Recent patterns in gastric cancer: A global overview

Authors

  • Paola Bertuccio,

    1. Istituto di Ricerche Farmacologiche “Mario Negri”, Via G. La Masa 19, 20156 Milan, Italy
    Search for more papers by this author
  • Liliane Chatenoud,

    Corresponding author
    1. Istituto di Ricerche Farmacologiche “Mario Negri”, Via G. La Masa 19, 20156 Milan, Italy
    • Istituto di Ricerche Farmacologiche “Mario Negri”, Via Giuseppe La Masa 19, 20156 Milan, Italy
    Search for more papers by this author
    • Fax: +39-0233200231.

  • Fabio Levi,

    1. Unite' d'Épidemiologie du Cancer et Registres Vaudois et Neuchâtelois des Tumeurs, Institut de Médecine Sociale et Préventive (IUMSP), Centre Hospitalier Universitaire Vaudois et Universite' de Lausanne, CHUV-Falaises 1, 1011 Lausanne, Switzerland
    Search for more papers by this author
  • Delphine Praud,

    1. Istituto di Ricerche Farmacologiche “Mario Negri”, Via G. La Masa 19, 20156 Milan, Italy
    2. Institut d'Informatique et de Mathématiques Appliquées, Université Joseph Fourier, BP N°53, 38041 Grenoble Cedex 9, France
    Search for more papers by this author
  • Jacques Ferlay,

    1. International Agency for Research on Cancer, 150 Cours Albert Thomas, 69372 Lyon Cedex 08, France
    Search for more papers by this author
  • Eva Negri,

    1. Istituto di Ricerche Farmacologiche “Mario Negri”, Via G. La Masa 19, 20156 Milan, Italy
    Search for more papers by this author
  • Matteo Malvezzi,

    1. Istituto di Ricerche Farmacologiche “Mario Negri”, Via G. La Masa 19, 20156 Milan, Italy
    2. Istituto di Statistica Medica e Biometria “G.A. Maccacaro”, Università degli Studi di Milano, Via Venezian 1, 20133 Milan, Italy
    Search for more papers by this author
  • Carlo La Vecchia

    1. Istituto di Ricerche Farmacologiche “Mario Negri”, Via G. La Masa 19, 20156 Milan, Italy
    2. Istituto di Statistica Medica e Biometria “G.A. Maccacaro”, Università degli Studi di Milano, Via Venezian 1, 20133 Milan, Italy
    Search for more papers by this author

Abstract

Until the mid-1990s, gastric cancer has been the first cause of cancer death worldwide, although rates had been declining for several decades and gastric cancer has become a relatively rare cancer in North America and in most Northern and Western Europe, but not in Eastern Europe, Russia and selected areas of Central and South America or East Asia. We analyzed gastric cancer mortality in Europe and other areas of the world from 1980 to 2005 using joinpoint regression analysis, and provided updated site-specific incidence rates from 51 selected registries. Over the last decade, the annual percent change (APC) in mortality rate was around −3, −4% for the major European countries. The APC were similar for the Republic of Korea (APC = −4.3%), Australia (−3.7%), the USA (−3.6%), Japan (−3.5%), Ukraine (−3%) and the Russian Federation (−2.8%). In Latin America, the decline was less marked, but constant with APC around −1.6% in Chile and Brazil, −2.3% in Argentina and Mexico and −2.6% in Colombia. Cancers in the fundus and pylorus are more common in high incidence and mortality areas and have been declining more than cardia gastric cancer. Steady downward trends persist in gastric cancer mortality worldwide even in middle aged population, and hence further appreciable declines are likely in the near future. © 2009 UICC

Until the mid-1990s, gastric cancer has been the most common cause of cancer death worldwide, although rates have been substantially declining for several decades, and gastric cancer has become a relatively rare cancer in North America and in most Northern and Western European countries.1, 2 There are, however, areas of the world where gastric cancer remains common. These include Japan, Russia and other countries of the former Soviet Union, and Chile and a few other Latin American countries. Therefore, gastric cancer still accounts for over 10% of cancer deaths worldwide, and is the second most frequent cause of cancer death after lung cancer.3, 4

We considered gastric cancer trends in Europe between 1980 and 1999, and found an almost 50% fall in rates in Western and Eastern Europe and a 40% fall in Russia.5 In proportional terms, the falls in the last decade considered were similar to previous calendar years, although in absolute terms the impact of the decline in gastric cancer mortality became smaller, particularly in Western Europe.

Over the last few decades, the incidence of non-cardia stomach cancer has substantially declined in developed regions of the world. However, non-cardia stomach cancer remains common in many geographic regions, including China, Japan, Eastern Europe and Central/South Americas.6, 7

To provide a global overview of recent rates of gastric cancer mortality and incidence in Europe and other areas of the world, we have considered gastric cancer mortality over the period 1980–2005 for 63 countries providing data, and the European Union (EU) as a whole, and updated site-specific incidence rates from 51 selected registries. We also used joinpoint regression8 to analyze trends for the 19 major countries considered.

Material and methods

Official death certification data from gastric cancer for 43 countries of the European Region (according to the World Health Organization (WHO) definition) and other 20 countries in the world over the period 1980–2005 were derived, whenever available, from the WHO database.9

For the Americas, only 13 countries with more than 2 million inhabitants and with sufficiently detailed age-stratified mortality and population figures were included (i.e., Canada, USA, Argentina, Brazil, Chile, Colombia, Costa Rica, Cuba, Ecuador, Mexico, Puerto Rico, Uruguay and Venezuela).

Mortality data from gastric cancer were also available for 4Asian countries (Hong Kong, Japan, Republic of Korea and Singapore); 1 from Africa (Mauritius), and for Australia and New Zealand.

During the 1980s up to the mid-1990s, all countries utilized the 9th Revision of the International Classification of Diseases (ICD-9)10 with the only exception of Northern European countries and Switzerland, that used ICD-811 (code no. 151). The ICD-1012 (code C16) was adopted by some countries (33) from mid-1990s; others (19) adopted it only around 2000. Since differences between various revisions were minor, gastric cancer deaths were recoded for all countries according to the ICD-9 (code no. 151), i.e., the most commonly used classification.10 Estimates of the resident populations for the corresponding calendar periods, based on official censuses, were extracted from the same WHO database.9 For the Americas, data from certain years onward were unavailable in the WHO database and were thus extracted from the Pan American Health Organization (PAHO).13 Since the PAHO publication provided sex and 5-year age groups (from 0–4 to 80+ years) for selected years only (i.e., 1995, 2000, 2005), sex- and age-specific data for missing years were estimated, by interpolation using the last year available in the WHO database (1986 for Brazil, 1990 for Uruguay, 1991 for Ecuador, 1992 for Puerto Rico, 1994 for Chile, Colombia and Venezuela, 1995 for Costa Rica and Mexico, 1996 for Argentina and Cuba).

From the matrices of certified deaths and resident population, we computed age-specific rates for each calender year and 5-year age group (from 0, 1–4 to 85+ years). For Latin American countries 80+ years was the last age group available. We computed age-standardized mortality rates per 100,000 men and women using the direct method on the basis of the world standard population14 at all ages and, for selected countries only, in 2 age groups (35–64 and 30–49), together with the corresponding percent change in rates over the 2 decades considered. Given the large number of deaths in most countries, the corresponding changes in rates are significant. The issue of significance in trends is formally considered by using joinpoint regression analysis8 (see below).

We also calculated gastric cancer mortality rates for the EU as a whole (defined as the 27 member states as in January 2007), excluding Cyprus whose data were not available and Belgium whose data were provided only up to 1997.

To identify significant changes in trends, we performed joinpoint regression analysis with the software provided by the Surveillance Research Program of the US National Cancer Institute (joinpoint regression program, version 3.3, http://srab.cancer.gov/joinpoint).15 With this analysis, it is possible to identify years where a significant change in the linear slope of the trend (in a log-scale) is detected.8 The best fitting points, called “joinpoints,” are chosen where the rate changes significantly. The analysis starts with the minimum number of joinpoints (e.g., 0 joinpoints, which is a straight line), and tests whether one or more joinpoints (up to 3) are significant and must be added to the model. Each significant joinpoint that indicates a change in the slope (if any) is retained in the final model. To describe linear trends by period, the estimated annual percent change (APC) in mortality rate is then computed for each of those trends by fitting a regression line to the natural logarithm of the rates using the calendar the year as the regressor variable (i.e., given y = a + bx, where y = ln(rate) and x = calendar year, the APC is estimated as 100*(eb−1)).

For 51 selected cancer registries, we also showed age-standardized (world population) incidence rates for the period 1998–2002, derived from “Cancer Incidence in Five Continents, Volume IX”16 by gender for gastric cancer as a whole and for 4 different anatomical subsites, i.e. cardia (ICD-10: C16.0), fundus and corpus (ICD-10: C16.1–C16.2), pyloric antrum and pylorus (ICD-10: C16.3–C16.4), and other subsites (lesser and greater curvature, ICD-10: C16.5–C16.6) together with the unspecified subsites (ICD-10: C16.8–C16.9).

Results

Figures 1a and 1b give trends in overall age-standardized gastric cancer mortality for men and women, respectively, in 4 geographic areas of Europe, i.e., East, South, West and North from 1980–1984 to 2000–2004. For both sexes, mortality rates declined during all the period and in the 4 geographic areas considered. Eastern Europe remains the area with the highest mortality rates, followed by Southern Europe, whereas Western and Northern Europe had the lowest values. Between the early 1980s and 2000–2004, rates decreased in Eastern Europe from 29 to 17/100,000 in men and from 12.5 to 7/100,000 in women, with a similar fall (−41 and −44% for men and women, respectively). In the same period, in southern Europe rates decreased from 22 to 12/100,000 in men (−45%) and from 11 to 5.5/100,000 (−50%) in women. Corresponding values for Northern Europe were from 17 to 7/1,00,000 in men and from 7.5 to 3/100,000 in women, with a fall of 57% for both sexes.

Figure 1.

Trends in overall age standardized (world population) death certification rates from gastric cancer in Eastern Europe (Bulgaria, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovakia, Slovenia), Southern Europe (Croatia, Greece, Italy, Malta, Portugal, Spain), Western-Central Europe (Austria, France, Germany, Luxembourg, The Netherlands, Switzerland), Northern Europe (Denmark, Finland, Ireland, Norway, Sweden, UK), from 1980 to 2004. (a) men; (b) women.

Table I gives for each European country considered the overall age-standardized mortality rates from gastric cancer per 100,000 inhabitants for men and women in 1994, 1999 and 2004 (when available), and the corresponding percent change.

Table I. Overall Age-Adjusted (World Population) Mortality Rates From Gastric Cancer Per 100,000 Men and Women in Europe in 1994, 1999, 2004 (Unless Indicated in Parentheses) and Corresponding Changes in Rates
 Calendar years1MenWomen
199419992004% Change 1999/1994% Change 2004/1999199419992004% Change 1999/1994% Change 2004/1999
  • 1

    Years not available: [1990–1991] for Albania, [1983–1984] for Armenia, Azerbaijan, Belarus, Estonia, Kazakhstan, Kyrgyzstan, Lithuania, Moldova, Tajikistan and Ukraine, [1983–1984, 1993] for Georgia, [1997–1998] for Poland, [2000] for United Kingdom, [1983–1984, 2001] for Uzbekistan.

EUROPE
 Albania1987–2004 16.2913.92 −14.59 7.547.73 2.55
 Armenia (2003)1981–200315.9113.3516.79−16.0825.766.985.827.09−16.6121.81
 Austria1980–200514.109.898.07−29.83−18.387.655.354.63−30.12−13.42
 Azerbaijan (2002)1981–200224.3225.2421.643.81−14.2810.2710.0611.06−2.039.90
 Belarus (2003)1981–200335.5132.9727.44−7.17−16.7615.0713.0210.51−13.63−19.22
 Bulgaria1980–200418.5514.9113.25−19.62−11.158.757.496.25−14.43−16.50
 Croatia1985–200521.1522.7916.387.77−28.128.277.496.21−9.50−17.09
 Czech Republic1986–200516.5711.7310.15−29.20−13.537.345.884.78−19.79−18.82
 Denmark (2001)1980–20016.405.195.13−18.81−1.203.022.832.75−6.25−2.85
 Estonia1981–200525.9223.4518.94−9.52−19.2413.4012.099.66−9.75−20.10
 Finland1980–200510.817.896.31−27.06−20.024.643.643.86−21.656.06
 France1980–20057.306.535.63−10.46−13.852.812.462.08−12.45−15.16
 Georgia (2001)1981–200111.2012.028.517.36−29.205.255.104.99−2.74−2.26
 Germany1980–200412.889.947.95−22.81−20.086.725.194.23−22.75−18.44
 Greece1980–2005 8.707.25 −16.68 3.733.65 −2.18
 Hungary1980–200521.6417.7913.49−17.79−24.169.457.506.49−20.63−13.51
 Iceland1980–200510.297.686.88−25.35−10.424.043.303.06−18.20−7.48
 Ireland1980–200510.268.326.90−18.87−17.045.514.292.64−22.17−38.55
 Italy (2003)1980–200315.0611.9510.43−20.64−12.737.235.465.04−24.46−7.68
 Israel (2003)1980–20038.877.497.81−15.594.345.074.333.41−14.50−21.42
 Kazakhstan1981–200533.9230.5725.71−9.88−15.9213.9412.6510.77−9.25−14.87
 Kyrgyzstan1981–200530.5824.4923.24−19.92−5.1310.6810.848.301.44−23.37
 Latvia1980–200526.3822.4920.55−14.74−8.6211.489.837.85−14.40−20.18
 Lithuania1981–200525.6722.8621.09−10.92−7.779.709.367.98−3.51−14.76
 Luxembourg1980–20059.047.527.03−16.83−6.554.403.552.86−19.33−19.54
 Macedonia TFYR (2003)1991–200322.0117.1717.77−22.003.5010.527.807.98−25.882.28
 Malta1980–20059.9912.258.0322.66−34.425.773.703.65−35.78−1.55
 Netherlands1980–200410.758.476.53−21.17−22.994.433.553.37−19.79−5.04
 Norway1980–20059.618.555.57−10.96−34.904.684.282.94−8.41−31.47
 Poland1980–200519.6816.5214.28−16.05−13.576.975.705.00−18.27−12.23
 Portugal (2003)1980–200321.0218.6916.36−11.09−12.469.668.937.30−7.57−18.24
 Republic of Moldova1981–200519.7416.5116.89−16.352.339.516.697.52−29.6312.41
 Romania1980–200417.7916.6716.00−6.33−4.026.836.086.29−10.953.43
 Russian Federation1980–200537.4631.7027.16−15.40−14.3215.5813.0211.18−16.43−14.14
 Slovakia1992–2005 16.6913.63 −18.34 5.845.78 −0.92
 Slovenia1985–200521.1817.8214.84−15.86−16.757.827.675.42−1.86−29.44
 Spain1980–200513.1610.909.02−17.14−17.265.694.673.79−17.99−18.74
 Sweden1980–20047.056.134.94−13.12−19.323.753.052.73−18.78−10.29
 Switzerland1980–20057.985.534.50−30.75−18.553.582.712.10−24.48−22.35
 Tajikistan1981–200419.2517.9914.88−6.53−17.2810.6310.829.261.80−14.38
 Ukraine1981–200528.8524.8621.23−13.84−14.6111.249.588.28−14.76−13.65
 United Kingdom1980–200510.418.086.09−22.36−24.684.223.262.75−22.80−15.65
 Uzbekistan1981–200515.9311.9812.34−24.772.937.115.566.22−21.8612.04
 EU 13.2210.859.09−17.90−16.285.984.814.17−19.42−13.45

For both sexes the falls in rates were similar in proportional terms in the 2 subsequent calendar periods considered. In absolute terms, the falls were larger in Eastern than in Western Europe over the last decade, due to their higher baseline mortality. For some countries, the falls continued to increase within the last decade (including Belarus, Croatia and Kazakhstan). In the EU as a whole, the falls were lower both in absolute and proportional terms over the last decade.

Table II considers other areas of the world. In all American countries, mortality rates declined steadily until recent years. In 2004, the lowest rates were observed in the USA (3/100,000 men and 2/100,000 women) and Canada (5/100,000 men and 2/100,000 women), but also in some Latin American countries, i.e., Cuba, Mexico and Puerto Rico, whereas gastric cancer mortality remained high in Chile, Costa Rica, Colombia and Ecuador, mainly in men with rates from 16 to 25/100,000. Also in Asia and Oceania, falls in rates were observed, but gastric cancer mortality remained high in Japan and in the Republic of Korea, particularly in men (with rates in 2004 of 23/100,000 in Japan and 31/100,000 in Korea, the highest rate among the 63 countries considered).

Table II. Overall Age-Adjusted (World Population) Mortality Rates From Gastric Cancer Per 100,000 Men and Women in Other Selected Areas of the World in 1994, 1999, 2004 (Unless Indicated in Parentheses) and Corresponding Changes in Rates
 Calendar years1MenWomen
199419992004% Change 1999/1994% Change 2004/1999199419992004% Change 1999/1994% Change 2004/1999
  • 1

    Years not available: [1981, 1983] for Chile, [1987–1991, 2002–2003] for Colombia, [1984] for Costa Rica and Mexico, [1983, 1989–1990, 1996] for Ecuador, [1991–1992, 2002–2003] for Uruguay, [1984, 1991, 1995] for Venezuela.

AMERICA
 Argentina1982–200410.8210.119.10−6.58−9.944.433.913.38−11.79−13.43
 Brazil1980–200412.5511.4510.89−8.80−4.905.424.814.50−11.26−6.53
 Canada1980–20046.155.354.52−13.04−15.542.902.612.17−10.12−16.69
 Chile1980–200432.2430.4425.36−5.60−16.6811.7310.889.15−7.21−15.89
 Colombia1984–200421.4318.5616.26−13.39−12.3912.999.868.82−24.11−10.53
 Costa Rica1980–200537.7825.2622.50−33.15−10.9216.6312.8011.22−23.01−12.37
 Cuba1980–20056.136.315.812.90−7.873.163.953.4124.69−13.64
 Ecuador1982–200520.9519.8215.58−5.39−21.3913.9112.9510.76−6.87−16.94
 Mexico1981–20059.558.337.26−12.78−12.866.826.115.31−10.40−13.02
 Puerto Rico (2003)1980–20038.186.725.10−17.90−24.134.142.782.56−33.04−7.78
 Uruguay1980–200516.7612.3210.59−26.47−14.086.074.754.78−21.780.61
 USA1980–20044.573.823.20−16.54−16.292.081.871.63−10.03−12.83
 Venezuela1980–200416.8513.7511.00−18.42−19.989.658.066.23−16.44−22.71
AFRICA, ASIA, OCEANIA
 Mauritius1980–200512.179.687.27−20.39−24.917.184.216.43−41.3352.72
 Hong Kong SAR1980–20059.799.877.510.86−23.905.695.044.13−11.37−18.07
 Japan1980–200530.7327.9023.35−9.19−16.3112.7011.109.19−12.60−17.18
 Republic of Korea1985–200550.4637.1030.68−26.48−17.2920.1314.6811.31−27.05−22.99
 Singapore1980–200515.5614.699.40−5.58−35.988.346.755.25−19.08−22.24
 Australia (2003)1980–20036.745.394.31−19.94−20.192.772.402.35−13.57−1.99
 New Zealand1980–2004 7.946.41 −19.32 3.573.08 −13.60

When truncated rates for 35–64 years were considered, the patterns of trends were similar to the one observed at all ages, and the falls were also similar, in proportional terms, to those observed at all ages (data not shown).

To further understand cohort patterns and most likely future trends in gastric cancer mortality, we also considered trends in men and women aged 30–49 years in the 18 largest countries and in the EU as a whole (Table III). Even if the rates were based on few deaths and thus changes must be interpreted with due caution, there has been a steady decline even in this age group, except for women in France, UK, the USA and Australia, which had, however, the lowest rates (around 1/100,000 women). In 2004 the Russian Federation, Ukraine, Republic of Korea and Japan had the highest rates for both sexes in the young adults (from 11 to 6/100,000 in men and from 8 to 5/100,000 in women).

Table III. Age-Adjusted (World Population) Mortality Rates From Gastric Cancer Per 100,000 Men and Women Aged 30–49 Years in Selected Countries in 1994, 1999, 2004 and Corresponding Changes in Rates
 MenWomen
199419992004% Change 1999/1994% Change 2004/1999199419992004% Change 1999/1994% Change 2004/1999
EUROPE
 France2.242.322.013.64−13.281.021.030.990.73−3.74
 Germany3.882.992.48−22.86−17.152.612.372.16−9.02−8.82
 Italy (2003)4.313.572.90−17.19−18.562.682.282.63−14.6215.03
 Poland6.715.424.44−19.18−18.123.613.072.66−14.99−13.36
 Romania7.368.055.449.44−32.513.292.902.73−11.80−5.87
 Spain5.423.733.35−31.07−10.412.952.361.99−19.90−15.82
 United Kingdom1.991.611.49−19.04−7.551.200.961.34−20.1840.33
 EU4.253.652.99−14.01−18.052.522.232.10−11.74−5.73
 Russian Federation16.7213.4610.14−19.48−24.678.397.206.15−14.23−14.46
 Ukraine14.6212.6911.22−13.22−11.557.456.676.41−10.44−3.93
 Uzbekistan6.706.025.39−10.10−10.494.174.133.90−1.14−5.43
AMERICA
 Argentina4.043.773.67−6.72−2.741.812.341.9529.82−16.73
 Brazil4.624.474.17−3.20−6.752.452.422.33−1.17−3.75
 Chile7.887.936.250.67−21.193.173.583.3912.67−5.08
 Colombia7.096.655.72−6.09−14.063.893.513.99−9.7513.63
 USA1.711.571.35−8.15−13.950.890.950.986.872.48
AFRICA, ASIA, OCEANIA
 Japan9.417.965.70−15.43−28.458.907.295.41−18.05−25.78
 Republic of Korea16.6912.199.74−26.95−20.0911.289.647.82−14.55−18.86
 Australia (2003)1.951.571.74−19.3810.901.171.061.37−9.1829.52

The main findings from joinpoint regression over the period 1980–2005 in the 19 largest countries and in the EU as a whole are given in Table IV and Figure 2. Trends were significantly downward in all countries and for both sexes, with, at the beginning of this century, values of APC around −3 to −4% for the major European countries, with the only exception of Romania, for which the decline was smaller (APC = −1%) over recent years. Similar trends were evident for the Russian Federation (−2.8% from 1997 in men), the Republic of Korea (APC = −4.3% from 1993), Australia (−3.7% from 1980), the USA (−3.6% from 1991), Japan (−3.5% from 1996) and Ukraine (−3% from 1991). In Latin America, the decline was steady but less marked, with APC ranging between −1.6 and −1.7% in Chile and Brazil and between −2.3 and −2.6% in Argentina and Colombia from the mid-1980s, and −2.3% from the end of the 1980s for Mexico. For most countries, the APC were similar in women.

Figure 2.

Joinpoint analysis for gastric cancer in selected countries at all ages (men and women), 1980–2005 (when available). Men ♂__♂; women ♀__♀.

Table IV. Joinpoint Analysis for Gastric Cancer in Selected Countries (Men and Women, All Ages), 1980–20051
 MenWomen
Trend 1Trend 2Trend 3Trend 4Trend 1Trend 2Trend 3Trend 4
YearsAPCYearsAPCYearsAPCYearsAPCYearsAPCYearsAPCYearsAPCYearsAPC
  • 1

    When available.

  • 2

    Significantly different from 0 (p < 0.05).

  • APC: Estimated annual percent change.

EUROPE
 France1980–1988−3.1321988–1992−4.9621992–2005−2.772  1980–1996−4.4321996–2005−2.642    
 Germany1980–1985−4.1021985–1995−3.2921995–1998−5.8821998–2004−4.3221980–1994−3.3721994–2004−4.312    
 Italy1980–2003−3.882      1980–2003−2.932      
 Poland1980–1987−2.6521987–2005−3.402    1980–2005−3.452      
 Romania1980–1984−5.7921984–2004−1.072    1980–1982−8.9721982–1989−3.2721989–2004−1.182  
 Spain1980–1982−6.4921982–1987−1.6821987–1997−2.7521997–2005−3.7621980–2005−3.542      
 UK1980–1988−2.5821988–2005−4.722    1980–2005−4.222      
 EU1980–1994−2.9621994–1997−5.5821997–2000−1.482000–2004−4.7721980–2004−3.462      
 Russian Federation1980–1993−2.2321993–1997−3.9721997–2005−2.842  1980–1990−2.3621990–2005−3.382    
 Ukraine1981–1991−0.7921991–2005−3.002    1981–1992−1.1321992–1996−4.9221996–2000−1.672000–2005−3.682
 Uzbekistan1981–1988−0.551988–2000−6.0022000–20051.16  1981–1988−1.191988–1999−5.8221999–20051.78  
OTHER COUNTRIES
 Argentina1982–19850.981985–2004−2.302    1982–2004−3.062      
 Brazil1980–2004−1.732      1980–1985−3.7321985–2004−1.902    
 Chile1980–2004−1.642      1980–1982−9.791982–2004−2.362    
 Colombia1984–2004−2.642      1984–1993−2.9221993–1997−7.071997–2004−1.53  
 Mexico1981–19891.4221989–2005−2.332    1981–1989−0.131989–2005−2.382    
 USA1980–1988−2.2221988–1991−0.191991–2005−3.612  1980–1984−1.9721984–1987−4.301987–19900.381990–2005−2.542
 Japan1980–1993−3.2921993–1996−0.621996–2005−3.472  1980–1990−4.4221990–2005−3.572    
 Republic of Korea1985–1990−4.2721990–19932.211993–2005−4.312  1985–1990−5.2821990–19932.001993–2005−5.442  
 Australia1980–2003−3.742      1980–2003−3.322      

Among the 19 countries considered, only in Uzbekistan there was a nonstatistically significant rise in the most recent years considered (2000–2005) for both sexes, despite a marked and significant decline in the previous decade.

The Supporting Information Appendix Figure 1 shows gastric cancer incidence rates in 51 selected cancer registries worldwide by anatomical subsite and gender. Incidence rates in men were higher than those in women in all the registries, but the geographical pattern was similar in the 2 sexes. Only a few registries reported incidence rates for gastric cancer as a whole higher than 30/100,000 men. These were Korea (66.1/100,000 men, 26 women), followed by Nagasaki (59.5/100,000 men, 22.3 women) and Osaka, Japan (51.5/100,000 men, 19.8 women), Valdivia, Chile (43.1/100,000 men, 15.9 women), Belarus (35.8/100,000 men, 15.2 women), Shanghai, China (34.2/100,000 men, 17.3 women), Costa Rica (33.1/100,000 men, 17.0 women) and St. Petersburg, Russia (32.6/100,000 men, 16.0 women). High gastric cancer incidence rates (from 25.1 to 29.5/100,000 in men and from 10.8 to 16.6 in women) were also reported from Lithuania and Estonia, in Eastern Europe, in Italy (Romagna Region) for Southern Europe, and from Brazil (Sao Paolo), Colombia (Cali), Peru (Trujillo), in the Americas. The lowest gastric cancer incidence rates worldwide (around 7/100,000 men and 3/100,000 women) were in Denmark, Sweden, Vaud, Switzerland and USA, SEER (14 Registries) with however higher values in Blacks.

Noncardia gastric cancers constitute the majority of stomach cancer cases worldwide. For cardia cancer, the incidence was roughly homogeneous all over the world, with values ranging from 0.2 to 5.6/100,000 men and from 0.1 to 1.6 women. In contrast, fundus incidence rates ranged from less than 2 to 19/100,000 men and from less than 1 to 8.1 women (but only 3 registries showed incidences greater than 10/100,000 men and 4.5 women). For pylorus, the range was from less than 2 to 22/100,000 men and from less than 1 to 8.2 women (with only 2 registries with rates greater than 8/100,000 men and 4.5 women). For the other subsites, lesser and greater curvature account for a few cases in all the registries considered, and only 4 registries (Ireland, Korea, Shanghai [China] and Nagasaki [Japan]) reported about 2/100,000 for men and 1 for women. However, in several registries, mainly for Asia and South America, over 50% of gastric cancers were from other and unspecified origin.

Cancers in the fundus and pylorus were more common in high gastric cancer incidence areas, in particular the highest rates were observed in Korea (13.5/100,000 for fundus and 22/100,000 for pylorus cancer), Nagasaki, Japan (19/100,000 for fundus and 15.2/100,000 for pylorus), Belarus (11.3/100,000 for fundus and 7.4/100,000 for pylorus) and Costa Rica (7.6/100,000 for pylorus).

Discussion

This work summarizes most available data on descriptive epidemiology of gastric cancer mortality worldwide. A limitation is the paucity of mortality data for Asia, and the availability of only one country in Africa, which limits the generalizability of the results.

The major finding from this analysis of recent trends in gastric cancer mortality is the presence of persisting and steady declines—in relative terms—in Europe as well as in most other areas of the world. The falls were still observed in countries with very low rates, such as Sweden or Switzerland in Europe, the USA and Australia, and were evident in middle-aged and young adults too, indicating that the downward trends are likely to persist in the future.

This is confirmed by joinpoint regression analysis, whose key message is that the fall in gastric cancer shows no evidence of leveling off in most Europe but also in other areas of the world. The estimated APC provided by the joinpoint analysis, however, is a measure of the proportional fall in gastric cancer death rates. Since gastric cancer mortality has long been declining, the absolute fall is now smaller and will become even smaller in the future. During the 1990s gastric cancer, together with colorectum and lung, was the major contributor to the fall in overall cancer mortality in Europe, with over 30,000 deaths avoided per year, as compared to the rates in the late 1980s in the 15 countries of the EU (defined as before 20045, 17). In the same countries, the absolute fall corresponds to less than 20,000 deaths avoided per year over the last years considered, due to lower underlying mortality rates.

At age 30–49 years, gastric cancer mortality has stopped falling among women in a few countries, including France, UK, the USA and Australia, with rates around 1/100,000. Those women were born around 1960. That generation may therefore represent the first one for which an asymptote in female gastric cancer mortality has been reached in selected countries.

There were, however, several areas of the world, including not only Eastern Asia and South America, but also most Central and Eastern Europe,5, 18 where higher incidence and mortality rates from gastric cancer were still observed.

Gastric cancers include adenocarcinomas of the fundus and corpus and distal stomach (pyloric antrum and pylorus), whose incidence has long been falling, but also adenocarcinomas of the cardia.19, 20 The latter are influenced by reflux disease and, like adenocarcinoma of the lower oesophagus, have been rising in Europe and mostly North America over the last 2 decades.21–24 Despite these recent trends, the geographic variation of gastric cardia cancer is smaller than that of other gastric subsites. The incidence of other rarer gastric cancer morphologies such as gastro–intestinal stromal tumors or lymphomas, have not been declining. The most recent incidence data reported in this article confirm that high incidence areas of gastric cancer are mainly related to high incidence of fundus and pylorus cancers.

Apart from problems of classification across subsites and histological types, diagnosis and certification of gastric cancers has long been sufficiently reliable to allow valid inference on trends for most countries from Europe, but also from America and few Asia providing data.1, 5 The consistency of trends at all ages and truncated at age 35–64 years is an additional indicator of the validity of gastric cancer death certification. It is in any case unlikely that any change in diagnosis and certification of gastric cancer explains more than a minor part of the widespread and substantial downward trends observed in gastric cancer mortality over the last decades.3, 25

The reasons for the generalized decline in gastric cancer rates are complex and not completely understood. Almost certainly, these include a more varied and affluent diet and better food conservation, including refrigeration, as well as the control of Helicobacter pylori (H. pylori) infection.

Throughout the world there is a strong and consistent correlation between consumption of salt and salted foods and stomach cancer incidence.26–32 A large number of studies that have examined this relationship have generally found an increased risk by approximately 2-fold for frequent consumption of salt and salted foods. The relationship is also biologically plausible, given that salt may lead to damage of the protective mucosal layer of the stomach. Other methods of food preservation, including curing and smoking foods, have also been associated with stomach cancer, although the evidence is not consistent.

An increased risk of gastric cancer is associated with H. pylori. The plausibility of a causal association is also supported by a strong association between H. pylori and precancerous lesions, including chronic and atrophic gastritis and dysplasia. Given that the prevalence of infection is very high, especially in developing countries and among older cohorts, H. pylori can explain over 50% of all new cases of gastric cancer, and over 5% of all cancer cases globally.33 There are, however, still some uncertainties regarding the quantification of this association in various populations and age groups. For instance, in Japan the fall in H. pylori prevalence between 1986 and 1996 was larger at younger age groups (20–39 years), though gastric cancer rates had long been declining.34 Different strains of H. pylori, moreover, have different carcinogenic potential. In fact, the presence of H. pylori DNA in gastric biopsies, the severity of precancerous lesions, and hence gastric cancer risk, appear to be specific of cagA positive strains.35–38

Another important cause of stomach cancer is tobacco smoking. Smokers have a 50–60% increased risk of stomach cancer, as compared to nonsmokers. This would indicate that smoking is responsible for approximately 10% of all cases,39 and the recent decline in smoking prevalence in men from developed countries may account for part of the fall in gastric cancer rates.40

The high prevalence of H. pylori in subsequent cohorts of the populations, several aspects of foods conservation, diet and nutrition as well as tobacco smoking partly or largely explain the smaller fall in gastric cancer rates in Eastern as compared to Western Europe and other areas of the developed world. In Russia and other former nonmarket economy countries, refrigeration was available later, there were large deficiencies in food distribution and availability, and tobacco smoking was more common, mainly in men.41

In conclusion, despite the encouraging trends on mortality, stomach cancer remains one of the major causes of death worldwide but further declines are likely. Primary prevention of stomach cancer by dietary means is feasible by encouraging high-risk populations to decrease consumption of cured meats and salt preserved foods, and increase the consumption of vegetables and fruit.42–47 Prevention may also be feasible through eradication of H. pylori infection, particularly in childhood and adolescence.

Whether improved diagnosis and treatment has also played some role on the favorable trends in gastric cancer, particularly over most recent calendar periods, remains open to quantification.3, 48 Screening and early detection of stomach cancer have been developed in Japan with use of X-ray photofluorography to identify early lesions, followed by gastroscopy, and may also be responsible for part of the decline in mortality, mostly in Japan but also in other developed areas of the world.49 In Japan, there were also advancements in surgical treatment of early disease, with consequent improved survival rates as compared to most other developed countries.50–52

Acknowledgements

The authors thank Ms. I. Garimoldi for editorial assistance. This work was conducted with contributions from the Italian Association for Cancer Research (AIRC), the Italian and Swiss Leagues Against Cancer, the Swiss Foundation for Research Against Cancer. PB was supported by a fellowship from the Italian Foundation for Cancer Research (FIRC). This work was undertaken while CLV was a senior fellow at the International Agency for Research on Cancer.

Ancillary