, lung, and kidney (Wilms tumor) have been reported in this population. Data on the incidence of NMSCs in this population are sparse; however, review of the literature suggests that the age of diagnosis is between the mid-20s and early 30s and that women seem to be affected more often than men.[160,161,162,163,164]
Individuals with this disease have increased susceptibility to DNA cross-linking agents (e.g., mitomycin-C or diepoxybutane) and ionizing and UV radiation. The diagnosis of this disease is made by observing increased chromosomal breakage, rearrangements, or exchanges in cells after exposure to carcinogens such as diepoxybutane.
Seventeen complementation groups have been identified for Fanconi anemia; details regarding the genes associated with these groups are listed in Table 7 below. Exome sequencing has revealed that a subset of individuals can carry multiple heterozygous pathogenic variants in Fanconi anemia genes, which may impact phenotypic presentation.
The proteins involved with DNA crosslink repairs have been termed the FANC pathway because of their involvement with Fanconi anemia. They interact with several other proteins associated with hereditary cancer risk, including those for Bloom syndrome and ataxia-telangiectasia. Further investigation has revealed that FANCD1 is the same gene as BRCA2, a gene that causes predisposition to breast and ovarian cancer. Other Fanconi anemia genes, FANCJ (BRIP1) and FANCN (PALB2), have also been identified as rare breast cancer susceptibility genes. (Refer to the PDQ summary on Genetics of Breast and Gynecologic Cancers for more information about BRCA2, BRIP1, PALB2, and RAD51.) Individuals who are heterozygous carriers of other Fanconi anemia–associated variants do not appear to have an increased risk of cancer, with the possible exception of a twofold increase in breast cancer incidence in carriers of FANCC pathogenic variants.
In 2018, a group reported a significant increase in SCC cases (OR, 1.69; 95% CI, 1.26–2.26) associated with a specific BRCA2 allele, which is relatively prevalent in the Icelandic population (K3326X; allele frequency, 1.1%). This allele results in normal production of an altered protein, and the authors hypothesized carriers have an increased sensitivity to environmental factors, which require DNA repair. This variant was also associated with an increased risk of small cell lung cancer, breast cancer, and ovarian cancer (but lower than the risk associated with the BRCA pathogenic variants that decrease protein levels).
Dyskeratosis congenita (Zinsser-Cole-Engman syndrome)
Dyskeratosis congenita, like Werner syndrome, results in premature aging and is thus considered a progeroid disease. The classic clinical triad for diagnosis includes nail dystrophy, reticular pigmentation of the chest and neck, and oral leukoplakia. In addition, individuals with this disorder are at markedly increased risk of myelodysplastic syndrome, acute leukemia, and bone marrow failure. Ocular, dental, neurologic, gastrointestinal, pulmonary, and skeletal abnormalities have also been described in conjunction with this disease, but clinical expressivity is variable. Developmental delay may also be present in variants of dyskeratosis congenita, such as Hoyeraal-Hreidarsson syndrome (HHS) and Revesz syndrome.
Approximately 10% of individuals with dyskeratosis congenita will develop nonhematologic tumors, often before the third decade of life.[173,174] Solid tumors may be the first manifestation of this disorder. Head and neck cancers were the most commonly reported, accounting for nearly half of the cancers observed. Cutaneous SCC occurred in about 1.5% of the subjects, and the median age at diagnosis was 21 years. These cancers are generally managed as any other SCC of the skin.
Several genes associated with telomere function (DKC1, TERC, TINF2, NHP2, NOP10, RTEL1 and TERT) have been implicated in dyskeratosis congenita; approximately one-half of the individuals with a clinical diagnosis of this disease have an identified pathogenic variant in one of these seven genes.[175,176,177,178,179,180,181,182]TERC and TINF2 are inherited in an autosomal dominant manner, whereas NHP2 (NOLA2) and NOP10 (NOLA3) show autosomal recessive inheritance, and RTEL1 and TERT can be either autosomal dominant or autosomal recessive. Recessive pathogenic variants in RTEL1 can also be associated with HHS. A study of more than 1,000 individuals of Ashkenazi Jewish ancestry identified a founder RTEL1 splice-site pathogenic variant, c.3791G>A (p.R1264H), that had a carrier frequency of 1% in Orthodox Ashkenazi Jewish individuals and 0.45% in the general Ashkenazi Jewish population.DKC1 shows an X-linked recessive pattern. Alterations in these genes result in shortening of telomeres, which in turn leads to defects in proliferation and spontaneous chromosomal rearrangements. Levels of TERC, the RNA component of the telomerase complex, are reduced in all dyskeratosis congenita patients. Missense pathogenic variants in WRAP53, a gene with a protein product that facilitates trafficking of telomerase, have also been associated with an autosomal recessive form of dyskeratosis congenita. Pathogenic variants in C16orf57 were identified in 6 of 132 families who did not have a variant detected in other known genes.C16orf57 pathogenic variants are also associated with poikiloderma with neutropenia. (Refer to the Rothmund-Thomson syndrome section of this summary for more information about poikiloderma congenitale.)
The recommended approach for diagnosis begins with a six-cell panel assay for leukocyte telomere length testing. If telomere length is in the lowest 1% for three or more cell types, molecular genetic testing is indicated. Testing of DKC1 may be performed first in male probands, as pathogenic variants in this gene account for up to 36% of those identified in dyskeratosis congenita to date. Pathogenic variants in TINF2 and TERT are responsible for 11% to 24% and 6% to 10% of cases, respectively.[172,179,180,191,192]
Rothmund-Thomson syndrome, also known as poikiloderma congenitale, is a heritable disorder characterized by chromosomal instability. The cutaneous presentation of this condition is an erythematous, blistering rash appearing on the face, buttocks, and extremities in early infancy. Other characteristics of this syndrome include telangiectasias, skeletal abnormalities, short stature, cataracts, and increased risk of osteosarcoma. Areas of hyperpigmentation and hypopigmentation of the skin develop later in life, and BCC or SCC can develop at an early age. Reports of multiple SCCs in situ have been reported in individuals as young as 16 years. The precise increased risk of skin cancer is not well characterized, but the point prevalence of keratinocyte carcinomas, including both BCC and SCC, is 2% to 5% in young individuals affected by this syndrome. This prevalence is clearly greater than that found in individuals in the same age group in the general population. Although increased UV sensitivity has been described, SCCs are also found in areas of the skin that are not exposed to the sun.
A pathogenic variant in the gene RECQL4 is present in 66% of clinically affected individuals. This gene is located at 8q24.3, and inheritance is believed to be autosomal recessive. RECQL4 encodes the ATP-dependent DNA helicase Q4, which promotes DNA unwinding to allow for cellular processes such as replication, transcription, and repair. A role for this protein in repair of DNA double-strand breaks has also been suggested. Pathogenic variants in similar DNA helicases lead to the inherited disorders of Bloom syndrome and Werner syndrome.
At least 19 different truncating pathogenic variants in this gene have been identified as deleterious. These pathogenic variants cause severe down-regulation of RECQL4 transcripts in this subset of individuals with Rothmund-Thomson syndrome. Cells deficient in RECQL4 have been found to be hypersensitive to oxidative stress, resulting in decreased DNA synthesis. Deficiencies in the RecQ helicases permit hyper-recombination, thereby leading to loss of heterozygosity. Loss of heterozygosity associated with deficiencies of this protein suggests that the helicases are caretaker-type tumor suppressor proteins.
Three of six families with Rothmund-Thomson syndrome were found to have homozygous pathogenic variants in the C16orf57 gene. Pathogenic variants in this gene have also been identified in individuals with dyskeratosis congenita and poikiloderma with neutropenia, suggesting that these syndromes are related;[188,189] however, skin cancer risk in these conditions is not well characterized. (Refer to the Dyskeratosis congenita [Zinsser-Cole-Engman syndrome] section of this summary for more information.)
Loss of genomic stability is also the major cause of Bloom syndrome. This disorder shows increased chromosomal breakage and is diagnosed by increased sister chromatid exchanges on chromosomal analysis. Clinical manifestations of Bloom syndrome include severe growth retardation, recurrent infections, diabetes, chronic pulmonary disease, and an increased susceptibility to cancers of many types. The typical skin lesion seen in this disorder is a photosensitive erythematous telangiectatic rash that occurs in the first or second year of life. Although it is most commonly found on the face, it can also be present on the dorsa of hands or forearms. SCC of the skin is the third most common malignancy associated with this disorder. Skin cancer accounts for approximately 9% of tumors in the Bloom Syndrome Registry. Skin cancers occur at an early age in this population, with a mean age of 31 years at the time of diagnosis.
The BLM gene, located on the short arm of chromosome 15, is the only gene known to be associated with Bloom syndrome. This gene encodes a 1,417-amino acid protein that is regulated by the cell cycle and demonstrates DNA-dependent ATPase and DNA duplex-unwinding activities. Its helicase domain shows considerable similarity to the RecQ subfamily of DNA helicases. Absence of this gene product is thought to destabilize other enzymes that participate in DNA replication and repair.[203,204]
This rare chromosomal breakage syndrome is inherited in an autosomal recessive manner and is characterized by loss of genomic stability. Sixty-four pathogenic variants described in the BLM gene include nucleotide insertions and deletions (41%), nonsense variants (30%), variants resulting in mis-splicing (14%), and missense variants (16%).[205,206] A specific pathogenic variant identified in the Ashkenazi Jewish population is a 6-bp deletion/7-bp insertion at nucleotide 2,281, designated as BLMASH. Many of these variants result in truncation of the C-terminus, which prevents normal localization of this protein to the nucleus. Absence of functional BLM protein can cause increased rates of pathogenic variants and recombination. This somatic hypermutability leads to an increased risk of cancer at an early age in virtually every organ, including the skin.
Cells from people with Bloom syndrome have been found to have abnormal responses to UV radiation. Normal nuclear accumulation of TP53 after UV radiation was absent in 2 of 11 primary cultures from individuals with Bloom syndrome; in contrast, responses in cultures from people who have XP and ataxia-telangiectasia were normal. The gene product of the BLM gene has also been found to complex with Fanconi proteins, raising the possibility of connections between the BLM and Fanconi anemia pathways for DNA stability.
Like Bloom syndrome, Werner syndrome is characterized by spontaneous chromosomal instability, resulting in increased susceptibility to cancer and premature aging. Diagnostic criteria, often in the setting of consanguinity, include cataracts, short stature, premature graying or thinning of hair, and a positive 24-hour urinary hyaluronic acid test. Cardinal cutaneous manifestations of this disorder consist of sclerodermatous skin changes, ulcerations, atrophy, and pigmentation changes. Individuals with this syndrome have an average life expectancy of fewer than 50 years. Cancers have an early onset and occur in up to 43% of these patients. The spectrum of tumors associated with this disorder has primarily been described in the Japanese population and includes an increased incidence of sarcoma, thyroid cancers, and skin cancers. Approximately 20% of the cancers reported in this syndrome are cutaneous, with melanoma and SCC of the skin accounting for 14% and 5%, respectively. A study of 189 individuals with Werner syndrome estimated melanoma risk to be elevated 53-fold in these individuals. SCC was less frequently diagnosed. Acral lentiginous melanomas are overrepresented, and SCCs may exhibit more aggressive behavior, with metastasis to lymph nodes and internal organs.[212,215]
Pathogenic variants in the WRN gene on chromosome 8p12-p11.2 have been identified in approximately 90% of individuals with this syndrome; no other genes are known to be associated with Werner syndrome.[211,216,217,218,219] Inheritance of this gene is believed to be autosomal recessive. The product of the WRN gene is a multifunctional protein including a DNA exonuclease and an ATP-dependent DNA helicase belonging to the RecQ subfamily. This protein may play a role in processes such as DNA repair, recombination, replication, transcription, and combined DNA functions.[220,221,222,223,224,225,226,227,228] Telomere dysfunction has been associated with premature aging and cancer susceptibility. Other helicases with similar function are altered in other chromosomal instability syndromes, such as BLM in Bloom syndrome and RecQL4 in Rothmund-Thomson syndrome.
Pathogenic variants described in the WRN gene include all types of variants; however, the 1136C→T variant is the most common and is found in 20% to 25% of the Japanese and white populations.[230,231] In the Japanese population, a founder pathogenic variant (IVS 25-1G→C) is present in 60% of affected individuals.
Pathogenic variants in the WRN gene causes loss of nuclear localization of the gene product. Intracellular levels of the mRNA and protein associated with the variant are also markedly decreased, compared with those of the wild type. Half-lives of the mRNA and protein associated with the variant are also shorter than those associated with the wild-type mRNA and protein.[231,233]
Prevention and treatment of skin cancers
A phase III, double-blind, placebo-controlled clinical trial evaluated the effects of oral nicotinamide (vitamin B3) in 386 individuals with a history of at least two NMSCs within 5 years before study enrollment. After 12 months of treatment, those taking nicotinamide 500 mg twice daily had a 30% reduction in the incidence of new SCCs (95% CI, 0%–51%; P = .05). A statistically significant reduction was also seen in actinic keratoses, the precursor skin lesions to SCCs. The rate of new NMSCs was 23% lower in the nicotinamide group (95% CI, 4–38, P = .02) than in the placebo group. No clinically significant differences in adverse events were observed between the two groups, and there was no evidence of benefit after discontinuation of nicotinamide. Of note, this study was not conducted in a population with an identified genetic predisposition to SCC.
Because many of the syndromes described above are rare, few clinical trials have been conducted in these specific populations. However, valuable information has been developed from the clinical management experience related to skin cancer risk and treatment in the XP population. Strict sun avoidance beginning in infancy, use of protective clothing, and close clinical monitoring of the skin are key components to management of XP. Full-body photography of the skin, conjunctivae, and eyelids is recommended to aid in follow-up. Although few studies on treatment of SCC in the XP population have been done, in most cases treatment is similar to what would be recommended for the general population. Actinic keratoses are treated with topical therapies such as fluorouracil (5-FU), cryotherapy with liquid nitrogen, or dermabrasion, whereas cutaneous cancers are generally managed surgically.
Oral isotretinoin has been used as chemoprevention in XP patients with promising results. A small study of daily use of isotretinoin (13-cis retinoic acid; given as 2 mg/kg/day) reduced NMSC incidence by 63% in a small number of people with XP. Toxicities associated with this treatment included mucocutaneous symptoms, abnormalities in liver function tests and triglyceride levels, and musculoskeletal symptoms such as arthralgias, calcifications of tendons and ligaments, and osteoporosis.[237,238] Dose reduction to 0.5 mg/kg/day reduced toxicity and decreased skin cancer frequency in three of seven subjects (43%); increasing the dose to 1 mg/kg/day resulted in decreased skin cancer frequency in three of the four subjects who did not respond at the lower dose. Oral isotretinoin use may be useful as a chemopreventive agent in other hereditary skin cancer syndromes, including basal cell nevus syndrome (BCNS), Rombo syndrome, EB, and epidermodysplasia verruciformis.[240,241]
Level of evidence (oral isotretinoin for XP): 3aii
Level of evidence (oral isotretinoin for BCNS, Rombo syndrome, epidermodysplasia verruciformis): 5
Topical T4N5 liposome lotion, containing the bacterial enzyme T4 endonuclease V, was also investigated as a chemopreventive agent in a randomized, placebo-controlled trial of 30 XP patients. Although no effect was seen on incidence of SCC, 17.7 fewer actinic keratoses per year were seen in the treatment group. Additionally, 1.6 fewer BCCs per year were observed in patients being treated with this therapy. Both of these results were statistically significant. The risk of BCC was reduced by 47%, which was of borderline statistical significance. No significant adverse effects of this agent were reported. To date, this agent has not been approved for use by the U.S. Food and Drug Administration.
Level of evidence: 1aii
For patients with XP and unresectable SCC, therapy with 5-FU has been investigated. Several treatment methods were used in this prospective study, including topical therapy to the lesions, short systemic infusion with folic acid, and continuous systemic infusion in combination with cisplatin. Topical 5-FU demonstrated some efficacy, but in some cases viable tumor remained in the deeper dermis. The systemic chemotherapy resulted in one complete response and three partial responses in a total of five patients, suggesting that this therapy may be an option for treatment of extensive lesions. A dose reduction of 30% to 50% has been recommended for systemic chemotherapeutic agents in this population because of the increased sensitivity of XP cells.
Level of evidence: 3diii
For patients with EB, wide local excision of SCC with 2 cm margins remains the treatment of choice. Amputation may be considered as an option to reduce disease recurrence, although it is not clear that this has an impact on survival. The role of sentinel lymph node biopsy remains unclear in this population.
Current guidelines recommend that individuals with EB and unresectable SCC be treated with radiation therapy, but the dose may need to be given in smaller fractions in order to decrease the risk of skin desquamation. Systemic therapy with epidermal growth factor receptor antagonists or tyrosine kinase inhibitors may also be considered for individuals with advanced SCC.
For people who have genetic disorders other than XP, data are lacking, but general sun-safety measures remain important. Careful protection of the skin and eyes is the mainstay of prevention in all patients with increased susceptibility to skin cancer. Key points include avoidance of sun exposure at peak hours, protective clothing and lenses, and vigilant use of sunscreen. Avoidance of x-ray therapy has also been advocated for some groups with hereditary skin cancer syndromes, such as those with epidermodysplasia verruciformis. However, XP patients with unresectable skin cancers or internal cancers, such as spinal cord astrocytoma or glioblastomas of the brain, have been treated successfully with standard therapeutic doses of x-ray radiation. Some experts recommend dermatologic evaluation every 6 months and ophthalmologic evaluation at least annually in these high-risk populations. Guidelines for the management of patients with EB recommend skin examinations every 3 to 6 months starting at age 10 years for individuals with the RDEB-sev gen subtype of the disease. For individuals with other subtypes of EB, skin examination every 6 to 12 months starting at age 20 years is recommended in the absence of an established SCC diagnosis. Dental examination every 6 months is also recommended in this population.
For individuals with DEB, wound care is paramount. Use of silver sulfadiazine cream, medical grade honey, and soft silicone dressings can be helpful in these settings. Attention to nutritional status, which may be compromised because of esophageal strictures, iron-deficiency anemia, infection, and inflammation, is another critical consideration for wound healing for these patients. Multivitamin supplementation, often at higher doses than those routinely recommended for the general population, may be warranted.
Bone marrow transplantation has been explored in patients with DEB; however, there is no evidence that this intervention results in a reduction of skin cancer. A double-blind, randomized, placebo-controlled trial of infusion of nonhematopoietic bone marrow stem cells with or without cyclosporine was conducted in 14 patients with recessive DEB. The rationale for this study was that mesenchymal stem cells (MSCs) have the potential to differentiate into dermal fibroblasts, the main expressor of type VII collagen. Seven subjects were randomly assigned to receive MSCs with 5 mg/kg/day of cyclosporine and an additional seven subjects received only MSCs. The number of new blisters and the rate of blister healing were significantly improved in both groups (P = .003 for the number of new blisters in the combination therapy group and P = .004 in the group receiving MSCs only; P < .001 for the rate of blister healing in both groups). However, no difference was seen between the groups.
Level of evidence (MSCs for blister prevention): 1b
Level of evidence (MSCs for blister treatment): 1
Future therapies for epidermolysis bullosa
Researchers are taking advantage of recent technological advances to study new strategies for the treatment of dominant and recessive EB.[248,249,250,251] Clustered regularly interspaced short palindromic repeats (CRISPR)-associated protein 9 (Cas9) is a technology that can be used to edit DNA. One research group used CRISPR/Cas9 to correct an inherited pathogenic variant in COL7A1 in keratinocytes isolated from a patient with RDEB. Keratinocytes that contained the corrected version of COL7A1 were successfully transplanted onto mice and staining of skin grafts after transplant showed normal skin. Another study used a different approach, retrovirus infection, to introduce normal COL7A1 into keratinocytes from four RDEB patients. The corrected keratinocytes were then assembled into epidermal graft sheets and transplanted onto six wound areas of each of the four patients. The grafts were well tolerated and showed greater healing capabilities than did noncorrected skin after further study. All of these therapies are still in early research stages and have not yet been evaluated in clinical trials.
Level of evidence: None assigned
Rare, high-penetrance and common, low-penetrance genetic factors for melanoma have been identified, and approximately 5% to 10% of all melanomas arise in multiple-case families. However, a significant fraction of these families do not have detectable pathogenic variants in specific susceptibility genes. The frequency with which multiple-case families are ascertained and specific genetic variants are identified differs substantially between populations and geographic regions. A major population-based study has concluded that the high-penetrance susceptibility gene CDKN2A does not make a large contribution to the incidence of melanoma.
Melanoma can also occur in other parts of the body, such as the eye and brain. (Refer to the PDQ summary on Intraocular (Uveal) Melanoma Treatment for more information about melanoma in the eye.)
Risk Factors for Melanoma
This section focuses on risk factors in individuals at increased hereditary risk of developing melanoma. (Refer to the PDQ summary on Skin Cancer Prevention for information about risk factors for melanoma in the general population.)
Sun exposure is well established as a major etiologic factor in all forms of skin cancer, although its effects differ by cancer type. The relationship between sun exposure, sunscreen use, and the development of skin cancer is complex. It is complicated by negative confounding (i.e., subjects who are extremely sun sensitive deliberately engage in fewer activities in direct sunlight, and they are more likely to wear sunscreen when they do). These subjects are genetically susceptible to the development of skin cancer by virtue of their cutaneous phenotype and thus may develop skin cancer regardless of the amount of sunlight exposure or the sun protection factor of the sunscreen.[2,3]
Pigmentary characteristics are important determinants of melanoma susceptibility. There is an inverse correlation between melanoma risk and skin color that goes from lightest skin to darkest skin. Dark-skinned ethnic groups have a very low risk of melanoma on pigmented skin surfaces; however, individuals in these groups develop melanoma on less-pigmented acral surfaces (palms, soles, nail beds) at the same frequency as light-skinned individuals. Among relatively light-skinned individuals, skin color is modified by genetics and behavior. Melanocortin 1 receptor (MC1R) is one of the major genes controlling pigmentation (refer to the MC1R section of this summary); other pigmentation genes are under investigation. A relatively new area of investigation is the evaluation of multiple variants together, often referred to as a polygenic risk score (PRS).[5,6,7] The integration of these models with known phenotype risk factors and their overall clinical utility in melanoma risk prediction have not been demonstrated. PRSs for melanoma are not clinically available at this time.
Clinically, several pigmentary characteristics are evaluated to assess the risk of melanoma and other types of skin cancer. These include the following:
Nevi (or moles) are sharply circumscribed benign pigmented lesions of the skin or mucosa composed of nest melanocytes. Patients with multiple nevi demonstrate increased risk of melanoma. While there is evidence that both the presence of multiple nevi and the presence of multiple clinically atypical nevi are associated with an increased risk of melanoma, most studies demonstrate a stronger risk of melanoma with the presence of atypical nevi.[10,11,12,13] In addition, patients with multiple atypical nevi, regardless of personal and/or family history of melanoma, are at significantly increased risk of developing melanoma than are patients without atypical nevi. A population-based study in the United Kingdom that identified genetic risk factors for the development of nevi showed that some of the same variants are modestly associated with melanoma risk.
The phenotype of multiple nevi has both familial and environmental affecters. The number of nevi can increase with childhood sun exposure.[16,17] The analysis of this association is complex because the use of sun protection strongly correlates with sun exposure. Inheritance of the specific phenotype of a high number of nevi, including clinically atypical nevi, was initially reported as an autosomal dominant trait under the names dysplastic nevus syndrome  and familial atypical multiple mole-melanoma syndrome. A portion of this inherited phenotype is attributed to the major melanoma risk gene CDKN2Adiscussed below. Even within gene carriers in high-risk families, sun exposure seems to affect nevus number.
Generally, a family history of melanoma appears to increase risk of melanoma by about twofold. A family cancer registry study assessed over 20,000 individuals with melanoma and found a standardized incidence ratio (SIR) of 2.62 for offspring of individuals with melanoma and 2.94 for siblings. A larger study of more than 200,000 individuals from the Nurses' Health Study, the Nurses' Health Study 2, and the Health Professionals Follow-up Study found that individuals with a family history of melanoma had an increased risk of melanoma (hazard ratio [HR], 1.74; 95% confidence interval [CI], 1.45–2.09). Slightly higher melanoma risks were found in a population-based study of 1,506,961 individuals in Western Australia; first-degree relatives (FDRs) of 5,660 individuals with melanoma showed an HR for melanoma of 3.58 (95% CI, 2.43–5.43). Another population-based study of more than 238,000 FDRs of 23,000 melanoma patients found a lifetime cumulative risk of melanoma of 2.5% to 3%, which is about double the risk of the general population. Risk based on family history is dependent not only on the number of individuals in the family who have a melanoma but also on the number of melanomas in each family member. For example, the familial risk of melanoma was found to increase 2.2-fold (95% CI, 2.2–2.3) with a single FDR who has one melanoma and up to 16.3-fold (95% CI, 9.5–26.1) with a single FDR who has five or more melanomas. When two or more family members were diagnosed with melanoma before age 30 years, the lifetime cumulative risk for the family members rose to 14%.
A study on the heritability of cancer among 80,309 monozygotic and 123,382 dizygotic twins showed that melanoma has a heritability of 58% (95% CI, 43%–73%), suggesting that more than half of the risk of melanoma is caused by inherited factors. A study looking at the contribution of family history to melanoma risk showed a population-attributable fraction ranging from less than 1% in northern Europe to 6.4% in Australia, suggesting that only a small percentage of melanoma cases are caused by familial factors. Rarely, however, in some families many generations and multiple individuals develop melanoma and are at much higher risk. For individuals from these families, the incidence of melanoma is higher for sun-protected rather than sun-exposed skin.
The major hereditary melanoma susceptibility gene, CDKN2A, is found to be altered in approximately 35% to 40% of families with three or more melanoma cases. To date, more than half of the families with multiple cases of melanoma have no identified pathogenic variant.[29,30]Multigene (panel) testing for pathogenic variants in multiple melanoma genes simultaneously is a consideration; however, a large study of extended panel testing (30 genes) in melanoma families without CDKN2A and CDK4 pathogenic variants had a diagnostic yield of 4%.
The definition of a familial cluster of melanoma varies by geographical region worldwide, because of the role played by ultraviolet (UV) radiation in melanoma pathogenesis. In heavily insolated regions (regions with high ambient sun exposure), three or more affected family members are required; in regions with lower levels of ambient sunlight, two or more affected family members are considered sufficient to define a familial cluster. The American College of Medical Genetics and Genomics and the National Society of Genetic Counselors recommend that an individual with any of the following characteristics be referred for a cancer genetics consultation:
Personal history of melanoma
A previous melanoma places one at high risk of developing additional primary melanomas, particularly for people with the most common risk factors for melanoma, such as cutaneous phenotype, family history, a pathogenic variant in CDKN2A, a great deal of early-life sun exposure, and numerous or atypical nevi. In the sporadic setting, approximately 5% of melanoma patients develop more than one primary cancer, while in the familial setting the corresponding estimate is 30%. This greater-than-expected rate of multiple primary cancers of the same organ is a common feature of hereditary cancer susceptibility syndromes; it represents a clinical finding that should raise the level of suspicion that a given patient's melanoma may be related to an underlying genetic predisposition. Risk of a second primary melanoma after diagnosis of a first primary melanoma is approximately 5% and is greater for males and older patients.[33,34,35,36] A study in Sweden of more than 65,000 individuals with melanoma found a SIR of 2.8 (95% CI, 2.3–3.4) for a second melanoma in individuals with a family history of melanoma and a SIR of 2.5 (95% CI, 2.3–2.7) in individuals with no family history. The risk of a second melanoma increased when the first melanoma was diagnosed before age 40 years (SIR, 4.7; 95% CI, 3.9–5.6%). The SIRs increased with increasing numbers of melanomas.
Personal history of BCC or SCC
Having a personal history of basal cell carcinoma (BCC) or squamous cell carcinoma (SCC) is also associated with an increase in risk of a subsequent melanoma.[38,39,40] Depending on the study, this risk ranges from a nonsignificant increase for melanoma with a previous SCC of 1.04 (95% CI, 0.13–8.18) to a highly significant risk of 7.94 (95% CI, 4.11–15.35).[41,42] It is likely that this relationship is the result of shared risk factors (of which sun exposure is presumably one), rather than a specific genetic factor that increases risk of both. Pigmentary characteristics are critically important for the development of melanoma, and cutaneous phenotype (described above), in combination with excessive sun exposure, is associated with an increased risk of all three types of skin cancers.
Major Genes for Melanoma
The major gene associated with melanoma is CDKN2A/p16, cyclin-dependent kinase inhibitor 2A, which is located on chromosome 9p21. This gene has multiple names (MTS1, INK4, and MLM) and is commonly called by the name of its protein, p16. It is an upstream regulator of the retinoblastoma gene pathway, acting through the cyclin D1/cyclin-dependent kinase 4 complex. This tumor suppressor gene has been intensively studied in multiple-case families and in population-based series of melanoma cases. CDKN2A controls the passage of cells through the cell cycle and provides a mechanism for holding damaged cells at the G1/S checkpoint to permit repair of DNA damage before cellular replication. Loss of function of tumor suppressor genes—a good example of which is CDKN2A—is a critical step in carcinogenesis for many tumor systems.
CDKN2A encodes two proteins, p16INK4a and p14ARF, both inhibitors of cellular senescence. The protein produced when the alternate reading frame (ARF) for exon 1 is transcribed instead of the standard reading frame exerts its biological effects through the p53 pathway. It mediates cell cycle arrest at the G1 and G2/M checkpoints, complementing p16's block of G1/S progression—thereby facilitating cellular repair of DNA damage.
Pathogenic variants in CDKN2A account for 35% to 40% of familial melanomas  and fewer than 1% of unselected melanoma cases. A study of more than 1,000 individuals in Spain showed that 6.6% of individuals with melanoma have a family history of two or more FDRs with melanoma, and up to 15% have a family history suggestive of familial melanoma that includes melanoma or pancreatic cancer diagnoses in FDRs or second-degree relatives (SDRs). A large case series from Britain found that CDKN2A pathogenic variants were present in 100% of families with seven to ten individuals affected with melanoma, 60% to 71% of families with four to six cases, and 14% of families with two cases. A similar study of Greek individuals with melanoma found CDKN2A pathogenic variants in 3.3% of single melanoma cases, 22% of familial melanoma cases, and 57% of individuals with multiple primary melanomas (MPM). A study of 92 sequential cases of Italian individuals with familial atypical multiple mole-melanoma syndrome (defined as three or more individuals with primary cutaneous melanoma or one individual with MPM) found CDKN2A pathogenic variants in 20% of individuals, including three unrelated individuals with a p.D84V variant.Cascade testing identified 14 of 40 unaffected family members undergoing testing who carried their family's CDKN2A pathogenic variant. However, a second study of 106 familial melanoma cases (defined as at least two melanoma cases) only found CDKN2A pathogenic variants in 8.3% of cases. The frequency of CDKN2A pathogenic variants is as high as 22% in families with two cases of melanoma who have other features of hereditary melanoma, such as an age at diagnosis younger than 50 years or one or more individuals diagnosed with MPM. A study of 587 individuals with a single primary melanoma or MPM found CDKN2A pathogenic variants in 19% of individuals with MPM relative to 4.4% of individuals with a single primary melanoma.CDKN2A pathogenic variants were found in 29.6% of individuals with three or more primary melanomas. Individuals with more than three primary melanomas and a family history of melanoma (undefined) had a frequency of CDKN2A pathogenic variants of 58.8%. Many pathogenic variants reported among families consist of founder variants, which are unique to specific populations and the geographic areas from which they originate.[50,51,52,53,54,55,56,57]
Depending on the study design and target population, melanoma penetrance related to CDKN2A pathogenic variants differs widely. One study of 80 multiple-case families demonstrated that the penetrance varied by country, an observation that was attributed to major differences in sun exposure. For example, in Australia, the penetrance was 30% by age 50 years and 91% by age 80 years; in the United States, the penetrance was 50% by age 50 years and 76% by age 80 years; in Europe, the penetrance was 13% by age 50 years and 58% by age 80 years. In contrast, a comparison of families with the CDKN2A pathogenic variant in the United Kingdom and Australia demonstrated the same cumulative risk of melanoma; for CDKN2A carriers, the risk of developing melanoma seemed independent of ambient UV radiation. Another study of individuals with melanoma identified in eight population-based cancer registries and one hospital-based sample obtained a self-reported family history and sequenced CDKN2A in all individuals. The penetrance was estimated as 14% by age 50 years and 28% by age 80 years. The explanation for these differences lies in the method of identifying the individuals tested, with penetrance estimates increasing with the number of affected family members. The method of family ascertainment in the latter study made it much less likely that "heavily loaded" melanoma families would be identified. Coinheritance of MC1R variants also increases CDKN2A penetrance; this genetic variant, described in further detail below, is therefore both a low-penetrance susceptibility gene and a modifier gene. (Refer to the MC1R section of this summary for more information.) Other modifier loci have also been assessed in CDKN2A carriers; interleukin-9 (IL9) and GSTT1 were the only loci with effects that reached statistical significance, suggesting that other minor risk factors may interact with major risk loci.[61,62]
One study reported a melanoma incidence rate of 9.9 per 1,000 person years among 354 FDRs and 2.1 per 1,000 person years among 391 SDRs of probands with a p16-Leiden (c.225-243del19) CDKN2A pathogenic variant (95% CIs of 7.4–13.3 and 1.2–3.8, respectively). These data indicate a melanoma rate that is much higher than that of the general population (12.9-fold increased incidence) for SDRs in untested relatives of carriers of CDKN2A pathogenic variants.
A comparison of clinical features from 182 patients with CDKN2A pathogenic variants and 7,513 individuals without variants found that individuals with CDKN2A pathogenic variants were statistically significantly younger at diagnosis (mean age at diagnosis, 39.0 y vs. 54.3 y; P < .001). There was also a 5-year cumulative incidence of a second melanoma of 23.4% in carriers of pathogenic variants and a rate of 2.3% in controls who were negative for a pathogenic variant. A study of pediatric patients with melanoma (aged 9–19 y) in melanoma-prone families reported a significant increase in melanoma prevalence (6-fold to 28-fold) relative to the general population. In this series, 7 of 21 patients (33%) with CDKN2A pathogenic variants were diagnosed with MPMs before age 20 years. An Italian study performed genotype-phenotype correlations in 100 families with familial melanoma to determine clinical features predictive of the identification of a CDKN2A pathogenic variant. Probands with MPM, at least one melanoma with Breslow thickness greater than 0.4 mm, and more than three affected family members had a greater than 90% likelihood of having a pathogenic variant; probands with none of these features had less than a 1% likelihood of having a CDKN2A pathogenic variant. The most predictive feature was MPM.
Melanomas in carriers of CDKN2A pathogenic variants largely resemble those found sporadically. A large study that compared melanoma pathology between CDKN2A carriers and individuals with sporadic melanoma found few significant differences, with a minor trend of increased pigmentation among pathogenic variant carriers. Another study of more than 670 carriers of CDKN2A pathogenic variants and 1,258 carriers of wild-type or benign CDKN2A variants found that participants with pathogenic variants were more likely to be diagnosed at an earlier age (median age, 38 y vs. 46 y) and have MPM (average number of melanomas, 2.3 vs. 1.4). A smaller study of 106 carriers of CDKN2A pathogenic variants and 199 noncarriers without a history of familial melanoma showed that carriers were more likely to have MPM but had no differences in overall or disease-specific survival compared with noncarriers. However, two pathogenic variants in CDKN2A (p.Arg112dup, p.Pro48Leu) may be prognostic factors in patients with melanoma. After adjusting for age, sex, and tumor classification, carriers of these CDKN2A pathogenic variants had poorer melanoma-specific survival than did non-CDKN2A carriers (HR, 2.5; 95% CI, 1.49–2.21). An early study suggested that somatic NRAS mutations occurred at a higher rate in melanomas diagnosed in Swedish families who carry CDKN2A pathogenic variants as compared with those with sporadic melanomas. However, subsequent studies have found that the rates of the common somatic mutations (BRAF, NRAS) in melanomas from CDKN2A carriers resemble or are lower than those described in the sporadic melanoma population.[72,73] Of note, melanomas from several patients with CDKN2A variants had coexisting BRAF and NRAS mutations, which is an uncommon occurrence in sporadic melanomas.
CDKN2A exon 1ß pathogenic variants (p14ARF) have been identified in a small percentage of families negative for p16INK4a pathogenic variants. In a study of 94 Italian families with two or more cases of melanoma, 3.2% of families had variants in p14ARF. A patient with a balanced translocation between chromosomes 9 and 22 that disrupted p14ARF had melanoma, DNA repair deficiency, and features of DiGeorge syndrome, including deafness and malformed inner ears.
CDKN2A, cutaneous phenotypes, and cancers other than melanoma
In a Melanoma Genetics Consortium (GenoMEL) study of 1,641 family members of melanoma probands, family members with a CDKN2A pathogenic variant were more likely to have atypical nevi than were family members of CDKN2A noncarriers (odds ratio [OR], 1.65; 95% CI, 1.18–2.28). Another study of individuals in Sweden with MPM and two or more cases of melanoma in their first-, second-, or third-degree relatives found CDKN2A pathogenic variants in 43 of 100 cases. Familial MPM cases with CDKN2A variants, familial MPM cases wild-type for CDKN2A, and nonfamilial MPM cases all showed increased risks of future cutaneous SCCs compared with controls (relative risk [RR], 4.8; 95% CI, 1.5–15.1).
Results from the Genes, Environment, and Melanoma study showed that FDRs of carriers of CDKN2A pathogenic variants with melanoma had an approximately 50% increased risk of cancers other than melanoma, compared with FDRs of other melanoma patients. Cancers with increased risk in this population included gastrointestinal cancers (RR, 2.4; 95% CI, 1.4–3.7), pancreatic cancers (RR, 7.4; 95% CI, 2.3–18.7), and Wilms tumor (RR, 40.4; 95% CI, 3.4–352.7). A Spanish study of the FDRs of 66 melanoma patients with known CDKN2A pathogenic variants also showed an increase in prevalence of other cancers, including pancreatic (prevalence ratio [PR], 2.97; 95% CI, 1.72–5.15), lung (PR, 3.04; 95% CI, 1.93–4.80), and breast cancers (PR, 2.19; 95% CI, 1.36–3.55). A large registry study from Sweden that included 27 families carrying the Arg112dup pathogenic variant in CDKN2A observed excess nonmelanoma cancers in both carriers (n = 120) and FDRs (n = 275). For carriers of CDKN2A pathogenic variants, increased risks relative to a control population were seen for pancreatic (RR, 43.8; 95% CI, 13.8–139), upper digestive (RR, 17.1; 95% CI, 6.3–46.5), respiratory (RR, 15.6; 95% CI, 5.4–46.0), and breast cancers (RR 3.0; 95% CI, 0.9–9.9), among others (all cancers: RR, 5.0; 95% CI, 3.7–7.3). The RRs in FDRs were 20.6 (95% CI, 11.6–36.7) for pancreatic cancers, 6.0 (95% CI, 2.8–13.1) for respiratory cancers, 3.3 (95% CI, 1.5–7.6) for upper digestive cancers, and 1.9 (95% CI, 0.9–4.0) for breast cancers, with a RR of all cancers of 2.1 (95% CI, 1.6–2.7). A lesser-increased cancer risk was seen among SDRs. They also observed a significant association between smoking and risk of pancreatic, respiratory, and upper digestive cancers, with an OR of 9.3 (95% CI, 1.9–44.7) for ever-smoking carriers compared with never-smoking carriers.
A few studies have identified individuals with sarcoma who have germline pathogenic variants in CDKN2A, but the number of cases is too small to determine the risk of sarcoma associated with this gene.[81,82] One patient with features of Li-Fraumeni syndrome did not carry a TP53 pathogenic variant, but a deletion of CDKN2A and CDKN2B. A whole-exome sequencing study of a Li-Fraumeni–like family with three individuals with soft tissue sarcoma identified a shared pathogenic CDKN2A variant. An evaluation of 474 melanoma families with cases of sarcoma and 190 TP53 variant–negative Li-Fraumeni–like families found eight additional individuals with sarcoma and pathogenic CDKN2A variants.
A subset of families carrying a CDKN2A pathogenic variant also displays an increased risk of pancreatic cancer.[83,84] The overall lifetime risk of pancreatic cancer in these families ranges from 11% to 17%. The RR has been reported as high as 47.8. Although at least 18 different variants in p16 have been identified in such families, specific pathogenic variants appear to have a particularly elevated risk of pancreatic cancer.[29,87] Pathogenic variants affecting splice sites or Ankyrin repeats were found more commonly in families with both pancreatic cancer and melanoma than in those with melanoma alone. The p16 Leiden variant is a 19-base pair deletion in CDKN2A exon 2 and is a founder pathogenic variant originating in the Netherlands. In one major Dutch study, 19 families with 86 members who had melanoma also had 19 members with pancreatic cancer in their families, a cumulative risk of 17% by age 75 years. In this study, the median age of pancreatic cancer onset was 58 years, similar to the median age at onset for sporadic pancreatic cancer. However, other reports indicate that the average age at diagnosis is 5.8 years earlier for these carriers of pathogenic variants than for those with sporadic pancreatic cancer. Geographic variation may play a role in determining pancreatic risk in these families carrying known pathogenic variants. In a multicontinent study of the features of germline CDKN2A pathogenic variants, Australian families carrying these variants did not have an increased risk of pancreatic cancer. It was also reported that similar CDKN2A variants were involved in families with and without pancreatic cancer; therefore, there are additional factors involved in the development of melanoma and pancreatic cancer. Some families with CDKN2A pathogenic variants may have a pattern of site-specific pancreatic cancer only.[92,93,94] Conversely, melanoma-prone families that do not have a CDKN2A pathogenic variant have not been shown to have an increased risk of pancreatic cancer.
In a review of 110 families with multiple cases of pancreatic cancer, 18 showed an association between pancreatic cancer and melanoma. Only 5 of the 18 families with cases of both pancreatic cancer and melanoma had individuals with multiple dysplastic nevi. These 18 families were assessed for pathogenic variants in CDKN2A; variants were identified in only 2 of the 18 families, neither of which had a dysplastic nevi phenotype.
The melanoma-astrocytoma syndrome is another phenotype caused by pathogenic variants in CDKN2A. The possible existence of this disorder was first described in 1993. A study of 904 individuals with melanoma and their families found 15 families with 17 members who had both melanoma and multiple types of tumors of the nervous system. Another study found a family with multiple melanoma and neural cell tumors that appeared to be caused by loss of p14ARF function or to disruption of expression of p16. Plexiform neurofibromas have also been reported in individuals with deleterious CDKN2A variants.[99,100,101,102]
CDK4 and CDK6
Cyclin-dependent kinases have important roles in progression of cells from G1 to S phase. CDK4 and CDK6 partner with the cyclin–D associated kinases to accelerate the function of the cell cycle. Phosphorylation of the retinoblastoma (Rb) protein in G1 by cyclin-dependent kinases releases transcription factors, inducing gene expression and metabolic changes that precede DNA replication, thus allowing the cell to progress through the cell cycle. These genes are of conceptual significance because they are in the same signaling pathway as CDKN2A.
Germline CDK4 pathogenic variants are very rare, being found in only a handful of melanoma kindreds.[103,104,105] All described families demonstrated a substitution of amino acid 24, suggesting this position as a variant hotspot within the CDK4 gene. Three Latvian families with melanoma have a R24H substitution arising on the same haplotype, which suggests that it could be a founder pathogenic variant in this population. A CDK4 pathogenic variant affects binding of p16 with its subsequent inhibition of CDK4 functionality. With constitutive activation of germline CDK4, CDK4 acts as a dominant oncogene. A small study showed that the melanoma cancer risk in 17 families with CDK4 pathogenic variants was similar to the risk seen in families with CDKN2A variants. (Refer to the CDKN2A/p16 and p14/ARF section of this summary for more information.) In addition, the melanomas found in CDK4 families appear to have similar rates of somatic BRAF mutations to those found in sporadic populations, although because of the rare nature of CDK4 germline variants, the data are necessarily limited.
Despite its functional similarity to CDK4, germline variants in CDK6 have not been identified in any melanoma kindreds.
Telomere maintenance genes
Telomerase reverse transcriptase (TERT)
Linkage of melanoma to a region of chromosome 5p was observed in a single, large kindred with multiple melanomas and other cancers. Sequencing demonstrated a pathogenic variant in the promoter region of a subunit of TERT, which demonstrated increased promoter activity in construct assays. This variant cosegregated with melanoma and other cancers (ovarian, renal, bladder, and lung), with multiple cancers observed in single individuals. At least one affected family member was observed to have numerous nevi. Somatic mutations in the same region were observed in 125 of 168 sporadic melanomas in the same report. A separate study reported pathogenic variants that also increased promoter activity in the same TERT promoter region in 50 of 70 sporadic melanomas. Similar pathogenic variants were seen in 16% of a diverse set of established cancer cell lines, suggesting this might be a common activation variant in multiple cancer types. The frequency of this variant in melanoma families has not been fully investigated, but one study of 273 families with three or more cases of melanoma identified only one family (with 7 melanoma cases) that carried a c.-57 T>G promoter variant. A study of 106 familial melanoma cases (defined as at least two melanoma cases or MPM in the proband) found that 47% of MPM cases and 58% of familial melanoma cases carried a risk-associated TERT promoter variant, rs2853669. The prevalence of this variant in the general population is estimated to be between 25% and 29%.
Exome and genome-sequencing in individuals from hereditary melanoma families led to the identification of missense pathogenic variants in POT1 that segregate with disease in numerous studies.[114,115] A POT1 Ser270Asn missense pathogenic variant was found in 5 of 56 unrelated melanoma families from Italy. This variant was not observed in over 2,000 Italian controls. Ser270Asn is thought to be a founder pathogenic variant, as all families with the variant shared a haplotype. Additional POT1 missense pathogenic variants, including Tyr89Cys, Arg137His, and Gln623His, were identified in other melanoma families and were not seen in unaffected controls.[114,115] Together, POT1 pathogenic variants were found in approximately 4% of melanoma families who lacked CDKN2A or CDK4 variants, suggesting it may be another gene in hereditary melanoma. POT1 binds to single-stranded telomeric repeat regions and is thought to aid in maintenance of telomere length. Most of the variants segregating in families occur in the two oligonucleotide/oligosaccharide-binding domains of the protein, which are the portion of the protein critical for binding DNA. Individuals carrying POT1 pathogenic variants showed longer telomere lengths than melanoma cases without the POT1 variants, suggesting a link between disruption in normal telomere length and melanoma.[114,115] The clinical utility of testing this gene has not yet been established.
In one study, 510 melanoma families were screened by next-generation sequencing for pathogenic variants in genes in the shelterin complex, which protects chromosomal ends. Six families were found to have variants in ACD, and four families had variants in TERF2IP. The ACD variants clustered in the POT1 binding domain. Because some of these variants did not lead to a truncated protein, the functional significance is not confirmed.
Xeroderma pigmentosum (XP) patients with defective DNA repair have a more than 1,000-fold increase in melanoma risk. These patients are diagnosed with melanoma at a significantly younger age than individuals in the general population; on average, melanoma diagnosis occurs at age 22 years in XP patients. The anatomic site distribution of melanomas in XP patients is similar to that of the general population.[118,119]
Genetic polymorphisms associated with DNA repair genes have been associated with mildly increased melanoma risk in the general population. A meta-analysis of eight case-control studies comprising more than 5,000 cases and 7,000 controls found that individuals carrying the Asp1104His polymorphism in XPG had an increased risk of melanoma (OR, 2.42; 95% CI, 2.26–2.60).
(Refer to the Xeroderma pigmentosum section in the Squamous Cell Carcinoma section of this summary for more information.)
BAP1 has recently emerged as a gene implicated both in sporadic and hereditary melanomas. Originally described in a cohort of uveal melanoma patients, BAP1 is a tumor suppressor gene that was found to be inactivated in 84% of uveal melanoma patients with metastases. Although the majority of these variants were somatic, one patient was found to have a germline frameshift variant. A phenotype associated with BAP1 pathogenic variants was subsequently described. Two families with multiple, elevated melanocytic tumors that were clinically and histopathologically distinct from other melanocytic neoplasms were found to have inactivating germline pathogenic variants of BAP1. These tumors, which have been termed melanocytic BAP1-mutated atypical intradermal tumors, or MBAITs, are found throughout the body, generally measure approximately 5 mm, and begin to appear in the second decade of life. MBAITs are 2 mm to 10 mm in diameter, and affected individuals (about 67% of BAP1 pathogenic variant carriers) can have 5 to more than 50 skin lesions.[124,125] Cases of cutaneous melanoma were present in these families, but the rate of malignant progression is thought to be low due to the relative lack of melanomas in comparison with the number of more papular tumors. This syndrome has been called BAP1 tumor syndrome or the COMMON (c utaneous and o cular m elanoma and atypical m elanocytic proliferation with o ther internal n eoplasms) syndrome, and it is inherited in an autosomal dominant pattern. Further investigation has supported the association between familial cutaneous melanoma and uveal melanoma in BAP1 carriers.[127,128,129,130,131] However, potentially pathogenic BAP1 germline variants occur in a low percentage of melanoma cases. One targeted sequencing study of 1,109 unselected cutaneous melanoma cases found only seven germline missense pathogenic variants (<1%). A second series of 1,977 melanoma cases and 754 controls identified 22 rare variants in BAP1 among cases and 5 rare variants among controls; three of the variants found only among cases were confirmed to disrupt BAP1 function and were associated with family histories of other BAP1-associated cancers. In support of a link between melanoma risk and BAP1, in one series, about 18% of individuals with a BAP1 pathogenic variant developed melanoma. In addition, although data are currently limited, patients with germline pathogenic variants in BAP1 may be at increased risk of lung adenocarcinoma, mesothelioma, BCC, and clear cell carcinoma of the kidney.[125,127,129,130,133,134]
Other studies have reported pathogenic variants in BAP1. A missense pathogenic variant (p.Leu570Val) in a family with multiple cases of melanoma was described to affect splicing and result in a frameshift. This family also had cases of uveal melanoma and paraganglioma. Another family with a Y646X BAP1 pathogenic variant had multiple cancers, including multiple cutaneous melanomas and BCCs, uveal melanoma, and mesotheliomas. The authors hypothesized that a gene-environment interaction between BAP1 pathogenic variants and UV radiation and asbestos exposure contributed to the high incidence of multiple cancers in this family.
PTENhamartoma tumor syndromes (including Cowden syndrome)
Cowden syndrome and Bannayan-Riley-Ruvalcaba syndrome (BRRS) are part of a spectrum of conditions known collectively as PTEN hamartoma tumor syndromes. Approximately 85% of patients diagnosed with Cowden syndrome, and approximately 60% of patients with BRRS have an identifiable PTEN pathogenic variant. In addition, PTEN pathogenic variants have been identified in patients with very diverse clinical phenotypes. The term PTEN hamartoma tumor syndromes refers to any patient with a PTEN pathogenic variant, irrespective of clinical presentation.
PTEN functions as a dual-specificity phosphatase that removes phosphate groups from tyrosine, serine, and threonine. Pathogenic variants of PTEN are diverse, including nonsense, missense, frameshift, and splice-site variants. Approximately 40% of variants are found in exon 5, which encodes the phosphatase core motif, and several recurrent pathogenic variants have been observed. Individuals with variants in the 5' end or within the phosphatase core of PTEN tend to have more organ systems involved.
Operational criteria for the diagnosis of Cowden syndrome have been published and subsequently updated.[140,141] These included major, minor, and pathognomonic criteria consisting of certain mucocutaneous manifestations and adult-onset dysplastic gangliocytoma of the cerebellum (Lhermitte-Duclos disease). An updated set of criteria based on a systematic literature review has been suggested  and is currently utilized in the National Comprehensive Cancer Network (NCCN) guidelines. Contrary to previous criteria, the authors concluded that there was insufficient evidence for any features to be classified as pathognomonic. With increased utilization of genetic testing, especially the use of multigene panels, clinical criteria for Cowden syndrome will need to be reconciled with the phenotype of individuals with documented germline PTEN pathogenic variants who do not meet these criteria. Until then, whether Cowden syndrome and the other PTEN hamartoma tumor syndromes will be defined clinically or based on the results of genetic testing remains ambiguous. The American College of Medical Genetics and Genomics (ACMG) suggests that referral for genetics consultation be considered for individuals with a personal history of or a first-degree relative with 1) adult-onset Lhermitte-Duclos disease or 2) any three of the major or minor criteria that have been established for the diagnosis of Cowden syndrome. Detailed recommendations, including diagnostic criteria for Cowden syndrome, can be found in the NCCN and ACMG guidelines.[32,143] Additionally, a predictive model that uses clinical criteria to estimate the probability of a PTEN pathogenic variant is available; a cost-effectiveness analysis suggests that germline PTEN testing is cost effective if the probability of a variant is greater than 10%.
Over a 10-year period, the International Cowden Consortium (ICC) prospectively recruited a consecutive series of adult and pediatric patients meeting relaxed ICC criteria for PTEN testing in the United States, Europe, and Asia. Most individuals did not meet the clinical criteria for a diagnosis of Cowden syndrome or BRRS. Of the 3,399 individuals recruited and tested, 295 probands (8.8%) and an additional 73 family members were found to harbor germline PTEN pathogenic variants. In addition to breast, thyroid, and endometrial cancers, the authors concluded that on the basis of cancer risk, melanoma, kidney cancer, and colorectal cancers should be considered part of the cancer spectra arising from germline PTEN pathogenic variants. A second study of approximately 100 patients with a germline PTEN pathogenic variant confirmed these findings and suggested a cumulative cancer risk of 85% by age 70 years.
The risk of melanoma in PTEN carriers is controversial. In the study of 100 patients referenced above, four women and four men were diagnosed with melanoma and less than one case was expected, for a SIR of 28.3 for women (95% CI, 7.6–35.4) and 39.4 for men (95% CI, 10.6–100.9) (P < .001). In the ICC study described above, an elevated SIR of 8.5 (95% CI, 4.1–15.6) was reported in 368 carriers of PTEN pathogenic variants. In this cohort, the estimated lifetime risk of melanoma in carriers of PTEN pathogenic variants was 6% (range, 1.6%–9.4%). However, it is important to recognize that a subsequent prospective study did not observe an elevated melanoma risk. In this study, only 1 of 180 carriers was diagnosed with melanoma. (Refer to the PDQ summaries on the Genetics of Colorectal Cancer and the Genetics of Breast and Gynecologic Cancers for more information about risks of other cancers in PTEN hamartoma tumor syndromes.)
O-6-methylguanine DNA methyltransferase (MGMT)
In one study of 64 families with familial melanoma that looked for germline genomic rearrangements of 34 tumor suppressor genes, a deletion of the promoter and exon 1 of the MGMT gene was found. The wild-type allele was lost in individuals with melanoma in this family. MGMT is an enzyme involved in DNA repair. Additional melanoma families with variants in this gene need to be identified before a definitive connection between MGMT and familial melanoma can be made.
Additional candidate regions for familial melanoma susceptibility
Several additional loci for familial melanoma have been identified through genome-wide studies. A melanoma susceptibility locus on 1p22 was identified through a linkage analysis of 49 Australian families who had at least three melanoma cases and who were negative for CDKN2A and CDK4 pathogenic variants. Deletion mapping in tumors shows a minimal region of loss of a 9-Mb interval within the peak linkage region, but none of the linkage families have pathogenic variants in the genes tested thus far. A GWAS of individuals from 34 high-risk melanoma families revealed three single nucleotide polymorphisms (SNPs) on 10q25.1 associated with melanoma risk. The ORs for risk for the SNPs ranged from 6.8 to 8.4. Subsequent parametric linkage analysis in one family showed logarithm of the odd scores of 1.5, whereas the other two families assessed did not show linkage. No obvious candidate gene was identified in the genomic region of interest. Two genome-wide linkage studies of 35 and 42 Swedish families identified evidence of linkage on chromosomal regions 3q29, 17p11-12, and 18q22.[152,153] An Italian linkage study showed a suggestive linkage to the same 3q29 locus. No causative genes have been confirmed, but candidates map to all of the loci. None of these loci have been confirmed in independent studies.
Several GWAS have suggested a risk locus for melanoma on chromosome 20q11, with an OR of 1.27.[155,156] This is the location of the ASIP locus that encodes the agouti signaling protein, which controls hair color during the hair growth cycle in some mammals. It acts as an antagonist to MC1R. Although ASIP variation has been associated with variation in human pigmentation, initial studies did not demonstrate an association with melanoma. Additionally, variants in a transcription factor for ASIP, NCOA6, which is also on chromosome 20, showed a maximum OR of 1.82. However, no interaction was seen between these variants and MC1R variants and melanoma risk. The mechanism by which variants at 20q11 cause an increased risk of melanoma remains unclear.
Other risk loci have been reported on chromosomes 2, 5, 6, 7, 9, 10, 11, 15, 16, and 22.[154,159,160,161,162,163,164] A GWAS of melanoma published in 2014 examined eight of the loci with a previous significant association with melanoma, but without a confirmed causal gene. Researchers were able to confirm seven of eight loci and found some evidence supporting the eighth. These included the chromosome 20 locus discussed above and a 9p21 locus distinct from CDKN2A. Candidate genes at these loci seem to be clustered in functional groups associated with skin pigmentation and nevus development, both traits with a known melanoma association. (Refer to the Risk Factors for Melanoma section of this summary for more information about these traits.) A multicenter meta-analysis of 11 GWAS and two data sets included 15,990 cutaneous melanoma cases and 26,409 controls. They reported five melanoma susceptibility loci that involved putative melanocyte regulatory elements, telomere biology, and DNA repair.
A publicly available database, MelGene, maintains lists of variants that have been associated with melanoma risk through GWAS. MelGene also includes network and potential functional relationships between these genes and variants.
When the first data linking CDKN2A pathogenic variants to melanoma risk became available, it was clear that these variants did not account for all the melanoma tumors in which 9p21 loss of heterozygosity could be demonstrated. In fact, 51% of informative cases had deletions that did not involve somatic mutations in CDKN2A. There are data that the golgi membrane protein 1 (GOLM1) gene, mapping to 9q21, may be involved in melanoma risk. Exome sequencing of DNA from 12 sets of cousins with cutaneous melanoma who were negative for known high-risk melanoma genes led to the identification of a rare GOLM1 variant (rs149739829) in three affected individuals in one pedigree. Two additional pairs of related melanoma cases with the putative risk allele were identified. Family-based case-control studies showed association with melanoma risk (OR, 9.81; P < .001). In a population-based case-control study of 1,534 melanoma cases, unselected for family history, and 1,146 controls, there was an increased risk of melanoma in individuals that carried the GOLM1 rs149739829 risk allele (OR, 2.45; P = .02).
Minor Genes (Genetic Modifiers) for Melanoma
The MC1R gene, otherwise known as the alpha melanocyte-stimulating hormone receptor, is located on chromosome 8. Partial loss-of-function pathogenic variants, of which there are at least ten, are associated not only with red hair, fair skin, and poor tanning, but also with increased skin cancer risk independent of cutaneous pigmentation.[169,170,171,172] A comprehensive meta-analysis of more than 8,000 cases and 50,000 controls showed the highest risk of melanoma in individuals with MC1R variants associated with red hair; however, alleles not associated with red hair have also been linked to increased melanoma risk. Additional phenotypic associations have been found. In different studies, MC1R variants were found to be associated with lentigo maligna melanoma (OR, 2.16; 95% CI, 1.07–4.37; P = .044)  and increased risk of melanoma for individuals with no red hair, no freckles, and Fitzpatrick type III or IV skin (summary OR, 3.14; 95% CI, 2.06–4.80). Pooled studies of 5,160 cases and 12,119 controls from 17 sites calculated that melanoma risk attributable to MC1R variants is 28%, suggesting that these variants may be an important contributor to melanoma risk in the general population.
A scoring system for MC1R polymorphisms has been proposed to identify associations between the degree of functional impairment of the melanogenesis pathway and the clinical characteristics of the patients and their melanoma presentation. The initial classification system designated MC1R variants that were strongly associated with red hair and fair skin as strong (R) red hair variants with an OR of 63.3 (95% CI, 31.9–139.6), whereas those with weaker association were designated weak (r) variants and had an OR of 5.1 (95% CI, 2.5–11.3). This work was expanded to evaluate additional MC1R variants and to add a summary score between zero and four, with a consensus sequence allele valued at zero, an r allele valued at one, and an R allele valued at two. In a study of 1,044 melanoma patients, those with a score of three or more were more likely to develop melanoma before age 50 years (OR, 1.47; 95% CI, 1.01–2.14). The MC1R score has been subsequently found to have implications for a survival benefit in melanoma patients. There was a lower risk of death in melanoma patients with no consensus MC1R alleles (HR, 0.78; 95% CI, 0.65–0.94) when compared with those with at least one consensus allele. An independent study found a similar survival benefit in individuals carrying two MC1R variants (HR, 0.60; 95% CI, 0.40–0.90).
A meta-analysis showed that the more MC1R variants an individual carried, the higher the risk of SCC and BCC. Individuals with two or more MC1R variants had a summary OR of 2.48 (95% CI, 1.96–3.15) for BCC and a summary OR of 2.80 (95% CI, 1.71–4.57) for SCC; these risks appeared to be stronger in individuals with red hair. Data from a study of individuals diagnosed with BCC before age 40 years also found a stronger association between BCC and MC1R variants in those with phenotypic characteristics not traditionally considered high risk.
Although variants in this gene are associated with increased risk of all three types of skin cancer, adding MC1R information to predictions based on age, sex, and cutaneous melanin density offers only a small improvement to risk prediction.[181,182] However, one study that examined predictors of early-onset melanoma in both population- and family-based studies showed that the addition of MC1R genotypes improved the area under the receiver operator curve (AUC) by 6% over demographic information alone (P < .001). When genotypes were combined with nevi and history of keratinocyte carcinoma (BCC or SCC), the AUC was 0.78 (95% CI, 0.75–0.82) for self-reported nevi and 0.83 (95% CI, 0.80–0.86) for physician-described nevi.
A study of 1,791 individuals assessed whether there are sex-specific effects of MC1R variants on melanoma risk. Carrying two or more MC1R variants was associated with an increased risk in both women (OR, 2.65; 95% CI, 1.86–3.79; P = .001) and men (OR, 1.65; 95% CI, 1.14–2.38; P = .007). In a multivariate analysis including other risk factors such as freckling, wrinkling, sunburns, and solar lentigines, the association between melanoma risk with MC1R variants in women remained significant, whereas the association did not remain significant in men.
MC1R variants can also modify melanoma risk in individuals with CDKN2A pathogenic variants. A study consisting of 815 carriers of CDKN2A pathogenic variants looked at four common non-synonymous MC1R variants and found that having one variant increased the melanoma risk twofold, but having two or more variants increased melanoma risk nearly sixfold. After stratification for hair color, the increased risk of melanoma appeared to be limited to subjects with brown or black hair. These data suggest that MC1R variants increase melanoma risk in a manner independent of their effect on pigmentation. A meta-analysis of individuals with CDKN2A pathogenic variants showed that those with greater than one variant in MC1R had approximately fourfold increased risk of melanoma. Individuals with one or more variants in MC1R showed an average 10-year decrease in age of onset from 47 to 37 years. In contrast, a large consortium study did not show as large a decrease in age at onset of melanoma. Another study of Norwegian melanoma cases and controls showed that carriers of CDKN2A pathogenic variants had an increased risk of melanoma when they carried either the Arg160Trp or Asp84Glu MC1R variants.
In addition to studies evaluating the relationship between germline variants and MC1R variants, multiple groups have assessed whether MC1R variants are associated with somatic BRAF mutations. Studies indicate that there may be an association between MC1R variants and BRAF V600E somatic mutations.[188,189,190,191] However, it is difficult to determine the impact of pigmentary influences on BRAF somatic mutations versus genetic effects.
Other pigmentary genes
Pathogenic variants in albinism genes may also account for a small proportion of familial melanoma. For example, variants in TYRP1, TYR, and OCA2 were observed at an increased frequency in one study of individuals with familial cutaneous melanoma compared with population controls. Further studies are needed to confirm these findings. (Refer to the Oculocutaneous albinism section in the Squamous Cell Carcinoma section of this summary for a discussion of the pigmentary genes OCA2, SLC45A2, TYR, and TYRP1, which have also been associated with melanoma.)
Whole-genome sequencing led to the identification of an E318K variant in the microphthalmia–associated transcription factor (MITF) gene in a family with seven cases of melanoma. MITF is a transcription factor that has been shown to regulate multiple genes important in melanocyte function and the E318K variant impairs the normal SUMOylation of MITF. The E318K variant was found in three of seven melanoma cases tested in this family and was present at a higher frequency in melanoma cases than controls. Six additional families among 182 families negative for CDKN2A and CDK4 pathogenic variants were found to carry the variant. Another study found six individuals with the E318K variant in a cohort of 168 individuals with melanoma (frequency, 0.018); no unaffected controls carried the variant. Individuals with the E318K variant were more likely to be fair skinned, with high nevus counts and high freckling scores, and all had MPM. There was also a high frequency of amelanotic melanomas. Another study showed that the E318K variant was associated with melanoma (OR, 1.7; 95% CI, 1.1–2.7) but that it had a stronger effect in individuals with dark hair (OR, 3.8; 95% CI, 1.5–9.6). Population-based studies in Australia and the United Kingdom consisting of 3,920 cases and 4,036 controls show a twofold increased risk of melanoma in carriers. Spanish and Italian case-control studies found ORs of approximately 3.0 for melanoma in carriers of the E318K variant.[196,197] However, the Spanish study also included melanoma cases from families with and without CDKN2A pathogenic variants. The prevalence of the MITF E318K variant was similar in families with and without CDKN2A pathogenic variants (2.9% and 1.9%, respectively). The MITF E318K variant may be associated with nodular melanoma, as 5 of 12 carriers (42%) in the Italian study had nodular melanoma compared with only 90 of 655 (14%) melanoma patients not carrying the variant. These data suggest that the E318K variant may be a moderate-risk allele for melanoma. However, these data remain controversial. Another study in a Polish population of 4,266 cancer patients and 2,114 controls found no association with melanoma.
The Breast Cancer Linkage Consortium found that pathogenic variants in BRCA2 were associated with a RR of melanoma of 2.58 (95% CI, 1.3–5.2). A second study reported a similar increase in risk, although the result fell short of statistical significance. In contrast, another large cohort study of carriers of BRCA2 pathogenic variants in the Netherlands showed a decreased risk of melanoma; however, the expected incidence of melanoma was rare in this population, and this result reflects a difference of only two melanoma cases. Ashkenazi Jewish melanoma patients have not been shown to have an increased prevalence of the three founder pathogenic variants in BRCA1 and BRCA2 that are commonly found in this population. Overall, the evidence for increased risk of melanoma in the BRCA2 population is inconsistent at this time.
(Refer to the BRCA1 and BRCA2 section in the PDQ summary on Genetics of Breast and Gynecologic Cancers for more information.)
Melanoma Risk Assessment
Patients with a personal history of melanoma or dysplastic nevi should be asked to provide information regarding a family history of melanoma and other cancers to detect the presence of familial melanoma. Age at diagnosis in family members and pathologic confirmation, if available, should also be sought. The presence of MPM in the same individual may also provide a clue to an underlying genetic susceptibility. Approximately 30% of affected individuals in hereditary melanoma kindreds have more than one primary melanoma, versus 4% of sporadic melanoma patients. Family histories should be updated regularly; an annual review is often recommended.
For individuals without a personal history of melanoma, several models have been suggested for prediction of melanoma risk. The models differ in performance with respect to sensitivity and specificity, including differences by sex in some models. Data from the Nurses' Health Study were used to create a model that included gender, age, family history of melanoma, number of severe sunburns, number of moles larger than 3 mm on the limbs, and hair color. The concordance statistic for this model was 0.62 (95% CI, 0.58–0.65). Another measure of baseline melanoma risk was derived from a case-control study of individuals with and without melanoma in the Philadelphia and San Francisco areas. This model focused on gender, history of blistering sunburn, color of the complexion, number and size of moles, presence of freckling, presence of solar damage to the skin, absence of a tan, age, and geographic area of the United States. Attributable risk with this model was 86% for men and 89% for women. This predictive tool, the Melanoma Risk Assessment Tool, is available online. However, this tool was developed using a cohort of primarily white individuals without a personal or family history of melanoma or NMSC. It is designed for use by health professionals, and patients are encouraged to discuss results with their physicians. Additional external validation is appropriate before this tool can be adopted for widespread clinical use. Professional organizations have published genetic counseling referral guidelines for individuals with a history of melanoma. (Refer to the Family history section of this summary for more information.)
Two models have been developed to predict the probability of identifying germline CDKN2A pathogenic variants in individuals or families for research purposes (Table 8). MelPREDICT  uses logistic regression and MelaPRO  uses a Mendelian modeling algorithm to estimate the chance of an individual carrying a pathogenic variant in CDKN2A.
Clinical testing is available to identify germline pathogenic variants in CDKN2A. Multiple centers in the United States and overseas offer sequence analysis of the entire coding region, and a number of centers perform deletion and duplication analysis. For information on genetic testing laboratories, refer to GeneTests: Laboratory Directory.
Expert opinion regarding testing for germline pathogenic variants of CDKN2A follows two divergent schools of thought. Arguments for genetic testing include the value of identifying a cause of disease for the individual tested, the possibility of improved compliance with prevention protocols in individuals with an identified pathogenic variant, and the reassurance of a negative testing result in individuals in a family carrying a pathogenic variant. However, a negative test result in a family that does not have a known pathogenic variant is uninformative; the genetic cause of disease in these patients must still be identified. It should also be noted that members of families carrying a CDKN2A pathogenic variant who do not carry the variant themselves may remain at increased risk of melanoma. At this time, identification of a CDKN2A pathogenic variant does not affect the clinical management of the affected patient or family members. Close dermatologic follow-up of these people is indicated, regardless of genetic testing result, and pancreatic cancer screening has unclear utility, as discussed below.
If genetic testing is undertaken in this population, experts suggest that it be performed after complete genetic counseling by a qualified genetics professional who is knowledgeable about the condition.
Refer to the Psychosocial Issues in Familial Melanoma section of this summary for information about psychosocial issues related to genetic testing for melanoma risk.
Management of members of melanoma-prone families
High-risk individuals, including first- and second-degree family members in melanoma-prone families, should be educated about sun safety and warning signs of melanoma. Regular examination of the skin by a health care provider experienced in the evaluation of pigmented lesions is also recommended. One guideline suggests initiation of examination at age 10 years and conducting exams on a semiannual basis until nevi are considered stable, followed by annual examinations. These individuals should also be taught skin self-examination techniques, to be performed on a monthly basis. Observation of lesions may be aided by techniques such as full-body photography and dermoscopy.[212,213] A cost-utility analysis has demonstrated the benefits of screening in the high-risk population.
Biopsies of skin lesions in the high-risk population should be performed using the same criteria as those used for lesions in the general population. Prophylactic removal of nevi without clinically worrisome characteristics is not recommended. The reasons for this are practical: many individuals in these families have a large number of nevi, and complete removal of them all is not feasible, since new atypical nevi continue to develop. In addition, individuals with increased susceptibility to melanoma may have cancer arise de novo, without a precursor lesion such as a nevus.
Standard recommendations for screening and management of patients with BAP1 germline pathogenic variants are not currently available, but one group of experts has recommended annual ocular examinations starting at age 16 years, full-body skin examinations starting at age 20 years, and consideration of annual renal ultrasound and/or abdominal magnetic resonance imaging every 2 years.
At present, chemoprevention of melanoma in high-risk individuals remains an area of active investigation; however, no medications are recommended for melanoma risk reduction at this time.
Pancreatic cancer screening inCDKN2Apathogenic variant carriers
Screening for pancreatic cancer remains an area of investigation and controversy for carriers of CDKN2A pathogenic variants. At present, no effective means of pancreatic cancer screening is available for the general population; however, serum and radiographic screening measures are under study in high-risk populations. One proposed protocol  suggested starting pancreatic screening in high-risk families at age 50 years or 10 years before the youngest age at diagnosis of pancreatic cancer in the family, whichever came first. In this algorithm, asymptomatic patients would be screened annually with serum cancer antigen 19-9 and endoscopic ultrasound, whereas symptomatic patients or those with abnormal test results would undergo endoscopic retrograde cholangiopancreatography (ERCP) and/or spiral computed tomography (CT) scanning. A study evaluating use of endoscopic ultrasound and ERCP in high-risk families concluded that these procedures were cost-effective in this setting.
The disadvantages of screening include the limitations of available noninvasive testing methods and the risks associated with invasive screening procedures. ERCP is the gold standard for identifying early cancers and precancerous lesions in the pancreas. However, serious complications such as bleeding, pancreatitis, and intestinal perforation can occur with this procedure. Implementation of pancreatic screening in carriers of CDKN2A pathogenic variants is further complicated by the apparent lack of increased incidence of pancreatic cancer in many of these families.
Most experts suggest that pancreatic cancer screening should be considered for carriers of CDKN2A pathogenic variants only if there is a family history of pancreatic cancer and, even then, only in the context of a clinical trial.
Screening for melanoma is not recommended by the U.S. Preventive Services Task Force (USPSTF), although the American Cancer Society, the Skin Cancer Foundation, and the American Academy of Dermatology recommend monthly skin self-examination and regular examination by a physician for people older than 50 years or those with multiple melanomas or dysplastic nevus syndrome. USPSTF does not recommend screening because they judge that the evidence for efficacy is not strong. On the other hand, the groups who recommend screening base their support on the logic that screening will find melanomas early in their development and that those melanomas will not progress further. This position is supported by the unusually detailed prognostic information that can be obtained through histopathology examination of primary melanoma tumors, in which a variety of features (lack of invasion through the basement membrane, thin cancers [≤ 0.76 mm], absence of vertical growth phase disease, ulceration, and histologic regression) have been solidly linked to favorable prognosis.
The question of whether the lesions found through screening are programmed to progress or whether they will grow very slowly and never progress to metastatic disease has not been answered. One study showed that skin self-examination might prevent the formation of melanomas and that skin self-examination was associated with reduced 5-year mortality. The primary preventive effect could be biased by the fact that healthy individuals who participate in studies are somewhat more likely to participate in screening activities. The 63% reduction in mortality observed in that study was not statistically significant. Therefore, until a randomized trial of screening and mortality is undertaken, the utility of general population screening remains uncertain.
Nonetheless, it is well documented that, when a patient is under the care of a dermatologist, his or her second melanoma is diagnosed at a thinner Breslow depth than the index melanoma.[221,222,223] As survival is inversely correlated with Breslow depth for melanoma, early diagnosis leads to better prognosis.
Primary prevention for melanoma consists of avoiding intense intermittent exposure to UV radiation, both solar and nonsolar. It should be stressed that the dose-response levels for such exposure are not defined, but that large, sporadic doses of UV radiation on skin are those epidemiologically most associated with later development of melanoma. Sunburn is a marker of that exposure, so that the amount of time spent in the sun should be calculated to avoid sunburn if at all possible. Tanning beds should be avoided, as studies suggest that they increase the risk of melanoma.[225,226] In an attempt to prevent skin cancer, more than 40 states have laws prohibiting tanning bed use by teenagers or requiring signed parental consent for teenagers who request tanning bed use.
Primary prevention should stress the need for caution in the sun and protection in the form of clothing, shade, and sunscreens when long periods of time are spent outdoors or at times of day when sunburn is likely. High-risk patients should understand that the application of sunscreens should not be used to prolong the time they spend in the sun because UV radiation makes its way through the sunscreen over time.[228,229] However, regular sunscreen use has been shown to reduce melanoma incidence in a prospective, randomized controlled trial.
As described in the PDQ summary on Melanoma Treatment, therapeutic options range widely from local excision in early melanoma to chemotherapy, radiation, and aggressive management in metastatic melanoma. The best defense against melanoma as a whole is to encourage sun-protective behaviors, regular skin examinations, and patient skin self-awareness in an effort to decrease high-risk behaviors and optimize early detection of potentially malignant lesions.
Brooke-Spiegler Syndrome, Multiple Familial Trichoepithelioma, and Familial Cylindromatosis
Brooke-Spiegler Syndrome (BSS), familial cylindromatosis, and multiple familial trichoepithelioma (MFT) are all autosomal dominant syndromes with overlapping clinical characteristics with allelic variance. Features of BSS include multiple skin appendage tumors such as cylindromas (tumors arising in the hair follicle stem cells), trichoepitheliomas (tumors arising in the hair follicle), and spiradenomas (benign tumors arising in the sweat gland). MFT is characterized by nonmalignant skin tumors, primarily trichoepitheliomas, and familial cylindromatosis manifests predominantly as cutaneous cylindromas. Onset of tumors for these syndromes is typically in late childhood or early adolescence, suggesting a hormonal influence. There is some evidence of greater severity in females than in males. UV radiation appears to be a major initiating factor for cylindromas. Typical tumor sites for cylindromas in familial cylindromatosis are the scalp (81% of carriers), the trunk (69% of carriers), and the pubic area (42% of carriers). Other tumors that can be associated with these syndromes include parotid gland tumors, basal cell adenomas, and basal cell carcinomas. Refer to Table 3, Basal Cell Carcinoma (BCC) Syndromes, for more information about BSS.
Because pathogenic variants in CYLD on16q12-q13 have been identified in individuals with each of these disorders, these syndromes are thought to represent different phenotypic manifestations of the same disease. Penetrance of pathogenic variants in CYLD is reported to be 60% to 100%.[3,5] In one study, 85% of the BSS families, 100% of familial cylindromatosis families, and only 44% of MFT families were found to have pathogenic variants in CYLD. A second locus for MFT maps to 9p21, but the gene for this locus remains unknown.
Given the potential for progressive enlargement, the preferred approach for cylindromas is ablation while the tumors are small and easily managed. Electrosurgery or Mohs micrographic surgery may be utilized for therapy, although excision of large lesions may require skin grafting for closure. Trichoepitheliomas and spiradenomas typically remain smaller in size; thus, after the diagnosis is confirmed by skin biopsy, unless there is impingement on critical structures, further intervention is not required. If therapy is deemed necessary and appropriate, either electrosurgery or ablative laser therapy is a valid option. Radiotherapy is not recommended for treatment of any of these tumors because a potential for increased tumor induction.
Level of evidence: 4
Cutaneous sebaceous neoplasms may be associated with Muir-Torre syndrome (MTS). Multiple types of sebaceous tumors including sebaceous adenomas, epitheliomas, carcinomas, and keratoacanthomas or BCCs with sebaceous differentiation have been described. A variant of Lynch syndrome /hereditary non-polyposis colorectal cancer syndrome, the MTS phenotype involves the synchronous or metachronous development of at least one cutaneous sebaceous neoplasm and at least one visceral malignancy. The visceral malignancies may be of gastrointestinal (colorectal, stomach, small bowel, liver, and bile duct) and/or genitourinary (endometrial and bladder) origin and typically demonstrate a less aggressive phenotype than non-MTS equivalent tumors.[9,10] MTS, inherited in an autosomal dominant fashion with high penetrance and variable expressivity, is associated with pathogenic variants in the mismatch repair genes MLH1, MSH2, and less commonly, MSH6.[11,12,13,14,15,16] In a study of 36 sebaceous lesions that included sebaceous carcinomas, sebaceous adenomas, and sebaceomas, 38.9% of lesions were missing one or more mismatch repair proteins by immunohistochemistry (IHC). Of the ten individuals with absent staining of one or more proteins, five had gene testing that confirmed a diagnosis of Lynch syndrome. This result suggests that routine screening of sebaceous lesions by IHC may be useful in identification of individuals with Lynch syndrome. One study of the incidence of skin cancer in Lynch syndrome suggests there is an increase in sebaceous carcinoma and squamous cell carcinoma in these patients.
While the commonly noted sebaceous hyperplasia has not been associated with MTS, any sebaceous lesion with atypical or difficult to classify histologic features should prompt further exploration of the patient's family and personal medical history. Consideration should be given to referring patients with sebaceous neoplasms to medical geneticists or gastroenterologists to evaluate further for Lynch syndrome. While the diagnosis of visceral malignancy precedes that of cutaneous sebaceous neoplasms in the majority of patients, 22% of patients develop cutaneous sebaceous neoplasms first, offering an opportunity for visceral malignancy screening. Current diagnosis of MTS is based upon clinical criteria but may be supported by immunohistochemical staining for MSH2, MLH1, and MSH6 as a screening mechanism before molecular genetic analysis.[12,14,15,16,20] Genetic counseling and testing for the patient and family members, with appropriate visceral malignancy screening regimens, should be pursued.
Level of evidence: 3
Hereditary Leiomyomatosis and Renal Cell Carcinoma (HLRCC)
Although cutaneous smooth muscle tumors (leiomyomas) are not themselves a form of skin cancer, multiple cutaneous leiomyomas are associated with renal cell cancer (RCC) in an inherited syndrome known as hereditary leiomyomatosis and renal cell cancer (HLRCC). Cutaneous leiomyomas present as firm, pink or reddish-brown papules and nodules distributed over the trunk and extremities and, occasionally, on the face. These lesions occur at a mean age of 25 years (age range, 10–47 y) and tend to increase in size and number with age. Lesions are sensitive to light touch and/or cold temperature and are, less commonly, painful. Pain is correlated with severity of cutaneous involvement. The presence of multiple cutaneous leiomyomas is associated with HLRCC until proven otherwise and should prompt a genetic workup; a solitary leiomyoma requires careful analysis of family history. (Refer to the HLRCC section in the PDQ summary on Genetics of Kidney Cancer (RCC) for more information.)
This section reviews the literature examining risk reduction and early-detection behaviors in individuals with heightened risk of melanoma resulting from their family history of the disease and in individuals from hereditary families who have been tested for melanoma high-risk pathogenic variant status. The review also addresses risk perception and communication in individuals at heightened risk of melanoma.
Interest in and Uptake of Genetic Testing for Risk of Melanoma
Currently, clinical testing for CDKN2A is not recommended outside the research context because most individuals from multiple-case families will not be identified as having a pathogenic variant in this gene, and because recommendations for those testing positive do not differ for multiple-case family members who test negative, or do not pursue testing.[1,2] Despite these cautions, CDKN2A testing is commercially available, and thus demand for the test will likely increase. Arguments for the availability of genetic testing include that the results of testing could provide psychological security and contribute to enhanced screening and prevention efforts for those testing positive for CDKN2A. (Refer to the Melanoma Risk Assessment section of this summary for more information about clinical genetic testing for melanoma susceptibility.)
Few studies have examined motivation and interest in genetic testing for melanoma risk. In summary, the findings include the following:
In Australia, a qualitative study (N = 40) found that almost all participants with a strong family history of melanoma were interested in genetic testing.[6,9] Genetic testing was favored by the participants as a means to gain information about their children's susceptibility to melanoma, to increase their understanding of their own risk, to advance melanoma research, and to provide increased motivation for sun-protective behavior.
A Dutch study examined interest in CDKN2A testing (p16-Leiden pathogenic variant). Of 510 letters sent to members of 18 p16-Leiden-positive families recruited from the Pigmented Lesions Clinic at the Leiden University Medical Center in the Netherlands, 488 individuals responded by attending clinic for physical examination; an additional 15 family members also accompanied these individuals. Of these, 403 individuals were eligible for genetic counseling. A total of 184 family members followed through with counseling, and 141 of them opted for genetic testing. After the counseling session, 94 individuals returned a completed questionnaire. Older age predicted higher interest in genetic testing; reasons for having genetic testing included learning personal risk (57%) and learning the risk of one's child carrying the pathogenic variant (69%). Most participants (88%) felt that genetic testing would make a contribution to diagnostics within their family. However, some individuals (40%) reported that they had not expected to receive risk information concerning pancreatic cancer, and half of the participants (49%) reported increased worry about the possibility of developing pancreatic cancer. Finally, in an Arizona qualitative study of 22 individuals with a strong family history of melanoma, none elected genetic testing even though it was provided as an option for them.
In an Australian study of 121 individuals with a strong family history of melanoma, participants completed questionnaires before genetic counseling and testing. Distress (melanoma-specific distress and general distress) levels were very low in this population. The most important predictors of distress included a personal history of melanoma, having concerns about the impact of melanoma on family, having a high information-seeking disposition (monitoring style), a perceived importance of sun exposure in causing melanoma, and not having children.
Testing in children
Among 61 people tested for CDKN2A pathogenic variants (52.5% tested positive) from two large melanoma kindreds, most (75.4%) had children or grandchildren younger than 18 years and expressed interest in testing of minors (73.8%). Among carriers of CDKN2A pathogenic variants, most (86.7%) wanted their children or grandchildren to be tested, and among noncarriers, half (50%) wanted testing for their own children or grandchildren. The most cited reason for testing children was to aid in risk awareness and to improve protection and screening behavior.
Risk Awareness and Risk Reduction in Individuals at Increased Familial Risk of Melanoma
A number of studies have been conducted examining risk reduction via adoption of sun protection (including the use of sunscreen and protective clothing and shade seeking behavior) in individuals with a family history of melanoma. Overall, these studies indicate inconsistent adoption and maintenance of these behaviors. Most of these studies have been conducted with clinic-based populations that might be more prone to risk reduction and screening behaviors than those with a similar risk profile in the general population.
In terms of sun protection, in a Swedish population, 87 young adults with dysplastic nevi were surveyed, and 70% estimated their melanoma risk to be equal or lower than that of the Swedish population in general, and one third reported frequent sunbathing behavior. Another study examined 229 first-degree relatives (FDRs) referred by melanoma patients attending clinic appointments; those who were older, female, and had greater confidence in their ability to practice sun-protection were most likely to do so, but the utilization of sun-protective behavior was inconsistent. Another study in the United States examined sun-protective behavior in 100 FDRs of melanoma clinic patients and found that less than one-third of patients use sunscreen routinely when in the sun and that more regular usage was related to higher education levels, higher self-efficacy for sun protection, and higher perceived melanoma risk. Perceived severity of melanoma and response-efficacy were not related to adoption of sun-protective behaviors.
A study that focused on 68 minor children (aged 17 years or younger) of melanoma survivors demonstrated that while overall rates of sun-protective behavior were high (near 80%), the rates of sunburn were also high (49%). The authors concluded that multiple methods of sun-protective behavior are warranted in these individuals. However, in the teenage years, there were significant reductions in sun protection indicating an even greater need for intervention in this group.
Another study based in the United Kingdom examined sunburn rates in 170 individuals with a family history of melanoma compared with 140 controls matched to age, sex, and geographical location. Of those with a melanoma family history, 31% reported sunburn in the previous summer (compared with 41% of controls); melanoma families reported better sun-protection behaviors than controls overall. Across controls and those with a family history of melanoma, younger males were more likely to report recent sunburns; also, across controls and those with a family history of melanoma, those relatives with atypical mole syndrome and a belief in their ability to prevent melanoma showed better sun protection.
One qualitative study of 20 FDRs of melanoma patients recruited from a high-risk clinic at the University of Arizona identified perceived unmet needs for physician communication of risk status, including greater consistency in communication, education for patients concerning the importance of family history to risk status, and needs and desire for more complex advice (e.g., reapplication of sunscreen and wearing clothing with ultraviolet protection factor).
A prospective study examined interest in and 3-month behavioral and psychosocial outcomes associated with disclosure of melanoma high-risk pathogenic variant research results in 19 individuals (three CDKN2A carriers). All of the variant carriers, but only four of the noncarriers, had a family history of melanoma. Carrier status did not affect risk perception, distress, or sun-protection behaviors.
A few intervention studies have targeted knowledge about melanoma, sun protection, and screening in family members of melanoma patients. In one study among siblings, participants drawn from a clinic population were randomly assigned to an intervention that included telephone messages and tailored print materials about risk reduction and screening recommendations. The usual care group received a standard physician-practice recommendation that patients notify family members about their diagnosis. The intervention group showed improvements in knowledge about melanoma, confidence in seeing a dermatologist and having a screening examination, and greater improvements in skin self-examination practices compared with control participants after 12 months; both groups showed twofold increases in physician examinations after 12 months; there was no change in sunscreen behaviors in either group.
In another study, 443 family members of melanoma patients were randomly assigned to either a generic or tailored intervention that consisted of three (untailored or tailored) print mailings and one (untailored or tailored) telephone counseling session. Overall, the tailored intervention group showed an almost twofold increase in frequency of total cutaneous skin examinations by a health care provider compared with the generic intervention. However, no differences were observed for skin self-examinations between intervention arms. In contrast to the previous study, which did not show improvements in sun protection habits, participants in this study who received the tailored intervention were significantly more likely to report improvements in sun protection habits than were those who received the generic intervention.
Screening Behaviors in Individuals at Increased Familial Risk of Melanoma
A number of studies have examined early-detection behaviors in individuals at increased risk of melanoma. In a U.S. sample of 404 siblings drawn from a clinic population of melanoma patients, only 42% of individuals had ever seen a dermatologist; 62% had engaged in skin self-examination; 27% had received a physician skin examination; and only 54% routinely used sunscreen. Female gender was related to greater sunscreen use; those older than age 50 years were more likely to have received a physician skin examination. Having a dermatologist was strongly related to all three outcomes (skin self-examination, physician examination, and sunscreen use). In a U.S. study of 229 FDRs referred by patients attending clinic, about half (55%) reported ever having a total cutaneous examination, and slightly more (71%) reported ever performing skin self-examination. Common predictors of skin examination (physician and self-examinations) included physician recommendation and low perceived barriers of screening. Interestingly, 14% of the sample had not told their primary care doctor about their sibling's melanoma diagnosis. One U.S. study showed that half (53%) of FDRs had never received a total cutaneous screening by a physician; only 27% had received a physician recommendation to have a screening. Early detection adherence was related to the following: higher education level, more melanoma risk factors, health care provider recommendation for screening, perceived risk of melanoma, and perceived severity of melanoma. Parents of melanoma patients were less likely to have pursued screening than siblings and children. A U.S. study examined intentions to receive a physician skin examination and to perform skin self-examination among FDRs of individuals diagnosed with melanoma who had not recently engaged in skin surveillance. Predictors of intentions included both benefits and barriers to screening and family support for screening, but not knowledge of recommended screening frequency.
A cross-sectional Australian study of 120 individuals from families with a known CDKN2A pathogenic variant found that in the past 12 months, 50% reported engaging in skin self-examinations at least four times, and 43% had undergone at least one clinical skin examination. In contrast, 15% had not performed a skin self-examination in the past 12 months, and 27% had never had a clinical skin examination. Correlates of skin cancer screening behaviors included having a history of melanoma, a physician's recommendation, and stronger behavioral intentions. Additional correlates for skin self-examination included self-efficacy, perceived efficacy of melanoma treatment, and melanoma-specific distress. Perceived risk of developing melanoma was not significantly associated with skin cancer screening behaviors.
Psychosocial Outcomes of Genetic Counseling and Genetic Testing
A few small studies have examined distress and behavioral factors associated with CDKN2A testing for melanoma. In a Swedish clinic for individuals at high risk of melanoma resulting from dysplastic nevus syndrome, 11 unaffected, untested individuals drawn from families in which a CDKN2A pathogenic variant has been identified were examined. Most (9 of 11) reported no worry about increased melanoma risk. In assessments after disclosure of results, there were no increasing trends towards depression, anxiety, or increased melanoma-risk perception by test results, and no systematic change in sun-related habits by test results.
A prospective study examined interest in and 3-month behavioral and psychosocial outcomes associated with disclosure of melanoma high-risk pathogenic variant research results in 19 individuals (three CDKN2A carriers). All of the pathogenic variant carriers, but only four of the noncarriers, had a family history of melanoma. Carrier status did not affect risk perception, distress, or sun-protection behaviors.
In a randomized controlled trial, 73 adults with a family history of melanoma were randomly assigned to receive either genetic counseling with genotyping results (CDKN2A and MC1R) or usual care. Overall, participants in the intervention group reported a significant increase in frequency of skin self-examinations, compared with a slight decrease among those in the control group. In addition, intervention participants reported a smaller decrease in frequency of wearing a shirt for sun protection compared with control participants. No other differences in sun protection habits were noted. These results should be interpreted with caution, as only five individuals (three in the intervention arm) had a pathogenic variant for one or both of the genes. Nonetheless, study results support the notion that genetic testing for melanoma does not lead to false reassurance and reduced sun protection behaviors among those who test negative.
Another study examined behavioral factors associated with CDKN2A carrier status among 64 individuals from two large Utah families in which a CDKN2A pathogenic variant had been identified. The individuals received extensive recommendations for sun protection and screening. Questionnaires conducted one month after receipt of genetic test results and recommendations showed increased intention for skin examinations (self-examinations and health care professional examinations), regardless of whether individuals were found to be CDKN2A carriers or noncarriers. Rates of over screening (>1 skin self-examination per month) also increased in CDKN2A carriers. In a follow-up study one month later with the same sample, CDKN2A carriers showed marginally increased intentions for sun-protective behaviors; CDKN2A noncarriers showed no increase in overall photoprotection but a shift to using sun-protective clothing rather than sun avoidance. Thirty-seven individuals from the same cohort were assessed for psychosocial and behavioral outcomes 2 years posttesting. Levels of anxiety, depression, melanoma worry, and pancreatic cancer worry were all low and decreased over time, with more perceived benefits of testing noted than drawbacks of testing. Adherence to annual total-body skin examinations significantly increased among unaffected carriers (from 40% at baseline to 70% at 2 years) but decreased among unaffected noncarriers (from 56% at baseline to 13% at 2 years). Affected carriers were adherent at both assessments (91% and 82%, respectively).
The PDQ cancer information summaries are reviewed regularly and updated as new information becomes available. This section describes the latest changes made to this summary as of the date above.
Revised text to state that the bottom layer of the skin is formed of basal keratinocytes abutting the basement membrane, along with interspersed melanocytes. Also revised text to state that as basal keratinocytes migrate toward the skin surface, they progressively differentiate, lose their nuclei and form the spinous cell layer; the granular cell layer; and the keratinized outer layer, or stratum corneum, which serves as a protective covering of the body.
Added text to state that melanocytes contain melanin, which is packaged into melanosomes and transported to nearby keratinocytes to induce pigmentation of the skin. Melanin provides a barrier for the nuclei of keratinocytes against ultraviolet radiation and also plays a role in the immune system (cited Lin et al. as reference 5).
Revised text to state that the dermis is largely composed of an extracellular matrix. Prominent cell types and organelles in this compartment are fibroblasts, endothelial cells, smooth muscle cells, transient immune system cells, blood vessels, and nerves. When malignant transformation occurs, fibroblasts form fibrosarcomas and endothelial cells form angiosarcomas, Kaposi sarcoma, or other vascular tumors.
Revised text to state that melanomas are characterized by dark pigment with asymmetry, border irregularity, color variation, a diameter of more than 6 mm, and evolution.
Basal Cell Carcinoma (BCC)
Added text to state that a family history of melanoma in the Nurses' Health Study, the Nurses' Health Study 2, and the Health Professionals Follow-Up Study showed an increased risk of BCC (cited Wei et al. as reference 4).
Added DNA repair genes as a new subsection.
Squamous Cell Carcinoma (SCC)
Added text to state that a study of 216,115 individuals from the Nurses' Health Study, the Nurses' Health Study 2, and the Health Professionals Follow-Up Study found that individuals with a family history of melanoma showed a 22% increased risk of SCC (cited Wei et al. as reference 39).
Added text to state that a cohort study found that 18 of 283 individuals with recessive dystrophic epidermolysis bullosa had a diagnosis of SCC with a median age of diagnosis of 22.6 years (cited Feinstein et al. as reference 132).
Added text to state that a large study of more than 200,000 individuals from the Nurses' Health Study, the Nurses' Health Study 2, and the Health Professionals Follow-Up Study found that individuals with a family history of melanoma had an increased risk of melanoma (cited Wei et al. as reference 22).
Added text to state that a study of 1,791 individuals assessed whether there are sex-specific effects of MC1R variants on melanoma risk. Carrying two or more MC1R variants was associated with an increased risk in both women and men. In a multivariate analysis including other risk factors such as freckling, wrinkling, sunburns, and solar lentigines, the association between melanoma risk with MC1R variants in women remained significant, whereas the association did not remain significant in men (cited Wendt et al. as reference 184).
This summary is written and maintained by the PDQ Cancer Genetics Editorial Board, which is editorially independent of NCI. The summary reflects an independent review of the literature and does not represent a policy statement of NCI or NIH. More information about summary policies and the role of the PDQ Editorial Boards in maintaining the PDQ summaries can be found on the About This PDQ Summary and PDQ® - NCI's Comprehensive Cancer Database pages.
Purpose of This Summary
This PDQ cancer information summary for health professionals provides comprehensive, peer-reviewed, evidence-based information about the genetics of skin cancer. It is intended as a resource to inform and assist clinicians who care for cancer patients. It does not provide formal guidelines or recommendations for making health care decisions.
Reviewers and Updates
This summary is reviewed regularly and updated as necessary by the PDQ Cancer Genetics Editorial Board, which is editorially independent of the National Cancer Institute (NCI). The summary reflects an independent review of the literature and does not represent a policy statement of NCI or the National Institutes of Health (NIH).
Board members review recently published articles each month to determine whether an article should:
Changes to the summaries are made through a consensus process in which Board members evaluate the strength of the evidence in the published articles and determine how the article should be included in the summary.
The lead reviewers for Genetics of Skin Cancer are:
Any comments or questions about the summary content should be submitted to Cancer.gov through the NCI website's Email Us. Do not contact the individual Board Members with questions or comments about the summaries. Board members will not respond to individual inquiries.
Levels of Evidence
Some of the reference citations in this summary are accompanied by a level-of-evidence designation. These designations are intended to help readers assess the strength of the evidence supporting the use of specific interventions or approaches. The PDQ Cancer Genetics Editorial Board uses a formal evidence ranking system in developing its level-of-evidence designations.
Permission to Use This Summary
PDQ is a registered trademark. Although the content of PDQ documents can be used freely as text, it cannot be identified as an NCI PDQ cancer information summary unless it is presented in its entirety and is regularly updated. However, an author would be permitted to write a sentence such as "NCI's PDQ cancer information summary about breast cancer prevention states the risks succinctly: [include excerpt from the summary]."
The preferred citation for this PDQ summary is:
PDQ® Cancer Genetics Editorial Board. PDQ Genetics of Skin Cancer. Bethesda, MD: National Cancer Institute. Updated <MM/DD/YYYY>. Available at: https://www.cancer.gov/types/skin/hp/skin-genetics-pdq. Accessed <MM/DD/YYYY>. [PMID: 26389333]
Images in this summary are used with permission of the author(s), artist, and/or publisher for use within the PDQ summaries only. Permission to use images outside the context of PDQ information must be obtained from the owner(s) and cannot be granted by the National Cancer Institute. Information about using the illustrations in this summary, along with many other cancer-related images, is available in Visuals Online, a collection of over 2,000 scientific images.
The information in these summaries should not be used as a basis for insurance reimbursement determinations. More information on insurance coverage is available on Cancer.gov on the Managing Cancer Care page.
More information about contacting us or receiving help with the Cancer.gov website can be found on our Contact Us for Help page. Questions can also be submitted to Cancer.gov through the website's Email Us.
Last Revised: 2020-05-15
Healthwise, Healthwise for every health decision, and the Healthwise logo are trademarks of Healthwise, Incorporated.
2615 Lake Drive
Raleigh, NC 27607
901 Ridgefield Drive
Raleigh, NC 27609
4420 Lake Boone Trail
Raleigh, NC 27607
934 Vandora Springs Road
Garner, NC 27529
1505 SW Cary Parkway
Cary, NC 27511