J Neurooncol (2014) 120:547–555 DOI 10.1007/s11060-014-1585-0
CLINICAL STUDY
Significant heterogeneity in the geographical distribution of diffuse grade II/III gliomas in France Ame´lie Darlix • Sonia Zouaoui • Jean-Marc Virion • Vale´rie Rigau • He´le`ne Mathieu-Daude´ • Marie Blonski • German Reyes-Botero • Faiza Bessaoud • Brigitte Tre´tarre • Fabienne Bauchet • Laurent Capelle Michel Fabbro • Christine Kerr • Dominique Figarella-Branger • Hugues Duffau • Luc Taillandier • Luc Bauchet
•
Received: 14 March 2014 / Accepted: 9 August 2014 / Published online: 24 August 2014 Ó Springer Science+Business Media New York 2014
Abstract Diffuse WHO grade II and III gliomas (DGII/ IIIG) are rare tumors, with few specific epidemiological studies. We aimed at describing the geographical distribution of a homogeneous series of histologically confirmed DGII/IIIG, over a four-year period (2006–2009), at a national level. The methodology is based on a multidisciplinary national network already established by the French Brain Tumor DataBase and data collected directly from every neuropathology department. Personal home addresses were collected for confirmed cases. For each region, the incidence of DGII/IIIG was analyzed and standardized on the age and sex distribution of the French population. The number of patients with newly diagnosed, histologically confirmed DGII/IIIG was 4,790. The overall crude rate was
19.4/106. To enable international comparisons, standardized rates were calculated as follows: 19.8/106, 18.8/106 and 16.0/ 106 (reference population, Europe, US and world, respectively). The geographical distribution by region showed significant differences, with higher incidence rates in Northeast and central parts of France. This work is the first studying the geographical distribution of a pure series of DGII/IIIG at a national level. It demonstrates significant heterogeneity in the distribution, and raises the question of the role of environmental and/or genetic risk(s) factor(s) for DGII/IIIG.
Ame´lie Darlix and Sonia Zouaoui equally contributed to the study.
Introduction
Electronic supplementary material The online version of this article (doi:10.1007/s11060-014-1585-0) contains supplementary material, which is available to authorized users. A. Darlix (&) M. Fabbro C. Kerr Departement of Medical Oncology and Department of Radiation Oncology, Institut Re´gional du Cancer de Montpellier - Val d’Aurelle, Montpellier, France e-mail: [email protected] A. Darlix S. Zouaoui H. Duffau L. Bauchet INSERM U1051, Montpellier, France S. Zouaoui H. Duffau L. Bauchet Department of Neurosurgery, Gui de Chauliac Hospital, Montpellier, France S. Zouaoui H. Mathieu-Daude´ F. Bessaoud B. Tre´tarre F. Bauchet Department of Epidemiology, French Brain Tumor Database, GNOLR, Registre des Tumeurs de l’He´rault, Institut Re´gional du Cancer de Montpellier - Val d’Aurelle, Montpellier, France
Keywords Brain tumor Neurooncology Glioma Epidemiology
Diffuse WHO grade II and III gliomas (DGII/IIIG) include diffuse astrocytomas, anaplastic astrocytomas, J.-M. Virion Department of Clinical Epidemiology and Evaluation, University Hospital, Nancy, France V. Rigau Department of Pathology, Gui de Chauliac Hospital, Montpellier, France M. Blonski Neuro-oncology Unit, Central Hospital, Nancy, France G. Reyes-Botero Department of Neurology, Pitie´-Salpe´trie`re Hospital, Paris, France L. Capelle Department of Neurosurgery, Pitie´-Salpe´trie`re Hospital, Paris, France
123
548
oligodendrogliomas, anaplastic oligodendrogliomas, and mixed gliomas [1]. They differ from the other gliomas (in particular the glioblastoma) in terms of clinical and radiological presentation, and natural history. They are rare, with an incidence B 20/106 person-years in the United States or in Europe [2–5]. Except for a few cases, their causes remain unknown. Environmental factors could be involved but to date the only one known to increase the risk of glioma is exposure to high-dose ionizing radiations [6– 9]. The epidemiology of DGII/IIIG is poorly known. Some studies have suggested heterogeneity in the geographical distribution of gliomas, at a worldwide or European level, with a trend towards higher incidence rates in highly developed countries (http://globocan.iarc.fr/) [2, 10–12]. However these studies have two major limits. First, they include all grades of gliomas or even different histological subtypes of primary central nervous system tumors (PCNSTs). Secondly, they compare incidence rates at an international or intercontinental level, resulting in an obvious bias due to differences among continents and European countries in terms of case registrations and access to healthcare. Here we aimed at studying the geographical distribution of a series of newly diagnosed and histologically confirmed DGII/IIIG on metropolitan France, over a four-year period.
Materials and methods Settings The present study was conducted in metropolitan France (including Corsica) representing 552,000 km2 and divided into 22 regions. The 2007 population in metropolitan France was 61,795,238. Population density and age distribution vary from one region to another (http://www.insee. fr/fr/themes/detail.asp?reg_id=99&ref_id=estim-pop). Data sources: French brain tumor database (FBTDB) Cases were identified by the FBTDB, which is the largest histological database for PCNSTs in Europe [13–17]. This multidisciplinary national network was established thanks to the collaboration of all scientific societies involved in
D. Figarella-Branger Department of Neuropathology and INSERM U911, Timone Hospital, Universite´ de la Me´diterrane´e, Marseille, France L. Taillandier Neuro-oncology Unit, University Hospital, Poitiers, France
123
J Neurooncol (2014) 120:547–555
the care of patients with PCNST in France (Association des Neuro-Oncologues d’Expression Franc¸aise, Socie´te´ Franc¸aise de Neurochirurgie, Socie´te´ Franc¸aise de Neuropathologie). It aims at prospectively registering all histologically-proven cases of PCNST in France. Its methodology was previously published [13, 16, 17]. Two different sources were used to collect data. The first source is a data sheet available in all operating rooms conducting PCNSTs surgery (n = 52), for prospective registration of cases. Secondly, to improve the exhaustiveness of data collection, an annual listing of all cases analyzed by each neuropathology department was collected to complete the FBTDB [17]. In some specific cases when data from the data sheet and listings did not match, or when the data sheet was not available for a case identified by the listing from one neuropathology department, a clinical research technician contacted directly the neuropathology department in order to verify and collect the missing data. All patients diagnosed with a histologically proven DGII/IIIG were included, i.e. patients with a WHO grade II or III astrocytoma (fibrillary astrocytoma, gemistocytic astrocytoma, protoplasmic astrocytoma or anaplastic astrocytoma, corresponding to ICD-O code 9420/3, 9411/3, 9410/3 and 9401/3, respectively), WHO grade II or III oligodendroglioma (ICD-O code 9450/3 and 9451/3, respectively) or WHO grade II or III mixed glioma (ICDO code 9382/3). Patients diagnosed with a ‘‘not otherwise specified’’ (NOS) astrocytoma (ICD-O code 9400/3, i.e. astrocytoma for which the WHO grading was unavailable, with the exclusion of glioblastoma) were included. By definition these tumors are classified as a DGIIG in the WHO classification and in the CBTRUS (Central Brain Tumor Registry of the United States) classification [1, 2]. Patients with a NOS mixed glioma (ICD-O code 9382/3) were also included. Since the French nomenclature (ADICAP 2003, http://www.adicap.asso.fr/THESAURUS/ Adicap_v5-03pdf) differentiates oligoastrocytoma from anaplastic oligoastrocytoma, the number of NOS mixed gliomas was expected to be low in our study and thus not to impact the repartition between GIIG and GIIIG. We chose of including NOS gliomas (ICD-O code 9380/3), i.e. diffuse gliomas for which both histological subtype and WHO grading were unavailable, since some among them are some DGII/IIIG, and their exclusion could result in an underestimation of the studied cases (see discussion). For this study, when the information collected from either the data sheet or listings from pathologists was NOS glioma, NOS mixed glioma or NOS astrocytoma, the pathological report was collected and further analysed by an expert pathologist (VR) in order to try and reach a diagnosis according to the WHO 2007 classification.
J Neurooncol (2014) 120:547–555
Diffuse WHO grade II astrocytomas for which the subtype was unknown (fibrillary, gemistocytic or protoplasmic) were called ‘‘WHO grade II astrocytomas’’. Diffuse WHO grade II and grade III gliomas for which the histological subtype was unknown (astrocytoma, oligodendroglioma or oligoastrocytoma) were called ‘‘NOS WHO grade II gliomas’’ and ‘‘NOS WHO grade III gliomas’’, respectively. Since differentiating between oligodendrogliomas and astrocytomas can, sometimes, be difficult, we chose to group all GIIG subgroups and all GIIIG subgroups in order to analyse their respective incidences [18–20]. Patients were included if the histological diagnosis had been made between January 1st, 2006 and December 31st, 2009, in one of the neuro-oncological centers in metropolitan France. Patients who had previous glioma surgery were excluded if the first diagnosis of glioma was made before January 1st, 2006. If the diagnosis was established during the inclusion period, data from the first surgery was collected and included in the study.
Geographical distribution The home address (city, zip code) of each patient at the time of the surgery was collected and matched to the corresponding five digits official geographical code (http:// insee.fr/fr/methodes/nomenclatures/cog/) as defined by the National Institute of Statistics and Economic Studies (INSEE). Patients for whom the home address at the time of surgery was unknown were not included in the analysis, neither were patients who lived abroad or in one of the overseas French departments or territories.
Statistical analysis Statistical analysis was performed with the SAS software (v9.3). Standardized incidences (with 95 % confidence intervals) were calculated for each region. Incidences were calculated and adjusted for age and sex based on the distribution of the French population (reference population: France INSEE 2007 http://www.insee.fr/fr/ themes/document.asp?ref_id=ip1170#inter1). To enable international comparisons, standardized rates for all DGII/IIIG were also calculated with Europe, USA and worldwide populations (2007) as references. Incidence ratios were defined as the standardized incidence for the area of interest divided by the national incidence. Differences in incidence between the area of interest and national incidence were tested using a Chi squared test. Maps of standardized incidences were drawn using the Geoclip website (http://www.geoclip.fr/fr/) approved by the INSEE.
549
Results Population characteristics Over the four-year period, 4,915 patients had surgery for newly diagnosed DGII/IIIG in metropolitan France. Among them 125 (2.5 %) were excluded from the study: 91 patients (1.8 %) lived abroad or in an overseas French department or territory, and the home address was unknown for 34 patients (0.7 %). Therefore, 4,790 patients were included (1,220, 1,190, 1,228, and 1,152 for the years 2006, 2007, 2008 and 2009, respectively). The distribution of patients according to the histological subtype is summarized in Table 1. Among the 4,790 patients included, 2,099 (43.82 %) were diagnosed with a WHO grade II glioma and 2,484 (51.86 %) with a WHO grade III glioma. Thirty-eight (0.79 %), 25 (0.52 %) and 144 (3.01 %) patients were diagnosed with a NOS astrocytoma, a NOS mixed glioma, or a NOS glioma, respectively. Median and mean ages at diagnosis were 51 (range 0–88) and 49.8 ± 17.6 (44.8 ± 16.6 for WHO grade II gliomas vs. 54.0 ± 16.9 for WHO grade III gliomas; p \ 0.001, F-test), respectively. Table 2 shows the distribution of patients according to their age at diagnosis. The [30–39] and [40–49] age classes were the most represented for WHO grade II gliomas while the [50–59] and [60–69] were the most represented for WHO grade III gliomas. The distribution of histological subtypes of tumors significantly varied (p \ 0.001, Chi squared test) according to the age class. Diffuse WHO grade II gliomas were the most represented among people aged 49 or less, while diffuse WHO grade III were the most represented among people over 50 (p \ 0.001). In our study 56.9 % of patients were male. There was no difference in terms of sex ratio between WHO grade II and WHO grade III gliomas (p = 0.8, Chi squared test). Mean age was 50.0 ± 16.9 for men and 49.5 ± 18.3 for women. Median age was 52 (range 0–86) for men and 51 (range 0–88) for women. The distribution of histological subtypes of tumors according to gender showed no significant difference between men and women. Standardized incidence rates The standardized average annual age and sex adjusted incidence rate for all DGII/IIIG (n = 4,790) was 19.4/106 (95 % CI 18.8–19.9). To enable international comparisons, standardized rates were calculated as follows: 19.8/106 (population of reference: Europe), 18.8/106 (reference: USA), and 16.0/106 (reference: world) (Table 3). When NOS gliomas were excluded (standardized incidence of NOS gliomas 0.6/106; 95 % CI 0.5–0.7), standardized rates
123
550
J Neurooncol (2014) 120:547–555
Table 1 Cases distribution among men and women according to histologic subtypes with the corresponding median age at diagnosis Histology
N
Males (n)
Females (n)
Median age (in years)
Diffuse WHO grade II astrocytoma (fibrillary, gemistocytic or protoplasmic astrocytoma, ‘‘WHO grade II astrocytoma’’)
286
165
121
47
Anaplastic astrocytoma
59
417
234
183
NOSa astrocytoma
38
21
17
Oligodendroglioma
1,354
763
591
44
Anaplastic oligodendroglioma
1,196
678
518
56
Oligoastrocytoma
412
231
181
42
Anaplastic oligoastrocytoma NOSa mixed glioma
827 25
485 13
342 12
56 44
NOSa WHO grade II glioma
47
34
13
51
NOSa WHO grade III glioma
44
24
20
51
144
77
67
57
4,790
2,725
2,065
51
NOSa glioma Total a
48.5
NOS not otherwise specified
Table 2 Distribution of patients according to their age at diagnosis Age class (years)
All DGII/IIIG (n = 4,790), n (%)
WHO grade II gliomas (n = 2,099), n (%)
WHO grade III gliomas (n = 2,484), n (%)
[0–19]
244 (5.1)
119 (5.7)
95 (3.8)
[20–29]
384 (8.0)
253 (12.0)
122 (4.9)
[30–39]
724 (15.1)
448 (21.3)
258 (10.4)
[40–49]
866 (18.1)
448 (21.3)
390 (15.7)
[50–59]
1,043 (21.8)
412 (19.6)
592 (23.8)
[60–69]
870 (18.2)
262 (12.5)
565 (22.7)
C70
659 (13.8)
157 (7.5)
462 (18.6)
were calculated as follows: 19.2/106 (reference: Europe), 18.2/106 (reference: USA), and 15.5/106 (reference: world). The standardized incidence rates for all diffuse WHO grade II gliomas (n = 2,099) and WHO diffuse grade III gliomas (n = 2,484) were 8.5/106 (95 % CI 8.1–8.9) and 10.0/106 (95 % CI 9.7–10.4), respectively. The standardized incidence rates for grade II and III astrocytomas, grade II and III oligodendrogliomas and grade II and III oligoastrocytomas were 3.0/106 (95 % CI 2.8–3.2), 10.3/106 (95 % CI 9.9–10.7) and 5.0/106 (95 % CI 4.7–5.3), respectively.
see Fig. 2). It was significantly lower than the national incidence in Iˆle-de-France, Basse-Normandie, Rhoˆne-Alpes and Provence-Alpes-Coˆte d’Azur (incidence rates 13.9/ 106 to 17.8/106). The incidence rates of each of the 22 regions are summarized in Table 4 (supplementary materials). We also found a significant heterogeneity in the geographical distribution of WHO grade II and WHO grade III gliomas (Fig. 3 and Fig. 4, supplementary materials). The standardized incidence rates for each of the 22 regions for WHO grade II and WHO grade III gliomas are summarized in Tables 5 and 6, respectively (supplementary materials).
Geographical distribution We found a significant heterogeneity in the geographical distribution by regions of all DGII/IIIG in metropolitan France (Fig. 1). The standardized incidence rate (reference: France) was significantly higher than the national incidence (i.e. the incidence ratio was significantly different from 1) in Champagne-Ardenne, Nord-Pas-de-Calais, Alsace, Bretagne and Auvergne (incidence rates 21.9/106 to 30.9/106,
123
Discussion This study is the first to analyze incidence rates and geographical distribution at a national level for DGII/IIIG. To our knowledge, this is the largest series that exhaustively includes homogeneous subtypes of histologically proven glioma, with 4,790 confirmed cases of DGII/IIIG.
J Neurooncol (2014) 120:547–555
551
Table 3 Standardized rates and international comparisons (Age-sex adjusted incidence rates, per 106 with 95 % confidence intervals, to 2007 Europe, USA and World standard population) Reference population France
Europe
USA
All DGII/IIIGa
19.4 [18.8–19.9]
19.8 [19.2–20.4]
18.8 [18.2–19.3]
16 [15.5–16.4]
All DGII/IIIGa excluding NOS gliomas
18.8 [18.3–19.3]
19.2 [18.6–19.8]
18.2 [17.7–18.7]
15.5 [15.0–16.0]
All DGIIGb All DGIIIGc
8.5 [8.1–8.9] 10.0 [9.7–10.4]
8.6 [8.3–9.0] 10.3 [9.9–10.7]
8.3 [8.0–8.7] 9.6 [9.3–10.0]
7.6 [7.3–8.0] 7.6 [7.3–8.0]
a
World
DGII/IIIG diffuse grade II or III glioma
b
DGIIG diffuse grade II glioma
c
DGIIG diffuse grade III glioma
Incidence data In our series the standardized incidence rate adjusted on the French population for all DGII/IIIG was 19.4/106 (19.8/ 106, 18.8/106 and 16.0/106 when standardized on European, USA and worldwide populations, respectively). These results are consistent with those from larger series of PCNSTs, i.e. data from the CBTRUS [2] (standardized incidence rate: 19.8/106, and higher proportion of NOS gliomas). For diffuse WHO grade II gliomas our results match those in recently published studies, with a standardized incidence rate of 8.5/106, lower than the one reported by the Austrian Brain Tumor Registry (incidence rate adjusted to the USA population: 12.2/106) but higher than in recent population-based studies in Europe [4, 10, 21]. In France, the only data available are based on a departmental registry and thus on a much smaller population (incidence rate for DGII/IIIG 13.7/106, excluding NOS gliomas, 2000-2007) [5]. Data from England (incidence rate for DGII/IIIG 21.2/106, 1999–2003, worldwide population) included about 30 % of NOS gliomas, resulting in an important selection bias [22]. Heterogeneity in geographical distribution of DGII/IIIG We showed significant differences in the metropolitan France geographical distribution for all DGII/IIIG, diffuse WHO grade II and diffuse WHO grade III gliomas, with higher incidence rates in the Northeast and central France. This uneven geographical distribution may suggest a role of environmental and/or genetic factors. However we cannot exclude some methodological bias. First, we cannot exclude a selection bias due to differences across regions in terms of access to healthcare. However, as metropolitan France has a unique healthcare system, access to healthcare is likely to be similar for all regions. The distribution of neurosurgical centers is roughly homogeneous on the metropolitan French territory,
with one or more neurosurgical center(s) (depending on the number of inhabitants) for each region (and no region without a neurosurgical center). Moreover, data from the national medical board (http://www.conseil-national.mede cin.fr/sites/default/files/Atlas__national_2013.pdf) shows that there is no major difference among regions as regards the distribution of neurosurgeons. Important differences in population sizes among the compared regions might also be a source of bias, by making the distinction between random and true variations in incidence rates difficult. Another concern is that we are unaware of how long patients have lived in the same city at the time of the surgical procedure. If a patient has moved from one region to another not long before tumor diagnosis, the home address collected might not appropriately reflect environmental exposure. A large number of such patients could also influence the regional incidence rates. However, data from the INSEE on the 2008 French population shows that the number of home residence changes between regions is low (http://www.insee.fr/fr/bases-de-donnees/default.asp? page=recensement/resultats/doc/presentation-fichiers-detailmigrations-residentielles.htm). We chose to include NOS gliomas, i.e. gliomas for which both histological subtype and WHO grading were unavailable due to neuropathological difficulties. Only 144 cases (3.1 %) of NOS gliomas could not be reclassified after analysis of the pathological report. We cannot exclude that some of these tumors were actually WHO grade IV gliomas, particularly because the median age (57) is closer to that of high-grade than low-grade glioma patients. Yet, due to their small number and the fact that for these 144 patients, home addresses at diagnosis were distributed all over metropolitan France, we believe that their inclusion was unlikely to bias our results. As the neuropathological diagnosis of the grade can be difficult in gliomas, we cannot exclude a bias due to inconsistency in tumor grading between regions. To this end, a centralized pathological review would be of great
123
552
J Neurooncol (2014) 120:547–555
Fig. 1 Standardized incidences for each region for all DGII/IIIG (n = 4,790)
value but is currently not possible to get at the national level in France, or, to our knowledge, in any other country. However, with regards to the differentiation between grade III and grade IV gliomas, it has been established by the Brain Tumor Epidemiology Consortium that glioblastomas have a good general agreement across regions [20]. However it would be interesting to find out if the geographical distribution of glioblastomas has the same pattern than that of DGII/IIIG. As the differential diagnosis between diffuse grade II and grade III gliomas can be challenging, we chose
123
to group grade II and II gliomas together to eliminate any possible bias due to misclassification between grade II and grade III gliomas. Our study only included tumors for which a histological confirmation was available, since the FBTDB does not record non-histologically confirmed cases. As there is no national PCNSTs registry in France, these non-histological cases are very difficult to record. We believe that their noninclusion does not bias our results in terms of geographical distribution for two main reasons. First, for the last
J Neurooncol (2014) 120:547–555
553
Fig. 2 Standardized incidence ratios for each region for all DGII/ IIIG (n = 4,790). Regions in red correspond to areas with a significantly higher standardized incidence compared to the mean
incidence in France, whereas regions in green correspond to areas with a significantly lower standardized incidence
10 years in France clinicians have promoted histological confirmation of tumors in patients \70. Secondly, it has been shown, based on a French departmental registry, that astrocytomas and oligodendrogliomas are histologically validated in more than 98 % and 93 % of cases, respectively (http://etudes.isped.u-bordeaux2.fr/REGISTRES-CAN CERS-AQUITAINE/Snc/document/Rtsnc_lettre_information_ 200905.pdf). We were not able to provide molecular data, as this piece of information is not currently collected by the FBTDB. It would be of great value and we will aim at
adding molecular markers to the data collection in a near future.
Heterogeneity in geographical distribution of DGII/IIIG: hypothesis The uneven geographical distribution of DGII/IIIG may suggest a role of environmental and/or genetic factors. Interestingly, a geographical heterogeneity in incidence rates has been highlighted for other cancers. In France,
123
554
incidence rates for all cancers were higher in the North and Northeast regions (data from the National Cancer Institute: http://www.e-cancer.fr/publications/69-epidemio logie/574-la-situation-du-cancer-en-france-en-2011). Especially, distribution of colorectal cancer (CRC) was comparable to that of DGII/IIIG, with similar higher incidence rates in the Northeast part of France. We can thus hypothesize that these two diseases may share some risk factors. Diet, has been validated as a risk factor in CRC with red and processed meat, alcohol, obesity or abdominal fat being strongly associated with a higher risk of CRC, while the ‘‘Mediterranean diet’’ prevents its occurrence [23, 24]. In France, the consumption of red and processed meat is higher in Northeast regions (CREDOC, research center for the study and the observation of living conditions, http://www.credoc.fr/pdf/4p/ 101.pdf). Moreover the prevalence of obesity varies across France in a similar way (http://www.insee.fr/fr/ffc/ ipweb/ip1123/ip1123.pdf). A working hypothesis could thus be that diet factors, and in particular meat-based diet and obesity, may have an impact on DGII/IIIG risk. Dietary sources of nitroso compounds (such as cured or processed meats) have been associated with an increased glioma risk in several population-based studies [25–27]. Some other studies however found no conclusive evidence for an association between gliomas (or PCNSTs) and processed meat [28–30]. So far, no epidemiological study has investigated the association between processed meat and DGII/IIIG risk specifically. Multiple sclerosis (MS) is another neurological disease with a higher incidence in Northeast France, comparable to what was found for DGII/IIIG [31–33]. As DGII/IIIG have the same pattern of geographical distribution than MS, these two diseases may share some risk factors. There is strong evidence that environmental factors influence the uneven distribution of MS. In particular, the amount of sunlight (and, consequently, outdoor activities) is significantly associated with MS prevalence (sunny areas being associated with low prevalence). This association probably involves vitamin D, as a low prevalence of MS has been reported in countries with a poor amount of sunlight but a high consumption of vitamin D (oily fish) [34, 35]. We could thus hypothesize that sunlight and vitamin D intake influence the risk of DGII/IIIG. Based on the map of the sunlight provided by the Joint Research center of the European Commission (available at: http://re.jrc.ec.europa.eu/pvgis/cmaps/eu_cmsaf_opt/G_ opt_FR.png), it seems that regions with low sunlight correlate with those with higher incidence rates of DGII/IIIG in our study, supporting the hypothesis of an association between sunlight and DGII/IIIG risk. This hypothesis is also supported by a study that found a reverse association between UVB irradiance and brain tumor risk [36].
123
J Neurooncol (2014) 120:547–555
Conclusion This work is, to our knowledge, the first to analyse the geographical distribution over a four-year period of a pure series of DGII/IIIG at a national level, thus studying a homogeneous population in terms of access to healthcare. This study demonstrates a significant heterogeneity in the geographical distribution of newly diagnosed and histologically confirmed DGII/IIIG in metropolitan France and raises the question of environmental and/or genetic risk factors for DGII/IIIG. The next step will be to compare environmental and genetic factors between regions with low and those with high incidence. Acknowledgments The authors wish to thank all pathologists, neurosurgeons and other stakeholders, for their help and input with this work (see the list in supplementary materials), and Be´ne´dicte Cle´ment, scientific translator, for valuable translation work. This work was conducted with the financial support in the form of grants from the Ligue Nationale Contre le Cancer, Association des NeuroOncologues d’Expression Franc¸aise, Socie´te´ Franc¸aise de Neurochirurgie, Associations pour la Recherche sur les Tumeurs Ce´re´brales (ARTC and ARTC Sud), Roche Laboratory, Sophysa Laboratory, De´partement de l’He´rault, and Groupe de Neuro-Oncologie du Languedoc Roussillon. Conflict of interest
None declared.
References 1. Louis DN, Ohgaki H, Wiestler OD et al (2007) The 2007 WHO classification of tumours of the central nervous system. Acta Neuropathol 114:97–109 2. Dolecek TA, Propp JM, Stroup NE, Kruchko C (2012) CBTRUS statistical report: primary brain and central nervous system tumors diagnosed in the United States in 2005-2009. Neuro Oncol 14(suppl 5):v1–v49 3. Ohgaki H, Kleihues P (2005) Epidemiology and etiology of gliomas. Acta Neuropathol 109:93–108 4. Wo¨hrer A, Waldho¨r T, Heinzl H et al (2009) The Austrian Brain Tumour Registry: a cooperative way to establish a populationbased brain tumour registry. J Neurooncol 95:401–411 5. Baldi I, Gruber A, Alioum A et al (2011) Descriptive epidemiology of CNS tumors in France: results from the Gironde Registry for the period 2000-2007. Neuro Oncol 13:1370–1378 6. Ron E, Modan B, Boice JD, Alfandary E, Stovall M, Katz L (1988) Tumors of the brain and nervous system after radiotherapy in childhood. N Engl J Med 319:1033–1039 7. Yonehara S, Brenner AV, Kishikawa M et al (2004) Clinical and epidemiologic characteristics of first primary tumors of the central nervous system and related organs among atomic bomb survivors in Hiroshima and Nagasaki, 1958-1995. Cancer 101:1644–1654 8. Bondy ML, Scheurer ME, Malmer B et al (2008) Brain tumor epidemiology: consensus from the Brain Tumor Epidemiology Consortium. Cancer 113(7 Suppl):1953–1968 9. McCarthy BJ, Rankin KM, Aldape K et al (2011) Risk factors for oligodendroglial tumors: a pooled international study. Neuro Oncol 13:242–250
J Neurooncol (2014) 120:547–555 10. Crocetti E, Trama A, Stiller C et al (2012) Epidemiology of glial and non-glial brain tumours in Europe. Eur J Cancer 48:1532–1542 11. Houben MP, Coebergh JW, Birch JM, Tijssen CC, van Duijn CM, McNally RJ (2005) Space–time clustering patterns of gliomas in the Netherlands suggest an infectious aetiology. Eur J Cancer 41:2917–2923 12. Houben MP, Coebergh JW, Birch JM, Tijssen CC, van Duijn CM, McNally RJ (2006) Space-time clustering of glioma cannot be attributed to specific histological subgroups. Eur J Epidemiol 21:197–201 13. Bauchet L, Rigau V, Mathieu-Daude´ H et al (2007) French brain tumor data bank: methodology and first results on 10,000 cases. J Neurooncol 84:189–199 14. Bauchet L, Rigau V, Mathieu-Daude´ H et al (2009) Clinical epidemiology for childhood primary central nervous system tumors. J Neurooncol 92:87–98 15. Bauchet L, Mathieu-Daude´ H, Fabbro-Peray P et al (2010) Oncological patterns of care and outcome for 952 patients with newly diagnosed glioblastoma in 2004. Neuro Oncol 12:725–735 16. Rigau V, Zouaoui S, Mathieu-Daude´ H et al (2012) French brain tumor database: 5- year histological results on 25756 cases. Brain Pathol 21:633–644 17. Zouaoui S, Rigau V, Mathieu-Daude´ H et al (2012) French brain tumor database: general results on 40,000 cases, main current applications and future prospects. Neurochirurgie 58:4–13 18. Hartmann C, von Deimling A (2009) Molecular pathology of oligodendroglial tumors. Recent Results Cancer Res 171:25–49 19. van den Bent MJ (2010) Interobserver variation of the histopathological diagnosis in clinical trials on glioma: a clinician’s perspective. Acta Neuropathol 120(3):297–304 20. Davis FG, Malmer BS, Aldape K et al (2008) Issues of diagnostic review in brain tumor studies: from the Brain Tumor Epidemiology Consortium. Cancer Epidemiol Biomarkers Prev 17(3): 484–489 21. Okamoto Y, Di Patre P-L, Burkhard C, Horstmann S, Jourde B, Fahey M et al (2004) Population-based study on incidence, survival rates, and genetic alterations of low-grade diffuse astrocytomas and oligodendrogliomas. Acta Neuropathol 108:49–56 22. Arora RS, Alston RD, Eden TO et al (2010) Are reported increases in incidence of primary CNS tumours real ? An analysis of longitudinal trends in England, 1979-2003. Eur J Cancer 46:1607–1616
555 23. Chan DS, Lau R, Aune D et al (2011) Red and processed meat and colorectal cancer incidence: meta-analysis of prospective studies. PLoS One 6(6):e20456 24. Aleksandrova K, Nimptsch K, Pischon T (2013) Influence of obesity and related metabolic alterations on colorectal cancer risk. Curr Nutr Rep 2(1):1–9 25. Boeing H, Schlehofer B, Blettner M, Wahrendorf J (1993) Dietary carcinogens and the risk for glioma and meningioma in Germany. Int J Cancer 53(4):561–565 26. Blowers L, Preston-Martin S, Mack WJ (1997) Dietary and other lifestyle factors of women with brain gliomas in Los Angeles County (California, USA). Cancer Causes Control 8:5–12 27. Kyritsis AP, Bondy ML, Levin VA (2011) Modulation of glioma risk and progression by dietary nutrients and antiinflammatory agents. Nutr Cancer 63(2):174–184 28. Huncharek M, Kupelnick B, Wheeler L (2003) Dietary cured meat and the risk of adult glioma: a meta-analysis of nine observational studies. J Environ Pathol Toxicol Oncol 22(2):129–137 29. Michaud DS, Holick CN, Batchelor TT, Giovannucci E, Hunter DJ (2009) Prospective study of meat intake and dietary nitrates, nitrites, and nitrosamines and risk of adult glioma. Am J Clin Nutr 90:570–577 30. Dubrow R, Daresky AS, Park Y et al (2010) Dietary components related to N-Nitroso compound formation: a prospective study of adult glioma. Cancer Epidemiol Biomarkers Prev 19(7):1709–1722 31. Fromont A, Binquet C, Sauleau EA et al (2010) Geographic variations of multiple sclerosis in France. Brain 133:1889–1899 32. Fromont A, Binquet C, Sauleau E et al (2012) National estimate of multiple sclerosis incidence in France (2001-2007). Mult Scler 18(8):1108–1115 33. Vukusic S, Van Bockstael V, Gosselin S, Confavreux C (2006) Regional variations in the prevalence of multiple sclerosis in French farmers. J Neurol Neurosurg Psychiatry 78:707–709 34. Ebers GC (2008) Environmental factors and multiple sclerosis. Lancet Neurol 7(3):268–277 35. Kampman MT, Wilsgaard T, Mellgren SI (2007) Outdoor activities and diet in childhood and adolescence relate to MS risk above the Arctic Circle. J Neurol 254(4):471–477 36. Mohr SB, Gorham ED, Garland CF, Grant WB, Garland FC (2010) Low ultraviolet B and increased risk of brain cancer: an ecological study of 175 countries. Neuroepidemiology 35(4): 281–290
123
CLINICAL STUDY
Significant heterogeneity in the geographical distribution of diffuse grade II/III gliomas in France Ame´lie Darlix • Sonia Zouaoui • Jean-Marc Virion • Vale´rie Rigau • He´le`ne Mathieu-Daude´ • Marie Blonski • German Reyes-Botero • Faiza Bessaoud • Brigitte Tre´tarre • Fabienne Bauchet • Laurent Capelle Michel Fabbro • Christine Kerr • Dominique Figarella-Branger • Hugues Duffau • Luc Taillandier • Luc Bauchet
•
Received: 14 March 2014 / Accepted: 9 August 2014 / Published online: 24 August 2014 Ó Springer Science+Business Media New York 2014
Abstract Diffuse WHO grade II and III gliomas (DGII/ IIIG) are rare tumors, with few specific epidemiological studies. We aimed at describing the geographical distribution of a homogeneous series of histologically confirmed DGII/IIIG, over a four-year period (2006–2009), at a national level. The methodology is based on a multidisciplinary national network already established by the French Brain Tumor DataBase and data collected directly from every neuropathology department. Personal home addresses were collected for confirmed cases. For each region, the incidence of DGII/IIIG was analyzed and standardized on the age and sex distribution of the French population. The number of patients with newly diagnosed, histologically confirmed DGII/IIIG was 4,790. The overall crude rate was
19.4/106. To enable international comparisons, standardized rates were calculated as follows: 19.8/106, 18.8/106 and 16.0/ 106 (reference population, Europe, US and world, respectively). The geographical distribution by region showed significant differences, with higher incidence rates in Northeast and central parts of France. This work is the first studying the geographical distribution of a pure series of DGII/IIIG at a national level. It demonstrates significant heterogeneity in the distribution, and raises the question of the role of environmental and/or genetic risk(s) factor(s) for DGII/IIIG.
Ame´lie Darlix and Sonia Zouaoui equally contributed to the study.
Introduction
Electronic supplementary material The online version of this article (doi:10.1007/s11060-014-1585-0) contains supplementary material, which is available to authorized users. A. Darlix (&) M. Fabbro C. Kerr Departement of Medical Oncology and Department of Radiation Oncology, Institut Re´gional du Cancer de Montpellier - Val d’Aurelle, Montpellier, France e-mail: [email protected] A. Darlix S. Zouaoui H. Duffau L. Bauchet INSERM U1051, Montpellier, France S. Zouaoui H. Duffau L. Bauchet Department of Neurosurgery, Gui de Chauliac Hospital, Montpellier, France S. Zouaoui H. Mathieu-Daude´ F. Bessaoud B. Tre´tarre F. Bauchet Department of Epidemiology, French Brain Tumor Database, GNOLR, Registre des Tumeurs de l’He´rault, Institut Re´gional du Cancer de Montpellier - Val d’Aurelle, Montpellier, France
Keywords Brain tumor Neurooncology Glioma Epidemiology
Diffuse WHO grade II and III gliomas (DGII/IIIG) include diffuse astrocytomas, anaplastic astrocytomas, J.-M. Virion Department of Clinical Epidemiology and Evaluation, University Hospital, Nancy, France V. Rigau Department of Pathology, Gui de Chauliac Hospital, Montpellier, France M. Blonski Neuro-oncology Unit, Central Hospital, Nancy, France G. Reyes-Botero Department of Neurology, Pitie´-Salpe´trie`re Hospital, Paris, France L. Capelle Department of Neurosurgery, Pitie´-Salpe´trie`re Hospital, Paris, France
123
548
oligodendrogliomas, anaplastic oligodendrogliomas, and mixed gliomas [1]. They differ from the other gliomas (in particular the glioblastoma) in terms of clinical and radiological presentation, and natural history. They are rare, with an incidence B 20/106 person-years in the United States or in Europe [2–5]. Except for a few cases, their causes remain unknown. Environmental factors could be involved but to date the only one known to increase the risk of glioma is exposure to high-dose ionizing radiations [6– 9]. The epidemiology of DGII/IIIG is poorly known. Some studies have suggested heterogeneity in the geographical distribution of gliomas, at a worldwide or European level, with a trend towards higher incidence rates in highly developed countries (http://globocan.iarc.fr/) [2, 10–12]. However these studies have two major limits. First, they include all grades of gliomas or even different histological subtypes of primary central nervous system tumors (PCNSTs). Secondly, they compare incidence rates at an international or intercontinental level, resulting in an obvious bias due to differences among continents and European countries in terms of case registrations and access to healthcare. Here we aimed at studying the geographical distribution of a series of newly diagnosed and histologically confirmed DGII/IIIG on metropolitan France, over a four-year period.
Materials and methods Settings The present study was conducted in metropolitan France (including Corsica) representing 552,000 km2 and divided into 22 regions. The 2007 population in metropolitan France was 61,795,238. Population density and age distribution vary from one region to another (http://www.insee. fr/fr/themes/detail.asp?reg_id=99&ref_id=estim-pop). Data sources: French brain tumor database (FBTDB) Cases were identified by the FBTDB, which is the largest histological database for PCNSTs in Europe [13–17]. This multidisciplinary national network was established thanks to the collaboration of all scientific societies involved in
D. Figarella-Branger Department of Neuropathology and INSERM U911, Timone Hospital, Universite´ de la Me´diterrane´e, Marseille, France L. Taillandier Neuro-oncology Unit, University Hospital, Poitiers, France
123
J Neurooncol (2014) 120:547–555
the care of patients with PCNST in France (Association des Neuro-Oncologues d’Expression Franc¸aise, Socie´te´ Franc¸aise de Neurochirurgie, Socie´te´ Franc¸aise de Neuropathologie). It aims at prospectively registering all histologically-proven cases of PCNST in France. Its methodology was previously published [13, 16, 17]. Two different sources were used to collect data. The first source is a data sheet available in all operating rooms conducting PCNSTs surgery (n = 52), for prospective registration of cases. Secondly, to improve the exhaustiveness of data collection, an annual listing of all cases analyzed by each neuropathology department was collected to complete the FBTDB [17]. In some specific cases when data from the data sheet and listings did not match, or when the data sheet was not available for a case identified by the listing from one neuropathology department, a clinical research technician contacted directly the neuropathology department in order to verify and collect the missing data. All patients diagnosed with a histologically proven DGII/IIIG were included, i.e. patients with a WHO grade II or III astrocytoma (fibrillary astrocytoma, gemistocytic astrocytoma, protoplasmic astrocytoma or anaplastic astrocytoma, corresponding to ICD-O code 9420/3, 9411/3, 9410/3 and 9401/3, respectively), WHO grade II or III oligodendroglioma (ICD-O code 9450/3 and 9451/3, respectively) or WHO grade II or III mixed glioma (ICDO code 9382/3). Patients diagnosed with a ‘‘not otherwise specified’’ (NOS) astrocytoma (ICD-O code 9400/3, i.e. astrocytoma for which the WHO grading was unavailable, with the exclusion of glioblastoma) were included. By definition these tumors are classified as a DGIIG in the WHO classification and in the CBTRUS (Central Brain Tumor Registry of the United States) classification [1, 2]. Patients with a NOS mixed glioma (ICD-O code 9382/3) were also included. Since the French nomenclature (ADICAP 2003, http://www.adicap.asso.fr/THESAURUS/ Adicap_v5-03pdf) differentiates oligoastrocytoma from anaplastic oligoastrocytoma, the number of NOS mixed gliomas was expected to be low in our study and thus not to impact the repartition between GIIG and GIIIG. We chose of including NOS gliomas (ICD-O code 9380/3), i.e. diffuse gliomas for which both histological subtype and WHO grading were unavailable, since some among them are some DGII/IIIG, and their exclusion could result in an underestimation of the studied cases (see discussion). For this study, when the information collected from either the data sheet or listings from pathologists was NOS glioma, NOS mixed glioma or NOS astrocytoma, the pathological report was collected and further analysed by an expert pathologist (VR) in order to try and reach a diagnosis according to the WHO 2007 classification.
J Neurooncol (2014) 120:547–555
Diffuse WHO grade II astrocytomas for which the subtype was unknown (fibrillary, gemistocytic or protoplasmic) were called ‘‘WHO grade II astrocytomas’’. Diffuse WHO grade II and grade III gliomas for which the histological subtype was unknown (astrocytoma, oligodendroglioma or oligoastrocytoma) were called ‘‘NOS WHO grade II gliomas’’ and ‘‘NOS WHO grade III gliomas’’, respectively. Since differentiating between oligodendrogliomas and astrocytomas can, sometimes, be difficult, we chose to group all GIIG subgroups and all GIIIG subgroups in order to analyse their respective incidences [18–20]. Patients were included if the histological diagnosis had been made between January 1st, 2006 and December 31st, 2009, in one of the neuro-oncological centers in metropolitan France. Patients who had previous glioma surgery were excluded if the first diagnosis of glioma was made before January 1st, 2006. If the diagnosis was established during the inclusion period, data from the first surgery was collected and included in the study.
Geographical distribution The home address (city, zip code) of each patient at the time of the surgery was collected and matched to the corresponding five digits official geographical code (http:// insee.fr/fr/methodes/nomenclatures/cog/) as defined by the National Institute of Statistics and Economic Studies (INSEE). Patients for whom the home address at the time of surgery was unknown were not included in the analysis, neither were patients who lived abroad or in one of the overseas French departments or territories.
Statistical analysis Statistical analysis was performed with the SAS software (v9.3). Standardized incidences (with 95 % confidence intervals) were calculated for each region. Incidences were calculated and adjusted for age and sex based on the distribution of the French population (reference population: France INSEE 2007 http://www.insee.fr/fr/ themes/document.asp?ref_id=ip1170#inter1). To enable international comparisons, standardized rates for all DGII/IIIG were also calculated with Europe, USA and worldwide populations (2007) as references. Incidence ratios were defined as the standardized incidence for the area of interest divided by the national incidence. Differences in incidence between the area of interest and national incidence were tested using a Chi squared test. Maps of standardized incidences were drawn using the Geoclip website (http://www.geoclip.fr/fr/) approved by the INSEE.
549
Results Population characteristics Over the four-year period, 4,915 patients had surgery for newly diagnosed DGII/IIIG in metropolitan France. Among them 125 (2.5 %) were excluded from the study: 91 patients (1.8 %) lived abroad or in an overseas French department or territory, and the home address was unknown for 34 patients (0.7 %). Therefore, 4,790 patients were included (1,220, 1,190, 1,228, and 1,152 for the years 2006, 2007, 2008 and 2009, respectively). The distribution of patients according to the histological subtype is summarized in Table 1. Among the 4,790 patients included, 2,099 (43.82 %) were diagnosed with a WHO grade II glioma and 2,484 (51.86 %) with a WHO grade III glioma. Thirty-eight (0.79 %), 25 (0.52 %) and 144 (3.01 %) patients were diagnosed with a NOS astrocytoma, a NOS mixed glioma, or a NOS glioma, respectively. Median and mean ages at diagnosis were 51 (range 0–88) and 49.8 ± 17.6 (44.8 ± 16.6 for WHO grade II gliomas vs. 54.0 ± 16.9 for WHO grade III gliomas; p \ 0.001, F-test), respectively. Table 2 shows the distribution of patients according to their age at diagnosis. The [30–39] and [40–49] age classes were the most represented for WHO grade II gliomas while the [50–59] and [60–69] were the most represented for WHO grade III gliomas. The distribution of histological subtypes of tumors significantly varied (p \ 0.001, Chi squared test) according to the age class. Diffuse WHO grade II gliomas were the most represented among people aged 49 or less, while diffuse WHO grade III were the most represented among people over 50 (p \ 0.001). In our study 56.9 % of patients were male. There was no difference in terms of sex ratio between WHO grade II and WHO grade III gliomas (p = 0.8, Chi squared test). Mean age was 50.0 ± 16.9 for men and 49.5 ± 18.3 for women. Median age was 52 (range 0–86) for men and 51 (range 0–88) for women. The distribution of histological subtypes of tumors according to gender showed no significant difference between men and women. Standardized incidence rates The standardized average annual age and sex adjusted incidence rate for all DGII/IIIG (n = 4,790) was 19.4/106 (95 % CI 18.8–19.9). To enable international comparisons, standardized rates were calculated as follows: 19.8/106 (population of reference: Europe), 18.8/106 (reference: USA), and 16.0/106 (reference: world) (Table 3). When NOS gliomas were excluded (standardized incidence of NOS gliomas 0.6/106; 95 % CI 0.5–0.7), standardized rates
123
550
J Neurooncol (2014) 120:547–555
Table 1 Cases distribution among men and women according to histologic subtypes with the corresponding median age at diagnosis Histology
N
Males (n)
Females (n)
Median age (in years)
Diffuse WHO grade II astrocytoma (fibrillary, gemistocytic or protoplasmic astrocytoma, ‘‘WHO grade II astrocytoma’’)
286
165
121
47
Anaplastic astrocytoma
59
417
234
183
NOSa astrocytoma
38
21
17
Oligodendroglioma
1,354
763
591
44
Anaplastic oligodendroglioma
1,196
678
518
56
Oligoastrocytoma
412
231
181
42
Anaplastic oligoastrocytoma NOSa mixed glioma
827 25
485 13
342 12
56 44
NOSa WHO grade II glioma
47
34
13
51
NOSa WHO grade III glioma
44
24
20
51
144
77
67
57
4,790
2,725
2,065
51
NOSa glioma Total a
48.5
NOS not otherwise specified
Table 2 Distribution of patients according to their age at diagnosis Age class (years)
All DGII/IIIG (n = 4,790), n (%)
WHO grade II gliomas (n = 2,099), n (%)
WHO grade III gliomas (n = 2,484), n (%)
[0–19]
244 (5.1)
119 (5.7)
95 (3.8)
[20–29]
384 (8.0)
253 (12.0)
122 (4.9)
[30–39]
724 (15.1)
448 (21.3)
258 (10.4)
[40–49]
866 (18.1)
448 (21.3)
390 (15.7)
[50–59]
1,043 (21.8)
412 (19.6)
592 (23.8)
[60–69]
870 (18.2)
262 (12.5)
565 (22.7)
C70
659 (13.8)
157 (7.5)
462 (18.6)
were calculated as follows: 19.2/106 (reference: Europe), 18.2/106 (reference: USA), and 15.5/106 (reference: world). The standardized incidence rates for all diffuse WHO grade II gliomas (n = 2,099) and WHO diffuse grade III gliomas (n = 2,484) were 8.5/106 (95 % CI 8.1–8.9) and 10.0/106 (95 % CI 9.7–10.4), respectively. The standardized incidence rates for grade II and III astrocytomas, grade II and III oligodendrogliomas and grade II and III oligoastrocytomas were 3.0/106 (95 % CI 2.8–3.2), 10.3/106 (95 % CI 9.9–10.7) and 5.0/106 (95 % CI 4.7–5.3), respectively.
see Fig. 2). It was significantly lower than the national incidence in Iˆle-de-France, Basse-Normandie, Rhoˆne-Alpes and Provence-Alpes-Coˆte d’Azur (incidence rates 13.9/ 106 to 17.8/106). The incidence rates of each of the 22 regions are summarized in Table 4 (supplementary materials). We also found a significant heterogeneity in the geographical distribution of WHO grade II and WHO grade III gliomas (Fig. 3 and Fig. 4, supplementary materials). The standardized incidence rates for each of the 22 regions for WHO grade II and WHO grade III gliomas are summarized in Tables 5 and 6, respectively (supplementary materials).
Geographical distribution We found a significant heterogeneity in the geographical distribution by regions of all DGII/IIIG in metropolitan France (Fig. 1). The standardized incidence rate (reference: France) was significantly higher than the national incidence (i.e. the incidence ratio was significantly different from 1) in Champagne-Ardenne, Nord-Pas-de-Calais, Alsace, Bretagne and Auvergne (incidence rates 21.9/106 to 30.9/106,
123
Discussion This study is the first to analyze incidence rates and geographical distribution at a national level for DGII/IIIG. To our knowledge, this is the largest series that exhaustively includes homogeneous subtypes of histologically proven glioma, with 4,790 confirmed cases of DGII/IIIG.
J Neurooncol (2014) 120:547–555
551
Table 3 Standardized rates and international comparisons (Age-sex adjusted incidence rates, per 106 with 95 % confidence intervals, to 2007 Europe, USA and World standard population) Reference population France
Europe
USA
All DGII/IIIGa
19.4 [18.8–19.9]
19.8 [19.2–20.4]
18.8 [18.2–19.3]
16 [15.5–16.4]
All DGII/IIIGa excluding NOS gliomas
18.8 [18.3–19.3]
19.2 [18.6–19.8]
18.2 [17.7–18.7]
15.5 [15.0–16.0]
All DGIIGb All DGIIIGc
8.5 [8.1–8.9] 10.0 [9.7–10.4]
8.6 [8.3–9.0] 10.3 [9.9–10.7]
8.3 [8.0–8.7] 9.6 [9.3–10.0]
7.6 [7.3–8.0] 7.6 [7.3–8.0]
a
World
DGII/IIIG diffuse grade II or III glioma
b
DGIIG diffuse grade II glioma
c
DGIIG diffuse grade III glioma
Incidence data In our series the standardized incidence rate adjusted on the French population for all DGII/IIIG was 19.4/106 (19.8/ 106, 18.8/106 and 16.0/106 when standardized on European, USA and worldwide populations, respectively). These results are consistent with those from larger series of PCNSTs, i.e. data from the CBTRUS [2] (standardized incidence rate: 19.8/106, and higher proportion of NOS gliomas). For diffuse WHO grade II gliomas our results match those in recently published studies, with a standardized incidence rate of 8.5/106, lower than the one reported by the Austrian Brain Tumor Registry (incidence rate adjusted to the USA population: 12.2/106) but higher than in recent population-based studies in Europe [4, 10, 21]. In France, the only data available are based on a departmental registry and thus on a much smaller population (incidence rate for DGII/IIIG 13.7/106, excluding NOS gliomas, 2000-2007) [5]. Data from England (incidence rate for DGII/IIIG 21.2/106, 1999–2003, worldwide population) included about 30 % of NOS gliomas, resulting in an important selection bias [22]. Heterogeneity in geographical distribution of DGII/IIIG We showed significant differences in the metropolitan France geographical distribution for all DGII/IIIG, diffuse WHO grade II and diffuse WHO grade III gliomas, with higher incidence rates in the Northeast and central France. This uneven geographical distribution may suggest a role of environmental and/or genetic factors. However we cannot exclude some methodological bias. First, we cannot exclude a selection bias due to differences across regions in terms of access to healthcare. However, as metropolitan France has a unique healthcare system, access to healthcare is likely to be similar for all regions. The distribution of neurosurgical centers is roughly homogeneous on the metropolitan French territory,
with one or more neurosurgical center(s) (depending on the number of inhabitants) for each region (and no region without a neurosurgical center). Moreover, data from the national medical board (http://www.conseil-national.mede cin.fr/sites/default/files/Atlas__national_2013.pdf) shows that there is no major difference among regions as regards the distribution of neurosurgeons. Important differences in population sizes among the compared regions might also be a source of bias, by making the distinction between random and true variations in incidence rates difficult. Another concern is that we are unaware of how long patients have lived in the same city at the time of the surgical procedure. If a patient has moved from one region to another not long before tumor diagnosis, the home address collected might not appropriately reflect environmental exposure. A large number of such patients could also influence the regional incidence rates. However, data from the INSEE on the 2008 French population shows that the number of home residence changes between regions is low (http://www.insee.fr/fr/bases-de-donnees/default.asp? page=recensement/resultats/doc/presentation-fichiers-detailmigrations-residentielles.htm). We chose to include NOS gliomas, i.e. gliomas for which both histological subtype and WHO grading were unavailable due to neuropathological difficulties. Only 144 cases (3.1 %) of NOS gliomas could not be reclassified after analysis of the pathological report. We cannot exclude that some of these tumors were actually WHO grade IV gliomas, particularly because the median age (57) is closer to that of high-grade than low-grade glioma patients. Yet, due to their small number and the fact that for these 144 patients, home addresses at diagnosis were distributed all over metropolitan France, we believe that their inclusion was unlikely to bias our results. As the neuropathological diagnosis of the grade can be difficult in gliomas, we cannot exclude a bias due to inconsistency in tumor grading between regions. To this end, a centralized pathological review would be of great
123
552
J Neurooncol (2014) 120:547–555
Fig. 1 Standardized incidences for each region for all DGII/IIIG (n = 4,790)
value but is currently not possible to get at the national level in France, or, to our knowledge, in any other country. However, with regards to the differentiation between grade III and grade IV gliomas, it has been established by the Brain Tumor Epidemiology Consortium that glioblastomas have a good general agreement across regions [20]. However it would be interesting to find out if the geographical distribution of glioblastomas has the same pattern than that of DGII/IIIG. As the differential diagnosis between diffuse grade II and grade III gliomas can be challenging, we chose
123
to group grade II and II gliomas together to eliminate any possible bias due to misclassification between grade II and grade III gliomas. Our study only included tumors for which a histological confirmation was available, since the FBTDB does not record non-histologically confirmed cases. As there is no national PCNSTs registry in France, these non-histological cases are very difficult to record. We believe that their noninclusion does not bias our results in terms of geographical distribution for two main reasons. First, for the last
J Neurooncol (2014) 120:547–555
553
Fig. 2 Standardized incidence ratios for each region for all DGII/ IIIG (n = 4,790). Regions in red correspond to areas with a significantly higher standardized incidence compared to the mean
incidence in France, whereas regions in green correspond to areas with a significantly lower standardized incidence
10 years in France clinicians have promoted histological confirmation of tumors in patients \70. Secondly, it has been shown, based on a French departmental registry, that astrocytomas and oligodendrogliomas are histologically validated in more than 98 % and 93 % of cases, respectively (http://etudes.isped.u-bordeaux2.fr/REGISTRES-CAN CERS-AQUITAINE/Snc/document/Rtsnc_lettre_information_ 200905.pdf). We were not able to provide molecular data, as this piece of information is not currently collected by the FBTDB. It would be of great value and we will aim at
adding molecular markers to the data collection in a near future.
Heterogeneity in geographical distribution of DGII/IIIG: hypothesis The uneven geographical distribution of DGII/IIIG may suggest a role of environmental and/or genetic factors. Interestingly, a geographical heterogeneity in incidence rates has been highlighted for other cancers. In France,
123
554
incidence rates for all cancers were higher in the North and Northeast regions (data from the National Cancer Institute: http://www.e-cancer.fr/publications/69-epidemio logie/574-la-situation-du-cancer-en-france-en-2011). Especially, distribution of colorectal cancer (CRC) was comparable to that of DGII/IIIG, with similar higher incidence rates in the Northeast part of France. We can thus hypothesize that these two diseases may share some risk factors. Diet, has been validated as a risk factor in CRC with red and processed meat, alcohol, obesity or abdominal fat being strongly associated with a higher risk of CRC, while the ‘‘Mediterranean diet’’ prevents its occurrence [23, 24]. In France, the consumption of red and processed meat is higher in Northeast regions (CREDOC, research center for the study and the observation of living conditions, http://www.credoc.fr/pdf/4p/ 101.pdf). Moreover the prevalence of obesity varies across France in a similar way (http://www.insee.fr/fr/ffc/ ipweb/ip1123/ip1123.pdf). A working hypothesis could thus be that diet factors, and in particular meat-based diet and obesity, may have an impact on DGII/IIIG risk. Dietary sources of nitroso compounds (such as cured or processed meats) have been associated with an increased glioma risk in several population-based studies [25–27]. Some other studies however found no conclusive evidence for an association between gliomas (or PCNSTs) and processed meat [28–30]. So far, no epidemiological study has investigated the association between processed meat and DGII/IIIG risk specifically. Multiple sclerosis (MS) is another neurological disease with a higher incidence in Northeast France, comparable to what was found for DGII/IIIG [31–33]. As DGII/IIIG have the same pattern of geographical distribution than MS, these two diseases may share some risk factors. There is strong evidence that environmental factors influence the uneven distribution of MS. In particular, the amount of sunlight (and, consequently, outdoor activities) is significantly associated with MS prevalence (sunny areas being associated with low prevalence). This association probably involves vitamin D, as a low prevalence of MS has been reported in countries with a poor amount of sunlight but a high consumption of vitamin D (oily fish) [34, 35]. We could thus hypothesize that sunlight and vitamin D intake influence the risk of DGII/IIIG. Based on the map of the sunlight provided by the Joint Research center of the European Commission (available at: http://re.jrc.ec.europa.eu/pvgis/cmaps/eu_cmsaf_opt/G_ opt_FR.png), it seems that regions with low sunlight correlate with those with higher incidence rates of DGII/IIIG in our study, supporting the hypothesis of an association between sunlight and DGII/IIIG risk. This hypothesis is also supported by a study that found a reverse association between UVB irradiance and brain tumor risk [36].
123
J Neurooncol (2014) 120:547–555
Conclusion This work is, to our knowledge, the first to analyse the geographical distribution over a four-year period of a pure series of DGII/IIIG at a national level, thus studying a homogeneous population in terms of access to healthcare. This study demonstrates a significant heterogeneity in the geographical distribution of newly diagnosed and histologically confirmed DGII/IIIG in metropolitan France and raises the question of environmental and/or genetic risk factors for DGII/IIIG. The next step will be to compare environmental and genetic factors between regions with low and those with high incidence. Acknowledgments The authors wish to thank all pathologists, neurosurgeons and other stakeholders, for their help and input with this work (see the list in supplementary materials), and Be´ne´dicte Cle´ment, scientific translator, for valuable translation work. This work was conducted with the financial support in the form of grants from the Ligue Nationale Contre le Cancer, Association des NeuroOncologues d’Expression Franc¸aise, Socie´te´ Franc¸aise de Neurochirurgie, Associations pour la Recherche sur les Tumeurs Ce´re´brales (ARTC and ARTC Sud), Roche Laboratory, Sophysa Laboratory, De´partement de l’He´rault, and Groupe de Neuro-Oncologie du Languedoc Roussillon. Conflict of interest
None declared.
References 1. Louis DN, Ohgaki H, Wiestler OD et al (2007) The 2007 WHO classification of tumours of the central nervous system. Acta Neuropathol 114:97–109 2. Dolecek TA, Propp JM, Stroup NE, Kruchko C (2012) CBTRUS statistical report: primary brain and central nervous system tumors diagnosed in the United States in 2005-2009. Neuro Oncol 14(suppl 5):v1–v49 3. Ohgaki H, Kleihues P (2005) Epidemiology and etiology of gliomas. Acta Neuropathol 109:93–108 4. Wo¨hrer A, Waldho¨r T, Heinzl H et al (2009) The Austrian Brain Tumour Registry: a cooperative way to establish a populationbased brain tumour registry. J Neurooncol 95:401–411 5. Baldi I, Gruber A, Alioum A et al (2011) Descriptive epidemiology of CNS tumors in France: results from the Gironde Registry for the period 2000-2007. Neuro Oncol 13:1370–1378 6. Ron E, Modan B, Boice JD, Alfandary E, Stovall M, Katz L (1988) Tumors of the brain and nervous system after radiotherapy in childhood. N Engl J Med 319:1033–1039 7. Yonehara S, Brenner AV, Kishikawa M et al (2004) Clinical and epidemiologic characteristics of first primary tumors of the central nervous system and related organs among atomic bomb survivors in Hiroshima and Nagasaki, 1958-1995. Cancer 101:1644–1654 8. Bondy ML, Scheurer ME, Malmer B et al (2008) Brain tumor epidemiology: consensus from the Brain Tumor Epidemiology Consortium. Cancer 113(7 Suppl):1953–1968 9. McCarthy BJ, Rankin KM, Aldape K et al (2011) Risk factors for oligodendroglial tumors: a pooled international study. Neuro Oncol 13:242–250
J Neurooncol (2014) 120:547–555 10. Crocetti E, Trama A, Stiller C et al (2012) Epidemiology of glial and non-glial brain tumours in Europe. Eur J Cancer 48:1532–1542 11. Houben MP, Coebergh JW, Birch JM, Tijssen CC, van Duijn CM, McNally RJ (2005) Space–time clustering patterns of gliomas in the Netherlands suggest an infectious aetiology. Eur J Cancer 41:2917–2923 12. Houben MP, Coebergh JW, Birch JM, Tijssen CC, van Duijn CM, McNally RJ (2006) Space-time clustering of glioma cannot be attributed to specific histological subgroups. Eur J Epidemiol 21:197–201 13. Bauchet L, Rigau V, Mathieu-Daude´ H et al (2007) French brain tumor data bank: methodology and first results on 10,000 cases. J Neurooncol 84:189–199 14. Bauchet L, Rigau V, Mathieu-Daude´ H et al (2009) Clinical epidemiology for childhood primary central nervous system tumors. J Neurooncol 92:87–98 15. Bauchet L, Mathieu-Daude´ H, Fabbro-Peray P et al (2010) Oncological patterns of care and outcome for 952 patients with newly diagnosed glioblastoma in 2004. Neuro Oncol 12:725–735 16. Rigau V, Zouaoui S, Mathieu-Daude´ H et al (2012) French brain tumor database: 5- year histological results on 25756 cases. Brain Pathol 21:633–644 17. Zouaoui S, Rigau V, Mathieu-Daude´ H et al (2012) French brain tumor database: general results on 40,000 cases, main current applications and future prospects. Neurochirurgie 58:4–13 18. Hartmann C, von Deimling A (2009) Molecular pathology of oligodendroglial tumors. Recent Results Cancer Res 171:25–49 19. van den Bent MJ (2010) Interobserver variation of the histopathological diagnosis in clinical trials on glioma: a clinician’s perspective. Acta Neuropathol 120(3):297–304 20. Davis FG, Malmer BS, Aldape K et al (2008) Issues of diagnostic review in brain tumor studies: from the Brain Tumor Epidemiology Consortium. Cancer Epidemiol Biomarkers Prev 17(3): 484–489 21. Okamoto Y, Di Patre P-L, Burkhard C, Horstmann S, Jourde B, Fahey M et al (2004) Population-based study on incidence, survival rates, and genetic alterations of low-grade diffuse astrocytomas and oligodendrogliomas. Acta Neuropathol 108:49–56 22. Arora RS, Alston RD, Eden TO et al (2010) Are reported increases in incidence of primary CNS tumours real ? An analysis of longitudinal trends in England, 1979-2003. Eur J Cancer 46:1607–1616
555 23. Chan DS, Lau R, Aune D et al (2011) Red and processed meat and colorectal cancer incidence: meta-analysis of prospective studies. PLoS One 6(6):e20456 24. Aleksandrova K, Nimptsch K, Pischon T (2013) Influence of obesity and related metabolic alterations on colorectal cancer risk. Curr Nutr Rep 2(1):1–9 25. Boeing H, Schlehofer B, Blettner M, Wahrendorf J (1993) Dietary carcinogens and the risk for glioma and meningioma in Germany. Int J Cancer 53(4):561–565 26. Blowers L, Preston-Martin S, Mack WJ (1997) Dietary and other lifestyle factors of women with brain gliomas in Los Angeles County (California, USA). Cancer Causes Control 8:5–12 27. Kyritsis AP, Bondy ML, Levin VA (2011) Modulation of glioma risk and progression by dietary nutrients and antiinflammatory agents. Nutr Cancer 63(2):174–184 28. Huncharek M, Kupelnick B, Wheeler L (2003) Dietary cured meat and the risk of adult glioma: a meta-analysis of nine observational studies. J Environ Pathol Toxicol Oncol 22(2):129–137 29. Michaud DS, Holick CN, Batchelor TT, Giovannucci E, Hunter DJ (2009) Prospective study of meat intake and dietary nitrates, nitrites, and nitrosamines and risk of adult glioma. Am J Clin Nutr 90:570–577 30. Dubrow R, Daresky AS, Park Y et al (2010) Dietary components related to N-Nitroso compound formation: a prospective study of adult glioma. Cancer Epidemiol Biomarkers Prev 19(7):1709–1722 31. Fromont A, Binquet C, Sauleau EA et al (2010) Geographic variations of multiple sclerosis in France. Brain 133:1889–1899 32. Fromont A, Binquet C, Sauleau E et al (2012) National estimate of multiple sclerosis incidence in France (2001-2007). Mult Scler 18(8):1108–1115 33. Vukusic S, Van Bockstael V, Gosselin S, Confavreux C (2006) Regional variations in the prevalence of multiple sclerosis in French farmers. J Neurol Neurosurg Psychiatry 78:707–709 34. Ebers GC (2008) Environmental factors and multiple sclerosis. Lancet Neurol 7(3):268–277 35. Kampman MT, Wilsgaard T, Mellgren SI (2007) Outdoor activities and diet in childhood and adolescence relate to MS risk above the Arctic Circle. J Neurol 254(4):471–477 36. Mohr SB, Gorham ED, Garland CF, Grant WB, Garland FC (2010) Low ultraviolet B and increased risk of brain cancer: an ecological study of 175 countries. Neuroepidemiology 35(4): 281–290
123
