- Email: [email protected]
Descriptive epidemiology of thyroid cancer in France: Incidence, mortality and survival
Annales d’Endocrinologie 71 (2010) 95–101
Original article
Descriptive epidemiology of thyroid cancer in France: Incidence, mortality and survival Épidémiologie descriptive du cancer thyroïdien en France : incidence, mortalité et survie M. Colonna a,∗,b , N. Bossard c,d , A.-V. Guizard b,e , L. Remontet c,d , P. Grosclaude b,f , le réseau FRANCIM1 a
Registre du cancer de l’Isère, 23, chemin-des-Sources, 38240 Meylan, France b FRANCIM, faculté de médecine, 31073 Toulouse cedex, France c Service de biostatistique, hospices civils de Lyon, 69003 Lyon, France d CNRS, UMR 5558, laboratoire biostatistique-santé, université Lyon-1, 69100 Villeurbanne, France e Registre des cancers du Calvados, centre Fran¸ cois-Baclesse, 14076 Caen cedex 5, France f Registre des cancers du Tarn, CHS Pierre-Jamet, 81001 Albi cedex, France Available online 29 December 2009
Résumé Objectifs. – Fournir les tendances de l’incidence et de la mortalité ainsi que les survies à un et cinq ans associées au cancer thyroïdien par groupe histologique dans des régions franc¸aises couvertes par des registres. Matériel et méthodes. – Un registre spécialisé et 11 généraux ont fourni des données pour 1975 à 2004. Les estimations de l’incidence sont des corrections des incidences des registres ; des projections 2008 en dérivent. Le registre spécialisé et neuf autres ont fourni les survies globales et relatives par sexe et âge (diagnostics 1989 à 1997 ; point au 1er janvier 2002). Des taux standardisés-monde ont permis des comparaisons entre régions et périodes. Résultats. – Entre 1980 et 2005, l’incidence a augmenté mais la mortalité a diminué chez hommes et femmes. Les cas annuels ont quintuplé (projections 2008 : 8 000 cas et 400 décès) avec maximum pour le carcinome papillaire. Les survies globales et relatives ont été de 92 et 94 % à un an et de 87 et 93 % à cinq ans (> 94 % pour le carcinome papillaire et < 15 % pour le carcinome anaplasique). Généralement, les femmes ont mieux survécu ; précisément, mieux pour les carcinomes papillaire et folliculaire mais pas pour les carcinomes médullaire et anaplasique. Avec l’âge, les survies ont augmenté pour les carcinomes médullaires mais restaient basses dans les deux sexes après carcinome anaplasique. Conclusion. – La fréquence du cancer thyroïdien est en augmentation dramatique mais la survie est en amélioration. Bien que le pronostic des plus fréquentes formes soit généralement bon, trouver les causes de l’augmentation est nécessaire pour une prévention adéquate. © 2009 Elsevier Masson SAS. Tous droits réservés. Mots clés : Cancer thyroïdien ; Incidence ; Survie ; Mortalité ; Type histologique
DOI of original article:10.1016/j.ando.2009.11.003. Corresponding author. E-mail address: [email protected] (M. Colonna). 1 Réseau FRANCIM : Dr M. Velten, registre des cancers du Bas-Rhin ; Pr G. Launoy, registre digestif du Calvados ; Pr X. Troussard, registre régional des hémopathies malignes de Basse-Normandie ; Dr A.V. Guizard, registre général des tumeurs du Calvados ; Pr J. Faivre, registre bourguignon des cancers digestifs ; M. Colonna, registre des cancers de l’Isère ; Dr P. Arveux, registre gynéco Côte-d’Or ; Pr M. Maynadié, registre hémato Côte-d’Or ; Dr A. Danzon, registre des cancers du Doubs ; Pr H. Gouerou, registre digestif du Finistère ; Dr A. Buemi, registre des cancers du Haut-Rhin ; Dr B. Tretarre, registre des tumeurs de l’Hérault ; Dr S. Bara, registre des cancers de la Manche ; Pr Ganry, registre des cancers de la Somme ; Dr P. Grosclaude, registre des cancers du Tarn ; Dr A.M. Kadi-Hanifi, registre des cancers de la Vendée ; Dr F. Molinié, registre des cancers de Loire-Atlantique ; Dr A. Monnereau, registre des hémopathies malignes de la Gironde ; Dr C. Schvartz, registre des cancers thyroïdiens de la Marne et des Ardennes ; Dr I. Baldi, registre des tumeurs du système nerveux de la Gironde ; Dr M. Saves, étude de faisabilité du registre général des cancers de la Gironde ; E. Guerrin-Tran, étude de faisabilité du registre des cancers en Île-de-France ; Dr K. Ligier, étude de faisabilité du registre des cancers du Nord ; Dr M. Dieye, registre de la Martinique ; Dr J. Plenet, registre de la Guyanne. ∗
0003-4266/$ – see front matter © 2009 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.ando.2009.11.006
96
M. Colonna et al. / Annales d’Endocrinologie 71 (2010) 95–101
Abstract Objectives. – Describe time trends of incidence and mortality associated with thyroid cancer and provide 1 and 5-year survivals by histological group in French areas covered by cancer registries. Material and methods. – Data for 1975 to 2004 were provided by one thyroid-dedicated and 11 general registries. Incidence estimates were obtained by correction of incidence from areas with registries, then projections for 2008 were derived. Overall and relative survivals by sex and age (diagnosis period 1989–1997; cut-off date 1st January 2002) were obtained from the dedicated and nine other registries. Comparisons between areas or time periods used world-standardized rates. Results. – Between 1980 and 2005, incidence increased but mortality decreased in men and women. Annual cases increased five times and projections for 2008 were 8,000 cases and 400 deaths. The main increasing subtype was papillary carcinoma. One-year overall and relative survivals were 92 and 94%, respectively. Five-year overall and relative survivals were 87 and 93%, respectively. The highest survival (> 94%) concerned papillary carcinomas and the lowest (< 15%) anaplastic carcinomas. Survivals were generally higher in women than in men; precisely, higher in women for papillary and follicular carcinomas but higher in men for medullary and anaplastic carcinomas. Survivals increased with age, but for medullary carcinomas. Survivals from anaplastic carcinomas were very low whatever the age. Conclusion. – The increase of thyroid cancer frequency is dramatic but survivals are improving. Though the prognosis of the most increasing histological subtype is generally good, it remains very important to identify the causes of this steady increase to implement adequate preventive measures. © 2009 Elsevier Masson SAS. All rights reserved. Keywords: Thyroid cancer; Incidence; Survival; Mortality; Histological type
1. Introduction Until recently, in France, the incidence and mortality associated with thyroid cancer were not very high. However, although the number of deaths from thyroid cancer remains low (less than 1% of all cancer deaths in both sexes), the incidence of thyroid cancer in women is now the 5th leading cancer site; in 2005, it accounted for 3.7% of all cancers in women, whereas, in 1980, it accounted for only 1.4% of all newly diagnosed cancers and was ranking the 13th [1]. These facts reflect major changes in the frequency of this cancer over the last 20 years. In the present article, we describe the main features of: • the incidence of and mortality due to all subtypes of thyroid cancer in France; • the incidence by subtypes in French départements covered by cancer registries. We also report survival estimates relative to the various morphologies of this cancer. 2. Material The data used for the above estimates correspond to the period 1975 to 2004 and were supplied by one registry dedicated to thyroid cancers (Marne-Ardennes) and 11 other general registries (Doubs, Calvados, Isère, Bas-Rhin, Somme, Tarn, Haut-Rhin, Hérault, Manche, Loire-Atlantique, and Vendée) [1]. The detailed data, with classification by subtypes, were obtained from the seven oldest registries (Marne-Ardennes, Calvados, Doubs, Isère, Bas-Rhin, Somme, and Tarn). National incidence estimates per thyroid cancer subtypes cannot be obtained because National mortality data do not always mention these subtypes. Thus, the distribution of thyroid cancers according to subtypes had to be based on the more detailed cancer registry records. Survivals of patients aged 15 and over were obtained from the thyroid-specialized and nine other general registries (Loire Atlantique and Vendée were excluded because their data collection was only recently started). The period of diagnosis was
1989 to 1997 and the cut-off date was set at the 1st January 2002 with active search for vital status. The overall percentage of “lost to follow-ups” was 6.9% and ranged from 3.4% (for anaplastic thyroid cancer) to 8.2% (for follicular thyroid cancer). 3. Methods In the absence of a nationwide cancer registry, French national incidence estimates were obtained by applying a correction factor to incidence data from areas with registries. This correction factor was the ratio of the mortality over the whole country to the mortality in all registry areas. This method was detailed in an article by Belot et al. [1] and used to obtain national estimates for the period 1980 to 2005. The above-cited estimation method made it possible to derive incidence projections for 2008 [2]. These projections are subject to a high degree of uncertainty because they are based on the assumption that past trends will continue. To compare results between geographical areas or different time periods differing by population age structures, the rates were world-standardized. Survival analyses were carried out in terms of overall survival (death whatever its cause) and relative survival; the latter indicator enabled estimating cancer-specific survival, i.e. survival that would be observed if thyroid cancer were the only cause of patient death. This estimation was made using an approach described elsewhere [3–5]. Survival was estimated by sex, age, and period of diagnosis on all thyroid cancer subtypes confounded but by sex and age only per each subtype. 4. Results 4.1. 1980 and 2005 trends in thyroid cancer in France and projection for 2008 National French estimates of thyroid cancer showed constant and regular increases in incidence (Table 1 and Fig. 1) but
M. Colonna et al. / Annales d’Endocrinologie 71 (2010) 95–101
97
Table 1 Thyroïd cancer incidence and mortality rates in France (1980–2008; World-standardized per 100,000). Données d’incidence et de mortalité par cancer de la thyroïde en France (1980–2008 ; standardisées monde pour 100 000 personnes). Year
Mean yearly trend (%)
1980
1985
1990
1995
2000
2005
2008a
Incidence Men Women
1.0 2.9
1.3 3.9
1.7 5.2
2.3 7.0
3.1 9.5
4.2 12.7
5.0 15.2
5.8 6.0
6.4 6.1
Mortality Men Women
0.4 0.6
0.4 0.6
0.4 0.5
0.3 0.4
0.3 0.4
0.3 0.3
0.2 0.3
−1.8 −3.1
−2.7 −3.9
a
1980 to 2005
2000 to 2005
Data for 2008 correspond to projections based on the assumption that the trend observed over the period 1980 to 2004 will continue until 2008.
Fig. 1. Trends for incidence and mortality (World-standardized per 100,000). Tendances de l’incidence et de la mortalité (standardisées monde pour 100 000 personnes).
also constant and regular decreases in mortality rates in men and women. The annual trends over the 2000 to 2005 period showed that the increase in incidence and the drop in mortality were especially marked by the end of that period (Table 1). Throughout the period, the incidence was three times higher in women than in men. Between 1980 and 2005, there was a five-fold increase in the annual number of cases diagnosed in both sexes (Table 2). The largest part of that rise (82%) corresponded to an increased risk and the remaining part (18%) to demographic changes: population increase between 1980 and 2005 (14%) and progressive ageing in both sexes (4%). According to projections for 2008 made on the basis of trends observed between 1980 and 2004, the number of new cases of thyroid cancer should exceed 8,000 and the number of deaths be about 400.
Table 2 Number of thyroid cancer cases and deaths estimated in France (1980–2008). Cas et décès liés au cancer de la thyroïde en France (1980–2008). Year 1980
1985
1990
1995
2000
2005
2008 a
Incidence Men Women
325 1027
421 1352
571 1844
794 2572
1119 3607
1599 5073
1980 6214
Mortality Men Women
151 345
153 339
152 324
151 302
150 280
147 256
140 244
a Data for 2008 correspond to projections based on the assumption that the trend observed over the period 1980 to 2004 will continue until 2008.
98
M. Colonna et al. / Annales d’Endocrinologie 71 (2010) 95–101
Table 3 Distribution of new thyroid cancer cases by subtype, time period, and sex. Distribution des nouveaux cas de cancer de la thyroïde par sous-type, période et sexe. Papillary Casesa (%) Men 1980–84 1985–89 1990–94 1995–99 2000–04
62 (43.7) 101 (54.9) 157 (60.9) 251 (70.3) 439 (72.9)
Total
1010 (65.5)
Women 1980–84 1985–89 1990–94 1995–99 2000–04
214 (47.4) 391 (62.3) 666 (72.4) 1093 (78.8) 1630 (83.9)
Total
3994 (74.9)
a b
Follicular
Incidence rateb 0.53 0.74 1.11 1.67 2.77
Casesa (%)
33 (23.2) 41 (22.3) 60 (23.3) 49 (13.7) 83 (13.8)
Medullary
Incidence rateb 0.26 0.28 0.42 0.30 0.50
14 (9.9) 21 (11.4) 24 (9.3) 34 (9.5) 50 (8.3)
266 (17.2)
1.76 2.88 4.57 7.19 10.01
124 (27.4) 151 (24.0) 164 (17.8) 177 (12.8) 202 (10.4)
Casesa (%)
Anaplastic
Incidence rateb 0.12 0.15 0.16 0.25 0.33
143 (9.3)
0.96 1.02 1.03 1.07 1.18
22 (15.5) 17 (9.2) 11 (4.3) 17 (4.8) 19 (3.2)
Incidence rateb 0.17 0.12 0.06 0.09 0.08
86 (5.6)
15 (3.3) 31 (4.9) 37 (4.0) 54 (3.9) 52 (2.7)
818 (15.4)
Casesa (%)
Other
0.11 0.19 0.25 0.35 0.33
189 (3.6)
39 (8.6) 36 (5.7) 36 (3.9) 45 (3.2) 35 (1.8)
Casesa (%)
11 (7.8) 4 (2.2) 6 (2.3) 6 (1.7) 11 (1.8)
Incidence rateb 0.09 0.02 0.04 0.03 0.06
38 (2.5)
0.19 0.16 0.12 0.13 0.10
191 (3.6)
60 (13.3) 19 (3.0) 17 (1.9) 19 (1.4) 23 (1.2)
0.37 0.07 0.07 0.07 0.10
138 (2.6)
Tarn and Somme registries had no data for the years 1980 and 1981. World-standardized per 100,000.
decreased sharply in people over 75 years of whom almost a quarter had anaplastic carcinomas. In the recent period (2000–2004), the incidence of papillary thyroid cancer varied widely between départements (Fig. 2) whereas the incidence of other subtypes (non-papillary) was more homogeneous. Thus, incidence differences between the eight départements can be attributed to differences in observed papillary carcinomas in both sexes.
4.2. French registry data on the incidence of thyroid cancer by histological subtype Data in Table 3 show that the main increase in thyroid cancer incidence concerned the papillary subtype whereas the incidence of other subtypes (follicular and medullary) increased only slightly and that of the anaplastic subtype even decreased. Over 20 years, the incidence of papillary thyroid cancer was multiplied by 5.2 in men and 5.7 in women and the worldstandardized rates increased from 0.53 to 2.77 in men and from 1.76 to 10.0 in women. Between 1985 and 1989, the percentages of the papillary subtype relative to all thyroid cancers were 55% in men and 62% in women. From 2000 to 2004, these percentages increased up to 73% in men and 84% in women. The previous period (1980–1984) provided poor information because of a high percentage of “subtype: unspecified”. The distribution of thyroid cancers by subtypes was highly dependent on the age at diagnosis (Table 4). Although papillary carcinoma was the most frequent in all age groups, its proportion
4.3. French registry data on survival from thyroid cancer In general, the prognosis of thyroid cancer was very good (Table 5). At 1-year, the overall survival was 92% and the relative survival 94%. At 5-years, these survival rates were 87% and 93%, respectively. However, the estimates included two extremely contrasting situations: very high survival rates for papillary carcinoma (5-year overall and relative survivals of 94% and 99%, respectively) and very low survival rates for anaplastic carcinoma (5-year overall and relative survivals of 10% and 15%, respectively).
Table 4 Distribution of thyroid cancer cases by subtype, age and sex (1980–2004). Distribution des cas de cancer de la thyroïde par sous-type, âge et sexe (1980–2004). Papillary
Follicular
Medullary
Anaplastic
Other
Age
Case
%
Case
%
Case
%
Case
%
Case
%
< 15 15–45 45–55 55–65 65–75 75+ Men Women
24 2004 1274 968 528 206 1010 3994
60.0 80.3 79.4 74.0 60.6 37.3 65.5 74.9
2 358 218 209 176 121 266 818
5.0 14.4 13.6 16.0 20.2 21.9 17.2 15.4
14 100 69 70 55 24 143 189
35.0 4.0 4.3 5.4 6.3 4.3 9.3 3.6
0 5 22 38 85 127 86 191
0.0 0.2 1.4 2.9 9.8 23.0 5.6 3.6
0 28 21 24 28 75 38 138
0.0 1.1 1.3 1.8 3.2 13.6 2.5 2.6
M. Colonna et al. / Annales d’Endocrinologie 71 (2010) 95–101
99
Fig. 2. Papillary and non-papillary thyroid cancers: world-standardized incidence rates for 2000 to 2004 per 100,000 persons by sex and département (men, left panel vs women, right panel). Cancers papillaires et non papillaires de la thyroïde: taux d’incidence standardisés monde pour 2000 à 2004 et 100 000 personnes par sexe et département (hommes, partie gauche vs femmes, partie droite).
Considering all histological subtypes combined, survival was higher in women than in men. However, whereas survival was higher in women as regards papillary and follicular carcinomas, it was higher in men as regards medullary and anaplastic carcinomas. In women with anaplastic carcinoma, the 5-year overall and relative survivals were extremely low (6% and 9%, respectively). However, because the number of anaplastic carcinomas was low (93, of whom 64 women), the confidence intervals of the survival estimates were very wide. Generally, survival decreased with age but for medullary carcinoma; however, the latter subtype was rare. This decrease, expected for overall survival, was also found for relative survival. Survival decreased markedly after age 75 years. Considering all subtypes, the 5-year relative survival ranged from 98% in the 15 to 44 years old category to 51% in category 75 years and over. This decrease was less marked for papillary carcinoma; the 5year relative survival ranged from 100% in the 15 to 44 years old category to 80% in category 75 years and over. Concerning anaplastic carcinomas, survival was very low at all ages. Survival from thyroid cancer, all subtypes confounded, is improving over time; the 5-year relative survival increased from 89% for the cases diagnosed between 1989 and 1991 to 96% for the cases diagnosed between 1995 and 1997. That increase was also true for one-year survival. 5. Discussion In France, as in most Western countries, the incidence of thyroid cancer has been dramatically increasing over the last 20 years or more [6–7]. In 2005, thyroid cancer became the 5th ranking cancer site in France in terms of incidence in women [1]. This increase concerned almost exclusively papillary carcinomas and is most probably linked to technical and diagnostic progresses [8–9]. A previous description of trends in the incidence of papillary carcinomas by tumor size based on data from six French cancer registries has implicitly confirmed this increase by showing a very marked increase in micropapillary carcinomas, particularly in those less than 5 mm diameter [10]. Improvements in diagnosis may have also been responsible for the increase in the incidence of follicular carcinomas.
Thus, the near-stability in the rate of follicular carcinomas may seem surprising but can be explained by changes in the anatomopathological definitions. Actually, in 1975, the WHO guidelines stipulated that any follicular carcinoma with a papillary component should be classified as papillary. Also, in 1988, nuclear criteria for papillary carcinomas were introduced resulting in tumors with a purely follicular architecture being classified as papillary [11]. These guidelines were gradually adopted in France by all anatomopathologists and used by registries in the late 1980s and early 1990s. Thus, a large proportion of carcinomas previously classified as follicular or mixed are now classified as papillary. Another factor that may specifically explain the decrease in follicular carcinomas is the disappearance of iodine deficiency. Among the hypotheses put forward to explain the increased incidence of thyroid cancer is the impact of the radioactive fallout from the Chernobyl nuclear accident. Although this hypothesis cannot be completely discarded, especially that results of several French case-control studies are awaited, it does not seem very plausible given the regular increases both in time (already before 1986) and in place (in areas less affected by the fallout, such as the Tarn Département in Southwestern France). A recent publication by Cardis et al. [12] has shown that, except in heavily contaminated areas, it is not easy or possible to show correlations between increased incidence and radioactive fallouts. The scientific litterature mentioned several other risk factors, essentially occupational and environmental. On this subject, one may refer to a recent review by Leux et Guénel [13] who underscored the genotoxicity of a number chemical products. The heterogeneity of cancer incidence in different French départements is mirrored worldwide. In the latest report on Cancer incidence in five continents that incorporated data from 300 registries and covered the period 1998 to 2002 [14], the registry of département Tarn, near the Spanish border, reported the 5th highest incidence, whereas that of département Bas-Rhin, on the German border, reported the 194th. In the present study, survival from papillary or follicular carcinomas was higher in women than in men but survival from medullary and anaplastic carcinomas was higher in men than
100
M. Colonna et al. / Annales d’Endocrinologie 71 (2010) 95–101
Table 5 One- and 5-year overall and relative survivals by sex and age. Survies globales et relatives à un an et à cinq ans par âge et sexe. 1-year survival
5-year survival
Overall (95% CI)
Relative (95% CI)
Overall (95% CI)
Relative (95% CI)
95 (92–97) 98 (98–99) 100 (99–100) 99 (98–99) 93 (89–96) 82 (72–88)
98 (95–99) 100 (100–100) 100 (100–100) 98 (96–99) 95 (90–97) 85 (73–92)
88 (84–91) 96 (95–97) 98 (97–99) 94 (91–95) 81 (75–86) 56 (44–67)
95 (91–98) 99 (99–100) 100 (99–100) 98 (89–99) 92 (83–96) 80 (53–93)
All
98 (97–98)
99 (99–100)
95 (93–96)
99 (98–100)
Follicular Men Women [15;55[ [55;65[ [65;75[ [75;++[
91 (85–94) 95 (92–97) 99 (97–100) 98 (97–99) 93 (91–95) 68 (53–80)
92 (85–96) 100 (99–100) 99 (96–100) 96 (88–99) 96 (93–97) 75 (58–86)
75 (65–82) 87 (83–91) 98 (94–99) 91 (84–95) 71 (62–79) 38 (24–52)
85 (74–92) 98 (95–99) 99 (84–100) 95 (55–100) 80 (69–87) 59 (31–78)
All
94 (91–96)
97 (94–98)
85 (81–88)
95 (91–97)
Medullary Men Women [15;55[ [55;65[ [65;75[ [75;++[
97 (94–98) 90 (83–94) 96 (89–98) 98 (95–99) 91 (85–95) 46 (19–70)
99 (96–100) 89 (81–94) 96 (89–99) 99 (94–100) 83 (63–93) 87 (68–95)
84 (73–91) 79 (69–86) 92 (82–96) 90 (77–96) 64 (45–78) 24 (5–50)
96 (81–99) 85 (73–91) 92 (81–97) 96 (73–99) 72 (40–89) 49 (15–77)
All
91 (85–95)
92 (86–96)
82 (74–88)
88 (79–93)
Anaplastic Men Women [55;65[ [65;75[ [75;++[
31 (17–47) 14 (7–22) 18 (4–40) 15 (5–30) 14 (7–23)
33 (18–49) 12 (6–20) 18 (4–40) 15 (5–31) 16 (8–27)
20 (8–36) 6 (2–14) 10 (0–46) 10 (1–32) 6 (2–15)
27 (11–47) 9 (0–49) 10 (0–51) 10 (1–38) 11 (3–23)
All
18 (12–26)
20 (13–28)
10 (5–17)
15 (8–24)
All subtypes Men Women [15;55[ [55;65[ [65;75[ [75;++[ All ages 1989–1991 1992–1994 1995–1997
88 (85–90) 93 (92–94) 99 (99–99) 93 (91–95) 83 (79–87) 53 (47–59) 92 (91–93) 88 (85–90) 92 (89–93) 95 (93–96)
91 (88–93) 96 (95–97) 99 (99–100) 94 (92–96) 85 (81–88) 58 (51–65) 95 (94–96) 90 (88–92) 94 (92–96) 97 (96–98)
78 (75–82) 89 (87–90) 97 (96–98) 89 (86–92) 69 (64–73) 33 (27–39) 87 (85–88) 81 (77–83) 87 (84–89) 91 (88–92)
88 (85–91) 95 (94–96) 98 (97–99) 94 (90–96) 77 (71–82) 51 (22–74) 94 (92–95) 89 (86–91) 94 (92–96) 96 (94–98)
All cases
92 (91–93)
94 (93–95)
87 (85–88)
93 (92–94)
Papillary Men Women [15;55[ [55;65[ [65;75[ 75;++[
Overall survival: observed survival; relative survival: survival that would be observed in the absence of non-cancer-related deaths; CI: confidence interval.
in women. Let us note however that the estimates relative to the latter two subtypes were not very accurate. Because most of these differences are not very marked, they are seldom mentioned in the scientific literature. Akslen et al. [15] did not find sex-specific differences concerning survival. Cunningham et al. [16] and Colonna et al. [17] have already mentioned a higher survival in women as regards papillary and follicular carcino-
mas. Guilliland et al. [18] observed similar survivals in men and women as to papillary and follicular carcinomas, poorer survivals in women than in men as to anaplastic carcinomas, and higher survivals in women than in men as to medullary carcinomas. Finally, contrarily to our observations on medullary carcinomas, Esik et al. [19] observed higher survivals in women than in men.
M. Colonna et al. / Annales d’Endocrinologie 71 (2010) 95–101
Age-related differences were much greater than sex-related ones. Moreover, the literature has confirmed that age is an important prognostic factor [15–19]; indeed, it was the only prognostic factor we kept in our analyses. However, another very influential prognostic factor, the cancer stage at the time of diagnosis, should have been considered as well but that information was not available for all analyzed cases. The effect of cancer stage at diagnosis is certainly associated with the very high survivals of patients with papillary carcinoma and the survival increase over the period 1987 to 1994 [5], during which small tumors were predominant [10]. 6. Conclusion The ever increasing frequency of thyroid cancer is becoming a real public health problem in women, even though the specific survival is extremely high. Thus, as some authors [20] seem to note a slow-down of that frequency in the French Rhône-Alpes Region (eight départements) whereas others report completely different results elsewhere, such as in the United States [21], it remains of utmost importance to closely monitor the disease trends and make better estimates of required resources. 7. Conflicts of interest The authors have not declared any conflict of interest. 8. French version A french version of this doi:10.1016/j.ando.2009.11.003.
article
is
available
at
Acknowledgements This study received funding from the Ligue Nationale Contre le Cancer, (National Cancer Research Fund) and the Comité de l’Isère (local branch of the National Cancer Research Fund). The authors thank Jean Iwaz (hospices civils de Lyon) for a thorough revision of the manuscript. References [1] Belot A, Grosclaude P, Bossard N, et al. Cancer incidence and mortality in France over the period 1980–2005. Rev Epidemiol Sante Publ 2008;56:159–75. [2] Hospices Civils de Lyon – Institut de veille sanitaire – Institut National du Cancer – Francim – Institut National de la Santé et de la Recherche médicale. Projections de l’incidence et de la mortalité par cancer en France pour l’année 2008. Rapport technique, mars 2009. (http://www.invs.sante.fr/surveillance/cancers).
101
[3] Remontet L, Bossard N, Belot A, Estève J, the French network of cancer registries (FRANCIM). An overall strategy based on regression models to estimate relative survival and model the effects of prognostic factors in cancer survival studies. Stat Med 2007;26:2214–28. [4] Bossard N, Velten M, Remontet L, et al. Survival of cancer patients in France: A population-based study from the Association of the French Cancer Registries (FRANCIM). Eur J Cancer 2007;43:149–60. [5] Réseau FRANCIM. Survie des patients atteints de cancer en France. Étude des registres de cancers du réseau FRANCIM. Paris: Springer-Verlag; 2007. [6] Colonna M, Grosclaude P, Remontet L, et al. Incidence of thyroid cancer in adults recorded by French cancer registries (1978–1997). Eur J Cancer 2002;38:1762–8. [7] Chérié-Challine L, Les membres du comité de rédaction. Surveillance sanitaire en lien avec l’accident de Tchernobyl en France – bilan actualisé sur les cancers thyroïdiens et études épidémiologiques en cours en 2006. Saint-Maurice: Institut de veille sanitaire; 2006. www.invs.sante.fr. [8] Verkooijen HM, Fioretta G, Pache JC, et al. Diagnostic changes as a reason for the increase in papillary thyroid cancer incidence in Geneva, Switzerland. Cancer Causes Control 2003;14:13–7. [9] Leenhardt L, Bernier MO, Boin-Pineau MH, et al. Advances in diagnosis practices affect thyroid cancer incidence in France. Eur J Endocrinol 2004;150:133–9. [10] Colonna M, Guizard AV, Schartz C, et al. Time trend analysis of papillary and follicular cancers as a function of tumour size: A study of data from six cancer registries in France (1983–2000). Eur J Cancer 2007;43:891–900. [11] Hedinger C, Williams ED, Sobin LH. The WHO histological classification of thyroid tumors: a commentary on the second edition. Cancer 1989;63:908–11. [12] Cardis E, Krewski D, Boniol M, et al. Estimates of the cancer burden in Europe from radioactive fallout from the Chernobyl accident. Int J Cancer 2006;119:1224–35. [13] Leux C, Guénel P. Facteurs de risque des cancers et nodules thyroïdiens : effet des polluants de l’environnement et risques professionnels. Med Clin Endocrinol Diabete 2009:1–7. [14] Curado P, Edwards B, Shin H, Storm H, Ferlay J, Heanue M, Boyle P. Cancer incidence in five continents, Vol. IX. IARC Scientific Publications 160, Lyon, International Agency for Research on Cancer, 2007. [15] Akslen L, Haldorsen T, Thoresen SO, Glattre E. Survival and causes of death in thyroid cancer: A population-based Study of 2479 Cases from Norway. Cancer Res 1991;51:1234–41. [16] Cunningham MP, Duda RB, Recant W, Cmiel JS, Sylvester JA, Fremgen A. Survival discriminants for differentiated thyroid cancer. Am J Surg 1990;160:344–7. [17] Colonna M, Grande E, Jonasson JG, the EUROCARE Working Group. Cancer Variation in relative survival of thyroid cancers in Europe: Results from the analysis on 21 countries over 4 the period 1983–1994 (EUROCARE-3 study). Eur J Cancer 2006;42:2598–608. [18] Guilliland FD, Hunt WC, Morris DM, Key XR. Prognostic factors for thyroid carcinoma. Cancer 1997;19:564–73. [19] Esik O, Tusnády G, Trón L, et al. Markov model-based estimation of individual survival probability for medullary thyroid cancer patients. Pathol Oncol Res 2002;8:93–104. [20] Sassolas G, Hafdi-Nejjari1 Z, Remontet L, et al. Thyroid cancer: is the incidence rise abating? Eur J Endocrinol 2008;159:1–10. [21] Ries L, Melbert D, Krapcho M, Stinchcomb DG et al. SEER Cancer Statistics Review, 1975–2005, National Cancer Institute. Bethesda, 2009. (http://seer.cancer.gov/csr/1975 2005/).
Original article
Descriptive epidemiology of thyroid cancer in France: Incidence, mortality and survival Épidémiologie descriptive du cancer thyroïdien en France : incidence, mortalité et survie M. Colonna a,∗,b , N. Bossard c,d , A.-V. Guizard b,e , L. Remontet c,d , P. Grosclaude b,f , le réseau FRANCIM1 a
Registre du cancer de l’Isère, 23, chemin-des-Sources, 38240 Meylan, France b FRANCIM, faculté de médecine, 31073 Toulouse cedex, France c Service de biostatistique, hospices civils de Lyon, 69003 Lyon, France d CNRS, UMR 5558, laboratoire biostatistique-santé, université Lyon-1, 69100 Villeurbanne, France e Registre des cancers du Calvados, centre Fran¸ cois-Baclesse, 14076 Caen cedex 5, France f Registre des cancers du Tarn, CHS Pierre-Jamet, 81001 Albi cedex, France Available online 29 December 2009
Résumé Objectifs. – Fournir les tendances de l’incidence et de la mortalité ainsi que les survies à un et cinq ans associées au cancer thyroïdien par groupe histologique dans des régions franc¸aises couvertes par des registres. Matériel et méthodes. – Un registre spécialisé et 11 généraux ont fourni des données pour 1975 à 2004. Les estimations de l’incidence sont des corrections des incidences des registres ; des projections 2008 en dérivent. Le registre spécialisé et neuf autres ont fourni les survies globales et relatives par sexe et âge (diagnostics 1989 à 1997 ; point au 1er janvier 2002). Des taux standardisés-monde ont permis des comparaisons entre régions et périodes. Résultats. – Entre 1980 et 2005, l’incidence a augmenté mais la mortalité a diminué chez hommes et femmes. Les cas annuels ont quintuplé (projections 2008 : 8 000 cas et 400 décès) avec maximum pour le carcinome papillaire. Les survies globales et relatives ont été de 92 et 94 % à un an et de 87 et 93 % à cinq ans (> 94 % pour le carcinome papillaire et < 15 % pour le carcinome anaplasique). Généralement, les femmes ont mieux survécu ; précisément, mieux pour les carcinomes papillaire et folliculaire mais pas pour les carcinomes médullaire et anaplasique. Avec l’âge, les survies ont augmenté pour les carcinomes médullaires mais restaient basses dans les deux sexes après carcinome anaplasique. Conclusion. – La fréquence du cancer thyroïdien est en augmentation dramatique mais la survie est en amélioration. Bien que le pronostic des plus fréquentes formes soit généralement bon, trouver les causes de l’augmentation est nécessaire pour une prévention adéquate. © 2009 Elsevier Masson SAS. Tous droits réservés. Mots clés : Cancer thyroïdien ; Incidence ; Survie ; Mortalité ; Type histologique
DOI of original article:10.1016/j.ando.2009.11.003. Corresponding author. E-mail address: [email protected] (M. Colonna). 1 Réseau FRANCIM : Dr M. Velten, registre des cancers du Bas-Rhin ; Pr G. Launoy, registre digestif du Calvados ; Pr X. Troussard, registre régional des hémopathies malignes de Basse-Normandie ; Dr A.V. Guizard, registre général des tumeurs du Calvados ; Pr J. Faivre, registre bourguignon des cancers digestifs ; M. Colonna, registre des cancers de l’Isère ; Dr P. Arveux, registre gynéco Côte-d’Or ; Pr M. Maynadié, registre hémato Côte-d’Or ; Dr A. Danzon, registre des cancers du Doubs ; Pr H. Gouerou, registre digestif du Finistère ; Dr A. Buemi, registre des cancers du Haut-Rhin ; Dr B. Tretarre, registre des tumeurs de l’Hérault ; Dr S. Bara, registre des cancers de la Manche ; Pr Ganry, registre des cancers de la Somme ; Dr P. Grosclaude, registre des cancers du Tarn ; Dr A.M. Kadi-Hanifi, registre des cancers de la Vendée ; Dr F. Molinié, registre des cancers de Loire-Atlantique ; Dr A. Monnereau, registre des hémopathies malignes de la Gironde ; Dr C. Schvartz, registre des cancers thyroïdiens de la Marne et des Ardennes ; Dr I. Baldi, registre des tumeurs du système nerveux de la Gironde ; Dr M. Saves, étude de faisabilité du registre général des cancers de la Gironde ; E. Guerrin-Tran, étude de faisabilité du registre des cancers en Île-de-France ; Dr K. Ligier, étude de faisabilité du registre des cancers du Nord ; Dr M. Dieye, registre de la Martinique ; Dr J. Plenet, registre de la Guyanne. ∗
0003-4266/$ – see front matter © 2009 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.ando.2009.11.006
96
M. Colonna et al. / Annales d’Endocrinologie 71 (2010) 95–101
Abstract Objectives. – Describe time trends of incidence and mortality associated with thyroid cancer and provide 1 and 5-year survivals by histological group in French areas covered by cancer registries. Material and methods. – Data for 1975 to 2004 were provided by one thyroid-dedicated and 11 general registries. Incidence estimates were obtained by correction of incidence from areas with registries, then projections for 2008 were derived. Overall and relative survivals by sex and age (diagnosis period 1989–1997; cut-off date 1st January 2002) were obtained from the dedicated and nine other registries. Comparisons between areas or time periods used world-standardized rates. Results. – Between 1980 and 2005, incidence increased but mortality decreased in men and women. Annual cases increased five times and projections for 2008 were 8,000 cases and 400 deaths. The main increasing subtype was papillary carcinoma. One-year overall and relative survivals were 92 and 94%, respectively. Five-year overall and relative survivals were 87 and 93%, respectively. The highest survival (> 94%) concerned papillary carcinomas and the lowest (< 15%) anaplastic carcinomas. Survivals were generally higher in women than in men; precisely, higher in women for papillary and follicular carcinomas but higher in men for medullary and anaplastic carcinomas. Survivals increased with age, but for medullary carcinomas. Survivals from anaplastic carcinomas were very low whatever the age. Conclusion. – The increase of thyroid cancer frequency is dramatic but survivals are improving. Though the prognosis of the most increasing histological subtype is generally good, it remains very important to identify the causes of this steady increase to implement adequate preventive measures. © 2009 Elsevier Masson SAS. All rights reserved. Keywords: Thyroid cancer; Incidence; Survival; Mortality; Histological type
1. Introduction Until recently, in France, the incidence and mortality associated with thyroid cancer were not very high. However, although the number of deaths from thyroid cancer remains low (less than 1% of all cancer deaths in both sexes), the incidence of thyroid cancer in women is now the 5th leading cancer site; in 2005, it accounted for 3.7% of all cancers in women, whereas, in 1980, it accounted for only 1.4% of all newly diagnosed cancers and was ranking the 13th [1]. These facts reflect major changes in the frequency of this cancer over the last 20 years. In the present article, we describe the main features of: • the incidence of and mortality due to all subtypes of thyroid cancer in France; • the incidence by subtypes in French départements covered by cancer registries. We also report survival estimates relative to the various morphologies of this cancer. 2. Material The data used for the above estimates correspond to the period 1975 to 2004 and were supplied by one registry dedicated to thyroid cancers (Marne-Ardennes) and 11 other general registries (Doubs, Calvados, Isère, Bas-Rhin, Somme, Tarn, Haut-Rhin, Hérault, Manche, Loire-Atlantique, and Vendée) [1]. The detailed data, with classification by subtypes, were obtained from the seven oldest registries (Marne-Ardennes, Calvados, Doubs, Isère, Bas-Rhin, Somme, and Tarn). National incidence estimates per thyroid cancer subtypes cannot be obtained because National mortality data do not always mention these subtypes. Thus, the distribution of thyroid cancers according to subtypes had to be based on the more detailed cancer registry records. Survivals of patients aged 15 and over were obtained from the thyroid-specialized and nine other general registries (Loire Atlantique and Vendée were excluded because their data collection was only recently started). The period of diagnosis was
1989 to 1997 and the cut-off date was set at the 1st January 2002 with active search for vital status. The overall percentage of “lost to follow-ups” was 6.9% and ranged from 3.4% (for anaplastic thyroid cancer) to 8.2% (for follicular thyroid cancer). 3. Methods In the absence of a nationwide cancer registry, French national incidence estimates were obtained by applying a correction factor to incidence data from areas with registries. This correction factor was the ratio of the mortality over the whole country to the mortality in all registry areas. This method was detailed in an article by Belot et al. [1] and used to obtain national estimates for the period 1980 to 2005. The above-cited estimation method made it possible to derive incidence projections for 2008 [2]. These projections are subject to a high degree of uncertainty because they are based on the assumption that past trends will continue. To compare results between geographical areas or different time periods differing by population age structures, the rates were world-standardized. Survival analyses were carried out in terms of overall survival (death whatever its cause) and relative survival; the latter indicator enabled estimating cancer-specific survival, i.e. survival that would be observed if thyroid cancer were the only cause of patient death. This estimation was made using an approach described elsewhere [3–5]. Survival was estimated by sex, age, and period of diagnosis on all thyroid cancer subtypes confounded but by sex and age only per each subtype. 4. Results 4.1. 1980 and 2005 trends in thyroid cancer in France and projection for 2008 National French estimates of thyroid cancer showed constant and regular increases in incidence (Table 1 and Fig. 1) but
M. Colonna et al. / Annales d’Endocrinologie 71 (2010) 95–101
97
Table 1 Thyroïd cancer incidence and mortality rates in France (1980–2008; World-standardized per 100,000). Données d’incidence et de mortalité par cancer de la thyroïde en France (1980–2008 ; standardisées monde pour 100 000 personnes). Year
Mean yearly trend (%)
1980
1985
1990
1995
2000
2005
2008a
Incidence Men Women
1.0 2.9
1.3 3.9
1.7 5.2
2.3 7.0
3.1 9.5
4.2 12.7
5.0 15.2
5.8 6.0
6.4 6.1
Mortality Men Women
0.4 0.6
0.4 0.6
0.4 0.5
0.3 0.4
0.3 0.4
0.3 0.3
0.2 0.3
−1.8 −3.1
−2.7 −3.9
a
1980 to 2005
2000 to 2005
Data for 2008 correspond to projections based on the assumption that the trend observed over the period 1980 to 2004 will continue until 2008.
Fig. 1. Trends for incidence and mortality (World-standardized per 100,000). Tendances de l’incidence et de la mortalité (standardisées monde pour 100 000 personnes).
also constant and regular decreases in mortality rates in men and women. The annual trends over the 2000 to 2005 period showed that the increase in incidence and the drop in mortality were especially marked by the end of that period (Table 1). Throughout the period, the incidence was three times higher in women than in men. Between 1980 and 2005, there was a five-fold increase in the annual number of cases diagnosed in both sexes (Table 2). The largest part of that rise (82%) corresponded to an increased risk and the remaining part (18%) to demographic changes: population increase between 1980 and 2005 (14%) and progressive ageing in both sexes (4%). According to projections for 2008 made on the basis of trends observed between 1980 and 2004, the number of new cases of thyroid cancer should exceed 8,000 and the number of deaths be about 400.
Table 2 Number of thyroid cancer cases and deaths estimated in France (1980–2008). Cas et décès liés au cancer de la thyroïde en France (1980–2008). Year 1980
1985
1990
1995
2000
2005
2008 a
Incidence Men Women
325 1027
421 1352
571 1844
794 2572
1119 3607
1599 5073
1980 6214
Mortality Men Women
151 345
153 339
152 324
151 302
150 280
147 256
140 244
a Data for 2008 correspond to projections based on the assumption that the trend observed over the period 1980 to 2004 will continue until 2008.
98
M. Colonna et al. / Annales d’Endocrinologie 71 (2010) 95–101
Table 3 Distribution of new thyroid cancer cases by subtype, time period, and sex. Distribution des nouveaux cas de cancer de la thyroïde par sous-type, période et sexe. Papillary Casesa (%) Men 1980–84 1985–89 1990–94 1995–99 2000–04
62 (43.7) 101 (54.9) 157 (60.9) 251 (70.3) 439 (72.9)
Total
1010 (65.5)
Women 1980–84 1985–89 1990–94 1995–99 2000–04
214 (47.4) 391 (62.3) 666 (72.4) 1093 (78.8) 1630 (83.9)
Total
3994 (74.9)
a b
Follicular
Incidence rateb 0.53 0.74 1.11 1.67 2.77
Casesa (%)
33 (23.2) 41 (22.3) 60 (23.3) 49 (13.7) 83 (13.8)
Medullary
Incidence rateb 0.26 0.28 0.42 0.30 0.50
14 (9.9) 21 (11.4) 24 (9.3) 34 (9.5) 50 (8.3)
266 (17.2)
1.76 2.88 4.57 7.19 10.01
124 (27.4) 151 (24.0) 164 (17.8) 177 (12.8) 202 (10.4)
Casesa (%)
Anaplastic
Incidence rateb 0.12 0.15 0.16 0.25 0.33
143 (9.3)
0.96 1.02 1.03 1.07 1.18
22 (15.5) 17 (9.2) 11 (4.3) 17 (4.8) 19 (3.2)
Incidence rateb 0.17 0.12 0.06 0.09 0.08
86 (5.6)
15 (3.3) 31 (4.9) 37 (4.0) 54 (3.9) 52 (2.7)
818 (15.4)
Casesa (%)
Other
0.11 0.19 0.25 0.35 0.33
189 (3.6)
39 (8.6) 36 (5.7) 36 (3.9) 45 (3.2) 35 (1.8)
Casesa (%)
11 (7.8) 4 (2.2) 6 (2.3) 6 (1.7) 11 (1.8)
Incidence rateb 0.09 0.02 0.04 0.03 0.06
38 (2.5)
0.19 0.16 0.12 0.13 0.10
191 (3.6)
60 (13.3) 19 (3.0) 17 (1.9) 19 (1.4) 23 (1.2)
0.37 0.07 0.07 0.07 0.10
138 (2.6)
Tarn and Somme registries had no data for the years 1980 and 1981. World-standardized per 100,000.
decreased sharply in people over 75 years of whom almost a quarter had anaplastic carcinomas. In the recent period (2000–2004), the incidence of papillary thyroid cancer varied widely between départements (Fig. 2) whereas the incidence of other subtypes (non-papillary) was more homogeneous. Thus, incidence differences between the eight départements can be attributed to differences in observed papillary carcinomas in both sexes.
4.2. French registry data on the incidence of thyroid cancer by histological subtype Data in Table 3 show that the main increase in thyroid cancer incidence concerned the papillary subtype whereas the incidence of other subtypes (follicular and medullary) increased only slightly and that of the anaplastic subtype even decreased. Over 20 years, the incidence of papillary thyroid cancer was multiplied by 5.2 in men and 5.7 in women and the worldstandardized rates increased from 0.53 to 2.77 in men and from 1.76 to 10.0 in women. Between 1985 and 1989, the percentages of the papillary subtype relative to all thyroid cancers were 55% in men and 62% in women. From 2000 to 2004, these percentages increased up to 73% in men and 84% in women. The previous period (1980–1984) provided poor information because of a high percentage of “subtype: unspecified”. The distribution of thyroid cancers by subtypes was highly dependent on the age at diagnosis (Table 4). Although papillary carcinoma was the most frequent in all age groups, its proportion
4.3. French registry data on survival from thyroid cancer In general, the prognosis of thyroid cancer was very good (Table 5). At 1-year, the overall survival was 92% and the relative survival 94%. At 5-years, these survival rates were 87% and 93%, respectively. However, the estimates included two extremely contrasting situations: very high survival rates for papillary carcinoma (5-year overall and relative survivals of 94% and 99%, respectively) and very low survival rates for anaplastic carcinoma (5-year overall and relative survivals of 10% and 15%, respectively).
Table 4 Distribution of thyroid cancer cases by subtype, age and sex (1980–2004). Distribution des cas de cancer de la thyroïde par sous-type, âge et sexe (1980–2004). Papillary
Follicular
Medullary
Anaplastic
Other
Age
Case
%
Case
%
Case
%
Case
%
Case
%
< 15 15–45 45–55 55–65 65–75 75+ Men Women
24 2004 1274 968 528 206 1010 3994
60.0 80.3 79.4 74.0 60.6 37.3 65.5 74.9
2 358 218 209 176 121 266 818
5.0 14.4 13.6 16.0 20.2 21.9 17.2 15.4
14 100 69 70 55 24 143 189
35.0 4.0 4.3 5.4 6.3 4.3 9.3 3.6
0 5 22 38 85 127 86 191
0.0 0.2 1.4 2.9 9.8 23.0 5.6 3.6
0 28 21 24 28 75 38 138
0.0 1.1 1.3 1.8 3.2 13.6 2.5 2.6
M. Colonna et al. / Annales d’Endocrinologie 71 (2010) 95–101
99
Fig. 2. Papillary and non-papillary thyroid cancers: world-standardized incidence rates for 2000 to 2004 per 100,000 persons by sex and département (men, left panel vs women, right panel). Cancers papillaires et non papillaires de la thyroïde: taux d’incidence standardisés monde pour 2000 à 2004 et 100 000 personnes par sexe et département (hommes, partie gauche vs femmes, partie droite).
Considering all histological subtypes combined, survival was higher in women than in men. However, whereas survival was higher in women as regards papillary and follicular carcinomas, it was higher in men as regards medullary and anaplastic carcinomas. In women with anaplastic carcinoma, the 5-year overall and relative survivals were extremely low (6% and 9%, respectively). However, because the number of anaplastic carcinomas was low (93, of whom 64 women), the confidence intervals of the survival estimates were very wide. Generally, survival decreased with age but for medullary carcinoma; however, the latter subtype was rare. This decrease, expected for overall survival, was also found for relative survival. Survival decreased markedly after age 75 years. Considering all subtypes, the 5-year relative survival ranged from 98% in the 15 to 44 years old category to 51% in category 75 years and over. This decrease was less marked for papillary carcinoma; the 5year relative survival ranged from 100% in the 15 to 44 years old category to 80% in category 75 years and over. Concerning anaplastic carcinomas, survival was very low at all ages. Survival from thyroid cancer, all subtypes confounded, is improving over time; the 5-year relative survival increased from 89% for the cases diagnosed between 1989 and 1991 to 96% for the cases diagnosed between 1995 and 1997. That increase was also true for one-year survival. 5. Discussion In France, as in most Western countries, the incidence of thyroid cancer has been dramatically increasing over the last 20 years or more [6–7]. In 2005, thyroid cancer became the 5th ranking cancer site in France in terms of incidence in women [1]. This increase concerned almost exclusively papillary carcinomas and is most probably linked to technical and diagnostic progresses [8–9]. A previous description of trends in the incidence of papillary carcinomas by tumor size based on data from six French cancer registries has implicitly confirmed this increase by showing a very marked increase in micropapillary carcinomas, particularly in those less than 5 mm diameter [10]. Improvements in diagnosis may have also been responsible for the increase in the incidence of follicular carcinomas.
Thus, the near-stability in the rate of follicular carcinomas may seem surprising but can be explained by changes in the anatomopathological definitions. Actually, in 1975, the WHO guidelines stipulated that any follicular carcinoma with a papillary component should be classified as papillary. Also, in 1988, nuclear criteria for papillary carcinomas were introduced resulting in tumors with a purely follicular architecture being classified as papillary [11]. These guidelines were gradually adopted in France by all anatomopathologists and used by registries in the late 1980s and early 1990s. Thus, a large proportion of carcinomas previously classified as follicular or mixed are now classified as papillary. Another factor that may specifically explain the decrease in follicular carcinomas is the disappearance of iodine deficiency. Among the hypotheses put forward to explain the increased incidence of thyroid cancer is the impact of the radioactive fallout from the Chernobyl nuclear accident. Although this hypothesis cannot be completely discarded, especially that results of several French case-control studies are awaited, it does not seem very plausible given the regular increases both in time (already before 1986) and in place (in areas less affected by the fallout, such as the Tarn Département in Southwestern France). A recent publication by Cardis et al. [12] has shown that, except in heavily contaminated areas, it is not easy or possible to show correlations between increased incidence and radioactive fallouts. The scientific litterature mentioned several other risk factors, essentially occupational and environmental. On this subject, one may refer to a recent review by Leux et Guénel [13] who underscored the genotoxicity of a number chemical products. The heterogeneity of cancer incidence in different French départements is mirrored worldwide. In the latest report on Cancer incidence in five continents that incorporated data from 300 registries and covered the period 1998 to 2002 [14], the registry of département Tarn, near the Spanish border, reported the 5th highest incidence, whereas that of département Bas-Rhin, on the German border, reported the 194th. In the present study, survival from papillary or follicular carcinomas was higher in women than in men but survival from medullary and anaplastic carcinomas was higher in men than
100
M. Colonna et al. / Annales d’Endocrinologie 71 (2010) 95–101
Table 5 One- and 5-year overall and relative survivals by sex and age. Survies globales et relatives à un an et à cinq ans par âge et sexe. 1-year survival
5-year survival
Overall (95% CI)
Relative (95% CI)
Overall (95% CI)
Relative (95% CI)
95 (92–97) 98 (98–99) 100 (99–100) 99 (98–99) 93 (89–96) 82 (72–88)
98 (95–99) 100 (100–100) 100 (100–100) 98 (96–99) 95 (90–97) 85 (73–92)
88 (84–91) 96 (95–97) 98 (97–99) 94 (91–95) 81 (75–86) 56 (44–67)
95 (91–98) 99 (99–100) 100 (99–100) 98 (89–99) 92 (83–96) 80 (53–93)
All
98 (97–98)
99 (99–100)
95 (93–96)
99 (98–100)
Follicular Men Women [15;55[ [55;65[ [65;75[ [75;++[
91 (85–94) 95 (92–97) 99 (97–100) 98 (97–99) 93 (91–95) 68 (53–80)
92 (85–96) 100 (99–100) 99 (96–100) 96 (88–99) 96 (93–97) 75 (58–86)
75 (65–82) 87 (83–91) 98 (94–99) 91 (84–95) 71 (62–79) 38 (24–52)
85 (74–92) 98 (95–99) 99 (84–100) 95 (55–100) 80 (69–87) 59 (31–78)
All
94 (91–96)
97 (94–98)
85 (81–88)
95 (91–97)
Medullary Men Women [15;55[ [55;65[ [65;75[ [75;++[
97 (94–98) 90 (83–94) 96 (89–98) 98 (95–99) 91 (85–95) 46 (19–70)
99 (96–100) 89 (81–94) 96 (89–99) 99 (94–100) 83 (63–93) 87 (68–95)
84 (73–91) 79 (69–86) 92 (82–96) 90 (77–96) 64 (45–78) 24 (5–50)
96 (81–99) 85 (73–91) 92 (81–97) 96 (73–99) 72 (40–89) 49 (15–77)
All
91 (85–95)
92 (86–96)
82 (74–88)
88 (79–93)
Anaplastic Men Women [55;65[ [65;75[ [75;++[
31 (17–47) 14 (7–22) 18 (4–40) 15 (5–30) 14 (7–23)
33 (18–49) 12 (6–20) 18 (4–40) 15 (5–31) 16 (8–27)
20 (8–36) 6 (2–14) 10 (0–46) 10 (1–32) 6 (2–15)
27 (11–47) 9 (0–49) 10 (0–51) 10 (1–38) 11 (3–23)
All
18 (12–26)
20 (13–28)
10 (5–17)
15 (8–24)
All subtypes Men Women [15;55[ [55;65[ [65;75[ [75;++[ All ages 1989–1991 1992–1994 1995–1997
88 (85–90) 93 (92–94) 99 (99–99) 93 (91–95) 83 (79–87) 53 (47–59) 92 (91–93) 88 (85–90) 92 (89–93) 95 (93–96)
91 (88–93) 96 (95–97) 99 (99–100) 94 (92–96) 85 (81–88) 58 (51–65) 95 (94–96) 90 (88–92) 94 (92–96) 97 (96–98)
78 (75–82) 89 (87–90) 97 (96–98) 89 (86–92) 69 (64–73) 33 (27–39) 87 (85–88) 81 (77–83) 87 (84–89) 91 (88–92)
88 (85–91) 95 (94–96) 98 (97–99) 94 (90–96) 77 (71–82) 51 (22–74) 94 (92–95) 89 (86–91) 94 (92–96) 96 (94–98)
All cases
92 (91–93)
94 (93–95)
87 (85–88)
93 (92–94)
Papillary Men Women [15;55[ [55;65[ [65;75[ 75;++[
Overall survival: observed survival; relative survival: survival that would be observed in the absence of non-cancer-related deaths; CI: confidence interval.
in women. Let us note however that the estimates relative to the latter two subtypes were not very accurate. Because most of these differences are not very marked, they are seldom mentioned in the scientific literature. Akslen et al. [15] did not find sex-specific differences concerning survival. Cunningham et al. [16] and Colonna et al. [17] have already mentioned a higher survival in women as regards papillary and follicular carcino-
mas. Guilliland et al. [18] observed similar survivals in men and women as to papillary and follicular carcinomas, poorer survivals in women than in men as to anaplastic carcinomas, and higher survivals in women than in men as to medullary carcinomas. Finally, contrarily to our observations on medullary carcinomas, Esik et al. [19] observed higher survivals in women than in men.
M. Colonna et al. / Annales d’Endocrinologie 71 (2010) 95–101
Age-related differences were much greater than sex-related ones. Moreover, the literature has confirmed that age is an important prognostic factor [15–19]; indeed, it was the only prognostic factor we kept in our analyses. However, another very influential prognostic factor, the cancer stage at the time of diagnosis, should have been considered as well but that information was not available for all analyzed cases. The effect of cancer stage at diagnosis is certainly associated with the very high survivals of patients with papillary carcinoma and the survival increase over the period 1987 to 1994 [5], during which small tumors were predominant [10]. 6. Conclusion The ever increasing frequency of thyroid cancer is becoming a real public health problem in women, even though the specific survival is extremely high. Thus, as some authors [20] seem to note a slow-down of that frequency in the French Rhône-Alpes Region (eight départements) whereas others report completely different results elsewhere, such as in the United States [21], it remains of utmost importance to closely monitor the disease trends and make better estimates of required resources. 7. Conflicts of interest The authors have not declared any conflict of interest. 8. French version A french version of this doi:10.1016/j.ando.2009.11.003.
article
is
available
at
Acknowledgements This study received funding from the Ligue Nationale Contre le Cancer, (National Cancer Research Fund) and the Comité de l’Isère (local branch of the National Cancer Research Fund). The authors thank Jean Iwaz (hospices civils de Lyon) for a thorough revision of the manuscript. References [1] Belot A, Grosclaude P, Bossard N, et al. Cancer incidence and mortality in France over the period 1980–2005. Rev Epidemiol Sante Publ 2008;56:159–75. [2] Hospices Civils de Lyon – Institut de veille sanitaire – Institut National du Cancer – Francim – Institut National de la Santé et de la Recherche médicale. Projections de l’incidence et de la mortalité par cancer en France pour l’année 2008. Rapport technique, mars 2009. (http://www.invs.sante.fr/surveillance/cancers).
101
[3] Remontet L, Bossard N, Belot A, Estève J, the French network of cancer registries (FRANCIM). An overall strategy based on regression models to estimate relative survival and model the effects of prognostic factors in cancer survival studies. Stat Med 2007;26:2214–28. [4] Bossard N, Velten M, Remontet L, et al. Survival of cancer patients in France: A population-based study from the Association of the French Cancer Registries (FRANCIM). Eur J Cancer 2007;43:149–60. [5] Réseau FRANCIM. Survie des patients atteints de cancer en France. Étude des registres de cancers du réseau FRANCIM. Paris: Springer-Verlag; 2007. [6] Colonna M, Grosclaude P, Remontet L, et al. Incidence of thyroid cancer in adults recorded by French cancer registries (1978–1997). Eur J Cancer 2002;38:1762–8. [7] Chérié-Challine L, Les membres du comité de rédaction. Surveillance sanitaire en lien avec l’accident de Tchernobyl en France – bilan actualisé sur les cancers thyroïdiens et études épidémiologiques en cours en 2006. Saint-Maurice: Institut de veille sanitaire; 2006. www.invs.sante.fr. [8] Verkooijen HM, Fioretta G, Pache JC, et al. Diagnostic changes as a reason for the increase in papillary thyroid cancer incidence in Geneva, Switzerland. Cancer Causes Control 2003;14:13–7. [9] Leenhardt L, Bernier MO, Boin-Pineau MH, et al. Advances in diagnosis practices affect thyroid cancer incidence in France. Eur J Endocrinol 2004;150:133–9. [10] Colonna M, Guizard AV, Schartz C, et al. Time trend analysis of papillary and follicular cancers as a function of tumour size: A study of data from six cancer registries in France (1983–2000). Eur J Cancer 2007;43:891–900. [11] Hedinger C, Williams ED, Sobin LH. The WHO histological classification of thyroid tumors: a commentary on the second edition. Cancer 1989;63:908–11. [12] Cardis E, Krewski D, Boniol M, et al. Estimates of the cancer burden in Europe from radioactive fallout from the Chernobyl accident. Int J Cancer 2006;119:1224–35. [13] Leux C, Guénel P. Facteurs de risque des cancers et nodules thyroïdiens : effet des polluants de l’environnement et risques professionnels. Med Clin Endocrinol Diabete 2009:1–7. [14] Curado P, Edwards B, Shin H, Storm H, Ferlay J, Heanue M, Boyle P. Cancer incidence in five continents, Vol. IX. IARC Scientific Publications 160, Lyon, International Agency for Research on Cancer, 2007. [15] Akslen L, Haldorsen T, Thoresen SO, Glattre E. Survival and causes of death in thyroid cancer: A population-based Study of 2479 Cases from Norway. Cancer Res 1991;51:1234–41. [16] Cunningham MP, Duda RB, Recant W, Cmiel JS, Sylvester JA, Fremgen A. Survival discriminants for differentiated thyroid cancer. Am J Surg 1990;160:344–7. [17] Colonna M, Grande E, Jonasson JG, the EUROCARE Working Group. Cancer Variation in relative survival of thyroid cancers in Europe: Results from the analysis on 21 countries over 4 the period 1983–1994 (EUROCARE-3 study). Eur J Cancer 2006;42:2598–608. [18] Guilliland FD, Hunt WC, Morris DM, Key XR. Prognostic factors for thyroid carcinoma. Cancer 1997;19:564–73. [19] Esik O, Tusnády G, Trón L, et al. Markov model-based estimation of individual survival probability for medullary thyroid cancer patients. Pathol Oncol Res 2002;8:93–104. [20] Sassolas G, Hafdi-Nejjari1 Z, Remontet L, et al. Thyroid cancer: is the incidence rise abating? Eur J Endocrinol 2008;159:1–10. [21] Ries L, Melbert D, Krapcho M, Stinchcomb DG et al. SEER Cancer Statistics Review, 1975–2005, National Cancer Institute. Bethesda, 2009. (http://seer.cancer.gov/csr/1975 2005/).