- Email: [email protected]
The global burden of cancer
seminars in C A N C E R B I OLOG Y, Vol 8, 1998: pp 219]235
Article No. sc980080
The global burden of cancer D. M. Parkin
The number of new cancer cases, and the annual incidence rates, of 25 different cancers have been estimated for the year 1990 for every country of the world. The distributions of the most common cancers in men and women are presented for 23 broad ‘ Areas’. The total number (excluding nonmelanoma skin cancer) was 8.1 million, just over half of which occur in the developing countries. The most common cancer is lung cancer, which accounts for 18% of cancers of men worldwide. Stomach cancer is second in frequency (almost 10% of all new cancers) and breast cancer}by far the most important cancer of women (21% of the total) is third. There are very large differences in the relative importance of the different cancers by world area; some of the factors, environmental and genetic, underlying the geographic distributions, are discussed.
This review concentrates upon describing the current cancer pattern and, for the major cancers, examines incidence by geographic location, briefly discusses temporal trends and, finally, indicates some of the underlying reasons for these patterns and trends, with an attempt at approximate quantitation. Global cancer patterns are described in terms of incidence, the number of new cancer cases, which, expressed as a rate, is an indicator of the average risk of developing cancer to individuals in the population. It is therefore the appropriate statistic to use if one is interested in the causes of cancer, or Žprimary. cancer prevention. Incidence data are available from cancer registries. Disease registers are not unique to cancer, but for this condition, they comprise a particularly well developed and extensive network. They may cover entire national populations, or sub-samples of selected regions. They also provide statistics on cancer survival, thereby enabling incidence to be estimated from mortality. Mortality data by cause are available for many countries through registration of vital events, although the degree of detail and quality of the data Žboth the accuracy of the recorded cause of death and the completeness of registration. vary considerably. With such data, it is possible to prepare estimates of the numbers of new cancer cases Žand deaths. by site, sex and age group. These are more or less accurate, for different countries, depending on the extent and accuracy of locally available data. We have estimated the incidence of cancer worldwide, at intervals of 5 years, the figures relating to 1975,1 1980,2 1985,3 and 1990.4 In this paper, the estimates of 1990, for 23 world ‘Areas’ Žas defined by the UN ŽFigure 1. are presented. However, the estimates are built up at country level, and the full dataset can be accessed either via a CD-ROM 5 or, with less flexibility in the analyses possible, on the Internet Žhttp:rrwww-dep.iarc.frrglobocanrglobocan.htm.. The discussion makes references to the more detailed national-level data as appropriate. No attempt has been made to estimate incidence of non-melanoma skin cancer, because of the difficulties of measurement, and consequent lack of data.
Key words: new cancer cases r annual incidence rates r most common cancers Q1998 Academic Press
Introduction STATISTICS DESCRIBING THE pattern of cancer occurrence in human populations are essential for two purposes. Firstly, by showing how risk of disease varies according to place, time and personal characteristics, the search for more defined environmental exposures, or differential susceptibility to them, can be refined in specially designed studies. Secondly, such data permit rational priority-setting of healthcare activities Žprevention, screening and therapy. and the monitoring the results of interventions through the healthcare system.
From the Unit of Descriptive Epidemiology, International Agency for Research on Cancer, 150 cours Albert Thomas, 69372 Lyon Cedex 08, France Q1998 Academic Press 1044-579Xr 98r 040219q 17 $30.00r 0
219
D. M. Parkin
Figure 1. Map showing the 23 world areas studied.
The total ‘All Cancer’ therefore excludes such tumours. v
Methods of estimation v
Incidence rates for five broad age groups Ž0]14, 15]44, 45]54, 55]64 and 65 and over. and sex were estimated for as many individual countries as possible. Age-standardized incidence rates ŽASRs. were calculated using the weights of the ’world standard’ population Ž0.31, 0.43, 0.11, 0.08 and 0.07. in these five age classes. The Area estimates were obtained by combining agersex-specific rates for component countries as a weighted average Žusing the corresponding country populations.. The sources of data for the countries of the 23 world areas and the methods used to produce estimates of incidence in each of them are summarised in Parkin et al.4 There are six categories Žin order of preference.: v National Incidence Data from good quality cancer registries. v National Mortality Data, with estimation of inci-
v
v
dence using sets of regression models, specific for site, sex, and age, derived from local cancer registry data Žincidence plus mortality.. Local (Regional) Incidence Data from one or more regional cancer registries within a country. Mortality surveys}for China, the mortality data derive from a sample survey of deaths at socalled ‘Disease Surveillance Points’ ŽDSP..6 Frequency Data: In some areas of Africa and Asia, only data on the relative frequency of different cancers Žby age and sex. are available. These are applied to an estimated ‘all sites’ incidence rate, derived from existing cancer registry results. No data: The country-specific rates are those of the corresponding Area Žcalculated from the other countries for which estimates could be made..
Results and discussion An estimated 8.1 million new cancer cases occurred 220
The global burden of cancer
Lung cancer
in 1990, divided almost exactly between developed and developing countries. Figure 2 Žmales. and Figure 3 Žfemales. show the number of cases and the ranking of the 25 sites in developed and developing countries. Table 1 shows the number of cases, by world Area, for the 12 most common cancers in males and in females. Table 2 shows the corresponding age standardised incidence rates. The number by world-area ranges from 1.4 million in China to approximately 1100 in MicronesiarPolynesia. Some caution is necessary in comparing the most recent estimates Žfor 1990. with those for earlier periods. This is because the nature and availability of different sources of data on cancer has changed over time. In general, this should result in progressively more accurate figures, but it does lead to some abrupt changes in estimated incidence for certain areas. Having noted this, the progressive increase in the estimated number of cancer cases worldwide since 1975 has been rather more rapid than the equivalent growth in the world population ŽFigure 4..
This was the most common cancer in 1990, with 1.04 million new cases, or 12.8% of the world total. Worldwide, it is by far the most common cancer of men Ž18% of the total. with the highest rates observed in North America and Europe Žespecially eastern Europe.. Moderately high rates are also seen in temperate South America, AustraliarNew Zealand, and, in Asia, Hong Kong, Singapore and the Philippines ŽFigure 5.. In females, incidence rates are lower Žoverall, the rate is 10.8 per 10 5 women, compared with 37.5 per 10 5 in men.. The highest rates are in the USA and Canada, UK, Denmark. It is of note that the incidence in China is rather high Žage-standardrate 13.4 per 10 5 ., similar to that in, for example Australia or the Netherlands. The major cause of lung cancer is tobacco smoking, and in general, incidence rates in a country closely reflect the past history of tobacco smoking.7 The proportion of lung cancer cases due to tobacco smoking can be estimated by comparing observed
Figure 2. Estimated numbers of new cases of 21 cancers in men in developed and developing areas of the world.
221
D. M. Parkin
Figure 3. Estimated numbers of new cases of 23 cancers in women in developed and developing areas in the world.
incidence in different areas with that expected based upon rates in non-smokers from several large cohort studies.8 Updating the results to 1990, some 86% of lung cancer cases in men and 49% in women are due to smoking, though there is considerable regional variation. Thus in countriesrregions with a long history of smoking, 90% or more cases in men are tobacco-related, while the fraction is much lower in Africa and southern Asia. The proportions are more variable in women, even in Europe, where they range from 80% in the UK, to virtually nil in Spain and Portugal. Our estimate of the numbers of cases worldwide has increased by 16% since 1985 Žq14% in men and q21% in women.; this represents an approximately 2.5% increase in the actual risk in men, and 9.5% in women Žthe remainder is due to population growth and ageing.. Trends vary enormously by individual countries. Thus, in men, several populations have now passed the peak of the tobacco-related epidemic, and incidence rates are now declining Že.g. USA, the countries of northern and western Europe.. In con-
trast, rapid increases are still seen in southern and eastern European countries.9 In women, the ‘epidemic’ is less advanced; most ‘Western’ countries are still showing a rising trend in incidence and mortality, although for some there is, so far, no evidence of this Že.g. Spain., while for others ŽUK., it seems that the peak of risk may already have been reached.10 Stomach cancer Cancer is the second in rank overall Ž798,000 new cases, 9.9% of the total., although only fourth in women. The estimated number of cases is just 6% more than in 1985 which, allowing for population increase and ageing, represents a decline of approximately 4]5% in age-adjusted risk. Thirty-eight percent of cases occur in China, where it remains the most common cancer in both sexes, as it is elsewhere in eastern Asia}the highest incidence is still found in Japan ŽASR 77.9 per 100,000 in men, 33.3 in women. followed by Korea Ž59.9 and 24.8.. The incidence is also high in much of Central America and 222
89.2 2.2 29.8 1.2 46.6 5.1 32.5 3.8 65.0 1.1 28.1 4.1 49.8 5.7 59.8 12.8 158.0 18.4 634.3 118.2 866.4 163.1 224.1 32.9 58.9 9.2 181.4 38.3 452.5 47.8 63.8 14.1 435.1 124.5 182.5 41.0 244.4 56.9 350.0 65.3 38.3 5.9 2.4 0.1 0.5 0.1 2108.7 444.7 2184.8 327.1 4293.5 771.8
1. Eastern Africa 2. Middle Africa 3. Northern Africa 4. Southern Africa 5. Western Africa 6. Caribbean 7. Central America 8. South America ŽTemperate. 9. South America ŽTropical. 10. North America 11. Eastern Asia: China 12. Eastern Asia: Japan 13. Eastern Asia: Other 14. South-eastern Asia 15. South-central Asia 16. Western Asia 17. Eastern Europe 18. Northern Europe 19. Southern Europe 20. Western Europe 21. AustraliarNew Zealand 22. Melanesia 23. MicronesiarPolynesia
Developed countries Developing countries
ALL AREAS
Lung
All sites
MALE
223 511.0
191.7 319.3
4.1 3.6 2.4 1.4 5.5 2.3 5.6 5.6 23.9 14.4 207.7 64.7 13.3 11.9 27.7 4.3 58.9 11.8 20.5 20.1 1.3 0.1 0.0
Stomach
401.9
257.3 144.6
3.6 0.4 2.5 1.4 2.3 2.2 2.7 6.4 11.7 77.1 64.9 32.9 5.2 15.9 21.0 4.0 40.8 23.8 28.2 48.9 5.6 0.2 0.0
Colonr rectum
396.1
300.5 95.7
6.4 4.3 1.9 3.5 8.4 6.0 7.0 5.5 20.5 170.0 4.7 7.3 1.1 6.8 16.6 2.8 22.1 26.2 17.7 50.7 6.5 0.1 0.0
Prostate
316.3
53.9 262.4
9.1 5.8 1.9 2.7 11.3 1.1 1.8 0.6 2.6 5.4 178.1 23.2 8.4 24.8 11.5 2.3 8.4 1.9 8.8 5.9 0.3 0.3 0.0
Liver
212.6
44.8 167.9
5.6 0.4 1.1 4.5 0.9 1.0 1.0 2.6 5.9 8.6 102.3 8.0 1.7 4.1 35.7 1.1 10.4 4.5 4.5 8.1 0.6 0.0 0.0
Oesophagus
Table 1. Estimated number of new cancer cases Žthousands. by site, sex and area, 1990
202.5
127.7 74.9
2.5 0.5 10.0 1.4 3.2 1.1 1.5 3.1 7.2 41.7 14.9 7.5 2.0 5.4 17.1 4.8 20.2 13.9 22.0 20.6 1.8 0.0 0.0
Bladder
141.2
57.5 83.7
3.6 0.9 1.8 2.0 2.8 1.1 1.2 1.6 6.8 12.7 5.8 2.5 0.9 5.2 47.6 1.8 15.1 3.2 9.4 12.7 1.8 0.5 0.0
Oral cavity
130.3
55.4 74.9
1.8 0.3 2.4 0.9 2.8 1.1 2.8 1.4 6.3 17.0 24.5 4.0 1.8 8.0 17.8 2.9 11.5 5.1 7.4 9.3 1.3 0.1 0.0
Leukaemia
126.3
64.8 61.5
5.4 1.2 2.3 0.9 6.4 0.8 1.8 1.1 6.6 25.1 6.9 5.5 1.2 7.3 16.4 3.1 7.7 5.9 8.6 10.5 1.5 0.2 0.0
NonHodgkin lymphoma
118.5
58.1 60.4
1.1 0.4 2.2 0.9 1.1 1.0 1.5 2.0 5.0 10.9 7.9 2.8 1.1 4.8 27.4 4.0 19.7 3.1 10.6 10.3 0.6 0.0 0.0
Larynx
91.5
62.5 29.0
1.0 0.4 0.8 0.4 0.9 0.4 1.2 2.0 3.8 17.9 6.9 4.4 0.9 2.5 6.4 1.3 15.4 5.2 6.0 12.6 1.0 0.0 0.0
Kidney, etc.
The global burden of cancer
All sites 85.5 28.3 45.9 29.2 65.1 26.6 68.3 64.4 166.5 571.5 528.1 162.7 59.5 192.4 488.4 57.1 405.6 185.6 197.2 324.1 34.2 2.9 0.6 1881.0 1908.8 3789.8
FEMALE
1. Eastern Africa 2. Middle Africa 3. Northern Africa 4. Southern Africa 5. Western Africa 6. Caribbean 7. Central America 8. South America ŽTemperate. 9. South America ŽTropical. 10. North America 11. Eastern Asia: China 12. Eastern Asia: Japan 13. Eastern Asia: Other 14. South-eastern Asia 15. South-central Asia 16. Western Asia 17. Eastern Europe 18. Northern Europe 19. Southern Europe 20. Western Europe 21. AustraliarNew Zealand 22. Melanesia 23. MicronesiarPolynesia
Developed countries Developing countries
ALL AREAS
Table 1. Ž Continued.
224 795.6
471.5 324.1
10.5 3.1 12.1 4.9 10.0 5.0 9.7 18.9 35.0 167.3 61.4 24.1 6.2 38.3 95.8 12.9 77.2 48.4 50.9 94.7 9.0 0.4 0.1
Breast
381.0
259.2 121.7
2.2 0.7 2.0 1.2 2.0 2.4 2.8 7.3 11.6 77.4 50.4 25.6 4.6 13.9 16.7 3.8 48.4 24.4 25.2 53.0 5.1 0.1 0.0
Colonr rectum
371.2
83.3 287.9
21.5 5.7 5.2 6.5 13.6 5.0 17.7 7.0 29.9 15.2 24.7 8.5 9.3 30.9 107.0 3.0 27.5 7.6 9.9 13.2 1.3 0.7 0.1
Cervix uteri
287.2
124.2 163.0
2.8 3.2 1.2 0.6 3.3 1.2 4.6 2.6 13.6 9.5 94.4 34.0 7.5 7.8 16.7 3.3 44.1 7.8 13.0 15.2 0.7 0.1 0.0
Stomach
265.1
151.8 113.3
1.0 0.2 1.1 1.2 0.5 1.5 2.7 2.2 6.2 70.7 65.5 12.2 3.8 14.0 10.8 2.6 26.6 17.9 8.8 13.2 2.3 0.0 0.0
Lung
165.5
86.7 78.8
4.6 0.7 1.4 0.8 3.7 1.0 2.9 2.5 6.2 22.0 14.5 5.4 1.6 11.0 24.4 3.3 25.3 9.3 8.5 14.9 1.2 0.1 0.0
Ovary, etc.
142.4
95.9 46.5
1.6 0.7 0.8 1.1 1.0 1.0 2.7 3.6 6.6 31.0 8.2 3.6 0.8 5.8 9.3 3.2 23.2 8.3 12.0 16.6 1.3 0.1 0.0
Corpus uteri
121.1
27.0 94.1
3.1 3.4 1.0 1.0 3.7 0.7 1.4 0.4 1.9 3.0 57.5 7.6 3.0 9.0 6.3 1.5 7.7 1.3 4.5 2.8 0.1 0.2 0.0
Liver
103.2
13.6 89.5
2.7 0.0 0.7 1.7 0.5 0.3 0.5 1.4 1.6 3.2 47.8 1.8 0.4 1.9 28.8 1.0 2.7 3.1 0.6 1.9 0.4 0.0 0.0
Oesophagus
100.9
45.1 55.8
1.4 0.3 1.6 0.6 2.2 0.9 2.5 1.2 5.1 12.5 18.6 2.8 1.2 6.6 11.7 1.9 10.8 4.2 5.7 8.2 0.9 0.1 0.0
Leukaemia
94.6
52.8 41.8
3.0 1.9 1.7 0.6 4.3 0.6 1.3 1.0 4.5 21.3 4.9 3.8 0.8 5.5 9.2 2.4 5.6 5.3 6.4 9.3 1.2 0.1 0.0
NonHodgkin lymphoma
78.9
53.6 25.3
0.8 0.9 0.2 0.3 0.8 0.5 2.3 1.5 3.3 15.3 6.2 6.1 1.0 1.9 4.3 1.1 13.2 5.9 4.6 7.9 0.8 0.0 0.0
Pancreas
D. M. Parkin
All sites 178.0 160.4 105.6 247.4 140.0 199.0 154.7 255.1 196.9 369.9 179.2 270.9 235.7 130.4 106.3 131.4 269.4 270.0 256.0 294.8 312.7 165.4 204.5 299.6 152.4 203.5
MALE
1. Eastern Africa 2. Middle Africa 3. Northern Africa 4. Southern Africa 5. Western Africa 6. Caribbean 7. Central America 8. South America ŽTemperate. 9. South America ŽTropical. 10. North America 11. Eastern Asia: China 12. Eastern Asia: Japan 13. Eastern Asia: other 14. South-eEastern Asia 15. South-central Asia 16. Western Asia 17. Eastern Europe 18. Northern Europe 19. Southern Europe 20. Western Europe 21. AustraliarNew Zealand 22. Melanesia 23. MicronesiarPolynesia
Developed countries Developing countries
225
ALL AREAS 37.5
62.6 24.1
4.9 6.5 12.8 29.1 2.2 29.8 19.3 55.1 24.1 69.6 34.7 38.9 40.4 29.7 11.9 30.5 75.9 59.1 58.8 54.1 47.6 7.8 52.5
Lung
24.5
27.1 23.1
9.0 20.0 5.9 11.5 12.4 16.2 18.6 23.8 31.1 8.4 43.6 77.9 54.3 9.0 6.7 9.4 36.3 16.9 20.8 16.4 10.8 6.5 12.3
Stomach
19.4
35.9 10.2
8.1 2.3 6.0 11.2 4.7 16.0 8.8 27.2 15.0 44.3 13.3 39.5 21.3 11.9 5.0 8.8 25.3 34.4 28.8 39.8 45.8 11.1 17.6
Colonr rectum
19.8
40.1 7.6
16.8 29.6 5.1 31.0 23.9 42.4 24.8 22.9 28.1 92.4 1.1 8.5 5.4 5.9 4.5 7.1 14.1 34.7 16.9 39.6 49.7 6.0 21.2
Prostate
14.7
7.6 17.9
18.5 28.4 4.7 20.1 22.1 8.0 5.4 2.7 3.2 3.2 35.8 27.6 32.2 17.6 2.8 5.0 5.2 2.7 8.9 4.9 2.8 20.1 10.9
Liver
10.2
6.4 12.2
12.5 2.1 2.8 32.6 2.1 7.5 3.4 11.1 7.7 5.2 21.6 9.5 7.3 3.2 9.0 2.4 6.4 6.7 4.9 6.9 4.6 3.6 4.5
Oesophagus
Table 2. Estimated age standardized rates of cancer incidence by site, sex and area, 1990
9.9
17.6 5.6
5.8 2.8 23.3 11.7 8.0 7.9 5.1 13.0 9.5 23.8 3.3 8.9 8.8 4.3 4.4 10.9 12.6 19.9 22.0 16.6 14.5 2.7 6.7
Bladder
6.6
8.4 5.8
7.6 5.3 4.7 16.0 6.4 8.1 4.1 6.9 8.5 7.8 1.1 3.1 3.5 3.8 11.4 3.8 9.3 5.0 10.2 11.3 15.1 36.9 6.0
Oral cavity
5.6
8.6 4.1
3.1 1.3 3.9 5.4 4.9 7.3 5.9 6.0 6.1 10.5 4.5 5.7 4.9 4.0 3.1 4.5 7.6 8.4 8.6 8.5 10.8 4.0 7.5
Leukaemia
5.6
9.6 3.7
8.8 5.1 4.3 5.8 11.0 5.2 4.7 4.6 7.2 15.6 1.3 7.0 4.0 4.6 3.4 5.2 4.9 9.6 9.7 9.2 12.9 9.0 5.4
NonHodgkin lymphoma
5.7
8.5 4.4
2.5 2.4 5.6 6.9 2.6 7.4 5.1 8.6 6.5 6.8 1.7 3.3 4.9 3.8 6.8 8.4 12.0 4.9 11.3 8.9 5.2 3.6 6.5
Larynx
4.3
9.1 2.0
1.6 1.5 1.8 2.7 1.6 3.1 3.9 8.4 4.7 11.0 1.5 5.4 3.8 1.6 1.4 2.6 9.5 8.1 6.4 10.6 8.1 1.3 3.4
Kidney, etc.
The global burden of cancer
All sites 145.3 124.9 94.6 187.8 118.5 175.5 180.1 230.1 185.0 277.5 105.3 166.8 179.9 115.7 109.9 108.3 172.8 234.5 177.6 210.4 254.0 185.6 205.2 208.9 122.5 154.6
FEMALE
1. Eastern Africa 2. Middle Africa 3. Northern Africa 4. Southern Africa 5. Western Africa 6. Caribbean 7. Central America 8. South America ŽTemperate. 9. South America ŽTropical. 10. North America 11. Eastern Asia: China 12. Eastern Asia: Japan 13. Eastern Asia: other 14. South-eastern Asia 15. South-central Asia 16. Western Asia 17. Eastern Europe 18. Northern Europe 19. Southern Europe 20. Western Europe 21. AustraliarNew Zealand 22. Melanesia 23. MicronesiarPolynesia
Developed countries Developing countries
ALL AREAS
Table 2. Ž Continued.
226 33.0
56.4 20.4
18.6 13.6 25.0 31.5 19.0 33.5 25.5 69.1 39.1 86.3 11.8 28.6 17.8 22.5 21.2 24.3 36.0 68.3 49.5 67.3 71.7 23.9 51.7
Breast
15.3
25.4 8.1
4.2 3.4 4.2 8.4 3.9 15.5 7.9 24.4 13.6 32.8 10.2 24.6 14.1 8.9 3.8 7.6 18.5 26.1 20.2 29.0 34.8 5.3 11.3
Colonr rectum
15.5
11.2 18.2
37.4 26.6 11.3 40.4 26.2 33.2 44.4 27.7 31.8 9.1 5.0 9.7 26.6 18.6 23.8 5.5 13.7 12.5 10.4 10.9 11.2 43.4 26.0
Cervix uteri
11.6
12.6 11.0
5.3 14.8 2.6 4.3 6.6 7.9 13.3 8.7 15.9 4.0 19.0 33.3 22.8 5.1 4.0 6.6 16.9 8.1 10.0 8.2 4.9 4.8 7.8
Stomach
10.8
15.5 7.8
2.0 0.8 2.6 7.7 0.9 10.1 7.9 7.6 7.2 32.9 13.4 11.2 11.6 9.3 2.6 5.2 10.3 20.2 7.3 8.2 16.1 3.7 17.3
Lung
6.8
10.3 4.9
7.8 2.9 2.8 5.3 6.5 6.3 7.6 9.1 6.9 11.2 2.7 6.3 4.6 6.2 5.4 6.3 11.5 12.7 8.0 10.0 9.6 7.7 8.4
Ovary, etc.
5.9
10.7 3.1
3.0 3.0 1.6 7.4 2.1 7.0 7.7 12.6 7.9 15.0 1.7 3.9 2.3 3.6 2.2 6.5 9.9 10.8 10.6 11.1 9.1 7.4 12.9
Corpus uteri
4.9
2.6 6.2
5.7 13.6 2.4 6.6 6.7 4.4 4.0 1.5 2.2 1.4 11.5 6.9 9.5 5.8 1.5 3.1 2.8 1.4 3.4 1.5 1.1 9.9 4.5
Liver
4.2
1.3 6.2
5.3 0.2 1.7 11.9 1.2 2.1 1.4 4.5 2.0 1.4 9.9 1.6 1.1 1.3 7.0 2.0 1.0 3.1 0.5 1.1 2.4 2.6 1.5
Oesophagus
4.0
5.5 3.0
1.9 1.0 2.7 3.5 3.8 5.6 5.0 4.4 4.7 6.3 3.5 3.7 3.4 3.2 2.2 2.9 5.0 5.7 5.8 5.6 6.5 3.3 5.6
Leukaemia
3.8
5.8 2.5
3.9 8.4 3.4 4.0 5.9 3.6 3.2 3.4 4.9 10.1 0.9 4.0 2.3 3.1 2.0 4.1 2.5 6.6 5.9 5.8 8.5 4.8 4.0
NonHodgkin lymphoma
3.1
5.0 1.7
1.7 4.1 0.6 1.9 1.6 3.5 6.9 4.9 4.0 6.4 1.2 5.4 3.0 1.3 1.1 2.3 4.8 6.0 3.5 4.2 4.9 0.5 2.9
Pancreas
D. M. Parkin
The global burden of cancer
Figure 4. The estimated total of new cancer cases Žin millions. and estimated world population Žin billions. in 1975, 1980, 1985, and 1990 Žshowing divisions between developing and developed countries..
South America Žespecially the Andean countries., and in eastern Europe ŽFigure 6.. Moderately high rates are observed in some countries of central and west Africa. Low rates are found in east and north Africa, North America, and south and south-east Asia ŽASR in men 5.9]9.0 per 100 000, and 2.6]5.3 in women.. There is clearly a strong environmental component to the risk differences. Migrant populations from high risk parts of the world show a marked diminution in risk when they move to a lower risk area, although this is quite gradual, and seems to depend
on the age at migration. The data fit with observations concerning the importance of childhood environment in determining risk.11 Most research has focused on the contribution of dietary factors to the risk Žas well as the adverse effects of tobacco smoking and alcohol.. Infection with the bacterium Helicobacter pylori appears also to be important, although the exact relationship is still the subject of debate Žsee discussion by Nyren in this issue.. In 1994, IARC 12 accepted H. pylori as carcinogenic to humans, based on ecological correlation studies Žsuch as the EURO-
Figure 5. Incidence of lung cancer: ASR Žworld. }male Žall ages..
227
D. M. Parkin
Figure 6. Incidence of stomach cancer: ASR Žworld. }male Žall ages..
GAST study 13 ., a host of case-control studies and, more usefully, several cohort studies. The combined odds ratio from these studies is 2.1.14 Assuming an average prevalence of infection of 50% in developed countries, and 80% in developing countries, the number of new cases of stomach cancer attributed to
the bacterium would be 337,800, representing 42% of the world total these cancers Ž47% in developing countries and 35% elsewhere.. A decline in gastric cancer incidence and mortality is evident in most countries. This presumably relates to improvements in preservation and storage of
Figure 7. Incidence of breast cancer: ASR ŽWorld. ŽAll Ages..
228
The global burden of cancer
foods.15 It may also reflect changes in the prevalence of H. pylori by birth cohort, perhaps as a result of reduced transmission in childhood, following a trend to improved hygiene and reduction of crowding.16
Asia, the world total in 2010 would be 1.45 million, an 82% increase on the figure for 1990.
Breast cancer
Cancers of the colon]rectum accounted for 782,000 new cases in 1990 Ž9.7% of the world total.. Numbers were similar in males and females Žratio 1.05:1.. The geographic distribution follows the division between Westernised versus developing countries. The highest rates are in N. America, AustraliarNew Zealand and Europe Žespecially GermanyrAustria., with moderately high incidence rates in southern South America. Rates in Africa, except for South Africa, are very low. The geographical distribution of colonic cancer and rectal cancer is similar, although the variation between countries is less for the rectum than for colon. Thus, in high-risk populations, the ratio colonrrectum is 2:1 or more Žrather more in females.; in low-risk countries, rates are similar Žwith even slight excess of rectal cancer in India.. These large geographic differences probably largely represent the effects of different environmental exposures, presumably mainly dietary. There are strong international correlations between risk of large bowel cancers and per capita consumption patterns of meat,18 fat Žspecifically animal fat.19 and fibre.20 That the risk of colon cancer is quite labile to environmental change is evident from the study of migrants. In Hawaii, first-generation Japanese migrants acquire incidence rates similar to, or even in excess of, the indigenous white population.21 Incidence rates have risen}especially in men}in most areas ŽN. America is an exception. since 1985, so that the estimated number of cases in 1990 is 15.5% greater Ž21% in men, 10% in women. than in 1985.
Colorectal cancer
Brease cancer is by far the most frequent cancer of women Ž21% of all cancers. and ranks third overall when both sexes are considered together. It is the most common cancer of women in all the ‘developed’ areas Žexcept for Japan, where it is third after stomach and colon]rectal cancer., and indeed on a world map ŽFigure 7., the pattern strongly resembles that of indicators of ‘affluence’. The incidence is more modest in north Africa, South America, east, south-east and western Asia, but it is still the most common cancer of women in these geographic regions. The age-standardized incidence is highest in the United States Ž87.1 per 100,000. and lowest in China Ž11.8.. Prevalence of carriers of the major susceptibility genes ŽBRCA1 and BRCA2. in the general population is low, and their differential distribution around the world can hardly account for much international or inter-ethnic variation in risk. Most must therefore be a consequence of different environmental exposures, and indeed there are quite marked changes in risk following migration, particularly if this takes place at young ages. The major influences on breast cancer risk appear to be certain reproductive factors, and, less certainly, diet. There have however, been few attempts to quantify the magnitude of risk differentials between populations which might be explained by such factors. In the USA, Brinton et al 17 calculated that the difference in incidence between whites and blacks, at least in women aged 40]54 Ž20%. was entirely explicable in terms of the different prevalence of certain reproductive and ‘lifestyle’ variables. Incidence rates of breast cancer are increasing in most countries, and the changes are usually greatest where rates were previously low. Since our previous estimates for 1985, there has been an overall increase in incidence rates of approximately 0.5% annually Žthis excludes China, where data sources were changed between estimates.. At this rate of growth, there would be approximately 1.35 million new cases in 2010. However, cancer registries in China are recording annual increases in incidence of over 5% and in those elsewhere in eastern Asia, increases are not much less. Assuming a modest 3% growth in east
Liver cancer Liver cancer is currently the fifth most common cancer in the world Ž437,000 new cases annually.. It is primarily a problem of developing countries where over 80% of cases occur. There are high incidence rates in west and central Africa Žwhere it accounts for almost one-quarter of cancer in men., eastern and south-eastern Asia, and in Melanesia. The incidence is low in developed areas}only in southern Europe Žand especially in Greece. is there any substantial risk. The overall sex ratio is approximately 2.6, rather greater in the high-risk areas and less in low-risk 229
D. M. Parkin
areas. The major risk factors for liver cancer are chronic infection with hepatitis viruses ŽHBV and HBC.. Both pose a substantial increase in risk Žmore than 20-fold 22 .. With information on prevalence of infection in different areas, the percentage of liver cancers attributable to these viruses can be estimated; it is 77% of cases worldwide, and 85% of cases in the developing countries.14
ethnic groups ŽWhite, Black, Asian. within the United States. Even before PSA screening, international differences reflected diagnostic bias to some extent. The prevalence of latent prostate cancer at autopsy shows much less variation than clinical prostate cancer, though the ethnic-specific ranks are much the same as for incidence.23 The frequency of latent carcinoma of prostate in Japan is increasing Žas is clinical prostate cancer. and approaching the prevalence for US Whites. This suggests that promotion of latent carcinoma may be important in explaining international differences. The differences in ethnic-specific risk are probably mediated via population differences in alleles of genes coding for enzymes involved in testosterone metabolism.24 The estimated number of cases is substantially higher Ž36%. than 5 years earlier; most of this increase is the result of the huge increase in incidence rates in North America from an age-standardised rate of 61.3 per 100,000 in 1985 to 92.4 per 100,000 in 1990 Žan annual increase of 8.5%..
Prostate cancer Prostate cancer is the fourth most frequent cancer of men overall Ž9.2% of all new cancer cases.. Incidence rates are now influenced by the diagnosis of latent cancers by screening asymptomatic individuals, so that where this practice is common, the ‘incidence’ may be very high Ž95.1 per 100,000 in the United States, for example, where it is now by far the most commonly diagnosed cancer in men.. Incidence is high also in northern and western Europe, and AustraliarNew Zealand. The age-adjusted rates show that several developing areas have relatively high incidence rates also}sub-Saharan Africa, Latin America and the Caribbean in particular. By contrast, the incidence rates in Asia and particularly in China are low. Even when comparison is made after age standardization, the difference in risk between China and North America is more than 80-fold. Similar, albeit less extreme, patterns are observed between different
Cervix cancer Cervix cancer is seventh in frequency overall, third in frequency in women, in whom it comprises 9.8% of all cancers Ž371,200 new cases per year.. The geographic pattern is a complete contrast to breast cancer ŽFigure 8.; the highest incidence is observed in
Figure 8. Incidence of cervix uteri cancer: ASR ŽWorld. ŽAll Ages..
230
The global burden of cancer
Latin America and the Caribbean, sub-Saharan Africa, and south and south-east Asia. In developed countries, the incidence rates are generally low, with agestandardized rates less than 14 per 100,000. It is quite clear that the major aetiological agents are oncogenic subtypes of the human papilloma virus ŽHPV.,25 indeed, it may be that the disease does not occur in the absence of infection. It is curious, therefore, that there is no systematic information on geographicrethnic variations in prevalence of persistent HPV infection with which to compare incidence data, although case control studies do suggest that it is greater in control subjects from higher risk countries Žapprox. 10]20%. than in areas of lower risk Ž5]10%.. The estimated number of cases for 1990 is some 15% lower than in 1985. This is consistent with the decline in incidence of cervix cancer observed in many countries. Although this is evident where there are well established screening programmes, there are also declines observed in several developing countries. This is especially marked in China where the estimated age-standardised incidence rate in 1990
was 5.0, compared with 17.8 in 1985. Although some of this difference reflects changing data sources, cancer registry results also indicate a fairly dramatic decline in rates in recent years. As a result of these trends, cervix cancer has ceded its place as the leading cancer in developing countries to breast cancer, and only in sub-Saharan Africa, Central America, south-central Asia and Melanesia is it now the main cancer of women. Oesophageal cancer Oesophageal cancer is the eighth most common cancer worldwide, responsible for 316,000 new cases in 1990 Ž3.9% of the total.. It is mainly a tumour of developing countries where it accounts for 7.7% of cancers in men, and 4.7% in women. China alone accounts for 47.5% of cases. Geographic variation in incidence is very striking. Incidence rates vary at least 40-fold between registries published in Cancer Incidence in Five Continents, Vol. VII ŽFigure 9.. In men, high risk areas include
Figure 9. Word Standardised Rate w0]85 q x, per 100,000.
231
D. M. Parkin
central Asia and China, east and South Africa, parts of South America and the Caribbean, and, in western Europe, France and Switzerland. The pattern is much the same in women, except in Europe, where high rates are found in Ireland and Scotland, but not in France and Switzerland. The Indian subcontinent is also an area of moderate risk in women. There are also very marked variations in incidence within countries in the high-risk areas. These must represent exposures to important carcinogens, but it seems that these are quite different in the various high-risk areas. Tobacco and alcohol are the main agents involved in Europe and North America, where over 90% of cases can be attributed to these causes. Chewing of tobacco Žand betel. is important in the Indian subcontinent. Hot beverages have been shown to increase risk, and drinking hot mate ´ is probably responsible for raised rates in Uruguay, southern Brazil and northern Argentina. Nutritional deficiencies Žspecifically of micronutrients. are thought to underlie the high risk in central Asia, China, and southern Africa. Here other factors such as pickled vegetables, nitrosamine-rich foods and mycotoxins may also be involved, as well as consumption of opium residues Žin Iran. or pipe stem residues Žin the Transkei of southern Africa. 26 .
America and AustraliarNew Zealand. We have not attempted to differentiate between the different leukaemia subtypes, though clearly the geographic patterns will not be the same. For example, chronic lymphocytic leukaemia, numerically the most important diagnosis in western Europe, is extremely rare in east and south-east Asia. Lymphomas Of the lymphomas, Non-Hodgkin lymphoma is by far the commonest, with 221,000 new cases annually, compared with 59,600 cases of Hodgkin’s disease and 56,600 cases of Myeloma. For the non-Hodgkin lymphomas, the highest incidence is observed in the developed areas of N. America, AustraliarNew Zealand and Europe. Currently the highest rates are observed in populations which are most affected by the epidemic of AIDS Žmales in San Francisco.. High incidence rates are also observed in the central part of Africa, in part the consequence of very high rates in the childhood age-range, due to Burkitt’s lymphoma. The lowest risk is in eastern and southern Asia. Incidence rates of NHL have increased in practically all populations since the 1970s and probably before this too; the environmental exposures Žor modified host responses to them. which underlies this are a mystery}the HIV epidemic can account for only part of it, in certain geographic areas.27 Hodgkin’s disease has a rather similar geographic distribution. Incidence is especially low in the eastern Asian areas, and the Pacific islands. For Myeloma, the highest rates are, again, observed in developed countries, but a few populations of African origin also display moderately elevated rates}for example: in males the Caribbean and southern Africa, in females Middle Africa.
Bladder cancer Bladder cancer, the seventh most frequent cancer of men, is considerably less common in women, in whom it appears only 14th in rank Žsex-ratio worldwide is approx. 3.5:1.. In general, the geography is similar to that of lung cancer, with the highest rates in developed areas Žwith the exception of Japan.. In fact, we estimate that 36% of bladder cancer cases in men and 17% in women might be attributable to tobacco smoking. However, there are areas of high risk in north Africa Žestd. ASR in Egypt 46.6. western Asia Žestd. ASR in Iraq 16.5. and some African countries wZimbabwe Ž16.0. and Zambia Ž15.0.x. Here, the high rates result from endemic urinary schistosomiasis, known to be associated with an increased risk of squamous cell carcinomas.12
Mouth and pharynx In previous estimates, all cancers of the mouth and pharynx were considered together as a single group. In the current estimates, we have divided this group into three categories}cancers of the oral cavity, nasopharynx and other pharynx. Of these, cancers of the oral cavity are the most frequent, with 212,000 new cases worldwide Ž2.6% of the total., followed by other pharyngeal cancers with 94,000 cases Ž1.2%. and cancers of the nasopharynx with 57,500 cases Ž0.7%.. All of these are much more common in males than females, the sex ratio ŽM:F. is 2.0 for mouth
Leukaemia Leukaemia is the 10th most frequent cancer, with an estimated 231,000 new cases. There is relatively little geographic variation, although rates appear to be low in sub-Saharan Africa Žpossibly related in part to under-recording .. The incidence is highest in North 232
The global burden of cancer
cancer 4.4 for pharynx cancer and 2.2 for nasopharynx cancer. There are some similarities to the geographic patterns for cancers of the mouth, and ‘other pharynx’. In men, both are high in western and southern Europe, and south Asia, while mouth cancers Žbut not pharynx. have high rates in Melanesia, southern Africa, and AustraliarNew Zealand. In females, both types of cancer have relatively high incidence rates in southern Asia and Žmouth cancers only. Melanesia and AustraliarNew Zealand. These patterns reflect prevalence of specific risk factors}tobaccoralcohol in western and southern Europe, and southern Africa, and the chewing of betel quid in south-central Asia and Melanesia. The high rate of oral cancer in Australia is due to lip cancer Žrelated to solar irradiation.. Although cancer of the nasopharynx is caused by infection by the Epstein]Barr virus,28 this ubiquitous agent cannot explain the tumour’s very singular geographic distribution. Although certain traditional dietary items are clearly implicated in the increased risk,29 it seems more likely that genetic predisposition is the explanation. Thus the incidence is highest in peoples from southern China ŽGuangdong, Guanxi, Hunan and Fujian.,30 migrants from these areas Že.g. in Singapore., or in areas where they have settled, for example, south-east Asia. Forty-four percent of the world cases occur in China and 23% in south-east Asia. A linkage study based on affected sib pairs of Chinese origin suggested that a gene Žor genes. closely linked to the HLA locus is associated with a marked elevation Ž20-fold. in the risk of nasopharynx cancer.31 There is also a high risk in Inuit populations in the Arctic Žpossibly due to common ancestry.. In the medium risk area of north Africa, the epidemiological profile is a little different, with a small peak in incidence in early adolescence.
developed countries. Ovarian cancer occurs in almost all with age-standardised rates in excess of seven per 100,000}the exceptions are in southern Europe ŽSpain, Portugal, Greece. and Japan. Incidence in temperate South America is moderately high Žaverage 9.1 per 100,000, Table 2.. The high risk areas for cancer of the Corpus uteri are North America ŽASR 15.0., central Europe Ž13.8., Scandinavia Ž13.6., and, temperate South America Ž12.6.. Corpus cancer is rare in Asia}including Japan}and in sub-Saharan Africa, with age standardised rates inferior to four per 100,000.. Kidney cancer Kidney cancer Ž150,000 new cases, 1.9% of the world total. has the highest rates in North America and western, northern and eastern Europe while incidence rates are low in Africa, Asia Žexcept Japanese males. and the Pacific. Larynx cancer Larynx cancer Ž136,000 new cases. is predominantly a cancer of men, in whom it comprises 2.8% of cases. The sex ratio Žalmost 7:1. is greater than for any other site and Figure 3 emphasises the rarity of larynx cancer in women, particularly in developed countries. For men, the high-risk areas are Europe Žeast, south, west., temperate South America and western Asia. In western Asia, larynx cancer accounts for more than 6% of cancers in men. Tobacco smoking is estimated to cause two-thirds of all cases in men. Brain and nervous system cancers Cancers of the brain and nervous system account for some 127,000 new cases annually Ž1.6% of new cancers.. The highest rates are observed in developed areas ŽAustraliarNew Zealand, North America, northern Europe. and lowest in Africa and the Pacific islands. This suggests that the availability of diagnostic facilities may well be important in determining geographic patterns, at least in part. The incidence is probably underestimated in many developing countries.
Pancreas cancer Pancreas cancer Ž170,000 cases per year, 2.1% of the total. is slightly more common in men than in women Žratio 1.2:1.. Geographic variation is not large; the higher rates in most developed Areas probably reflect diagnostic capacity rather than aetiology. Ovarian cancer Ovarian cancer Ž165,000 cases, 2.0% of all cancers, 4.4% of cancers in women. and cancer of the corpus uteri Ž142,000 cases, 1.8% of all cancer, 3.8% of cancer in women. are, in general, tumours of women in
Malignant melanoma Malignant melanoma of skin accounts for 105,000 new cases annually, with slightly more occurring in 233
D. M. Parkin
women than in men Žsex-ratio 0.9..The highest rates are observed in AustraliarNew Zealand ŽASR 27.9 per 100,000 in men, 25.0 in women., North America, and northern Europe Žespecially Scandinavia. }these regions have seen annual increases of between 1.5 and 4.5% in incidence Žage-adjusted . of melanoma since 1895. The lowest rates ŽASR less than 0.6 per 100,000. are observed in east and south-east Asia, and in south Asia.
tions of identified causes Ženvironmental and genetic. to the observed patterns. In this way, epidemiology contributes directly to public health, in guiding the selection of priorities in prevention of disease.
References 1. Parkin DM, Stjernsward ¨ J, Muir CS Ž1984. Estimates of the worldwide frequency of twelve major cancers. Bull WHO 62:163]182 2. Parkin DM, Laara ¨¨ ¨ E, Muir CS Ž1988. Estimates of the worldwide frequency of sixteen major cancers in 1980. Int J Cancer 41:184]197 3. Parkin DM, Pisani P, Ferlay J Ž1993. Estimates of the worldwide incidence of eighteen major cancers in 1985. Int J Cancer 54:594]606 4. Parkin DM, Pisani P, Ferlay J Ž1998a. Estimates of the worldwide incidence of twenty-five major cancers in 1990. Int J Cancer, in press 5. Ferlay J, Parkin DM, Pisani P Ž1998. GLOBOCAN: Cancer Incidence and Mortality Worldwide, IARC CancerBase No. 3, in press 6. World Health Organization Ž1993. Cause of Death Statistics and Vital Rates, Civil Registration Systems and Alternative Sources of Information, World Health Statistics Annual 1993, WHO, Geneva 7. Doll R Peto R Ž1981. The Causes of Cancer, Oxford University Press, Oxford 8. Parkin DM, Pisani P, Lopez AD, Masuyer E Ž1994. At least one in seven cases of cancer is cause by smoking. Global estimates for 1985. Int J Cancer 59:494]504 9. Kubik AK, Parkin DM, Plesko I, Zatonski W, Kramarova ´ ´ E, Mohner M, Friedl HP, Juhasz L, Tzvetansky CG, Reissigova J ¨ Ž1995. Patterns of cigarette sales and lung cancer mortality in some Central and Eastern European Countries, 1960]1989. Cancer 75:2452]2460 10. Lopez AG, Pollan M, de-la Iglesia P, Ruiz M Ž1995. Characterization of the lung cancer epidemic in the European Union Ž1970]1990.. Cancer Epidemiol Biomarkers Prev 4:813]820 11. Coggon D, Osmond C, Barker DJ Ž1990. Stomach cancer and migration within England and Wales. Br J Cancer 61:573]574 12. IARC Monograph on the Evaluation of Carcinogenic Risks to Humans, vol. 61 Ž1994. Schistosomes, Liver Flukes and Helicobacter pylori, International Agency for Research on Cancer, Lyon, France 13. Eurogast Study Group Ž1993. An international association between Helicobacter pylori infection and gastric cancer. Lancet 341:1329]1362 14. Parkin DM, Pisani P, Munoz ˜ N, Ferlay J Ž1998b. The global health burden of infection associated cancers, in: Infections and Human Cancer, vol 33 ŽWeiss RA, Beral V, Newton R, eds. Žin press.. Cancer Surveys 15. Howson CP, Hiyama T, Wynder EL Ž1986. The decline in gastric cancer: epidemiology of an unplanned triumph. Epidemiol Rev 8:1]27 16. Banatvala N, Mayo K, Megraud F, Jennings R, Deeks JJ, Feldman RA Ž1993. The cohort effect and Helicobacter pylori. J Infect Dis 168:219]221 17. Brinton LA, Benichou J, Gammon MD, Brogan DR, Coates R, Schoenberg JB Ž1997. Ethnicity and variation in breast cancer incidence. Int J Cancer 73Ž3.:349]355 18. Armstrong B, Doll R Ž1975. Environmental factors and cancer incidence and mortality in different countries, with special reference to dietary practices. Int J Cancer 15:617]631
Cancer of the thyroid Thyroid cancer Ž87,000 new cases. is the only cancer considered here for which the incidence is much higher in females than in males wsex ratio ŽM:F. is 0.33x. It comprises 1.7% of cancers in women. The highest estimated incidence is in MicronesiarPolynesia Žan area well known for a high frequency of these cancers 32 . and in Central America. Testis cancer Testis cancer is relatively rare}36,000 new cases annually or 0.8% of cancers in men. The highest rates are observed in west and north Europe ŽASR in Denmarks 9.5, in Germanys 8.4.. The highest risk is in the age-group 15}44, so that testis cancer is the most common cancer of men in developed countries in this age-group Ž11.5% of new cases, compared with 9.4% due to lung cancer..
Conclusions Completing the global overview of cancer, in terms of its overall impact, and how the risk of different cancers varies between populations, has several objectives. Traditionally, for the epidemiologist, variation in risk between populations and over time has been a useful starting point for investigating the causes of cancer. Almost always, these were thought of in terms of environmental Žincluding lifestyle. exposures, and the failure to identify adequate explanations was a source of frustration, whilst more recently, the importance of genetic variation as a cause of interpopulation differences in risk has been increasingly recognised. In addition, knowledge of the cancer burden in different countriesrareas, together with information on past trends and likely future evolution, are essential elements in planning and the evaluation of cancer control programmes. In this context it is particularly valuable to estimate the contribu234
The global burden of cancer
26. Munoz ˜ N, Day N Ž1996. Esophageal cancer, in Cancer Epidemiology and Prevention Ž2nd ed.. ŽSchottenfeld D, Fraumeni JF, eds. Oxford University Press, New York, USA 27. Hartge P, Devesa SS Ž1992. Quantification of the impact of known risk factors on time trends in non-Hodgkin’s lymphoma incidence. Cancer Res 52:5566s]5569s 28. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, vol. 70 Ž1997. Epstein-Barr Virus and Kaposi’s Sarcoma Herpes virusrHuman Herpes virus 8, International Agency for Research on Cancer, Lyon, France 29. Hildesheim A, Levine PH Ž1993. Etiology of nasopharyngeal carcinoma: a review. Epidemiol Rev 15:466]485 30. Li J, Liu B, Li G, Rong S, Cao D, 25 others Ž1979. Atlas of Cancer Mortality in the People’s Republic of China, China Map Press, Shanghai 31. Lu SJ, Day NE, Degos L, Lepage V, Wang PC, Chan SH, Simons M, McKnight B, Easton D, Zeng Y et al Ž1990. Linkage of a nasopharyngeal carcinoma susceptibility locus to the HLA region. Nature, 346:470]471 32. Paksoy N Ž1992. Frequency of thyroid cancer in pacific populations. J Natl Cancer Inst 84:1124]1125
19. Prentice RL, Sheppard L Ž1990. Dietary fat and cancer: consistency of the epidemiologic data, and disease prevention that may follow from a practical reduction in fat consumption. Cancer Causes Control 1:81]87 20. McKeown-Eyssen G Ž1994. Epidemiology of colorectal cancer revisited: are serum triglycerides andror plasma glucose associated with risk? Cancer Epidemiol Biomarkers Prev 3:687]695 21. Kolonel LN, Hirohata T, Chappell BV, Viola FV, Harris DE Ž1980. Cancer mortality in a cohort of naval shipyard workers in Hawaii: early findings. J Natl Cancer Inst 64:739]743 22. Donato F, Boffetta P, Puoti MA Ž1998. meta-analysis of epidemiological studies on the combined effect of Hepatitis B and C virus infections in causing hepatocellular carcinoma. Int J Cancer 75Ž3.:347]354 23. Yatani R, Chigusa I, Akazaki K, Stemmerman GN, Welsh RA, Correa P Ž1982. Geographic pathology of latent prostate carcinoma. Int J Cancer 29:611]616 24. Shibata A, Whittemore A Ž1997. Genetic predisposition to prostate cancer: possible explanations for ethnic differences in risk. The Prostate 32:65]72 25. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, vol. 64 Ž1995. Human Papillomavirus, International Agency for Research on Cancer, Lyon, France
235
Article No. sc980080
The global burden of cancer D. M. Parkin
The number of new cancer cases, and the annual incidence rates, of 25 different cancers have been estimated for the year 1990 for every country of the world. The distributions of the most common cancers in men and women are presented for 23 broad ‘ Areas’. The total number (excluding nonmelanoma skin cancer) was 8.1 million, just over half of which occur in the developing countries. The most common cancer is lung cancer, which accounts for 18% of cancers of men worldwide. Stomach cancer is second in frequency (almost 10% of all new cancers) and breast cancer}by far the most important cancer of women (21% of the total) is third. There are very large differences in the relative importance of the different cancers by world area; some of the factors, environmental and genetic, underlying the geographic distributions, are discussed.
This review concentrates upon describing the current cancer pattern and, for the major cancers, examines incidence by geographic location, briefly discusses temporal trends and, finally, indicates some of the underlying reasons for these patterns and trends, with an attempt at approximate quantitation. Global cancer patterns are described in terms of incidence, the number of new cancer cases, which, expressed as a rate, is an indicator of the average risk of developing cancer to individuals in the population. It is therefore the appropriate statistic to use if one is interested in the causes of cancer, or Žprimary. cancer prevention. Incidence data are available from cancer registries. Disease registers are not unique to cancer, but for this condition, they comprise a particularly well developed and extensive network. They may cover entire national populations, or sub-samples of selected regions. They also provide statistics on cancer survival, thereby enabling incidence to be estimated from mortality. Mortality data by cause are available for many countries through registration of vital events, although the degree of detail and quality of the data Žboth the accuracy of the recorded cause of death and the completeness of registration. vary considerably. With such data, it is possible to prepare estimates of the numbers of new cancer cases Žand deaths. by site, sex and age group. These are more or less accurate, for different countries, depending on the extent and accuracy of locally available data. We have estimated the incidence of cancer worldwide, at intervals of 5 years, the figures relating to 1975,1 1980,2 1985,3 and 1990.4 In this paper, the estimates of 1990, for 23 world ‘Areas’ Žas defined by the UN ŽFigure 1. are presented. However, the estimates are built up at country level, and the full dataset can be accessed either via a CD-ROM 5 or, with less flexibility in the analyses possible, on the Internet Žhttp:rrwww-dep.iarc.frrglobocanrglobocan.htm.. The discussion makes references to the more detailed national-level data as appropriate. No attempt has been made to estimate incidence of non-melanoma skin cancer, because of the difficulties of measurement, and consequent lack of data.
Key words: new cancer cases r annual incidence rates r most common cancers Q1998 Academic Press
Introduction STATISTICS DESCRIBING THE pattern of cancer occurrence in human populations are essential for two purposes. Firstly, by showing how risk of disease varies according to place, time and personal characteristics, the search for more defined environmental exposures, or differential susceptibility to them, can be refined in specially designed studies. Secondly, such data permit rational priority-setting of healthcare activities Žprevention, screening and therapy. and the monitoring the results of interventions through the healthcare system.
From the Unit of Descriptive Epidemiology, International Agency for Research on Cancer, 150 cours Albert Thomas, 69372 Lyon Cedex 08, France Q1998 Academic Press 1044-579Xr 98r 040219q 17 $30.00r 0
219
D. M. Parkin
Figure 1. Map showing the 23 world areas studied.
The total ‘All Cancer’ therefore excludes such tumours. v
Methods of estimation v
Incidence rates for five broad age groups Ž0]14, 15]44, 45]54, 55]64 and 65 and over. and sex were estimated for as many individual countries as possible. Age-standardized incidence rates ŽASRs. were calculated using the weights of the ’world standard’ population Ž0.31, 0.43, 0.11, 0.08 and 0.07. in these five age classes. The Area estimates were obtained by combining agersex-specific rates for component countries as a weighted average Žusing the corresponding country populations.. The sources of data for the countries of the 23 world areas and the methods used to produce estimates of incidence in each of them are summarised in Parkin et al.4 There are six categories Žin order of preference.: v National Incidence Data from good quality cancer registries. v National Mortality Data, with estimation of inci-
v
v
dence using sets of regression models, specific for site, sex, and age, derived from local cancer registry data Žincidence plus mortality.. Local (Regional) Incidence Data from one or more regional cancer registries within a country. Mortality surveys}for China, the mortality data derive from a sample survey of deaths at socalled ‘Disease Surveillance Points’ ŽDSP..6 Frequency Data: In some areas of Africa and Asia, only data on the relative frequency of different cancers Žby age and sex. are available. These are applied to an estimated ‘all sites’ incidence rate, derived from existing cancer registry results. No data: The country-specific rates are those of the corresponding Area Žcalculated from the other countries for which estimates could be made..
Results and discussion An estimated 8.1 million new cancer cases occurred 220
The global burden of cancer
Lung cancer
in 1990, divided almost exactly between developed and developing countries. Figure 2 Žmales. and Figure 3 Žfemales. show the number of cases and the ranking of the 25 sites in developed and developing countries. Table 1 shows the number of cases, by world Area, for the 12 most common cancers in males and in females. Table 2 shows the corresponding age standardised incidence rates. The number by world-area ranges from 1.4 million in China to approximately 1100 in MicronesiarPolynesia. Some caution is necessary in comparing the most recent estimates Žfor 1990. with those for earlier periods. This is because the nature and availability of different sources of data on cancer has changed over time. In general, this should result in progressively more accurate figures, but it does lead to some abrupt changes in estimated incidence for certain areas. Having noted this, the progressive increase in the estimated number of cancer cases worldwide since 1975 has been rather more rapid than the equivalent growth in the world population ŽFigure 4..
This was the most common cancer in 1990, with 1.04 million new cases, or 12.8% of the world total. Worldwide, it is by far the most common cancer of men Ž18% of the total. with the highest rates observed in North America and Europe Žespecially eastern Europe.. Moderately high rates are also seen in temperate South America, AustraliarNew Zealand, and, in Asia, Hong Kong, Singapore and the Philippines ŽFigure 5.. In females, incidence rates are lower Žoverall, the rate is 10.8 per 10 5 women, compared with 37.5 per 10 5 in men.. The highest rates are in the USA and Canada, UK, Denmark. It is of note that the incidence in China is rather high Žage-standardrate 13.4 per 10 5 ., similar to that in, for example Australia or the Netherlands. The major cause of lung cancer is tobacco smoking, and in general, incidence rates in a country closely reflect the past history of tobacco smoking.7 The proportion of lung cancer cases due to tobacco smoking can be estimated by comparing observed
Figure 2. Estimated numbers of new cases of 21 cancers in men in developed and developing areas of the world.
221
D. M. Parkin
Figure 3. Estimated numbers of new cases of 23 cancers in women in developed and developing areas in the world.
incidence in different areas with that expected based upon rates in non-smokers from several large cohort studies.8 Updating the results to 1990, some 86% of lung cancer cases in men and 49% in women are due to smoking, though there is considerable regional variation. Thus in countriesrregions with a long history of smoking, 90% or more cases in men are tobacco-related, while the fraction is much lower in Africa and southern Asia. The proportions are more variable in women, even in Europe, where they range from 80% in the UK, to virtually nil in Spain and Portugal. Our estimate of the numbers of cases worldwide has increased by 16% since 1985 Žq14% in men and q21% in women.; this represents an approximately 2.5% increase in the actual risk in men, and 9.5% in women Žthe remainder is due to population growth and ageing.. Trends vary enormously by individual countries. Thus, in men, several populations have now passed the peak of the tobacco-related epidemic, and incidence rates are now declining Že.g. USA, the countries of northern and western Europe.. In con-
trast, rapid increases are still seen in southern and eastern European countries.9 In women, the ‘epidemic’ is less advanced; most ‘Western’ countries are still showing a rising trend in incidence and mortality, although for some there is, so far, no evidence of this Že.g. Spain., while for others ŽUK., it seems that the peak of risk may already have been reached.10 Stomach cancer Cancer is the second in rank overall Ž798,000 new cases, 9.9% of the total., although only fourth in women. The estimated number of cases is just 6% more than in 1985 which, allowing for population increase and ageing, represents a decline of approximately 4]5% in age-adjusted risk. Thirty-eight percent of cases occur in China, where it remains the most common cancer in both sexes, as it is elsewhere in eastern Asia}the highest incidence is still found in Japan ŽASR 77.9 per 100,000 in men, 33.3 in women. followed by Korea Ž59.9 and 24.8.. The incidence is also high in much of Central America and 222
89.2 2.2 29.8 1.2 46.6 5.1 32.5 3.8 65.0 1.1 28.1 4.1 49.8 5.7 59.8 12.8 158.0 18.4 634.3 118.2 866.4 163.1 224.1 32.9 58.9 9.2 181.4 38.3 452.5 47.8 63.8 14.1 435.1 124.5 182.5 41.0 244.4 56.9 350.0 65.3 38.3 5.9 2.4 0.1 0.5 0.1 2108.7 444.7 2184.8 327.1 4293.5 771.8
1. Eastern Africa 2. Middle Africa 3. Northern Africa 4. Southern Africa 5. Western Africa 6. Caribbean 7. Central America 8. South America ŽTemperate. 9. South America ŽTropical. 10. North America 11. Eastern Asia: China 12. Eastern Asia: Japan 13. Eastern Asia: Other 14. South-eastern Asia 15. South-central Asia 16. Western Asia 17. Eastern Europe 18. Northern Europe 19. Southern Europe 20. Western Europe 21. AustraliarNew Zealand 22. Melanesia 23. MicronesiarPolynesia
Developed countries Developing countries
ALL AREAS
Lung
All sites
MALE
223 511.0
191.7 319.3
4.1 3.6 2.4 1.4 5.5 2.3 5.6 5.6 23.9 14.4 207.7 64.7 13.3 11.9 27.7 4.3 58.9 11.8 20.5 20.1 1.3 0.1 0.0
Stomach
401.9
257.3 144.6
3.6 0.4 2.5 1.4 2.3 2.2 2.7 6.4 11.7 77.1 64.9 32.9 5.2 15.9 21.0 4.0 40.8 23.8 28.2 48.9 5.6 0.2 0.0
Colonr rectum
396.1
300.5 95.7
6.4 4.3 1.9 3.5 8.4 6.0 7.0 5.5 20.5 170.0 4.7 7.3 1.1 6.8 16.6 2.8 22.1 26.2 17.7 50.7 6.5 0.1 0.0
Prostate
316.3
53.9 262.4
9.1 5.8 1.9 2.7 11.3 1.1 1.8 0.6 2.6 5.4 178.1 23.2 8.4 24.8 11.5 2.3 8.4 1.9 8.8 5.9 0.3 0.3 0.0
Liver
212.6
44.8 167.9
5.6 0.4 1.1 4.5 0.9 1.0 1.0 2.6 5.9 8.6 102.3 8.0 1.7 4.1 35.7 1.1 10.4 4.5 4.5 8.1 0.6 0.0 0.0
Oesophagus
Table 1. Estimated number of new cancer cases Žthousands. by site, sex and area, 1990
202.5
127.7 74.9
2.5 0.5 10.0 1.4 3.2 1.1 1.5 3.1 7.2 41.7 14.9 7.5 2.0 5.4 17.1 4.8 20.2 13.9 22.0 20.6 1.8 0.0 0.0
Bladder
141.2
57.5 83.7
3.6 0.9 1.8 2.0 2.8 1.1 1.2 1.6 6.8 12.7 5.8 2.5 0.9 5.2 47.6 1.8 15.1 3.2 9.4 12.7 1.8 0.5 0.0
Oral cavity
130.3
55.4 74.9
1.8 0.3 2.4 0.9 2.8 1.1 2.8 1.4 6.3 17.0 24.5 4.0 1.8 8.0 17.8 2.9 11.5 5.1 7.4 9.3 1.3 0.1 0.0
Leukaemia
126.3
64.8 61.5
5.4 1.2 2.3 0.9 6.4 0.8 1.8 1.1 6.6 25.1 6.9 5.5 1.2 7.3 16.4 3.1 7.7 5.9 8.6 10.5 1.5 0.2 0.0
NonHodgkin lymphoma
118.5
58.1 60.4
1.1 0.4 2.2 0.9 1.1 1.0 1.5 2.0 5.0 10.9 7.9 2.8 1.1 4.8 27.4 4.0 19.7 3.1 10.6 10.3 0.6 0.0 0.0
Larynx
91.5
62.5 29.0
1.0 0.4 0.8 0.4 0.9 0.4 1.2 2.0 3.8 17.9 6.9 4.4 0.9 2.5 6.4 1.3 15.4 5.2 6.0 12.6 1.0 0.0 0.0
Kidney, etc.
The global burden of cancer
All sites 85.5 28.3 45.9 29.2 65.1 26.6 68.3 64.4 166.5 571.5 528.1 162.7 59.5 192.4 488.4 57.1 405.6 185.6 197.2 324.1 34.2 2.9 0.6 1881.0 1908.8 3789.8
FEMALE
1. Eastern Africa 2. Middle Africa 3. Northern Africa 4. Southern Africa 5. Western Africa 6. Caribbean 7. Central America 8. South America ŽTemperate. 9. South America ŽTropical. 10. North America 11. Eastern Asia: China 12. Eastern Asia: Japan 13. Eastern Asia: Other 14. South-eastern Asia 15. South-central Asia 16. Western Asia 17. Eastern Europe 18. Northern Europe 19. Southern Europe 20. Western Europe 21. AustraliarNew Zealand 22. Melanesia 23. MicronesiarPolynesia
Developed countries Developing countries
ALL AREAS
Table 1. Ž Continued.
224 795.6
471.5 324.1
10.5 3.1 12.1 4.9 10.0 5.0 9.7 18.9 35.0 167.3 61.4 24.1 6.2 38.3 95.8 12.9 77.2 48.4 50.9 94.7 9.0 0.4 0.1
Breast
381.0
259.2 121.7
2.2 0.7 2.0 1.2 2.0 2.4 2.8 7.3 11.6 77.4 50.4 25.6 4.6 13.9 16.7 3.8 48.4 24.4 25.2 53.0 5.1 0.1 0.0
Colonr rectum
371.2
83.3 287.9
21.5 5.7 5.2 6.5 13.6 5.0 17.7 7.0 29.9 15.2 24.7 8.5 9.3 30.9 107.0 3.0 27.5 7.6 9.9 13.2 1.3 0.7 0.1
Cervix uteri
287.2
124.2 163.0
2.8 3.2 1.2 0.6 3.3 1.2 4.6 2.6 13.6 9.5 94.4 34.0 7.5 7.8 16.7 3.3 44.1 7.8 13.0 15.2 0.7 0.1 0.0
Stomach
265.1
151.8 113.3
1.0 0.2 1.1 1.2 0.5 1.5 2.7 2.2 6.2 70.7 65.5 12.2 3.8 14.0 10.8 2.6 26.6 17.9 8.8 13.2 2.3 0.0 0.0
Lung
165.5
86.7 78.8
4.6 0.7 1.4 0.8 3.7 1.0 2.9 2.5 6.2 22.0 14.5 5.4 1.6 11.0 24.4 3.3 25.3 9.3 8.5 14.9 1.2 0.1 0.0
Ovary, etc.
142.4
95.9 46.5
1.6 0.7 0.8 1.1 1.0 1.0 2.7 3.6 6.6 31.0 8.2 3.6 0.8 5.8 9.3 3.2 23.2 8.3 12.0 16.6 1.3 0.1 0.0
Corpus uteri
121.1
27.0 94.1
3.1 3.4 1.0 1.0 3.7 0.7 1.4 0.4 1.9 3.0 57.5 7.6 3.0 9.0 6.3 1.5 7.7 1.3 4.5 2.8 0.1 0.2 0.0
Liver
103.2
13.6 89.5
2.7 0.0 0.7 1.7 0.5 0.3 0.5 1.4 1.6 3.2 47.8 1.8 0.4 1.9 28.8 1.0 2.7 3.1 0.6 1.9 0.4 0.0 0.0
Oesophagus
100.9
45.1 55.8
1.4 0.3 1.6 0.6 2.2 0.9 2.5 1.2 5.1 12.5 18.6 2.8 1.2 6.6 11.7 1.9 10.8 4.2 5.7 8.2 0.9 0.1 0.0
Leukaemia
94.6
52.8 41.8
3.0 1.9 1.7 0.6 4.3 0.6 1.3 1.0 4.5 21.3 4.9 3.8 0.8 5.5 9.2 2.4 5.6 5.3 6.4 9.3 1.2 0.1 0.0
NonHodgkin lymphoma
78.9
53.6 25.3
0.8 0.9 0.2 0.3 0.8 0.5 2.3 1.5 3.3 15.3 6.2 6.1 1.0 1.9 4.3 1.1 13.2 5.9 4.6 7.9 0.8 0.0 0.0
Pancreas
D. M. Parkin
All sites 178.0 160.4 105.6 247.4 140.0 199.0 154.7 255.1 196.9 369.9 179.2 270.9 235.7 130.4 106.3 131.4 269.4 270.0 256.0 294.8 312.7 165.4 204.5 299.6 152.4 203.5
MALE
1. Eastern Africa 2. Middle Africa 3. Northern Africa 4. Southern Africa 5. Western Africa 6. Caribbean 7. Central America 8. South America ŽTemperate. 9. South America ŽTropical. 10. North America 11. Eastern Asia: China 12. Eastern Asia: Japan 13. Eastern Asia: other 14. South-eEastern Asia 15. South-central Asia 16. Western Asia 17. Eastern Europe 18. Northern Europe 19. Southern Europe 20. Western Europe 21. AustraliarNew Zealand 22. Melanesia 23. MicronesiarPolynesia
Developed countries Developing countries
225
ALL AREAS 37.5
62.6 24.1
4.9 6.5 12.8 29.1 2.2 29.8 19.3 55.1 24.1 69.6 34.7 38.9 40.4 29.7 11.9 30.5 75.9 59.1 58.8 54.1 47.6 7.8 52.5
Lung
24.5
27.1 23.1
9.0 20.0 5.9 11.5 12.4 16.2 18.6 23.8 31.1 8.4 43.6 77.9 54.3 9.0 6.7 9.4 36.3 16.9 20.8 16.4 10.8 6.5 12.3
Stomach
19.4
35.9 10.2
8.1 2.3 6.0 11.2 4.7 16.0 8.8 27.2 15.0 44.3 13.3 39.5 21.3 11.9 5.0 8.8 25.3 34.4 28.8 39.8 45.8 11.1 17.6
Colonr rectum
19.8
40.1 7.6
16.8 29.6 5.1 31.0 23.9 42.4 24.8 22.9 28.1 92.4 1.1 8.5 5.4 5.9 4.5 7.1 14.1 34.7 16.9 39.6 49.7 6.0 21.2
Prostate
14.7
7.6 17.9
18.5 28.4 4.7 20.1 22.1 8.0 5.4 2.7 3.2 3.2 35.8 27.6 32.2 17.6 2.8 5.0 5.2 2.7 8.9 4.9 2.8 20.1 10.9
Liver
10.2
6.4 12.2
12.5 2.1 2.8 32.6 2.1 7.5 3.4 11.1 7.7 5.2 21.6 9.5 7.3 3.2 9.0 2.4 6.4 6.7 4.9 6.9 4.6 3.6 4.5
Oesophagus
Table 2. Estimated age standardized rates of cancer incidence by site, sex and area, 1990
9.9
17.6 5.6
5.8 2.8 23.3 11.7 8.0 7.9 5.1 13.0 9.5 23.8 3.3 8.9 8.8 4.3 4.4 10.9 12.6 19.9 22.0 16.6 14.5 2.7 6.7
Bladder
6.6
8.4 5.8
7.6 5.3 4.7 16.0 6.4 8.1 4.1 6.9 8.5 7.8 1.1 3.1 3.5 3.8 11.4 3.8 9.3 5.0 10.2 11.3 15.1 36.9 6.0
Oral cavity
5.6
8.6 4.1
3.1 1.3 3.9 5.4 4.9 7.3 5.9 6.0 6.1 10.5 4.5 5.7 4.9 4.0 3.1 4.5 7.6 8.4 8.6 8.5 10.8 4.0 7.5
Leukaemia
5.6
9.6 3.7
8.8 5.1 4.3 5.8 11.0 5.2 4.7 4.6 7.2 15.6 1.3 7.0 4.0 4.6 3.4 5.2 4.9 9.6 9.7 9.2 12.9 9.0 5.4
NonHodgkin lymphoma
5.7
8.5 4.4
2.5 2.4 5.6 6.9 2.6 7.4 5.1 8.6 6.5 6.8 1.7 3.3 4.9 3.8 6.8 8.4 12.0 4.9 11.3 8.9 5.2 3.6 6.5
Larynx
4.3
9.1 2.0
1.6 1.5 1.8 2.7 1.6 3.1 3.9 8.4 4.7 11.0 1.5 5.4 3.8 1.6 1.4 2.6 9.5 8.1 6.4 10.6 8.1 1.3 3.4
Kidney, etc.
The global burden of cancer
All sites 145.3 124.9 94.6 187.8 118.5 175.5 180.1 230.1 185.0 277.5 105.3 166.8 179.9 115.7 109.9 108.3 172.8 234.5 177.6 210.4 254.0 185.6 205.2 208.9 122.5 154.6
FEMALE
1. Eastern Africa 2. Middle Africa 3. Northern Africa 4. Southern Africa 5. Western Africa 6. Caribbean 7. Central America 8. South America ŽTemperate. 9. South America ŽTropical. 10. North America 11. Eastern Asia: China 12. Eastern Asia: Japan 13. Eastern Asia: other 14. South-eastern Asia 15. South-central Asia 16. Western Asia 17. Eastern Europe 18. Northern Europe 19. Southern Europe 20. Western Europe 21. AustraliarNew Zealand 22. Melanesia 23. MicronesiarPolynesia
Developed countries Developing countries
ALL AREAS
Table 2. Ž Continued.
226 33.0
56.4 20.4
18.6 13.6 25.0 31.5 19.0 33.5 25.5 69.1 39.1 86.3 11.8 28.6 17.8 22.5 21.2 24.3 36.0 68.3 49.5 67.3 71.7 23.9 51.7
Breast
15.3
25.4 8.1
4.2 3.4 4.2 8.4 3.9 15.5 7.9 24.4 13.6 32.8 10.2 24.6 14.1 8.9 3.8 7.6 18.5 26.1 20.2 29.0 34.8 5.3 11.3
Colonr rectum
15.5
11.2 18.2
37.4 26.6 11.3 40.4 26.2 33.2 44.4 27.7 31.8 9.1 5.0 9.7 26.6 18.6 23.8 5.5 13.7 12.5 10.4 10.9 11.2 43.4 26.0
Cervix uteri
11.6
12.6 11.0
5.3 14.8 2.6 4.3 6.6 7.9 13.3 8.7 15.9 4.0 19.0 33.3 22.8 5.1 4.0 6.6 16.9 8.1 10.0 8.2 4.9 4.8 7.8
Stomach
10.8
15.5 7.8
2.0 0.8 2.6 7.7 0.9 10.1 7.9 7.6 7.2 32.9 13.4 11.2 11.6 9.3 2.6 5.2 10.3 20.2 7.3 8.2 16.1 3.7 17.3
Lung
6.8
10.3 4.9
7.8 2.9 2.8 5.3 6.5 6.3 7.6 9.1 6.9 11.2 2.7 6.3 4.6 6.2 5.4 6.3 11.5 12.7 8.0 10.0 9.6 7.7 8.4
Ovary, etc.
5.9
10.7 3.1
3.0 3.0 1.6 7.4 2.1 7.0 7.7 12.6 7.9 15.0 1.7 3.9 2.3 3.6 2.2 6.5 9.9 10.8 10.6 11.1 9.1 7.4 12.9
Corpus uteri
4.9
2.6 6.2
5.7 13.6 2.4 6.6 6.7 4.4 4.0 1.5 2.2 1.4 11.5 6.9 9.5 5.8 1.5 3.1 2.8 1.4 3.4 1.5 1.1 9.9 4.5
Liver
4.2
1.3 6.2
5.3 0.2 1.7 11.9 1.2 2.1 1.4 4.5 2.0 1.4 9.9 1.6 1.1 1.3 7.0 2.0 1.0 3.1 0.5 1.1 2.4 2.6 1.5
Oesophagus
4.0
5.5 3.0
1.9 1.0 2.7 3.5 3.8 5.6 5.0 4.4 4.7 6.3 3.5 3.7 3.4 3.2 2.2 2.9 5.0 5.7 5.8 5.6 6.5 3.3 5.6
Leukaemia
3.8
5.8 2.5
3.9 8.4 3.4 4.0 5.9 3.6 3.2 3.4 4.9 10.1 0.9 4.0 2.3 3.1 2.0 4.1 2.5 6.6 5.9 5.8 8.5 4.8 4.0
NonHodgkin lymphoma
3.1
5.0 1.7
1.7 4.1 0.6 1.9 1.6 3.5 6.9 4.9 4.0 6.4 1.2 5.4 3.0 1.3 1.1 2.3 4.8 6.0 3.5 4.2 4.9 0.5 2.9
Pancreas
D. M. Parkin
The global burden of cancer
Figure 4. The estimated total of new cancer cases Žin millions. and estimated world population Žin billions. in 1975, 1980, 1985, and 1990 Žshowing divisions between developing and developed countries..
South America Žespecially the Andean countries., and in eastern Europe ŽFigure 6.. Moderately high rates are observed in some countries of central and west Africa. Low rates are found in east and north Africa, North America, and south and south-east Asia ŽASR in men 5.9]9.0 per 100 000, and 2.6]5.3 in women.. There is clearly a strong environmental component to the risk differences. Migrant populations from high risk parts of the world show a marked diminution in risk when they move to a lower risk area, although this is quite gradual, and seems to depend
on the age at migration. The data fit with observations concerning the importance of childhood environment in determining risk.11 Most research has focused on the contribution of dietary factors to the risk Žas well as the adverse effects of tobacco smoking and alcohol.. Infection with the bacterium Helicobacter pylori appears also to be important, although the exact relationship is still the subject of debate Žsee discussion by Nyren in this issue.. In 1994, IARC 12 accepted H. pylori as carcinogenic to humans, based on ecological correlation studies Žsuch as the EURO-
Figure 5. Incidence of lung cancer: ASR Žworld. }male Žall ages..
227
D. M. Parkin
Figure 6. Incidence of stomach cancer: ASR Žworld. }male Žall ages..
GAST study 13 ., a host of case-control studies and, more usefully, several cohort studies. The combined odds ratio from these studies is 2.1.14 Assuming an average prevalence of infection of 50% in developed countries, and 80% in developing countries, the number of new cases of stomach cancer attributed to
the bacterium would be 337,800, representing 42% of the world total these cancers Ž47% in developing countries and 35% elsewhere.. A decline in gastric cancer incidence and mortality is evident in most countries. This presumably relates to improvements in preservation and storage of
Figure 7. Incidence of breast cancer: ASR ŽWorld. ŽAll Ages..
228
The global burden of cancer
foods.15 It may also reflect changes in the prevalence of H. pylori by birth cohort, perhaps as a result of reduced transmission in childhood, following a trend to improved hygiene and reduction of crowding.16
Asia, the world total in 2010 would be 1.45 million, an 82% increase on the figure for 1990.
Breast cancer
Cancers of the colon]rectum accounted for 782,000 new cases in 1990 Ž9.7% of the world total.. Numbers were similar in males and females Žratio 1.05:1.. The geographic distribution follows the division between Westernised versus developing countries. The highest rates are in N. America, AustraliarNew Zealand and Europe Žespecially GermanyrAustria., with moderately high incidence rates in southern South America. Rates in Africa, except for South Africa, are very low. The geographical distribution of colonic cancer and rectal cancer is similar, although the variation between countries is less for the rectum than for colon. Thus, in high-risk populations, the ratio colonrrectum is 2:1 or more Žrather more in females.; in low-risk countries, rates are similar Žwith even slight excess of rectal cancer in India.. These large geographic differences probably largely represent the effects of different environmental exposures, presumably mainly dietary. There are strong international correlations between risk of large bowel cancers and per capita consumption patterns of meat,18 fat Žspecifically animal fat.19 and fibre.20 That the risk of colon cancer is quite labile to environmental change is evident from the study of migrants. In Hawaii, first-generation Japanese migrants acquire incidence rates similar to, or even in excess of, the indigenous white population.21 Incidence rates have risen}especially in men}in most areas ŽN. America is an exception. since 1985, so that the estimated number of cases in 1990 is 15.5% greater Ž21% in men, 10% in women. than in 1985.
Colorectal cancer
Brease cancer is by far the most frequent cancer of women Ž21% of all cancers. and ranks third overall when both sexes are considered together. It is the most common cancer of women in all the ‘developed’ areas Žexcept for Japan, where it is third after stomach and colon]rectal cancer., and indeed on a world map ŽFigure 7., the pattern strongly resembles that of indicators of ‘affluence’. The incidence is more modest in north Africa, South America, east, south-east and western Asia, but it is still the most common cancer of women in these geographic regions. The age-standardized incidence is highest in the United States Ž87.1 per 100,000. and lowest in China Ž11.8.. Prevalence of carriers of the major susceptibility genes ŽBRCA1 and BRCA2. in the general population is low, and their differential distribution around the world can hardly account for much international or inter-ethnic variation in risk. Most must therefore be a consequence of different environmental exposures, and indeed there are quite marked changes in risk following migration, particularly if this takes place at young ages. The major influences on breast cancer risk appear to be certain reproductive factors, and, less certainly, diet. There have however, been few attempts to quantify the magnitude of risk differentials between populations which might be explained by such factors. In the USA, Brinton et al 17 calculated that the difference in incidence between whites and blacks, at least in women aged 40]54 Ž20%. was entirely explicable in terms of the different prevalence of certain reproductive and ‘lifestyle’ variables. Incidence rates of breast cancer are increasing in most countries, and the changes are usually greatest where rates were previously low. Since our previous estimates for 1985, there has been an overall increase in incidence rates of approximately 0.5% annually Žthis excludes China, where data sources were changed between estimates.. At this rate of growth, there would be approximately 1.35 million new cases in 2010. However, cancer registries in China are recording annual increases in incidence of over 5% and in those elsewhere in eastern Asia, increases are not much less. Assuming a modest 3% growth in east
Liver cancer Liver cancer is currently the fifth most common cancer in the world Ž437,000 new cases annually.. It is primarily a problem of developing countries where over 80% of cases occur. There are high incidence rates in west and central Africa Žwhere it accounts for almost one-quarter of cancer in men., eastern and south-eastern Asia, and in Melanesia. The incidence is low in developed areas}only in southern Europe Žand especially in Greece. is there any substantial risk. The overall sex ratio is approximately 2.6, rather greater in the high-risk areas and less in low-risk 229
D. M. Parkin
areas. The major risk factors for liver cancer are chronic infection with hepatitis viruses ŽHBV and HBC.. Both pose a substantial increase in risk Žmore than 20-fold 22 .. With information on prevalence of infection in different areas, the percentage of liver cancers attributable to these viruses can be estimated; it is 77% of cases worldwide, and 85% of cases in the developing countries.14
ethnic groups ŽWhite, Black, Asian. within the United States. Even before PSA screening, international differences reflected diagnostic bias to some extent. The prevalence of latent prostate cancer at autopsy shows much less variation than clinical prostate cancer, though the ethnic-specific ranks are much the same as for incidence.23 The frequency of latent carcinoma of prostate in Japan is increasing Žas is clinical prostate cancer. and approaching the prevalence for US Whites. This suggests that promotion of latent carcinoma may be important in explaining international differences. The differences in ethnic-specific risk are probably mediated via population differences in alleles of genes coding for enzymes involved in testosterone metabolism.24 The estimated number of cases is substantially higher Ž36%. than 5 years earlier; most of this increase is the result of the huge increase in incidence rates in North America from an age-standardised rate of 61.3 per 100,000 in 1985 to 92.4 per 100,000 in 1990 Žan annual increase of 8.5%..
Prostate cancer Prostate cancer is the fourth most frequent cancer of men overall Ž9.2% of all new cancer cases.. Incidence rates are now influenced by the diagnosis of latent cancers by screening asymptomatic individuals, so that where this practice is common, the ‘incidence’ may be very high Ž95.1 per 100,000 in the United States, for example, where it is now by far the most commonly diagnosed cancer in men.. Incidence is high also in northern and western Europe, and AustraliarNew Zealand. The age-adjusted rates show that several developing areas have relatively high incidence rates also}sub-Saharan Africa, Latin America and the Caribbean in particular. By contrast, the incidence rates in Asia and particularly in China are low. Even when comparison is made after age standardization, the difference in risk between China and North America is more than 80-fold. Similar, albeit less extreme, patterns are observed between different
Cervix cancer Cervix cancer is seventh in frequency overall, third in frequency in women, in whom it comprises 9.8% of all cancers Ž371,200 new cases per year.. The geographic pattern is a complete contrast to breast cancer ŽFigure 8.; the highest incidence is observed in
Figure 8. Incidence of cervix uteri cancer: ASR ŽWorld. ŽAll Ages..
230
The global burden of cancer
Latin America and the Caribbean, sub-Saharan Africa, and south and south-east Asia. In developed countries, the incidence rates are generally low, with agestandardized rates less than 14 per 100,000. It is quite clear that the major aetiological agents are oncogenic subtypes of the human papilloma virus ŽHPV.,25 indeed, it may be that the disease does not occur in the absence of infection. It is curious, therefore, that there is no systematic information on geographicrethnic variations in prevalence of persistent HPV infection with which to compare incidence data, although case control studies do suggest that it is greater in control subjects from higher risk countries Žapprox. 10]20%. than in areas of lower risk Ž5]10%.. The estimated number of cases for 1990 is some 15% lower than in 1985. This is consistent with the decline in incidence of cervix cancer observed in many countries. Although this is evident where there are well established screening programmes, there are also declines observed in several developing countries. This is especially marked in China where the estimated age-standardised incidence rate in 1990
was 5.0, compared with 17.8 in 1985. Although some of this difference reflects changing data sources, cancer registry results also indicate a fairly dramatic decline in rates in recent years. As a result of these trends, cervix cancer has ceded its place as the leading cancer in developing countries to breast cancer, and only in sub-Saharan Africa, Central America, south-central Asia and Melanesia is it now the main cancer of women. Oesophageal cancer Oesophageal cancer is the eighth most common cancer worldwide, responsible for 316,000 new cases in 1990 Ž3.9% of the total.. It is mainly a tumour of developing countries where it accounts for 7.7% of cancers in men, and 4.7% in women. China alone accounts for 47.5% of cases. Geographic variation in incidence is very striking. Incidence rates vary at least 40-fold between registries published in Cancer Incidence in Five Continents, Vol. VII ŽFigure 9.. In men, high risk areas include
Figure 9. Word Standardised Rate w0]85 q x, per 100,000.
231
D. M. Parkin
central Asia and China, east and South Africa, parts of South America and the Caribbean, and, in western Europe, France and Switzerland. The pattern is much the same in women, except in Europe, where high rates are found in Ireland and Scotland, but not in France and Switzerland. The Indian subcontinent is also an area of moderate risk in women. There are also very marked variations in incidence within countries in the high-risk areas. These must represent exposures to important carcinogens, but it seems that these are quite different in the various high-risk areas. Tobacco and alcohol are the main agents involved in Europe and North America, where over 90% of cases can be attributed to these causes. Chewing of tobacco Žand betel. is important in the Indian subcontinent. Hot beverages have been shown to increase risk, and drinking hot mate ´ is probably responsible for raised rates in Uruguay, southern Brazil and northern Argentina. Nutritional deficiencies Žspecifically of micronutrients. are thought to underlie the high risk in central Asia, China, and southern Africa. Here other factors such as pickled vegetables, nitrosamine-rich foods and mycotoxins may also be involved, as well as consumption of opium residues Žin Iran. or pipe stem residues Žin the Transkei of southern Africa. 26 .
America and AustraliarNew Zealand. We have not attempted to differentiate between the different leukaemia subtypes, though clearly the geographic patterns will not be the same. For example, chronic lymphocytic leukaemia, numerically the most important diagnosis in western Europe, is extremely rare in east and south-east Asia. Lymphomas Of the lymphomas, Non-Hodgkin lymphoma is by far the commonest, with 221,000 new cases annually, compared with 59,600 cases of Hodgkin’s disease and 56,600 cases of Myeloma. For the non-Hodgkin lymphomas, the highest incidence is observed in the developed areas of N. America, AustraliarNew Zealand and Europe. Currently the highest rates are observed in populations which are most affected by the epidemic of AIDS Žmales in San Francisco.. High incidence rates are also observed in the central part of Africa, in part the consequence of very high rates in the childhood age-range, due to Burkitt’s lymphoma. The lowest risk is in eastern and southern Asia. Incidence rates of NHL have increased in practically all populations since the 1970s and probably before this too; the environmental exposures Žor modified host responses to them. which underlies this are a mystery}the HIV epidemic can account for only part of it, in certain geographic areas.27 Hodgkin’s disease has a rather similar geographic distribution. Incidence is especially low in the eastern Asian areas, and the Pacific islands. For Myeloma, the highest rates are, again, observed in developed countries, but a few populations of African origin also display moderately elevated rates}for example: in males the Caribbean and southern Africa, in females Middle Africa.
Bladder cancer Bladder cancer, the seventh most frequent cancer of men, is considerably less common in women, in whom it appears only 14th in rank Žsex-ratio worldwide is approx. 3.5:1.. In general, the geography is similar to that of lung cancer, with the highest rates in developed areas Žwith the exception of Japan.. In fact, we estimate that 36% of bladder cancer cases in men and 17% in women might be attributable to tobacco smoking. However, there are areas of high risk in north Africa Žestd. ASR in Egypt 46.6. western Asia Žestd. ASR in Iraq 16.5. and some African countries wZimbabwe Ž16.0. and Zambia Ž15.0.x. Here, the high rates result from endemic urinary schistosomiasis, known to be associated with an increased risk of squamous cell carcinomas.12
Mouth and pharynx In previous estimates, all cancers of the mouth and pharynx were considered together as a single group. In the current estimates, we have divided this group into three categories}cancers of the oral cavity, nasopharynx and other pharynx. Of these, cancers of the oral cavity are the most frequent, with 212,000 new cases worldwide Ž2.6% of the total., followed by other pharyngeal cancers with 94,000 cases Ž1.2%. and cancers of the nasopharynx with 57,500 cases Ž0.7%.. All of these are much more common in males than females, the sex ratio ŽM:F. is 2.0 for mouth
Leukaemia Leukaemia is the 10th most frequent cancer, with an estimated 231,000 new cases. There is relatively little geographic variation, although rates appear to be low in sub-Saharan Africa Žpossibly related in part to under-recording .. The incidence is highest in North 232
The global burden of cancer
cancer 4.4 for pharynx cancer and 2.2 for nasopharynx cancer. There are some similarities to the geographic patterns for cancers of the mouth, and ‘other pharynx’. In men, both are high in western and southern Europe, and south Asia, while mouth cancers Žbut not pharynx. have high rates in Melanesia, southern Africa, and AustraliarNew Zealand. In females, both types of cancer have relatively high incidence rates in southern Asia and Žmouth cancers only. Melanesia and AustraliarNew Zealand. These patterns reflect prevalence of specific risk factors}tobaccoralcohol in western and southern Europe, and southern Africa, and the chewing of betel quid in south-central Asia and Melanesia. The high rate of oral cancer in Australia is due to lip cancer Žrelated to solar irradiation.. Although cancer of the nasopharynx is caused by infection by the Epstein]Barr virus,28 this ubiquitous agent cannot explain the tumour’s very singular geographic distribution. Although certain traditional dietary items are clearly implicated in the increased risk,29 it seems more likely that genetic predisposition is the explanation. Thus the incidence is highest in peoples from southern China ŽGuangdong, Guanxi, Hunan and Fujian.,30 migrants from these areas Že.g. in Singapore., or in areas where they have settled, for example, south-east Asia. Forty-four percent of the world cases occur in China and 23% in south-east Asia. A linkage study based on affected sib pairs of Chinese origin suggested that a gene Žor genes. closely linked to the HLA locus is associated with a marked elevation Ž20-fold. in the risk of nasopharynx cancer.31 There is also a high risk in Inuit populations in the Arctic Žpossibly due to common ancestry.. In the medium risk area of north Africa, the epidemiological profile is a little different, with a small peak in incidence in early adolescence.
developed countries. Ovarian cancer occurs in almost all with age-standardised rates in excess of seven per 100,000}the exceptions are in southern Europe ŽSpain, Portugal, Greece. and Japan. Incidence in temperate South America is moderately high Žaverage 9.1 per 100,000, Table 2.. The high risk areas for cancer of the Corpus uteri are North America ŽASR 15.0., central Europe Ž13.8., Scandinavia Ž13.6., and, temperate South America Ž12.6.. Corpus cancer is rare in Asia}including Japan}and in sub-Saharan Africa, with age standardised rates inferior to four per 100,000.. Kidney cancer Kidney cancer Ž150,000 new cases, 1.9% of the world total. has the highest rates in North America and western, northern and eastern Europe while incidence rates are low in Africa, Asia Žexcept Japanese males. and the Pacific. Larynx cancer Larynx cancer Ž136,000 new cases. is predominantly a cancer of men, in whom it comprises 2.8% of cases. The sex ratio Žalmost 7:1. is greater than for any other site and Figure 3 emphasises the rarity of larynx cancer in women, particularly in developed countries. For men, the high-risk areas are Europe Žeast, south, west., temperate South America and western Asia. In western Asia, larynx cancer accounts for more than 6% of cancers in men. Tobacco smoking is estimated to cause two-thirds of all cases in men. Brain and nervous system cancers Cancers of the brain and nervous system account for some 127,000 new cases annually Ž1.6% of new cancers.. The highest rates are observed in developed areas ŽAustraliarNew Zealand, North America, northern Europe. and lowest in Africa and the Pacific islands. This suggests that the availability of diagnostic facilities may well be important in determining geographic patterns, at least in part. The incidence is probably underestimated in many developing countries.
Pancreas cancer Pancreas cancer Ž170,000 cases per year, 2.1% of the total. is slightly more common in men than in women Žratio 1.2:1.. Geographic variation is not large; the higher rates in most developed Areas probably reflect diagnostic capacity rather than aetiology. Ovarian cancer Ovarian cancer Ž165,000 cases, 2.0% of all cancers, 4.4% of cancers in women. and cancer of the corpus uteri Ž142,000 cases, 1.8% of all cancer, 3.8% of cancer in women. are, in general, tumours of women in
Malignant melanoma Malignant melanoma of skin accounts for 105,000 new cases annually, with slightly more occurring in 233
D. M. Parkin
women than in men Žsex-ratio 0.9..The highest rates are observed in AustraliarNew Zealand ŽASR 27.9 per 100,000 in men, 25.0 in women., North America, and northern Europe Žespecially Scandinavia. }these regions have seen annual increases of between 1.5 and 4.5% in incidence Žage-adjusted . of melanoma since 1895. The lowest rates ŽASR less than 0.6 per 100,000. are observed in east and south-east Asia, and in south Asia.
tions of identified causes Ženvironmental and genetic. to the observed patterns. In this way, epidemiology contributes directly to public health, in guiding the selection of priorities in prevention of disease.
References 1. Parkin DM, Stjernsward ¨ J, Muir CS Ž1984. Estimates of the worldwide frequency of twelve major cancers. Bull WHO 62:163]182 2. Parkin DM, Laara ¨¨ ¨ E, Muir CS Ž1988. Estimates of the worldwide frequency of sixteen major cancers in 1980. Int J Cancer 41:184]197 3. Parkin DM, Pisani P, Ferlay J Ž1993. Estimates of the worldwide incidence of eighteen major cancers in 1985. Int J Cancer 54:594]606 4. Parkin DM, Pisani P, Ferlay J Ž1998a. Estimates of the worldwide incidence of twenty-five major cancers in 1990. Int J Cancer, in press 5. Ferlay J, Parkin DM, Pisani P Ž1998. GLOBOCAN: Cancer Incidence and Mortality Worldwide, IARC CancerBase No. 3, in press 6. World Health Organization Ž1993. Cause of Death Statistics and Vital Rates, Civil Registration Systems and Alternative Sources of Information, World Health Statistics Annual 1993, WHO, Geneva 7. Doll R Peto R Ž1981. The Causes of Cancer, Oxford University Press, Oxford 8. Parkin DM, Pisani P, Lopez AD, Masuyer E Ž1994. At least one in seven cases of cancer is cause by smoking. Global estimates for 1985. Int J Cancer 59:494]504 9. Kubik AK, Parkin DM, Plesko I, Zatonski W, Kramarova ´ ´ E, Mohner M, Friedl HP, Juhasz L, Tzvetansky CG, Reissigova J ¨ Ž1995. Patterns of cigarette sales and lung cancer mortality in some Central and Eastern European Countries, 1960]1989. Cancer 75:2452]2460 10. Lopez AG, Pollan M, de-la Iglesia P, Ruiz M Ž1995. Characterization of the lung cancer epidemic in the European Union Ž1970]1990.. Cancer Epidemiol Biomarkers Prev 4:813]820 11. Coggon D, Osmond C, Barker DJ Ž1990. Stomach cancer and migration within England and Wales. Br J Cancer 61:573]574 12. IARC Monograph on the Evaluation of Carcinogenic Risks to Humans, vol. 61 Ž1994. Schistosomes, Liver Flukes and Helicobacter pylori, International Agency for Research on Cancer, Lyon, France 13. Eurogast Study Group Ž1993. An international association between Helicobacter pylori infection and gastric cancer. Lancet 341:1329]1362 14. Parkin DM, Pisani P, Munoz ˜ N, Ferlay J Ž1998b. The global health burden of infection associated cancers, in: Infections and Human Cancer, vol 33 ŽWeiss RA, Beral V, Newton R, eds. Žin press.. Cancer Surveys 15. Howson CP, Hiyama T, Wynder EL Ž1986. The decline in gastric cancer: epidemiology of an unplanned triumph. Epidemiol Rev 8:1]27 16. Banatvala N, Mayo K, Megraud F, Jennings R, Deeks JJ, Feldman RA Ž1993. The cohort effect and Helicobacter pylori. J Infect Dis 168:219]221 17. Brinton LA, Benichou J, Gammon MD, Brogan DR, Coates R, Schoenberg JB Ž1997. Ethnicity and variation in breast cancer incidence. Int J Cancer 73Ž3.:349]355 18. Armstrong B, Doll R Ž1975. Environmental factors and cancer incidence and mortality in different countries, with special reference to dietary practices. Int J Cancer 15:617]631
Cancer of the thyroid Thyroid cancer Ž87,000 new cases. is the only cancer considered here for which the incidence is much higher in females than in males wsex ratio ŽM:F. is 0.33x. It comprises 1.7% of cancers in women. The highest estimated incidence is in MicronesiarPolynesia Žan area well known for a high frequency of these cancers 32 . and in Central America. Testis cancer Testis cancer is relatively rare}36,000 new cases annually or 0.8% of cancers in men. The highest rates are observed in west and north Europe ŽASR in Denmarks 9.5, in Germanys 8.4.. The highest risk is in the age-group 15}44, so that testis cancer is the most common cancer of men in developed countries in this age-group Ž11.5% of new cases, compared with 9.4% due to lung cancer..
Conclusions Completing the global overview of cancer, in terms of its overall impact, and how the risk of different cancers varies between populations, has several objectives. Traditionally, for the epidemiologist, variation in risk between populations and over time has been a useful starting point for investigating the causes of cancer. Almost always, these were thought of in terms of environmental Žincluding lifestyle. exposures, and the failure to identify adequate explanations was a source of frustration, whilst more recently, the importance of genetic variation as a cause of interpopulation differences in risk has been increasingly recognised. In addition, knowledge of the cancer burden in different countriesrareas, together with information on past trends and likely future evolution, are essential elements in planning and the evaluation of cancer control programmes. In this context it is particularly valuable to estimate the contribu234
The global burden of cancer
26. Munoz ˜ N, Day N Ž1996. Esophageal cancer, in Cancer Epidemiology and Prevention Ž2nd ed.. ŽSchottenfeld D, Fraumeni JF, eds. Oxford University Press, New York, USA 27. Hartge P, Devesa SS Ž1992. Quantification of the impact of known risk factors on time trends in non-Hodgkin’s lymphoma incidence. Cancer Res 52:5566s]5569s 28. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, vol. 70 Ž1997. Epstein-Barr Virus and Kaposi’s Sarcoma Herpes virusrHuman Herpes virus 8, International Agency for Research on Cancer, Lyon, France 29. Hildesheim A, Levine PH Ž1993. Etiology of nasopharyngeal carcinoma: a review. Epidemiol Rev 15:466]485 30. Li J, Liu B, Li G, Rong S, Cao D, 25 others Ž1979. Atlas of Cancer Mortality in the People’s Republic of China, China Map Press, Shanghai 31. Lu SJ, Day NE, Degos L, Lepage V, Wang PC, Chan SH, Simons M, McKnight B, Easton D, Zeng Y et al Ž1990. Linkage of a nasopharyngeal carcinoma susceptibility locus to the HLA region. Nature, 346:470]471 32. Paksoy N Ž1992. Frequency of thyroid cancer in pacific populations. J Natl Cancer Inst 84:1124]1125
19. Prentice RL, Sheppard L Ž1990. Dietary fat and cancer: consistency of the epidemiologic data, and disease prevention that may follow from a practical reduction in fat consumption. Cancer Causes Control 1:81]87 20. McKeown-Eyssen G Ž1994. Epidemiology of colorectal cancer revisited: are serum triglycerides andror plasma glucose associated with risk? Cancer Epidemiol Biomarkers Prev 3:687]695 21. Kolonel LN, Hirohata T, Chappell BV, Viola FV, Harris DE Ž1980. Cancer mortality in a cohort of naval shipyard workers in Hawaii: early findings. J Natl Cancer Inst 64:739]743 22. Donato F, Boffetta P, Puoti MA Ž1998. meta-analysis of epidemiological studies on the combined effect of Hepatitis B and C virus infections in causing hepatocellular carcinoma. Int J Cancer 75Ž3.:347]354 23. Yatani R, Chigusa I, Akazaki K, Stemmerman GN, Welsh RA, Correa P Ž1982. Geographic pathology of latent prostate carcinoma. Int J Cancer 29:611]616 24. Shibata A, Whittemore A Ž1997. Genetic predisposition to prostate cancer: possible explanations for ethnic differences in risk. The Prostate 32:65]72 25. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, vol. 64 Ž1995. Human Papillomavirus, International Agency for Research on Cancer, Lyon, France
235