@article{3119549, title = "A multicentre study of naevus-associated melanoma vs. de novo melanoma, tumour thickness and body site differences*", author = "Dessinioti, C. and Geller, A.C. and Stergiopoulou, A. and Dimou, N. and Lo, S. and Keim, U. and Gershenwald, J.E. and Haydu, L.E. and Dummer, R. and Mangana, J. and Hauschild, A. and Egberts, F. and Vieira, R. and Brinca, A. and Zalaudek, I. and Deinlein, T. and Evangelou, E. and Thompson, J.F. and Scolyer, R.A. and Peris, K. and Garbe, C. and Stratigos, A.J.", journal = "British Journal of Dermatology", year = "2021", volume = "185", number = "1", pages = "101-109", publisher = "John Wiley and Sons Inc", issn = "0007-0963, 1365-2133", doi = "10.1111/bjd.19819", keywords = "adult; aged; Article; Breslow thickness; cancer localization; cancer patient; cancer prognosis; cancer regression; clinical feature; cutaneous melanoma; de novo melanoma; female; head; histopathology; human; human tissue; lower limb; major clinical study; male; melanoma; neck; nevus associated melanoma; retrospective study; sex difference; superficial spreading melanoma; trunk; tumor thickness; upper limb; Australia; clinical trial; Europe; melanoma; multicenter study; skin tumor, Australia; Europe; Female; Humans; Melanoma; Retrospective Studies; Skin Neoplasms", abstract = "Background: Whether melanoma in histological contiguity with a naevus [naevus-associated melanoma (NAM)] is distinctly different from melanoma arising de novo remains unclear. Objectives: To determine whether the characteristics of de novo melanoma differ from NAM and are not due to naevus obliteration in thicker tumours. Methods: We conducted a multicentre retrospective study of de novo melanoma and NAM in seven referral centres in Europe, Australia and the USA between 2006 and 2015. Results: In a total of 9474 localized melanomas, de novo melanoma was associated with thicker tumours and body site differences compared with NAM. In the subset of T1 melanomas (n = 5307), similar body site differences were found in multivariate analysis by body site. When compared with NAM, de novo melanoma was more likely to affect older individuals (≥ 70 years) when located on the head/neck [odds ratio (OR) 4·65, 95% confidence interval (CI) 2·55–8·46], the trunk (OR 1·82, 95% CI 1·40–2·36) or the upper extremity (OR 1·69, 95% CI 1·14–2·50), was more likely to affect female patients when located on the lower extremities (OR 1·36, 95% CI 1·03–1·80), and was more likely to be of the nodular melanoma subtype (OR 2·23, 95% CI 1·14–4·35) when located on the trunk. De novo melanoma was less likely to have regression present compared with NAM. Conclusions: Clinicopathological and body site differences between de novo melanoma and NAM support the divergent pathway model of development. These differences were also found in thin melanomas, suggesting that de novo melanomas are different from NAM and their differences are not due to the obliteration of naevus remnants in thicker tumours. © 2021 British Association of Dermatologists" }