ORIGINAL ARTICLES AND REVIEWS
Cancer incidence in Italian contaminated sites
Pietro CombaI; Paolo RicciII,III; Ivano IavaroneI; Roberta PirastuIV; Carlotta BuzzoniIII,VI; Mario FuscoIII,VII; Stefano FerrettiIII,VIII; Lucia FazzoI; Roberto PasettoI; Amerigo ZonaI; Emanuele CrocettiIII, V; for ISS-AIRTUM Working Group for the study of cancer incidence in contaminated sites*
IDipartimento di Ambiente e Connessa Prevenzione Primaria, Istituto Superiore di Sanità, Rome, Italy
IIRegistro Tumori di Mantova, Mantua, Italy
IIIAIRTUM, Associazione Italiana dei Registri Tumori, Italy
IVDipartimento di Biologia e Biotecnologie "Charles Darwin", Sapienza Università di Roma, Rome, Italy
VUnità di Epidemiologia Clinica e Descrittiva, Istituto per lo Studio e la Prevenzione Oncologica, Florence, Italy
VIRegistro Tumori di Napoli (ASL Napoli 4), Naples, Italy
VIIRegistro Tumori di Ferrara, Dipartimento di Chirurgia e Medicina Sperimentale, Università degli Studi di Ferrara, Ferrara, Italy
ABSTRACT
INTRODUCTION. The incidence of cancer among residents in sites contaminated by pollutants with a possible health impact is not adequately studied. In Italy, SENTIERI Project (Epidemiological study of residents in National Priority Contaminated Sites, NPCSs) was implemented to study major health outcomes for residents in 44 NPCSs.
METHODS. The Italian Association of Cancer Registries (AIRTUM) records cancer incidence in 23 NPCSs. For each NPCSs, the incidence of all malignant cancers combined and 35 cancer sites (coded according to ICD-10), was analysed (1996-2005). The observed cases were compared to the expected based on age (5-year period,18 classes), gender, calendar period (1996-2000; 2001-2005), geographical area (North-Centre and Centre-South) and cancer sites specific rates. Standardized Incidence Ratios (SIR) with 90% Confidence Intervals were computed.
RESULTS. In both genders an excess was observed for overall cancer incidence (9% in men and 7% in women) as well as for specific cancer sites (colon and rectum, liver, gallbladder, pancreas, lung, skin melanoma, bladder and Non Hodgkin lymphoma). Deficits were observed for gastric cancer in both genders, chronic lymphoid leukemia (men), malignant thyroid neoplasms, corpus uteri and connective and soft-tissue tumours and sarcomas (women).
DISCUSSION. This report is, to our knowledge, the first one on cancer risk of residents in NPCSs. The study, although not aiming to estimate the cancer burden attributable to the environment as compared to occupation or life-style, supports the credibility of an etiologic role of environmental exposures in contaminated sites. Ongoing analyses focus on the interpretation of risk factors for excesses of specific cancer types overall and in specific NPCSs in relation to the presence of carcinogenic pollutants.
Key words: contaminated sites. environmental pollution. cancer incidence
INTRODUCTION
According to WHO contaminated sites are "areas hosting or having hosted human activities which have produced or might produce environmental contamination of soil, surface or groundwater, air, food-chain, resulting or being able to result in human health impacts" [1]. About 127 000 sites have been identified as contaminated in 33 European countries [2], a legacy of early industrialization and poor management practices. Characterizing the health impact of contaminated sites involves a multidisciplinary approach, with a relevant role of epidemiology both in the analysis and evolving pattern of the health profiles of populations, and in the study of the associations between environmental exposures and health effects, according to specific hypotheses [1]. Protecting human health from exposure to environ mental pollutants is a public health priority, but the burden of tumours among residents in areas contaminated by carcinogens is not adequately studied and the specific epidemiological literature on cancer risk in polluted areas is limited. In Italy, SENTIERI Project (Epidemiological study of residents in National Priority Contaminated Sites, NPCSs), funded by the Italian Ministry of Health, studied mortality among residents of 44 NPCSs included in the "National Environmental Remediation Programme" [3, 4].
The original feature of SENTIERI Project is the a priori evaluation of the epidemiological evidence on the causal association between 63 causes of death and residence in proximity of environmental sources of emission/release (chemical industries, petrochemicals and refineries, steel plants, power plants, mines and/ or quarries, harbour areas, asbestos or other mineral fibres, landfills and incinerators). The evaluation, based on standardized methodology [3, 4] was carried out to minimize the risk for researchers to be data-driven when interpreting results. Results of mortality analyses were published for each NPCS and for all 44 NPCSs combined [5]. In the period 1995-2002 for a total of 403 692 deaths (men and women combined), an excess of 9969 deaths was observed, 4309 out of 9969 excess cases were from cancer deaths.
A recent development of SENTIERI Project is the study of cancer incidence in NPCSs served by cancer registries. As compared to mortality, cancer incidence allows to investigate shorter latency periods, is mainly based on pathologically confirmed data not influenced by differential survival rates possibly related to socioeconomic deprivation, it accounts also for non lethal neoplasms. The Italian Association of Cancer Registries (AIRTUM, www.registri-tumori.it) records cancer incidence through population-based cancer registries. AIRTUM data are included in Cancer incidence in five continents published by the International Agency for Research on Cancer [6]. AIRTUM, funded by the Italian Ministry of Health, is involved in national [7] and international scientific collaborations [8].
The aim of the present study is to describe cancer incidence in populations living in NPCSs, as compared to populations not residing in these areas, thus integrating the previously reported cancer mortality data [4, 5]. Cancer incidence in all NPCSs combined is presented in this paper.
METHODS
Italian NPCSs mainly include industrial sites characterized by the presence of chemical, petrochemical and steel industries, toxic waste dump sites and sites contaminated by asbestos and fluoro-edenite fibres [9, 10]. AIRTUM cancer registries are active in 23 out of 44 NPCSs included in SENTIERI Project: 17 covering all age classes and six specialised in childhood cancer (0-19 years). The present investigation includes more than 2000 000 people living in 23 NPCSs.
For each NPCSs, the incidence of all malignant cancers combined and 35 cancer sites, or groups of cancer sites, was computed. Cancer cases were coded according to the 10th International Classification of Diseases (ICD-10). Figure 1 shows, for the 23 NPCSs, the geographical distribution, population size and years of activity of each cancer registry. The observation period is 1996-2005. In 5 out of 23 areas the number of years of observation is less than 10 years, ranging from 3 to 7.
The number of observed cases for all malignancies and specific cancer sites was compared to the expected figures based on age in 5-years period (18 classes), gender, period (1996-2000; 2001-2005), area (NorthCentre and Centre-South, not overlapping) and cancer sites specific rates. Two sets of geographical specific rates were calculated in light of a decreasing trend from north to south for the majority of cancer sites in Italy [11]. The populations from each NPCS and from two registries not participating to the study were excluded when calculating the expected. Standardized Incidence Ratios (SIRs) and 90% Confidence Intervals (90% CIs) were computed.
RESULTS
The overall study findings are reported in Table 1. In the overall analysis of 23 Italian contaminated sites served by cancer registries, on the basis of 57 391 and 49 058 incident cases respectively among men and women, an excess of cancer incidence (9% in men and 7% in women) was observed. The excess of all malignant neoplasms present in both genders was mainly explained by malignant neoplasms of colon and rectum (in men and women respectively 7204 cases, 5% increase and 6460 cases 6% increase), liver (in men and women respectively 3024 cases, 39% increase and 1419 cases 32% increase), gallbladder (in men and women respectively 589 cases, 12% increase and 951 cases 12% increase), pancreas (in men and women respectively 1543 cases, 12% increase and 1635 cases 8% increase), lung (in men and women respectively 9396 cases, 7% increase and 3046 cases 24% increase), skin melanoma (in men and women respectively 1280 cases, 24% increase and 1285 cases 14% increase), bladder (in men and women respectively 5872 cases, 5% increase and 1601 cases 15% increase) and Non Hodgkin lymphoma (in men and women respectively 1926 cases, 6% increase and 1886 cases 7% increase). Excesses of mesothelioma, and malignant neoplasms of prostate, testis, kidney and brain were present among men, among women increased incidence was present for breast cancer, lymphohaematopoietic system and in particular chronic myeloid leukemia. A deficit of gastric cancer was observed in both genders, deficits of total leukemias and specifically of (chronic) lymphoid leukemia were observed in men and, in women, deficits of malignant neoplasms of thyroid, corpus uteri and connective and soft-tissue tumours and sarcomas.
DISCUSSION
The assessment of the possible health impact of contaminated sites is a challenging issue, especially in the case of industrially contaminated sites involving multiple and heterogeneous human exposures. The methods used in the SENTIERI Project has been defined by WHO as a first level approach to describe population health profiles in contaminated areas [1].
The present report is, to our knowledge, the first one on cancer risk of communities resident in areas defined as National Priority Contaminated Sites. The excess of cancer incidence in the pooled analysis confirmed the previous observation of cancer mortality excess in 44 sites, 4.8% in men and 2.6% in women [4, 12].
Etiological hypotheses of a causal link between cancer incidence and residence in contaminated areas are present in the international scientific literature. There is general agreement, though, that an evaluation of sufficient evidence of causality cannot be formulated, the exception being pleural mesothelioma and residence near asbestos mines or asbestos industries, such as asbestos-cement plants. Recently, increasing evidence has been provided for associations between residence near:
i) processing and storage facilities and lung cancer and Non Hodgkin lymphoma [13, 14]; ii) refineries, thermal power plants, steel and pulp mills and breast cancer [15]; iii) metal industries and prostate cancer [16] and digestive tract cancer [17]; iv) sites contaminated by toxic waste dumps and bladder and kidney cancer [18]. Increased incidence of haematological cancers, was observed in ZIP codes containing benzene waste sites [19]. A detailed review of this body of evidence has been recently published [20].
The present study did not aim to estimate the burden of tumours attributable to environmental exposures as compared to occupational exposures or life-style factors. However our results support the credibility of an etiologic role of environmental exposures, also in reason of several cancer excesses detected in both genders. Analytical studies, with improved exposure assessment, are needed to reach more firm conclusions.
The plausibility of the observed cancer excesses is supported by their consistency with the a priori hypothesized increases [3] and by the detailed report currently in press [21] on each of the 18 contaminated sites included in the present article.
In SENTIERI Project 90% CI were computed for risk estimators to present the range of uncertainty. Estimators were not accompanied by a hypothesis test to discriminate statistical significant from non statistical significant results. The choice of 90% level was made to minimize the acritical use of CI as surrogate of hypothesis testing; such use could lead to consider relevant only those estimators for which the CI exclude the null value, i.e. the ones customarily defined as "statistical significant" [22, 23]. The discriminating use of statistical significance in the evaluation of causal associations in epidemiology has been discussed since 1965 [24] and recently re-proposed [25]. This is particularly the case in SENTIERI Project where an a priori evaluation of the epidemiological evidence was completed to identify those diseases for which the causal association with the environmental exposures was either ascertained or probable. In SENTIERI the primary interest is on excesses or defects, i.e. direction of the risk estimators, and after on their size and precision, the latter inferred from the range of CI.
The strengths of the present study are the validity of cancer registration and the elevated number of cases included in the investigation, which ensures a high statistical power. The quality indicators of AIRTUM registries comply with the international standards [6]. In addition, specific checks are routinely performed using also a software developed by AIRTUM (CheckAIRTUM, www.registri-tumori.it). Furthermore, the selection of the study areas and the definition of a priori hypotheses has followed standardized and explicit criteria and procedures [3-5]. Major limitations of the investigation are the lack of a quantitative indicator of population exposure and the use of municipality as the smallest level of data aggregation.
Another limitation is that risk estimates are not adjusted for potential confounders such as alcohol consumption, smoking and socioeconomic status (SES), this last both positively and negatively associated with some cancers [3]. This lack of adjustment is most often the case in ecological studies. The overall estimates in the present paper were not adjusted for socioeconomic status (SES) although populations living in polluted areas have low SES [26]. This decision was based on a sensitivity analysis of ad hoc SENTIERI Deprivation Index which did not show substantial differences between unadjusted and adjusted risk estimates. As far as smoking and alcohol consumption are concerned, most cancer sites in excess have Sufficient or Limited evidence of association with tobacco or alcohol consumption [3]. There is no evidence that the environmental exposures in NPCSs are associated to either exposure, therefore the lack of adjustment should not lead to biased estimates.
Some underestimation of risk may derive from the incomplete coverage of cancer registration in some NPCSs. Developments of the multi-outcome SENTIERI approach include morbidity (as estimated by hospital discharge records), occurrence of congenital anomalies (pathology registries) and a specific focus on children health [27]. Ongoing analyses are focusing on the interpretation of cancer incidence in single NPCS, in relation to industrial emissions and/or releases from dumping sites as well as the presence of chemical carcinogens with evidence of association to specific tumours. Studies on the geographical distribution of cancer incidence based on individual data and environmental exposure modelling are also under way.
SENTIERI introduces a multiple outcome analytical model based on updated health outcomes in order to establish a permanent observation system to monitor the state of health of residents in contaminated areas. This will pave the way for more in-depth analytical epidemiological investigations and support the establishment and continued monitoring of primary prevention projects. In SENTIERI particular attention is devoted to issues of information and communication.
Acknowledgements
SENTIERI Project was partially financed by the Ministry of Health's Project CCM 2009 "Epidemiological surveillance of populations living in contaminated sites".
Conflict of interest statement
There are no potential conflicts of interest or any financial or personal relationships with other people or organizations that could inappropriately bias conduct and findings of this study.
ISS-AIRTUM Working Group for the study of cancer incidence in contaminated sites
M. Autelitano (Registro tumori Milano), E. Beccaloni (Istituto Superiore di Sanità), M. Benedetti (Istituto Superiore di Sanità), L. Benfatto (Registro tumori Genova), A. Biggeri (Università di Firenze), A. Binazzi (INAIL), F. Bianconi (Registro tumori Umbria), E. Bidoli (Registro tumori Friuli Venezia Giulia), C. Bruno (Istituto Superiore di Sanità), C. Buzzoni (Registro tumori Regione Toscana), G. Candela (Registro tumori Provincia di Trapani), M. Carere (Istituto Superiore di Sanità), D. Catelan (Università di Firenze), M. Cocchioni (Registro tumori infantili e degli adolescenti regione Marche), P. Comba (Istituto Superiore di Sanità), S. Conti (Istituto Superiore di Sanità), M. Corfiati (INAIL), E. Coviello (Registro tumori Barletta-Trani), L. Cremone (Registro tumori della Provincia di Salerno), E. Crocetti (Registro tumori Regione Toscana), A.P. Dei Tos (Registro tumori del Veneto), M. De Santis (Istituto Superiore di Sanità), F. Falcini (Registro tumori della Romagna), F. Falleni (Istituto Superiore di Sanità), L. Fazzo (Istituto Superiore di Sanità), M. Federico (Registro tumori della Provincia di Modena), S. Ferretti (Registro tumori della Provincia di Ferrara), M. Fusco (Registro tumori di popolazione Regione Campania), A. Giacomin (Registro tumori Piemonte, Provincia di Biella), G. Gola (Registro tumori della Provincia di Como), L. Grisotto (Università di Firenze), S. Guzzinati (Registro tumori Veneto), I. Iavarone (Istituto Superiore di Sanità), F. La Rosa (Registro tumori umbro di popolazione), L. Lillini (Registro tumori Genova), A. Madeddu (Registro tumori di Siracusa), M. Magoni (Registro tumori ASL di Brescia), L. Mangone (Registro tumori Reggiano, ASMN-IRCCS), V. Manno (Istituto Superiore di Sanità), I. Marcello (Istituto Superiore di Sanità), A. Marinaccio (INAIL), G. Marsili (Istituto Superiore di Sanità), S. Maspero (Registro tumori della Provincia di Sondrio), M. Maule (Registro tumori Infantili Piemonte), G. Mazzoleni (Registro tumori dell'Alto Adige-Tumorregister SüdTirol), F. Merletti (Registro dei tumori infantili del Piemonte), G. Minelli (Istituto Superiore di Sanità), A. Minerba (Registro tumori della Provincia di Taranto), M. Michiara (Registro tumori della Provincia di Parma), C. Nicita (ASP 7 Ragusa), F. Pannozzo (Registro tumori della Provincia di Latina), R. Pasetto (Istituto Superiore di Sanità), A. Piccardi (Istituto Superiore di Sanità), S. Piffer (Registro tumori della Provincia di Trento), R. Pirastu (Sapienza Università di Roma), P. Pisani (Registro tumori Infantili Piemonte), P. Ricci (Registro tumori Mantova), M. Santoro (Consiglio Nazionale delle Ricerche), F. Scaini (Istituto Superiore di Sanità), S. Sciacca (Registro tumori Messina-Catania), O. Sechi (Registro tumori della Provincia di Sassari), D. Serraino (Registro tumori del Friuli Venezia Giulia), M.E. Soggiu (Istituto Superiore di Sanità), F. Stracci (Registro tumori Umbria), A. Sutera Sardo (Registro tumori Catanzaro), G. Tagliabue (Registro tumori Lombardia, Provincia di Varese), F. Tisano (Registro tumori Siracusa), M. Usala (Registro tumori Nuoro), Vercelli (Registro tumori Genova), F. Vitale (Registro tumori provincia di Palermo), S. Vitarelli (Registro tumori di Macerata), P. Zambon (Registro tumori Veneto), A. Zona (Istituto Superiore di Sanità).
REFERENCES
1. World Health Organization. Contaminated sites and health. Report of two WHO workshops: Syracuse, Italy, 18 November 2011; Catania, Italy, 21-22 June 2012. Pasetto R, Martin Olmedo P, Martuzzi M (Eds). Copenhagen, Denmark, WHO Regional Office for Europe, 2013. Available from: www.euro.who.int/ data/assets/ pdf_file/0003/186240/e96843e.pdf.
2. Panagos P, Van Liedekerke M, Yigini Y, et al. Contaminated sites in Europe. Review of the current situation based on data collected through a European Network. J Environ Pub Health 2013. DOI: 10.1155/2013/158764.
3. Pirastu R, Ancona C, Iavarone I, et al. Mortality study of residents in Italian polluted sites: evaluation of the epidemiological evidence. [Article in Italian] Epidemiol Prev 2010;34(Suppl. 3). Available from: www.epiprev.it/pubblicazione/epidemiol-prev-2010-34-5-6-suppl-3.
4. Pirastu R, Pasetto R, Zona A, et al. The health profile of population living in contaminated sites: SENTIERI approach. J Environ Pub Health 2013. DOI:10.1155/2013/939267.
5. Pirastu R, Iavarone I, Pasetto R, et al. [SENTIERI Project. Mortality study of residents in Italian polluted sites: results]. Epidemiol Prev 2011;35(Suppl. 4). Available from: www.epiprev.it/pubblicazione/epidemiol-prev2011-35-5-6-suppl-4.
6. World Health Organization International Agency for Research on Cancer. Cancer incidence in five continents. Volume IX. Curado MP, Edwards B, Shin HR, et al (Eds). IARC Scientific Publications No. 160 Lyon, 2007.
7. AIRTUM and AIEOP working group. Italian cancer figures Report 2012. Cancer in children and adolescents. Epidemiol Prev 2013; 37(1 Suppl).
8. Crocetti E, De Angelis R, Buzzoni C, et al. AIRTUM Working group. Cancer prevalence in United States, Nordic Countries, Italy, Australia, and France: an analysis of geographic variability. Br J Cancer 2013;109:219-28. DOI: 10.1038/bjc.2013.311.
9. Musmeci L, Bellino M, Falleni F, et al. Environmental characterization of the National contaminated sites in SENTIERI project. [Article in Italian]. In: Pirastu R, Iavarone I, Pasetto R, Zona A, Comba P. SENTIERI Project. Mortality study of residents in Italian polluted sites: results. Epidemiol Prev 2011;35(Suppl. 4);20-3.
10. Comba P, Gianfagna A, Paoletti L. Pleural mesothelioma cases in Biancavilla are related to a new fluoro-edenite fibrous amphibole. Arch Environ Health 2003;58:229-32.
11. AIRTUM working group. Italian cancer figures Report 2009. Cancer trends (1998-2005) Epidemiol Prev 2009;33(Suppl. 1).
12. Martuzzi M, Mitis F, Pirastu R, et al. Global burden of mortality in Italian polluted sites. [Article in Italian]. In: Pirastu R, Iavarone I, Pasetto R, Zona A, Comba P. SENTIERI Project. Mortality study of residents in Italian polluted sites: results. Epidemiol Prev 2011;35(Suppl. 4):153-62.
13. Zusman M, Dubnov J, Barchana M, et al. Residential proximity to petroleum storage tanks and associated cancer risks: Double Kernel Density approach vs. zonal estimates. Sci Total Environ 2012;441:265-76. DOI: 10.1016/j.scitotenv.2012.09.054.
14. Ramis R, Fernandez-Navarro P, Garcia-Perez J, et al. Risk of cancer mortality in spanish towns lying in the vicinity of pollutant industries. ISRN Oncol 2012;2012:614198. DOI: 10.5402/2012/614198.
15. Pan SY, Morrison H, Gibbons L, et al. Canadian Cancer Registries Epidemiology Research Group. Breast cancer risk associated with residential proximity to industrial plants in Canada. J Occup Environ Med 2011;53:522-9. DOI: 10.1097/JOM.0b013e318216d0b3.
16. Ramis R, Diggle P, Cambra K, et al. Prostate cancer and industrial pollution Risk around putative focus in a multi-source scenario. Environ Int 2011;37:577-85. DOI: 10.1016/j.envint.2010.12.001
17. García-Pérez J, López-Cima MF, Boldo E, et al. Leukemia-related mortality in towns lying in the vicinity of metal production and processing installations. Environ Int 2010;36:746-53. DOI: 10.1016/j.envint.2010.05.010.
18. Gensburg LJ, Pantea C, Kielb C, et al. Cancer incidence among former Love Canal residents. Environ Health Perspect 2009;117:1263-71. DOI: 10.1289/ehp.0800153
19. Boberg E, Lessner L, Carpenter DO. The role of residence near hazardous waste sites containing benzene in the development of hematologic cancers in upstate New York. Int J Occup Med Environ Health 2011;24:327-38. DOI: 10.2478/s13382-011-0037-8
20. Pascal M, Pascal L, Bidondo ML, et al. A review of the epidemiological methods used to investigate the health impacts of air pollution around major industrial areas. J Environ Public Health 2013;2013:737926. DOI: 10.1155/2013/737926. Epub2013Jun 2.
21. Pirastu R, Comba P, Conti S. SENTIERI Epidemiological Study of Residents in National Priority Contaminated Sites: mortality, cancer incidence and hospital discharges in National Priority Contaminated Sites. Epidemiol Prev 2014;35(5-6).Suppl.4:1-204.
22. Sterne J, Davey-Smith G. Sifting the evidence-what's wrong with significance tests? BMJ 2001;322:226-31. DOI: 10.1136/bmj.322.7280.226
23. Biggeri A, Catelan D, Barbone F. Reporting and interpreting uncertainty in epidemiological studies. Epidemiol Prev 2011;35:51-52.
24. Hill Bradford A. The environment and disease: Association or causation? Proc Royal Soc Med 1965;58:295-300.
25. Willett WC. The search for truth must go beyond statistics. Epidemiology 2008;19:655-56. DOI: 10.1097/ EDE.0b013e318181b877
26. Pasetto R, Sampaolo L, Pirastu R. Measures of material and social circumstances to adjust for deprivation in small-area studies of environment and health: review and perspectives. Ann Ist Super Sanità 2010;46(2):185-97. DOI: 10.4415/ANN_10_02_13.
27. Iavarone I, Pirastu R, Minelli G, et al. Children's health in Italian polluted sites. Epidemiol Prev 2013;37(1 Suppl.):255-60.
Address for correspondence:
Pietro Comba, Dipartimento di Ambiente e Connessa Prevenzione Primaria, Istituto Superiore di Sanità, Viale Regina Elena 299, 00161 Rome, Italy.
E-mail: pietro.comba@iss.it.
Received on 13 January 2014.
Accepted on 3 April 2014.
* the Working Group members are cited before the References