Article Text

Download PDFPDF

Original article
Lymphoma risk and occupational exposure to pesticides: results of the Epilymph study
  1. Pierluigi Cocco1,
  2. Giannina Satta1,
  3. Stefania Dubois1,
  4. Claudia Pili1,
  5. Michela Pilleri1,
  6. Mariagrazia Zucca2,
  7. Andrea Martine ‘t Mannetje3,
  8. Nikolaus Becker4,
  9. Yolanda Benavente5,
  10. Silvia de Sanjosé5,13,
  11. Lenka Foretova6,
  12. Anthony Staines7,
  13. Marc Maynadié8,
  14. Alexandra Nieters9,
  15. Paul Brennan10,
  16. Lucia Miligi11,
  17. Maria Grazia Ennas2,
  18. Paolo Boffetta12,14
  1. 1Department of Public Health, Clinical and Molecular Medicine, Occupational Health Section, University of Cagliari, Monserrato, Italy
  2. 2Department of Biomedical Sciences, University of Cagliari, Monserrato, Italy
  3. 3Centre for Public Health Research, Massey University, Wellington, New Zealand
  4. 4German Cancer Research Center, Heidelberg, Germany
  5. 5Unit of Infections and Cancer, Cancer Epidemiology Research Program, Catalan Institute of Oncology, Hospitalet de Llobregat, Spain
  6. 6Department of Cancer Epidemiology and Genetics, Masaryk Memorial Cancer Institute, Brno, Czech Republic
  7. 7School of Nursing and Human Sciences, Dublin City University, Dublin, Ireland
  8. 8Dijon University Hospital, Dijon, France
  9. 9Centre of Chronic Immunodeficiency, University of Freiburg, Freiburg, Germany
  10. 10International Agency for Research on Cancer, Lyon, France
  11. 11ISPO Cancer Prevention and Research Institute, Florence, Italy
  12. 12The Tisch Cancer Institute and Institute for Translational Epidemiology, Mount Sinai School of Medicine, New York, New York, USA
  13. 13Centro de Investigación Biomédica en Red de Epidemiología y Salud Pública (CIBERESP), Spain
  14. 14International Prevention Research Institute, Lyon, France
  1. Correspondence to Professor Pierluigi Cocco, Department of Public Health, Occupational Health Section, University of Cagliari, Asse Didattico – Policlinico Universitario, SS 554, km 4,500, 09042 Monserrato (Cagliari), Italy; coccop{at}medicina.unica.it

Abstract

Objectives We investigated the role of occupational exposure to specific groups of agrochemicals in the aetiology of lymphoma overall, B cell lymphoma and its most prevalent subtypes.

Methods In 1998–2003, 2348 incident lymphoma cases and 2462 controls were recruited to the EPILYMPH case-control study in six European countries. A detailed occupational history was collected in cases and controls. Job modules were applied for farm work including specific questions on type of crop, farm size, pests being treated, type and schedule of pesticide use. In each study centre, industrial hygienists and occupational experts assessed exposure to specific groups of pesticides and individual compounds with the aid of agronomists. We calculated the OR and its 95% CI associated with lymphoma and the most prevalent lymphoma subtypes with unconditional logistic regression, adjusting for age, gender, education and centre.

Results Risk of lymphoma overall, and B cell lymphoma was not elevated, and risk of chronic lymphocytic leukaemia (CLL) was elevated amongst those ever exposed to inorganic (OR=1.6, 95% CI 1.0 to 2.5) and organic pesticides (OR=1.5, 95% CI 1.0 to 2.1). CLL risk was highest amongst those ever exposed to organophosphates (OR=2.7, 95% CI 1.2 to 6.0). Restricting the analysis to subjects most likely exposed, no association was observed between pesticide use and risk of B cell lymphoma.

Conclusions Our results provide limited support to the hypothesis of an increase in risk of specific lymphoma subtypes associated with exposure to pesticides.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

What this paper adds

  • Inconsistent opinions exist about the evidence linking occupational exposure to pesticides with lymphoma risk.

  • The complex array of chemicals comprised in the pesticide definition and the heterogeneity of the pathological diagnoses included in the lymphoma or non-Hodgkin's lymphoma definitions might contribute to the controversy.

  • We used the WHO classification of lymphoma to identify specific lymphoma entities, and state of the art retrospective exposure assessment for occupational exposure to chemical classes of pesticides and specific agrochemicals in a population-based case-control study.

  • Our results provide limited evidence of an increase in risk of chronic lymphocytic leukaemia associated with exposure to organophosphates, and no association for other lymphoma subtypes.

Introduction

Among hundreds of agents and groups of agents examined in 35 years of International Agency for Research on Cancer (IARC) Monographs (volumes 1–99)1 pesticides account for two dozens; only a few of those are still in use worldwide, some are obsolete but still in use in developing countries, and most have been banned or abandoned for some decades. Only arsenic and arsenical pesticides are group 1 human carcinogens, while occupational exposure in the spraying and application of non-arsenical insecticides overall is included in group 2A, because of limited evidence from epidemiological studies. Group 2A also includes two active ingredients, namely the fungicide captafol, which uses have been restricted in the USA and most world countries from 1999,2 and ethylene dibromide, which is used as a grain fumigant. As for the rest, the insufficient evidence from human studies is coupled with the sufficient, limited or unavailable evidence from experimental animal studies. Nowadays, thousands of chemicals are available to farmers to treat plant diseases and protect their crops; their use changes year by year, across countries and within each country, and by type of crop and type of disease being treated: the difficulty of conducting epidemiological studies of the long term effects of agrochemicals is reflected in the poor information on their human carcinogenicity and the absence of evaluation by international scientific and regulatory agencies.

Reviews of the scientific literature reported inconsistent opinions about the association between occupational exposure to pesticides and non-Hodgkin's lymphoma (NHL).3 ,4 In fact, while several meta-analyses have come to positive conclusions on NHL risk,5–10 particularly for prolonged exposures,11–13 or for exposure in the years relatively close to the diagnosis,14 risk has been shown to vary by gender,15 or specific jobs,16 and by specific chemicals.17 Besides, the causal link is not always recognised,18 and negative studies have also been published.19–24 In some instances, interpretation of findings is limited by imprecise definition of either exposure or disease entity22 or a small study size.23 Geographical variation in NHL mortality has also been reported in relation to the prevalent type of crop, and therefore the pesticide used:25–28 for instance, NHL mortality in the female population was elevated in an area of Minnesota where wheat, corn and soy crops were prevalent.28

Methods

The EPILYMPH study, a multicentre case-control study on environmental exposures and lymphoid neoplasms, was conducted in Czech Republic, France, Germany, Italy, Ireland and Spain from 1998 to 2004. Details about the study have been described elsewhere.29 Briefly, cases were all consecutive adult patients first diagnosed with lymphoma during the study period, resident in the referral area of the participating centres. The diagnosis was classified according to the 2001 WHO classification of lymphoma,30 and slides of about 20% of cases from each centre were reviewed centrally by a panel of pathologists, coordinated by MM. Controls from Germany and Italy were randomly selected by sampling from the general population, matched to cases on gender, 5-year age-group, and residence area. The rest of the centres used matched hospital controls, with eligibility criteria limited to diagnoses other than cancer, infectious diseases and immunodeficient diseases. Approval by the relevant Ethics Committees was obtained in all centres. Informed consent was obtained for the 2348 lymphoma cases and 2462 controls who participated to the study. Overall, the participation rate was 88% in cases, 81% in hospital controls and 52% in population controls.

Trained interviewers conducted in person interviews with cases and controls, using the same structured questionnaire translated into the local language. Questions sought information on sociodemographic factors, lifestyle, health history and a list of all full time jobs held for 1 year or longer. Industrial hygienists in each participating centre coded the occupations and industries using the 5-digit 1968 International Labour Office International Standard Classification of Occupations31 and the 4-digit codes of the 1996 European Statistical Classification of Economic Activities, revision 1 (NACE, rev. 1).32 Study subjects who reported having worked in agriculture were given a job-specific module inquiring in detail into the following: detailed description of the tasks; kind of the crops and size of the cultivated area; type of pests being treated; pesticides used, and procedures of crop treatment; use of personal protective equipment; re-entry after treatment; frequency of the treatment in days/year.

Occupational exposure assessment

With the support of a local agronomist, and the support of a crop-exposure matrix, created by LM, to supplement the available information, industrial hygienists and occupational experts in each participating centre reviewed the general questionnaires and job modules to assess exposure to pesticides classified into inorganic (mainly sulphur and arsenic salts) and organic (carbamates, organophosphates, organochlorines, triazines and triazoles, phenoxyacids, and chlorophenols). Exposure was classified according to the following exposure metrics:

  • confidence, representing the industrial hygienist's degree of certainty that the worker had been truly exposed to the agent, based upon two criteria: 1. a summary evaluation of the probability of the given exposure (1= possible, but not probable; 2=probable; and 3=certain); and 2. the proportion of workers exposed in the given job (1≤40%; 2=40–90%; 3≥90%);

  • intensity of exposure, expressed in relation to the circumstances of use (personal preparation of the pesticide mixture, use of hand pump or tractor, size of the area being treated, re-entry after treatment) and use of personal protective equipment. Semiquantitative estimates of exposure were derived from the publicly available EUROPOEM programme,33 and then categorised on a 4-point scale (0=unexposed; 1=low; 2=medium; 3=high);

  • frequency of exposure, expressed in annual days of pesticide use reported in the questionnaire or estimated based on the type of plant disease and the size of the crop or the livestock being treated (low≤ 50 days/year; medium 51–100 days/year; high≥101 days/year).

A cumulative exposure score was calculated for each pesticide group as follows: Embedded Image where C is the cumulative exposure score; i the study subject; j the jth job in the work history of study subject i; y the duration of exposure (in years); x the exposure intensity level f the exposure frequency level.

Cumulative exposure scores for each pesticide group were then categorised by tertiles of their distribution among the exposed (cases and controls combined).

Consistency in the occupational coding and exposure assessments was optimised through several meetings of the industrial hygienists.

Statistical methods

We assessed risk of B cell lymphoma, and its most prevalent subtypes, diffuse large B cell lymphoma (DLBCL) and chronic lymphocytic leukaemia (CLL), associated with ever exposure to inorganic and organic pesticides (all types), and the organic pesticide groups listed in table 1. The analysis was led by PC, supported by GS, SD, MP and TN, both on all exposed subjects, and after restriction to subjects whose exposure was assessed with high confidence. Linear trends in all exposure metrics were also estimated. The OR was calculated using unconditional logistic regression, adjusted for age, gender, education and centre. Two-tailed 95% CI for the OR were estimated using the Wald statistics (eβ±(zα/2× seβ)). Subjects unexposed to any pesticide comprised the reference category used for all the analyses. Trends in the ORs were assessed using the Wald test for trend.

Table 1

Prevalence of exposed to the individual pesticide groups by country in the EPILYMPH study

Role of the funding sources

The private and public institutions that sponsored this study did not influence or intervene in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; and in the decision to submit the paper for publication.

Results

Details on the study size, number of cases and controls by participating centre, and their frequency distribution by selected variables of interest in the occupational analyses were reported elsewhere.34 Table 1 shows the frequency distribution of exposure to pesticide groups for which exposure was assessed in the EPILYMPH study, by country. In a footnote, the active ingredients within each group are reported, selected among those reported by study subjects or suggested by the collaborating agronomists. Overall, the prevalence of exposure to pesticides in our study was low, with only 3.7% of participants exposed to inorganic pesticides and 6.4% exposed to organic pesticides, and it was lowest for triazines and triazoles and phenoxy acids. The prevalence of exposed was highest in Spain and lowest in the Czech Republic. The prevalence of exposure to the specific groups of pesticides varied by country. Use of inorganic pesticides was widespread, but it mainly consisted of copper sulphide or other sulphur compounds as reported by study subjects, indicated by the agronomist or by the crop-exposure matrix. Arsenicals were mainly used in Spain and, to a smaller extent, in Ireland. Among organic pesticides, chlorophenols were most frequently represented, and their prevalence was highest in Spain, Italy and Germany. Organophosphates were the most prevalent group of organic pesticides in Ireland. The most variegate pattern of pesticide use was described in Italy.

Table 2 shows risk of lymphoma overall, B cell lymphoma, DLBCL and CLL, amongst those ever exposed to each type of pesticide considered in this study. No excess risk of lymphoma (all types), B cell lymphoma and DLBCL was observed in association with ever exposure to inorganic or organic pesticide, nor to any of the organic pesticide groups assessed in this study. Risk of CLL was significantly associated with ever exposure to organic pesticides and particularly to organophosphates (OR=2.6, 95% CI 1.2 to 6.0). An elevated risk of CLL was also associated with ever exposure to inorganic pesticides (OR=1.6, 95% CI 1.0 to 2.5), but not with arsenical pesticides. Because of the a priori hypothesis of an association, we cite the moderate excess risk of B cell lymphoma associated with ever exposure to phenoxyacids (OR=1.4, 95% CI 0.6 to 3.1). No excess risk was observed in association with ever exposure to the other pesticide groups. The results did not change when exploring risk for NHL, thus excluding CLL and multiple myeloma, but including T-cell lymphomas. Further adjustment for ever exposure to solvents or contact with livestock did not virtually change the risk estimates.

Table 2

Risk of lymphoma and major subtypes associated with ever exposure to pesticide groups in the Epilymph study

Table 3 shows risk of CLL by intensity of exposure. The excess risk associated with ever exposure to inorganic pesticides was limited to the lowest category of intensity of exposure. Risk for medium-high intensity of exposure to organophosphates showed a 2.6-fold excess (CI 95% 0.7 to 9.2), matching that observed for the low intensity category; however, the CI was wide at either level, and there was no trend in risk (Wald test for trend=0.14; p=0.44).

Table 3

Chronic lymphocytic leukaemia risk and intensity of exposure to pesticide groups (all levels of confidence)

Risk of B cell lymphoma associated with ever exposure to organophosphates did not vary according to whether exposure started before 1980 or from 1980 onwards and it did not increase by cumulative exposure tertiles (Wald test for trend p=0.09). Instead, CLL risk was highest when exposure started from 1980 onwards (OR=4.0, 95% CI 0.9 to 16.6), and it increased significantly by increasing cumulative exposure tertile (Wald test for trend p=0.02).

Only a few individual agrochemicals were represented by a sizable number of study subjects, and the exposed cases of DLBCL, CLL and even B cell lymphoma overall were too few for any meaningful inference to be drawn. Exposure to the three most frequently identified individual organophosphate pesticides, namely dimethoate and parathion, among the most commonly used agricultural insecticides, and glyphosate, an organophosphorous herbicide, was more prevalent among B cell lymphoma cases, while exposure to 2,4 dichlorophenoxyacetic acid (2,4 D) was not (table 4). Four cases and no controls had been exposed to methylchloro-phenoxyacetic acid (MCPA) (p=0.13); these were one case of diffuse large B cell lymphoma, one case of follicular lymphoma and two cases of unspecified non-Hodgkin's lymphoma. Three cases and one control to other phenoxy herbicides (p=0.28) not shown in the tables.

Table 4

Risk of B cell lymphoma and occupational exposure to selected specific active ingredients of pesticides

When limiting the analysis to the only study subjects whose exposure was assessed with a high degree of confidence, numbers became smaller and CIs wider. The excess risk of CLL associated with exposure to organophosphates and that of B cell lymphoma associated with exposure to phenoxyacids were no longer observed. Risk was not significantly elevated for the B cell lymphomas overall among study subjects with high confidence of exposure to organophosphates, and CLL risk was moderately increased among study subjects with high confidence of exposure to organochlorines . Overall, these results were not interpretable because of the small number of cases and the rarity of the exposed.

Discussion

Our results provide limited support to the hypothesis of an association between occupational exposure to organophosphorous pesticides and risk of CLL. We did not find evidence of an association with lymphoma overall, B cell lymphoma as a group of different subtype entities and DLBCL. The low prevalence of exposed in our community based study did not allow to explore the association with other less prevalent lymphoma subgroups, nor to detect unquestionable associations with specific agrochemicals. Also, we were unable to confirm the repeatedly reported association between exposure to phenoxyacids and lymphoma. It is worth reporting, however, that while we did not observe any indication of a higher prevalence of exposure to 2,4 D among cases in respect to controls, four B cell lymphoma cases and no controls were identified as exposed to MCPA.

Organophosphate insecticides were introduced for agricultural use in Europe mainly in the early 1970s, when insect resistance to organochlorines became manifest. Their use was associated with an almost twofold increase in risk of NHL in a Nebraska case-control study;35 women appeared to be at greater risk.15 Similar findings were reported in Italy and China.36 ,37 An increase in NHL risk was also reported in Australia for exposures defined as substantial, although no increasing trend in risk was observed with frequency, intensity level, probability, duration and period of exposure.38 Specific organophosphates were investigated in several studies. Malathion, one the most frequently used organophosphorous insecticide, showed an association in a Canadian case-control study,39 and in another study conducted in Iowa and Minnesota.40 Diazinon and dichlorvos also showed an association in the Minnesota study.40 The positive association with exposure to diazinon was confirmed, but limited to lymphocytic lymphoma, in one study,41 while results of the US Agricultural Health Study were negative for malathion.42 Studies were negative for phorate,43 and positive for terbufos, scoring the fourth in the US sales of organophosphates, although again no trend was observed by exposure metrics.44 Selected lymphoma subtypes, such as multiple myeloma45 and hairy cell leukaemia46 were reported in association with exposure to glyphosate. The four B cell lymphoma cases exposed to glyphosate in our study included one case each of DLBCL, CLL, multiple myeloma and unspecified B cell lymphoma.

2,4 D is the best known phenoxy acid. Its association with NHL risk was first reported in Sweden47 and thereafter confirmed in the US,48 with a sixfold increase among farmers using it for more than 20 days/year, and significant increases in risk for direct use and lack of use of personal protective equipment. Further positive results were reported in case-control studies,35 ,49 ,50 while a Danish follow-up study51 was negative for an association, and a multicentre mortality study of 19 000 European workers exposed to phenoxy acids and chlorophenols52–54 confirmed the association in presence of concurrent exposure to 2,2,4,4 tetrachlorodibenzodioxin, a frequent contaminant of phenoxy herbicides and chlorophenols, but not when tetrachlorodibenzodioxin exposure was excluded. Recent updates of historical cohorts of phenoxyacids manufacturers and trichlorophenol manufacturers provided negative or conflicting findings.55–57 On the other hand, a dose-related increase in NHL mortality by semiquantitative indicators of exposure to pentachlorophenol was observed in a cohort of woodworkers in British Columbia, Canada58 and two out of three NHL cases in a small cohort of Swedish woodworkers exposed to phenoxyacid herbicides belonged to the highest exposure subgroup.59 Although the most recent case-control studies, with a more accurate exposure assessment, tend to confirm the association of NHL risk with phenoxy herbicides, and particularly 2,4D, 4-chlor-2-MCPA, and 4-chlor-2-methyl phenoxypropionic acid (MCPP or meclorpop),14 ,39 ,44–46 ,60 ,61 reviews still underline the uncertainties and inconsistency in the results.62 A specific effect of 2,4D on the haemolymphopoietic tissue was supported by the observation of an increase of the lymphocyte proliferation index in workers exposed to the herbicide.63

Our results did not find an association with phenoxyacid herbicides and chlorophenols, and provide only limited support to an increase in risk associated with exposure to organophosphate insecticides and herbicides. We did not observe an association with exposure to carbamates and thiocarbamates, organochlorines, and triazines and triazoles.

Use of carbamates in general, and carbaryl in particular, was associated with an increase in NHL risk in Italy,36 Canada39 and the USA40 and the association was inconsistent by exposure metric in the US Agricultural Health Study,64 while, in this large survey, NHL risk showed an increasing trend with increasing exposure to butilate, a thiocarbamate,65 and sevin was the only carbamate associated with a dose-related increase in NHL risk in a Chinese study.66 We did not find an association between exposure to carbamates and risk of lymphoma or its most prevalent subtypes.

A first suggestion of an increasing NHL risk among exposed to organochlorine insecticides, namely chlordane, toxafene, aldrin, lindane and DDT, came from several international case-control studies, where concurrent exposure to numerous other pesticides also occurred.36 ,39 ,40 Multiple myeloma,67 CLL38 and hairy cell leukaemia68 were more frequently associated. However, detailed analyses of US case-control studies found out that the positive association with exposure to DDT or lindane disappeared after adjusting for exposure to organophosphates and phenoxyacids.69 ,70 Positive findings have been more recently published on lindane,71 ,72 but they keep being inconclusive for the entire class of organochlorines.38

Suggestions of a positive association between triazine herbicides and NHL risk provided limited evidence because of multiple concurrent exposures and the small study size;68 ,73 besides, the pooled analysis of three case-control studies74 and the analysis of the repeatedly cited US Agricultural Health Study75 ,76 did not support the hypothesis of a role of atrazine and cyanazine. Less relevant in this regard seems to be the increased risk associated with exposure to metribuzin in one study, as all lymphopoietic malignancies were considered altogether and the CI included unity.77

Our study presented the advantage of a very detailed exposure assessment, coupled with an up-to-date pathological definition of disease entities. Such conditions represent substantial improvements in the assessment of occupational exposures and in lessening exposure and disease misclassification in population-based studies, which would help in revealing true associations. While this differentiates our effort from most previous population-based case-control studies, loss of statistical power is the unavoidable consequence of the gain in specificity. In fact, the overall prevalence of exposed was small, with 3.8% participants exposed to inorganic pesticides and 7.1% exposed to organic pesticides, and when restricting the analysis to study subject with high confidence exposure to pesticides in general, or when investigating individual chemicals, numbers were further reduced and the CI of the risk estimates widened. For the same reason, no analysis on the effects of specific combined pesticide exposures was conducted. On the other hand, since biological effects are likely to vary by individual chemicals in a chemical class or functional class, when exploring associations with such classes results may be diluted, thereby missing true effects related to individual chemicals.

The use of hospital controls in several centres contributing to this multinational effort, and the low response rate in the two centres where controls were population-based, may have introduced selection bias, further limiting the interpretation of our study results.

We included age, gender, education, and centre as covariates in our regression models to adjust our risk estimates, as a set of core factors potentially relevant in subgroup analyses. We selected education as a surrogate for lifestyle factors potentially acting as confounders of the association between pesticide exposure and risk of B cell lymphoma. In turn, lifestyle might surrogate exposure to endotoxin, a component of the outer membrane of Gram-negative bacteria, a common contaminant associated with poverty, crowding, pets, household cleanliness and the rural environment.78 ,79 However, the endotoxin role in lymphomagenesis remains to be investigated. Other potential confounders would include occupational exposure to solvents and livestock, and household use of insecticides. We did not observe any change in the risk estimates for B cell lymphoma and CLL associated with ever exposure to organophosphates after adjusting by ever exposure to solvents or contact with livestock. Information on household insecticide use was self-reported by the study subjects; however, the low prevalence of occupationally exposed in our population based study, and the poor ability of study subjects to identify the chemical class of the household insecticides, did not allow the use of such information. We cannot exclude that bias might have resulted, although it seems unlikely that it would have acted only on the specific association between occupational exposure to organophosphates and CLL.

Caution is therefore required in interpreting our findings. Among pesticides considered in the IARC Monographs for their potential human carcinogenicity,1 subjects in our study mentioned having used arsenicals, DDT, chlorophenols and phenoxyacids. In most instances, the use of these chemicals date to early periods in the work histories of study subjects, while the limited evidence of an association with CLL risk was related to still popular organophosphorous insecticides and phenoxy herbicides that did not undergo specific IARC evaluations thus far. The lack of consistent dose response trends with all the exposure metrics might support chance as the explanation for the observed associations, or it might imply some mechanism different from a direct intervention in the carcinogenic process. For instance, dimethoate was shown to have the lowest cytotoxic and genotoxic potential in cultured cells, compared to other three organophosphates and the organochlorine endosulfan;80 however, its administration in experimental female mice caused a decrease in total immunoglobulins and IgM and in the number of plaque forming cells;81 the same effects were observed over three generations following repeated administration of low doses dimethoate in outbred Wistar rats.82 Functional activity of Th1 lymphocytes, immune reactions associated with these cells, and interferon-γ  production were impaired after subacute malathion intoxication in albino rats,83 while thymic atrophy and reduction in splenic germinal centres followed methylparathion administration in rabbits.84 Such immunosuppressive effects do not seem related to acetylcholinesterase inhibition, the typical toxicological mechanism of organophosphate poisoning, and cover a large number of pesticides, including organochlorines, organophosphates, carbamates and pyrethroids.85 ,86 It is unclear whether the typically toxicological criterion of dose-response in establishing causal association would apply also in mechanisms involving the immune system.

In conclusion, our analysis of a large European data set provides no support to a role of occupational exposure to several specific agrochemicals in the aetiology of B cell lymphoma, and limited support in the aetiology of CLL. Further multicentre studies in international settings coupling state of the art exposure assessment in farm work and availability of detailed pathological diagnoses with a larger study size might provide the proper setting to further test the hypothesis.

References

Footnotes

  • Funding (1) European Commission, 5th Framework Programme, Quality of Life (QLK4-CT-2000-00422); (2) European Commission, 6th Framework Programme, FP6-2003-FOOD-2-B (contract No. 023103); (3) the Spanish Ministry of Health (grant No. 04-0091, RCESP 09-10); (4) the German Federal Office for Radiation Protection (grants No. StSch4261 and StSch4420); (5) La Fondation de France; (6) the Italian Ministry for Education, University and Research (PRIN 2007 prot. 2007WEJLZB and PRIN 2009 prot. 20092ZELR2); (7) the Italian Association for Cancer Research (IG 2011/11855).

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval Local Ethics Committees in each of the six study centres.

  • Provenance and peer review Not commissioned; externally peer reviewed.