Skin precancer.

Squamous cell carcinoma of the skin and melanoma are the rare progeny of precancerous lesions that usually remain stable or regress. For SCC the sequence appears to include TP53 mutant clones in normal skin; dysplasia; carcinoma in situ; and SCC. When such lesions are contiguous, their TP53 mutations are consistent with a single clonal lineage. The set of TP53 mutations in tumours is more restricted than in precancers, suggesting additional selection. Melanoma lies at the end of a continuum including mole, dysplastic naevus, radial growth melanoma and vertical growth. The genetics of melanoma is less clear. Basal cell carcinomas seem to arise without a precancer and contain mutations in TP53 and PTCH. Childhood sunlight exposure directs the location and frequency of precancers. For melanoma, its effects on intermittently exposed body sites are superimposed on the effect at sites chronically exposed. SCC precancers and tumours, BCC tumours and melanoma cell lines contain UV induced mutations. Sun exposed skin of normal individuals contains thousands of small clones of TP53 mutated cells. Predisposition to sunlight induced precancer is a multigenic trait involving factors such as hair and skin color, DNA repair proficiency and mole type and number. These each contribute a relative risk on the order of two to four. Familial predisposition to dysplastic naevi carries a larger risk. The cell of origin for melanoma is uncontroversial, and the proposed hair follicle origin of BCC is consistent with the presence of stem cells in the bulge region. The origin of SCCs and the arrangement of interfollicular stem cell compartments are less clear. Clonal expansion of the initial mutated cell may also be driven by sunlight. When a mutation confers apoptosis resistance, as TP53 mutations do, subsequent UV exposure will be more likely to kill normal cells than mutants. The latter can expand into a clone, only one cell of which need be mutated again. Immunosuppressant drugs may have the same effect as UV, facilitating the clonal expansion of precancers. In the absence of exogenous influences, mutant clones and precancers tend to regress. There is little evidence that regression of precancers is immunological, though regression of melanoma appears to be. The chemotherapeutic agent 5-FU causes regression of dysplasias by removing initiated cells, perhaps by enhancing apoptosis. In contrast, retinoic acid temporarily suppresses clonal expansion. Most sunscreens are mutagenic, with as yet unknown consequences. Mice develop dysplasias and SCCs after UV irradiation. Initiation and clonal expansion of dysplasias is UV driven, but conversion to SCC and subsequent growth involve spontaneous events. With chemical carcinogens mice develop papillomas that usually regress and thus are precancers. Tumour promotion yields abundant low risk papillomas that contain Hras1 mutations but rarely progress to SCC. High risk papillomas are infrequent but do convert to SCC, particularly if re-treated with mutagens. Conversion to SCC is associated with TP53 mutations. The mechanisms of multiple mutation and clonal expansion observed in human and mouse systems, respectively, are beginning to converge into a coherent understanding of precancerous events in skin.