Annals of Oncology Advance Access originally published online on February 21, 2008
Annals of Oncology 2008 19(5):1009-1018; doi:10.1093/annonc/mdm593
epidemiology |
Changes in breast cancer incidence and mortality in middle-aged and elderly women in 28 countries with Caucasian majority populations
Data Analysis and Interpretation Group and Epidemiology Methods and Support Group, International Agency for Research on Cancer, Lyon, France
* Correspondence to: Dr P. Autier, International Agency for Research on Cancer, 150 Cours Albert Thomas, F-69372 Lyon Cedex 08, France. Tel: +33-472-73-81-63; Fax: +33-472-73-83-51; E-mail: autierp{at}iarc.fr
 |
Abstract |
|---|
 Top Abstract introductionmethodsresultsdiscussionfundingAcknowledgementsReferences |
|---|
Background: Mammography screening and menopause hormone therapy is essentially offered to women 50–69 years old.
Methods: In 28 European ancestry countries, we quantified changes in breast cancer incidence and mortality using a joinpoint regression analysis from 1960 until last year of available data.
Results: Since 1960, increases in incidence often in the order of 2%–3% per year occurred in all countries, mainly in women 50–69 years old whose incidence in eight countries surpassed the incidence in women 70 years old and more. In 10 countries, a decrease in incidence in women 
70 years was noticeable in the last years of observation, but the magnitude of this decrease was far from matching the magnitude of the increases observed in the 50–69 age-group. In the beginning of years 2000s, a persistent decrease in mortality of 
2% per year was observed in women 50–69 years old in most countries and parallel declines in mortality were observed in women 70 years or more.
Conclusions: In years 2000s, in a number of countries, the incidence of breast cancer has become greater in middle-aged women than in older women. If trends remain unchanged, the same phenomenon is likely to happen in other countries.
Key words: breast cancer, epidemiology, temporal trends
|
introduction |
|---|
|
TopAbstractintroductionmethodsresultsdiscussionfundingAcknowledgementsReferences |
|---|
Breast cancer is the first cause of cancer related-death in women 45 years old and more in the world with 915 000 new cases and >335 000 deaths in 2002 [1].
Since 1985, considerable changes in breast cancer detection and management have occurred [2], with the availability of efficient hormonal and chemotherapeutic adjuvant and neo-adjuvant treatments for both pre- and postmenopausal women, implementation of mammography screening, introduction of breast-conserving surgery and sentinel node procedures, new indications for radiotherapy and the advent of multidisciplinary management of breast cancer in specialized centers.
In most European countries, breast cancer screening was made available mainly to women 50–69 years old. Treatments have been different for breast cancer occurring in pre- or postmenopausal women, and in 1990, hormonal replacement therapy (HRT) became widely recommended for postmenopausal women [3].
Description of incidence or mortality trends across a long period of time could inform on how changes in breast cancer detection and management may have affected the burden of this disease in countries with mainly European ancestry. This paper describes and quantifies changes in breast cancer incidence and mortality in women 50 years old and more in 28 countries in Europe (For the text clarity, England and Wales on one hand and Scotland on the other hand were considered as two countries), North America, Australia, New Zealand and Israel for which both incidence and mortality data were available.
|
methods |
|---|
|
TopAbstractintroductionmethodsresultsdiscussionfundingAcknowledgementsReferences |
|---|
data sources
We considered 28 countries with mainly European ancestry for which incidence data of invasive breast cancer were available in successive volumes of Cancer Incidence in Five Continent [4] (Table 1). Incidence data were not available for seven European countries where there is neither national nor subnational cancer registry with sufficient historical series (Belgium, Greece, Hungary, Luxembourg, Poland, Portugal and Romania). Incidence and mortality data from Malta were not included in the analysis because of small population size and short period of cancer registration.
|
More recent years for cancer registries in Europe were extracted from the EUROCIM database of the European Network of Cancer Registries [5] and from the NORDCAN database of the Association of Nordic Cancer Registries [6]. For the United States, additional data were obtained from the Surveillance, Epidemiology and End Results (SEER) nine registry database [7]. Registry data were passed through the multiple primary check program used by International Agency for Research on Cancer (IARC). The SEER definition of a ‘second primary’ cancer is more liberal than that used by the IARC: in SEER data, counting of second cancer cases is time dependent and bilateral tumors are counted twice. Also classification of tumors can be made in more histological groups than in the IARC system, and two breast cancers occurring in the same patient may be counted as two different cancers according to SEER but as only one cancer according to IARC. Additional information was found in various national cancer registry reports available through the Internet [8–18].
For countries with no national cancer registration system (e.g. France, Germany, Italy, Spain and Switzerland) or for which the national data do not cover a sufficient time period (United States of America and Australia), local cancer registries were aggregated to obtain estimates of national incidence (Table 1).
We extracted breast cancer mortality data from the World Health Organization database [19] also available at IARC Web site [20] for the 28 countries with available incidence data. Changes in the classification or in the codification process of death certificates may affect mortality trends. For instance, the change from International Classification of Diseases (ICD)-8 to ICD-10 in 1995 in Switzerland caused discontinuities in trends and the Swiss data before 1995 have been corrected accordingly [21]. In France, the effect of the introduction of the ICD-10 on mortality statistics has been studied [22] and mortality data before 2000 have been adjusted. A similar study was conducted in England and Wales [23] and corresponding data before 2000 and after 1993 have been corrected. Corrections for changes in classification brought only small changes to trends.
In UK, incidence data were available for England (without Wales) and Scotland, but mortality data were available for England and Wales and for Scotland. For the sake of text clarity, England with or without Wales on one hand and Scotland on the other hand were considered as two countries.
statistical analysis
We computed the age-specific rates for each 5-year age-group and calendar year and estimated the age-standardized rate (ASR) of incidence and mortality rates per 100 000 person-years of two different age-groups (50–69 and 70 years of age), using the World Standard Population as defined by Segi [24]. A systematic analysis of temporal trends was done with joinpoint regression using the software from the Surveillance Research Program of USA National Cancer Institute [25]. The joinpoint regression was carried out for the data since 1960 or on earliest data after 1960, until the last year of available data. Joinpoint regression can identify periods with distinct linear, or log-linear, trends of rates that can be separated by inflexion points called joinpoint. Annual percentage changes (APC) were also estimated for each period and statistically tested for the difference with a slope of zero. An APC was labeled as statistically significant when its 95% confidence interval did not include zero. The method follows the principle of minimization of the weighted sum of squared errors and the choice of the number of joinpoints is on the basis of permutation tests [26, 27]. A maximum of three joinpoints has been allowed for each model fitted. Joinpoint regression was defined as a log-linear model of the rate with heteroscedastic errors of the count based on a Poisson distribution. Because available period of interest for Ireland was short, we used a minimum of one point between two joinpoints.
|
results |
|---|
|
TopAbstractintroductionmethodsresultsdiscussionfundingAcknowledgementsReferences |
|---|
Table 1 displays the available data in the 28 countries and the ASR, incidence and mortality rates at the beginning and at the end of the observed period. Table 2 displays the quantitative parameters relative to temporal trends derived from joinpoint analysis, i.e. years of joinpoint inflexion and APC that were used for drawing incidence and mortality curves.
|
incidence
In the majority of the 28 countries (Table 2 and Figure 1), breast cancer incidence was increasing since incidence data are recorded. Increases in incidence rates have essentially concerned the 50–69 age-group with high APC
3% found in Belarus, Finland, Ireland, Norway, Sweden, England, Scotland, Germany, The Netherlands, United States of America, Australia and Israel. For this age-group, eight countries were characterized by an important raise in incidence during 1980–1995, followed by a lower but still increasing incidence rate (APC usually <1.5%). The incidence in the 50–69 age-group has reached or has surpassed the incidence rates in the 70+ age-group in 1997 in Spain, Iceland and Norway, in 1998 in Australia, in 1999 in Finland, in 2000 in Sweden and Germany and in 2002 in Ireland. If trends remain unchanged over next years, the same phenomenon is likely to happen in six others countries (Italy, Slovenia, Austria, France, Switzerland and New Zealand).
|
In the majority of countries, the incidence increase in the 70+ age-group was of lower magnitude, with declining trends at the end of the registration period in Latvia, Slovenia, Spain, Norway, Austria, Germany, Canada, United States of America, Australia and Israel. However, because numbers of women 70 years old or more represent
55%–65% of women 50–69 years old, the decreases in the 70+ age-group were far from matching the increases observed in the 50–69 age-group. For instance, compared with 1990, the incidence in the 50–69 years group in Norway in 2005 corresponded to 620 more breast cancer cases, while the decline in women 70 years old and more in 2005 corresponded to only 35 breast cancer cases. In Australian areas covered by registries, compared with 1990, there were 376 more breast cancer cases in women 50–69 years old in 2002 and only seven less breast cancer cases in women 70 years old and more. In the United States of America, after a continuous increase since 1980, an abrupt significant inversion of trends occurred in 2001 with a significant decline of trends until the end of the period for the two age-groups. Similar (but nonsignificant) swift declines in incidence among women 50 years old and more is noticeable in Israel after 1998.
mortality
In the majority of countries, mortality from breast cancer was on the rise until the end of the 1980s—beginning of the 1990s (Table 2), except in Sweden, where mortality in the 50–69 age-group decreases since 1960. From 1978 (Norway) until 1996 (France), mortality rates in the 50–69 age-group started to substantially decrease in 15 countries. Declines were most important in this age-group than in women 70 years old and more, but they started about the same year in both age-groups. Decreases in mortality occurred either after a period of gradual levelling off or abruptly from 1 year to another. In Norway and in UK, decrease in mortality started about the same year when steep raises in incidence rates occurred. In several Central and Eastern European countries, mortality remained steady or was still rising in both age-groups (Belarus, Estonia and Latvia).
|
discussion |
|---|
|
TopAbstractintroductionmethodsresultsdiscussionfundingAcknowledgementsReferences |
|---|
The main aim of this study was to describe and quantify breast cancer changes in several countries during a long period of time for postmenopausal women. Some studies of time trends in Caucasian population were published [28, 29] but they did not describe trends over a period of time long enough for assessing changes in breast cancer detection and management that took place after 1985 [30].
While the age–period–cohort model is widely used to examine temporal trends and to assess and estimate the effects of three factors on the rates (age, period and cohort) [31, 32], modifications in temporal trends and intensity of changes over time are best described by joinpoint analysis. The joinpoint analysis performs a statistical evaluation of changes over time, indicating when they occured and leaving no room for subjective assessment of trends. Some apparent inconsistencies in results could, however, not be avoided: for instance, small fluctuations in mortality rates observed in Austria and in Sweden 1970 were probably due to changes in coding practice of information from death certificates.
Our results point out that in many countries, the likelihood of women 50–69 years old diagnosed with a breast cancer has become similar or higher than for women 70 years of age or more. The marked increases in women 50–69 years old were often followed by slowdowns, but never by returns to incidence rates prevailing before the raises. The decreases in incidence sometimes observed in women 70 years old or more were far from compensating the increases in numbers of breast cancers detected in 50–69 years old women. In Belarus, Latvia and Lithuania, under registration of breast cancer in women >70 years old may explain that women 50–69 old had a higher incidence than women 70 years old and more [33].
In high resource countries and in countries knowing sustained economic growth, there is homogenization of lifestyle. Numbers of breast cancer patients has profoundly changed over time because of changes in reproductive factors, in breast cancer awareness and in access to mammography screening. For instance, from 1980 until 2000, we quantified an increasing of number of breast cancer patients of 47% in Sweden, 55% in Czech Republic, 60% in England and Wales and 56% in Norway.
The marked raises in slopes of incidence rates curves in women 50–69 years old corresponded to rapid nationwide implementation of mammography screening. Menopause hormonal therapy is another factor that has contributed to increasing the breast cancer incidence after menopause [34, 35], but its influence was highly variable according to the levels of use. For instance, HRT use was high in United States of America and Denmark but low in Italy [36]. Decrease of breast cancer incidence in women 50 years old and more is observed in United States since 2001. This decline is deemed to be due to the reduction in use of hormone therapy among postmenopausal women that occurred following the publication of the results from the Women Health Initiative study [37].
Few cancer registries published data on breast cancer incidence according to the stage at diagnosis. These data show that before 1980, the increase in incidence already mainly concerned localized breast cancers [38, 39]. After 1980, in Australia, The Netherlands, UK and United States of America, most of the increase in invasive breast cancer incidence consisted in small, localized breast cancer usually of good prognosis (i.e. <T2N0M0) [40–44]. In contrast, incidence rates of more advanced breast cancers (T2 or N1 or M1 and TXN1) remained stable or showed moderate declines in the order of 10%–15% 8 years or more after start of mammography screening activities. These data indicate that the considerable increase in incidence observed in most countries essentially concerned breast cancers of low malignant potential, a proportion of which would have probably never become life threatening.
Most recent mortality data show that the decrease in breast cancer mortality that started in many countries in the 1990s [28, 29] persisted and even seemed to amplify in the beginning of years 2000s with a steep decrease of 2%–3% per year in 15 countries [45]. The changes in mortality trends taking place about the same year across age categories are strongly indicative of a period effect such as the implementation of more effective management of breast cancer. Differences among countries in the generalization of effective management may, however, explain differences in mortality changes. In several Central and Eastern European countries, increasing or limited decreases in mortality could still be due to economical reasons that would reduce access to screening and treatments [46].
|
funding |
|---|
|
TopAbstractintroductionmethodsresultsdiscussionfundingAcknowledgementsReferences |
|---|
European Commission (Directorate of SANCO, Luxembourg, Grand Duchy of Luxembourg), for the European Network for Information in Cancer Epidemiology project; Cancéropôle Lyon Auvergne Rhône-Alpes, France.
|
Acknowledgements |
|---|
|
TopAbstractintroductionmethodsresultsdiscussionfundingAcknowledgementsReferences |
|---|
The authors declare no conflict of interests.
Received for publication October 31, 2007. Revision received December 11, 2007. Accepted for publication December 11, 2007.
|
References |
|---|
|
TopAbstractintroductionmethodsresultsdiscussionfundingAcknowledgementsReferences |
|---|
1. Ferlay J, Bray F, Pisani P, et al. GLOBOCAN 2002: Cancer Incidence, Mortality and Prevalence Worldwide (2004) Lyon, France: IARC Press.
2. Veronesi U, Boyle P, Goldhirsch A, et al. Breast cancer. Lancet (2005) 365:1727–1741.[CrossRef][Web of Science][Medline]
3. Stefanick ML. Estrogens and progestins: background and history, trends in use, and guidelines and regimens approved by the US Food and Drug Administration. Am J Med (2005) 118(Suppl 12B):64–73.[CrossRef][Medline]
4. Parkin DM, Whelan S, Ferlay J, et al. Cancer Incidence in Five Continents, vol. I to VIII [IARC CancerBase No. 7, Lyon, 2005.] (2005) Lyon, France: IARC Press.
5. Eurocim Version 4.0. European Incidence Database. European Network of Cancer Registries [v2.4, ICD-10 Dictionary] (2003) Lyon, France: IARC Press.
6. Engholm G, Storm H, Ferlay J, et al. NORDCAN: Cancer Incidence and Mortality in the Nordic Countries [Version 2.2] (2006) Copenhagen, Denmark: Danish Cancer Society.
7. Surveillance, Epidemiology, and End Results (SEER) Program. Public-Use Data (1973–2004) (2007) National Cancer Institute,DCCPS,Surveillance Research Program,Cancer Statistics Branch USA.
8. New Zealand Health Information Service. The New Zealand Cancer Register. http://www.nzhis.govt.nz (June 2006, date last accessed).
9. Ondrusova M, Plesko I, Safaei Diba C. Cancer Incidence in the Slovak Republic 2003 (2006) Bratislava, Slovakia: National Health Information Centre. 210pp.
10. Austria: Statistik Austria. http//:www.statistik.at (June 2006, date last accessed).
11. Scottish Cancer Registry. http://www.isdscotland.org (June 2006, date last accessed).
12. Cancer Registry of Bulgaria. http://www.onco-bg.com (June 2006, date last accessed).
13. England and Wales. UK: National Statistics. http://www.statistics.gov.uk (June 2006, date last accessed).
14. Cancer Registry of Norway. Cancer Incidence in Norway. http://www.kreftregisteret.no (June 2006, date last accessed).
15. National Cancer Registry of Ireland. http://www.ncri.ie/ncri (June 2006, date last accessed).
16. Association of Comprehensive Cancer Centres. Utrecht: The Netherlands. http://www.ikcnet.nl (June 2006, date last accessed).
17. Epidemiology of Malignant Tumors in the Czech Republic. National Cancer Registry of the Czech Republic. http://www.svod.cz (June 2006, date last accessed).
18. Hertl K, Primic-Zakelj M, Zgajnar J, et al. Performance of opportunistic breast cancer screening in Slovenia. Neoplasma (2006) 53:237–41.[Medline]
19. WHO Statistical Information System. WHO Databank (2007) World Health Organization. http://www.who.int/whosis (August 2006, date last accessed).
20. International Agency for Cancer Research. IARC. http://www.iarc.fr (June 2006, date last accessed).
21. Lutz JM, Pury P, Fioretta G, et al. The impact of coding process on observed cancer mortality trends in Switzerland. Eur J Cancer Prev (2004) 13:77–81.[CrossRef][Web of Science][Medline]
22. Pavillon G, Boileau J, Renaud G, et al. Conséquences des changements de codage des causes médicales de décès sur les données nationales de mortalité en France, à partir de l'année 2000. BEH (2005) 4:13–16.
23. Brock A, Griffiths C, Rooney C. The effect of the introduction of ICD-10 on cancer mortality trends in England and Wales. Health Stat Q (2004) 23:7–17.[Medline]
24. Segi M. Cancer mortality for Selected Sited in 24 Countries (1950–57) (1960) Tohoku University, Sendai, Japan.
25. National Cancer Institute. Joinpoint Regression Program,version 2.7. http://srab.cancer.gov/joinpoint/ (August 2006, date last accessed).
26. Lerman PM. Fitting segmented regression models by grid search. Appl Stat (1980) 29:77–84.[CrossRef]
27. Kim HJ, Fay MP, Feuer EJ, et al. Permutation tests for joinpoint regression with applications to cancer rates. Stat Med (2000) 19:335–351.[CrossRef][Web of Science][Medline]
28. Hermon C, Beral V. Breast cancer mortality rates are levelling off or beginning to decline in many western countries: analysis of time trends, age-cohort and age-period models of breast cancer mortality in 20 countries. Br J Cancer (1996) 73:955–960.[Web of Science][Medline]
29. Botha JL, Bray F, Sankila R, et al. Breast cancer incidence and mortality trends in 16 European countries. Eur J Cancer (2003) 39:1718–1729.[CrossRef][Web of Science][Medline]
30. IARC Handbooks of Cancer Prevention. Breast Cancer Screening. IARC Press 2002.
31. Clayton D, Schifflers E. Models for temporal variation in cancer rates. I: age-period and age-cohort models. Stat Med (1987) 6:449–467.[Web of Science][Medline]
32. Clayton D, Schifflers E. Models for temporal variation in cancer rates. II: age-period-cohort models. Stat Med (1987) 6:469–481.[Web of Science][Medline]
33. Parkin DM, Whelan S, Ferlay J, et al. Cancer Incidence in Five Continents, vol. VII (2002) Lyon, France: IARC Scientific Publication No 143.
34. National Center for Health Statistics. Health, United States, 2005 with Chartbook on Trends in the Health of Americans (2005) Hyattsville, Maryland. http://www.cdc.gov/nchs/hus.htm. (August 2006, date last accessed).
35. Chlebowski RT, Hendrix SL, Langer RD, et al. Influence of estrogen plus progestin on breast cancer and mammography in healthy postmenopausal women: the Women's Health Initiative Randomized Trial. JAMA (2003) 289:3243–3253.
36. Jolleys JV, Olesen F. A comparative study of prescribing of hormone replacement therapy in USA and Europe. Maturitas (1996) 23:47–53.[CrossRef][Web of Science][Medline]
37. Ravdin PM, Cronin KA, Howlader N, et al. The decrease in breast-cancer incidence in 2003 in the United States. N Engl J Med (2007) 356:1670–1674.
38. De Bono AM, Pillers EM. Carcinoma of the breast in East Anglia 1960–1975: a changing pattern of presentation? J Epidemiol Community Health (1978) 32:178–182.
39. Joensuu H, Toikkanen S. Comparison of breast carcinomas diagnosed in the 1980s with those diagnosed in the 1940s to 1960s. BMJ (1991) 303:155–158.
40. McCann J, Stockton D, Day N. Breast cancer in East Anglia: the impact of the breast screening programme on stage at diagnosis. J Med Screen (1998) 5:42–48.
41. Schootman M, Jeffe D, Reschke A, et al. The full potential of breast cancer screening use to reduce mortality has not yet been realized in the United States. Breast Cancer Res Treat (2004) 85:219–222.[CrossRef][Medline]
42. Kricker A, Farac K, Smith D, et al. Breast cancer in New South Wales in 1972–1995: tumor size and the impact of mammographic screening. Int J Cancer (1999) 81:877–880.[CrossRef][Medline]
43. Fracheboud J, Otto SJ, van Dijck JA, et al. Decreased rates of advanced breast cancer due to mammography screening in The Netherlands. Br J Cancer (2004) 91:861–867.[Web of Science][Medline]
44. Anderson WF, Jatoi I, Devesa SS. Assessing the impact of screening mammography: breast cancer incidence and mortality rates in Connecticut (1943–2002). Breast Cancer Res Treat (2006) 99:333–340.[CrossRef][Web of Science][Medline]
45. Tyczynski JE, Berkel HJ. Long-term all-sites cancer mortality time trends in Ohio, USA, 1970–2001: differences by race, gender and age. BMC Cancer (2005) 5:136.[CrossRef][Medline]
46. Tyczynski JE, Plesko I, Aareleid T, et al. Breast cancer mortality patterns and time trends in 10 new EU member states: mortality declining in young women, but still increasing in the elderly. Int J Cancer (2004) 112:1056–64.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||