Seminars in Surgical Oncology 8345-352 (1992)
Cutaneous Malignant Melanoma: Epidemiological Considerations SlLVlA FRANCESCHI, MD, AND MARIO CRISTOFOLINI, MD From the Epidemiology Unit, Aviano Cancer Center, Aviano (PN) (S.F.); Department of Dermatology, 5. Chiara Hospital, Trento, Italy (M.C.)
Large geographic and ethnic variations in incidence and mortality rates of cutaneous malignant melanoma (CMM) suggest inverse correlations with the degree of skin pigmentation and latitude. Increases in CMM incidence and mortality rates have occurred in virtually all white populations, more markedly in those that permanently (immigrants) or temporarily (tourists/vacationers) reside in very sunny areas outside their primitive living environment. The interpretation of data from case-control studies is, however, hampered by the difficulties in quantifying retrospectively, but reliably, CMM risk correlates (e.g., host factors, sun exposure, clothing habits, sunburns) in various periods during the life span. There is, thus, a need for large prospective studies on the development of CMM and naevi, ideally in different contexts as concerns host factors, sun exposure, and other life-style habits within and across different populations. 0 1992 WiIey-Liss, Inc. KEY WORDS:cutaneous malignant melanoma, incidence, mortality, risk factors
INTRODUCTION
The incidence of cutaneous malignant melanoma (CMM) in white populations for the last forty years has been rising more rapidly than that of most other cancers [ 1-71. Although the prognosis has been improving, this has not been sufficient to prevent a steady rise in the death rates from the disease as well. The clarification of the extent to which CMM increases are related to the influence of more hazardous, modern lifestyles is, therefore, essential in order to establish a public counselling strategy with the aim of reversing the present upward temporal trends for CMM. The purpose of this review is to present evidence from the descriptive and analytical epidemiology of CMM, cognisant of the problems and limitations of the studies conducted so far. MATERIALS AND METHODS Mortality data are generally available for a longer period of time than are incidence data for the purpose of studying the descriptive epidemiology of a 0 1992 Wiley-Liss, Inc.
disease. Furthermore, mortality data tend to be more stable than incidence data since they exist on a national basis, whereas incidence data are available primarily on a regional basis. However, in a highly curable disease such as CMM [7], mortality data provide information only on a small proportion of patients who actually develop the disease, and are very sensitive to improvements in survival rates. Most of all, mortality data have often been shown to be inadequate in distinguishing between different histological types (e.g., melanomatous and nonmelanomatous skin cancers). Thus, although geographical differentials and temporal trends of both CMM mortality and incidence rates are considered, respectively, on the basis of the World Health Organization (WHO) mortality data bank and various editions of Cancer Incidence in Five Continents [l-51, incidence data are discussed to a greater extent. Special attention is paid to the comparison of Address reprint requests to Silvia Franceschi, M.D., Epidemiology Unit, Aviano Cancer Center, Via Pedemontana Occ., 3308 1 Aviano (PN), Italy.
346 Franceschi and Cristofolini TABLE I. Age-Standardized (World) Incidence Rates of Cutaneous Malignant Melanoma per 100,OOO in the 10 Cancer Registry Regions With Highest Rates Worldwide, 1978-1982 [Sl Males Rank (from highest to lowest) 1
2 3 4 5 6 7 8 9 10
Region
Incidence I 100.000 per year
Region
Incidence l100,000 per year
Australia. Queensland US, Hawaii. White Australia. West Australia. New South Wales New Zealand, Non Maori Australia, Victoria Australia, South Australia, Cap. Terr. US, Los Angeles. White US. Atlanta, White
30.9 22.7 18.3 17.1 15.6 13.9 13.8 12.8 12.4 12.3
Australia, Queensland New Zealand, Non Maori Australia, West Hawaii, White Australia, Cap. Terr. Australia, South Australia, New South Wales Australia, Victoria Israel, Jews born in Israel Tasmania
28.5 21.4 20.8 18.8 18.1 16.4 16.1 16.0 14.5 11.7
age-adjusted incidence rates in European countries
PI.
Females
As regards analytical epidemiology studies, three groups of potential risk factors are examined in detail: host factors, ultraviolet (UV) light exposure, and hormonal and reproductive factors. Most of the epidemiological evidence on a relatively rare disease such as CMM derives from case-control studies. Although a large number of case-control studies has been reviewed [8-251, the focus here is on four of the largest and most representative investigations on CMM, i.e., those conducted in Western Australia [26-301, Western Canada [31-371, Denmark [38-411, and England [42]. Details of such investigations are provided elsewhere [26-421. Briefly, the study from Western Australia includes 51 1 cases under 80 years of age, diagnosed between January 1980 and November 1981, and the same number of population controls matched by age in 5-year intervals, randomly selected from the electoral roll. The investigation from Western Canada provided information on 67 1 cases, aged 20-79, diagnosed between April 1979 and March 1981 and 671 population controls, matched by sex and age i2 years, drawn from medical insurance plans of the same province. The Danish case-control investigation deals with a total of 474 CMM patients, aged 20- 79, diagnosed between October 1982 and March 1985, and 926 controls, matched by sex and age in 5-year groups, randomly selected from the National Population Register of East Denmark. The most recent case-control investigation considered, the English study [42], includes 195 CMM cases aged 20-79 and diagnosed between January 1984 and December 1986 and 195 age- and sex-matched controls drawn from all inpatients and outpatients in hospitals serving the defined source population.
RESULTS Descriptive Epidemiology CMM is a disease largely confined to Caucasians [43]. The incidence in Whites is approximately three times higher than in nonwhites [ 5 ] , although Africans and American blacks tend to show higher CMM rates than do Asians [ 5 ] . The incidence of CMM varies more than 100-fold around the world. Areas covered by Cancer Registries in which the highest age-standardized incidence rates in both sexes have been recorded [5] include parts of the United States @., whites in Los Angeles, Atlanta, New Mexico, and Hawaii), Israel (Israeli-born Jews), New Zealand, and Australia, where the area with the highest CMM rate in the world (i.e., Queensland, incidence rates 30.9 and 28.5 in males and females, respectively) is located (Table I). Conversely, very low incidence rates (i.e.,
Cutaneous Malignant Melanoma: Epidemiological Considerations SlLVlA FRANCESCHI, MD, AND MARIO CRISTOFOLINI, MD From the Epidemiology Unit, Aviano Cancer Center, Aviano (PN) (S.F.); Department of Dermatology, 5. Chiara Hospital, Trento, Italy (M.C.)
Large geographic and ethnic variations in incidence and mortality rates of cutaneous malignant melanoma (CMM) suggest inverse correlations with the degree of skin pigmentation and latitude. Increases in CMM incidence and mortality rates have occurred in virtually all white populations, more markedly in those that permanently (immigrants) or temporarily (tourists/vacationers) reside in very sunny areas outside their primitive living environment. The interpretation of data from case-control studies is, however, hampered by the difficulties in quantifying retrospectively, but reliably, CMM risk correlates (e.g., host factors, sun exposure, clothing habits, sunburns) in various periods during the life span. There is, thus, a need for large prospective studies on the development of CMM and naevi, ideally in different contexts as concerns host factors, sun exposure, and other life-style habits within and across different populations. 0 1992 WiIey-Liss, Inc. KEY WORDS:cutaneous malignant melanoma, incidence, mortality, risk factors
INTRODUCTION
The incidence of cutaneous malignant melanoma (CMM) in white populations for the last forty years has been rising more rapidly than that of most other cancers [ 1-71. Although the prognosis has been improving, this has not been sufficient to prevent a steady rise in the death rates from the disease as well. The clarification of the extent to which CMM increases are related to the influence of more hazardous, modern lifestyles is, therefore, essential in order to establish a public counselling strategy with the aim of reversing the present upward temporal trends for CMM. The purpose of this review is to present evidence from the descriptive and analytical epidemiology of CMM, cognisant of the problems and limitations of the studies conducted so far. MATERIALS AND METHODS Mortality data are generally available for a longer period of time than are incidence data for the purpose of studying the descriptive epidemiology of a 0 1992 Wiley-Liss, Inc.
disease. Furthermore, mortality data tend to be more stable than incidence data since they exist on a national basis, whereas incidence data are available primarily on a regional basis. However, in a highly curable disease such as CMM [7], mortality data provide information only on a small proportion of patients who actually develop the disease, and are very sensitive to improvements in survival rates. Most of all, mortality data have often been shown to be inadequate in distinguishing between different histological types (e.g., melanomatous and nonmelanomatous skin cancers). Thus, although geographical differentials and temporal trends of both CMM mortality and incidence rates are considered, respectively, on the basis of the World Health Organization (WHO) mortality data bank and various editions of Cancer Incidence in Five Continents [l-51, incidence data are discussed to a greater extent. Special attention is paid to the comparison of Address reprint requests to Silvia Franceschi, M.D., Epidemiology Unit, Aviano Cancer Center, Via Pedemontana Occ., 3308 1 Aviano (PN), Italy.
346 Franceschi and Cristofolini TABLE I. Age-Standardized (World) Incidence Rates of Cutaneous Malignant Melanoma per 100,OOO in the 10 Cancer Registry Regions With Highest Rates Worldwide, 1978-1982 [Sl Males Rank (from highest to lowest) 1
2 3 4 5 6 7 8 9 10
Region
Incidence I 100.000 per year
Region
Incidence l100,000 per year
Australia. Queensland US, Hawaii. White Australia. West Australia. New South Wales New Zealand, Non Maori Australia, Victoria Australia, South Australia, Cap. Terr. US, Los Angeles. White US. Atlanta, White
30.9 22.7 18.3 17.1 15.6 13.9 13.8 12.8 12.4 12.3
Australia, Queensland New Zealand, Non Maori Australia, West Hawaii, White Australia, Cap. Terr. Australia, South Australia, New South Wales Australia, Victoria Israel, Jews born in Israel Tasmania
28.5 21.4 20.8 18.8 18.1 16.4 16.1 16.0 14.5 11.7
age-adjusted incidence rates in European countries
PI.
Females
As regards analytical epidemiology studies, three groups of potential risk factors are examined in detail: host factors, ultraviolet (UV) light exposure, and hormonal and reproductive factors. Most of the epidemiological evidence on a relatively rare disease such as CMM derives from case-control studies. Although a large number of case-control studies has been reviewed [8-251, the focus here is on four of the largest and most representative investigations on CMM, i.e., those conducted in Western Australia [26-301, Western Canada [31-371, Denmark [38-411, and England [42]. Details of such investigations are provided elsewhere [26-421. Briefly, the study from Western Australia includes 51 1 cases under 80 years of age, diagnosed between January 1980 and November 1981, and the same number of population controls matched by age in 5-year intervals, randomly selected from the electoral roll. The investigation from Western Canada provided information on 67 1 cases, aged 20-79, diagnosed between April 1979 and March 1981 and 671 population controls, matched by sex and age i2 years, drawn from medical insurance plans of the same province. The Danish case-control investigation deals with a total of 474 CMM patients, aged 20- 79, diagnosed between October 1982 and March 1985, and 926 controls, matched by sex and age in 5-year groups, randomly selected from the National Population Register of East Denmark. The most recent case-control investigation considered, the English study [42], includes 195 CMM cases aged 20-79 and diagnosed between January 1984 and December 1986 and 195 age- and sex-matched controls drawn from all inpatients and outpatients in hospitals serving the defined source population.
RESULTS Descriptive Epidemiology CMM is a disease largely confined to Caucasians [43]. The incidence in Whites is approximately three times higher than in nonwhites [ 5 ] , although Africans and American blacks tend to show higher CMM rates than do Asians [ 5 ] . The incidence of CMM varies more than 100-fold around the world. Areas covered by Cancer Registries in which the highest age-standardized incidence rates in both sexes have been recorded [5] include parts of the United States @., whites in Los Angeles, Atlanta, New Mexico, and Hawaii), Israel (Israeli-born Jews), New Zealand, and Australia, where the area with the highest CMM rate in the world (i.e., Queensland, incidence rates 30.9 and 28.5 in males and females, respectively) is located (Table I). Conversely, very low incidence rates (i.e.,
