Sarcomas are a heterogeneous group of rare cancers that consist of

Transcription

1 ORIGINAL ARTICLE Soft-Tissue Sarcoma and Pesticides Exposure in Men Results of a Canadian Case-Control Study Punam Pahwa, PhD, Chandima P. Karunanayake, PhD, James A. Dosman, MD, John J. Spinelli, PhD, John R. McLaughlin, PhD, and Cross-Canada Group Objectives: The objective was to investigate the putative associations of specific pesticides with soft-tissue sarcoma (STS). Methods: A Canadian population-based case-control study conducted in six provinces was used in this analysis. The study design consisted of two stages: a self-administered postal questionnaire and a telephone interview for those reporting pesticides exposure of 10 hours per year or more; and a 15% random sample of the remainder. Conditional logistic regression was used to fit the statistical models. Results: A positive history of cancer among first-degree relatives and exposure to aldrin and diazinon were statistically significant independent predictors of an increased risk for STS, whereas diagnosis of whopping cough lowered the risk of STS. Conclusions: The incidence of STS was associated with specific insecticides after adjustment for other independent predictors. Sarcomas are a heterogeneous group of rare cancers that consist of neoplasms of connective tissues. 1 3 Soft-tissue sarcoma (STS) is a rare cancer; however, it is one of the most frequent among various sarcomas and accounts for approximately 1% of all malignant neoplasms and about 1% of all cancer deaths in adults. 4 Soft-tissue sarcoma can originate from soft tissues such as fat, muscle, nerves, fibrous tissues, blood vessels, or deep skin tissues. 1 2 In Canada, age-standardized 3-year moving average incidence rates for STS among young men have risen since 2001, after a steep decline for many years, but these rates have been stable in women since Survival rates for the period 2001 to 2004 in Canada were greater than that for 1992 to 1995 (72% vs 63%) for STS. 5 Several risk factors for STS have been identified. These include occupational exposure to certain chemicals and specific pesticides, 6 15 a positive personal history of cancer or other diseases, 16,17 certain types of medical treatments for cancer, and lifestyle factors. 2,18 22 External radiation therapy is also a wellestablished risk factor for STS, and it has been shown that the incidence of sarcomas is increased in patients treated with radiation therapy, with the average time between radiation exposure and diagnosis being approximately 10 years. 2 There has not been consistency in the detection of an association between cigarette smoking or tobacco use and STS Some studies support an association between phenoxyacetic acids and chlorophenols used as pesticides in agriculture and other From the Department of Community Health & Epidemiology (Dr Pahwa) and Canadian Centre for Health and Safety in Agriculture (Drs Pahwa, Karunanayake, and Dosman), University of Saskatchewan, Saskatoon, Saskatchewan, Canada; Cancer Control Research (Dr Spinelli), British Columbia Cancer Agency and School of Population and Public Health, University of British Columbia, Vancouver, British Columbia, Canada; and Prevention & Cancer Control (Dr McLaughlin), Cancer Care, Ontario, Dalla Lana School of Public Health, University of Toronto, and Samuel Lunenfeld Research Institute, Toronto, Ontario, Canada. The authors declare no competing financial interests. Address correspondence to: Punam Pahwa, PhD, Canadian Centre for Health and Safety in Agriculture, University of Saskatchewan, Royal University Hospital, 103 Hospital Drive, Saskatoon, Saskatchewan, S7N 0W8, Canada. pup165@mail.usask.ca. Copyright C 2011 by American College of Occupational and Environmental Medicine DOI: /JOM.0b013e applications, and STS, 6,7,9 12 whereas other studies have reported no associations. 19,21,23 25 Phenoxyacetic acids and the structurally related chlorophenols are widely used for the control of weeds and fungi as well as for wood preservation and other purposes. A strong association with STS and chlorophenol exposure was reported by Wingren et al 6 and Hoppin et al. 7 Several Swedish studies by Hardell and others reported that exposure to phenoxyacetic acids and exposure to chlorophenol gave a roughly 3-fold increased risk for STS In 1953, 64% of the pesticide products available in Canada containing 2,4,5-trichlorophenoxy acetic acid (2,4,5-T) also contained 2,4-dichlorophenoxyacetic acid (2,4-D). The pesticide 2,4,5- T is no longer manufactured or sold in Canada or the United States. A significant excess of STS has been reported in populations that are largely agricultural Our study included six provinces of Canada with diverse agricultural practices to test the pesticide-exposure hypothesis related to four rare tumors: STS; multiple myeloma (MM); Hodgkin s lymphoma (HL); or non-hodgkin s lymphoma (NHL). The objective of this article is to investigate the putative associations of specific pesticides with STS. Soft-tissue sarcoma was defined according to the classification of International Classification of Diseases for Oncology, Third Edition (ICD-O-3) morphology coding (and included M , M , M , M8840, M , M , M , M , M , M , M , M9503, M , and M ). 26 METHODOLOGY Details of the study design and methodology have been published elsewhere. 27 Briefly, we conducted a matched populationbased case-control study of male residents in six Canadian provinces (Quebec, Ontario, Manitoba, Saskatchewan, Alberta, and British Columbia) to test whether there are associations between pesticide exposures and four rare tumors (HL, NHL, MM, and STS). There were total 1528 cases diagnosed with the tumors under investigation (HL = 316, MM = 342, NHL = 513, and STS = 357). Of the STS cases, 36.4% were diagnosed in Ontario, 31.4% in Prairies, 19.6% in British Columbia, and 12.6% in Quebec. The study design aimed to have at least one control participant matched for each case according to age and province of residence. However, difficulties occurred in recruiting controls in the older age groups (eg, 69 years and older); thus, the total number of control participants was To analyze data for each specific tumor type, all controls were used to provide a consistent comparator. For example, in this report, statistical analysis is based on 357 (60.8% of those contacted) STS cases and all 1506 (48.0% of those contacted) controls. The main reasons for nonparticipation for both cases and controls were death, change of address, and refusal. The study had an average of four to five matched controls for each STS case. Incident cases were men aged 19 years or older, with first diagnosis of STS (ICD-O ) from September 1, 1991, to December 31, Cases were ascertained from provincial cancer registries, except in Quebec, where hospital ascertainment was used. After physician consent was received, postal questionnaires and informed consent forms were mailed to potential cases. Control subjects were matched by age ± 2 years to be comparable with the JOEM Volume 53, Number 11, November

2 Pahwa et al JOEM Volume 53, Number 11, November 2011 TABLE 1. Comparison of Demographic, Antecedent Personal Medical, General Pesticide Exposures, and Cigarette Smoking History Between Cases of STS and Control Subjects Based on the Postal Questionnaire STS, N = 357 Controls, N = 1506 n % n % Age, yr > > > > Mean ± SD 54.7 ± ± 16 Education level Elementary/high school (reference) University ( ) Technical/vocational ( ) University and technical/vocational ( ) Residence on a farm at any time Yes ( ) No (reference) Pesticide exposure (screening question) <10 h/yr (reference) h/yr ( ) Smoking history Nonsmoker (reference) Ex-smoker ( ) Current smoker ( ) Missing data Medical history Measles (yes) ( ) Rheumatoid arthritis (yes) ( ) Mononucleosis (yes) ( ) Whooping cough (yes) ( ) Family history of cancer any first degree relative (yes) ( ) *Odds ratio (and 95% confidence interval ) stratified by age group and province of residence. Factors found to be significant at P < 0.20 included in table. Also tested and found to be not associated at P < 0.20: mumps, allergies, acne; asthma; celiac disease; chickenpox; diabetes; hay fever; mononucleosis; rheumatic fever; glioma, Shingles, syphilis, ringworm; shingles; syphilis; tuberculosis; urinary tract infections; whooping cough; allergies; drug treatment for overactive thyroid; treatment for head lice, body lice, or scabies; medical implants; drug treatment for epilepsy; tonsillectomy; skin-prick allergy test; allergy desensitization shots; positive allergy prick skin test, patch skin test, or positive patch skin test for allergy. age distribution of the entire case group (HL, NHL, MM, and STS) within each province of residence. Potential controls (men, aged 19 years and older) were selected at random within age constraints from provincial health insurance records (Alberta, Saskatchewan, Manitoba, and Quebec), computerized telephone listings (Ontario), or voters lists (British Columbia). Postal questionnaires and informed consent forms were mailed to potential controls, and information collected from all participating control subjects was used in the statistical analyses of each cancer site. Deceased subjects were ineligible as either cases or controls. Data were collected in two stages: by self-administered postal questionnaire in stage 1; and detailed pesticide exposure information was collected via telephone interview in stage 2. The pesticide exposure questionnaire developed by Hoar et al 28 was modified (with permission) for use in our Canadian context. 27 The postal questionnaire captured demographic details, personal medical history, cigarette smoking history, lifetime occupational history, and specific occupational exposures of interest. Each subject who reported 10 hours per year or more of exposure to pesticides (in any combination of compounds), as defined by the screening questions, and a 15% random sample of the remainder were mailed a list of pesticides (both chemical and brand names) and an information letter. Each subject was subsequently telephoned to obtain details of pesticide use. Exposure to an individual herbicide, insecticide, fumigant, or fungicide was coded as Yes if they responded affirmatively to Did you use it at work, in your home garden or as a hobby? The telephone questionnaire was used as a source of information to characterize exposure to individual pesticides. Pesticide data were collected at several levels, beginning with the broadest categories (eg, minimal exposure, occupations with potential pesticide exposure) and progressing sequentially from major classes (eg, herbicides) to chemical groups (eg, phenoxy herbicides) and finally to individual components (eg, 2,4-D, 2-methyl-4-chlorophenoxyacetic acid, and 2,4,5-T). The listed pesticides were chosen for inclusion for the following reasons: (a) whether the compound was ever registered for use in Canada and reviewed by the International Agency 1280 C 2011 American College of Occupational and Environmental Medicine

3 JOEM Volume 53, Number 11, November 2011 Soft-Tissue Sarcoma and Pesticide TABLE 2. Herbicides: Frequency of Exposure to Herbicides Classified into Major Chemical Classes and as Individual Compounds (Among Compounds Reported by 1% or More of Responders) STS, N = 357 Controls, N = 1506 Major Chemical Classes n Exposed % Exposed n Exposed % Exposed Phenoxyherbicides exposed ( ) 1.09 ( ) Individual phenoxyherbicides 2,4-D ( ) 0.98 ( ) Mecoprop ( ) 1.34 ( ) MCPA ( ) 1.11 ( ) Diclofop-methyl ( ) 1.21 ( ) Phosphonic acid exposed ( ) 0.91 ( ) Individual phosphonic herbicides Glyphosate (round-up) ( ) 0.90 ( ) Thiocarbamates exposed ( ) 0.85 ( ) Individual thiocarbamate herbicides Diallate ( ) 1.28 ( ) Phenols: Bromoxynil exposed ( ) 0.92 ( ) Dicamba# exposed ( ) 1.26 ( ) Individual Dicamba herbicides Dicamba (Banvel or Target) ( ) 1.31 ( ) Dinitroaniline exposed ( ) 0.44 ( ) Individual dinitroanilinc herbicides Trifluralin ( ) 0.44 ( ) 2,4-D, 2,4-dichlorophenoxyacetic acid; MCPA, 2-methyl-4-chlorophenoxyacetic acid. *Odds ratio (and 95% confidence interval ) stratified by age group and province of residence. Odds ratio adjusted for statistically significant medical history variables (history of measles, rheumatoid arthritis, mononucleosis, whooping cough, and a positive family history of cancer in a first-degree relative) and with strata for the variables of age group and province of residence. Phenoxyherbicides include the phenoxyacetic acids (eg, 2,4-D and MCPA), the phenoxy-2-propionic acids (eg, mecoprop), the phenoxybutanoic acids (eg, 2,4-DB), and other phenoxyalkanoic acids (eg, diclofopmethyl). Glyphosate is the only phosphonic acid herbicide reported by more than 1% of responders. Round-up, Touchdown, Victor, Wrangler, Laredo do not include dicamba, and Rustler is a mixture of dicamba and glyphosate. Thiocarbamate herbicides include diallate and triallate. Bromoxynil is the only phenol herbicide included. #Dicamba as a major chemical class includes Banvel and Target, and a mixture of dicamba and glyphosate (Rustler), or mixtures of dicamba, 2,4-D, and mecoprop (Dynel DS, Killex). **Dinitroaniline herbicides include ethalfluralin and trifluralin. for Research on Cancer; (b) whether the pesticide was recently banned or restricted in Canada by the federal licensing agency; or (c) whether the pesticide was commonly used in Canada for specific purposes Merged data from the postal and telephone interview questionnaires were used for the statistical analysis by using SAS (SAS Institute, Cary, NC). 33 Results were reported as odds ratios (OR) and 95% confidence intervals (CI) derived from conditional logistic regression analyses. Conditional logistic regression models were utilized to conduct bivariable (to select variables for multivariable model) and multivariable analyses by adjusting for age group and province of residence (matching variables). Based on bivariable analysis, statistically significant variables with P < 0.20 were selected as candidates for multivariable models. All variables that were statistically significant (P < 0.05), as well as biologically important, were retained in the final multivariable model. Subgroup analyses of the associations between pesticides and STS were also performed for the three largest histological subtypes: fibromatous; myomatous; and lipomatous. RESULTS Data from the postal questionnaire based on responses from the 357 STS cases and 1506 control subjects were analyzed. The distribution of STS histology (based on ICD-O-3) 26 types was fibromatous ( ) 108 (30.3%), myomatous ( ) 62 (17.4%), lipomatous ( ) 53 (14.8%), Ewing s sarcoma (9260) 6 (1.7%), endometrial stromal (8930) 1(0.3%), other sarcoma 68 (19.0%), and not otherwise specified ( ) 59 (16.5%). C 2011 American College of Occupational and Environmental Medicine 1281

4 Pahwa et al JOEM Volume 53, Number 11, November 2011 TABLE 3. Insecticides: Frequency of Exposure to Insecticides Classified into Major Chemical Classes and as Individual Compounds (Among Compounds Reported by 1% or More of Responders) STS, N = 357 Controls, N = 1506 Major Chemical Classes N exposed % exposed N exposed % exposed Carbamates exposed ( ) 0.93 ( ) Individual carbamate insecticides Carbaryl ( ) 1.26 ( ) Carbofuran ( ) 1.10 ( ) Methomyl ( ) 0.78 ( ) Organochlorine, (1) exposed ( ) 0.86 ( ) Individual organochlorine, (1) insecticides Chlordane ( ) 0.94 ( ) Lindane ( ) 0.89 ( ) Aldrin ( ) 3.71 ( ) Organochlorine (2) ( ) 1.05 ( ) diphenylchlorides exposed Individual organochlorine (2) diphenylchlorides Methoxychlor ( ) 0.97 ( ) DDT ( ) 1.52( ) Organophosphorus exposed ( ) 1.30 ( ) Individual organophosphorus insecticides Malathion ( ) 1.23 ( ) Dimethoate ( ) 1.16 ( ) Diazinon ( ) 3.31( ) DDT, dichlorodiphenyltrichloroethane. *Odds ratio (and 95% confidence interval ) stratified by age group and province of residence. Odds ratio adjusted for statistically significant medical history variables (history of measles, rheumatoid arthritis, mononucleosis, whooping cough, and a positive family history of cancer in a first-degree relative), and with strata for the variables of age group and province of residence. Carbamate insecticides include carbaryl, carbofuran, and methomyl. Organochlorine insecticides class one includes aldrin; chlordane; dieldrin; endrin; heptachlor; lindane; and a mixture of lindane, carbathiin, and thiram (Vitavex). Organochlorine(2) diphenylchloride insecticides include DDT and methoxychlor. Organophosphorus insecticides include malathion, chlorpyrifos, diazinon, dimethoate, parathion, methidathion, and trichlorfon. Comparisons between STS cases and control subjects based on selected demographic variables, education level, general pesticide exposure, cigarette smoking, antecedent medical history, and familial history of cancer in any first-degree relative are presented in Table 1. Smoking patterns were similar in both cases and controls. No significant differences were observed between STS cases and controls with respect to their education level and response to ever lived on a farm variable (Table 1). Cases were less likely to have a history of measles (OR [95% CI] = 0.80 [ ]), rheumatoid arthritis (OR [95% CI] = 0.63 [ ]), mononucleosis (OR [95% CI]= 0.53 [ ]), and whooping cough (OR [95% CI] = 0.74[ ]) at P < 0.20 (criteria to determine variables for multivariable analysis) but more likely to have a positive family history of cancer (OR [95% CI] = 1.30 [ ]) (P < 0.05) as compared with controls. Cases and controls were similar in their cumulative exposure to pesticides of 10 hours per year or more. Herbicides were classified by major chemical classes: phenoxy; phosphonic acid; thiocarbamates; phenols; dicamba; and dinitroaniline. The frequency of exposure to herbicides classified into major classes and individual compounds for which at least 1% respondents reported exposure along with ORs adjusted for statistically significant (P < 0.20) medical history variables observed are shown in Table 2. No statistically significant differences of frequency of exposure to herbicides were observed between STS cases and control subjects. Insecticide analyses for major chemical classes (carbamates, Organochlorine (1), organochlorine (2), and organophosphorous) and by individual compounds found that exposure to aldrin was significantly (OR [95% CI] = 3.71 [ ]) (P < 0.05) associated with the incidence of STS (Table 3). The other individual insecticide that was significantly associated with STS incidence was diazinon (OR [95% CI] = 3.31 [ ]). Among fungicides, only exposure to formaldehyde came close to being significantly (OR [95% CI] = 2.07 [ ]) (borderline) associated with STS (Table 4). These OR estimates and their 95% CIs were adjusted for statistically significant (P < 0.20) medical history variables, as shown in Table C 2011 American College of Occupational and Environmental Medicine

5 JOEM Volume 53, Number 11, November 2011 Soft-Tissue Sarcoma and Pesticide TABLE 4. Fungicides: Frequency of Exposure to Fungicides Classified into Major Chemical Classes and as Individual Compounds (Among Compounds Reported by 1% or More of Responders) STS, N = 357 Controls, N = 1506 Major Chemical Classes n Exposed % Exposed n Exposed % Exposed Amide exposed ( ) 0.82 ( ) Individual amide fungicides Captan ( ) 0.89 ( ) Vitavax ( ) 0.71 ( ) Aldehyde exposed ( ) 2.07 ( ) Individual aldehyde fungicides Formaldehyde ( ) 2.07 ( ) Mercury containing exposed ( ) 1.25 ( ) Mercury-containing fungicides Mercury dust (n exposed) ( ) 1.44 ( ) Mercury liquid (n exposed) ( ) 0.78 ( ) Sulfur compounds ( ) 1.54 ( ) *Odds ratio (and 95% confidence interval) stratified by age group and province of residence. Odds ratios adjusted for statistically significant medical history variables ((history of measles, rheumatoid arthritis, mononucleosis, whooping cough, and a positive family history of cancer in a first-degree relative), and with strata for the variables of age group and province of residence. Amide fungicides include captan and a mixture of carbathiin, thiram, and lindane (Vitavax). Aldehyde fungicides include formaldehyde and a mixture of formaldehyde and iprodione (Rovral Flo). Mercury-containing fungicides include mercury dusts (Ceresan, Reytosan, and Agrox) and mercury liquids (Panogen, Leytosol, and PMAS). TABLE 5. Most Parsimonious Model: Conditional Logistic Regression Analyses that Contained Individual Pesticides and Important Covariates (P < 0.10) Variable Odds Ratio adj * Whooping cough (yes) 0.64 ( ) First-degree relative with cancer (yes) 1.32 ( ) Aldrin (user) 3.35 ( ) Diazinon (user) 3.19 ( ) *Odds ratio (and 95% confidence interval) stratified by age group and province of residence. TABLE 6. Exposure Frequency of Exposure to Selected Insecticides Stratified by the Number of Days per Year of STS Controls Individual compound Days/yr N % N % Insecticides Aldrin Unexposed >0and ( ) 4.08 ( ) > ( ) 2.91 ( ) Diazinon Unexposed >0and ( ) 4.28 ( ) > ( ) 2.68 ( ) *Odds ratio (and 95% confidence interval) stratified by age group and province of residence. Odds ratios adjusted for statistically significant medical history variables (history of measles, rheumatoid arthritis, mononucleosis, whooping cough, and a positive family history of cancer in a first-degree relative), and with strata for the variables of age group and province of residence. C 2011 American College of Occupational and Environmental Medicine 1283

6 Pahwa et al JOEM Volume 53, Number 11, November 2011 TABLE 7. Odds Ratios for Aldrin and Diazinon by Histological Subtype of STS Subtype/Individual STS Controls, N = 1506 Chemicals n Exposed % Exposed n Exposed % Exposed STS subtype fibromatous (n = 108) Aldrin ( ) 2.42 ( ) Diazinon ( ) 4.57 ( ) STS subtype- myomatous (n = 62) Aldrin ( ) 5.08 ( ) Diazinon ( ) 4.81 ( ) STS subtype lipomatous (n = 53) Aldrin Diazinon STS subtype- sarcomas not otherwise specified (n = 59) Aldrin Diazinon ( ) 4.70 ( ) STS subtype- other sarcomas (n = 68) Aldrin Diazinon *Odds ratio (and 95% confidence interval) stratified by age group and province of residence. Odds ratios adjusted for statistically significant medical history variables (history of measles, rheumatoid arthritis, mononucleosis, whooping cough, and a positive family history of cancer in a first-degree relative), and with strata for the variables of age group and province of residence. Because of small numbers odds ratios were not reported. The final parsimonious, multivariable logistic regression model (Table 5) that included individual pesticides (with P < 0.10) contained the following variables: whooping cough (OR [95% CI] = 0.64 [ ]), first-degree relative with cancer (OR [95% CI] = 1.32 [ ]), and exposures to aldrin (OR [95% CI] = 3.35 [ ]) and diazinon (OR [95% CI] = 3.19 [ ]). When the effect of frequency of exposure was examined for aldrin and diazinon (ie, the significant compounds in the multivariable analysis of Table 5), there was not a pattern of increasing risk with increasing number of days per year of exposure (Table 6). For specific STS subtypes, statistically significant associations with diazinon were detected for fibromatous (OR [95% CI] = 4.57 [ ]), myomatous (OR [95% CI] = 4.81 [ ]), and for sarcomas not otherwise specified (OR [95% CI] = 4.70 [ ]); although there was no significant heterogeneity between subgroups for either aldrin (P = 0.554) or diazinon (P = 0.261) (Table 7). DISCUSSION This was the first Canadian population-based case-control study that collected detailed information on pesticides to investigate the association between pesticides exposures and four specified cancer sites (NHL, STS, HL, and MM). In this report, we observed that there was an association between increased risk of STS and the following insecticides: aldrin (OR [95% CI] = 3.71 [ ]), and diazinon (OR [95% CI] = 3.31 [ ]) when adjusted for significant (at P < 0.20) medical history variables. Our findings are consistent with those of Zahm et al, 14 who reported that exposure to insecticides on animals was associated with increased risk of fibromatous and myomatous STS. They also reported that longer duration of usage of insecticides on animals significantly increased the risk of myomatous STS. 14 In our study, diazinon was associated with increased risk of fibromatous, and myomatous subtypes, as well as sarcomas not otherwise specified. Previous studies identified an association between phenoxy herbicides and STS 6,7,9 12 ; however, we found no association between herbicides (individual compound or major chemical class) and STS. An increased risk for two STS subtypes (malignant fibrous histiocytoma and leiomyosarcoma) and exposure to chlorophenol and cutting oil have been reported. 7 Regarding factors other than pesticides, no association was observed between cigarette smoking and STS, which is consistent with some, 21 but not all, 22 previous studies. Association between various medical conditions and the incidence of STS has been controversial. Our results showed that previous diagnosis of whooping cough (OR [95% CI] = 0.74 [ ]) or measles (OR [95% CI] = 0.80 [ ]) has borderline significantly protective effects on the incidence of STS. Having a first-degree relative diagnosed with cancer was also associated with a significantly increased risk of STS. Our data did not show associations between chicken pox, mumps, or any other medical history variables and incidence of STS C 2011 American College of Occupational and Environmental Medicine

7 JOEM Volume 53, Number 11, November 2011 Soft-Tissue Sarcoma and Pesticide These results differ from those of Serraino et al, 34 who reported that previous diagnoses of herpes zoster infection (OR = 2.4), chicken pox (OR = 2.2), and mumps (OR = 2.0) were significantly associated with increased STS risk. 34 It was also reported that a history of diabetes and previous diagnoses of measles and whooping cough were associated with the increased risk of STS, but this association was not significant. 34 Results based on our study indicated that STS cases and controls were comparable regarding education, which was consistent with previous findings. 21,34 Our study showed that higher level of education is associated (not significantly) with decreased risk of STS (Table 1); however, this may have been due to a response bias. In our multivariable model, the significant variables related to medical history were previous diagnosis of whooping cough (OR [95% CI] = 0.64 [ ]) and first-degree relative diagnosed with cancer (OR [95% CI] = 1.32 [ ]). Our study has strengths and limitations that are common to epidemiological case-control studies. Strengths include the large number of cases and controls. In addition, a pilot study was conducted to determine a suitable definition of pesticides exposure, and all study procedures were tested before implementation. A reference pathologist reviewed the 119 tumor slides and pathology reports that could be obtained and confirmed the diagnosis of 108 (91%) STS cases. Finally, the study included men who had any type of pesticide exposure and thus included both indoor, such as in animal confinement buildings, grain elevators, and manufacturing, and outdoor, such as in grain farming and commercial applications, exposures. Some of the results of our study are consistent with previous findings of an association between exposure to insecticides and the development of STS, but our study did not support the previously published results, which showed an association between exposure to herbicides and STS. We observed association between exposure to aldrin and diazinon and increased risk of STS. Some of the insecticides reported here have been restricted or banned because of concern over their environmental persistence and possible health effects in humans. 35 Dichlorodiphenyltrichloroethane and chlordane were widely used in Canada and the United States in the 1970s. Dichlorodiphenyltrichloroethane was banned 36 in the 1980s and chlordane use ceased 37 in 1995 by both countries. Registration of aldrin and dieldrin in Canada was discontinued in Our study further demonstrates that there is a need for larger studies to evaluate the relationship between specific pesticides and STS subtypes. ACKNOWLEDGMENTS The authors thank the collaborators Drs G. Theriault, D. Robson, S. Fincham, L. Skinnider, D. White, and N.W. Choi. Special thanks go to Dr H. McDuffie. The authors are indebted to the members of the Advisory Committee for this project for the sharing of their experiences (Drs G. B. Hill, A. Blair, L. Burmeister, H. Morrison, R. Gallagher, and D. White) and to the provincial coordinators, data managers, and students (T. Switzer, M. Gantefor, J. Welykholowa, J. Ediger, I. Fan, M. Ferron, E. Houle, S. de Freitas, K. Baerg, L. Lockinger, E. Hagel, P. Wang, G. Dequiang, J. Hu C. Maguire, and J. Thiessen). They extend their gratitude to Dr L. Skinnider for reviewing the tumor tissues and to the hospitals that provided the tissues. They sincerely thank each of the study participants who gave freely their time and shared personal details with them and the organizations who permitted access to their records. REFERENCES 1. Hoppin JA, Tolbert PE, Flanders WD, et al. Occupational risk factors for sarcoma subtypes. Epidemiology. 1999;10: Cormier JN, Pollock RE. Soft tissue sarcomas. CA Cancer J Clin. 2004;54: Olsson H. An updated review of the epidemiology of soft tissue sarcomas. Acta Orthop Scand. 2004;75(suppl 311): Morrison BA. Soft tissue sarcomas of the extremities. Baylor Univ Med Center Proc. 2003;16: Canadian Cancer Society. Canadian Cancer Statistics Toronto, Canada: Canadian Cancer Society/ National Cancer Institute of Canada; Wingren G, Fredrikson M, Brage HN, Nordenskjold B, Axelson O. Soft tissue sarcoma and occupational exposures. Cancer. 1990;66: Hoppin JA, Tolbert PE, Herrick RF, et al. Occupational chlorophenol exposure and soft tissue sarcoma risk among men aged years. Am J Epidemiol. 1998;148: Cancer and the Environment. Published Accessed January 5, Hardell L. Pesticides, soft-tissue sarcoma and non-hodgkin lymphoma historical aspects on the precautionary principle in cancer prevention. Acta Oncol. 2008;47: Eriksson M, Hardell L, Berg NO, Moller T, Axelson O. Sof-tissue sarcomas and exposure to chemical substances: a case-referent study. Br J Ind Med. 1981;38: Eriksson M, Hardell L, Adami HO. Exposure to dioxins as a risk factor for soft tissue sarcoma: a population-based case-control study. J Nat Cancer Inst. 1990;82: Hardell L, Eriksson M. Association between soft tissue sarcomas and exposure to phenoxyacetic acids: a new case-referent study. Cancer. 1988;62: Hansen ES, Hasle H, Lander F. A cohort study on cancer incidence among Danish gardeners. Am J Ind Med. 1992;21: Zahm SH, Blair A, Holmes FF. A Case-referent study of soft-tissue Sarcoma and Hodgkin s disease: farming and insecticide use. Scand J Work Environ Health. 1988;14: Balarajan R, Acheson ED. Soft-tissue sarcomas in agricultural and forestry workers. J Epidemiol Comm Health. 1984;38: McDuffie HH, Pahwa P, Karunanayake CP, Spinelli JJ, Dosman JA. Clustering of cancer among families of cases with Hodgkin lymphoma (HL), multiple myeloma (MM), non-hodgkin s lymphoma (NHL), soft tissue sarcoma (STS) and control subjects. BioMed Central Cancer. 2009;9:70. DOI: / Moutou C, Le Bihan C, Chopret A, et al. Genetic transmission of susceptibility to cancer in families of children with soft tissue sarcomas. Cancer. 1996;78: Brady MS, Gaynor JJ, Brennan MF. Radiation-associated sarcoma of bone and soft tissue. Arch Surg. 1992;127: Zahm SH, Fraumeni JF Jr. The epidemiology of soft tissue sarcoma. Semin Oncol. 1997;24: Berwick M. Soft Tissue Sarcoma. In: Schottenfeld D, Jr, Fraumeni JF, eds. Cancer Epidemiology and Prevention. New York, NY: Oxford University Press; Serraino D, Franceschi S, La Vecchia C, Carbone A. Occupation and soft-tissue sarcoma in northeastern Italy.Cancer Causes Control. 1992;3: Zahm SH, Heineman EF, Vaught JB. Soft tissue sarcoma and tobacco use: data from a prospective cohort study of United States veterans. Cancer Causes Control. 1992;3: Pahwa P, McDuffie HH, Dosman JA, et al. Hodgkin lymphoma, multiple myeloma, soft tissue sarcomas, insect repellents, and phenoxyherbicides. J Occup Environ Med. 2006;48: Woods JS, Polissar L, Severson RK, Heuser LS, Kulander EG. Soft tissue sarcoma and non-hodgkin s lymphoma in relation to phenoxyherbicide and chlorinated phenol exposure in western Washington. J Natl Cancer Inst. 1987;78: Smith AH, Pearce NE, Fisher DO, Giles HJ, Teague CA, Howard JK. Soft tissue sarcoma and exposure to phenoxyherbicides and chlorophenols in New Zealand. J Nat Cancer Inst. 1984;73: ICD-O-3: International Classification of Diseases for Oncology. 3rded. Geneva, Switzerland: World Health Organization; McDuffie HH, Pahwa P, McLaughlin JR, et al. Non-Hodgkin s lymphoma and specific pesticide exposures in men: cross-canada study of pesticides and health. Cancer Epidemiol Biomarkers Prev. 2001;10: Hoar SK, Blair A, Holmes FF, et al. Agricultural herbicides use and risk of lymphoma and soft tissue sarcoma.jama. 1986;48: IARC Working Group. An evaluation of chemicals and industrial processes associated with cancer in humans based on human and animal data. Cancer Res. 1980;40:1 12. C 2011 American College of Occupational and Environmental Medicine 1285

8 Pahwa et al JOEM Volume 53, Number 11, November IARC Working Group. Some halogenated hydrocarbons and pesticides exposures. In: Monographs on the Evaluation of the Carcinogenic Risk of Chemicals to Humans. Lyon, France: IARC Press; 1986: IARC Working Group. Overall Evaluation of Carcinogenicity: An Updating of IARC Monographs. Lyon, France: IARC Press; 1987: IARC Working Group. Occupational exposures in insecticide application and some pesticides. In: Monographs on the Evaluation of Carcinogenic Risks to Humans. Lyon, France: IARC Press; 1991: SAS Institute. The SAS System for Windows, Release Cary, NC: SAS Institute; Serraino D, Franceschi S, Talamini R, Frustaci S, La Vecchia C. Nonoccupational risk factors for adult soft-tissue sarcoma in northern Italy. Cancer Causes Control. 1991;2: Purdue MP, Hoppin JA, Blair A, Dosemeci M, Alavanja MC. Occupational exposure to organochlorine insecticides and cancer incidence in the Agricultural Health Study. Int J Cancer. 2007;120: Commission for Environmental Cooperation of North America. Fact Sheet: DDT no longer used in North America Storage/50/4285_DDT_en.pdf. Accessed July 15, Commission for Environmental Cooperation of North America. Fact Sheet: chlordane no longer used in North America Storage/50/4282_chlordane_en.pdf. Accessed on July 15, Agriculture Canada. CAPCO note on chlorinated hydrocarbon insecticides: aldrin, chlordane, DDT, dieldrin, and endrin. Report No , Food Production and Inspection Branch, Pesticides Directorate; Ottawa, Canada: Agriculture Canada; C 2011 American College of Occupational and Environmental Medicine