Risk factors for melanoma include environmental (particularly ultraviolet exposure) and genetic factors. In rare families, susceptibility to melanoma is determined by high penetrance mutations in the genes CDKN2A or CDK4, with more common, less penetrant genes also postulated. A further, potent risk factor for melanoma is the presence of large numbers of melanocytic nevi so that genes controlling nevus phenotype could be such melanoma susceptibility genes. A large Australian study involving twins aged 12 y of predominantly U.K. ancestry showed strong evidence for genetic influence on nevus number and density. We carried out essentially the same study in the U.K. to gain insight into gene-environment interactions for nevi. One hundred and three monozygous (MZ) and 118 dizygous (DZ) twin pairs aged 10-18 y were examined in Yorkshire and Surrey, U.K. Nevus counts were, on average, higher in boys (mean = 98.6) than girls (83.8) (p = 0.009) and higher in Australia (110.4) than in the U.K. (79.2, adjusted to age 12 y, p < 0.0001), and nevus densities were higher on sun-exposed sites (92 per m2) than sun-protected sites (58 per m2) (p < 0.0001). Correlations in sex and age adjusted nevus density were higher in MZ pairs (0.94, 95%CI 0.92-0.96) than in DZ pairs (0.61, 95%CI 0.49-0.72), were notably similar to those of the Australian study (MZ = 0.94, DZ = 0.60), and were consistent with high heritability (65% in the U.K., 68% in Australia). We conclude that emergence of nevi in adolescents is under strong genetic control, whereas environmental exposures affect the mean number of nevi.

We report a study of 221 teenage twin pairs to examine the genetic and environmental determinants of nevi representing the most potent phenotypic risk factor for melanoma. Our published heritability analysis estimated that nevi are mainly genetically determined. In this paper we examine the role of sun exposure. We report a correlation between nevus density and sun exposure, particularly that acquired in hotter countries than in the UK (mean nevus density 41 per m2 in those in the highest quartile of exposure vs 24 per m2 in those with no exposure, p<0.0001). We were not able to demonstrate a protective effect for either sun protection cream or shirt wearing. By including phenotypic variables and reported sun exposure into the heritability analysis, we conclude that 66% of the total variance of nevus count is attributable to genetic effects: 7% associated to eye color, 6% to hair color, and 1% to reported skin type, which leaves 52% as to yet unidentified genetic factors. Of the 25% of variation attributable to environmental influences, one-third is estimated to be because of sun exposure on hot holidays.

Melanocytic nevi are the most potent risk factors for melanoma yet identified. Variation in the nevus phenotype within a population is predominantly genetically determined. Genes that determine nevus expression may therefore act as low penetrance melanoma susceptibility genes. Rare germline mutations in CDKN2A predispose to melanoma and appear to be nevogenic, although the correlation between nevus phenotype and mutation status is poor. It is plausible that more common CDKN2A variants may influence both melanoma susceptibility and nevus susceptibility. Ala148Thr is a G to A missense polymorphism of CDKN2A, which is found in 4%-6% of the general population. We have investigated the role of Ala148Thr as a low penetrance melanoma or nevus susceptibility allele in two separate groups of individuals. The first was a sample of 488 adults recruited from 179 families of patients with the atypical nevus phenotype and/or a family history of melanoma, and the second was a population-based sample of 599 women. Similar prevalences of Ala148Thr (4.9% and 5.2%) were found in both samples but significant variation in the prevalence of the polymorphism was seen across geographic areas within England. There was no association between Ala148Thr status and nevus number or history of melanoma, and therefore the results did not support the hypothesis that the Ala148Thr variant is a low penetrance melanoma or nevus susceptibility allele. A significant protective role of Ala148Thr on the number of atypical nevi was observed in the family sample (mean of 1 atypical nevus in those with the allele and 3.5 nevi in those without, p = 0.02). After allowing for potential confounders this was not evident in the population-based sample.

The inheritance of a G allele in position 61 in the 5'UTR of the epidermal growth factor (EGF) gene has been reported to increase melanoma susceptibility, a finding we have investigated in this study. The most potent phenotypic risk factor for melanoma is the atypical mole syndrome (AMS) phenotype. Our hypothesis is that the AMS is genetically determined and that nevus genes are also low penetrance melanoma susceptibility genes. We report that the G allele frequencies were the same in 697 healthy women and 380 melanoma cases (OR 0.97, 95% CI 0.8-1.2 p=0.76). We therefore found no evidence that this polymorphism is a melanoma susceptibility gene. Furthermore, we found no evidence that the polymorphism controls the nevus phenotype (nevus number, number atypical nevi or AMS phenotype). We did find some evidence that the G allele may be associated with decreased tumor Breslow thickness (OR 0.5, 95% CI 0.3-0.9) for the A/A genotype versus A/G and G/G combined in tumors of thickness >3.5 vs < or =3.5 mm and may therefore act as a predictor of survival, although this finding is not in accord with the original report. This is the second study to find no association between EGF +61 and melanoma susceptibility.