Abstract

The Clark model for melanoma progression emphasizes a series of histopathological changes beginning from benign melanocytic nevus to melanoma via dysplastic nevus. Several models of the genetic basis of melanoma development and progression are based on this Clark's multi-step model, and predict that the acquisition of a BRAF mutation can be a founder event in melanocytic neoplasia. However, our recent investigations have challenged this view, showing the polyclonality of BRAF mutations in melanocytic nevi. Furthermore, it is suggested that many melanomas, including acral and mucosal melanomas, arise de novo, not from melanocytic nevus. While mutations of the BRAF gene are frequent in melanomas on non-chronic sun damaged skin which are prevalent in Caucasians, acral and mucosal melanomas harbor mutations of the KIT gene as well as the amplifications of cyclin D1 or cyclin-dependent kinase 4 gene. Amplifications of the cyclin D1 gene are detected in normal-looking 'field melanocytes', which represent a latent progression phase of acral melanoma that precedes the stage of atypical melanocyte proliferation in the epidermis. Based on these observations, we propose an alternative genetic progression model for melanoma.