1 15 1 https://www.nal.usda.gov/exhibits/speccoll/files/original/c7b4c2ff56c104ff469c33776b3e4eec.pdf 6842e2a58724a67bf7193de9c2e8fd57 PDF Text Text Item ID Number 01769 Author Woods, James S. Corporate Author Roport/Artldo TitlB Soft Tissue Sarcoma and Non-Hodgkin's Lymphoma in Relation to Phenoxyherbicide and Chlorinated Phenol Exposure in Western Washington JOIirnal/BOOk TltlB Journal of the National Cancer Institute Year 1987 Month/Day Ma Color NumbOT oflmaD.es v n 12 Descripton Notes Monday, June 11, 2001 Page 1770 of 1793 �Soft Tissue Sarcoma and Non-Hodgkin's Lymphoina In Relation to Phenoxyherbicide and Chlorinated Phenol Exposure In sg^a*!*! Vllaohtin/v$rtn 1-2 .•S'v/ s 8. Wood.-*, Ph.D., 3 Lincoln Pcilssar, P/j.D., * Richard K. Sewr,<m, P/7.D.,4 Linda S. Heuser, MA,4 and Bruce G., Kuiander, Af.D. 5>9 propionic acid (Silvex), which have been reported to be contaminated with TCDD as high as 30 ppm (16). Moreover, TCDD is the prototype of a larger series of halogenated aromatic hydrocarbons including other PCDDs, dibenzofurans, and biphenyls, all of which share similar biochemical properties and which may also contaminate phenoxyherbicide or chlorophenol products in commercial use. Hence concern exists regarding the possible health risks from exposure to chemical products in which any such contaminants may be present. The present study was conducted to investigate the relationship between the incidence of STSs and NHLs and past exposure to phenoxyhevbicides and chlorinated phenols with the use of population-based case-control approach. Subjects were drawn from the male population of western Washington State, where phenoxyacetic acid herbicides and chlorophenols have been widely utilized during the past 40 years by agricultural, forestry, and wood products industries. Specific emphasis was placed on identification of intervening factors or conditions that may influence human susceptibility to the risk of cancer in association with exposure to phenoxyacetic acids, chlorophenols, or PCDDs. ABSTRACT—A population-based case-control sfcudy was conducted in western Washington Stale io evaluais the relationship between occupations) exposure of men aged 20-79 to phenoxyacetic acid herbicides and chlorinated phenols and the risks of developing soft tissue sarcoma (STS) and non-Hodgkin's iymprioma (NHL). Occupational histories and other data were obtained by persona! interviews for 128 STS cases and 576 NHL cases, diagnosed between 1981 and 1984, and for 894 randomly selected controls without cancer. Among the study subjects with any past occupational exposure to phenoxyherbicides, ihe estimated relative risk ar.d 95% confidence intern! of developing STS was 0.80 (0,5-1.25, and of developing NHL, 1.07 (0.8-1.4). Risk estimates of developing STS and NHL associated with past crtlorophenoi exposure were 0.99 (0.7-1.5) and 0.99 (0.8-1.2), respectively. No increasing risk of eirfiar cancer was associated with overall duration or infensity of chemical exposure or with exposure to any specific pher.oxyherbicide per se. However, estimated <-isks of NHL were elevated among men who had been farmers, 1.33 {1.03-1.7), foresiry herbicide applicators, 4.80 (12-19.4), and for those potentially exposed to pr»enoxyherbicic!es in any occupation for 15 years or more ouring the period prior to 15 years before cancer diagnosis, 1.71 (1.04-2.8). increased risks of NHL were also observed among those with occupational exposure to organocriiorine insecticides, such as DOT [1,82 {1.04-3.2)} and organic solvents [1 35 (1.08-1.7)), and to other chemicals typically encountered in the agricultural, forestry, or wood products industries. These results demonstrate small but significantly increased risks of developing NHL in association with some occupational activities where phenoxyrterbicirjes have been used in combination with other types of chemicals, particularly for prolonged periods. They do not demonstrate a positive association between increased cancer risks and exposure to any specific phenoxyter&icide product alone. Moreover, these findings provide no evidence of increased risks of developing NHL associated with chlorinated phenol exposure or of developing STS associated with exposure to either class of chemicai.—JNC! 1987; 78:399-910. ABBREVIATIONS USED: CSS=Cancer Surveillance System; 2,4-0=2.4(dichlorophenoxy;acetkacid; DDT= l,l'-(2,2.2-trichloroethylidcoe;bis [•1-chIorobenz.enel; !CD-O = International Classification of Diseases for Oncology, First Revision; NHI. = non-Hodgkin's lymphoma; PCDD— poll/chlorinated dibenzo-p-dioxin; STS—soft tissue sarcoma; 2,4,5-T= (2,4.5-mchlorophcnoxy(acetic acid; TCOD=2,3,7,8-tetrachlorodibenzop-dioxin. 1 Recent studies from Sweden (7-5) and elsewhere (6-10} have reported an increased risk of STSs or NHLs in association with occupational exposures to phenoxyacetic acid herbicides and/ or chlorinated phenols. Negative studies of this association have also appeared (11-15). Implicated as the putative carcinogen in most of these studies is TCDD (CAS: 1746-01-6), a trace contaminant formed during the manufacture of chemicals that are derived from alkaline hydrolysis of 1,2,4,5tetrachlorobenzene. Of particular concern in this respect are 2.4,5-trichlorophenol and the phenoxyherbicides 2,4,5-T (CAS: 93-76-5) and 2-<2,<t,5-trichlorophenoxy)- Received June 30, 1986: accepted October 2!, 1986. ^Supported by Public Health Service grant CA-29900 from the Division ot Extramural Activities, National Cancer Institute. 'Batielle Human Affairs Research Centers, 4000 N.E. -Hst St., Scattie, WA 98105. 'The Fred Hutchinson Cancer Research Center, 1124 Columbia St., Seattle. WA 98 KM. Swedish Hospital Medical Center. Seattle. WA 98104. 6 We are indebted to the staff of the Epidemioiostic Research Unit of the Fred Hutchinson Cancer Research Center for their assistance in the conduct of this study. Also, we gratefully thank Frans M. Enzinger, M.D.. Chairman, Department of Soft Tissue Pathology, Armed Forces Institute of Pathology, Washington, DC, for assisting in the confirmation and histologic classification of soft tissue sarcoma case materials in this study. 899 JNCI. VOL. 78, NO. 5, MAY 1987 �900 Woods, Polissar, Severson, et ai. SUBJECTS AND METHODS Acquisition oj cases and controls.—From 1983 through 1985 men with either STS or NHL were identified through the CSS, a population-based turner registry that has covered 13 counties of western Washington State since 1974. Men eligible to participate in the study were between the ages of 20 and 79 at diagnosis and had been diagnosed during the years 1981-84 as having either STS or NHL, as classified hisiologically by the World Health Organization ICD-O (17). Case definitions of eligibility for STS by OCD-O codes are 8800-8804, 8810-8813, 8830-8832, 8840, 8850-8860, 8890-8920, 8990-8991, 904090-14, 9120, 9130, 9150, 9170, 9251, 9260, 9503, 9540-9560, and 9580-9581. Case definitions of eligibility for NHL by ICD-O codes are 9590-9642, 9690-9701, and 9750. The CSS obtains data, from all hospitals throughout a 13-county region, and essentially complete population coverage is attained. During 1983-85 a control group without STS or NHL was also selected. Live control subjects aged 20-64 were chosen by means of a random-digit-dialing procedure as described by Waksberg (18). Because the elderly are relatively sparse in the general population, it was not economical to locate all older controls by random digit dialing. Thus additional live coutrol subjects aged 65-79 were chosen at random from data supplied by the Health Care Financing Administration covering Social Security recipients in the study area. Controls were group matched to NHL cases by vital status and 5-year age group with a ratio of 1.2 controls per case. Deceased controls exclusive of homicides and suicides were identified from noncancer death certificates for persons aged 20-79 with a date of death occurring during the study period and with a residence within the 13-county study area. The disposition of case and control subjects according to vital status, interview outcome, and pathologic review of STS is presented in table 1. Interview procedure.—All study subjects, including proxies, were interviewed in person by experienced interviewers at a time and location of their choice. The interview lasted approximately 1 hour and covered the participant's residential, military, and medical histories with a detailed section on occupational exposures to herbicides, chlorophenois, and other chemicals prior to diagnosis or interview. Detailed information on known or suspected risk factors for STS and NHL, including past history of iraraunosuppressive disorders, use of immunosuppressant medication, or family history of cancer or immune disease, was also obtained. Additional information, including pathology, stage, extent of dasease, and date of diagnosis, was collected for each case from data supplied by the CSS, "Exposure" for all variables refers to that occurring prior to the date of diagnosis for the cases and prior to the date of interview for controls. The effect of recall differences on cancer risk owing to the difference in lapse time between diagnosis and interview for the cases {^1 yr) versus initial contact and interview for controls (•=*! mo) was evaluated and found to be negligible. Risk estimates among living and deceased cohorts were comparable for essentially al! associations; hence living and deceased groups were combined in all analyses. The questionnaire utilized in the interview was designed after identifying the principal occupations and activities undertaken within the study area that have involved the manufacture or use of phenoxyherbicides and/or chlorophenois. Such occupations and activities were identified in consultation with local industrial and university representatives who had had long-term experience with forestry, wood products, and agricultural industries in the Pacific Northwest. This consultation resulted in identification of 34 specific job titles (e.g., wood products worker; herbicide applicator) and 37 specific job activities (e.g., spray weeds with phenoxyherbicides; manufacture chlorophenois) involving potential exposure to either phenoxyherbicides or chlorophenois. In addition, 14 specific phenoxyherbicide or chloropheriol preparations in common commercial use in the study area were identified. Each job title or activity was then assigned to a "high," "medium," "low," or "no" exposure category for both phenoxyherbicides and chiorophenols, reflecting the consensus of the consultant group of the likely intensity of exposure to each TABLE 1.—A ease-control study of STS and NHL in relation to phetioxyherkicide a>id cM&ro-phmol Subject Total identified as eligible Physician refusal Respondent refusal Other reasons for noninterview Total interviewed Excluded by pathologic review Excluded as noneligihle post interview 4 Total subjects for analysis Dead Living No. Percent 150 20 -1 6 120 21 2 13 3 4 80 18" _ 97 _ JNCI. VOL. 78. NO. 5, MAY 1987 Dead Living Dead Living No. Percent No. Percent No. Percent No. Percent 527 80 15 24 408 15 3 o 77 219 23 8 10 178 — 11 4 5 81 622 — 95 51 476 — —4 — —1 288 — 29 40 219 — 0 10 14 76 12 0 Percent _-. 13 -I 7 77 28" — 31 — 402 174 — 475 219 — No. 56 7 2 4 43 " Percent of total interviewed. Based on diagnosis, age, residence, or date of diagnosis. 6 Controls NHL STS Subject disposition e m western Washington State: —6 ._. __ 15 8 77 — — — �STS and NHL and Chemical Exposure in WA 901 chemical received in that occupation. Examples oC occupational activities in- each exposure category are : given in the tables. In the interview, cue questions refeiring to each specific job tide, activity, or chemical preparation were asked to determine whether a subject had worked in an occupation involving probable exposure to phenoxyherbicides or chlorophenols. If an affirmative response to any cue question was elicited, a series of additional inquiries was made to acquire detailed information regarding the extent of exposure to specific chemicals of interest and the precise lime intervals during which each exposure episode had occurred. Among the additional questions asked was an inquiry regarding the name and address of the supervisor or a close co-worker in eacli job episode where possible chemical exposure had occurred. This information was used with subject's permission to corroborate self-reported exposure to specific chemicals and to verify exposure histories in instances where the subject or his proxy was uncertain regarding the actual exposure circumstances in that job. Additional measures taken to reduce the likelihood of recall bias with respect to the reporting of pherioxyherbicide or chlorophenol exposure included avoiding a deliberate focus on these specific chemicals in the interview and conducting the interview during a period (1983-85) of relatively low local or national media attention to die herbicide-cancer issue, The .-study was not advertised locally, and subjects were not made aware of the focus of the investigation except that it had broad implications with respect to environmental factors related to cancer etiology. In coding occupational exposure to phenoxyherbicides or chlorophenols based on job descriptions, the inclusive dates during which a person was employed in any specific occupation or activity were recorded in the questionnaire. In the analysis, however, occupational exposure to phenoxyherbicides and chlorophenols was considered to have begun no earlier than 1945 and 1937, respectively, the years when these chemicals first came into widespread commercial use in the study area. The coding of each job episode held by a study subject according to intensity and duration of exposure permitted evaluation of the exposure history of each subject in terms of duration of continuous or cumulative exposure at each dose level. Thus a complete exposure profile on each subject for each class of chemical under evaluation was obtained. Quality control procedures.—Several quality control procedures were employed 10 verify the accuracy of the data collected by interview. First, 145 respondents from a total of 1,444 who completed interviews were randomly selected and were recontacted by telephone approximately 1 month after the interview date. Interview and reinterview responses to seven questions were compared. Five of the seven questions reflected an 88% or higher agreement with the original interview, and one item had an 81* agreement rate. The remaining item, cups of coffee consumed per week, had a 34% agreement rate, although 55% of the discrepancies involved minor disagreements regarding the precise number of cups consumed. A second quality control procedure consisted of independent receding of a random sample of 238 coded interviews. The overall agreement rate between codes and recedes was 99%. The agreement rate was 98.9% for codes pertaining to phenoxyherbicides and 98.3% for codes pertaining to chlorophenols. Pathologic review of soft tissue sarcoma case materials,—Pathologic review of histologic slides and/or sections prepared from tissue blocks of STS cases was conducted to verify die initial diagnosis of STS. Reviews v/ere performed by a pathologist (B.K.) with expertise in soft tissue tumors. For the review of STS, specimen materials were acquired from each participating hospital where the initial diagnosis of STS was made and were then coded and sent "blind" to the project pathologist. In all, histologic material from 155 cases was obtained for review. Twenty percent of the slides or sections were resubmitted to the pathologist as an internal check of his first diagnosis. In a few circumstances the project pathologist disagreed with either the CSS or himself regarding the diagnosis of either STS per se or the specific histologic classification of STS. In these instances a sample of the disparate slides was sent to the Armed Forces Institute of Pathology in Washington, DC, for resolution of STS and/or histologic-type classification. This three-tier STS evaluation procedure resulted in retention of 45% of the STS cases in the study with histologic diagnosis as originally determined, and an additional 34% of cases retained with STS were confirmed although they were at a different histologic classification than originally made. Of the original cases 21% were determined either not likely to be STS, or of indeterminate pathology, and, consequently, were excluded from the study. Many of these were primary bone tumors. In light of the widely accepted high level of reliability associated with the accurate identification of NHL according to broad histologic classifications, pathologic materials from NHL cases were not reviewed for verification in this study. Statistical methods.—The data analysis includes estimates of relative risks as odds ratios and their 95% confidence intervals for a number of variables of interest. The most common confounding variable was age; thus all relationships were age adjusted by 5- or 10-year age groups, where possible. Pooled odds ratios were calculated by means of the Mantel-Haenszel method (19). Test-based confidence intervals were calculated by means of the method of Miettinen (20). In addition, logistic regression analysis (21) was used to evaluate the joint effect of several variables and possible interactions between chemical exposure and other potential risk factors on cancer risks. The total number of subjects represented in each analysis varies according to the rate of unknown responses: for occupational assessments unknowns were omitted from the analyses. In all other analyses unknowns were included in the "none" or "no exposure" category. The percentage of the total study population responding affirmatively with respect to specific job titles, exposures, or other characteristic is presented in the tables. "Significance" in all analyses was determined at the 5% level. JNCI, VOL. 78, NO. S, MAY 1987 �902 Woods, PoSlssar, Scverson, et at. - - , •• -."•< .£!,*'&4:$^': • • " ' ' "'' "•' TABLE 2.~Seteeted eharaeteriatics of 128 STS cose men, 576 NHL cose me-it {diagnosed 1981-84),. and 604 interviewed population control men (1933-85) Specification- •.•"'- ;• •• • -&;$••V.STS; % ' * NHL, % . - - - - : VU,, T'-' • Controls, % — -vw High school graduate:," ^g.78.6.v54,:,72.0 " U n k n o w n •-'»'..• '•'. -'-i/ :'5^'-V^ 1.7 ! Race .---. 4;;';^-'-: |,^*&.'•; "' • . -.White , ;.V X89.8s^'^'•.94.4"Black ife, 2.3 ;':*iV'i'2.1 Other , • • ; • ' •ff;.7.0. '^'.^•."•'^•o 4 Unknown , • -., . o.;i ^v'0.8. • • '•' Annual income level >'.•>*;-•• •'• .'•n, ' <$15,000/yr ::!25.8 • '-''••:' 28.0' $I5,000-$30,000/yr . . •V 32,8 ' -':!' 37.3 >$30,000/yr ••-33.6 "" 33.5 Refused '.:. 3.9 O.S Unkown •' 3.9 ' ""' "0.7 :;^-.ii;ao. 26.4 70.6 3.0 95.1 2.3 2.6 39.2 34-3 0.7 ' 1.7 w ' V RESULTS Characteristics of the Study Population Table 2 presents a comparison of NHL and STS cases and controls for selected demographic attributes. The 3 groups were similar in most respects. Notable differences included a higher proportion of STS cases than controls among subjects in the 20-39 age group and a higher percentage of STS cases in the "other" race category. Asians account principally for this difference. Risks Related to Intensity of Chemical Exposure When estimates of relative risks of both SI'S and NHL were determined for various levels of phenoxyherbicide or chlorophenol exposure among the entire study population, no increasing trend or risk estimates significantly different from unity for any exposure level were seen (table 3). However, elevations in risk were observed for several specific occupations involving chemical exposure. Estimates of relative risks of STS and NHL for various occupations involving exposure principally to phenoxyherbicides are shown in table 4. When viewed across job categories involving increasing levels of exposure, no clear indication of increasing risk is observed. No significantly increased risks of either cancer were seen among those in occupations involving low exposure to phenoxyherbicides, although a nearly significant 1.70 (0.9-3.1) odds ratio for NHL was observed among landscapers. Among those in medium-exposure occupations, no significantly elevated risks of STS were seen. A small but significantly increased risk of NHL [1.33 (1.03-1.7)] was observed among farmers, an occupation traditionally associated with regular use of weed killers. This observation is consistent with reports from other studies that have identified agricultural occupational groups at increased NHL risk (22-25). Further evaluation of this JNCf. VOL. 78. NO. 5, MAY 1987 observation with regard to duration of exposure and to use of specific substances indicated that the risk of NHL increased from 1.02 (0,7-1.5) among those working as farmers from 1'to-9 years to 1.62 (0.9-2.9) among farmers^ with, 10-19 years in that occupation. However,' no" increased risks of developing NHL were observed among those with more than 20 years as farmers, 0.92(0.5-1.6).. Additionally, there were no increased risks of NHL among farmers who reported having "regularly worked with" 2,4-D [0.68 (0.3-1.4); CAS: 94-75-7], 2,4,5-T [0.74 (0.3-2.1)], or phenoxyherbicides per se [0.71 (0.3-1.5)] when compared with study subjects reporting no phenoxyherbicide exposure. The estimated risk of developing NHL among fanners who reported having regularly ; sprayed weed killers by backpack, tractor, or aircraft was H.13 (0.7-1.9). A risk of developing NHL equal to 1.46 (0.8-2.8) was observed among farmers who reported having worked with the specific organochlorine insecticides DDT and chlordane. In occupations associated with high herbicide exposure, none was associated with a statistically significant increased risk of developing STS. A substantially increased risk of developing NHL [4.80 (1.2-19.4)] was noted for persons who claimed to have worked regularly in jobs involving the spraying of weed killers in national, state, or commercial forests. However, increased risks of tins magnitude were not observed among those engaged in other herbicide spraying activities or among those who worked as herbicide formulators or applicators per se. Further evaluation of the risk of NHL observed among forestry herbicide sprayers with respect to specific chemicals used and duration of exposures indicated that all forestry sprayers reported the combined use of 2,4-D and 2,4,5-T as well as various -commercial herbicide preparations containing these and other chemicals. An infinite risk estimate was attained for developing NHL among forestry herbicide applicators specifically in association with phenoxyherbicide TABLE 3. —Risk (pooled odd-s ratio) of developing STS or NHL in men aged 20--79 by estimated intensity of past occupational exposure to phenoxyherbicides or chloro-phenols: 128 STS cases and 576 NHL cases (diagnosed 1981-84) and 694. papulatwn control men (interviewed 1983-85) in western Washington State" Exposure category STS NHL OR (95% CD OR (95% CD Percent study population Phenoxyherbicides None Low Medium High 1.0 0.56(0.3-1.1) 0.99(0,6-1.7) 0.89(0.4-1.9) 1.0 0.90(0.6-1.3) 0.95(0.7-1.3) 1.24(0.8-1.9) 63.3 12.0 16.5 8.2 Chlorophenols None Low Medium High 1.0 0.90 (0.5-1,.6) 0.93 (0.6-1..5) 0.93 (0.5-1..8) 1.0 0.97 (0.7- i .3) 0.92 (0.7-1.2) 0.92 (0.9-1.4) 41.5 16.5 31.8 10.2 "The percentage of the total study population in each exposure category is also presented. �SYS and NHL and Chemical Exposure in WA TABLE 4.—Risk (pooled odds ratio) of developing STS or NHL in men aged. 20-79 for specific occupations and activities involving potential pkenoxyherbicide exposure: ISS men toith STS and-576 men with, NHL in western Washington State (diagnosed 1981-84) ' ' ' -population control men (interviewed 1983-85)" Phenoxy herbicides Occupation or activity ^ __ OR (95% CI) _ _NHL Percent study population OR (95% CD - Low exposure Landscaper Railway worker Telephone lineman Medium exposure Gardener-groundskeeper Fanner Working in sprayed area High exposure Herbicide formulator-mixer Herbicide applicator Spraying herbicides from backpack Spraying herbicides from tractor or aircraft Spraying farmlands with herbicides Spraying forests with herbicides Spraying near utility lines or railroad tracks 3.6 0.92 (0.3-2.8) 1.14 (0.6-2.2) 0.73 (0.1-3.6) 1.70 (0.9-3.1) 1.06 (0.7-1.5) 1.28 (0.6-2.6) 10.7 1.07 (0.5-2.2) 1.25 (0.8-1.9) 1.34 (0.7-2.6) 0.83 (0.5-1.4) 1.33* (1.03-1.7) 1.33 (0.9-2.0) 30.0 1.24 1.77 0.80 1.27 1.35 1.53 (0.7-3.3) 1.33 (0.8-3.9) 0.82 (0.4-1.5) 1.51 (0.9-2.5) 1.35 (0.8-2.4) 4.80* (1.2-19.4) 1.03 (0.3-3.1) 2.1 2.2 3.3 4.9 4.3 1.0 0.9 (0.3-5.3) (0.5-6.6) (0.3-2.4) (0.5-3.1) (0.5-3.3) — — 2.4 6.1 8.4 "The percentage of the total study population in each occupation or activity is also presented. exposure, inasmuch as there were no control subjects who served as forestry herbicide sprayers and did not use phenoxyherbicides. This association was statistically significant with P=.OQ4. However, only a very small number of exposed subjects (7) were involved. The risks of STS and NHL associated with all occupations involving potential exposure to phenoxyherbicides considered together were 0.80 (0.5-1.2) and 1.07 (0.8-1.4), respectively. The risks of developing NHL associated with exposure specifically to 2,4-D and 2,4,5-T and to phenoxyherbicides in general among the entire study population were 0.73 (0.4-1.3), 0.98 (0.5-2.0), and 0.87 (0.51.5), respectively. Table 5 presents risk estimates of SI'S and NHL for various occupations and activities involving chlorophenol exposure. As in the case of phenoxyherbicides, no clear indication of increasing risks of either STS or NHL is apparent when viewed across job categories in. volving increasing levels of exposure to chlorophenols. Somewhat elevated risks of developing STS are suggested for lumber graders [2.66 (1.1-6.4)] and log-lumber inspectors [4.83 (0.6-38.2)], although other jobs involving comparable chlorophenol exposure offer no suggestion of substantially increased STS risks. No specific jobs involving chiorophenol exposure were associated with increased risks of developing NHL. When all TABLE 5.—Risk (pooled odds ratio) of developing STS or NHL in men aged £0-79 for specific occupations and activities involving potential chlorophenol exposure: ISS men ivith STS and 576 mm with NHL in western Washington State (diagnosed 1981-84) and «'&4 population control men (interviewed Chlorophenol Occupation or activity OR (95% CI) Low exposure Planer mill worker Feeder man Bander man Medium exposure: Log-lumber inspector Sawmill worker Wood products worker Fork lift driver in mill High exposure Lumber grader Wood preserver Manufacturer of chlorophenols Percent study population STS OR (95% CI) 1.55 (0.5-4.7) 1.31 (0.4-4.7) 0.90 (0.2-4.4) 1.39 (0.7-2.7) 1.44 (0.7-2.81) 1.45 (0.6-3.4) 3.1 2.9 1.8 4.83 0.97 1.27 1.52 0.40 (0.0-3.6) 1.03(0.7-1.4) 0.88 (0.6-1.3) 1.44 (0.8-2.6) 0.4 15.0 13.0 4.2 0.94(0.5-1.9) 1.64 (0.6-4.2) 1.72 (0.9-3.4) 3.1 (0.6-38,2) (0.5-1.8) (0.7-2.3) (0.6-3.6) 2.66* (1.1-6.4) 0.79 (0.1-5.9) 1.37 (0.4-4.3) 1.4 3.0 "The percentage of the total study population in each occupation or activity is also presented. P<.05. 6 JNCI. VOL 78. NO. 5. MAY 1987 903 �904 Woods, Poilssar, Severson, et al. occupations involving potential exposure to chlorophenois were considered together, the risks of developing STS and NHL were 0.99 (0.7-1.5) and 0.99 (0.8-1.2), respectively. Results of Supervisor-Co-worker Survey Corroboration of self-reported occupational exposure to pherioxyherbicides, chlorophenols, and other chemicals was sought through telephone contact with employer supervisors or co-workers of subjects who had been employed in jobs involving potential exposure to such substances. Confirmation of subjects' responses regarding exposure was provided in essentially all instances where employers or co-workers could be reached. There were no significant differences between agreement rates for cases or controls. The ability to acquire corroborative evidence of exposure (==80% of contacts attempted) was greatest for most recent occupations held. Risks Related to Duration of Chemical Exposure Inasmuch as cancer risks may vary with length of exposure to specific carcinogens or tumor-promoting agents, it was of interest to estimate the relative risks of STS and NHL for various lengths of cumulative exposure to either phenoxyherbicides or chlorophenols. Hence risk estimates of both cancer types for 10-year durations (1-10, 11-20, >20) of exposure to phenoxyherbicides or chlorophenols were calculated. In this context, exposure refers to any level of occupational phenoxyherbicide or chloropbeaol exposure during any portion of the time period since 1946 or 1937, respectively, up to the date of diagnosis (cases) or interview (controls). "Duration" refers to the total length of cumulative exposure during this time period. From these calculations, it was determined that the risk of neither STS nor NHL increased with the duration of phenoxyherbicide or chlorophenol exposure for periods of 20 years or more when all levels of exposure are considered concomitantly. Moreover, no increased risks of either cancer were seen when increasing lengths of exposure to only high levels of phenoxyherbicides or chlorophenols were considered. Evaluation of a Latent Period tor Cancer Development The average latent period for a carcinogenic effect of aromatic hydrocarbons in humans has been postulated to be on the order of 15-30 years (26). It was, therefore, of interest to determine if exposure to either phenoxyherbicides or chlorophenols prior to an assumed latent period of 15 years before cancer diagnosis was associated with an increased risk of STS or NHL. For these determinations, the risk of each cancer type were calculated for two durations of cumulative exposure (1-14 and >15 yr) during the period between 1946 or 1937 for phenoxyherbicides and chlorophenols, respectively, and 15 years prior to diagnosis or interview. The results revealed a significant increase in the risk of NHL [3.71 (1.04-2.8)] among those with cumulative exposures to phenoxyherbicides of more than 15 years during the period preceding 15 years before diagnosis. When a 5-year latent period for cancer development was assumed, the risk of NHL among those with 15 years or more of prelatency phenoxyherbicide exposure dropped to 1.29 (0.9-2.0). In contrast, the risk of developing NHL was 2.51 (0.5-13.0) among subjects with more than 15 years of herbicide exposure prior to a 25-year assumed latent period. No increased risks of STS for either chemical or of NHL for chlorophenols were observed in relation to chemical exposure for any latent period or cumulative exposure assumption. Evaluation of Other Risk Factors for STS and NHL Since autoimmune diseases, primary immunodeficiency syndromes, and other conditions that compromise immune competence may act as independent risk factors of STS or NHL, it was of interest to determine the risks of either type of cancer associated with such conditions existing prior to the year of cancer diagnosis (or interview). Table 6 presents risk estimates of STS TABLE fj.—Risk {pooled odds ratio) of developing STS or NHL among men aged 20-70 with/actors or conditions associated with compromise of Immune competence: 128 men with STS and 576 men with NHL (diagnosed 1981-84) and 69l> population controls (interviewed 1983-85) in western Washington, State" Risk factor Malaria Preexisting cancer Nonskin Skin Corticosteroids Rheumatoid arthritis Low gamma or imrnunoglobulin Irnrnunosuppressant drug therapy 6 Immune deficiency in a blood relative STS NHL OR (95% CD OR (95% CD 1.14(0.4-3.1) 1.47 (0.9-2.5) 4.7 0,88 (0.3-2.5) 1.47 (0,7-3.1) 0.70 (0.4-1.1) 1.24 (0.7-2.1) 1.57r( 1.03-2.4) 0.91 (0.7-1.2) 1.38 (0.9-2.2) 1.59 (0.5-4.ti) 10.97r (2.1-57.3) 1.51 (0.4-o.ti) 4.8 7.5 27.5 5.9 1.0 0.7 0.6 "The percentage of the total study population with each risk factor is also presented. *Azathioprine, cyclophosphamide, chlorarnbucil, and/or mercaptopurine. ' P<.05. JNCI. VOL, 78, NO. 5, MAV 1987 Percent study population �STS and NHL and Chemical Exposure in WA 905 TABLE 7. —Risk (pooled odds ratio) of developing STS or NHL among men aged 20-79 associated with miscellaneous occupational factors or conditions among 128 men with STS and 576 men vrilh NHL (diayaosed 1981-84) and 694 population controls (intervieujed 1983-85) in western Washington State" Condition Insecticides Chlordane DDT Industrial chemicals Organic solvents Lead-lead arsenate Welding-metal fumes Chloracne Skin blisters from chemicals Cigarette smoking Coffee drinking J!TS NHL OR (95% CI) OR (95% CI) 0.96 (0.2-4.8) 1.10 (0.4-3.2) 1.61 (0.7-3.8) 1.82* (1.04-3.2) 1.6 4.0 1.10 1.51 1.30 3.32 1.72 0.93 0.49 1.35* (1.06-1.7) 1.60* (1.1-2.3) 1.31* (1.03-1.7) 2.12 (0.6-7.0) 1.06 (0.7-1.6) 0.85 (0.8-1.1) 1.28 (0.9-1.9) 29.8 12.0 31.1 1.0 7.8 73.5 89,8 (0.7-1.7) (0.9-2.6) (0.9-2.0) (0.8-14.0) (0.9-3.2) (0.6-1.4) (0.4-1.2) Percent study population " The percentage of the total study population with each condition or factor is also presented. * P<-.ftfi and NHL for a number of such conditions or factors. All assessments were made from subject interviews, not from medical records, Immunosuppressant drug therapy received for any reason prior to the year of diagnosis of NHL was the greatest risk factor for either cancer. Immune deficiency in a blood relative was also associated with an increased (although nonsignificant) risk of NHL. In contrast, the use of cordcosteroids for observed periods of use up to 29 years was not associated with an increased risk of either cancer type. Of specific note is the 1.38-fold increased risk of NHL observed in relation to a history of rheumatoid arthritis. Although not statistically significant, this observation is interesting in light of the current controversy surrounding the question of increased risks of NHL among persons afflicted with autoimmune diseases (27). Preexisting skin cancer was also associated with a slightly increased risk of both STS and NHL. Although the type of skin cancer was not ascertained in this study, several histologic types of STS and NHL can have dermal manifestations that could have been misreponed as "skin cancer." Among these are the soft tissue tumors, dermatofibrosarcoma (8832/3) and Kaposi's sarcoma (9140/3), and the non-Hodgkin's lymphorna, mycosis fungoides (9700-9701/3). Kaposi's sarcoma was excluded from evaluation in this study, and none of those reporting skin cancer was found to have this form of STS. Among those 12 STS cases reporting a prior history of skin cancer, none was found to have dermatofibrosarcorna; among 53 NHL. cases, none had confirmed mycosis fungoides. Malaria was evaluated as a potential risk factor for STS and NHL in light of studies suggesting an etiologic association between malarial infection and the risks of developing sarcomas or lymphomas through mechanisms involving piasmodial suppression of immune defenses against malignant diseases (28, 29). The risk of developing neither STS nor NHL was significantly elevated in association with a self-reported prior history of malaria alone. Finally, table 7 presents risk estimates associated with various occupational and/or lifestyle factors that were observed in the present study to independently alter the risk of STS or NHL and that might, therefore, be considered as potential modifiers of the effect of chemical exposure on cancer risks. Of particular note is the significantly increased risk of NHL seen among men with previous exposure to organic solvents, lead or lead arsenate pesticides, and welding and metal fumes. The risk of NHL was also elevated among men reporting previous exposure to the organochiorine insecticides chlordane and DDT. Also of note is the increased risk of both STS and NHL observed among those reporting a prior incidence of chioracne. Although this condition was not clinically confirmed in this study, this observation is of considerable interest inasmuch as chioracne is an important clinical manifestation of exposure to high levels of chlorinated dibenzodioxins and furans in humans (30). The statistical association of reported chioracne with phenoxyherbicide exposure was of borderline significance (P<.075) in this study. Neither cigarette smoking nor coffee drinking was a risk factor for either cancer. Not shown in table 7 are a number of other factors or conditions that were also evaluated as potential risks factors, but for which no significant or suggestive associations with either STS or NHL were found. These include: exposure to radiation, x-rays, or radioactive materials; exhaust fumes from motorized equipment; home use of phenoxyherbicides; residency near areas sprayed with weed killers; tuberculosis vaccination; consumption of multiple vitamins; eating fish caught in Puget Sound; and use of alcohol. Logistic regression analysis was used to estimate the potential interaction between phenoxyherbicides or chlorophenois and various other single variables appearing in table 6 or 7 as well as others of interest. Intervening variables for which the interaction with phenoxyherbicides or chlorophenois were determined included: organochiorine pesticides (DDT+chlordane), lead-lead arsenate, welding or metal fumes, inherited or acquired diseases of the immune system, any prior cancer, prior skin cancer, home use of phenoxyherbicides, family hisJNCI. VOL. 78. NO. 5, MAY 1987 �903 Woods, Polissar, Severson, et ai. lory of cancer, and family history of diseases of the immune system. A1J analyses were controlled for age. Results of the logistic regression analysis for these variables confirmed the magnitude of the risks shown in tables 6 and 7. None of the interactions between chlorinated phenol or phenoxyherbicide exposure and these variables was statistically significant, with the exception of that between phenoxyherbicides and organic solvents as a. risk factor for NHL. The odds ratio for joint exposure compared with exposure to neither substance was 1.50 with a 95% confidence interval of 1.03-2.18. The odds ratio estimates from the model for exposure solely to organic solvents or phenoxyherbicides were nonsignificandy 1.12 and 0.85, respectively. The significance of the joint exposure in this case may be a result of the large number of comparisons made in this study or, possibly, due to a genuine synergistic effect. Positive although nonsignificant interactions were also observed between exposure to either phenoxyherbicides or chlorophcnols and coexisting or preexisting autoimmune diseases or immune deficiency syndromes when those listed on table 6 and others (mononucleosis, celiac sprue, and Sjogren's syndrome) were considered jointly. T'his interaction was most notable with chlorophenols where the odds ratio for joint exposure (compared with neither) was 1.40 (0.95-2.07) as compared with 0.91 (0.72-1.14) and 1.32 (0.99-1.78), respectively, for chlorophenol or immune deficiency alone. These observations are worthy of note in light of the widely held theory (31) that the etiology of malignant lyniphomas involves failure of immunoregulation in the face of a persistent stimulus for lymphocyte proliferation. In autoimmune diseases, chronic amigenic stimulation is provided by the constant exposure to self-antigens, whereas in primary immunodeficiency syndromes, recurrent infections are the likely source for antigenic stimulation. In contrast, TCDD is well recognized as a suppressant of both humoral and cell-mediated immunity in animals (32-34) and has been recently characterized as a depressant of cell-mediated immunity in humans during prolonged, low-level exposure (35). Immunosuppression by other PCDDs has also been described (36). Therefore, further consideration should be given to the possibility that the statistical interactions, however slight, observed in this study between phenoxyherbicides or chlorophenols and immunosuppressive disorders have biologic relevance with respect to the etiology of NHL in humans. DJSCUSSJON The results of the present study demonstrate significantly increased risks of developing NHL among men in occupations involving farming and forestry herbicide spraying and for occupations involving prolonged phenoxyacedc acid herbicide exposure when a latent period of 15 or more years before cancer diagnosis is considered. However, neither phenoxyherbicides nor chlorophenols alone appear to constitute a sufficient cause of either NHL or STS when evaluated within a populaJNCI, VOL. 78, NO. 5, MAY 1987 tion residing in western Washington State, since increased cancer risks were not observed for numerous other occupations or activities involving comparable opportunity for exposure to these substances. In this regard, the present findings are not consistent with results of studies conducted in Swedish and other populations, which report consistent and substantially increased risks of both types of cancer in association with occupational exposure to specific chlorophenols or phenoxyacetic acids or to combinations of these chemicals. In consideration of possible reasons underlying the apparent lack of consistency between the results of Swedish studies and those conducted here and elsewhere, three issues that have not received major attention with regard to this question include: a) differences in the intensities or dosages of chemicals received by workers in comparable occupational activities, b) differences in the extent of environmental (nonoccupational) exposure to the chemicals received by the respective study populations, and c) differences between study populations with respect to the proportional distribution of other risk factors that in combination with phenoxyherbicides or chlorophenols contribute causally to the cancers putatively associated with chemical exposure alone. In addressing the question of differences in dosages of chemicals received by workers engaged in comparable job activities, it is likely that, should the chemicals under investigation independently increase cancer risks in humans, this effect should be more apparent among persons receiving higher dosages. In considering this possibility as it pertains specifically to application of phenoxyherbicides, it is known that spraying activities, as well as work in sprayed areas, extend over substantially shorter periods of the year in Sweden than occur in western Washington State, owing largely to climatic differences in the length of the growing season. Thus Swedish workers participate in activities in which phenoxyherbicide exposure is typically consolidated within a 2- to 3-month period annually, during which spraying activities involving intensive herbicide exposures that extend over several consecutive weeks at a time are not uncommon (/, 37). In contrast, the annual spraying season in the Pacific Northwest extends over 6-7 months, during which individual spraying episodes usually span only a few days at a time and may be separated by weeks or even months during which no spraying is performed. Such differences in herbicide use patterns could conceivably lead to Swedish applicators receiving appreciably higher cumulative exposures to biologically active substances than occurs among workers engaged in less intensive use patterns, as supported by several studies of this question (38, 39). To analyze this possibility quantitatively, we have employed the phannacokinetic model developed by Gehring et al. (40) from the oral study of 2,4,5-T in humans to calculate maximum absorbed daily dosages of herbicide received by Swedish and American workers. The workers in question engaged in activities with comparable job description, namely, herbicide applica- �STS and NHL and Chemical Exposure in WA 907 tor with the use of tractor-drawn equipment. Calculations, based on urinary herbicide concentrations with the use of data derived from studies of applicators under actual field conditions (37, 41), indicate that the maximum daily dose of 2,4,5-T absorbed by American workers in an application operation involving two sprayings, 2 weeks apart, is in the range of 12-86 /ug/kgbody weight, with a mean maximum daily dose of 45 MS^kg-In contrast, calculated maximum daily dosages received by Swedish workers, doing the same type of work but involving spraying for 3-4 hours/day over a consecutive 2-week period, ranged from 11 to 315 ftg/kg, with a mean maximum daily dosage of 90 /*g/kg. Maximum daily dosages received by American workers involved in other modes of herbicide application were backpack crew, 19-104 Mg/kg; helicopter crew, 17-23 Mg/kg; and mixers, 12-138 pg/kg. These results suggest that maximum daily dosages of herbicides received by Swedish applicators could substantially exceed those received by American counterparts as well as those engaged in other modes of spraying operations. From the pharmacokinetic studies in humans {•/#), it is known that 2,4,5-T is absorbed and excreted in the urine with a half-life of about 1 day following a single oral exposure. Moreover, from measurements of urinary 2,4,5-T, the maximum dose that can be absorbed on repeated exposures is estimated to be about 100 jug/kg/day, with the expected maximum concentration in plasma of individuals receiving repeated daily doses reaching a plateau after 3 days of such exposure. Based on these considerations, the calculated mean maximum daily dose of 2,4,5-T received by Swedish workers under the occupational circumstances described above (90 ng/kg) would approximate that required to produce the maximum plasma concentration of 2,4,5-T that could be sustained during spraying operations. Thus Swedish applicators would experience higher sustained tissue levels during repeated frequent exposure episodes, as well as higher tissue concentrations of any PCDDs or other contaminants that would be concomitantly absorbed than would be experienced by American counterparts, who receive lower exposures on a more sporadic basis. Although the implications of these calculations with respect to human cancer risks are difficult to estimate, they are nevertheless of interest in light of findings from recent studies of Swedish subjects (42) involving analysis of PCDDs in abdominal fat from both cases of STS and NHL as well as control subjects. These studies reported that cancer cases exposed to phenoxyherbicides 16-31 years previously had levels of highly chlorinated PCDDs significantly higher than control subjects unexposed to phenoxyherbicides. Interestingly, no differences in case or control TCDD levels were seen. These findings are consistent with the observations from both Swedish (2) and Danish (7) studies of increased cancer risks among persons exposed to phenoxyherbicides and chlorophenols that do not contain detectable levels of TCDD per se but that are most likely contaminated with a variety of other chlorinated dibenzodioxins and furans (43, 44), some of which have been shown to have carcinogenic potential (45, 46). Moreover, the capacity of TCDD, and presumably its approximate isostereoisomers, to act as a cocarcinogen (47) and a tumor promoter (48, 49) have been well characterized. These properties are highly dose dependent. Thus the exposure of Swedish workers to potentially higher concentrations of biologically active substances associated with the use or manufacture of phenoxyherbicides or chlorinated phenols is a consideration possibly consistent with the higher cancer risks observed in studies on such subjects. A recent study by Hoar et al. (25) showing that the relative risk of NHL dramatically increased with the number of days of phenoxyherbicide use per year among agricultural workers in the United States supports this view. Differences in risk estimates observed between this and the Swedish studies might also be accounted for on the basis of variation in the extent of nonoccupational exposure received by the general populations in areas where the studies were conducted. Several investigators (50, 51) have recently reported widespread contamination of the general population in the United States and Canada with PCDDs and PCDFs, based on analysis of human fat samples. The findings indicate that, although higher levels of total dioxins and other contaminants may be seen in some exposed persons, there is considerable overlap in actual tissue concentrations of such substances between some persons with confirmed occupational exposures and others who are not previously known to have been exposed through job-related activities. These observations suggest dial epidemiologic studies conducted in areas where the extensive use of phenoxyherbicides and chlorophenols has occurred may have inadvertently included subjects who have experienced significant exposure to the chemicals of concern outside of the occupational setting. Should this be the case, it is possible that estimates of actual risk based on recall of occupational exposures alone may be underestimated, owing to nondifferential misciassification of subjects according to exposure status (52). To estimate the extent to which nonoccupational exposure to phenoxyherbicides may have occurred in the present investigation, we have evaluated data from several air-monitoring studies (53, 54) conducted during the spraying season in the Pacific Northwest. These data indicate that phenoxyacetic acids as well as PCDDs can be transported in the atmosphere, either as vapor or adsorbed on particles, for distances ranging from several hundred feet up to a mile from die application area (54), depending on weather conditions and mode of dispersion. The maximum concentration of 2,4,5-T, for example, found in 24-hour collections from sampling stations in one study (55) was 3.4 MS/m3- Concentrations of up to 10 /ig/rn3 of other phenoxyherbicides have been detected at sampling sites in agricultural regions of Washington State (56). If it is assumed that a 70-kg person inhales 30 m3 of air per day, a person residing in the proximity of a sprayed area could conceivably receive a dose of 2,4,5-T equal to 1.5 /ig/kg/day during the spraying operation solely from atmospheric sources. These levels are on the order of 10-50 times less than those JNQ. VOL. 78, NO. 5. MAY 1987 �908 Woods, Poiissar, Severson, et ai. received from occupational sources, as described above, and moreover are received via a different route of exposure (inhalation versus dermal), which could alter absorption rales appreciably. However, 24% of STS cases, 23% of NHL cases, and 21% of control subjects in the present study responded positively to the question "Have you ever lived in an area where weed spraying was routinely done by truck or airplane?" This response suggests that considerable population exposure to phenoxyherbicides and their contaminants from environmental sources could have occurred over the 40-year exposure-assessment period. Eliminating subjects who reported residential or home use exposures to phenoxyherbicides or chlorophenols in the present study did not alter the estimated risks of developing STS or NHL. Hence it is unlikely that bias due to such exposures could account for the large differences in risk estimates observed between these and the Swedish studies. Nevertheless, should phenoxyherbicides and/or their contaminants increase the risk of cancer at environmental exposure levels, or, as recently suggested, produce subclinical immune system alterations that may predispose to sucli risks (35), it is possible that risk estimates based solely on assessment of occupational exposures could be attenuated as a result of exposure misclassification. Confirmation of this possibility awaits further investigations based on direct analysis of tissue chemical content or the development of a reliable surrogate measure of past chemical exposure. A third possible reason for a lack of consistency between the results of this and the Swedish studies may be differences between the study populations with respect to the proportional distribution of factors or conditions other than chemical exposure that contribute to the cancers under evaluation. If, for example, a specific inherited, lifestyle, or environmental condition that independently predisposes to increased risks of the cancer(s) associated widi chemical exposure is more prevalent among Scandinavians than among other populations, increased cancer risks could be observed in the former group, even if the prevalence of phenoxyherbicide or chlorophenol exposure were the same or even less than that occurring elsewhere. In this regard it is interesting to note in the present study that exposure to the insecticides DDT and lead arsenate as well as to agricultural and industrial chemicals, such as organic solvents and welding-rnetal furnes, are associated with significantly increased risks of developing NHL (table 7). Compromise of the immune system (31), exposure to zoonotic viruses (24, 57), and chronic mitogenic stimuli (2-f, 31) have also been suggested as etiologic factors for NHL. The extent to which differences in the proportional distribution of such factors between Swedish and the local study populations might account for the inconsistencies observed between the results of this and the Swedish studies cannot be currently estimated, since neither the specific conditions that modify the effects of chemical exposure on cancer risks nor their prevalence among the respective populations have as yet been idenJNCI. VOL. 78, NO. 5. MAY 1987 tified. However, information based on interview responses from existing studies (4) indicates that Swedish populations may have had as much as twice the frequency of exposure to DDT (5.8-7.8 vs. 3.8% locally) as well as to total insecticides (14.6 vs. 7.9%), and this higher exposure may underlie some increased risk of developing cancer, particularly NHL, independently of or in combination with the chemicals currently under study. On the other hand, the frequency of exposure to organic solvents, for which a small but significant interaction with phenoxyherbicides was observed in this study, was approximately equal (28.2 vs. 29.9%) between Swedish and local study populations. Similarly, the frequency of cigarette smoking, now or ever, among Swedish and local study populations was comparable (71 vs. 73%). Little or no data are available regarding population differences in nutritional, medical, or other lifestyle factors that might serve as component causes of STS or NHL or that modify the effect of chemical exposures on cancer risks. The prospect that inherited factors or conditions among Scandinavians might contribute to increased risks of developing cancer in that population when exposed to the chemicals under investigation is another possibility that might account for differences in risk estimates observed. Although specific studies of this question have as yet to be accomplished, laboratory investigations have shown that the binding affinity of the TCDD receptor in biological tissues, as well as subsequent receptor-mediated events, are genetically determined and may vary considerably even among different strains of the same animal species (58). That such variations are also a characteristic of human genealogy is suggested by recent studies (59) that demonstrate both the presence of the Ah (TCDD) receptor in human lung as well as considerable heterogeneity in the human population in regard to lung Ah receptor concentrations. . Moreover, evidence of heritable differences in aryl hydrocarbon hydroxylase induction among humans (60) has been presented. Recently, a hereditary predisposition for the development of STS has been described in the study of Danish phenoxyherbicide manufacturers (7). However, no attempt has been made to determine the frequency of this condition among Scandinavian or other populations or to investigate the extent to which the presence of such a condition could modify the effect of chemical exposure on cancer risks. In the present investigation, we have taken advantage of the fact that approximately 6% of the population of the study area is of Scandinavian heritage (61) to make a crude evaluation as to the extent to which this factor (Scandinavian heritage) might constitute an increased risk of STS or NHL in association with phenoxyherbicide or chlorophenol exposure. For this assessment, the surnames ot all study subjects were segregated into Scandinavian or "other" categories by a member of the University of Washington Department of Scandinavian Languages and Literature who had expertise in Scandinavian genealogy. Through this effort 169 subjects with Scandinavian surnames were identified including �STS and NHL and Chemical Exposure In WA 909 15 STS cases, 66 NHL cases, and 88 controls. No increased cancer risks were associated with having Scandinavian as compared with non-Scandinavian names. However, when the analysis was restricted to Scandinavians only, the risk estimates for STS in relation to past occupational chemical exposures were substantially greater than those observed among the study population as a whole both for high-level phenoxyherbicide [2.8 (0.5-15.6)] and high-level chlorophenol [7.2 (2.1-24.7)] exposures. These estimates are comparable in magnitude to those reported among subjects in Swedish (1-5) and Danish (7) studies. Moreover, the distribution of predominant histologic types of STS was comparable to that reported from the Swedish studies (2). No increased risks of NHL in relation to chemical exposures were observed among persons with Scandinavian surnames. Although the assignment of Scandinavian ancestry in these studies remains unconfirmed, the results are interesting inasmuch as they suggest that factors specific to Scandinavian descent, as opposed to residency or occupation in Scandinavian countries, may contribute to increased risks of STS when exposed to the chemicals under evaluation in this study. Further investigation of this issue may be warranted to help resolve the inconsistency between Swedish studies and those conducted elsewhere. In conclusion, the results of the present investigation demonstrate increased risks of NHL in several specific occupations in which phenoxyacetic acid herbicides are used, as well as for prolonged occupational exposure to these substances. However, they are not consistent with results from Swedish and other studies reporting substantially increased risks of either STS or NHL associated with phenoxyherbicides or chlorophenols as sole or major component causes of these diseases. Since methods of study design and analysis in this and the Swedish studies were similar in most respects, it is possible that factors specific to the populations under evaluation account for the inconsistencies observed. Concerns that bear further consideration in this regard include possible differences in the intensity or distribution of chemical exposures between the study populations and variations in the proportional distribution of specific inherited, life-style, and/or environmental factors that modify the effect of chemical exposure on the risks of cancer development. REFERENCES (1) HARDELL L, SANDSTROM A. Case-control study: Soft-tissue sarcomas and exposure to phenoxyacetic acids or chlorophenols. Br J Cancer 1979:39:711-717. (2) ERIKSSON M, HARDELL L, BF.RG NO, et al. Soft-tissue sarcomas and exposure to chemical substances: A case-referent study. Br J Ind Med 1981; 38:27-33. (}) HARDF.U. L, ERIKSSON M. Soft-tissue sarcomas, phenoxyherbicides, and chlorinated phenols. Lancet 1981; 2:250. (•f) HARDELL L. Relation of soft-tissue sarcoma, malignant lymphoina and colon canter to phenoxyacids, chlorophenols and other agents. Scand J Work Environ Health 1981; 7:119-130. (5) HARDELL L, ERIKSSON M, LENNER P, etal. Malignant lymphoma and exposure to chemicals, especially organic solvents, chlorophenols and phenoxyacids: A case-control study. Br J Cancer 1981; 43:169-176. (6) Brsnop CM, JONES AH. Non-Hodgkin's lymphoma of the scalp in workers exposed to dioxins. Lancet 1981; 1:369. (7) LYNGE £. A follow-up of cancer incidence among workers in manufacture of phenoxyherbicides in Denmark. Br J Cancer 1985; 52:259-270. (8) FINGERHUT MA, HALPERIN WE. Dioxin exposure and sarcoma. • JAMA 1983; 249:3176-3177. (9) MOSES M, SELIKOFF IJ. Soft tissue sarcomas, phcnoxyherbicides, and chlorinated phenols. Lancet 1981; 1:1370. (10) OTT MG, HOLDER BB, OLSON RD. A mortality analysis of employees engaged in the manufacture of 2,4,5-trichlorophenoxyacetic acid. J Occup Med 1980; 22:47-50. (11) WIKLUND K, HOLM LE. Soft tissue sarcoma risk in Swedish agricultural and forestry workers. JNCI 1986; 76:229-234. (12) SMITH AH, PEARCE NE, FISHER DO, et al. Soft tissue sarcoma and exposure to phenoxyherbicides and chlorophenols in New Zealand. JNCI 1984; 73:1111-1117. (13) RIIHIMAKI V, Asp S, PUKKALA E, et al. Mortality and cancer morbidity among chlorinated phenoxy acid applicators in Finland. Chemosphere 1983; 12:779-784. (14) SUSKIND R. Long-term health effects of exposure ot 2,4,5-T and/or its contaminants. Chemosphere 1983; 12:769-778. (15) PEARCE NE, SMITH AM, HOWARD JK, et al. Non-Hodgkin's lymphoma and exposure to phenoxyherbicides, chlorophenols, fencing work, and meal works employment A case-control study. Br J Indust Med 1986; 43:75-83. (16) RAPPE C, BUSER HR. Occupational exposure to polychlorinated dioxins and dibenzofurans. In: Choudhary G, ed. Chemical hazards in the workplace, measurement and control. ACS Symp Ser Vol 149.. Washington, DC: Am Chem Soc, 1981:319-342. (17) World Health Organization, International classification of diseases for oncology. 1st ed. Geneva: WHO, 1976. (18) WAKSBERG J. Sampling methods for random digit dialing. J Am Stat Assoc 1978; 73:40-46. . (19) MANTEL N, HAENSZEL W. Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst 1959; 22:719-748. (20) MIETTINEN OS. Estimabiluy and estimation in case-referent studies. Am j Epidemiol 1976; 103:226-235. (21) PRENTICE RL. Use of the logistic model in retrospective studies. Biometrics 1976; 32:597-606. (22) CANTOR KP. Farming and mortality from non-Hodgkin's lymphoma: A case-control study. Inl J Cancer 1982; 29:239-247. (23) SCHUMACHER MC. Farming occupations and mortality from nonHodgkin's lymphoma in Utah. J Occup Med 1985; 27:580-584. (24) PEARCE NE, SMITH AH, FISHER DO. Malignant lymphoma and multiple myeloma linked with agricultural occupations in a New Zealand cancer registry-based study. Am J Epidemiol 1985; 121:225-237. (25) HOAR SE, BLAIR A, HOLMES FF, et al. Agricultural herbicide use and risk of lymphoma and soft-tissue sarcoma. JAMA 1986; 256:1141-1147. (26) HUEPER WC, CONWAY WD. Chemical carcinogenesis and cancer. Springfield: Thomas, 1964:40-13. (27) PRIOR P. Cancer and rheumatoid arthritis: Epidemiologic considerations. Am J Med 1985; 78(suppl 1A):15-21. (25) HARGIS BJ, MALKIEI. S. Sarcomas induced by injection of simian virus 40 into neonatal CFW mice. JNCI 1979: 63:965-967. (29) WEDDERBURN N, CAMPA M, TOSTA CE, et al. The effect of malaria on the growth of two syngeneic transplantable murine tumors. Ann Trop Med Parasitol 1981: 75:597-605. (30) MOSES M, LUJS R, CROW KD, et al. Health status of workers with past exposure to 2,3,7,8-<eirachlorodit>enzo-p-dioxin in the manufacture of 2,4,5-tetrachiorophenoxy acetic acid: Comparison of findings with and without chloracne. Am J Ind Med 1984; 5:161-182. (31) ROBBINS SL, COTRAN RS, KUMAR V. Diseases of white cells, lymph nodes, and spleen. In: Pathologic basis of disease. 3d ed. Philadelphia: Saunders, 1984:653-704. (32) ZINKL J, Vos JG, MOORE JA. el al. Hematologic and clinical JNCI. VOL. 78. NO. 5, MAY 1987 �910 Woods, Polissar, Severson, et al. (33) (34) (35) (36) (37) (38) (39) (40) (41) (42) (43) (44) (45) (46) chemistry effects of 2,3,7,8-tetrachloiodibcn/o-p-dioxin in laboratory animals. Environ Health Perspect 1973; 5:111-118. CLARKE DA, GAUI.DIE J, SZKWCZUK MR, ei al. Enhanced suppressor cell activity as a mechanism of immunosuppression by 2,3,7,8-tetrachlorodibenzo-p-dioxin. Proc Soc Expt Biol Med 1981; 168:290-299. VF.CCHI A, SIRONI M, CANEGRATI MA, et al. Immunosuppressive effects of 2,3(7,8-tetrachlorodibenzo-£-dioxin in strains of mice with different susceptibility to induction of aryl hydrocarbon hydroxyiase. Toxicol Appl Pharmacol 1983; 68:434-441. HOFFMAN RE, STEHR-GREEN PA, WEBB RB, et al. Health effecis of long-term exposure to 2,3,7,8-ietrachlorodibenzo-p-dioxin. JAMA 1986; 255:2031-2038. HOLSAPPLE MP, McC'AY JA, BARNES DN. Immunosuppression without liver induction by subchronic exposure to 2,7-dichIorodibenzo-]t>-dioxin in adult female B6C3F] mice. Toxicol Appl Pharmacol 1986; 83:445-455. KOLMODIN-HEDMAN B, ERNE K, ENGQVisT A. Testing of professional exposure to phenoxyacids (2,4-D and 2,4,5-T). Arbete och Halsa, Vetenskaplig Skriftserie 1979; 17 (in Swedish). NEWTON N, NORRIS LA. Potential exposure of humans to 2,4,5-T and TCDD in the Oregon coastal ranges. Fund Appl Toxicol 1981; 1:339-346. TASKAR PK, DAS YT, TROUT JR, et al. Measurement of 2,4dichlorophenoxyacetic acid (2,4-D) after occupational exposure. Bull Environ Contam Toxicol 1982; 29:586-591. GEHRING PJ, KRAMER CG, SCHWETZ BA, et al. The fate of 2,4,5trichlorophenoxyacetic acid (2,4,5-T) following oral administration to man. Toxicol Appl Pharmacol 1973; 26:352-361. LENG ML, RAMSEY JC, BRAUN WH. Review of studies with 2,4,5trichlorophenoxyacetic acid in humans including applicators under field conditions. ACS Symp Ser Vol 182. Washington, DC: Am Chem Soc, 1982:135-156. HARDELL L, AXEI.SON O. Phenoxyherbicides and other pesticides in the etiology of cancer. Some comments on the Swedish experiences in cancer prevention. Strategies in the workplace. San Francisco: University of California, J984. NORSTROM A, RAPPE C, LINDALL R, et al. Analysis of some older Scandinavian formulations of 2,4-dichlorophenoxyacetic acid and 2,4,5-trichlorophenoxyacetic acid for contents of chlorinated dibenzo-p-dioxins and dibenzofurans. Scand J Work Environ Health 1979; 5:375-378. DICKSON D. PCP dioxins found to pose health risks. Nature 1980; 283:418. National Toxicology Program. NIH bioassay of 2,7-dichlorodibenzo-p-dioxin for possible carcinogenicity. Technical Report Series No. 123, 1979. . NIH bioassay of a mixture of 1,2,3,6,7,8- and 1,2,3,7,8,9hexachlorodibenzo-£-dioxins for possible carcinogenicity (gavage study). Technical Report Series No. 198, 1980. JNCI. VOL. 78, NO. 5, MAY 1987 (47) Koi!Rl RE, Rw>£ TH, JOGLEKAR R, et al. 2,3,7,8-Tetrachlorodiben/o-/>dioxin as carcinogen causing 3-methylcholanthreneinilinted subcutaneous tumors in mice genetically "nonresponsive" at Ah locus. Cancer Res 1978; 38:2777-2783. (48) PITOT HC, GOLDSWORTHY T, CAMPBELL HA, et al. Quantitative evaluation of the promotion by 2,3,7,8-letrachlorodibenzo-pdioxin of hepatocarcinogenesis from diethylniirosaminc. Cancer Res 1980; 40:3616-3620. (49) ABERNATHY DJ, GREENLEE WF, HUBAND JC, et al. 2,3,7,8-Tetrachlorodibenzo-jt>-dioxin (TCDD) promotes the transformation of C3H/10T1/2 cells. Carcinogenesis 1985; 6:651-653. (50) RYAN JJ, WILLIAMS DT, LAU BP, et al. In: Keith LH, Choudhary G, Rappc C, cds. Chlorinated dioxins and dibenzofurans in the total environment. II. Boston: Butterworth 1984:205-214. (51) GROSS ML, LAY JO, LYON PA, et al. 2,3,7,8-Tetrachlorodibenzo/>-dioxin levels in adipose tissues of Vietnam veterans. Environ Res 1984; 33:261-268. (52) COPELAND KT, QlECKOWAY H, HOLBROOK RM, et al. Bias due to misclassificatiori in the estimate of relative risk. Am J Epidemiol 1977; 105:488-495. (53) ADAMS DF, JACKSON CM, BAMESBERGER WL. Quantitative studies of 2,4-D esters in the air. Weeds 1974; 12:180-283. (54) RENNE DS, WOLF MA. Experimental studies of 2,4-D herbicide drift characteristics. Ag Meterol 1979; 20:7-24. (55) BAMESBERGER WL, ADAMS DF. An atmospheric survey for aerosol and gaseous 2,4-D compounds. In: Gould RF, ed. Organic pesticides in the environment. Chapt 18. Adv Chem Ser. Washington, DC: Am Chem Soc, 1966:219-227. (56) BOVEY RW. Uses of phenoxyherbicides and iheir methods of application. Chapt 3. In: Bovey RW, Young AL, eds. The science of 2,4,5-T and associated phenojcyherbicides. New YorkWiley, 1980:49-69. (57) BLAIR A, HAYES HM. Mortality patterns among US veterinarians. 1947-77: An expanded study. Int J Epidemiol 1982; 11:391-397 (55) GASIEWICZ TA. Evidence for a homologous nature of Ah receptor among various mammalian species. In: Poland A, Kimbroug:. RD, eds. Biological mechanisms of dioxin action. Banbur report No. 18. Cold Spring Harbor, NY: Cold Spring Harbi.; Laboratory, 1984:161-175. (59) ROBERTS EA, GOLAS CL, OKF.Y AB. Ah receptor mediating indu' lion of aryl hydrocarbon hydroxyiase: Detection in huma . lung by binding of [2,3,7,8-3H]tetrachlorodibenzo-j!J-dioxi-'. Cancer Res 1986; 46:3739-3743. (60) NEBERT DW, JENSEN NM. The Ah locus: Genetic regulation :,' the metabolism of carcinogens, drugs and other environment, chemicals by cytochrome P-450-mediated monooxygenasc • CRC Grit Rev Biochem 1979; 6:401-437. (61) 1980 Census of population. Vol 1. Characteristics of the popui. tion. Chapt C. General social and economics characterisn*. Part 49. Washington, DC: U.S. Dept of Commerce, 1983. � Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Alvin L. Young Collection on Agent Orange Description An account of the resource <p style="margin-top: -1em; line-height: 1.2em;">The Alvin L. Young Collection on Agent Orange comprises 120 linear feet and spans the late 1800s to 2005; however, the bulk of the coverage is from the 1960s to the 1980s and there are many undated items. The collection was donated to Special Collections of the National Agricultural Library in 1985 by Dr. Alvin L. Young (1942- ). Dr. Young developed the collection as he conducted extensive research on the military defoliant Agent Orange. The collection is in good condition and includes letters, memoranda, books, reports, press releases, journal and newspaper clippings, field logs and notebooks, newsletters, maps, booklets and pamphlets, photographs, memorabilia, and audiotapes of an interview with Dr. Young.</p> <p>For more about this collection, <a href="/exhibits/speccoll/exhibits/show/alvin-l--young-collection-on-a">view the Agent Orange Exhibit.</a></p> Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Box The box containing the original item. 066 Folder The folder containing the original item. 1769 Series The series number of the original item. Series III Subseries III Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Creator An entity primarily responsible for making the resource Woods, James S. Lincoln Polissar Richard K. Severson Linda S. Heuser Bruce G. Kulander Source A related resource from which the described resource is derived Journal of the National Cancer Institute Date A point or period of time associated with an event in the lifecycle of the resource 1987-05-01 Title A name given to the resource Soft Tissue Sarcoma and Non-Hodgkin's Lymphoma in Relation to Phenoxyherbicide and Chlorinated Phenol Exposure in Western Washington Subject The topic of the resource cancer risk assessment lymphomas dioxin industrial exposure ao_seriesIII