Epidemiology and genetics: New results in melanoma etiology and their application to melanoma prevention.

CS13-01

Despite a decade of sun protection programs, melanoma incidence is rising, and thick tumors of 4 mm or greater with poor prognosis are still increasing in men aged 50 years or older. Novel strategies to curtail this disease are of crucial importance. Epidemiological studies strongly implicate sun exposure in the development of melanoma. Moreover, sun exposure is associated with nevi number, a strong melanoma risk factor. However, the association between sun exposure and melanoma is complex and appears to vary by melanoma subtypes. Short burning episodes of sun exposure are a major risk factor, but chronic sun exposure may also contribute, particularly in the case of lentigo maligna melanomas. Melanomas on skin with few signs of solar elastosis occur in younger individuals and have frequent mutations in the BRAF oncogene. BRAF is part of the Ras/Raf/MAPK/ERK pathway, critical for responding to ultraviolet radiation (UVR), regulating apoptosis, and controlling cell growth; KIT somatic mutations have been found in melanomas from sun protected areas of the body. By contrast, melanomas on skin with signs of solar elastosis affect older individuals, have different patterns of chromosomal aberrations, and have a lower frequency of BRAF mutations. Pigmentation strongly influences UVR-sensitivity and exposure to sunlight. There is also some evidence of increased melanoma survival in subjects exposed to UVR after melanoma diagnosis; vitamin D has been proposed as responsible for these apparent protective effects. Both total number of common nevi and atypical nevi are important and independent risk factors for melanoma. Based on the different prevalence of nevi between patients with head and neck melanomas and patients with melanomas of the trunk, some investigators have hypothesized that cutaneous melanomas may arise through two pathways, one associated with melanocyte proliferation and the other with chronic exposure to sunlight. Others, have postulated that the different prevalence may be due to an age effect (melanomas of head and neck tend to arise in older ages, when nevi tend to regress), or to due to aberrant melanogenesis and cumulative sun exposure among susceptible individuals. Genetic factors increase the complexity of these associations. Approximately 3-12 percent of melanomas develop in families with multiple cases of melanoma. Familial aggregation may occur because family members share environmental factors (such as excessive sun exposure) or common genetic factors (such as strongly penetrant melanoma susceptibility alleles). Worldwide studies of melanoma-prone families have demonstrated linkage to at least two chromosome loci and identified candidate genes accounting for approximately 25% of melanoma kindreds worldwide. The major melanoma susceptibility gene identified to date is CDKN2A, a tumor suppressor gene encoding two distinct proteins, p16 (or p16INK4a) and p14^ARF, translated in alternative reading frames from alternatively spliced transcripts. The second melanoma susceptibility gene, accounting for a very limited number of melanoma kindreds is CDK4. Both CDKN2A and CDK4 are involved in cell cycle regulation. One low penetrance gene that has been associated with melanoma risk is the melanocortin-1 receptor gene (MC1R). Its protein product binds to α-melanocyte-stimulating hormone (α-MSH) and adrenocorticotropin (ACTH), thereby regulating pigmentation in humans. MC1R is highly polymorphic, and morethan 80 human MC1Ralleles have been identified to date. Studies have identified several MC1R variants that are associated with red hair, low tanning ability, light skin color,freckling, and increased melanoma risk. MC1R can also act as a modifier of melanoma risk within CDKN2A-positive melanoma-prone families. Variant alleles of MC1R have been associated with BRAF mutations, particularly in skin with no sign of chronic solar damage. Different distributions of melanomas by anatomic site have been observed among Mediterranean, Australian, and American populations, which could determine differences in MC1R variants and related effects on the BRAF oncogene. A possible mechanism mediating this susceptibility is increased generation of reactive oxygen species (ROS) in MC1R variant carriers. Mutations in the BRAF oncogene are frequent in common melanocytic nevi, where they could trigger senescence, and only in the presence of inactivated tumor suppressor gene(s), they may lead to a malignant phenotype. The complexity of melanoma etiology underscores the importance of identifying biological or molecular pathways that might prove amenable to preventive intervention in man. Public awareness and education, self-examinations, sun avoidance, sunscreen use, physical examination (which should be performed by physicians during routine care), periodic lifetime surveillance of patients at high risk, dermoscopy, and biopsy of suspect lesions comprise the elements of the approach to melanoma detection. Although incidence rates are generally higher among lightly-pigmented than darkly-pigmented individuals, rates of melanoma have been increasing in individuals once-considered at lower risk for skin cancer due to darker pigmentation, such as Mediterranean populations and California Hispanics, challenging traditional methods of risk assessment based largely on skin color phenotype. Moreover, the identification of increased melanoma risk in subjects carrying MC1R variants, regardless of subjects' pigmentation, suggests that identifying those at risk for skin cancer solely by pigmentation characteristics might miss individuals with an elevated risk. Furthermore, the increasing risk of thick (poor prognosis) melanomas in older men emphasizes the need for a re-examination of melanoma prevention practices. Several controversies exist in relation to the benefits of melanoma screenings and sunscreen use. Subjects who use sunscreen are generally the same subjects who are at higher risk for melanoma, thus controlling for confounding factors is a challenge in their epidemiological evaluation. Moreover, case-controlstudies may be open to recall bias of sunscreen use by casesand controls, and sunscreen use may increase time spent in the sun which in turn may increase melanoma risk. Randomized controlled trials of the use of sunscreens toprevent the development of nevi as a surrogate end point for riskof developing melanoma indicate a protective effect, and thus indirectly suggest that preventive measures against melanoma shouldinclude the proper use of sunscreens in addition to avoiding midday sun and wearing protective clothing. The increasing evidence of an important role of oxidative damage in melanoma development through increased genomic instability and mutation rates, suggest that more pharmacologically active compounds could be useful to reduce UV damage. Although based on sparse data in animal models, sunscreens including nitric oxide and ROS inhibitors have shown to reduce sunburns, immunosuppression and photocarcinogenesis in treated mice. Given the ability of UVA rays to penetrate deeper into the skin and induce ROS, sunscreens that could more effectively filter UVA are being investigated. Sun exposure cannot and should not be totally avoided due to the beneficial health effects of Vitamin D₃ production (which has been shown to decrease cell proliferation and increase apoptosis) and induction of alpha-MSH. Different strategies are being discussed to provide balanced public health recommendations, including targeting different ages and at risk subjects with "ad hoc" messages, and estimating the appropriate ranges of UV exposure based on subjects' skin type. Studies on the optimal levels of Vitamin D, taking into account genetic variants of the Vitamin D receptor, are needed in light of the possibility of using Vitamin D supplementation as a prevention strategy against melanoma. Very recent investigations have proposed analogs of alpha-MSH that function as MC1R agonists to protect human melanocytes from UV-induced DNA damage and cytotoxicity. Moreover, a potentially important new strategy is suggested by the recent discovery in mice that pigmentation can be rescued by topical application of the cyclic AMP agonist forskolin, without the need for UVR and functional MC1R, with a consequent decrease in DNA damage and tumorigenesis. Although far from being applicable in humans, this work has yielded new clues about tanning that should eventually help in the fight against melanoma. Chemoprevention of melanoma have also been proposed for subjects at high risk for melanoma, in particular those belonging to melanoma-prone families, with mutations in CDKN2A or CDK4, or with multiple nevi with atypical features. Intraepithelial neoplasia and atypical nevi have been proposed as intermediate endpoints to monitor the effects of these treatments. Data are still sparse, and often based on small studies or studies not directly designed to test efficacy on melanoma. Among the most promising, the lipid lowering drugs statins have generated significant interest as anti-cancer agents. Statins target Ras signaling and promote apoptosis, and could potentially be useful for melanomas with activation of the Ras/Raf/MAPK/ERK pathway. In conclusion, while the current prevention efforts are promising but far from being conclusive, major strides have been made in the understanding of the epidemiological, molecular, and genetic underpinnings of melanoma, which are likely to translate into new opportunities for efficient and targeted melanoma prevention strategies in the near future.