RESEARCH AND METHODOLOGIES
Studi ecologici dell'incidenza tumorale in un'area interessata da siti di smaltimento di rifiuti in Campania, Italia
Lucia FazzoI; Marco De SantisI; Francesco MitisII; Marta BenedettiI; Marco MartuzziII; Pietro CombaI; Mario FuscoIII
IDipartimento di Ambiente e Connessa Prevenzione Primaria, Istituto Superiore di Sanità, Rome, Italy
IIWorld Health Organization, Centro Europeo Ambiente e Salute, Rome, Italy
IIIRegistro Tumori della Regione Campania c/o ASL Napoli 4, Brusciano, Italy
Cancer incidence was investigated in an area which has been affected by the illegal practices of dumping hazardous waste and setting fire to mismanaged waste. For the 35 municipalities of this area that are served by a Cancer Registry, municipal standardized incidence ratios (SIR) and hierarchical Bayesian estimators (BIR) were computed. Moreover, municipal spatial clustering and a Poisson regression by municipality index of waste-related exposure were performed for 10 cancer types. Increased municipality SIRs were found for some cancer types. The BIRs confirmed the increases for liver cancer in two municipalities. Statistically significant clusters were detected for liver, lung, leukaemia and soft tissue sarcomas. In the regression analysis, testis cancer showed significant trend with the index of waste-related exposure (RR = 1.18).
Key words: cancer incidence, waste dumping sites, testicular neoplasms, liver neoplasms.
È stata investigata l'incidenza oncologica in un'area interessata da pratiche illegali di smaltimento e di combustione incontrollata di rifiuti. Per i 35 comuni di quest'area serviti da un Registro Tumori sono stati calcolati i rapporti standardizzati di incidenza (standardized incidence ratios, SIR) e gli stimatori bayesiani (BIR). Per 10 sedi tumorali sono state eseguite, inoltre, analisi di cluster spaziali e di regressione con un indice municipale di esposizione a rifiuti. SIR in eccesso sono stati osservati per diverse sedi tumorali; i BIR hanno confermato gli eccessi di tumore epatico in due comuni. I tumori epatici e polmonari, le leucemie e i sarcomi dei tessuti molli hanno mostrato cluster significativi. Nell'analisi di regressione il tumore del testicolo ha mostrato un trend in eccesso significativo (RR = 1,18).
Parole chiave: incidenza tumorale, rifiuti, neoplasie del testicolo, neoplasie del fegato.
The territory of Naples and Caserta Provinces in Campania Region (Southern Italy), consisting of 196 municipalities, has been extensively affected, since the eighties, by the illegal practices of dumping hazardous waste and setting fire to both hazardous and solid urban mismanaged waste. A part of this area (77 municipalities) has been declared "site of national interest for remediation" by the Ministry of Environment, because of the presence of dumping waste sites, together with other sources of environmental pollution.
In the last decades, several studies on the possible health impact of waste management and disposal were published. A WHO Report on these issues was published in 2007; it concluded that the scientific literature provides some indication of the association between residence near a landfill site and adverse health effects, even though the evidence (stronger for reproductive outcomes than for cancer) is not conclusive. Similar conclusions were reached for incinerators, with specific reference to a possible role of dioxins in the increased risk in soft tissue sarcomas (STS) and non-Hodgkin's lymphoma (NHL) . A more recent review of epidemiological studies published between 1983 and 2008 about health effects of solid urban waste management reached comparable conclusions .
For some cancer types a priori hypotheses of an association with residence near waste management sites appear to be more consistent, even if with the limitations indicated by the above-mentioned reviews. Associations with residence in the neighbourhood of landfill sites have been suggested for following cancer types: liver, lung, stomach, kidney, bladder and leukaemia [3-9].
In light of the widespread practice of setting fire to waste deposits in parts of the study area and the subsequent detection of high levels of dioxins in animal milk , even if no waste incinerator was operating in the area, literature about the health impact of incinerators was considered. Increased risks for liver and lung cancer, NHL and STS were reported in the neighbourhood of incinerators .
Even if there is not direct epidemiological evidence of an increase risk of testis cancer in areas with waste dumping sites, testis cancer was included in the study. The a priori etiological hypotheses were derived by awareness of the possible etiological role of exposure to endocrine disruptors . The endocrine disruptors are also taken into account in the etiology of congenital malformations  and in the assessment of congenital malformations risk around landfills . Similarly, biliary tract cancer was included for its suspected association with exposure to endocrine disruptors .
With regard to the territory of Naples and Caserta provinces, since 2006 several studies have been carried out on cause-specific mortality and prevalence of congenital malformations in this area, in order to investigate the health status of the population and its potential relationship with the presence of dumping waste sites. The first study showed increased risks in cancer mortality in the southern part of Caserta province and in the northern part of Naples province . Subsequently, municipal cluster analysis of cancer mortality and congenital malformations occurrence showed significant clusters of municipalities located in the same areas evidenced in the previous study . More recently, a study has been published on the correlation between cancer mortality, prevalence of congenital anomalies at birth and a municipal waste-related exposure index . The findings, adjusted for socioeconomic deprivation, show statistically significant increased relative risks in high exposure compared to low exposure index municipalities for all causes, all cancer and liver cancer mortality in both genders, stomach cancer and lung cancer in men, and for the occurrence of congenital anomalies of the internal urogenital tract and of the central nervous system.
The aim of the present study is to describe the spatial distribution of neoplastic disease incidence in this area and to generate hypotheses of etiological interest with respect to the aforementioned waste-related exposures, that might subsequently be tested by ad hoc analytical studies. Cancer incidence is in fact a more valid outcome than mortality, because it is based on pathology data and it is not affected by differential survival patterns reflecting differences in access to appropriate diagnostic and therapeutic procedures. Furthermore, the use of incidence data enables the study of neoplastic diseases characterized by low mortality rates. The present study was prompted by the availability of cancer incidence data in a subarea of the Province of Naples, comprising 35 municipalities, that is served by a cancer registry (Figure 1).
MATERIALS AND METHODS
The study area is constituted by the municipalities of Naples Province, served by a Cancer Registry; the time-window of analysis is from 1997 to 2005, corresponding to that of the Registry's time coverage.
The Cancer Registry of Local Health Authority "Naples 4", the only cancer registry operating in Caserta and Naples Provinces, constituted the source of the study subjects. The Registry is part of the Italian Network of Cancer Registries (AIRTUM) and is certified by International Agency for Research on Cancer (IARC) at the international level .
The Cancer Registry database includes 17 300 cases of cancer diagnosed in the period 1997-2005 and resident at the time of diagnosis in the 35 municipalities of area of the Local Health Authority "Naples 4". According to the coding rules adopted in cancer registration, International Classification Diseases for Oncology 3rd edition (ICD-O-3) codes based on topographic criteria were used. The SIR and BIR analyses included 14 892 cases diagnosed in the period 1997-2005 with the following 25 cancer types: head and neck, esophagus, stomach, colon-rectum, liver, biliary ducts, pancreas, larynx, lung, skin (melanoma), mesothelioma, breast, uterine cervix, uterus, prostate, testis, kidney, bladder, brain, thyroid, Hodgkin's lymphoma, non Hodgkin's lymphoma, myeloma, leukaemia and malignant tumours of the connective and soft tissue (the category of the topographical classification in which most of the soft-tissue sarcomas are included); analyses were also performed for all cancers (17 300 cases) and all cancers excluding skin (16 156 cases).
Given the specific interest of investigating soft-tissue sarcomas (STS), that have been associated with exposure to dioxins , possibly released by uncontrolled combustions, an additional analysis based on morphological classification was performed. This analysis was possible because all cancer cases considered by the Registry were diagnosed by pathology laboratories and histological examinations of primary tumour were available; subsequently Cancer Registry experts checked each report with respect to compliance with morphological classification in ICD 10th revision code. For adult STS (age > 19 years), histologically-confirmed first-incident STS cases, originating in non-visceral and visceral locations, were enrolled in the study. Inclusion criteria were based on the histological classification of soft tissue tumours of Weiss & Goldblum (2007) , very similar to the WHO classification (2002) . Kaposi's sarcoma were excluded. For childhood (age 0-14 years) and adolescent (age 15-19 years) STS cases were those with a first-incident histologically confirmed diagnosis of malignant STS, originating in non-visceral and visceral locations. We selected all malignant types included in the IX site group (soft tissue and other extraosseous sarcomas) of the third edition of the International Classification of Childhood Cancer (ICCC-3), excluding benign/intermediate types and Kaposi sarcoma.
The total resident population in the study area is constituted by 538 243 people (at 2001 census), on a surface of 425 kmq. The most populated municipality includes 47 940 inhabitants and the smallest 1769 subjects (at 2001 census). For each municipality, a database of cases specific by cancer type, gender and age (five-year classes) was prepared.
Standardized incidence ratios (SIRs) and hierarchical Bayesian estimators (BIRs) were computed for each municipality, using the whole population of the 35 municipalities area as reference. SIRs with their 95% confidence intervals were computed using the quadratic approximation to the Poisson log-likelihood for the log rate parameters, in STATA software package. BIRs, bayesian smoothed relative risks, with their 95% credibility intervals were computed with the Rapid Inquiry Facility (RIF), using WinBugs [21, 22] based on the procedures by Besag et al, . For STS morphologic classification only, 90% confidence intervals were used, in order to balance the fairly low statistical power due to small sample, also in view of the relatively stronger available a priori hypotheses.
Clustering and regression analyses were performed only for 10 of all the 25 selected cancer types, for which a priori hypotheses of an association with residence near waste dumping sites were available (stomach, liver, biliary ducts, lung, soft-tissue sarcomas, kidney, bladder, non Hodgkin's lymphoma, leukaemia, testis) and for all cancers. The selection of these cancer types was based on the examination of abovementioned systematic reviews and other recent sources [1, 24, 2].
Cluster analysis was performed in order to investigate spatial aggregation of cases resident in neighbouring municipalities in the whole study area. Municipal spatial cluster of cases were identified by using SatScan software (version 6) assuming a Poisson model for the distribution of cases in each municipality. The adopted procedure spatial scan statistics  employs a circular window whose radius was fixed with a maximum of 10 km and was centred at each step on one municipality identified by x, y coordinates of its townhall. We selected a significance threshold of p < 0.10. The underlying etiological hypothesis was that of a diffuse air, soil and water contamination in some subareas of the territory included in the study.
Clustering analysis took into account socioeconomic status. We calculated municipal deprivation index (DI) based on five 1991 census variables (education, unemployment, housing ownership, surface of the dwelling and family structure) . The municipalities were subdivided in quintiles with respect to the value of deprivation index in the study area (1 = least deprived; 5 = most deprived); the value of DI quintile of each municipality is used in the analysis as covariate, and the findings are adjusted by DI. The use of DI by index classes rather than using the basic index is commonly endorsed in environmental epidemiology , and has previously been adopted in this study setting [15, 16].
In order to evaluate the possible association between cancer incidence and environmental exposure to waste dumping sites, we used Poisson regression method (in STATA software package) applying to each municipality a waste-related exposure index (WEI). The latter was built within the previous mortality study  and it is described in detail in a recent paper .
The WEI was constructed on the basis of a database of legal waste landfills and illegal dumping sites present in the region (years 1997-2003), created by the regional Environmental Protection Agency. Many of these sites have been present for at least two decades, and have not yet been cleaned up . When the WEI was calculated, routinely environmental monitoring data in the area of interest were not geocoded. The municipal WEI took into account the potential hazard of each dumping waste site and the percentage of population resident in impact areas of the single waste site .
In order to use the index in regression analysis, municipalities were categorized in five groups of WEI, using the so-called natural breaks, that maximize homogeneity within groups and heterogeneity between groups (1 = least at risk; 5 = most at risk).
With regard to the area of the present study, 16.5% of it, comprising 5.5% of resident population (total population, at 2001 census: 29 434 people), is interested by waste impact areas. Table 1 shows the number of impact areas, interested territory and population in each municipality. The distribution of 35 municipalities in the different classes of DI and WEI is shown in Figure 2. Only one municipality is in the fourth and fifth WEI classes respectively. This distribution creates some limitations in the computation of correlation coefficients, which was taken into account in the interpretation of findings. Poisson regression analyses to investigate WEI effects were adjusted by DI.
The annual average population and total cases for cancer types in the 1997-2005 period, by gender and by classes of deprivation index and waste exposure index are shown in Table 2.
SIR and BIR
Table 3 shows the number of municipalities where SIRs showed 95% confidence interval lower limit superior than 1. Only for morphologically classified soft-tissue sarcomas, we used 90% confidence interval.
Significantly increased SIRs were found for liver and lung cancer, non Hodgkin's lymphoma (NHL) and total leukaemia (ICD 10 rev: C91-95, including NAS leukaemia), in both genders separately; testis, esophagus, larynx, pancreas, thyroid gland cancer and morphologic soft-tissue sarcomas in men; stomach and topographic soft tissue sarcomas in the overall population; kidney, biliary ducts, brain cancer and myeloma in women. Only the increases for liver cancer were confirmed in two municipalities by BIRs with 95% credibility intervals lower limit superior than 1 (observed cases: 112, BIR = 1.41, 95% CrI = 1.17-1.68; observed cases: 100, BIR = 1.57; 95% CrI = 1.28-1.89) (findings not shown).
Statistically significant clusters were detected in the total population for liver cancer (observed cases: 191, RR = 1.64, p-value = 0.0003; observed cases: 73, RR = 1.70, p-value = 0.0027), lung cancer (observed cases: 737, RR = 1.15, p-value = 0.08), total leukaemia (observed cases: 197, RR = 1.33, p-value = 0.05) and topographic soft tissue sarcomas (observed cases: 49, RR = 2.02, p-value = 0.08) (Figures 3-6). Statistically significant clusters were detected for liver cancer also among men and women separately: two clusters among men (observed cases: 127, RR = 1.64, p-value = 0.0007; observed cases: 62, RR = 1.60, p-value = 0.0470, respectively) and one cluster in women (observed cases: 63, RR = 1.684, p-value = 0.0213) (data not shown).
The two significant clusters detected for liver cancer in the total population substantially correspond to those detected among men; one of them is significant also among women. These clusters are located in the northern part of the study area, but they are not contiguous.
The leukaemia cluster included one municipality in the WEI 4th category; the only municipality of the WEI 5th category was not included in any cluster.
Other cancer types didn't show significant clusters among two genders separately.
The findings of correlation analysis are shown in Table 4.
Relative risks of each WEI group with respect to the first WEI group are reported. The linear risk trends across the five WEI groups are also shown. The results with 95% CI lower limit > 1 are reported in bold.
Only testis cancer incidence showed a statistically significant increasing trend with the WEI categories (RR = 1.18; 95% CI: 1.03-1.35).
Significantly increased relative risks for some cancer types occurred in specific WEI categories, with respect to the first class, used as reference: for lung cancer (RR = 1.44; 1.01-2.05) and bladder cancer (RR = 1.85; 1.13-3.03) in the fourth class and for the latter cancer type also in the second class (RR = 1.19; 1.01-1.41) in men; biliary ducts cancer in the third class (RR = 1.63; 1.00-2.63) and leukaemia in the fifth class (RR = 1.82; 1.08-3.07) in women. In women, the relative risk for liver cancer was found significantly decreased (RR = 0.68; 0.53-0.86) in the second class.
The study findings warrant some comments.
Liver cancer showed a significant increase in risk in terms of SIR, BIR and cluster analysis in a group of municipalities located in the northern part of the study area. Three of them also showed significantly increased mortality from liver cancer in the previous study . In the regression analysis, a positive trend across the five WEI groups was found, but statistical significance was not reached, while it had been obtained in the mortality analysis. Liver cancer is a multifactorial disease, whose most important etiological factor is infection by B and C hepatitis virus, which has been shown to be endemic in the area , but does not seem to completely explain the geographic distribution of liver cancer incidence. Alcohol consumption is also a well ascertained risk factor , while the role of socioeconomic conditions is controversial . The possibility of an interaction between hepatitis virus and hepatotoxic agents in liver carcinogenesis has also been raised, in particular if these chemicals have a fibrogenic action, as vinyl chloride .
Taking into account the consistent a priori hypothesis of a possible association between liver cancer and waste disposal and the findings of the several studies in the area of interest, specific ad hoc studies in this area are warranted.
Two neighbouring municipalities showed significant results for SIR and clustering with regards to lung cancer, one of them also showed an increased incidence of pleural mesothelioma. Lung cancer showed a positive trend in males in correlation analysis, also if not statistically significant, while significance was reached in the mortality study in males. Lung cancer has a well known multifactorial etiology; in this case the presence of an asbestos cement factory in the town with excess incidence of both lung cancer and mesothelioma points to a causal role of occupational asbestos exposure .
A significant cluster of leukaemia in total population and a non significant correlation with WEI in women were detected; the latter was partly due to a significant IRR in the highest score WEI category. These findings reflect the role of chance or different constellations of risk factors in the two genders, in light of the multifactorial etiology of this disease.
The present study detected a significant cluster of soft-tissue sarcoma (topographic classification), in the absence of correlation with WEI. The lack of correlation with WEI, together with the observation that the STS cluster is located in an area with high prevalence of greenhouse floriculture, may rather point to an etiologic role of herbicides and fungicides [35, 36].
The finding of a positive correlation between the waste-related exposure index (WEI) and the incidence of testicular cancer deserves some comments. The possible etiologic role of environmental exposure to endocrine disruptors has been suggested in some review articles, taking into account the multifactorial etiology of this disease [37, 11, 38]. The suspected association between endocrine disruptors and testicular cancer may be viewed within the frame of the Testicular Dysgenesis Syndrome, whose underlying cause might reflect a hormonal imbalance due to environmental or life-style factors during early foetal development [39, 40]. The possibility of an etiological role of endocrine disruptors ingested with milk during puberty was also recently raised .
In the present study, one municipality showed a significantly increased incidence of testicular cancer (based on 19 cases; SIR = 1.75, 95% CI = 1.05-2.73), and it contributed to the results of the regression analysis since it corresponded to the highest score of the waste-related exposure index. In the previous study a significant increase of congenital malformations of urogenital tract was detected in the same municipality . An ad hoc analytical study is required in order to evaluate a possible environmental etiology of testicular cancer in this area and a possible role of in utero exposure.
As mentioned in the Introduction section, the study area is characterized by a major problem of hazardous waste dumping, but also by the widespread practice of setting fire to both urban and hazardous waste. In this context, the a priori hypotheses of etiological interest have been derived from the available evidence concerning the possible health effects associated with residence in the neighbourhood of both landfills and incinerators. The complexity of the setting thus impairs full comparability of the findings of the present study with those of most previously published investigations.
Besides the comment of the study findings, some considerations about validity issues are required.
The data provided by the Cancer Registry have been extensively validated prior to their inclusion in the database endorsed by the International Agency for Research on Cancer .
Several methods of spatial analysis were applied. SIR analysis was characterized by high sensitivity and detected a number of significant increases of different cancer types. A subgroup of these signals was evidenced by cluster analysis, while BIR analysis was more specific and less sensitive with respect to both SIR and cluster analysis, since expected figures are computed based on the average values of contiguous municipalities.
Two main limitations affect the waste-related exposure index (WEI).
The first limitation derives from the inherent difficulty in the enumeration of the dumping sites distributed in the study area, often not authorized if not openly illegal; furthermore, the abovementioned illegal practice of setting fire to urban and hazardous waste has not been adequately monitored and mapped. Although the hazard characterization procedure of dumping sites has certainly been hampered by lack of information on the complex mix of chemicals released in the environment, still landfill density may be seen as an indirect indication of exposure to landfill emissions when a more detailed modelling is precluded . The index was developed in order to be used at the municipality level, subsuming an ecological approach, and it cannot, thus, predict exposure at individual level.
The second limitation regards the difference between the context in which the WEI was developed (196 municipalities encompassing two provinces) and the setting of the present study (35 municipalities located in a subarea of the Naples province). The five categories of the score were based on a "natural break" approach in the previous mortality study . In the present incidence study, based on 35 municipalities, only two of them have been allocated, respectively, to the fourth and fifth WEI category, thus hampering the regression approach. The regression analysis was anyhow performed in order to foster comparability with the previous mortality study, but the smaller number of municipalities included in the present investigation, that affected contrastability between WEI categories, and the resulting lower statistical power, determined a modest consistency between the two correlation studies.
As far as control of confounding is concerned, we took into account the socio-economic status and the results of cluster and regression analyses were adjusted by the municipal Deprivation Index, but a residual effect cannot be ruled out. We did not consider other possible confounding factors, as occupational exposures, lifestyle, tobacco consumption, that operate at the individual level, since the present study adopted an ecological design.
The area at study is not very large (424.48 km2) but it is densely populated (538 443 people, at 2001 census) and presents a variety of occupational settings, including chemical and metal industries, quarries and rural areas with intensive agricultural activity and greenhouse productions. The possible contribution of the resulting environmental contamination to the cancer burden in the area remains at the moment an open question.
The issue of multiple testing should also be addressed in this frame. There is general agreement on the risk of generating false-positive results in geographical analyses  and surveillance projects  because of multiple testing of a same null hypothesis. It should be stressed, though, that no irreversible decision follows the observation of a statistically significant increase of the incidence of a given cancer in a particular municipality; the study findings, as previously stated, are essentially used to select cancer types and locations for subsequent analytical investigations.
The present study adopted an ecological approach, that is not apt to conclusively corroborate or refute specific etiologic hypotheses. Still, the study, based on a validated cancer registry data-base and employing an appropriate set of methodological procedures for spatial analysis, detected several signals requiring an ad hoc analytical investigation.
In particular, the study findings suggested an association between the aforementioned waste-related exposure index and the incidence of testicular cancer, which might require investigation by an analytical study. Clusters of liver cancer, leukaemia and soft-tissue sarcomas were detected in the study area, but due to the absence of a clear association with the waste-related exposure index and to poor knowledge of locally relevant occupational exposures, it is difficult to draw conclusions regarding possible associations.
The authors wish to thank the Working Group of Cancer Registry of Campania Region, Caterina Bellatalla, Rosaria Di Napoli, Maria Fusco, Margherita Panico, Carmela Perrotta and Biagio Vassante, for their contribution in collection and acquisition data.
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.
1. Mitis F, Martuzzi M (Ed.). Population health and waste management: scientific data and policy options. Report of a WHO workshop. Rome, Italy, 29-30 March 2007. Copenhagen: WHO Regional Office for Europe Publications; 2007.
2. Porta D, Milani S, Lazzarino AI, Perucci CA, Forastiere F. Systematic review of epidemiological studies on health effects associated with management of municipal solid waste. Environ Health 2009;8:60.
3. Goldberg MS, Alhomsi N, Goulet L, Riberdy H. Incidence of cancer among persons living near a municipal solid waste landfill site in Montreal, Quebec. Arch Environ Health 1995;50:416-24.
4. Goldberg MS, Siematycki J, Dewar R, Désy M, Riberdy H. Risks of developing cancer relative to living near a municipal solid waste landfill site in Montreal, Quebec, Canada. Arch Environ Health 1999;54:291-6.
5. Griffith J, Duncan RC, Riggan WB, Pellom AC. Cancer mortality in US counties with hazardous waste sites and ground water pollution. Arch Environ Health 1989;44:69-74.
6. Budnick LD, Logue JN, Sokal DC, Fox JM, Falk H. Cancer and birth defects near the Drake Superfund site, Pennsylvania. Arch Environ Health 1984;39:409-13.
7. Mallin K. Investigation of a bladder cancer cluster in northwestern Illinois. Am J Epidemiol 1990;132(1):S96-S106.
8. Lewis-Michl EL, Kallenbach LR, Geary NS, Melius JM, Ju CL, Orr MF, Forand SP. Investigation of cancer incidence and residence near 38 landfills with soil gas migration conditions: New York State, 1980-1989. ATSDR/HS-98-93. Atlanta: Agency for Toxic Substances and Disease Registry, 1998.
9. Cutler JJ, Parker GS, Rosen S, Prenney B, Healey R, Caldwell GG. Childhood leukaemia in Woburn, Massachusetts. Public Health Rep 1986;101:201-5.
10. Esposito M, Cavallo M, Serpe FP, D'Ambrosio R, Gallo P, Colarusso G, Pellicanò R, Baldi L, Guarino A, Serpe L. Levels and congener profiles of polychlorinated dibenzo-p-dioxins, polychlorinated dibenzofurans and dioxin-like polychlorinated biphenyls in cow's milk collected in Campania, Italy. Chemosphere 2009;77:1212-6.
11. Sarma AV, McLaughlin JC, Schottenfeld D. Testicular cancer. In: Schottenfeld D & Fraumeni JF Cancer Epidemiology and Prevention, 3. ed. Oxford: Oxford University Press; 2006. p. 1151-65.
12. Landrigan P, Garg A, Droller DBJ. Assessing the effects of endocrine disruptors in the national children's study. Environ Health Perspect 2003;111:1678-82.
13. Ahrens W, Mambetova C, Bourdon-Raverdy N, Llopis-Gonzalez A, Guénel P, Hardell L, Merletti F, Morales-Suàrez-Varela M, Olsen J, Olsson H, Vyberg M, Zambon P. Occupational exposure to endocrine-disrupting compounds and biliary tract cancer among men. Scand. J Work Environ Health 2007;33(5):387-96.
14. Comba P, Bianchi F, Fazzo L, Martina L, Menegozzo M, Minichilli F, Mitis F, Musmeci L, Pizzuti R, Santoro M, Trinca S, Martuzzi M and "Health Impact of waste management Campania" Working Group. Cancer mortality in an area of Campania (Italy) characterized by multiple toxic dumping sites. Ann NY Acad Sci 2006;1076:449-61.
15. Fazzo L, Belli S, Minichilli F, Mitis F, Santoro M, Martina L, Pizzuti R, Comba P, Martuzzi M, Bianchi F and the Working Group. Cluster analysis of mortality and malformations in the Provinces of Naples and Caserta (Campania Region). Ann Ist Super Sanità 2008;44(1):99-111.
16. Martuzzi M, Mitis F, Bianchi F, Minichilli F, Comba P, Fazzo L. Cancer mortality and congenital anomalies in a region of Italy with intense environmental pressure due to waste. Occup Environ Med 2009;66:725-32.
17. Curado MP, Edwards B, Shin HR, Ferley J, Heanue M, Boyle P, Storm H (Ed.). Cancer incidence in five continents. Vol. IX. Lyon: Scientific Publication n. 160; 2009. IARC.
18. International Agency for Research on Cancer. IARC Monographs on the evaluation of carcinogenic risks to human, 69. Polychlorinated Dibenzo-para-dioxins and polychlorinated Dibenzofurans. Lyon: IARC Monographs; 1997.
19. Weiss SW & Goldblum (Ed.). Enzinger and Weiss's soft tissue tumors. St Louis London Philadelphia Sydney Toronto: The CV Mosby Company; 2007.
20. World Health Organization Classification of Tumours. Fletcher C, Unni KK, Mertens F (Ed.). Pathology and genetics of tumours of soft tissues and bone. Lyon: WHO IARC Press; 2002.
21. Beale L, Abellan JJ, Hodgson S, Jarup L. Methodologic issues and approaches to spatial epidemiology. Environ Health Perspect 2008;116:1105-10.
22. Beale L, Hodgson S, Abellan JJ, LeFevre S, Jarup L. Evaluation of spatial relationship between health and environment: the rapid inquiry facility. Environ Health Perspect 2010;118:1306-12.
23. Besag J, York J, Mollié A. Bayesian image restoration, with two applications in spatial statistics. Annals of the Institute of Statistical Mathematics 1991;43:1-59.
24. Associazione Italiana di Epidemiologia. Waste processing and health. A position document of the Italian Association of Epidemiology (AIE) - May 2008. Ann Ist Super Sanità 2008;44(3):301-6.
25. Kuldorff M. A spatial scan statistic. Communications in statistics: theory and methods 1997;26:481-96.
26. Cadum E, Costa G, Biggeri A, Martuzzi M. Deprivation and mortality: a deprivation index suitable for geographical analysis of inequalities. Epidemiol Prev 1999;23:175-87.
27. Elliott P, Briggs D, Morris S, de Hoog C, Hurt C, Kold Jensen T, Maitland I, Richardson S, Wakefield J, Jarup L. Risk of adverse birth outcomes in populations living near landfill sites. BMJ 2001;323:363-8. Erratum in: BMJ 2001;323:1182.
28. Musmeci L, Bellino M, Cicero MR, Falleni F, Piccardi A, Trinca S. The impact measure of solid waste management on health: the hazard index. Ann Ist Super Sanità 2010;46(3):293-8.
29. Legambiente-Osservatorio Ambiente e legalità. Rapporto Ecomafia 2009. Le storie e i numeri della criminalità ambientale. Milano: Edizioni Ambiente; 2009.
30. Fusco M, Girardi E, Piselli P, Palombino R, Polesel J, Maione C, Scognamiglio P, Pisanti FA, Solmone M, Di Cicco P, Ippolito G, Franceschi S, Serraino D. Epidemiology of viral hepatitis infections in an area of southern Italy with high incidence rates of liver cancer. Eur J Cancer 2008;44(6):847-53.
31. Baan R, Straif K, Grosse Y, Secretan B, El Ghissassi F, Bouvard V, Altieri A, Cogliano V. Carcinogenicity of alcoholic beverages. Lancet Oncol 2007;8:292-3.
32. Faggiano E, Partanen T, Kogevinas M, Moffetta P. Socioeconomic differences in cancer incidence and mortality. In: Kogevinas M, Pearce N, Susser M, Boffetta P (Ed.). Social inequalities and cancer. Lyon: International Agency for Research on Cancer; 1997. IARC Scientific Publications n. 138.
33. Mastrangelo G, Fedeli U, Fadda E, Valentini F, Agnesi R, Magarotto G, Marchì T, Buda A, Pinzani M, Martines D. Increased risk of hepatocellular carcinoma and liver cirrhosis in vinyl chloride workers: synergistic effetcs of occupational exposure with alcohol intake. Environ Health Perspect 2004;112:1188-92.
34. Straif K, Bernbrahim-Tallaa L, Baan R, Grasse Y, Secretan B, El Ghissasi F, Bouvard V, Guha N, Freeman C, Galichet L, Cogliano V. A review of human carcinogens-Part C: metals, arsenic, dusts and fibres. The Lancet Oncol 2009;10:453-5.
35. Lynge E. Cancer incidence in Danish phenoxy herbicides workers, 1947-1993. Environ Health Perspect 1998;106(Suppl.2):683-8.
36. Kogevinas M, Becher H, Benn T, Bertazzi PA, Boffetta P, Bueno-de-Mesquita HB, Coggon D, Colin D, Flesch-Janys D, Fingerhut M, Green L, Kauppinen T, Littorin M, Lynge E, Mathews JD, Newberger M, Pearce N, Saracci R. Cancer mortality in workers exposed to phenoxy herbicides, chlorophenols and dioxins: an expanded and update international cohort study. Am J Epid 1997;45(2):1061-75.
37. Garner MJ, Turner MC, Ghadirian P, Krewski D. Epidemiology of testicular cancer: an overview. Int J Cancer 2005:116:331-9.
38. Garner MJ, Turner MC, Ghadirian P, Krewski D, Wade M. Testicular cancer and hormonally active agents. J Toxicol Environm Health 2008;11:260-75.
39. Skakkebaek NE, Rajpert-De Meyts E, Main KM. Testicular dysgenesis syndrome: an increasingly common developmental disorder with environmental aspects. Human Reproduction 2001;16(5):972-8.
40. Martin OV, Shialis T, Lester JN, Scrimshaw MD, Boobis AR, Voulvoulis N. Testicular dysgenesis syndrome and the estrogen hypothesis: a quantitative meta-analysis. Environ Health Perspect 2008;116:149-57.
41. Richiardi L, Pettersson A, Akre O. Genetic and environmental risk factors for testicular cancer. Int J Andrology 2007;30:230-41.
42. Elliott P, Richardson S, Abellan JJ, Thomson A, de Hoogh C, Jarup L. and Briggs DJ. Geographic density of landfill sites and risk of congenitals anomalies in England: authors' response. Occup Environ Med 2009;66(2):140.
43. Elliott P, Wakefield LC, Best NG, Briggs DJ. Spatial epidemiology: methods and applications. Oxford: Oxford University Press; 2000.
44. Rolka H, Burkom H, Cooper G, Kulldorff M, Madigan D, Wong WK. Issues in applied statistics for public health bioterrorism surveillance using multiple data streams: research need. Stat Med 2007;26:1834-56.
Received on 3 January 2011.
Accepted on 14 March 2011.