Melanoma survival: sex does matter, but we do not know how.

The objective of this study was to provide further insights into the prognostic role of female sex in skin melanoma. The prognostic effect of sex in a population-based case series of 3900 skin melanomas in central Italy has been evaluated considering the possible confounding role of many demographic and clinical variables (age, period of diagnosis, Breslow's thickness, Clark level, ulceration, lymph node status, metastasis, histological type, skin site, and pathological T and N). Multiple imputations, according to chained equations, have been used for imputing incomplete values. A Cox proportional hazards model on the risk of death caused by melanoma was fitted. Univariate and multivariate effects of sex and of other variables were computed. The 5-year cause-specific survival was 87% (95% confidence interval: 86-89%) for women and 80% (78-82%) for men. Women had higher rates at any time since diagnosis. After adjustment for other confounders, women had a 34% reduced risk compared with men of dying from skin melanoma (hazard ratio=0.66, 95% confidence interval: 0.56-0.79). The present study confirmed a strong protective effect of female sex on skin melanoma mortality. The protective factor is still unknown.