Cancer incidence and mortality in France over the 1980–2012 period: Solid tumors

Available online at

ScienceDirect www.sciencedirect.com Revue d’E´pide´miologie et de Sante´ Publique 62 (2014) 95–108

Original article

Cancer incidence and mortality in France over the 1980–2012 period: Solid tumors Incidence et mortalite´ des cancers en France durant la pe´riode 1980–2012 : tumeurs solides F. Binder-Foucard a,b,*,c, N. Bossard d,e,f, P. Delafosse a,g, A. Belot d,e,f,h, A.-S. Woronoff a,i, L. Remontet d,e,f the French network of cancer registries (Francim) b

a Re´seau franc¸ais des registres des cancers, Francim, 31073 Toulouse, France Registre des cancers du Bas-Rhin, laboratoire d’e´pide´miologie et de sante´ publique, EA 3430, faculte´ de me´decine, universite´ de Strasbourg, 4, rue Kirschleger, 67085 Strasbourg cedex, France c Service de sante´ publique, hoˆpitaux universitaires de Strasbourg, 67000 Strasbourg, France d Service de biostatistique, hospices civils de Lyon, centre hospitalier Lyon-Sud, 69424 Lyon, France e Universite´ Lyon-I, 69622 Villeurbanne, France f UMR 5558, laboratoire biostatistique sante´, CNRS, 69495 Pierre-Be´nite cedex, France g Registre du cancer du de´partement de l’Ise`re, CHU de Grenoble, 38000 Grenoble, France h De´partement des maladies chroniques et traumatismes, Institut de veille sanitaire, 94410 Saint-Maurice, France i Registre des tumeurs du Doubs et du Territoire de Belfort, EA 3181, CHRU de Besanc¸on, 25030 Besanc¸on cedex, France

Received 28 June 2013; accepted 24 November 2013 Available online 7 March 2014

Abstract Background. – Cancer incidence and mortality estimates for 19 cancers (among solid tumors) are presented for France between 1980 and 2012. Methods. – Incidence data were collected from 21 local registries and correspond to invasive cancers diagnosed between 1975 and 2009. Mortality data for the same period were provided by the Institut national de la sante´ et de la recherche me´dicale. The national incidence estimates were based on the use of mortality as a correlate of incidence. The observed incidence and mortality data were modeled using an age-period-cohort model. The numbers of incident cases and deaths for 2010–2012 are the result of short-term projections. Results. – In 2012, the study estimated that 355,000 new cases of cancer (excluding non-melanoma skin cancer) and 148,000 deaths from cancer occurred in France. The incidence trend was not linear over the study period. After a constant increase from 1980 onwards, the incidence of cancer in men declined between 2005 and 2012. This recent decrease is largely related to the reduction in the incidence of prostate cancer. In women, the rates stabilized, mainly due to a change in breast cancer incidence. Mortality from most cancer types declined over the study period. A combined analysis of incidence and mortality by cancer site distinguished cancers with declining incidence and mortality (e.g., stomach) and cancers with increasing incidence and mortality (e.g., lung cancer in women). Some other cancers had rising incidence but declining mortality (e.g., thyroid). Conclusion. – This study reveals recent changes in cancer incidence trends, particularly regarding breast and prostate cancers. # 2014 Elsevier Masson SAS. All rights reserved. Keywords: Cancer; Incidence; Mortality; Trends; Registry

Re´sume´ Position du proble`me. – Cette e´tude pre´sente les estimations d’incidence et de mortalite´ par cancer en France entre 1980 et 2012. Me´thodes. – Parmi les tumeurs solides, 19 localisations cance´reuses ont e´te´ analyse´es. Les donne´es d’incidence proviennent de 21 registres de´partementaux et correspondent aux cancers invasifs diagnostique´s entre 1975 et 2009. Les donne´es de mortalite´ par cancer sur cette pe´riode ont e´te´ fournies par l’Institut national de la sante´ et de la recherche me´dicale. L’estimation nationale de l’incidence est base´e sur l’utilisation de la mortalite´ comme corre´lat de l’incidence. Les donne´es d’incidence et de mortalite´ ont e´te´ mode´lise´es a` l’aide d’un mode`le aˆge-pe´riode-cohorte. Les nombres de nouveaux cas de cancer et de de´ce`s entre 2010 et 2012 sont issus de projections a` court terme.

* Corresponding author. E-mail address: [email protected] (F. Binder-Foucard). 0398-7620/$ – see front matter # 2014 Elsevier Masson SAS. All rights reserved. http://dx.doi.org/10.1016/j.respe.2013.11.073

96

F. Binder-Foucard et al. / Revue d’E´pide´miologie et de Sante´ Publique 62 (2014) 95–108

Re´sultats. – En 2012, le nombre de nouveaux cas de cancer a e´te´ estime´ a` 355 000 et le nombre de de´ce`s par cancer a` 148 000. L’e´volution des taux d’incidence n’e´tait pas line´aire sur l’ensemble de la pe´riode. Apre`s une augmentation continue depuis 1980, l’incidence du cancer chez l’homme a diminue´ entre 2005 et 2012. Cette diminution re´cente est en grande partie lie´e a` la baisse de l’incidence du cancer de la prostate. Chez la femme, on observe une stabilisation essentiellement due a` la modification de l’e´volution de l’incidence du cancer du sein. La mortalite´ est en diminution sur l’ensemble de la pe´riode pour la majorite´ des localisations. L’analyse conjointe de l’incidence et de la mortalite´ par localisation permet de distinguer les cancers dont l’incidence et la mortalite´ ont diminue´ (ex. : estomac) des cancers dont l’incidence et la mortalite´ ont augmente´ (ex. : poumon chez la femme). D’autres cancers combinent une augmentation de l’incidence et une diminution de la mortalite´ (ex. : thyroı¨de). Conclusion. – Ce travail met en e´vidence des modifications re´centes et marque´es de l’e´volution de l’incidence, plus particulie`rement pour les cancers du sein et de la prostate. # 2014 Elsevier Masson SAS. Tous droits re´serve´s. Mots cle´s : Cancer ; Incidence ; Mortalite´ ; Tendance ; Registre

1. Introduction The availability of updated cancer incidence and mortality indicators is of paramount importance for decision-making and public health policy implementation. Such indicators allow to describe accurately the epidemiological features of cancer and their progression over time. They are used to develop and assess programs for primary prevention, screening, and care. They also provide public healthcare policy-makers with the necessary information to assess the needs of the population in terms of cancer management, prioritize prevention strategies, and evaluate therapeutic progress. This article is a follow-up of a previous study on cancer incidence and mortality published in 2008 on the 1980–2005 period [1,2]. This collaborative work involved the French network of cancer registries (Francim), the Department of Biostatistics of the Hospices Civils de Lyon (HCL), the Institut de veille sanitaire (InVS) and the Institut national du cancer (Inca). It is an integral part of the national cancer program (Plan Cancer 2009–2013) relating to the observation of cancer epidemiology [3]. Cancer registries provide a continuous and exhaustive recording of all cancer cases in a predefined geographical area. Thus, registry data are essential to estimate cancer incidence at a national level. The Francim network has created a common database that brings together all cancer data from participating registries, some dating back to 1975. We report here estimates of the trends in incidence of cancer and mortality from cancer (solid tumors only) in France between 1980 and 2012. The trends in incidence of hematological malignancies will be published separately. 2. Materials and methods 2.1. Incident cases Data on incidence were taken from the common database of all Francim network registries. For the purposes of the present study, data from 21 registries having at least 5 years of recorded data were used (Table 1). Among solid tumors, 19 cancer sites were analyzed. These sites were defined according to the International classification of diseases for oncology, 3rd edition (ICD-O3) [4] and are

shown in Table 2. For each site, the following rules were applied: only invasive tumors were considered and, unless otherwise specified, all morphologies were included except those that corresponded to hematopoietic malignancies. Skin cancers other than melanoma were also excluded. The study used incidence data from 925,242 cases diagnosed between 1 January 1975 and 31 December 2009 (Table 1). 2.2. Cancer-related deaths Data on mortality were provided by the Centre d’e´pide´miologie sur les causes me´dicales de de´ce`s (Ce´piDc), which is affiliated to the Institut national de la sante´ et de la recherche me´dicale (Inserm). All deaths from cancer (4,979,498 deaths) that occurred in France between 1 January 1975 and 31 December 2009 were stored in the database with the following information: year of death, place of residence (de´partement), sex, age at death, and cancer site (coded according to the International classification of diseases 8th, 9th, or 10th edition depending on the year of death, Table 2). Since many deaths were attributed to ‘‘uterine cancer’’ without any other specific indication, it was necessary to apply statistical methods to estimate separately the numbers of cervix and corpus uteri cancer deaths. The method used here was based on the proportions of deaths attributable to cancer of the cervix uteri and corpus uteri among all deaths from uterine cancer [5]. This proportion was obtained by modeling incidence and survival data available from cancer registries per cancer site [6,7]; incidence data from six registries over the 1982–2009 period and survival data from 12 registries on patients diagnosed between 1989 and 2004 and followed up until 31 December 2007. For the present study, more information was available than for the previous study [2]; therefore more elaborated models were used to provide more accurate incidence and survival estimates. The present mortality estimates could therefore differ slightly from those provided earlier [2]. These differences concern mainly mortality estimates for 1980. 2.3. Population data Population data were provided by the Institut national de la statistique et des e´tudes e´conomiques (Insee) for each

F. Binder-Foucard et al. / Revue d’E´pide´miologie et de Sante´ Publique 62 (2014) 95–108

97

Table 1 Incidence data used to estimate cancer incidence and mortality. Registry

De´partement

Type of registry

Years with available data

1 2 3 4 5 6 7 8 9 10 11 12

Calvados

Digestive tract Generala Digestive tract Hematopoieticc Breast General Digestive tract Hematopoieticc Central nervous system General General Breast, colon, rectum General Generala Thyroid

1978–2009 1978–2009 1976–2009 1980–2009 1982–2009 1978–2009 2005–2009 2002–2009 2000–2009 1987–2009 1979–2009 1991–2009 1998–2009 1994–2009 1975–2009

Hematopoieticc General General Digestive tract General General General

2002–2009 1975–2009 1988–2009 1982–2009 1982–2009 1982–2009 1997–2009

Coˆte-d’Or

Doubs Finiste`re Gironde He´rault Ise`re Loire-Atlantique b

13 14

Manche Marne Ardennes Ornea Bas-Rhin Haut-Rhin Saoˆne-et-Loire Somme Tarn Vende´e

15 16 17 18 19 20 21 a b c

Population in 2012 692,049 529,967

533,494 912,347 1,476,927 1,062,393 1,237,946 1,315,880 502,092 568,318 282,753 293,256 1,117,803 756,230 556,993 575,153 385,356 650,594

The registry of hematopoietic cancers of Basse-Normandie covers the following French de´partements: Calvados, Manche, and Orne for the 2002–2009 period. The Loire-Atlantique registry was specialized in breast and colorectal cancer until 1997, and became a general cancer registry in 1998. Registries recording hematopoietic cancers only contributed to the estimations of all cancers.

Table 2 Codes from the International Classification of Diseases used for the cancer sites reported. Incidence

Mortality a

a

Topography

Morphology

1975–1978

1979–1999

2000–2009

Cancer site

ICD-O-3

ICD-O-3

(ICD-8)

(ICD-9)

(ICD-10)

Lip, oral cavity, pharynx Esophagus Stomach Colon-rectum Liver Pancreas Larynx Lung Skin melanoma Breast Cervix uterib Corpus uterib Ovary

C00–C14 C15 C16 C18, C19, C20, C21 C22 C25 C32 C33, C34 C44 C50 C53 C54 C56, C570, C571, C572 C573, C574

140–149 150 151 153, 154 155, 1978 157 161 162 172 174 180–182 180–182 183

140–149 150 151 153, 154 155 157 161 162 172 174 179–182 179–182 183

C00–C14 C15 C16 C18, C19, C20, C21 C22 C25 C32 C33, C34 C43 C50 C53, C54, C55 C53, C54, C55 C56, C570, C571, C572 C573, C574

Prostate Testis Bladder Kidney Central nervous system Thyroid

C61 C62 C67 C64, C65, C66; C68 C70, C71, C72 C73

All All All All All All All All 87203 to 87803 All All All All except {84423; 84513; 84613; 84623; 84723; 84733} All All All All  91103 or  91800 All

185 186 188 189 191, 192 193

185 186 188 189 191, 192 193

C61 C62 C67 C64, C65, C66; C68 C70, C71, C72 C73

a b

Lymphomas were excluded from solid tumors. Deaths from cervix uteri and corpus uteri cancers were re-estimated.

F. Binder-Foucard et al. / Revue d’E´pide´miologie et de Sante´ Publique 62 (2014) 95–108

98

de´partement and each year from 1975 to 2013. The data included local population estimates up to 2008 followed by projections from 2009 onwards. These projections were obtained with the Omphale model developed by the Insee. These data give estimates of the number of persons alive on 1 January of each year according to age, year, sex, and de´partement.

multiplying the ratio of incidence to mortality observed in the registry area by the mortality for all of France. Thus, if l denotes incidence, m mortality, ZR the area covered by the registries, and FR all of France, the national incidence for a given age and a given cohort will be written:

2.4. Rate calculation

Each of the three components on the right-hand side of Equation (1) is obtained by a distinct age-period-cohort model using the same strategy for specifying the models as previously described [2]. The last year for which incidence and mortality data were observed is 2009. Thus, short-term projections were necessary to provide estimates up to 2012. These projections were calculated using the above-described statistical models with the numbers of person-years available up to 2012. For breast and prostate cancer, the incidence in the registry area has varied widely in recent years. To account for these complex trends, flexible models were used for these two cancers (manuscript in progress): each of the three components mentioned in Equation (1) was modeled using a twodimensional smoothing spline [9]. Tables 4 and 5 present the estimated incidence rates lFR (obtained with Equation (1)), age-standardized using the adapted standard weights (Table 3). They also present the annual rates of change for 1980–2012 as calculated from the age-standardized incidence rates for 1980 (l1980) and 2012 (l2012). This annual rate of change (noted ‘‘x’’ below) satisfies the following equation:

Individual incidence and mortality data were aggregated by 1-year classes according to age and cohort of birth (consequently, according to the year of cancer diagnosis or death). The number of person-years was the average between the population aged A years on 1 January of year P and the population aged A + 1 years on 1 January of year P + 1. This made it possible to estimate smooth rates for a given age and a given cohort using the statistical models described below. The age at diagnosis (or death) was calculated as the difference between the year of diagnosis (or death) and the year of birth. Thus, the incidence rate for a population in a 5-year age group actually corresponds to that of a population half a year younger than the same age group as standardized to the world standard age distribution. To allow for this difference, the weights attributed to each age class of the world population were slightly modified [8] (Table 3). All the rates reported in this paper are standardized on the age structure of the modified world population and are expressed per 100,000 person-years. 2.5. Estimation methods The details regarding the statistical method used to estimate cancer incidence can be found in the previous report [2]. Briefly, the national incidence of cancer was obtained by Table 3 Weights used for rate standardization according to the age structures of the European and the world population. Classical standard weights

Adapted standard weights

Age group

European

World

European

World

[00;04] [05;09] [10;14] [15;19] [20;24] [25;29] [30;34] [35;39] [40;44] [45;49] [50;54] [55;59] [60;64] [65;69] [70;74] [75;79] [80;84] [85;++]

0.08 0.07 0.07 0.07 0.07 0.07 0.07 0.07 0.07 0.07 0.07 0.06 0.05 0.04 0.03 0.02 0.01 0.01

0.12 0.1 0.09 0.09 0.08 0.08 0.06 0.06 0.06 0.06 0.05 0.04 0.04 0.03 0.02 0.01 0.005 0.005

0.072 0.071 0.07 0.07 0.07 0.07 0.07 0.07 0.07 0.07 0.07 0.061 0.051 0.041 0.031 0.021 0.011 0.011

0.108 0.102 0.091 0.09 0.081 0.08 0.062 0.06 0.06 0.06 0.051 0.041 0.04 0.031 0.021 0.011 0.0055 0.0055

lFR ¼ lZR  mFR =mZR




(1)

l2012 ¼ l1980  ð1 þ xÞ32 To obtain the annual rate of change for 2005–2012, the analyst should replace l1980 with l2005 and 32 with 7 in this equation. Table 6 presents the estimated number of incident cases in France in 2012 with the 95% confidence intervals as obtained using a jack-knife approach as described in a previous publication [10]. 2.6. Note of caution regarding prostate cancer The incidence trends presented for prostate cancer are based on 1980–2009 and not 1980–2012. Indeed, any projections for trends in prostate cancer seem too uncertain due to the determinant effect of medical practices on the incidence of this cancer. After a steady increase since 1980, the incidence decreased sharply from 2005 to 2009 (the last year for which data are available in the registries), but this decline cannot be assumed, with a reasonable degree of certainty, to have continued up to 2012. Nonetheless, we provided an estimate of incidence for 2012 on a reasonable but debatable hypothesis that the incidence rate remained constant between 2009 and 2012; we also provided an estimate based on the hypothesis that incidence continued to decline from 2009 to 2012. For mortality, the level of uncertainty is lower; projections from the model are provided up to 2012.

F. Binder-Foucard et al. / Revue d’E´pide´miologie et de Sante´ Publique 62 (2014) 95–108

99

Table 4 Estimated annual world age-standardized incidence and mortality rates by cancer type and annual rates of change (1980–2012 and 2005–2012) in men. Incidence

Mortality

Estimated incidence rate per 100,000 person-years Type of cancer

1980

1990

2000

2005

Annual rate of change (%) 2010

2012

Lip, oral cavity, 40.0 37.1 28.9 23.6 18.2 16.1 pharynx Esophagus 16.4 13.5 10.2 8.5 6.8 6.2 Stomach 14.2 11.6 9.2 8.2 7.3 7.0 Colon-rectum 34.7 38.1 39.4 39.2 38.8 38.4 Liver 4.4 7.2 10.0 11.0 11.8 12.1 Pancreas 4.9 5.1 6.3 7.5 9.3 10.2 Larynx 14.2 12.3 9.2 7.6 6.0 5.4 Lung 50.0 52.2 52.9 52.8 52.2 51.7 Skin melanoma 2.5 4.6 7.4 8.9 10.3 10.8 24.8 46.9 79.8 127.1 Prostatea Testis 3.3 4.3 5.7 6.5 7.0 7.2 Bladder 16.5 17.5 17.0 16.2 15.2 14.7 Kidney 7.7 9.7 11.7 12.8 14.0 14.5 Central nervous 4.5 5.4 5.9 6.1 6.3 6.3 system Thyroid 1.1 1.7 2.9 3.8 5.0 5.5 All cancers b 283.5 317.8 351.7 396.1 364.6 362.6

Estimated mortality rate per 100,000 person-years

1980–2012 2005–2012 1980

Annual rate of change (%)

1990

2000

2005

2010

2012

1980–2012 2005–2012

2.8

5.3

15.9

13.5

9.7

7.5

5.4

4.7 3.7

6.5

3.0 2.2 0.3 3.2 2.3 2.9 0.1 4.7

4.4 2.2 0.3 1.3 4.5 4.7 0.3 2.9

13.9 13.1 19.9

11.1 9.0 18.2

8.0 6.3 16.0

6.5 5.4 14.9

5.1 4.6 13.8

4.6 3.4 4.4 3.4 13.3 1.2

5.0 2.8 1.5

2.4 0.4 2.0 1.1

1.6 1.4 1.8 0.4

11.4 43.2 0.9 16.3 0.7 6.9 4.0 3.2

7.6 47.7 1.3 18.1 0.4 7.0 4.6 3.9

4.1 46.3 1.6 15.3 0.3 6.3 4.6 4.0

2.7 43.3 1.7 13.3 0.3 5.8 4.4 3.9

1.7 39.0 1.7 10.9 0.2 5.1 4.2 3.7

1.4 37.0 1.7 10.2 0.2 4.9 4.0 3.6

6.4 0.5 1.9 1.5 3.5 1.1 0.0 0.4

9.3 2.2 0.1 3.7 2.6 2.4 1.2 1.1

5.2 0.8

5.4 1.3

0.4 0.4 0.3 0.3 0.2 0.2 1.9 214.6 209.3 183.2 164.0 142.3 133.6 1.5

2.7 2.9

a

For prostate cancer, see the specific comments in Materials and methods and Results sections. Estimates for ‘‘All cancers’’ were obtained by calculating the sum of the estimates relative to cancer subtypes (with the assumption that incidence rates remained constant between 2009 and 2012 for prostate cancer) plus the estimates for hematopoietic cancers (defined according to the classification of Belot et al. in 2008) plus the estimates for ‘‘other cancers’’. b

Table 5 Estimated annual world age-standardized incidence and mortality rates by cancer type and annual rates of change (1980–2012 and 2005–2012) in women. Incidence

Mortality

Estimated incidence rate per 100,000 person-years Type of cancer

1980

1990

2000

2005

Annual rate of change (%) 2010

2012

Lip, oral cavity, 3.5 4.0 4.7 5.2 5.5 5.6 pharynx Esophagus 1.1 1.2 1.3 1.4 1.5 1.5 Stomach 6.2 4.6 3.4 3.0 2.7 2.6 Colon-rectum 23.0 24.5 24.6 24.3 23.9 23.7 Liver 0.8 1.2 1.7 2.0 2.3 2.4 Pancreas 2.0 2.6 3.8 4.8 6.2 6.9 Larynx 0.6 0.7 0.8 0.8 0.8 0.9 Lung 3.5 5.7 9.7 12.8 16.8 18.6 Skin melanoma 4.0 6.2 8.7 9.8 10.7 11.0 Breast 56.3 75.3 92.6 97.8 90.9 88.0 Cervix uteri 15.0 9.9 7.8 7.3 6.8 6.7 Corpus uteri 10.5 10.3 10.4 10.6 10.7 10.8 Ovary 9.1 9.2 8.6 8.2 7.8 7.6 Bladder 2.8 2.5 2.3 2.3 2.4 2.5 Kidney 3.4 4.2 5.0 5.3 5.7 5.8 Central nervous 3.1 3.8 4.1 4.2 4.2 4.2 system Thyroid 2.8 5.5 9.4 11.5 13.3 13.8 All cancers a 176.4 203.2 234.9 248.8 251.2 252.0

Estimated mortality rate per 100,000 person-years

1980–2012 2005–2012 1980 1.5

1.1

1.1 2.6 0.1 3.5 3.9 1.1 5.3 3.2 1.4 2.5 0.1 0.6 0.4 1.7 0.9

1.1 2.0 0.3 3.0 5.4 0.5 5.4 1.7 1.5 1.2 0.3 1.2 0.9 1.4 0.2

5.1 1.1

2.7 0.2

Annual rate of change (%)

1990 2000 2005 2010 2012 1980–2012 2005–2012 1.3

1.3

1.2

1.1

1.0

0.5

2.2

1.1 1.1 5.7 3.6 12.5 10.6

1.0 2.4 9.2

1.0 2.0 8.6

0.9 1.7 8.1

0.9 1.7 7.9

0.5 3.8 1.4

1.1 2.7 1.1

0.4 0.3 4.0 5.2 0.8 1.0 19.2 20.2 5.0 3.2 3.0 2.6 5.6 6.0 1.4 1.3 1.8 1.9 2.0 2.5

0.3 7.6 1.1 18.9 2.3 2.3 5.4 1.1 1.7 2.6

0.2 9.4 1.1 17.5 2.1 2.3 4.8 1.1 1.6 2.5

0.2 11.8 1.0 16.2 1.9 2.2 4.1 1.0 1.4 2.3

0.2 12.9 1.0 15.7 1.8 2.2 3.8 1.0 1.4 2.2

2.5 3.7 0.8 0.6 3.2 1.0 1.2 1.0 0.9 0.4

4.5 4.6 1.8 1.5 2.0 0.6 3.3 0.9 2.1 1.9

0.6 0.5 100.4 93.6

0.3 85.2

0.3 80.6

0.2 75.3

0.2 73.2

3.4 1.0

4.2 1.4

1.2

a Estimates for ‘‘All cancers’’ were obtained by calculating the sum of the estimates relative to cancer subtypes, plus the estimates for hematopoietic cancers (defined according to the classification of Belot et al. in 2008) plus the estimates for ‘‘other cancers’’.

100

F. Binder-Foucard et al. / Revue d’E´pide´miologie et de Sante´ Publique 62 (2014) 95–108

Table 6 Estimated number of incident cancer cases in France in 2012. Type of cancer

New cases in men [95% CI]

New cases in women [95% CI]

Lip, oral cavity, pharynx Esophagus Stomach Colon-rectum Liver Pancreas Larynx Lung Skin melanoma Breast Cervix uteri Corpus uteri Ovary Prostate a Testis Bladder Kidney Central nervous system Thyroid

8033 [6270; 9797]

3283 [2932; 3633]

3503 [3090; 3916] 4308 [3993; 4628] 23226 [21667; 24781] 6867 [5559; 8177] 5963 [4652; 7273] 2821 [2388; 3252] 28211 [25744; 30683] 5429 [3968; 6889] – – – – 53465 [46840; 60090]a 2317 [1777; 2853] 9549 [8398; 10696] 7781 [6795; 8762] 2814 [2551; 3078]

1129 [1038; 1222] 2248 [2026; 2470] 18926 [17205; 20647] 1856 [1525; 2183] 5699 [4659; 6742] 501 [388; 614] 11284 [10195; 12371] 5747 [4587; 6909] 48763 [46605; 50921] 3028 [2589; 3471] 7275[6796; 7755] 4615 [4095; 5136] – – 2416 [2097; 2739] 3792 [3340; 4246] 2185 [1963; 2408]

2324 [1364; 3282]

5887 [2687; 9083]

a

The estimates for prostate cancer are given for 2009 (see specific comments in sections Materials and methods and Results).

2.7. Note of caution for liver and pancreas cancers The lack of precision on the death certificates made accurate estimations of mortality difficult for these two cancers and the trends over time are difficult to interpret. However, the problem of mortality data quality does not preclude the use of these data to estimate incidence because one can reasonably assume that the lack of precision is of similar magnitude for the area covered by the registries and for all of France. Indeed, if we assume that the lack of precision leads to a 15% overestimation of mortality, given that this affects both areas, it will not affect the factor of adjustment, e.g., the ratio of mortality in France to mortality in the registry area because (1.15  mFR)/(1.15  mZR) = mFR/mZR. 2.8. Estimates for ‘‘all cancers’’ The estimate for ‘‘all cancers’’ is the sum of the estimates per cancer site (assuming that incidence rates remained stable between 2009 and 2012 for prostate cancer, see ‘‘Note of caution regarding prostate cancer’’ above), to which we added the estimates for hematopoietic malignancies (as previously defined in [2], which are not included in this study), plus the estimate for the ‘‘other cancers’’ category. 3. Results All details, results, and the full report (in pdf format) are available at: http://www.invs.sante.fr/fr/Dossiers-thematiques/ Maladies-chroniques-et-traumatismes/Cancers/Surveillanceepidemiologique-des-cancers/Estimations-de-l-incidence-etde-la-mortalite/Estimation-de-l-incidence-et-de-la-mortalitepar-cancer-en-France-entre-1980-et-2012-Tumeurs-solides.

Age-standardized incidence rates are essential to analyze the trends in risk, whereas the number of incident cases is necessary to make previsions for healthcare resources and to evaluate the overall burden of cancer. Incidence and mortality rates in 2012 in men and women are shown in Fig. 1. The numbers of new cases and deaths per cancer site in 2012 in men and women are presented in Fig. 2. Between 1980 and 2012, the number of incident cases of ‘‘all cancers’’ increased by 108.8% (170,000 cases estimated in 1980 vs. 355,000 estimated cases in 2012). The increase of 107.6% in the number of incident cases in men can be broken down into 30.8% due to the increase in population size, 33.7% due to population aging, and 43.1% due to changes in specific incidence rates per age group, e.g., changes in cancer risks [11]. The increase of 111.4% in the number of incident cases in women can similarly be broken down into 33.8%, 22.5%, and 55.1%, respectively. Over the same period, the age-standardized incidence rates increased by 27.9% in men and 42.9% in women. In 2012, the incidence rates were 362.6 in men and 252.0 in women. The incidence rates increased by approximately 1% per year from 1980 to 2012, but this does not reflect the trends observed in the most recent years. Indeed, in men, the incidence rate increased by 0.8% per year over the entire study period but decreased by 1.3% between 2005 and 2012 (Table 4). In women, there was no reversal in trend but rather a slowing down of the progression rate from 1.1% to 0.2% per year on average (Table 5). This phenomenon is primarily due to the decrease in the incidence of prostate cancer in men and to recent changes in the incidence of breast cancer in women. The number of deaths increased by 14.7% between 1980 and 2012 (129,000 vs. 148,000, respectively). The increase of 11.0% in the number of deaths from cancer in men may be explained by a 16.4% increase due to an increase in population size, a 43.3% increase due to population aging, and a 48.7% decrease in cancer risk. Similarly, the 20.3% increase in women can be broken down into +19.3%, +34.5%, and 33.5%, respectively. This leads to the conclusion that the observed overall increase is mainly the result of demographic trends, whereas, in fact, the risk of death from cancer decreased substantially. The age-standardized mortality rates decreased in men and women between 1980 and 2012, with an acceleration of this trend in recent years. The trends in the age-standardized incidence and mortality rates per cancer site are described and discussed below (see also Tables 4 and 5). The confidence intervals for the number of incident cases in 2012 are shown in Table 6. 3.1. Lip-Oral cavity-Pharynx In 2012, there were 11,316 incident cases (of which 71% occurred in men) and 3192 deaths (77% in men); the incidence and mortality relative to these cancer sites have been declining sharply in men over several decades. The incidence rate was 16.1 in men and 5.6 in women and the mortality rates were 4.7 in men and 1.0 in women. The incidence in men has been decreasing since 1980, the greatest decrease being observed in recent years (5.3% decrease per year between 2005 and 2012).

[(Fig._1)TD$IG]

F. Binder-Foucard et al. / Revue d’E´pide´miologie et de Sante´ Publique 62 (2014) 95–108

Men: Incidence rates

Men: Mortality rates

Prostate

Lung

Lung

Large bowel

Large bowel

Prostate

Lip-OC-P

Liver

Bladder

Pancreas

Kidney

Bladder

Liver

Lip-OC-P

Melanoma

Oesophagus

Pancreas

Stomach

Testis

Kidney

Stomach

Brain, CNS

Brain, CNS

Melanoma

Oesophagus

Larynx

Thyroid

Thyroid

Larynx

Testis

0

20

40

60

80

101

100

0

10

20

30

40

Incidence per 100 000 person-years

Mortality per 100 000 person-years

Women: Incidence rates

Women: Mortality rates

Breast

50

Breast

Large bowel

Lung

Lung

Large bowel

Thyroid

Pancreas

Melanoma

Ovary

Corpus uteri

Liver

Ovary

Brain, CNS

Pancreas

Corpus uteri

Cervix uteri

Cervix uteri

Kidney

Stomach

Lip-OC-P

Kidney

Brain, CNS

Bladder

Stomach

Lip-OC-P

Bladder

Melanoma

Liver

Oesophagus

Oesophagus

Thyroid

Larynx

Larynx

0

20

40

60

80

100

Incidence per 100 000 person-years

0

10

20

30

40

50

Mortality per 100 000 person-years

Fig. 1. Estimates of incidence and mortality rates in France for 2012, age-standardized to the world standard population and expressed per 100,000 person-years. For prostate cancer, the incidence rate in 2012 was based on the assumption of constant rate since 2009. OC: oral cavity; P: pharynx; CNS: central nervous system.

The trends in mortality in men followed a similar pattern with an even more pronounced decline (average annual decrease, 6.5%) between 2005 and 2012. Conversely, in women, the incidence has increased but slowed since 2005. Mortality among women decreased slightly over the whole period with a pronounced decline since 2005. These trends in incidence and mortality should be interpreted in light of the exposure to the main risk factors for these types of cancer, namely alcohol and tobacco consumption. The inverse trends in tobacco use (decrease in men and increase in women) may explain these discrepancies in incidence [12]. In women, the contrast between the increase in incidence and the decrease in mortality could be related to progress in therapy or to the more frequent

presence of human papillomavirus (HPV) in oropharyngeal cancers, as the prognoses of HPV-positive cancers is more favorable than tobacco-related cancers [13,14]. 3.2. Esophagus In 2012, with 4632 incident cases (of which 76% occurred in men) and 3444 deaths, cancer of the esophagus is the cancer type whose incidence has decreased the most in men since 1980, especially since 2005. The trends in mortality in men followed the same pattern. The incidence rates were 6.2 in men and 1.5 in women (male/female ratio: 4.1). The incidence in women increased slightly but steadily, whereas mortality

[(Fig._2)TD$IG]102

F. Binder-Foucard et al. / Revue d’E´pide´miologie et de Sante´ Publique 62 (2014) 95–108

Men: Number of cases

Men: Number of deaths

Prostate

Lung

Lung

Large bowel

Large bowel

Prostate

Bladder

Liver

Lip-OC-P

Pancreas

Kidney

Bladder

Liver

Stomach

Pancreas

Oesophagus

Melanoma

Kidney

Stomach

Lip-OC-P

Oesophagus

Brain, CNS

Larynx

Melanoma

Brain, CNS

Larynx

Thyroid

Thyroid

Testis

Testis

0

10000

20000

30000

40000

50000

0

10000

20000

30000

40000

50000

Number of cases

Number of deaths

Women: Number of cases

Women: Number of deaths

Breast

Breast

Large bowel

Lung

Lung

Large bowel

Corpus uteri

Pancreas

Thyroid

Ovary

Melanoma

Liver

Pancreas

Corpus uteri

Ovary

Stomach

Kidney

Kidney

Lip-OC-P

Brain, CNS

Cervix uteri

Bladder

Bladder

Cervix uteri

Stomach

Oesophagus

Brain, CNS

Lip-OC-P

Liver

Melanoma

Oesophagus

Thyroid

Larynx

Larynx

0

10000

20000

30000

40000

50000

Number of cases

0

10000

20000

30000

40000

50000

Number of deaths

Fig. 2. Estimated numbers of incident cancer cases and cancer-related deaths in men and women in France, 2012. For prostate cancer, the number of cancer cases in 2012 was based on the assumption of constant rate since 2009. OC: oral cavity; P: pharynx; CNS: central nervous system.

declined slightly over the whole period. The decrease in incidence observed in men is related to a reduction in risky behavior (alcohol and tobacco consumption) [12,15] and, to a lesser extent, to changes in diet [16]. Regarding the histological types, there was a reduction in epidermoid carcinomas in men and an increase in adenocarcinomas in both sexes, the main risk factors for the latter type being gastroesophageal reflux and obesity [17].

mortality rates were 4.4 in men and 1.7 in women. In both sexes, incidence and mortality has declined regularly over the last 30 years in France (as in the majority of Western countries). With the exception of incidence in men, these reductions in rates have leveled off over the last few years. This decrease might be explained by the fall in the prevalence of Helicobacter pylori infection according to the cohort of birth, the parallel improvement of living conditions and hygiene [18], and the changes in dietary behavior.

3.3. Stomach 3.4. Colon and rectum For 2012, the number of incident cases of stomach cancer was estimated at 6556, of which 66% occurred in men. The incidence rates were 7.0 in men and 2.6 in women. The

With 42,152 incident cases in 2012, of which 55% occurred in men, colorectal cancer is the third most frequent type of

F. Binder-Foucard et al. / Revue d’E´pide´miologie et de Sante´ Publique 62 (2014) 95–108

cancer in men and the second most frequent in women. The incidence rates were 38.4 in men and 23.7 in women. With 17,722 deaths, of which 52% occurred in men, this cancer is the second cause of cancer-related death in men and the third in women. The mortality rates were 13.3 in men and 7.9 in women. In both sexes, the incidence increased over about 20 years then stabilized before decreasing since 2005. This reduction in incidence has also been observed in the United States since the mid-1980s [19]. The access to screening programs and resection of precancerous lesions partially explain this decrease. Mortality declined steadily in both sexes between 1980 and 2012. 3.5. Liver For 2012, the number of incident cases of primary liver cancer was estimated at 8723, of which 79% occurred in men. The incidence rates were 12.1 in men and 2.4 in women (male/female ratio: 5.0). In both sexes, the incidence of liver cancer increased substantially between 1980 and 2005 and continued to increase, but to a lesser extent, up until 2012, particularly in men. These trends in the incidence of primary liver cancer can be partially explained by the increased incidence of chronic liver disease related to alcohol consumption, hepatitis B and C viruses, and steatohepatitis [20], as well as by the improvement in the management of patients with cirrhosis, which leaves more time for cancer to develop. Metabolic syndrome appears to be gaining ground as a cause of primary liver cancer [21].

103

incidence of larynx cancer mainly reflect trends in alcohol use and smoking, with a delay of 20–30 years [24]. 3.8. Lung With 29,949 deaths in 2012, of which 71% occurred in men, lung cancer caused the highest number of deaths. The number of new cases was 39,495, of which 71% occurred in men. The incidence rates were 51.7 in men and 18.6 in women (male/ female ratio: 2.8). The mortality rates were 37.0 in men and 12.9 in women. In men, the trend towards a rise in incidence since 1980 started to reverse in 2005, with an annual decrease of 0.3%. In women, the incidence strongly increased (by 5.3% per year on average between 1980 and 2012), which led to a corresponding increase in mortality, particularly in recent years (by 4.6% per year on average between 2005 and 2012). This discrepancy in incidence trends between men and women is primarily due to the decrease in smoking in men but its increase in women [12]. Similar trends were observed in other industrialized countries, such as the United States, where the peak of lung cancer incidence was reached in 2006 [25]. Accordingly, it seems unlikely that a reversal of this trend in women will occur soon in France. 3.9. Skin melanoma

For 2012, the number of incident cases of pancreatic cancer was estimated at 11,662, of which 51% occurred in men. The incidence rates were 10.2 in men and 6.9 in women. Since 1980, the incidence of this cancer has been increasing considerably in men and even more in women, and this rise has become more pronounced in recent years. These trends in incidence may be explained by changes in the diagnostic methods that resulted from the progress in imaging techniques as well as by trends in the exposure to some of the main risk factors such as obesity and diabetes [22,23].

In 2012, the number of new cases of skin melanoma was estimated at 11,176, of which 49% occurred in men. The number of deaths was estimated at 1672. The incidence rates were similar in men and women (10.8 and 11.0, respectively). In both sexes, the incidence of skin melanoma has risen sharply since 1980, although there has been a trend toward a decrease since 2005. In men, trends in mortality matched those of incidence, with a leveling off of the increasing trend from 2005 to 2012. In women, after an initial rise, there was a decrease in mortality. These trends are partially due to the campaigns for early detection of this type of cancer but they still need to be confirmed. However, although the French population is well aware of the risk of cancer incurred by prolonged exposure to the sun, there are still wide differences in behavior in terms of exposure avoidance. Prevention campaigns should be specifically oriented toward the youngest population [26].

3.7. Larynx

3.10. Breast

In 2012, with 3322 incident cases and 906 deaths, of which approximately 85% occurred in men, cancer of the larynx was among the cancer types whose incidence has decreased substantially in men since 1980, especially in recent years. Mortality in men followed this trend more markedly, with an average annual decrease of 9.3% between 2005 and 2012. The incidence rates were 5.4 in men and 0.9 in women and the mortality rates 1.4 and 0.2, respectively. Conversely, the incidence has been slightly increasing in women since 2005. Mortality in women followed the pattern seen in men, albeit to a lesser extent (an average decrease of 4.5% per year since 2005). Like other alcohol- and tobacco-related cancers, the trends in

Breast cancer is the most frequent type of cancer in women; it accounted for one-third of all new cancer cases in 2012. It is also the leading cause of cancer-related death in women. In 2012, there were 48,763 new cases and 11,886 deaths. The incidence rate was 88.0 and the mortality rate 15.7. The incidence of breast cancer, which rose considerably between 1980 and 2000, has been falling since 2005. These wide variations complicated the projections and the estimations of the incidence rates for 2012; these projections should therefore be considered with caution. The fall in incidence concerned mainly the tumors diagnosed at early stages (T1/T2-N0-M0) and was especially observed in women aged 50–74 years [27].

3.6. Pancreas

104

F. Binder-Foucard et al. / Revue d’E´pide´miologie et de Sante´ Publique 62 (2014) 95–108

In the area covered by registries, a strong decrease was observed in the 50 to 74 years old age group regardless of the tumor stage (data not shown). An important factor of this decrease in incidence since 2005 could be the decrease in the use of hormone-replacement therapy after menopause [28]. However, other factors, such the saturation of individual or collective screening programs, should also be considered [27,29]. Mortality remained relatively constant until about 1995 but declined thereafter. 3.11. Cervix uteri/Corpus uteri In 2012, the number of new cases of cervical cancer was estimated at 3028 and the number of related deaths at 1102. The incidence and mortality rates were 6.7 and 1.8, respectively. These rates have been declining for several decades, but this trend has slowed down since 2005. This decline was mainly due to the widespread implementation of screening programs (cervical smear tests) since the 1960s. A reduction in the prevalence of Papillomavirus, whose presence is required for the development of precancerous lesions [30], may have contributed to the observed trends. In 2012, there were 7275 new cases of cancer of the corpus uteri and 2025 deaths. The incidence was constant over the entire study period and the incidence rate was 10.8 in 2012. Mortality, which was already low, declined regularly over the study period; however, this trend has leveled off since 2000. In 2012, the mortality rate was 2.2. 3.12. Ovary In 2012, the number of new cases of invasive ovarian cancer was estimated at 4615 and the number of deaths at 3140. The incidence and mortality rates were 7.6 and 3.8, respectively. Both indicators have declined since the early 1990s, but the mortality rate has declined more rapidly since 2005. The incidence trend is probably related to the use of oral contraceptives, which protect against ovarian cancer [31]. Other factors that reduce the number of ovulations over the lifetime (such as late puberty, early menopause, and periods of pregnancy and breastfeeding) are also protectors against ovarian cancer. 3.13. Prostate It should be noted that the estimations for prostate cancer incidence represent the 1980–2009 period only. Indeed, wide variations in incidence were observed in the last few years of this period, making projections of future trends difficult, even short-term trends (e.g., 2009–2012). Thus, only estimates for the period of available observed data are provided. With 53,465 new cases estimated in 2009, prostate cancer was the most frequent cancer type in men, nearly a quarter of all cancers diagnosed. The incidence rate was 99.4. With 8876 deaths in 2012, prostate cancer ranks third in terms of cancerrelated mortality in men. In 2012, the mortality rate was 10.2. After a strong rise between 1980 and 2005, a considerable reduction in incidence was observed. Conversely, mortality has

been declining consistently since the late 1990s. The trend in the incidence of prostate cancer depended primarily on the impact of the screening via prostate-specific antigen (PSA) testing. The widespread use of PSA in the US led to a twofold increase in the incidence rates over 5 years, followed by an abrupt decline [32]. In several European countries (Sweden, Finland, the Netherlands), declines in incidence were also reported for 2005 [33,34]. This reduction involves two aspects: (i) epidemiological: after several years of active screening, a proportion of prevalent cancers were diagnosed; (ii) societal: an awareness of a risk of overdiagnosis among physicians and in the population prompted caution regarding PSA-based screening. Estimation of incidence with different scenarios may be proposed for 2012. For example, under the hypothesis of a constant incidence from 2009 to 2012, the number of new cases in 2012 would be estimated at 56,841, whereas under the hypothesis of a continued decline from 2009 to 2012, the number of new cases would be 41,881. 3.14. Testis Testicular cancer is rare; the estimations for 2012 were 2317 new cases and 85 deaths. The incidence and mortality rates were 7.2 and 0.2, respectively. The incidence has been increasing regularly since the 1980s, but it has slowed down since 2005. Mortality, which was already low, has been declining since 1980. This decline has been observed since the introduction of platinum salt therapy [35,36], but it may also be related to the increase in seminoma testicular cancers, which have a more favorable prognosis [37]. 3.15. Bladder Data on the incidence of bladder cancer should be interpreted with caution because the international classifications of this cancer have evolved over time and may have impacted the estimations, particularly those relative to the earliest periods (1980–1990). In 2012, 11,965 new cases were diagnosed (80% in men) and 4772 deaths occurred (nearly 75% in men). The incidence rates were 14.7 in men and 2.5 in women (male/female ratio: 5.9). In men, after an increase up to 1990, the incidence decreased between 2005 and 2012. In women, the incidence declined from 1980 to 2000 but has increased slightly since 2005. The mortality rates were 4.9 in men and 1.0 in women. Similar to incidence, mortality among men has been decreasing since 1990 and has decreased even more in recent years. Mortality in women declined consistently over the whole study period. In countries without schistosomiasis, smoking and occupational exposure to chemicals are the main risk factors. Changes in cigarette composition, with the introduction of potentially carcinogenic additives, could also explain the trends observed since the 1980s [38]. 3.16. Kidney The number of new cases and deaths estimated for 2012 were, respectively, 11,573 and 3957 (with 67% of cases and

F. Binder-Foucard et al. / Revue d’E´pide´miologie et de Sante´ Publique 62 (2014) 95–108

deaths occurring in men). The incidence rates were 14.5 in men and 5.8 in women (male/female ratio: 2.5). Since 1980, the incidence increased in both men and women. The mortality rates were 4.0 and 1.4 for men and women, respectively. Mortality has been falling since the 1990s. With smoking and hypertension, obesity is a major incidence risk factor for this cancer; the increasing prevalence of obesity may have contributed to the increase in incidence [22]. Furthermore, the increasingly widespread use of imaging examinations for various reasons has contributed to increasing incidence and decreasing mortality through fortuitous discoveries of tumors at early stages [39]. 3.17. Central nervous system The number of new cases in 2012 (malignant tumors only, Table 2) was 4999, of which 56% occurred in men. The incidence rates were 6.3 in men and 4.2 in women (male/ female ratio: 1.5). The number of deaths in 2012 was 3052, of which 58% were in men. The mortality rates were 3.6 in men and 2.2 in women. In both sexes, the incidence of central nervous system cancers has been rising since 1980, although slowing down since 2005. Similar findings were also observed in recent registry data from the Scandinavian countries [40]. Mortality increased marginally until 2000, and then showed a slight tendency towards a decrease. 3.18. Thyroid The number of new cases in 2012 was 8211, of which 72% occurred in women. The incidence rates were 5.5 in men and 13.8 in women (male/female ratio: 0.4). In both sexes, the incidence of thyroid cancer increased strongly from 1980 onwards, but this increase has slowed down since 2005 in women. The increase concerned mainly papillary thyroid cancers [41] whose prognoses are very favorable. The number of deaths in 2012 was 375. The mortality rates were low: 0.2 in men and women. Mortality decreased over the whole study period in women and from 1990 onwards in men. 4. Discussion The present results are an update of previous publications [2,10] based on data from a larger number of cancer registries that cover a longer period (1980–2009). The participating registries cover a population of over 8.8 million inhabitants (over 10.8 million for digestive tract cancers) and the oldest registries have been recording data for over 30 years. Although the French cancer registry network covers only 16–20% of the population (the proportion varies according to cancer site), the use of mortality data as a correlate of incidence makes it possible to estimate incidence for the whole country reliably. Indeed, the relevance of this method has been supported by a recent study [42] where national incidence for 2004–2006 was estimated using

105

medico-administrative data such as the PMSI (Programme de me´dicalisation des syste`mes d’information) and ALD 30 (Affections de longue dure´e) as a correlate of incidence. The results were found very close to national incidence values obtained using mortality as a correlate, showing that the three methods are reliable in estimating the national incidence for a recent period. However, the method that uses mortality as a correlate should be preferred because it has the additional advantage of providing smoothed estimation and trend analyses [42]. Previous publications have shown that, in France, all cancers combined, there were incidence increases and mortality decreases [2,10]. Here, the results show changes in incidence trends in both men and women. Indeed, after a consistent incidence increase for more than 25 years; since 2005, and for the first time, cancer incidence is declining in men and stabilizing in women. In men, the recent ‘‘all cancers’’ incidence trends are significantly affected by the incidence of prostate cancer (note, however, that the estimates for 2012 for prostate cancer are based on the assumption that the incidence rate remained constant between 2009 and 2012, although this remains to be confirmed). From 1980 to 2012, the number of new cancer cases increased by 108.8%. This increase is due to changes in population and risks. Over the same period, the incidence rates increased by 27.9% in men and 42.9% in women. However, these increases are not linear over the whole period; the trend has reversed since 2005. This is largely related to the reduced incidence of prostate cancer and the recent changes in breast cancer incidence. The number of cancer-related deaths increased by 14.7% over the same period, whereas the mortality rates declined by 37.7% in men and 27.1% in women. The rise in the number of deaths is essentially linked to demographic changes because the risk of cancer-related death declined. The mortality rates were higher in men than in women but the decrease was more pronounced in men. This drop is mainly related to the reduction in alcohol and tobacco consumption in men [12,15]. The use of mortality data as a correlate of incidence makes it possible to estimate the French national incidence, but this method comes at the cost of a limited accuracy in cancer description. Indeed, using the general ICD classification instead of the specific ICD-O allows an analysis of mortality data by tumor site but not by histological type. When all the studied sites are considered together, the ‘‘all cancers’’ category represents a highly heterogeneous set regarding risk factors, management, and prognosis. Therefore, the results concerning ‘‘all cancers’’ should not be interpreted alone. Furthermore, incidence and mortality should be analyzed jointly. While an increase in incidence sometimes reflects a real change in the risk of a given cancer, it may also reflect advances in medical technology that ensure early diagnoses of tumors that would otherwise go unnoticed. Mortality, on the other hand, depends on the number of incident cases, the proportion of cases seen at advanced stages at diagnosis, and treatment efficacy. Accordingly, an increase in incidence can be

106

F. Binder-Foucard et al. / Revue d’E´pide´miologie et de Sante´ Publique 62 (2014) 95–108

accompanied by an increase in mortality when the proportion of advanced cancer stages at diagnosis remains constant and/or when there is no major progress in therapy. Conversely, an increase in incidence can be accompanied by a reduction in mortality when the proportion of advanced cancer stages at diagnosis decreases and/or when there are major advancements in cancer management. The analysis of the trends over the last 30 years has shown that, for many cancer sites, there has been an increase in incidence associated with a decrease or stabilization of mortality. This may result from improved cancer detection (e.g., breast cancer), overdiagnosis (though hard to quantify in the case of prostate cancer), and even, though to a lesser extent, from an increase in risk with a decrease in mortality due to improvement in survival. Up to 2005, breast plus prostate cancers accounted for the largest share of the increase in diagnosed cancer cases. Some cancers showed a favorable trend, e.g., decreases in both incidence and mortality. This was the case of stomach cancer in men and women, which may be explained by changes in dietary habits and a lower prevalence of H. pylori infection [18]. In men, the same favorable pattern was also observed in cancers of the esophagus, the lip-oral cavity-pharynx, and the larynx, which may be explained by less alcohol and tobacco consumption, the two main risk factors for these cancers [12,15]. In women, a favorable trend was observed for cervical cancer, which may be primarily explained by more frequent diagnoses of precancerous lesions and early-stage cancers by a screening program. In these cases, the decrease in mortality is not due to an improvement in survival but to a decrease in incidence. On the other hand, the present study revealed several worrying trends, e.g., increases in both incidence and mortality. The first concern is lung cancer in women (because of its poor prognosis). Skin melanomas and the cancers of the central nervous system also showed concomitant increases in incidence and mortality up to 2000; however, the mortality declined or stabilized afterwards. The analysis of the recent trends over the 2005–2012 period has shown that in women the incidence and mortality of lung cancer continue to increase; lung cancer now ranks third in incidence and second in mortality. However, breast cancer and prostate cancer have shown significant decreases in incidence since 2005. For breast cancer, this fall is probably due to fewer prescriptions of hormone replacement therapies after menopause. For prostate cancer, the decrease should be linked to the lively debate on the place of PSA testing for prostate cancer screening. It is also possible that after many years of screening, a large proportion of prevalent cancers have now been diagnosed. In both cancers, it is impossible to predict the duration of this decrease. Note: more detailed results for each cancer site are available at: http://www.invs.sante.fr/fr/Dossiers-thematiques/Maladieschroniques-et-traumatismes/Cancers/Surveillance-epidemiolo gique-des-cancers/Estimations-de-l-incidence-et-de-la-mortalite/ Estimation-de-l-incidence-et-de-la-mortalite-par-cancer-en-Franceentre-1980-et-2012-Tumeurs-solides.

Disclosure of interest The authors declare that they have no conflicts of interest concerning this article. Acknowledgements The authors thank all those who contributed to the recording of cancer data in the registries, most particularly the laboratories and departments of anatomy, cytology, and pathology; the departments of medical information of the public and private hospitals; the local offices of the national social security service; and general practitioners and specialists. The authors also thank Jacques Este`ve for his methodological support, Fiona Ecarnot and Jean Iwaz for the translation of the original French manuscript and their editorial assistance. This research was carried out in the context of a fourinstitute research-program partnership involving the Institut national du cancer (INCa), the Institut de veille sanitaire (InVS), Francim, and Hospices Civils de Lyon (HCL). Appendix A List of contributors:  M. Velten, Registre des cancers du Bas-Rhin, France  A.-V. Guizard, Registre ge´ne´ral des tumeurs du Calvados, France  G. Coureau, Registre des cancers en Gironde, France  E. Marrer, Registre des cancers du Haut-Rhin, France  B. Tre´tarre, Registre des tumeurs de l’He´rault, France  M. Colonna, Registre du cancer de l’Ise`re, France  K. Je´hannin-Ligier, Registre des cancers de Lille et de sa re´gion, France  F. Molinie´, A. Cowppli-Bony, Registre des tumeurs de LoireAtlantique et Vende´e, France  S. Bara, Registre des cancers de la Manche, France  B. Lapoˆtre-Ledoux, Registre du cancer de la Somme, France  P. Grosclaude, L. Daubisse-Marliac, Registre des cancers ge´ne´raux du Tarn, France  N. Le´one, Registre des cancers en re´gion Limousin, France  A.-M. Bouvier, Registre bourguignon des cancers digestifs, France  G. Launoy, Registre des cancers digestifs du Calvados, France  J.-B. Nousbaum, Registre finiste´rien des tumeurs digestives, France  X. Troussard, Registre he´mopathies malignes de BasseNormandie, France  M. Maynadie´, Registre he´mopathies malignes de Coˆte-d’Or, France  A. Monnereau, Registre des he´mopathies malignes de Gironde, France  J. Clavel, Registre national des he´mopathies malignes de l’enfant, France  B. Lacour, Registre national des tumeurs solides de l’enfant, France

F. Binder-Foucard et al. / Revue d’E´pide´miologie et de Sante´ Publique 62 (2014) 95–108

 F. Galateau-Salle´, N. Le Stang, Registre multicentrique du me´sothe´liome a` vocation nationale, France  P. Arveux, Registre des cancers du sein et des cancers gyne´cologiques de Coˆte-d’Or, France  I. Baldi, Registre des tumeurs primitives du syste`me nerveux en Gironde, France  C. Schvartz, S. Dabakuyo, Registre des cancers de la thyroı¨de Marne-Ardennes, France  G. Rey, Institut national de la sante´ et de la recherche me´dicale, Centre d’e´pide´miologie sur les causes me´dicales de de´ce`s, France References [1] Belot A, Velten M, Grosclaude P, Bossard N, Launoy G, Remontet L, et al. Estimation nationale de l’incidence et de la mortalite´ par cancer en France entre 1980 et 2005. Saint-Maurice (France): Institut de veille sanitaire; 2008, 132 p. [2] Belot A, Grosclaude P, Bossard N, Jougla E, Benhamou E, Delafosse P, et al. Cancer incidence and mortality in France over the period 1980–2005. Rev Epidemiol Sante Publique 2008;56:159–75. [3] Plan cancer 2009-2013. Paris; 2010. Available from: http://www.e-cancer.fr/plancancer-2009-2013. [4] Fritz A, Percy C, Jack A, Shanmugaratram K, Sobin L, Parkin DM. International classification of diseases for oncology. 3rd ed, World Health Organisation; 2000. [5] Rogel A, Belot A, Suzan F, Bossard N, Boussac M, Arveux P, et al. Reliability of recording uterine cancer in death certification in France and age-specific proportions of deaths from cervix and corpus uteri. Cancer Epidemiol 2011;35:243–9. [6] Roche L, Danieli C, Belot A, Grosclaude P, Bouvier AM, Velten M, et al. Cancer net survival on registry data: use of the new unbiased Pohar-Perme estimator and magnitude of the bias with the classical methods. Int J Cancer 2013;132:2359–69. [7] Monnereau A, Troussard X, Belot A, Guizard AV, Woronoff AS, Bara S, et al. Unbiased estimates of long-term net survival of hematological malignancy patients detailed by major subtypes in France. Int J Cancer 2013;132:2378–87. ˆ ge en anne´es re´volues ou aˆge atteint dans l’anne´e. Rev [8] Hill C, Doyon F. A Epidemiol Sante Publique 2005;53:205–8. [9] Clements MS, Armstrong BK, Moolgavkar SH. Lung cancer rate predictions using generalized additive models. Biostatistics 2005;6: 576–89. [10] Remontet L, Esteve J, Bouvier AM, Grosclaude P, Launoy G, Menegoz F, et al. Cancer incidence and mortality in France over the period 1978–2000. Rev Epidemiol Sante Publique 2003;51:3–30. [11] Bashir S, Esteve J. Analysing the difference due to risk and demographic factors for incidence or mortality. Int J Epidemiol 2000;29:878–84. [12] Beck F, Guignard R, Richard JB, Wilquin JL, Peretti-Watel P. Augmentation re´cente du tabagisme en France : principaux re´sultats du barome`tre sante´, France, 2010. Bull Epidemiol Hebd 2011;20–21:230–3. [13] Chung CH, Gillison ML. Human papillomavirus in head and neck cancer: its role in pathogenesis and clinical implications. Clin Cancer Res 2009;15:6758–62. [14] Ang KK, Harris J, Wheeler R, Weber R, Rosenthal DI, Nguyen-Tan PF, et al. Human papillomavirus and survival of patients with oropharyngeal cancer. N Engl J Med 2010;363:24–35. [15] Observatoire franc¸ais des drogues et des toxicomanies : se´ries Statistiques Alcool. Available from: http://www.ofdt.fr/ofdtdev/live/produits/alcool/ conso.html. Data of 2011. [16] Castellsague X, Munoz N, De Stefani E, Victora CG, Quintana MJ, Castelletto R, et al. Smoking and drinking cessation and risk of esophageal cancer (Spain). Cancer Causes Control 2000;11:813–8. [17] Lepage C, Drouillard A, Jouve JL, Faivre J. Epidemiology and risk factors for oesophageal adenocarcinoma. Dig Liver Dis 2013;45:625–9.

107

[18] Tonkic A, Tonkic M, Lehours P, Megraud F. Epidemiology and diagnosis of Helicobacter pylori infection. Helicobacter 2012;17(Suppl. 1):1–8. [19] Cress RD, Morris C, Ellison GL, Goodman MT. Secular changes in colorectal cancer incidence by subsite, stage at diagnosis, and race/ ethnicity, 1992–2001. Cancer 2006;107:1142–52. [20] Trichopoulos D, Bamia C, Lagiou P, Fedirko V, Trepo E, Jenab M, et al. Hepatocellular carcinoma risk factors and disease burden in a European cohort: a nested case-control study. J Natl Cancer Inst 2011;103:1686–95. [21] Welzel TM, Graubard BI, Zeuzem S, El-Serag HB, Davila JA, McGlynn KA. Metabolic syndrome increases the risk of primary liver cancer in the United States: a study in the SEER-Medicare database. Hepatology 2011;54:463–71. [22] Pigeyre M, Dauchet L, Simon C, Bongard V, Bingham A, Arveiler D, et al. Effects of occupational and educational changes on obesity trends in France: the results of the MONICA-France survey 1986–2006. Prev Med 2011;52:305–9. [23] Kusnik-Joinville O, Weill A, Salanave B, Ricordeau P, Allemand H. Prevalence and treatment of diabetes in France: trends between 2000 and 2005. Diabetes Metab 2008;34:266–72. [24] Guerin S, Hill C. L’e´pide´miologie des cancers en France en 2010 : comparison avec les E´tats-Unis. Bull Cancer 2010;97:47–54. [25] Center for Desease Control and Prevention (CDC). State-specific trends in lung cancer incidence and smoking–United States, 1999–2008. Morb Mortal Wkly Rep 2011;60:1243–7. [26] Beck F, Gautier A. Barome`tre cancer 2010 : Barome`tres sante´. Inpes, coll; 2012 [272 p. Available from: http://www.inpes.sante.fr/CFESBases/catalogue/pdf/1405.pdf]. [27] Daubisse-Marliac L, Delafosse P, Boitard JB, Poncet F, Grosclaude P, Colonna M. Breast cancer incidence and time trend in France from 1990 to 2007: a population-based study from two French cancer registries. Ann Oncol 2011;22:329–34. [28] Ringa V, Fournier A. La diminution de l’utilisation du traitement hormonal de la me´nopause a-t-elle fait baisser l’incidence du cancer du sein en France (et ailleurs) ? Rev Epidemiol Sante Publique 2008;56:297–301. [29] Weedon-Fekjaer H, Bakken K, Vatten LJ, Tretli S. Understanding recent trends in incidence of invasive breast cancer in Norway: age-period-cohort analysis based on registry data on mammography screening and hormone treatment use. BMJ 2012;344:e299. [30] Walboomers JM, Jacobs MV, Manos MM, Bosch FX, Kummer JA, Shah KV, et al. Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J Pathol 1999;189:12–9. [31] Rose PG. Endometrial carcinoma. N Engl J Med 1996;335:640–9. [32] Howlader N, Noone AM, Krapcho M, Neyman N, Aminou R, Waldron W, et al. SEER cancer statistics review. 1975–2009 (vintage 2009 populations). Institute NC (ed); 2012, Available from: http://seer.cancer.gov/csr/ 1975_2009_pops09/. [33] Bray F, Lortet-Tieulent J, Ferlay J, Forman D, Auvinen A. Prostate cancer incidence and mortality trends in 37 European countries: an overview. Eur J Cancer 2010;46:3040–52. [34] Ferlay J, Steliarova-Foucher E, Lortet-Tieulent J, Lortet-Tieulent J, Rosso S, Coebergh JW, et al. Cancer incidence and mortality patterns in Europe: estimates for 40 countries in 2012. Eur J Cancer 2013;49: 1374–403. [35] Chia VM, Quraishi SM, Devesa SS, Purdue MP, Cook MB, McGlynn KA. International trends in the incidence of testicular cancer, 1973–2002. Cancer Epidemiol Biomarkers Prev 2010;19:1151–9. [36] Verhoeven RH, Gondos A, Janssen-Heijnen ML, Saum KU, Brewster DH, Holleczek B, et al. Testicular cancer in Europe and the USA: survival still rising among older patients. Ann Oncol 2013;24:508–13. [37] Karim-Kos HE, de Vries E, Soerjomataram I, Lemmens V, Siesling S, Coebergh JW. Recent trends of cancer in Europe: a combined approach of incidence, survival and mortality for 17 cancer sites since the 1990. Eur J Cancer 2008;44:1345–89. [38] Freedman ND, Silverman DT, Hollenbeck AR, Schatzkin A, Abnet CC. Association between smoking and risk of bladder cancer among men and women. JAMA 2011;306:737–45. [39] Chow WH, Dong LM, Devesa SS. Epidemiology and risk factors for kidney cancer. Nat Rev Urol 2010;7:245–57.

108

F. Binder-Foucard et al. / Revue d’E´pide´miologie et de Sante´ Publique 62 (2014) 95–108

[40] Deltour I, Johansen C, Auvinen A, Feychting M, Klaeboe L, Schuz J. Time trends in brain tumor incidence rates in Denmark, Finland, Norway, and Sweden, 1974–2003. J Natl Cancer Inst 2009;101: 1721–4. [41] Colonna M, Bossard N, Guizard AV, Remontet L, Grosclaude P. Descriptive epidemiology of thyroid cancer in France: incidence, mortality and

survival. Ann Endocrinol (Paris) 2010;71:95–101 [Available from: http:// opac.invs.sante.fr/doc_num.php?explnum_id=7044]. [42] Uhry Z, Belot A, Colonna M, Bossard N, Rogel A, Iwaz J, et al. National cancer incidence is estimated using the incidence/mortality ratio in countries with local incidence data: is this estimation correct? Cancer Epidemiol 2013;37:270–7.