Background: Skin cancers are known to be associated with sun exposure, whereas sunlight through the production of vitamin D may protect against some cancers. The aim of this study was to assess whether patients with skin cancer have an altered risk of developing other cancers. Methods: The study cohort consisted of 416,134 cases of skin cancer and 3,776,501 cases of non-skin cancer as a first cancer extracted from 13 cancer registries. 10,886 melanoma and 35,620 non-melanoma skin cancer cases had second cancers. The observed numbers (O) of 46 types of second primary cancer after skin melanoma, basal cell carcinoma or non-basal cell carcinoma, and of skin cancers following non-skin cancers were compared to the expected numbers (E) derived from the age, sex and calendar period specific cancer incidence rates in each of the cancer registries (O/E = SIR, standardised incidence ratios). Rates from cancer registries classified to sunny countries (Australia, Singapore and Spain) and less sunny countries (Canada, Denmark, Finland, Iceland, Norway, Scotland, Slovenia and Sweden) were compared to each other. Results: SIR of all second solid primary cancers (except skin and lip) after skin melanoma were significantly lower for the sunny countries (SIR(S) = 1.03; 95% CI 0.99-1.08) than in the less sunny countries (SIR(L) = 1.14; 95%CI 1.11-1.17). The difference was more obvious after non-melanoma skin cancers: after basal cell carcinoma SIR(S)/SIR(L) = 0.65 (95%CI = 0.58-0.72); after non-basal cell carcinoma SIR(S)/SIR(L) = 0.58 (95%CI = 0.50-0.67). In sunny countries, the risk of second primary cancer after non-melanoma skin cancers was lower for most of the cancers except for lip, mouth and non-Hodgkin lymphoma. Conclusions: Vitamin D production in the skin seems to decrease the risk of several solid cancers (especially stomach, colorectal, liver and gallbladder, pancreas, lung, female breast, prostate, bladder and kidney cancers). The apparently protective effect of sun exposure against second primary cancer is more pronounced after non-melanoma skin cancers than melanoma, which is consistent with earlier reports that non-melanoma skin cancers reflect cumulative sun exposure, whereas melanoma is more related to sunburn.
We acknowledge the work of Didier Colin, IARC, for initial preparation of the dataset, and to Aage Andersen for participation in early phase of this study. The study was supported by grants from the Tampere University Hospital, the Finnish Cancer Foundation and the Academy of Finland (P. Tuohimaa), Grant R03 CA101442 from the National Cancer Institute (P. Brennan) and Grant R03 CA101442-02 from the National Cancer Institute (P. Boffetta). G. Scélo worked on this study during the tenure of a Special Training Award from the International Agency for Research on Cancer. During the publication process of this paper two groups described similar results on prostate cancer incidence 42 and on some solid cancers 43 after the primary skin cancer.