Cutaneous melanoma attributable to solar radiation in Cali, Colombia.

Estimating the population attributable fraction (PAF) of melanomas due to sun exposure is challenging as there are no unexposed population nor reliable exposure data. In high incidence countries, a historic cohort of the South Thames cancer registry was used as a minimally exposed population using the formula PAF = (observed incidence-incidence in minimally exposure)/observed incidence. In this study, we apply this method, constructing a minimally exposed cohort for Colombia and also using the historical South Thames data, using melanoma incidence data from the population-based cancer registry of Cali, Colombia for the period 1967-2012. The historic cohort incidence rates were very similar to those of Thames, but cohort effects were smaller for women and nonexistent for men. Age-specific incidence rates of these minimally exposed cohorts were applied to recent population numbers. For females, PAFs were 19% using the historic Thames cohort and 25% using the historic Cali cohort, corresponding numbers for males were 62% (vs. Thames) and 0% (vs. Cali). Taking into account the incidence rates of acral melanomas, which are not sun related, the PAF increased in women to 26% (vs. Thames) and 34% (vs. Cali) and for men 77% (vs. Thames). This exercise shows the modest contribution of exposure to ultraviolet radiation in the burden of melanoma in low-incidence countries, as well as the importance to take into consideration the acral lentiginous melanomas.