Changing incidence and mortality from cutaneous malignant melanomaBMJ 1997; 315 doi: https://doi.org/10.1136/bmj.315.7116.1106 (Published 01 November 1997) Cite this as: BMJ 1997;315:1106
The reasons are not yet clear
- Jane Melia, Epidemiologista
Some evidence of a decline or stabilising of death rates from malignant melanoma is now seen in several countries with white populations, at least in some age groups.1 2 3 4 Knowing why these changes have occurred may help plan primary and secondary strategies to prevent skin cancer. The most likely explanations for the changing mortality from melanoma are increased awareness about early detection, improved sun protection, changes in other environmental factors, and changing natural history (rather than artifacts from the recording of deaths or changes in treatment). The data from McKie's group in Scotland published in this week's issue (p 1117)1 are unique to Britain because the Scottish melanoma register has ensured good quality population based data with which to study trends in the incidence of melanoma, assessed by Breslow thickness, alongside those for mortality since 1979.
Early detection has been considered an attractive method for prevention because of the strong association between survival and Breslow thickness of the lesion. There has, however, been no large scale randomised controlled trial to evaluate the benefits of early detection of melanoma in the general population. Screening benefits those with familial melanoma but these represent less than 5% of all cases.5 In a previous report from McKie's group a decline in mortality from and incidence of thick melanomas in women from 1985 to 1987 was suggested to have resulted from a campaign for early detection.6 However, such a campaign is unlikely to have had an impact on mortality so soon after its introduction. The trends in Scotland in 1979–94 show a levelling in mortality which started before the launch of the early detection campaign in 1985. If early detection has contributed to this trend, the high profile approach of 1985, which created a large hospital workload, may not be the best approach. Alternative approaches to early detection of melanoma should be explored to find the most cost effective.
The relation between mortality and incidence is complex. A reduction in mortality might be expected to be preceded by a decline in the incidence of thick melanomas, although this appears not to be so in Scotland. The considerable increase seen in the incidence of thin melanomas in Scotland, which undoubtedly is related to improved early detection, could lead to overdiagnosis of malignancy7 and the diagnosis of a slow growing or non-progressive melanoma.8 It is also possible that the natural history of melanoma and its rates of progression could change in response to changes in exposure to environmental factors such as ultraviolet radiation. The improved survival adjusted for histological factors could be a further indication of a changing natural history in diagnosed melanoma.
Improved sun protection might help to reduce both the incidence of and mortality from melanoma. Research into the causes of melanoma strongly supports the role of ultraviolet radiation,9 although the relation is not fully understood. Most British sun protection programmes have been launched in the 1990s, although awareness about primary prevention will have been increasing in the 1980s, partly through early detection activities.6 10 If the health benefits from primary prevention are not expected for up to 20 years, they are unlikely to have caused the levelling of incidence rates in Scotland. However, if ultraviolet radiation acts as a promoter as well as an initiator of skin cancer, reduced exposure to sun could slow down the progression of early lesions and their precursors in a shorter period and thus contribute to reduced incidence. Other health factors such as diet might be linked to skin cancer.11 Conceivably changes in these factors might be related to both the incidence and the rate of progression of melanoma, thus affecting trends in incidence of melanomas of different Breslow thicknesses.
The results from Scotland are encouraging, but their explanation remains intriguing. Trends in mortality and incidence rates help to generate hypotheses, but more research is needed into the natural history of the disease, the factors that affect the progression of melanoma and its precursors, and the evaluation of different options for primary and secondary interventions. Baseline data with which to monitor trends are essential. Individual studies may not produce large changes in attitudes and behaviour but overall increased awareness may contribute towards population changes needed to improve health.12