The analysis of the GLOBOCAN 2008 data adds to earlier reports of geographical variation of melanoma incidence, with North–South and West–East gradients observed since the 1990s.[6,7,9] As GLOBOCAN 2008 includes more CEE countries than previous studies, these estimates for 2008 reveal even wider variation of melanoma incidence across the continent, with more than eightfold differences between Switzerland and Romania or Greece.
Varying rates of melanoma incidence by geography might be explained by the presence of darker skin phenotypes in Eastern Europe and greater affluence allowing for intense recreational sun-exposure in North and Western Europe. However, rate differences appear to be greater than expected between neighbouring countries such as Poland and Germany, or Romania and Hungary, which share many factors related to melanoma risk, e.g. latitude, ultraviolet index distribution, cutaneous phototype and pigmentation distribution.[14–16] Moreover, comparable ethnic, cultural, religious and socioeconomic patterns[17–19] among these nations make wide differences of sun-exposure behaviours less likely.
Future comparisons should examine the variation of early sun-exposure behaviours, recreational and vacation histories, and naevus phenotypes, skin phototypes and melanoma risk genotypes across the continent. However, it is likely that the wide disparity in melanoma incidence throughout Europe is related in part to significant discrepancies in the registration of melanoma and the availability of information on melanoma in CEE. In CEE, there are suboptimal registration systems for melanoma: up to a third of countries have no functional population-based cancer registries, and for such countries GLOBOCAN estimates provide the only official updated information, usually based on cause-of-death statistics and information from neighbouring countries. As the gap between the number of cases and deaths, or the incidence/mortality ratio, differs across the countries, using deaths to predict the number of cases may have some imprecision.
Concerning melanoma mortality, Europe's rate of 1·5/100 000 is the third highest in the World, after Australia/New Zealand (3·5/100 000) and North America (1·7/100 000). Among the four geographical regions of Europe, the CEE countries have the largest share (35·5%) of the more than 20 000 melanoma deaths estimated to occur annually throughout the continent. Thus, stabilizing or reducing mortality rates here could have a profound effect across Europe. Mortality is higher for men than women in all European regions, and this may be related to later detection of thicker melanoma among men,[1,8,20,21] and to the recently indicated survival benefit for women.[22,23]
The GLOBOCAN 2008 and Cancer in Five Continents databases do not provide information on the stage distribution of melanoma in Europe. Accurate reports of melanoma thickness in CEE are scarce and most of the evidence comes from single-centre or local reports. However, the available data indicate that a far higher proportion of melanomas are diagnosed in their later stages in CEE[6,24–26] compared with Northern and Western Europe. In the CEE countries for which the reported incidence rates are very low, thick melanomas are commonly diagnosed and it appears that the shift towards early diagnosis has not yet occurred. For example, in Bulgaria and Romania, an estimated 25% of patients present with stage III and IV disease,[24,27] while thin melanomas under 1 mm thick comprise less than 10% of newly diagnosed cases. High rates of advanced-stage melanomas have been reported in the Russian Federation, and a median tumour thickness of 4 mm has been reported in Serbia. This is in sharp contrast to Western Europe, where up to 70% of newly diagnosed melanomas are <1 mm thick[1,2,8,30] and account for most of the increase in incidence. The relative proportion of histological subtypes does not appear to be different in CEE countries compared with Western Europe,[2,26,27] although the subtype is often unreported. Overdiagnosis of very early melanomas might be contributing to the dramatic increase in melanoma incidence in Western European countries,[1,2] although recent reports have shown that melanoma incidence rates in some Western European countries have actually increased across all of the thickness categories.[8,32]
Later diagnosis of advanced stage melanoma in CEE compared with other parts of Europe may stem in part from differences in general educational status, reduced government expenditures for education and lack of specific health education, resulting in lower awareness among the general population and physicians alike. Skin cancer education and prevention campaigns were introduced to occidental Europe in the 1980s, whereas many countries in CEE have only recently launched prevention initiatives such as Euromelanoma.
Physicians' ability to diagnose melanoma correctly may also be influenced by differences in healthcare systems in Eastern vs. Western Europe. Health expenditure per capita varies greatly in Europe, from 304 ppp$/head in the Republic of Moldova to 5734 ppp$/head in Luxembourg (ppp$: purchasing power parity dollars, 1 ppp$ is equivalent to 1 U.S. dollar spent in the U.S.A). Lower healthcare budgets in CEE translate to a lower availability of diagnostic tools such as dermoscopy, the understaffing of medical centres, and fewer technical and human resources for pathology laboratories. The large disparities in access to primary and specialist medical care between rural and urban areas in CEE countries[19,34] may be a further challenge for the correct diagnosis of melanoma. In some CEE countries, such as the Czech Republic, Bulgaria or Romania, the density of physicians in the capital cities is more than twice that of the average density of physicians for the nation.[19,34] Many CEE nations have large rural populations (46% in Romania). While rural areas generally report lower incidence rates for melanoma than the cities, in some Eastern European countries the rural areas provide the majority (up to 70%) of thick melanomas.[24,27]
In many European countries, melanoma is reported to the national and regional cancer registries, and the establishment of comprehensive cancer registries is a priority for the cancer control strategy of the European Union. The registries meeting sufficient quality criteria provide the basis for the major global and European databases.[4,11–13]
However, quality cancer registries currently exist in only 29 of 41 European countries, comprising only 33% of the European population. Several countries have only local registries, with highly variable coverage, such as in Poland, whose four local registries cover only 10% of the population. The quality of data collection and reporting by cancer registries is also highly variable. Based on the quality of their cancer registries, only four of 10 countries in CEE (Czech Republic, Poland, Slovakia and Slovenia) have been included in European studies.[4,11]
Many of the CEE and Southern European countries do not have centralized cancer registries of high quality, and their incidence data for melanoma were only estimated in GLOBOCAN 2008 through modelling from neighbouring countries (Table 1, Fig. 4), resulting in the lowest incidence estimates in Europe. However, recent data from a pilot regional cancer registry in Romania reported incidence rates of 12/100 000, more than four times higher than the GLOBOCAN estimates and the rates recorded by neighbouring countries. This pilot study points to the possibility of a large discrepancy between the modelled estimates and the actual data, and draws attention to the possible under-reporting of cases in the neighbouring CEE countries. Also, when new registries begin, they tend to register new cases as well as a number of cases from prior years. Under-reporting of melanoma may be a serious concern in CEE, particularly for earlier-stage cases that are frequently diagnosed in outpatient and private settings, which in some countries are less reliably covered by cancer registration.
Methods of cancer registration and incidence estimation for melanoma in Europe (as used by GLOBOCAN 2008). Red: countries with no cancer registry, incidence estimate based on neighbouring countries. Yellow: countries with regional cancer registries. Light green: countries with national cancer registries, incidence data available up to 2002. Dark green: countries with national cancer registries, incidence data available up to 2007.
In contrast, the recording of deaths from melanoma is less influenced by health expenditure and exists in all countries. Eastern and Western Europe have a similar estimate of mortality rates (1·5/100 000), which is also the average mortality rate of melanoma for the whole European Union. However, within each European region mortality rates vary, with up to twofold differences between neighbouring countries (e.g. Hungary and Romania). Nonetheless, recording deaths from melanoma is not without its problems and under-reporting may occur, especially in CEE countries with large rural areas that may have fewer resources to correctly identify melanoma as the cause of death.
The study has some limitations. Herein we report the melanoma incidence and mortality data as estimated in GLOBOCAN 2008. No separate analyses or revisions to the GLOBOCAN methods were employed. There is the possibility of under-reporting from certain countries, overdiagnosis from others and an over-reliance on estimates derived from neighbouring countries combined with national mortality statistics. Because of suboptimal reporting in earlier periods, comparisons over time would be imprecise. Despite these inherent flaws, establishing and documenting baseline rates is paramount to melanoma control in Europe. GLOBOCAN 2008 is an important improvement from earlier years and control efforts must embrace improved registration as being central to any control effort.
Based on the UN classification, GLOBOCAN categorized the European countries into four geographical regions. However, we are aware that beyond geographical location, the historical, political, economic and cultural backgrounds of European countries might also be important criteria for epidemiological analysis.
To conclude, both melanoma incidence and mortality rates, as well as data availability, appear to vary significantly between European countries. Our findings and subsequent review of the existing literature highlight the compelling need to improve melanoma registration in CEE, to understand the variation of melanoma risk factors across countries, to train physicians on recording and reporting data on melanoma, and to conduct an objective assessment of the capacity of current healthcare systems to detect, record and manage melanoma. Such studies are needed to provide the basis for building integrated strategies of promoting healthcare and education in CEE, as well as to plan for coordinated cancer control programmes to reduce disparities in care and survival across the continent.
A.M.F. has received grant support from Arbeitsgemeinschaft Dermatologische Prävention (ADP) e.V.
The British Journal of Dermatology. 2012;167(5):1124-1130. © 2012 Blackwell Publishing