7
Cancer
INTRODUCTION
Cancer is the second leading cause of death in the United States. Among males aged 45-64, the group that describes most Vietnam veterans, the risk of dying from cancer nearly equals the risk from heart disease, the overall leading cause of death in the United States (U.S. Census, 1997). Almost one-half of all men and slightly more than one in three women in the United States will develop an invasive cancer at some time in their lives; approximately one in five Americans will die from cancer (Ries, 1997).
In this chapter, the committee summarizes and reaches conclusions about the strength of the evidence in epidemiologic studies regarding associations between exposure to herbicides and 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) and each type of cancer under consideration in this report. The cancer types are, with minor exceptions, discussed in the order in which they are listed in the International Classification of Diseases, Ninth Edition (ICD·9). ICD·9 is a standardized means of classifying medical conditions used by physicians and researchers around the world. Appendix B lists ICD·9 codes for the major forms of cancer.
In assessing a possible relation between herbicide exposure and risk of cancer, one key issue is the level of exposure of those included in a study. As noted in Chapter 5, the detail and accuracy of exposure assessment varies widely among the studies reviewed by the committee. A small number of studies use a biomarker of exposure, for example, the presence of dioxin in serum or tissues; some develop an index of exposure from employment or activity records; and others use a surrogate measure of exposure, such as being present when herbicides were
used. Inaccurate assessment of exposure can obscure the presence or absence of exposure-disease associations and thus make it less likely that a true risk will be identified. A second key issue for herbicide exposure and cancer risks is latency, the effect of timing of exposure on subsequent risk of disease. Chapter 8 addresses this issue in detail.
The outcomes reviewed in this chapter follow a common format. Each section begins by providing some background information about the cancer under discussion, including data concerning its incidence in the general U.S. population. A brief summary of the scientific evidence described in the first two Agent Orange reports—Veterans and Agent Orange (1994; hereafter referred to as VAO), and Veterans and Agent Orange: Update 1996 (hereafter, Update 1996)—is then presented, followed by a discussion of the most recent scientific literature, and a synthesis of the material reviewed. Where appropriate, reviews are separated by the type of exposure (occupational, environmental, Vietnam veteran) being addressed. Each section concludes with the committee's finding regarding the strength of the evidence in epidemiologic studies, biologic plausibility, and evidence regarding Vietnam veterans.
The Department of Veterans Affairs (DVA) asked the committee to specifically address the classification of chondrosarcomas of the skull as part of its work. This is done in the discussion of bone cancer below.
Expected Number of Cancer Cases Among Vietnam Veterans in the Absence of Any Increase in Risk Due to Herbicide Exposure
To provide some background for the consideration of cancer risks in Vietnam veterans, this chapter also reports information on cancer incidence in the general U.S. population. Incidence rates are reported for individuals between the ages of 45 and 59 because most Vietnam era veterans are in this age group. The data, which were collected as part of the Surveillance, Epidemiology, and End Results (SEER) Program of the National Center for Health Statistics (NCHS), are categorized by sex, age, and race because these factors can have a profound effect on the estimated level of risk. Prostate cancer incidence, for example, is 14 times higher in men age 55-59 than in 45-49 year olds and nearly twice as high in African Americans age 45-59 than in whites of this age group (NCI, 1998). The figures presented for each cancer are estimates for the entire U.S. population, not precise predictions for the Vietnam veteran cohort. It should be remembered that numerous factors may influence the incidence reported here—including personal behavior (e.g., smoking and diet), genetic predisposition, and other risk factors such as medical history. These factors may make a particular individual more or less likely than average to contract a given cancer. Incidence data are reported for all races and also separately for African Americans and whites. The data reported are for the years 1990-1994, the most recent available at the time this report was written.
Given the large uncertainties that remain about the magnitude of potential risk from exposure to herbicides in the occupational, environmental, and veteran studies that have been reviewed, inadequate control for important confounders in these studies, and the lack of information needed to extrapolate from the level of exposure in the studies to that of individual Vietnam veterans, the committee cannot quantify the degree of risk likely to have been experienced by Vietnam veterans due to exposure to herbicides in Vietnam.
GASTROINTESTINAL TRACT TUMORS
Background
As a group, the category of gastrointestinal (GI) tract tumors includes some of the most common cancers in the United States and the world. The committee reviewed the data on stomach cancer (ICD·9 151.0-151.9), colon cancer (ICD·9 153.0-153.9), rectal cancer (ICD·9 154.0-154.1), and pancreatic cancer (ICD·9 157.0-157.9). According to American Cancer Society estimates, approximately 183,000 individuals will be diagnosed with these cancers in the United States in 1998 and some 99,000 individuals will die from them (ACS, 1998). Colon cancer accounts for about half of these diagnoses and deaths. The cases are divided approximately equally between men and women. Collectively, GI tract tumors are expected to account for 15 percent of new diagnoses and 18 percent of cancer deaths in 1998.
Average Annual Cancer Incidence (per 100,000 individuals) in the United Statesa Selected Gastrointestinal Cancers |
|||||||||
|
45-49 years of age |
50-54 years of age |
55-59 years of age |
||||||
|
all races |
white |
black |
all races |
white |
black |
all races |
white |
black |
Stomach |
|||||||||
males |
6 |
5 |
10 |
12 |
10 |
23 |
20 |
17 |
39 |
females |
3 |
2 |
4 |
5 |
4 |
7 |
9 |
7 |
14 |
Colon |
|||||||||
males |
18 |
16 |
27 |
37 |
35 |
49 |
69 |
68 |
92 |
females |
16 |
14 |
21 |
28 |
24 |
52 |
15 |
49 |
77 |
Rectal |
|||||||||
males |
9 |
9 |
15 |
15 |
15 |
20 |
21 |
20 |
23 |
females |
7 |
6 |
11 |
7 |
7 |
6 |
12 |
12 |
14 |
Pancreatic |
|||||||||
males |
5 |
5 |
10 |
12 |
11 |
23 |
21 |
20 |
40 |
females |
4 |
3 |
7 |
8 |
8 |
10 |
14 |
13 |
25 |
a SEER nine standard registries crude, age-specific rate, 1990-1994. |
The incidence of stomach, colon, rectal, and pancreatic cancers increases with age for individuals between 45 and 59 years of age. In general, incidence in males is higher than in females, and incidence in African Americans exceeds that
of whites. Risk factors besides age and race vary for these cancers, but always include family history of the same form of cancer, certain diseases of the affected organ, and dietary factors. Cigarette smoking is a risk factor for pancreatic cancer and may also increase the risk of stomach cancer (Miller et al., 1996). Infection with the bacterium Helicobacter pylori also increases the risk of stomach cancer.
Summary ofVAO andUpdate 1996
Numerous studies were considered in VAO and Update 1996 that examined one or more gastrointestinal tract cancers; no consistent associations between herbicide exposure and these cancers were found. These included studies of chemical production workers in the United States and other countries (Lynge, 1985; Coggon et al., 1986; Thomas, 1987; Bond et al., 1988; Zober et al., 1990; Fingerhut et al., 1991; Manz et al., 1991; Saracci et al., 1991; Bloemen et al., 1993; Bueno de Mesquita et al., 1993; Collins et al., 1993; Kogevinas et al., 1993); agricultural workers (Burmeister, 1981; Hardell, 1981; Burmeister et al., 1983; Wiklund, 1983; Hoar et al., 1986; Alavanja et al., 1988, 1989; Wigle et al., 1990; Hansen et al., 1992; Ronco et al., 1992; Blair et al., 1993; Garry et al., 1994; Asp et al., 1994); pesticide appliers (Axelson et al., 1980; Blair et al., 1983; Swaen et al., 1992); paper and pulp workers (Robinson et al., 1986; Henneberger et al., 1989; Solet et al., 1989); the population of Seveso, Italy (Bertazzi et al., 1989a,b; Pesatori et al., 1992; Bertazzi et al., 1993); others subjected to environmental exposure (Lampi et al., 1992); and Vietnam veterans (Kogan and Clapp, 1985; Lawrence et al., 1985; Anderson et al., 1986a,b; Boyle et al., 1987; Breslin et al., 1988; CDC, 1988; Dalager et al., 1995a; Visintainer et al., 1995).
VAO and Update 1996 found that, with rare exceptions, studies of GI cancers and herbicide exposure reported relative risks (RRs) close to 1.0, providing no evidence of any increased risk. They concluded that there is limited/suggestive evidence of no association between exposure to the herbicides (2,4-dichlorophenoxyacetic acid [2,4-D]; 2,4,5-trichlorophenoxyacetic acid [2,4,5-T] and its contaminant TCDD; cacodylic acid; and picloram) and GI cancers (stomach, pancreatic, rectal, and colon cancers). The evidence regarding association is drawn from occupational and other studies in which subjects were exposed to a variety of herbicide and herbicide components.
Update of the Scientific Literature
Occupational Studies
In an update and expansion of the International Agency for Research on Cancer (IARC) cohort study, Kogevinas et al. (1997) examined cancer mortality in a cohort of 26,615 male and female workers engaged in the production or application of phenoxy herbicides. The workers for this cohort were assembled
from 12 countries, drawn from national studies that followed the same core protocol developed jointly by the participants and coordinated by IARC.
No excess risk of death from GI cancer of any site was observed among the group comprised of all workers exposed to any phenoxy herbicide or chlorophenol (stomach: standardized mortality ratio [SMR] = 0.9, 95 percent confidence interval [95% CI] 0.7-1.1, 72 deaths; colon: SMR = 1.1, CI 0.8-1.3, 86 deaths; rectum: SMR = 1.1, CI 0.8-1.4, 44 deaths; pancreas: SMR = 0.9, CI 0.7-1.2, 47 deaths). When this group was divided into those exposed and unexposed to TCDD or higher chlorinated dioxins, slight elevations were seen for certain cancers, but they were found in both the exposed and the unexposed groups, and none achieved statistical significance. More detailed analysis by exposure variables such as duration and time since first exposure was not conducted for GI cancers.
Although the study includes large numbers of workers who were likely to be exposed at levels substantially higher than general population exposures, the lack of information about actual exposures limits the investigator's ability to examine exposure-response relationships within the cohort. In addition, the inclusion of workers in the exposed group based on ever having worked in a job considered exposed makes it impossible to distinguish heavily exposed workers from those with very minor exposures.
Becher et al. (1996) examined cancer mortality among workers in four German facilities that produced phenoxy herbicides and chlorophenols. The population included workers who had a least one month of employment, resulting in a cohort consisting of 2,479 male workers. The cohort was assembled from four plants; analysis was conducted on the total cohort, divided into four subcohorts that corresponded to each plant considered separately.
Based on production information and limited blood dioxin measurements, subcohorts I and II are supposed to have higher TCDD exposures than subcohorts III and IV. Of the four subcohorts, only group I had at least one observed or expected death at each of the GI cancer sites. Of these, stomach (SMR = 1.3, 95% CI 0.7-2.2, 12 cases) and rectum (SMR = 1.9, CI 0.7-4.0, 6 cases) were nonsignificantly elevated. The small numbers in the other three subcohorts resulted in few reportable results for GI cancers, and none of these demonstrated a significant excess risk.
Ott and Zober (1996) updated the experience of workers exposed to TCDD during the cleanup of a trichlorophenol (TCP) reactor, which exploded in 1953 at a BASF plant in Ludwigshafen, Germany. They studied cancer incidence and mortality up to 1992 for the group of 243 men and developed TCDD dose estimates based on work activity information, blood TCDD determinations on a subset of the population, and estimates of TCDD elimination rates. The analysis of cancer mortality showed a nonsignificant elevation of death for cancer of the digestive organs in the total cohort (SMR = 1.2, CI 0.6-2.1, 11 cases) with the lowest risk (SMR = 0.6) observed in the lowest-dose group (<0.1 µg/kg
body weight). The medium-dose group (0.1-0.99 µg/kg body weight) reported an SMR of 1.7 (CI 0.5-4.3), whereas the highest dose group (>1 µg/kg body weight) had an SMR of 1.5 (CI 0.5-3.4). Incident cancer cases showed a similar pattern, with a nonsignificant elevation of cancer of the digestive tract in the total cohort (standardized incidence ratio [SIR] = 1.1, CI 0.6-1.9, 12 cases). The lowest risk (SIR = 0.7) was observed in the lowest-dose group (<0.1 µg/kg body weight). The medium-dose group (0.1-0.99 µg/kg body weight) reported an SIR of 1.4 (CI 0.4-3.6), whereas the highest-dose group (>1 µg/kg body weight) showed an SIR of 1.2 (CI 0.4-2.9). Separate analysis for incidence of stomach cancer within this class revealed three cases overall (SIR = 1.0, CI 0.2-2.9), with no cases observed in the lowest-dose group, one case in the medium-dose group (SIR = 1.3, CI 0.0-7.0), and two cases in the highest-dose group (SIR = 1.7, CI 0.2-6.2).
Internal analysis of the cohort by proportional hazard analysis showed that TCDD dose was significantly associated with both death from cancer of the digestive tract (conditional regression ratio 1.5, CI 1.1-1.9) and incidence of digestive cancer (conditional risk ratio 1.4, CI 1.1-1.7). Review of these cases revealed that three had occurred among the four most highly exposed workers in the cohort and that one (stomach cancer) involved a worker with an estimated TCDD dose of 6.8 µg/kg and a second (pancreatic cancer), with an estimated dose of 6.1 µg/kg. However, information provided for the third case classified as a digestive cancer raises concerns. This case was a primary liver cancer with a dose estimate of 8.3 µg/kg. It is unclear why this case was not classified as a hepatobiliary cancer. Had it been, its exclusion from the digestive cancer analysis would probably have weakened the reported association in the proportional hazard analysis.
Ramlow et al. (1996) examined mortality in a cohort of workers exposed to pentachlorophenol (PCP), as part of a larger study of Dow chemical manufacturing workers exposed to the higher chlorinated dioxins. The study cohort was assembled from company records, starting with a cohort of 2,192 workers ever employed in a department with potential polychlorinated dibenzodioxin (PCDD) exposure between 1937 and 1980.
In the study analysis, U.S. white male death rates (five-year age and calendar specific) and the non-PCP and PCDD male Dow Michigan employees for 1940 to 1989 were both used as reference values to calculate expected deaths. Four exposure groups were developed for TCDD (1 unit = very low, 1-1.9 units = low, 2-2.9 units = medium, 3 units = high). Calculation of SMRs with exposure lagged by 15 years using both the U.S. and the Dow referent populations found no significant excess mortality for digestive cancer (SMR = 0.9, CI 0.4-1.6, 10 deaths). The vast majority of deaths observed were among the unexposed group, leaving very few deaths distributed over the four categories of cumulative exposure (unexposed, 517 deaths from digestive cancer; very low exposure, 1 death; low exposure, 5 deaths; medium exposure, 4 deaths; high exposure, none).
For the hepta- or octa-chlorinated dibenzodioxin (H/OCDD) groups, 2 units was the low-exposure category; 2-2.9, units medium; and 3 or more units, high. SMRs for cancers of the digestive system (15-year lagged exposure) showed no significant excess compared to the U.S. and Dow referent populations. As for TCDD, a very small proportion of the population was considered to have any exposure, and very few cases were observed.
Gambini et al. (1997) investigated cancer mortality among a cohort of rice growers in northern Italy. Using a set of registered farm owners consisting of 1,493 males who worked on farms from 1957 to 1992, they examined the cause of death for 958 subjects and compared this with expected numbers calculated from national rates. No direct exposure information was available, so employment on the farm was used as a surrogate for exposure to the range of phenoxy herbicides employed during the study period. Cancer mortality was evaluated for three GI sites (stomach, pancreas, and intestines), and observed and expected deaths did not differ significantly for any of these sites in the overall cohort (stomach: SMR = 0.9, CI 0.7-1.3, 39 deaths; pancreas: SMR = 0.9, CI 0.4-1.9, 7 deaths; intestines: SMR = 1.1, CI 0.7-1.6, 27 deaths). Stratification by age at death and duration of exposure (employment as a farmer) did not change the finding of nonsignificant differences. Although the study population is small, it does describe the experience of a cohort with good follow-up (99 percent) and long latency (37 percent of deaths observed beyond the age of 80). It is limited by very crude exposure assessment, however, and the degree to which the study subjects were actually exposed to phenoxy herbicides cannot be established with any certainty.
Environmental Studies
Bertazzi et al. (1997) continued the follow-up of people environmentally exposed to TCDD in Seveso, Italy. The events that led to the exposure and the methods used to study this population have been described fully in the earlier reports. This report updates the population after 15 years follow-up. Death from cancer of the rectum was significantly elevated for men in zone B (SMR = 2.9, CI 1.2-5.9, 7 observed deaths). No other significant elevation of death from digestive cancer overall or at any GI sites was observed in men or women in any exposure zone. More detailed investigation of subjects in zone B showed that there were nonsignificant elevations of death from digestive cancers overall for women in the longest-latency (>10 years; overall digestive cancer: SMR = 1.5, CI 0.7-2.7, 10 deaths; stomach: SMR = 2.4, CI 0.8-5.7, 5 deaths) and length of stay (>10 years; overall digestive cancer: SMR = 1.6, CI 0.8-2.9, 9 deaths; stomach: SMR = 2.3, CI 0.6-6.0, 4 deaths) groups, whereas men showed significant excesses of rectal cancer in the group with the longest length of stay (SMR = 7.2, CI 1.9-18.4, 4 deaths) and the longest-latency group (SMR = 6.2, CI 1.7-15.9, 4 deaths).
Svensson et al. (1995) studied mortality and cancer incidence in two cohorts of Swedish fishermen. One group (2,896 men) resided on the east coast of Swe-
den and consumed fish from the Baltic Sea. These fatty fish (particularly salmon and herring) are reported to contain elevated levels of polychlorinated biphenyls (PCBs), polychlorinated dibenzofurans (PCDFs), and PCDDs. The other group of fishermen (8,477) resided on the west coast of Sweden and was presumed to have a higher intake of lean (and less contaminated) fish, including cod and flat fish. This distinction in exposure by place of residence is reportedly confirmed by the finding that blood levels of ''dioxin-like compounds" were two times higher among east coast than west coast fishermen; however, no data are provided to support this point. East coast fishermen were found to have nonsignificantly increased mortality from stomach cancer (SMR = 1.4, CI 0.8-2.2, 17 deaths) and increased incidence of stomach cancer (SIR = 1.6, CI 1.0-2.4, 24 cases) compared to Swedish national rates. No significant excess incidence or mortality was seen among west coast fishermen. When the two groups were compared directly, the east coast excess incidence of stomach cancer was 2.2 (incidence rate ratio [IRR], CI 1.3-3.5). East coast fishermen were found to have significantly decreased incidence and mortality of cancer of the colon, which was not observed in west coast fishermen. The degree to which the difference in stomach cancer between east and west coast fishermen can be attributed to organochlorine exposure is limited by the lack of any direct information on exposure, aside from the above (unreferenced) statement that blood levels of dioxin-like compounds were twice as high among east coast as west coast fishermen. The authors also report that east coast fishermen consume smoked fish twice as often as west coast fishermen, so the role of confounding exposures cannot be discounted.
Vietnam Veteran Studies
In a comparison of mortality between Army Chemical Corps Vietnam and non-Vietnam veterans, Dalager et al. (1997) reported a significant excess of death from digestive diseases, primarily cirrhosis, among Vietnam veterans. The study compared 2,872 Vietnam veterans with 2,737 non-Vietnam veterans (all of whom served in Chemical Corps specialties). All study subjects served at least 18 months' active duty between 1965 and 1973, and vital status ascertainment was complete for both groups. Nonsignificant decreases in deaths from digestive system cancer were observed in both groups when compared to general U.S. population rates. When Vietnam and non-Vietnam cohorts were compared directly, the crude rate ratio of GI cancer death was 4.8 (Vietnam versus non-Vietnam). The adjusted RR calculated by proportional hazards modeling to include the effect of race, military rank, duration of service, and age at entry to follow-up was 2.2 (CI 0.2-19.8). Direct exposure information on the two cohorts was not available, and the presumption that Vietnam veterans had potential for higher levels of dioxin exposure because of their duties involving Agent Orange and other dioxin-contaminated herbicides (compared to non-Vietnam Chemical Corps veterans) has not been verified.
The Australian study Mortality of Vietnam Veterans: The Veteran Cohort Study (Crane et al., 1997a) examined the mortality experience of male Australian Vietnam veterans from 1980 to 1994. This cohort consists of 59,036 male veterans, who were followed for a period ranging from 22 to 32 years. There were 2067 deaths recorded among this group from 1980 to 1994, and vital status was ascertained for 96.9 percent of the cohort. There was a statistically significant excess of death for all cancer (SMR = 1.2, CI 1.1-1.3) in the cohort, by comparison to the Australian white male population and by calculation of a standardized relative mortality ratio (SRMR), which is the ratio of the cause-specific SMR and the SMR for all other causes combined (SRMR = 1.2, CI 1.1-1.3). No excess mortality was observed from cancer at any of the four GI sites. Death from cancer of the colon (SMR = 1.2, CI 1.0-1.5), rectum (SMR = 0.6, CI 0.4-1.0), stomach (SMR = 1.1, CI 0.7-1.5), and pancreas (SMR = 1.4, CI 1.0-1.9) did not exceed expected numbers for all military Vietnam veterans, and when analyzed separately by branch of service, only Navy veterans were reported to have excess mortality for any GI site (colon cancer: SMR = 1.8, CI 1.0-2.8), which was not statistically significant. The study authors have described the strengths and limitations of this cohort study of Australian veterans, including virtually complete identification of the study population, a period of follow-up ranging from 22 to 32 years, and vital status ascertainment of 96.9 percent. Among the weaknesses of the study are the possibility of underascertainment of death and the uncertain quality of exposure assessment regarding a variety of risk factors, including smoking and alcohol consumption, as well as herbicide and dioxin exposure. The examination of mortality among Australian National Service Vietnam veterans (Crane et al., 1997b) reported similar findings for GI cancer.
Synthesis
With rare exceptions, studies on GI cancers and exposure to herbicide in production, in agricultural use, from environmental sources, and among veteran populations found RRs close to 1.0, providing no evidence of any increase in risk. A reported statistically significant positive association in the BASF cohort is rendered problematic by the inclusion of a liver cancer in the analysis. The positive association reported for rectal cancer in males in zone B of the Seveso cohort is not by itself compelling, and is not supported by findings in any other zone or for females in the cohort.
Conclusions
Strength of Evidence in Epidemiologic Studies
This report, like its predecessors, concludes that there is limited/suggestive evidence of no association between exposure to the herbicides (2,4-D, 2,4,5-T
TABLE 7-1 Selected Epidemiologic Studies—Stomach Cancer
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
OCCUPATIONAL |
|||
New Studies |
|||
Kogevinas et al., 1997 |
IARC cohort |
|
|
|
Workers exposed to TCDD (or higher chlorinated dioxins) |
42 |
0.9 (0.6-1.2) |
|
Workers not exposed to TCDD (or higher chlorinated dioxins) |
30 |
0.9 (0.6-1.3) |
|
Workers exposed to any phenoxy herbicide or chlorophenol |
72 |
0.9 (0.7-1.1) |
Becher et al., 1996 |
German chemical production workers |
|
|
|
Plant I |
12 |
1.3 (0.7-2.2) |
|
Plant II |
0 |
|
|
Plant III |
0 |
|
|
Plant IV |
2 |
0.6 (0.1-2.3) |
Gambini et al., 1997 |
Italian rice growers |
39 |
0.9 (0.7-1.3) |
Ott and Zober, 1996 |
BASF cleanup workers |
3 |
1.0 (0.2-2.9) |
|
TCDD <0.1 µg/kg body wt |
0 |
|
|
TCDD 0.1-0.99 µg/kg body wt |
1 |
1.3 (0.0-7.0) |
|
TCDD >1 µg/kg body wt |
2 |
1.7 (0.2-6.2) |
Ramlow et al., 1996 |
Pentachlorophenol production workers |
|
|
|
0 year latency |
4 |
1.7 (0.4-4.3) |
|
15 year latency |
3 |
1.8 (0.4-5.2) |
Studies reviewed in Update 1996 |
|||
Blair et al., 1993 |
U.S. farmers in 23 states |
|
|
|
White males |
657 |
1.0 (1.0-1.1) |
|
Nonwhite females |
23 |
1.9 (1.2-2.8) |
Bueno de Mesquita et al., 1993 |
Phenoxy herbicide workers |
|
NS |
Collins et al., 1993 |
Monsanto 2,4-D production workers |
|
NS |
Kogevinas et al., 1993 |
Female herbicide spraying and production workers |
|
NS |
Studies reviewed in VAO |
|||
Ronco et al., 1992 |
Danish male self-employed farm workers |
286 |
0.9 |
Swaen et al., 1992 |
Dutch herbicide applicators |
1 |
0.5 (0-2.7)c |
Fingerhut et al., 1991 |
NIOSH cohort |
10 |
1.0 (0.5-1.9) |
Manz et al., 1991 |
German production workers |
12 |
1.2 (0.6-2.1) |
Saracci et al., 1991 |
IARC cohort |
40 |
0.9 (0.6-1.2) |
Wigle et al., 1990 |
Canadian farmers |
246 |
0.9 (0.8-1.0) |
Zober et al., 1990 |
BASF production workers—basic cohort |
3 |
3.0 (0.8-11.8) |
Alavanja et al., 1989 |
USDA forest/soil conservationists |
9 |
0.7 (0.3-1.3) |
Henneberger et al., 1989 |
Paper and pulp workers |
5 |
1.2 (0.4-2.8) |
Solet et al., 1989 |
Paper and pulp workers |
1 |
0.5 (0.1-3.0) |
Alavanja et al., 1988 |
USDA agricultural extension agents |
10 |
0.7 (0.4-1.4) |
Bond et al., 1988 |
Dow 2,4-D production workers |
0 |
— (0.0-3.7) |
Thomas, 1987 |
Flavor and fragrance chemical production workers |
|
1.4 |
Coggon et al., 1986 |
British MCPA production workers |
26 |
0.9 (0.6-1.3) |
Robinson et al., 1986 |
Paper and pulp workers |
17 |
1.2 (0.7-2.1) |
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
Lynge, 1985 |
Danish male production workers |
12 |
1.3 |
Blair, 1983 |
Florida pesticide applicators |
4 |
1.2 |
Burmeister et al., 1983 |
Iowa residents |
|
|
|
Farming exposures |
|
1.3 (p < .05) |
Wiklund, 1983 |
Swedish agricultural workers |
2,599 |
1.1 (1.0-1.2)b |
Burmeister, 1981 |
Farmers in Iowa |
338 |
1.1 (p < .01) |
Axelson et al., 1980 |
Swedish railroad workers—total exposure |
3 |
2.2 |
ENVIRONMENTAL |
|||
New Studies |
|||
Bertazzi et al., 1997 |
Seveso residents |
|
|
|
Males—zone A |
0 |
|
|
Males—zone B |
10 |
0.8 (0.4-1.5) |
|
Males—zone R |
76 |
0.9 (0.7-1.1) |
|
Females—zone A |
1 |
0.9 (0.0-5.3) |
|
Females—zone B |
7 |
1.0 (0.4-2.1) |
|
Females—zone R |
58 |
1.0 (0.8-1.3) |
Svensson et al., 1995 |
Swedish fishermen mortality |
|
|
|
East coast |
17 |
1.4 (0.8-2.2) |
|
West coast |
63 |
0.9 (0.7-1.2) |
|
Swedish fishermen incidence |
|
|
|
East coast |
24 |
1.6 (1.0-2.4) |
|
West coast |
71 |
0.9 (0.7-1.2) |
Studies reviewed in Update 1996 |
|||
Bertazzi et al., 1993 |
Seveso male residents—zone B |
7 |
1.0 (0.5-2.1) |
|
Female residents—zone B |
2 |
0.6 (0.2-2.5) |
|
Seveso male residents—zone R |
45 |
0.9 (0.7-1.2) |
|
Female residents—zone R |
25 |
1.0 (0.6-1.5) |
Studies reviewed in VAO |
|||
Pesatori et al., 1992 |
Seveso male residents—zones A and B |
7 |
0.9 (0.4-1.8) |
|
Female residents—zones A and B |
3 |
0.8 (0.3-2.5) |
Bertazzi et al., 1989a |
Seveso male residents—zones A, B, R |
40 |
0.8 (0.6-1.2) |
|
Female residents—zones A, B, R |
22 |
1.0 (0.6-1.5) |
Bertazzi et al., 1989b |
Seveso male residents—zone B |
7 |
1.2 (0.6-2.6) |
VIETNAM VETERANS |
|||
New Studies |
|||
Crane et al., 1997a |
Australian military veterans |
32 |
1.1 (0.7-1.5) |
Crane et al., 1997b |
Australian national service veterans |
4 |
1.7 (0.3->10) |
Studies reviewed in VAO |
|||
Breslin et al., 1988 |
Army Vietnam veterans |
88 |
1.1 (0.9-1.5) |
|
Marine Vietnam veterans |
17 |
0.8 (0.4-1.6) |
Anderson et al., 1986a |
Wisconsin Vietnam veterans |
3 |
— |
Anderson et al., 1986b |
Wisconsin Vietnam veterans |
1 |
— |
a Given when available. b 99% CI. c Risk estimate is for stomach and small intestine. |
TABLE 7-2 Selected Epidemiologic Studies—Colon Cancer
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
OCCUPATIONAL |
|||
New Studies |
|||
Kogevinas et al., 1997 |
IARC cohort |
|
|
|
Workers exposed to TCDD (or higher chlorinated dioxins) |
52 |
1.0 (0.8-1.3) |
|
Workers not exposed to TCDD (or higher chlorinated dioxins) |
33 |
1.2 (0.8-1.6) |
|
Workers exposed to any phenoxy herbicide or chlorophenol |
86 |
1.1 (0.8-1.3) |
Becher et al., 1996 |
German chemical production workers |
|
|
|
Plant I |
2 |
0.4 (0.0-1.4) |
|
Plant II |
0 |
|
|
Plant III |
1 |
2.2(0-12) |
|
Plant IV |
0 |
|
Gambini et al., 1997 |
Italian rice growers |
27 |
1.1 (0.7-1.6) |
Ott and Zober, 1996 |
BASF cleanup workers |
5 |
1.0 (0.3-2.3)b |
|
TCDD < 0.1 µg/kg body wt |
2 |
1.1 (0.1-3.9)b |
|
TCDD 0.1-0.99 µg/kg body wt |
2 |
1.4 (0.2-5.1)b |
|
TCDD > 1 µg/kg body wt |
1 |
0.5 (0.0-3.0)b |
Ramlow et al., 1996 |
Pentachlorophenol production workers |
|
|
|
0 year latency |
4 |
0.8 (0.2-2.1) |
|
15 year latency |
4 |
1.0 (0.3-2.6) |
Studies reviewed in Update 1996 |
|||
Blair et al., 1993 |
U.S. farmers in 23 states |
|
|
|
White males |
2,291 |
1.0 (0.9-1.0) |
Bueno de Mesquita et al., 1993 |
Phenoxy herbicide workers |
|
NS |
Collins et al., 1993 |
Monsanto 2,4-D production workers |
|
NS |
Studies reviewed in VAO |
|||
Swaen et al., 1992 |
Dutch herbicide applicators |
4 |
2.6 (0.7-6.5) |
Ronco et al., 1992 |
Danish male self-employed farm workers |
277 |
0.7 (p < .05) |
Fingerhut et al., 1991 |
NIOSH cohort |
25 |
1.2 (0.8-1.8) |
Manz et al., 1991 |
German production workers |
8 |
0.9 (0.4-1.8) |
Saracci et al., 1991 |
IARC cohort |
41 |
1.1 (0.8-1.5) |
Zober et al., 1990 |
BASF production workers—basic cohort |
2 |
2.5 (0.4-14.1)b |
Alavanja et al., 1989 |
USDA forest conservationists |
|
1.4 (0.7-2.8) |
|
USDA soil conservationists |
|
1.2 (0.7-2.0) |
Henneberger et al., 1989 |
Paper and pulp workers |
9 |
1.0 (0.5-2.0) |
Solet et al., 1989 |
Paper and pulp workers |
7 |
1.5 (0.6-3.0) |
Alavanja et al., 1988 |
USDA agricultural extension agents |
|
1.0 (0.7-1.5) |
Bond et al., 1988 |
Dow 2,4-D production workers |
4 |
2.1 (0.6-5.4) |
Thomas, 1987 |
Flavor and fragrance chemical production workers |
|
0.6 |
Coggon et al., 1986 |
British MCPA production workers |
19 |
1.0 (0.6-1.6) |
Hoar et al., 1986 |
Kansas residents |
|
|
|
No herbicide use |
|
1.6 (0.8-3.6) |
|
Herbicide use |
|
1.5 (0.6-4.0) |
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
Robinson et al., 1986 |
Paper and pulp workers |
7 |
0.4 (0.2-0.9) |
Lynge, 1985 |
Danish male production workers |
10 |
1.0 |
Blair, 1983 |
Florida pesticide applicators |
5 |
0.8 |
Wiklund, 1983 |
Swedish agricultural workers |
1,332 |
0.8 (0.7-0.8)c |
Thiess et al., 1982 |
BASF production workers |
|
0.4 |
Burmeister, 1981 |
Farmers in Iowa |
1,064 |
1.0 (NS) |
Hardell, 1981 |
Residents of Sweden |
|
|
|
Exposed to phenoxy acids |
11 |
1.3 (0.6-2.8) |
|
Exposed to chlorophenols |
6 |
1.8 (0.6-5.3) |
ENVIRONMENTAL |
|||
New studies |
|||
Bertazzi et al., 1997 |
Seveso male residents |
|
|
|
zone A |
0 |
|
|
zone B |
5 |
0.8 (0.3-2.0) |
|
zone R |
34 |
0.8 (0.6-1.1) |
|
Seveso female residents |
|
|
|
zone A |
2 |
2.6 (0.3-9.4) |
|
zone B |
3 |
0.6 (0.1-1.8) |
|
zone R |
33 |
0.8 (0.6-1.1) |
Svensson et al., 1995 |
Swedish fishermen mortality |
|
|
|
East coast |
4 |
0.1 (0.0-0.7) |
|
West coast |
58 |
1.0 (0.8-1.3) |
|
Swedish fishermen incidence |
|
|
|
East coast |
5 |
0.4 (0.1-0.9) |
|
West coast |
82 |
0.9 (0.8-1.2) |
Studies reviewed in Update 1996 |
|||
Bertazzi et al., 1993 |
Seveso male residents—zone B |
2 |
0.5 (0.1-2.0) |
|
Female residents—zone B |
2 |
0.6 (0.1-2.3) |
|
Seveso male residents—zone R |
32 |
1.1 (0.8-1.6) |
|
Female residents—zone R |
23 |
0.8 (0.5-1.3) |
Studies reviewed in VAO |
|||
Lampi et al., 1992 |
Finnish community exposed to chlorophenol contamination |
9 |
1.1 (0.7-1.8) |
Bertazzi et al., 1989a |
Seveso male residents—zones A, B, R |
20 |
1.0 (0.6-1.5) |
|
Female residents—zones A, B, R |
12 |
0.7 (0.4-2.2) |
Pesatori et al., 1992 |
Seveso male residents—zones A and B |
3 |
0.6 (0.2-1.9) |
|
Female residents—zones A and B |
3 |
0.7 (0.2-2.2) |
VIETNAM VETERANS |
|||
New studies |
|||
Crane et al., 1997a |
Australian military veterans |
78 |
1.2 (1.0-1.5) |
Crane et al., 1997b |
Australian national service veterans |
6 |
0.6 (0.2-1.5) |
Studies reviewed in Update 1996 |
|||
Dalager et al., 1995a |
Women Vietnam veterans |
|
2.8 (0.8-10.2) |
|
Nurses |
|
5.7 (1.2-27.0) |
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
Studies reviewed in VAO |
|||
Breslin et al. 1988 |
Army Vietnam veterans |
209 |
1.0 (0.7-1.3)d |
|
Marine Vietnam veterans |
33 |
1.3 (0.7-2.2)d |
Anderson et al., 1986a |
Wisconsin Vietnam veterans |
4 |
— |
Anderson et al., 1986b |
Wisconsin Vietnam veterans |
6 |
1.0 (0.4-2.2) |
a Given when available. b Colon and rectal cancer results are combined in this study. c 99% CI. d Intestinal and other GI cancer results are combined in this study. |
TABLE 7-3 Selected Epidemiologic Studies—Rectal Cancer
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
OCCUPATIONAL |
|||
New studies |
|||
Kogevinas et al., 1997 |
IARC cohort |
|
|
|
Workers exposed to TCDD (or higher chlorinated dioxins) |
29 |
1.3 (0.9-1.9) |
|
Workers not exposed to TCDD (or higher chlorinated dioxins) |
14 |
0.7 (0.4-1.2) |
|
Workers exposed to any phenoxy herbicide or chlorophenol |
44 |
1.1 (0.8-1.4) |
Becher et al., 1996 |
German chemical production workers |
|
|
|
Plant I |
6 |
1.8 (0.7-4.0) |
|
Plant II |
0 |
|
|
Plant III |
0 |
|
|
Plant IV |
1 |
0.9 (0.0-4.9) |
Ramlow et al., 1996 |
Pentachlorophenol production workers |
|
|
|
0 year latency |
0 |
|
|
15 year latency |
0 |
|
Studies reviewed in Update 1996 |
|||
Blair et al., 1993 |
U.S. farmers in 23 states |
|
|
|
White males |
367 |
1.0 (0.9-1.1) |
Bueno de Mesquita et al., 1993 |
Phenoxy herbicide workers |
|
NS |
Studies reviewed in VAO |
|||
Ronco et al., 1992 |
Danish male self-employed farmers |
309 |
0.8 (p < .05) |
Fingerhut et al., 1991 |
NIOSH cohort |
5 |
0.9 (0.3-2.1) |
Saracci et al., 1991 |
IARC cohort |
24 |
1.1 (0.7-1.6) |
Alavanja et al., 1989 |
USDA forest/soil conservationists |
9 |
1.0 (0.5-1.9) |
Henneberger et al., 1989 |
Paper and pulp workers |
1 |
0.4 (0.0-2.1) |
TABLE 7-3 Selected Epidemiologic Studies—Rectal Cancer
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
Alavanja et al., 1988 |
USDA agricultural extension agents |
5 |
0.6 (0.2-1.3) |
Bond et al., 1988 |
Dow 2,4-D production workers |
1 |
1.7 (0.0-9.3) |
Thomas, 1987 |
Flavor and fragrance chemical production workers |
|
2.5 |
Coggon et al., 1986 |
British MCPA chemical workers |
8 |
0.6 (0.3-1.2) |
Lynge, 1985 |
Danish male production workers |
14 |
1.5 |
Blair, 1983 |
Florida pesticide applicators |
2 |
1.0 |
Wiklund, 1983 |
Swedish agricultural workers |
1,083 |
0.9 (0.9-1.0)b |
ENVIRONMENTAL |
|||
New studies |
|||
Bertazzi et al., 1997 |
Seveso male residents |
|
|
|
zone A |
0 |
|
|
zone B |
7 |
2.9 (1.2-5.9) |
|
zone R |
19 |
1.1 (0.7-1.8) |
|
Seveso female residents |
|
|
|
zone A |
0 |
|
|
zone B |
2 |
1.3 (0.1-4.5) |
|
zone R |
12 |
0.9 (0.5-1.6) |
Svensson et al., 1995 |
Swedish fishermen mortality |
|
|
|
East coast |
4 |
0.7 (0.2-1.9) |
|
West coast |
31 |
1.0 (0.7-1.5) |
|
Swedish fishermen incidence |
|
|
|
East coast |
9 |
0.9 (0.4-1.6) |
|
West coast |
59 |
1.1 (0.8-1.4) |
Studies reviewed in Update 1996 |
|||
Bertazzi et al., 1993 |
Seveso male residents—zone B |
3 |
1.4 (0.4-4.4) |
|
Female residents—zone B |
2 |
1.3 (0.3-5.4) |
|
Seveso male residents—zone R |
17 |
1.1 (0.7-1.9) |
|
Female residents—zone R |
7 |
0.6 (0.3-1.3) |
Studies reviewed in VAO |
|||
Pesatori et al., 1992 |
Seveso male residents—zones A and B |
3 |
1.2 (0.4-3.8) |
|
Female residents—zones A and B |
2 |
1.2 (0.3-4.7) |
Bertazzi et al., 1989a |
Seveso male residents—zones A, B, R |
10 |
1.0 (0.5-2.0) |
|
Female residents—zones A, B, R |
7 |
1.2 (0.5-2.7) |
Bertazzi et al., 1989b |
Seveso male residents—zone B |
2 |
1.7 (0.4-7.0) |
VIETNAM VETERANS |
|||
New studies |
|||
Crane et al., 1997a |
Australian military veterans |
16 |
0.6 (0.4-1.0) |
Crane et al., 1997b |
Australian national service veterans |
3 |
0.7 |
Studies reviewed in VAO |
|||
Anderson et al., 1986a |
Wisconsin Vietnam veterans |
1 |
— |
Anderson et al., 1986b |
Wisconsin Vietnam veterans |
1 |
— |
a Given when available. b 99% CI. |
TABLE 7-4 Selected Epidemiologic Studies—Pancreatic Cancer
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
OCCUPATIONAL |
|||
New studies |
|||
Kogevinas et al., 1997 |
IARC cohort |
|
|
|
Workers exposed to TCDD (or higher chlorinated dioxins) |
30 |
1.0 (0.7-1.4) |
|
Workers not exposed to TCDD (or higher chlorinated dioxins) |
16 |
0.9 (0.5-1.4) |
|
Workers exposed to any phenoxy herbicide or chlorophenol |
47 |
0.9 (0.7-1.2) |
Becher et al., 1996 |
German chemical production workers |
|
|
|
Plant I |
2 |
0.6 (0.1-2.3) |
|
Plant II |
0 |
|
|
Plant III |
0 |
|
|
Plant IV |
2 |
1.7 (0.2-6.1) |
Gambini et al., 1996 |
Italian rice growers |
7 |
0.9 (0.4-1.9) |
Ramlow et al., 1996 |
Pentachlorophenol production workers |
|
|
|
0 year latency |
2 |
0.7 (0.1-2.7) |
|
15 year latency |
2 |
0.9 (0.1-3.3) |
Studies reviewed in Update 1996 |
|||
Blair et al., 1993 |
U.S. farmers in 23 states |
|
|
|
White males |
1,133 |
1.1 (1.1-1.2) |
Bueno de Mesquita et al., 1993 |
Phenoxy herbicide workers |
|
NS |
Studies reviewed in VAO |
|||
Ronco et al., 1992 |
Danish self-employed male farm workers |
137 |
0.6 (p < .05) |
Swaen et al., 1992 |
Dutch herbicide applicators |
3 |
2.2 (0.4-6.4) |
Fingerhut et al., 1991 |
NIOSH cohort |
10 |
0.8 (0.4-1.6) |
Saracci et al., 1991 |
NIOSH cohort |
26 |
1.1 (0.7-1.6) |
Alavanja et al., 1989 |
USDA forest conservationists |
|
1.2 (0.4-3.4) |
|
USDA soil conservationists |
|
1.1 (0.5-2.2) |
Henneberger et al., 1989 |
Paper and pulp workers |
9 |
1.9 (0.9-3.6) |
Solet et al., 1989 |
Paper and pulp workers |
1 |
0.4 (0.0-2.1) |
Alavanja et al., 1988 |
USDA agricultural extension agents |
21 |
1.3 (0.8-1.9) |
Thomas, 1987 |
Flavor and fragrance chemical production workers |
|
1.4 |
Coggon et al., 1986 |
British MCPA production workers |
9 |
0.7 (0.3-1.4) |
Robinson et al., 1986 |
Paper and pulp workers |
4 |
0.3 (0.1-1.1) |
Lynge, 1985 |
Danish male production workers |
3 |
0.6 |
Blair, 1983 |
Florida pesticide applicators |
4 |
1.0 |
Wiklund, 1983 |
Swedish agricultural workers |
777 |
0.8 (0.8-0.9)b |
Burmeister, 1981 |
Farmers in Iowa |
416 |
1.1 |
ENVIRONMENTAL |
|||
New studies |
|||
Bertazzi et al., 1997 |
Seveso residents |
|
|
|
Males—zone A |
1 |
1.9 (0.0-10.5) |
|
Males—zone B |
2 |
0.6 (0.1-2.0) |
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
|
Males—zone R |
20 |
0.8 (0.5-1.2) |
|
Females—zone A |
0 |
|
|
Females—zone B |
1 |
0.5 (0.0-3.1) |
|
Females—zone R |
11 |
0.7 (0.4-1.3) |
Svensson et al., 1995 |
Swedish fishermen mortality |
|
|
|
East coast |
5 |
0.7 (0.2-1.6) |
|
West coast |
33 |
0.8 (0.6-1.2) |
|
Swedish fishermen incidence |
|
|
|
East coast |
4 |
0.6 (0.2-1.6) |
|
West coast |
37 |
1.0 (0.7-1.4) |
Studies reviewed in VAO |
|||
Pesatori et al., 1992 |
Seveso male residents—zones A and B |
2 |
1.0 (0.3-4.2) |
|
Female residents—zones A and B |
1 |
1.6 (0.2-12.0) |
Bertazzi et al., 1989b |
Seveso male residents—zone B |
2 |
1.1 (0.3-4.5) |
Bertazzi et al., 1989a |
Seveso male residents—zones A, B, R |
9 |
0.6 (0.3-1.2) |
|
Female residents—zones A, B, R |
4 |
1.0 (0.3-2.7) |
VIETNAM VETERANS |
|||
New studies |
|||
Crane et al., 1997a |
Australian military veterans |
38 |
1.4 (1.0-1.9) |
Crane et al., 1997b |
Australian national service veterans |
6 |
1.5 |
Studies reviewed in Update 1996 |
|||
Visintainer et al., 1995 |
Michigan Vietnam veterans |
|
|
Studies reviewed in VAO |
|||
Thomas et al., 1991 |
Women Vietnam veterans |
5 |
2.7 (0.9-6.2) |
Breslin et al., 1988 |
Army Vietnam veterans |
82 |
0.9 (0.6-1.2) |
|
Marine Vietnam veterans |
18 |
1.6 (0.5-5.8) |
Anderson et al., 1986a |
Wisconsin Vietnam veterans |
6 |
5.5 (2.8-10.9) |
Anderson et al., 1986b |
Wisconsin Vietnam veterans |
4 |
— |
a Given when available. b 99% CI |
and its contaminant TCDD, cacodylic acid, and picloram) and gastrointestinal cancers (stomach, pancreatic, rectal, and colon cancers). The evidence regarding association was drawn from occupational and other studies in which subjects were exposed to a variety of herbicides and herbicide components.
Biologic Plausibility
Although a possible association between these exposures and cancer at gastrointestinal sites is considered plausible given the current knowledge of ways in which dioxin and herbicides affect human systems, the literature reviewed for
this update does not support a change from the previous conclusion of limited/ suggestive evidence of no association. A thorough discussion of biologic plausibility with respect to exposure to TCDD or herbicides and gastrointestinal cancers is contained in Chapter 3; a summary is presented in the conclusion to this chapter.
HEPATOBILIARY CANCERS
Background
This category includes cancers of the liver (ICD·9 155.0,155.2) and intrahepatic bile duct (ICD·9 155.1). According to American Cancer Society estimates, 9,300 men and 4,600 women will be diagnosed with liver cancer in the United States in 1998; 7,900 men and 5,100 women will die from the disease (ACS, 1998). Liver cancer is expected to account for about 1 percent of new diagnoses and 2 percent of cancer deaths in the United States in 1998. This disparity may be due to misclassification of metastatic cancers as primary liver cancer, leading to overreporting of deaths due to liver cancer (Percy et al., 1990). In developing countries, especially sub-Saharan Africa and Southeast Asia, liver cancers are common and are among the leading causes of death. The known risk factors for liver cancer include chronic infection with hepatitis B or hepatitis C virus and exposure to the carcinogens aflatoxin and vinyl chloride. In the general population, the incidence of liver and intrahepatic bile duct cancer increases slightly with age, and remains greater for men than women and greater for African Americans than whites throughout ages 45-59 years.
Average Annual Cancer Incidence (per 100,000 individuals) in the United Statesa Liver and Intrahepatic Bile Duct Cancers |
|||||||||
|
45-49 years of age |
50-54 years of age |
55-59 years of age |
||||||
|
all races |
white |
black |
all races |
white |
black |
all races |
white |
black |
males |
5 |
3 |
11 |
8 |
5 |
15 |
12 |
9 |
19 |
females |
2 |
1 |
3 |
2 |
1 |
4 |
4 |
3 |
4 |
a SEER nine standard registries crude, age-specific rate, 1990-1994. |
Summary of VAO and Update 1996
Occupational Studies
Among studies of occupational groups, mortality studies by Lynge (1985), Collins et al. (1993), Fingerhut et al. (1991), and Saracci et al. (1991) found no elevation of risk. In studies of agricultural or forestry workers, Wiklund (1983), Hardell et al. (1984), Ronco et al. (1992), Asp et al. (1994), and Blair et al. (1993) observed no evidence of increased liver cancer. In a study of mortality among
pulp and paper workers, Solet et al. (1989) observed a nonsignificant excess of deaths from liver cancer, based on two deaths.
Environmental Studies
A follow-up of the population involved in the Seveso incident (Bertazzi et al., 1989b and Bertazzi et al., 1993) reported a statistically significant excess in liver cancer mortality for female residents of zone B (RR = 3.3, CI 1.3-8.1). Nonsignificant excesses and deficits in the risk of death due to liver cancer were observed among males in the three exposure zones and among females living in the other two zones. Additional liver cancer incidence data show similar results (Pesatori et al., 1992). Data from U.S. populations living in contaminated areas do not add any useful information. Residents of the Quail Run trailer park in Times Beach, Missouri, the site of a major release of TCDD-contaminated oil, were free from diagnosed liver cancer, whereas 1.5 cases were expected (Hoffman et al., 1986; Stehr-Green et al., 1987).
A case-control study by Cordier et al. (1993) described the hepatocellular carcinoma risk among 152 North Vietnamese cases and 241 controls, in relation to viral infections and chemical exposures. The dominant risk factor was found to be positivity for hepatitis B surface antigen, which carried an odds ratio (OR) of 0.62 (95% CI 0.30-1.28). Use of organochlorine pesticides did not indicate any statistically significant trend, but military service in South Vietnam did. Those who served for more than 10 years in South Vietnam (N = 11) had an OR for hepatocellular carcinoma of 8.8 (CI 1.9-41). However, direct contact with aerial spraying resulted in a slight, nonsignificantly increased OR of 1.3.
Vietnam Veteran Studies
Studies of liver cancer among Vietnam veterans have not found a significant excess of mortality from liver cancer; however, the studies are hampered by the small size. Studies include those of Wisconsin Vietnam veterans by Anderson et al. (1986a,b), and the mortality component of the Centers for Disease Control and Prevention's (CDC's) Vietnam Experience Study (Boyle et al., 1987). In a larger mortality study among U.S. Army and Marine Corps Vietnam veterans, Breslin et al. (1988) identified 34 liver cancer deaths among Army veterans; the proportional mortality ratio (PMR) was 1.0 (CI 0.8-1.4). The data from Marines are consistent with this result, although there were fewer deaths among the latter group.
The CDC's Selected Cancers Study (CDC, 1990) included a pathologic review of studies to confirm the diagnosis of 130 men with primary liver cancer. After adjusting for design and a range of established risk factors, the RR was 1.2 (CI 0.5-2.7). The risk for Vietnam veterans was slightly lower than for men who served elsewhere in the military.
Update of the Scientific Literature
Occupational Studies
Production Workers In an update and expansion of the IARC cohort study, Kogevinas et al. (1997) examined cancer mortality in a cohort of 26,615 male and female workers engaged in the production or application of phenoxy herbicides. These workers were assembled from 12 countries, drawn from national studies that followed the same core protocol developed jointly by the participants and coordinated by IARC.
No excess risk of death from hepatobiliary cancer was observed among the group of all workers exposed to any phenoxy herbicide or chlorophenol. When this group was divided into those exposed and unexposed to TCDD or higher chlorinated dioxins, the TCDD-exposed group had a higher risk (SMR = 0.87, CI 0.45-1.52) than the unexposed group (SMR = 0.41, CI 0.09-1.22), and none achieved statistical significance. More detailed analysis by exposure variables such as duration and time since first exposure was not conducted for hepatobiliary cancers.
Although the study includes large numbers of workers likely to be exposed at levels substantially higher than the general population exposures, the lack of information about actual exposures limits the investigator's ability to examine exposure-response relationships within the cohort. In addition, the inclusion of workers in the exposed group based on ever having worked in a job considered exposed makes it impossible to distinguish heavily exposed workers from those with very minor exposures.
Becher et al. (1996) examined cancer mortality among workers in four German facilities that produced phenoxy herbicides and chlorophenols. The population included workers who had a least one month of employment, resulting in a cohort consisting of 2,479 male workers. The cohort was assembled from four plants; analysis was conducted on the total cohort, divided into four subcohorts that corresponded to each plant considered separately.
Based on production information and limited blood dioxin measurements, subcohorts I and II are supposed to have higher TCDD exposures than subcohorts III and IV. Of the four subcohorts, only group IV had at least one observed death for hepatobiliary cancer. In this group, one death was observed (0.8 expected) for an SMR of 1.2 (CI 0-6.9).
Agricultural Workers Cancer mortality among a cohort of rice growers in northern Italy was investigated by Gambini et al. (1997). Using a set of registered farm owners consisting of 1,493 males who worked on farms from 1957 to 1992, they examined the cause of death for 958 subjects and compared this with expected numbers calculated from national rates. No direct exposure information was available, so employment on the farm was used as a surrogate for exposure to
the range of phenoxy herbicides employed during the study period. Mortality was evaluated for liver cancer, and observed and expected deaths did not differ significantly in the overall cohort (SMR = 1.3, CI 0.5-2.6). Stratification by age at death and duration of exposure (employment as a farmer) did not change the finding of nonsignificant differences. Although the study population is small, it does describe the experience of a cohort with good follow-up (99 percent) and long latency (37 percent of deaths observed beyond age 80). It is limited by a very crude exposure assessment, however, and the degree to which study subjects were actually exposed to phenoxy herbicides cannot be established with any certainty.
Environmental Studies
Bertazzi et al. (1997) continued the follow-up of the people environmentally exposed to TCDD in Seveso, Italy. The events that led to the exposure and the methods used to study this population have been fully described in earlier reports. This report updates the population after 15 years' follow-up. Death from liver cancer showed nonsignificant decreases in all three exposure groups except for women in zone B, who had a nonsignificant elevation (SMR = 1.3, CI 0.3-3.8, 3 cases). More detailed investigation of exposed subjects in zone B was not conducted for liver cancer.
Svensson et al. (1995) studied mortality and cancer incidence in two cohorts of Swedish fishermen. One group (2,896 men) resided on the east coast of Sweden and consumed fish from the Baltic Sea. These fatty fish (particularly salmon and herring) are reported to contain elevated levels of PCB, PCDD, and PCDF. The other group of fishermen (8,477) resided on the west coast of Sweden and were presumed to have a higher intake of lean (and less contaminated) fish, including cod and flat fish. This distinction in exposure by place of residence is reportedly confirmed by the study's finding that blood levels of dioxin-like compounds were two times higher among east coast than west coast fishermen; however, no data were provided to support this point. East and west coast fishermen were found to have nonsignificantly decreased mortality from liver cancer. East coast fishermen had a nonsignificantly increased incidence of liver cancer (SIR = 1.31, CI 0.48-2.85, 6 cases) compared to national Swedish rates. A nonsignificantly decreased incidence was seen among west coast fishermen.
Vietnam Veteran Studies
The Australian study Mortality of Vietnam Veterans: The Veteran Cohort Study (Crane et al., 1997a) examined the mortality experience of male Australian Vietnam veterans from 1980 to 1994. The cohort consists of 59,036 male veterans, who were followed from 22 to 32 years. There were 2,067 deaths recorded among this group from 1980 to 1994, and vital status was ascertained for 96.9
percent of the cohort. There was a statistically significant excess of death for all cancer (SMR = 1.21, CI 1.11-1.31) in this cohort compared to the Australian white male population and according to the SRMR (1.2, CI 1.1-1.3). No excess mortality was observed from cancer of the liver in the military population overall (SMR = 0.6, CI 0.3-1.2) or when analyzed separately by branch of service. The authors have described the strengths and limitations of the Australian veterans cohort study, including virtually complete identification of the study population, a period of follow-up ranging from 22 to 32 years, and vital status ascertainment of 96.9 percent. Among the weaknesses of the study are the possibility of under-ascertainment of death, and uncertain quality of exposure assessment regarding a variety of risk factors, including smoking and alcohol consumption, as well as herbicide and dioxin exposure.
The examination of mortality among Australian National Service Vietnam veterans (Crane et al., 1997b) reported similar findings for hepatobiliary cancer.
Synthesis
VAO and Update 1996 found that there were relatively few occupational, environmental, or veteran studies of liver cancer (Table 7-5), and most of these are small in size and have not controlled for risk factors related to lifestyle. In summary, given the methodological difficulties associated with most of the few existing studies, the evidence regarding liver cancer was considered inadequate or insufficient to determine whether an association with herbicides exists. The earlier conclusion is unchanged in this report.
Conclusions
Strength of Evidence in Epidemiologic Studies
This committee finds no change in the conclusion of inadequate or insufficient evidence to determine whether an association exists between exposure to the herbicides (2,4-D, 2,4,5-T and its contaminant TCDD, cacodylic acid, and picloram) and hepatobiliary cancer.
The evidence regarding association is drawn from occupational and other studies in which subjects were exposed to a variety of herbicides and herbicide components. Most of these studies are small in size and have not fully controlled for lifestyle-related risk factors.
Biologic Plausibility
Although a possible association between the exposures considered here and hepatobiliary cancer is considered plausible given the current knowledge of ways
TABLE 7-5 Selected Epidemiologic Studies—Hepatobiliary Cancer
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
OCCUPATIONAL |
|||
New studies |
|||
Kogevinas et al., 1997 |
IARC cohort |
|
|
|
Workers exposed to TCDD (or higher chlorinated dioxins) |
12 |
0.9 (0.4-1.5) |
|
Workers not exposed to TCDD (or higher chlorinated dioxins) |
3 |
0.4 (0.1-1.2) |
|
Workers exposed to any phenoxy herbicide or chlorophenol |
15 |
0.7 (0.4-1.2) |
Gambini et al., 1996 |
Italian rice growers |
7 |
1.3 (0.5-2.6) |
Ott and Zober, 1996 |
BASF cleanup workers |
2 |
2.1 (0.3-8.0) |
|
TCDD <0.1 µg/kg body wt |
1 |
2.8 (0.1-15.5) |
|
TCDD 0.1-0.99 µg/kg body wt |
0 |
|
|
TCDD >1 µg/kg body wt |
1 |
2.8 (0.1-15.5) |
Becher et al., 1996 |
German chemical production workers |
1 |
1.2 (0.0-6.9) |
Ramlow et al., 1996 |
Pentachlorophenol production workers |
|
|
|
0 year latency |
0 |
|
|
15 year latency |
0 |
|
Studies reviewed in Update 1996 |
|||
Asp et al., 1994 |
Finnish herbicide applicators |
2 |
0.6 (0.1-2.2) |
Blair et al., 1993 |
U.S. farmers in 23 states |
326 |
1.0 (0.9-1.1) |
Collins et al., 1993 |
Monsanto 2,4-D production workers |
2 |
1.4 (0.2-5.2) |
Studies reviewed in VAO |
|||
Ronco et al., 1992 |
Danish and Italian farm workers |
|
|
|
Danish male self-employed farmers |
23 |
0.4 |
|
Employees of Danish farmers |
9 |
0.8 |
|
Female family workers |
5 |
0.5 |
Fingerhut et al., 1991 |
NIOSH cohort |
6 |
1.2 (0.4-2.5) |
|
20 years latency |
1 |
0.6 (0.0-3.3) |
Saracci et al., 1991 |
IARC cohort |
4 |
0.4 (0.1-1.1) |
Solet et al., 1989 |
Paper and pulp workers |
2 |
2.0 (0.2-7.3) |
Bond et al., 1988 |
Dow 2,4-D production workers |
|
1.2 |
Lynge, 1985 |
Danish production workers |
3 |
1.0 |
Hardell et al., 1984 |
Male residents of northern Sweden |
102 |
1.8 (0.9-4.0) |
Wiklund, 1983 |
Swedish agricultural workers |
103 |
0.3 (0.3-0.4)b |
Zack and Suskind, 1980 |
Monsanto production workers |
0 |
— |
ENVIRONMENTAL |
|||
New studies |
|||
Bertazzi et al., 1997 |
Seveso male residents |
|
|
|
zone A |
0 |
|
|
zone B |
4 |
0.6 (0.2-1.4) |
|
zone R |
35 |
0.7 (0.5-1.0) |
|
Seveso female residents |
|
|
|
zone A |
0 |
|
|
zone B |
4 |
1.1 (0.3-2.9) |
|
zone R |
25 |
0.8 0.5-1.3) |
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
Svensson et al., 1995 |
Swedish fishermen mortality |
|
|
|
East coast |
1 |
0.5 (0.0-2.6) |
|
West coast |
9 |
0.9 (0.4-1.7) |
|
Swedish fishermen incidence |
|
|
|
East coast |
6 |
1.3 (0.5-2.8) |
|
West coast |
24 |
1.0 (0.6-1.5) |
Studies reviewed in Update 1996 |
|||
Bertazzi et al., 1993 |
Seveso male residents—zone B |
5 |
1.8 (0.7-4.4) |
|
Female residents—zone B |
5 |
3.3 (1.3-8.1) |
|
Seveso male residents—zone R |
11 |
0.5 (0.3-1.0) |
|
Female residents—zone R |
12 |
0.9 (0.5-1.7) |
Cordier et al., 1993 |
Military service in South Vietnam for >10 years after 1960 |
11 |
8.8 (1.9-41.0) |
Studies reviewed in VAO |
|||
Pesatori et al., 1992 |
Seveso male residents—zones A and B |
4 |
1.5 (0.5-4.0) |
|
Female residents—zones A and B |
1 |
1.2 (0.2-9.1) |
Bertazzi et al., 1989b |
Seveso male residents—zone B |
3 |
1.2 (0.4-3.8) |
|
Male zone R residents |
7 |
0.4 (0.2-0.8) |
Stehr et al., 1986 |
Missouri residents |
0 |
— |
Hoffman et al., 1986 |
Residents of Quail Run Mobile Home Park |
0 |
— |
VIETNAM VETERANS |
|||
New studies |
|||
Crane et al., 1997a |
Australian military veterans |
8 |
0.6 (0.3-1.2) |
Crane et al., 1997b |
Australian national service veterans |
1 |
— |
Studies reviewed in VAO |
|||
CDC, 1990c |
U.S. men born between 1921 and 1953 |
8 |
1.2 (0.5-2.7) |
Breslin et al., 1988 |
Army Vietnam veterans |
34 |
1.0 (0.8-1.4) |
|
Marine Vietnam veterans |
6 |
1.2 (0.5-2.8) |
Anderson et al., 1986a,b |
Wisconsin Vietnam veterans |
0 |
— |
a Given when available. b 99% CI. |
in which dioxin and herbicides affect human systems, the literature reviewed for this update does not support a change from the previous conclusion of inadequate or insufficient evidence to determine whether an association exists. A thorough discussion of biologic plausibility with respect to exposure to TCDD or herbicides and hepatobiliary cancer is contained in Chapter 3; a summary is presented in the conclusion to this chapter.
Nasal/Nasopharyngeal Cancer
Background
There are many types of nasal (ICD·9 160.0-160.9) and nasopharyngeal (ICD·9 147.0-147.9) cancer, although undifferentiated carcinoma, squamous
cell carcinoma and lymphomas account for the vast majority of maligancies. The epithelium of the nasal and nasopharyngeal cavities is partly squamous, partly columnar, and partly ciliated pseudostratified columnar. There are also serous and mucous glands and lymphoid aggregates in close association with the epithelium.
The American Cancer Society estimates that approximately 4,100 men and 1,200 women will be diagnosed with nasal, pleural, tracheal, and other respiratory system cancers in the United States in 1998 and that some 700 men and 500 women will die from the diseases (ACS, 1998). Roughly speaking, nasal and nasopharyngeal cancers account for between one-third and one-half of these totals. The American Cancer Society (1998) estimates suggest that approximately 6,500 men and 2,100 women will be diagnosed with cancers of the pharynx (including nasopharynx, tonsil, oropharynx, hypopharynx, and buccal cavity) and that 1,500 men and 600 women will die from them. Nasopharyngeal cancers make up approximately one in five of these cancers. The incidence rates reported below show that men are at a greater risk than women for these diseases and that incidence increases with age, although the very small number of cases indicates that care should be exercised in interpreting the numbers.
Nasopharyngeal cancer is relatively common in China and Southeast Asia. It is also more common in Chinese and Vietnamese Americans than in whites, African Americans, or other groups, suggesting that genetic factors may play a role in this disease (Miller et al., 1996). There is no similar association for nasal cancer. Reported risk factors for nasal cancer include occupational exposure to nickel and chromium compounds (Hayes, 1997), wood dust (Demers et al., 1995), and formaldehyde (Blair and Kazerouni, 1997). Studies of nasopharyngeal cancer have reported associations with the consumption of salt-preserved foods (Miller et al., 1996), cigarette smoking (Zhu et al., 1995), and Epstein-Barr virus (Mueller, 1995).
Average Annual Cancer Incidence (per 100,000 individuals) in the United Statesa Nasal and Nasopharyngeal Cancers |
|||||||||
|
45-49 years of age |
50-54 years of age |
55-59 years of age |
||||||
|
all races |
white |
black |
all races |
white |
black |
all races |
white |
black |
Nose, nasal cavity, and mid ear |
|||||||||
males |
1.0 |
0.8 |
1.3 |
1.3 |
1.3 |
1.7 |
1.8 |
1.8 |
2.2 |
females |
0.5 |
0.5 |
1.1 |
0.9 |
0.7 |
1.8 |
0.8 |
0.7 |
2.2 |
Nasopharynx |
|||||||||
males |
1.0 |
0.3 |
1.6 |
2.3 |
1.3 |
2.6 |
2.8 |
1.7 |
3.2 |
females |
0.5 |
0.3 |
0.3 |
0.6 |
0.4 |
0.9 |
0.6 |
1.7 |
|
a SEER nine standard registries crude, age-specific rate, 1990-1994. b Insufficient data to provide meaningful incidence rate. |
Summary ofVAO and Update 1996
Studies that specifically considered nasal and nasopharyngeal cancers were very limited. Occupational studies included phenoxy herbicide production workers and sprayers, herbicide applicators, agricultural workers, and paper and pulp workers. These studies had very few cases of nasal or nasopharyngeal cancers and the results were inconsistent. The sole environmental study was the Seveso follow-up, with no cases reported in the two exposed zones, A and B. Vietnam veteran studies found no significant associations for these cancers.
Update of the Scientific Literature
Occupational Studies
IARC (Kogevinas et al., 1997) has brought together almost all of the phenoxy herbicide production workers in 36 cohorts in 12 countries (Fingerhut et al., 1991; Sarraci et al., 1991; Manz et al., 1991; Flesch-Janys et al., 1995; Becher et al., 1996) for a joint analysis. This cohort contains 26,976 workers and was divided into those who were exposed to TCDD or higher chlorinated dioxins and those who were not so. The combined cohort study showed no effect of phenoxy herbicide exposure on oral cavity and pharyngeal cancers (26 cases coded ICD·9 140-149), RR = 1.1 (CI 0.7-1.6). None of the three deaths from cancer of the nose and nasal sinuses (ICD·9 160), RR = 1.6 (CI 0.3-4.7) were in the TCDD-exposed group.
Environmental Studies
In the earlier Seveso population study (Bertazzi et al., 1993) no cases of nasal or nasopharyngeal cancer were observed in zones A and B. Two cases were seen in zone R, with 2.8 expected. In the more recent follow-up of the Seveso population (Bertazzi et al., 1997) nasopharyngeal cancers are not specifically mentioned.
Vietnam Veteran Studies
The mortality experience of Australian Vietnam veterans was compared to that of the general public for 1964-1979 and 1980-1994 (Crane et al., 1997a). During 1964-1979, no cases of nasal or nasopharyngeal cancer deaths were found, whereas 0.8 was expected. For 1980-1994, there were two deaths due to nasal and two to nasopharyngeal cancers, whereas 1.7 and 3.9 were expected, respectively. A companion study comparing conscripted Australian veterans of Vietnam with military personnel who did not serve there reported an RR of 1.3 for nasopharyngeal cancer based on a single death in each of the populations between 1982 and 1994 (Crane et al., 1997b). There was one death due to nasal cancer in the comparison population and none in Vietnam veterans over the same period.
Synthesis
Nasal and nasopharyngeal cancers are relatively rare in many parts of the world and thus difficult to study epidemiologically. Scientific evidence on the association between herbicide exposure and nasopharyngeal cancer continues to be too sparse to make a definitive statement.
TABLE 7-6 Selected Epidemiologic Studies—Nasal/Nasopharyngeal Cancer
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
OCCUPATIONAL |
|||
New studies |
|||
Kogevinas et al., 1997 |
IARC cohort |
|
|
|
Oral cavity and pharynx cancer (ICD·9 140-9) |
26 |
1.1 (0.7-1.6) |
|
Nose and nasal sinuses cancer (ICD·9 160) |
3 |
1.6 (0.3-4.7) |
Studies reviewed in Update 1996 |
|||
Asp et al., 1994 |
Finnish herbicide applicators |
1 |
0.5 (0.01-2.9) |
Studies reviewed in VAO |
|||
Ronco et al., 1992 |
Danish and Italian farm workers |
|
0.6 (NS) |
Saracci et al., 1991 |
IARC cohort |
3 |
2.9 (0.6-8.5) |
Coggon et al., 1986 |
British MCPA production workers |
3 |
4.9 (1.0-14.4) |
Robinson et al., 1986 |
Paper and pulp workers |
0 |
— |
Wiklund, 1983 |
Swedish agricultural workers |
64 |
0.8 (0.6-1.2) |
Hardell et al., 1982 |
Residents of northern Sweden |
|
|
|
Phenoxy acid exposure |
8 |
2.1 (0.9-4.7) |
|
Chlorophenol exposure |
9 |
6.7 (2.8-16.2) |
ENVIRONMENTAL |
|||
Studies reviewed in VAO |
|||
Bertazzi et al., 1993 |
Residents in Seveso (zone R) |
2 |
2.6 (0.5-13.3) |
VIETNAM VETERANS |
|||
New studies. |
|||
Crane et al., 1997a |
Australian military veterans |
|
|
|
Nasal cancer |
2 |
1.2 (0.2-4.4) |
|
Nasopharyngeal cancer |
2 |
0.5 (0.1-1.9) |
Crane et al., 1997b |
Australian national service veterans |
|
|
|
Nasal cancer |
0 |
0 (0.0->10) |
|
Nasopharyngeal cancer |
1 |
1.3 (0.0->10) |
Studies reviewed in VAO |
|||
CDC, 1990 |
U.S. men born between 1921 and 1953 |
|
|
|
Vietnam veterans |
2 |
0.7 (0.1-3.0) |
NOTE: NS = not significant. a Given when available. |
Conclusions
Strength of Evidence in Epidemiologic Studies
There is inadequate or insufficient evidence to determine whether an association exists between exposure to the herbicides (2,4-D, 2,4,5-T and its contaminant TCDD, cacodylic acid, and picloram) and nasal or nasopharyngeal cancer. The evidence regarding association is drawn from occupational and other studies in which subjects were exposed to a variety of herbicides and herbicide components.
Biologic Plausibility
A thorough discussion of biologic plausibility with respect to exposure to TCDD or herbicides and nasal or nasopharyngeal cancer is contained in Chapter 3; a summary is presented in the conclusion to this chapter.
LARYNGEAL CANCER
Background
According to American Cancer Society estimates, 9,000 men and 2,100 women will be diagnosed with cancer of the larynx (ICD·9 161.0-161.9) in the United States in 1998, and 3,400 men and 900 women will die from the disease (ACS, 1998). These numbers represent approximately 1 percent of new cancer diagnoses and deaths. Cancer of the larynx is more common in men than women, with an overall ratio in the United States of about 5:1. Incidence also increases with age in the 45-59 age group.
Risk factors include tobacco and alcohol, which act individually and synergistically. Research suggests that gastroesophageal reflux, human papillomavirus, a weakened immune system, and occupational exposure to asbestos and certain chemicals and dusts may also increase incidence (ACS, 1998).
Average Annual Cancer Incidence (per 100,000 individuals) in the United Statesa Laryngeal Cancer |
||||||||||
|
45-49 years of age |
50-54 years of age |
55-59 years of age |
|
||||||
|
all races |
white |
black |
all races |
white |
black |
all races |
white |
black |
|
males |
6 |
6 |
16 |
13 |
13 |
25 |
22 |
20 |
43 |
|
females |
1 |
1 |
3 |
3 |
3 |
8 |
6 |
6 |
10 |
|
a SEER nine standard registries crude, age-specific rate, 1990-1994. |
Summary ofVAO andUpdate 1996
In nearly all studies analyzing respiratory cancers, the authors either group all of the different types of cancer in this broad group together (ICD·9 161-165, which include trachea, bronchus, lung, and larynx) or present data for the largest category within the group (ICD·9 162, which includes trachea, bronchus, and lung). However, in a few studies, data are separated to allow assessment of laryngeal cancer.
Of note are five studies of production workers in which data for laryngeal cancer (ICD·9 161) are presented separately (Coggon et al., 1986; Bond et al., 1988; Fingerhut et al., 1991; Manz et al., 1991; Saracci et al., 1991). Although the numbers are too small to draw strong conclusions, the consistency of a mild elevation in RR is suggestive of an association for laryngeal cancer. Pooling all but the Coggon data (Coggon et al., 1986, 1991) yields an OR of 1.8 (CI 1.0-3.2). As mentioned above, the potential confounders of an occupational risk for this cancer include tobacco and alcohol consumption. These studies did not directly control for smoking, although its magnitude in Manz et al. (1991) and Fingerhut et al. (1991) is not likely to be large. There is no information on alcohol consumption in any of the studies.
Other than these studies of production workers, only one study reported separate results for cancer of the larynx: a proportional cancer mortality ratio (PCMR) study was performed for farmers in 23 states, using occupational information from death certificates (Blair et al., 1993). Based on 162 deaths from laryngeal cancer in white male farmers, the PCMR was significantly decreased, at 0.7 (CI 0.6-0.8). This is consistent with a significant decrease in lung cancer in the same subgroup. The PCMR for laryngeal cancer in nonwhite male farmers was 1.1 (CI 0.8-1.5), based on 32 deaths. There were no deaths from this cancer in female farmers.
Update of the Scientific Literature
Occupational Studies
The IARC joint analysis (Kogevinas et al., 1997) of phenoxy herbicide production workers reported an SMR for laryngeal cancer of 1.6 (CI 1.0-2.5), based on 21 deaths among 21,863 workers in cohorts from 12 countries. Workers exposed to TCDD or higher chlorinated dioxins had an SMR of 1.7 (CI 1.0-2.8), based on 12 deaths among 13,831 workers.
Ramlow et al. (1996) examined the mortality of 770 workers potentially exposed to PCP between 1937 and 1980. Dioxin is an unintended by-product of PCP production. Exposure was estimated on the basis of job description information and industrial hygiene characterizations of the job environment. The SMR, calculated without any factoring for latency, was 2.9 (CI 0.3-10.3), based on two deaths.
Gambini et al. (1997) investigated cancer mortality in rice growers in northern Italy for 1957-1992. Employment on the farm was used as a surrogate for exposure to the range of phenoxy herbicides including 2,4-D and 2,4,5-T. An SMR of 0.9 (CI 0.4-1.9) was calculated, based on seven cases among 1,493 subjects.
Veterans Studies
Watanabe and Kang (1996) compare laryngeal cancer rates among Army and Marine Vietnam veterans with veterans who did not serve in Vietnam. Among Army veterans, the PMR for cancer of the larynx was 1.3 for Vietnam veterans and 0.9 for non-Vietnam veterans, based on 50 and 34 deaths, respectively. The corresponding numbers for Marines are 0.7 versus 1.4, based on four deaths in each group.
The Australian Vietnam veteran study (Crane et al., 1997a) found an SMR of 1.3 (0.7-2.3) based on 12 laryngeal cancer deaths. A second study examining the
TABLE 7-7 Selected Epidemiologic Studies—Laryngeal Cancer
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
OCCUPATIONAL |
|||
New Studies |
|||
Gambini et al., 1997 |
Italian rice growers |
7 |
0.9 (0.4-1.9) |
Kogevinas et al., 1997 |
IARC cohort |
21 |
1.6 (1.0-2.5) |
|
Workers exposed to TCDD (or higher chlorinated dioxins) |
12 |
1.7 (1.0-2.8) |
Ramlow et al., 1996 |
Pentachlorophenol production workers |
2 |
2.9 (0.3-10.3) |
Studies reviewed in Update 1996 |
|||
Blair et al., 1993 |
U.S. farmers in 23 states |
|
|
|
White males |
162 |
0.7 (0.6-0.8) |
|
Nonwhite males |
32 |
1.1 (0.8-1.5) |
Studies reviewed in VAO |
|||
Fingerhut et al., 1991 |
NIOSH cohort |
|
|
|
1 year exposure, 20 years latency |
3 |
2.7 (0.6-7.8) |
Manz et al., 1991 |
German production workers |
2 |
2.0 (0.2-7.1) |
Saracci et al., 1991 |
IARC cohort—exposed subcohort |
8 |
1.5 (0.6-2.9) |
Bond et al., 1988 |
Dow 2,4-D production workers |
1 |
3.0 (0.4-16.8) |
Coggon et al., 1986 |
British MCPA production workers |
4 |
2.3 (0.5-4.5) |
VIETNAM VETERANS |
|||
New Studies |
|||
Crane et al., 1997a |
Australian military veterans |
12 |
1.3 (0.7-2.3) |
Crane et al., 1997b |
Australian national service veterans |
0 |
0 (0->10) |
Watanabe and Kang, 1996 |
Army Vietnam veterans |
50 |
1.3 |
|
Marine Vietnam veterans |
4 |
0.7 |
mortality experience of conscripted Australian veterans relative to military personnel who did not serve in the conflict reported no laryngeal cancer deaths among Vietnam veterans and one in the comparison population between 1982 and 1994 (Crane et al., 1997b).
Synthesis
Studies published since Update 1996 continue to support the conclusion that there is limited/suggestive evidence of an association. The committee concluded that the evidence for this association was limited/suggestive rather than sufficient because of the inconsistent pattern of positive findings across populations with various degrees and types of exposure and because the most important risk factors for laryngeal cancers—cigarette smoking and alcohol consumption—were not fully controlled for or evaluated in the studies.
Conclusions
Strength of Evidence in Epidemiologic Studies
There is limited/suggestive evidence of an association between exposure to the herbicides (2,4-D, 2,4,5-T and its contaminant TCDD, cacodylic acid, and picloram) and laryngeal cancer. The evidence regarding association is drawn from occupational and other studies in which subjects were exposed to a variety of herbicides and herbicide components.
Biologic Plausibility
A thorough discussion of biologic plausibility with respect to exposure to TCDD or herbicides and laryngeal cancer is contained in Chapter 3; a summary is presented in the conclusion to this chapter.
LUNG CANCER
Background
Lung cancer (carcinomas of the lung and bronchus, ICD·9 162.2-162.9) is the leading cause of cancer death in the United States. According to American Cancer Society estimates, 91,400 men and 80,100 women will be diagnosed with this cancer in the United States in 1998, and approximately 93,100 men and 67,000 women will die from the disease (ACS, 1998). These numbers represent roughly 14 percent of new cancer diagnoses and 28 percent of cancer deaths in 1998. The principal types of lung neoplasms are identified collectively as bronchogenic carcinoma (''bronchus" is the term used to describe either of the two
main branches of the trachea) or carcinoma of the lung. The lung is also a common site for the development of metastatic cancer.
Lung cancer incidence can vary greatly in the age groups that describe most Vietnam veterans. For men and women, the incidence of lung cancer increases rapidly beginning about age 40. Incidence in 50-54 year olds is double that of 45-49 year olds; and it doubles again for 55-59 year olds. The rate for African-American males is consistently higher than for females or white males.
The American Cancer Society estimates that more than 90 percent of lung cancers in males are the result of tobacco smoking (ACS, 1998). Tobacco smoke may include both tumor initiators and promoters. Among the other risk factors are occupational exposure to asbestos, chromium, nickel, aromatic hydrocarbons, and radioactive ores.
Average Annual Cancer Incidence (per 100,000 individuals) in the United Statesa Lung and Bronchus Cancer |
||||||||||
|
45-49 years of age |
50-54 years of age |
55-59 years of age |
|
||||||
|
all races |
white |
black |
all races |
white |
black |
all races |
white |
black |
|
males |
40 |
35 |
86 |
90 |
82 |
196 |
171 |
161 |
320 |
|
females |
30 |
29 |
46 |
65 |
66 |
86 |
113 |
116 |
133 |
|
a SEER nine standard registries crude, age-specific rate, 1990-1994. |
Summary of VAO and Update 1996
Numerous studies were considered in the earlier reports that specifically evaluated lung cancer and its relationship to herbicides. These included both cohort and case-control studies, which were primarily occupational. The strongest data in support of a lung cancer association come from the phenoxy herbicide formulators (Fingerhut et al., 1991; Manz et al., 1991). In both studies, attempts were made to compare the exposed workers to comparable worker groups with assumed similar smoking patterns. The large studies of farmers that did not show an association were not particularly informative due to lower smoking rates among farmers compared to the general population. The environmentally exposed Seveso population had relatively few lung cancers but had too short a follow-up period to conclude anything about lung cancer and TCDD.
Most studies either considered respiratory cancers as a group (ICD·9 161-165) or specifically addressed cancer of the lung and bronchus.
Update of the Scientific Literature
Occupational Studies
IARC (Kogevinas et al., 1997) has brought together almost all of the phenoxy herbicide production workers in 36 cohorts in 12 countries (Fingerhut et al., 1991; Manz et al., 1991; Sarraci et al., 1991; Flesch-Janys et al., 1995; Becher et al., 1996) for a joint analysis. This cohort contains 26,976 workers and was divided into those
who were exposed to TCDD or higher chlorinated dioxins and those who were not exposed. There were 225 lung cancer deaths in the exposed group, and 148 lung cancer deaths among unexposed workers. Lung cancer deaths were slightly elevated for the entire worker cohort, with an SMR of 1.1 (CI 1.0-1.2). No relationship was observed between incidence and years since first exposure or duration of exposure. For the TCDD-exposed group, the SMR was 1.1 (CI 1.0-1.3), whereas the non-TCDD exposed group had an SMR of 1.0 (CI 0.9-1.2). Becher et al. (1996) separately report 47 deaths from lung cancer in workers from four phenoxy herbicide plants examined as part of the IARC cohort (SMR = 1.4, CI 1.1-1.9). No association was observed between mortality and time since first exposure.
Ott and Zober (1996) updated their research on 243 male workers exposed to TCDD during the cleanup of a TCP reactor that exploded in 1953 at a BASF plant in Ludwigshafen, Germany. Dose estimates for these individuals were developed from work activity information, blood TCDD determinations on a subset of the population, and estimates of TCDD elimination rates: 11 deaths from respiratory cancers were identified during 1953-1992, 7 of them among the 69 workers with the highest estimated TCDD dose (SMR = 2.4, CI 1.0-5.0). There was also a trend of increased incidence with increased estimated dose for 8 observed cases of lung or bronchus cancer in the highest-dose group (out of eleven total) between 1960 and 1993, yielding an SIR of 2.2 (CI 1.0-4.3).
Ramlow et al. (1996) examined the mortality of 770 workers potentially exposed to PCP between 1937 and 1980. Dioxin is an unintended by-product of PCP production. Exposure was estimated on the basis of job description information and industrial hygiene characterizations of the job environment. Sixteen lung cancers were observed in the cohort, twelve of them in the higher estimated exposure group (RR = 1.1, CI 0.7-2.0).
The Svensson et al. (1995) study of Swedish fishermen reported a nonsignificant decrease in lung and larynx cancer mortality relative to the general population (SMR = 0.8, CI 0.5-1.3, based on 16 cases) in a group hypothesized to have greater exposure to PCB, PCDD, and PCDF through consumption of contaminated fish. This same group had a nonsignificantly elevated incidence of lung cancer (SIR = 1.2, CI 0.8-1.8, based on 24 cases).
Gambini et al. (1997) investigated cancer mortality among a cohort of 1,493 rice growers in northern Italy during 1957-1992. Employment on the farm was used as a surrogate for exposure to the range of phenoxy herbicides including 2,4-D and 2,4,5-T. Researchers reported an SMR for lung cancer of 0.8 (CI 0.6-1.1) based on 46 cases. Lower lung cancer mortality has also been observed in studies of farmers in the United States (e.g., Blair et al., 1993).
Environmental Studies
Studies of the population exposed to TCDD in a 1976 industrial accident in Seveso, Italy, included estimates of lung cancer risk (Bertazzi et al., 1989b;
The Seveso studies followed the population for 15 years after the accident. If the release of TCDD did increase the risk for lung cancers, one might not expect to see a full impact on tumor incidence for some years to come. At least another five years is needed before the impact of the accident on cancer incidence can be meaningfully assessed.
Vietnam Veteran Studies
Watanabe has updated his work (Watanabe and Kang, 1995, 1996) comparing lung cancer rates among Army and Marine Vietnam veterans with non-Vietnam veterans. Among Army veterans, the PMR for lung cancer was 1.1 for Vietnam veterans and 1.1 for non-Vietnam veterans based on 1,139 and 1,141 lung cancer deaths, respectively. The corresponding numbers for Marines are 1.2 and 0.9 based on 215 and 77 deaths, respectively. Twenty-two U.S. Army Chemical Corps units assigned to Vietnam between 1966 and 1971 have been followed for vital status. In a review of Chemical Corps units by Dalager and Kang (1997), 11 lung cancer deaths among Vietnam units were found, compared to 3 among units not in Vietnam, for an RR of 1.4 (CI 0.4-5.4). A case-control study of lung cancer mortality in Vietnam veterans (Mahan et al., 1997) found an increased OR of 1.4 (CI 1.0-1.9) based on 111 cases.
In a study (Crane et al., 1997a) of the mortality of all Australian Vietnam service personnel compared with the general population, 212 lung cancer deaths were found with an adjusted 164 expected for an SMR of 1.3 (CI 1.1-1.5). A second study (Crane et al., 1997b) compared cases among Vietnam veterans with those who did not serve in Vietnam. Lung cancer mortality showed an RR of 2.2 (CI 1.1-4.3) based on 27 deaths in the Vietnam group.
Synthesis
Since Update 1996, there have been several studies of respiratory cancer among occupationally exposed groups and Vietnam veterans. Newly published
studies of phenoxy herbicide production workers (Kogevinas et al., 1997) and workers exposed as a result of an industrial accident (Ott and Zober, 1996) show small but statistically significant excesses of lung cancer mortality. Results in both studies indicate higher estimated risk for individuals with higher estimated exposure. A third study addressing workers potentially exposed to dioxins including TCDD and other congeners (Ramlow et al., 1996) reported a RR indistinguishable from 1. A study of rice farmers in Italy found lower lung cancer incidence than observed in the general population, which is similar to results found in studies of U.S. farmers and may reflect lower smoking incidence in this occupational group.
Newer data from Seveso do not indicate any increase in lung cancer mortality in this environmentally exposed group, but an insufficient number of years have passed since exposure to draw conclusions about any effect the accidental exposure may have had.
Although increases in respiratory cancers were seen in new studies of Vietnam veterans, there is evidence that cigarette smoking was more prevalent among Vietnam veterans than among non-Vietnam veterans (McKinney et al., 1997). Also, veterans smoked at a higher rate than the general public both in the United States and Australia (O'Toole et al., 1996; McKinney et al., 1997). Current studies of lung cancer risk in veterans are of limited usefulness in evaluating the effect of herbicide exposure because it is not possible to identify the soldiers likely to have been exposed to herbicides and smoking information is not available. The studies do, however, show a consistent pattern of increased lung cancer mortality among those who served in Vietnam.
In summary, the most recently published studies continue to support the placement of lung and trachea cancer in the category "limited/suggestive evidence of an association." Several studies suggest a higher rate of these cancers in individuals with known exposure to phenoxy herbicides or dioxin, and there is some evidence of a dose-response relationship. Whereas smoking undoubtedly plays a role in these cancers, the consistency of the finding across several studies argues against the notion that it is the sole explanatory factor.
Conclusions
Strength of Evidence in Epidemiologic Studies
There is limited/suggestive evidence of an association between exposure to herbicides and/or dioxin and cancers of the lung/bronchus and trachea. The primary evidence comes from studies of individuals occupationally exposed to phenoxy herbicides. Although studies show an increase in lung cancer among Vietnam veterans in the United States and Australia, the lack of information regarding herbicide exposure and cigarette smoking limits their usefulness.
TABLE 7-8 Selected Epidemiologic Studies—Lung/Bronchus Cancer
Reference |
Study Population |
Exposed Cases |
Estimated Risk (95% CI) |
OCCUPATIONAL |
|||
New Studies |
|||
Gambini et al., 1997 |
Italian rice growers |
45 |
0.8 (0.6-1.1) |
Kogevinas et al., 1997 |
Phenoxy herbicides: 36 cohorts |
|
|
|
Exposed to TCDD or higher PCDD |
225 |
1.1 (1.0-1.3) |
|
Exposed to no or lower PCDD |
148 |
1.0 (0.9-1.2) |
Becher et al., 1996 |
German chemical production workers |
47 |
1.4 (1.1-1.9) |
Ott and Zober, 1996 |
BASF cleanup workers |
6 |
3.1 (1.1-6.7) |
Ramlow et al., 1996 |
Pentachlorophenol production workers |
18 |
1.0 (0.6-1.5) |
Studies reviewed in Update 1996 |
|||
Asp et al., 1994 |
Finnish herbicide applicators |
37 |
1.0 (0.7-1.4) |
Blair et al., 1993 |
U.S. farmers in 23 states |
|
|
|
White males |
6,473 |
0.9 (0.9-0.9) |
|
Nonwhite males |
664 |
1.0 (0.9-1.1) |
Bloemen et al., 1993 |
Dow 2,4-D production workers |
9 |
0.8 (0.4-1.5) |
Kogevinas et al., 1993 |
Female herbicide spraying and production workers |
2 |
1.4 (0.2-4.9) |
Lynge, 1993 |
Danish male production workers |
13 |
1.6 (0.9-2.8) |
Studies reviewed in VAO |
|||
Bueno de Mesquita et al., 1993 |
Phenoxy herbicide workers |
9 |
1.7 (0.5-6.3) |
Swaen et al., 1992 |
Herbicide applicators |
12 |
1.1 (0.6-1.9) |
Coggon et al., 1991 |
Phenoxy herbicide production workers |
19 |
1.3 (0.8-2.1) |
|
|
14 |
1.2 (0.7-2.1) |
Fingerhut et al., 1991 |
TCDD-exposed workers |
89 |
1.1 (0.9-1.4) |
|
>1 year exposure; >20 years latency |
40 |
1.4 (1.0-1.9) |
Green, 1991 |
Herbicide sprayers in Ontario |
5 |
1.1 (0.4-2.5) |
Manz et al., 1991 |
Phenoxy herbicide production workers |
26 |
1.7 (1.1-2.4) |
Saracci et al., 1991 |
Herbicide spraying and production workers |
173 |
1.0 (0.9-1.2) |
|
Probably exposed subgroup |
11 |
2.2 (1.1-4.0) |
McDuffie et al., 1990 |
Saskatchewan farmers applying herbicides |
103 |
0.6 |
Zober et al., 1990 |
TCDD reactor accident workers |
6 |
1.6 |
|
High exposure |
4 |
2.0 (0.6-5.2) |
|
Chloracne |
6 |
1.8 (0.7-4.0) |
Wiklund et al., 1989 |
Pesticide applicators in Sweden |
38 |
0.5 (0.4-0.7) |
Bond et al., 1988 |
Dow 2,4-D production workers (15 years latency) |
9 |
1.2 (0.6-2.3) |
|
Low cumulative exposure |
1 |
0.7 |
|
Medium cumulative exposure |
2 |
1.0 |
|
High cumulative exposure |
5 |
1.7 |
Coggon et al., 1986 |
MCPA production workers |
101 |
1.2 (1.0-1.4) |
|
Background exposure |
39 |
1.0 (0.7-1.4) |
|
Low-grade exposure |
35 |
1.1 (0.8-1.6) |
|
High-grade exposure |
43 |
1.3 (1.0-1.8) |
Reference |
Study Population |
Exposed Cases |
Estimated Risk (95% CI) |
Lynge, 1985 |
Danish production workers |
|
|
|
Males |
38 |
1.2 |
|
Females |
6 |
2.2 |
|
Manufacture and packing only—males |
11 |
2.1 (1.0-3.7) |
Blair et al., 1983 |
Licensed pesticide applicators in Florida, lawn and ornamental herbicides only |
7 |
0.9 (0.4-1.9) |
Axelson et al., 1980 |
Herbicide sprayers in Sweden |
3 |
1.4 (0.3-4.0) |
Bender et al., 1989 |
Herbicide sprayers in Minnesota |
54 |
0.7 (0.5-0.9) |
ENVIRONMENTAL |
|||
New Studies |
|||
Bertazzi et al., 1997 |
Seveso residents |
|
|
|
zone A—males |
4 |
0.8 (0.2-2.1) |
|
zone A—females |
0 |
0.0 (0.0-5.8) |
|
zone B—males |
40 |
1.2 (0.9-1.7) |
|
zone B—females |
2 |
0.5 (0.1-1.8) |
|
zone R—males |
208 |
0.9 (0.8-1.1) |
|
zone R—females |
35 |
1.1 (0.8-1.5) |
Svensson et al., 1995 |
Swedish fishermen mortality |
|
|
|
East coast |
16 |
0.8 (0.5-1.3) |
|
West coast |
77 |
0.9 (0.7-1.1) |
Studies reviewed in VAO |
|||
Bertazzi, 1993 |
Seveso residents first ten years after accident |
|
|
|
zone A—males |
2 |
0.8 (0.2-3.4) |
|
zone A—females |
0 |
— |
|
zone B—males |
18 |
1.1 (0.7-1.8) |
|
zone B—females |
0 |
— |
|
zone R—males |
96 |
0.8 (0.7-1.0) |
|
zone R—females |
16 |
1.5 (0.8-2.5) |
VIETNAM VETERANS |
|||
New Studies |
|||
Crane et al., 1997a |
Australian military veterans |
212 |
1.3 (1.1-1.5) |
Crane et al., 1997b |
Australian national service veterans |
27 |
2.2 (1.1-4.3) |
Dalager and Kang, 1997 |
Army Chemical Corps Veterans |
11 |
1.4 (0.4-5.4) |
Mahan et al., 1997 |
Case-control |
111 |
1.4 (1.0-1.9) |
Watanabe and Kang, 1996 |
Vietnam service Army |
1,139 |
1.1 |
|
Non-Vietnam |
1,141 |
1.1 |
|
Vietnam service Marines |
215 |
1.2 |
|
Non-Vietnam |
77 |
0.9 |
Watanabe and Kang, 1995 |
Vietnam service Marines v. non-Vietnam |
42 |
1.3 (0.8-2.1) |
Biologic Plausibility
A thorough discussion of biologic plausibility with respect to exposure to TCDD or herbicides and cancers of the lung/bronchus and trachea is contained in Chapter 3; a summary is presented in the conclusion to this chapter.
BONE CANCER
Background
According to the American Cancer Society, approximately 1,300 men and 1,100 women will be diagnosed with bone or joint cancer (ICD·9 170.0-170.9) in the United States in 1998, and 800 men and 600 women will die as a result of this cancer (ACS, 1998). Primary bone cancers are among the least common malignancies. The bones are, however, frequent sites for secondary tumors of other cancers that have metastasized (i.e., have spread from another site). Only the primary cancers are considered here.
Bone cancer is more common in teenagers than adults. The incidence among individuals in the age groups that characterize most Vietnam veterans is quite low, and care should be exercised when interpreting the numbers presented below.
Among the risk factors for adults contracting bone and joint cancer are exposure to ionizing radiation from treatment for other cancers and a history of certain noncancerous bone diseases.
Average Annual Cancer Incidence (per 100,000 individuals) in the United Statesa Bone and Joint Cancer |
|||||||||
|
45-49 years of age |
50-54 years of age |
55-59 years of age |
||||||
|
all races |
white |
black |
all races |
white |
black |
all races |
white |
black |
males |
1.0 |
1.1 |
0.3 |
0.8 |
0.8 |
0.4 |
1.2 |
1.2 |
1.1 |
females |
0.5 |
0.5 |
0.5 |
0.6 |
0.6 |
0.4 |
0.7 |
0.6 |
0.4 |
a SEER nine standard registries crude, age-specific rate, 1990-1994. |
Summary of VAO and Update 1996
Studies of bone cancer and herbicide exposure have included chemical production workers (Coggon et al., 1986; Bond et al., 1988; Zober et al., 1990; Fingerhut et al., 1991; Collins et al., 1993); agricultural workers (Burmeister, 1981; Wiklund, 1983; Ronco et al., 1992; Blair et al., 1993); and Vietnam veterans (Lawrence et al., 1985; Anderson et al., 1986a,b; Breslin et al., 1988). Very few of these studies reported more than a handful of cases, typically resulting in very wide confidence intervals, and there was little evidence of elevated risk from any single study. Point estimates of risk are relatively evenly distributed on both sides of the "no-effect" estimated risk of 1.0.
Update of the Scientific Literature
Occupational Studies
The most important new study is of the IARC combined occupational cohorts (Kogevinas et al., 1997). This study found five cases of bone cancer and an SMR of 1.2 (CI 0.4-2.8) for all workers, with the SMR lower (1.1) in those exposed to TCDD than in those not exposed (1.4). Ramlow and colleague's (1996) study of 770 pentachlorophenol workers reported no deaths from bone cancer. A study of rice growers in northern Italy (Gambini et al., 1997) identified one death. Hertzman's study of 26,000 Canadian sawmill workers presumptively exposed to dioxin-contaminated chlorophenate (Hertzman et al., 1997) reported five bone cancer deaths, with an SMR of 1.3 (CI 0.5-2.7). The SIR, based on four cases, was 1.1 (CI 0.4-2.4). These ratios were based on outcomes determined by record linkage to national data bases.
Environmental Studies
In Bertazzi and colleague's continued follow-up of individuals exposed to TCDD as the result of a 1976 industrial accident near Seveso, Italy (Bertazzi et al., 1997), two deaths occurred in men from the lowest exposure zone, zone R (SMR = 0.5). Among women, one death occurred in contaminated zone B and seven in zone R (SMR = 2.4, CI 1.0-4.9).
Vietnam Veteran Studies
No bone cancer results are reported in any of the DVA Vietnam veteran studies (Dalager et al., 1995a; Watanabe and Kang, 1995, 1996). The most recent study of Ranch Hands (Air Force Health Study, 1996) reports no bone cancer deaths. The Australian veteran studies (Crane et al., 1997a,b) do not specifically report bone cancer outcomes. Only Clapp's update of his study of Massachusetts veterans reports bone cancer outcomes: four cases observed, with an OR of 0.9 (CI 0.1-11.3), comparing Vietnam veterans to Vietnam era veterans who did not serve in Vietnam. No bone cancer results are reported in any of the other Vietnam veteran studies.
Synthesis
There is minimal new information on this very rare disease, for which few data existed before. The new studies do not change the conclusions of VAO and Update 1996 concerning bone cancer.
Conclusions
Strength of Evidence in Epidemiologic Studies
There is inadequate or insufficient evidence to determine whether an association exists between exposure to the herbicides (2,4-D, 2,4,5-T and its contaminant TCDD, cacodylic acid, and picloram) and bone cancer. The evidence regarding association is drawn from occupational and other studies in which subjects were exposed to a variety of herbicides and herbicide components.
Biologic Plausibility
A thorough discussion of biologic plausibility with respect to exposure to TCDD or herbicides and bone cancer is contained in Chapter 3; a summary is presented in the conclusion to this chapter.
Chondrosarcomas of the Skull
The DVA asked the committee responsible for this report to give special attention to the issue of classification of chondrosarcomas of the skull. "Chondrosarcoma" is the general name for a class of neoplasms derived from cartilage cells or their precursors. Most chondrosarcomas develop in the interior of the bone or on the bone surface. Certain rare chondrosarcomas involve soft tissue.
The studies used by the committee in its consideration of the effects of herbicide or dioxin exposure relied on ICD·9 codes to classify health outcomes. ICD·9 code 170 is the general classification for "malignant neoplasms of bone and articular cartilage." More specifically, chondrosarcomas arising in the skull, mandible, or vertebral column are classified under ICD·9 codes 170.0-170.2, the same codes used for bone cancers at these sites. Soft-tissue sarcomas are classified under ICD·9 codes 171.0-171.9 and 164.1 (heart neoplasms). The definition of code 171, "malignant neoplasm of connective and other soft tissue," explicitly excludes tumors involving articular cartilage but includes cartilage in areas such as the eye and ear.
The committee did not identify any studies that specifically examined herbicide or dioxin exposure and chondrosarcoma incidence. The classifications discussed above mean that any association between herbicide or dioxin exposure and skull chondrosarcomas will be subsumed under the discussion of bone cancer. However, this epidemiologic classification is not intended to be a substitute for the expert judgment of pathologists in individual cases.
SOFT-TISSUE SARCOMAS
Background
The American Cancer Society estimates that 3,700 men and 3,300 women will be diagnosed with soft-tissue sarcoma (STS) (ICD·9 171.0-171.9, 164.1) in
TABLE 7-9 Selected Epidemiologic Studies—Bone Cancer
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
OCCUPATIONAL |
|||
New Studies |
|||
Gambini et al., 1997 |
Italian rice growers |
1 |
|
Hertzman et al., 1997 |
British Columbia sawmill workers |
|
|
|
Mortality |
5 |
1.3 (0.5-2.7) |
|
Incidence |
4 |
1.1 (0.4-2.4) |
Kogevinas et al., 1997 |
IARC cohort |
5 |
1.2 (0.4-2.8) |
|
Workers exposed to TCDD (or higher chlorinated dioxins) |
|
1.1 |
|
Workers not exposed to TCDD (or higher chlorinated dioxins) |
|
1.4 |
Ramlow et al., 1996 |
Pentachlorophenol production workers |
0 |
|
Studies reviewed in Update 1996 |
|||
Blair et al., 1993 |
U.S. farmers in 23 states |
49 |
1.3 (1.0-1.8) |
Collins et al., 1993 |
Monsanto 2,4-D production workers |
2 |
5.0 (0.6-18.1) |
Studies reviewed in VAO |
|||
Ronco et al., 1992 |
Danish male self-employed farm workers |
9 |
0.9 |
Fingerhut et al., 1991 |
NIOSH cohort |
2 |
2.3 (0.3-8.2) |
Zober et al., 1990 |
BASF production workers |
0 |
— (0.0-70.0) |
Bond et al., 1988 |
Dow 2,4-D production workers |
0 |
— (0.0-31.1) |
Coggon et al., 1986 |
British MCPA production workers |
1 |
0.9 (0.0-5.0) |
Wiklund, 1983 |
Swedish agricultural workers |
44 |
1.0 (0.6-1.4)b |
Burmeister, 1981 |
Farmers in Iowa |
56 |
1.1 (NS) |
ENVIRONMENTAL |
|||
New Studies |
|||
Bertazzi et al., 1997 |
Seveso |
|
|
|
Men in zone R |
2 |
0.5 |
|
Women in zone B |
1 |
— (0.0-14.4) |
|
Women in zone R |
7 |
2.4 (1.0-4.9) |
VIETNAM VETERANS |
|||
New Studies |
|||
Clapp, 1997 |
Massachusetts Vietnam veterans |
4 |
0.9 (0.1-11.3) |
AFHS, 1996 |
Ranch Hands |
0 |
|
Studies reviewed in VAO |
|||
Breslin et al., 1988 |
Army Vietnam veterans |
27 |
0.8 (0.4-1.7) |
|
Marine Vietnam veterans |
11 |
1.4 (0.1-21.5) |
Anderson et al., 1986a |
Wisconsin Vietnam veterans |
1 |
— |
Anderson et al., 1986b |
Wisconsin Vietnam veterans |
1 |
— |
Lawrence et al., 1985 |
New York Vietnam veterans |
8 |
1.0 (0.3-3.0) |
NOTE: NS = not significant. a Given when available. b 99% CI. |
the United States in 1998 and that 2,000 men and 2,300 women will die from these cancers (ACS, 1998). STSs arise in the soft somatic tissues that occur within and between organs. Three of the most common types of STS—liposarcoma, fibrosarcoma, and rhabdomyosarcoma—occur in similar numbers in men and women. Because of the diverse characteristics of STS, accurate diagnosis and classification can be difficult.
There is no consistent pattern to the incidence of STSs over the age groups that describe most Vietnam veterans.
Among the risk factors for these cancers are exposure to ionizing radiation from treatment for other cancers and certain inherited conditions including Gardner's syndrome, Li-Fraumeni syndrome, and neurofibromatosis. Several chemical exposures have also been identified as possible risk factors (Zahm and Fraumeni, 1997).
Average Annual Cancer Incidence (per 100,000 individuals) in the United Statesa Soft-Tissue Sarcomas (including malignant neoplasms of the heart) |
|||||||||
|
45-49 years of age |
50-54 years of age |
55-59 years of age |
||||||
|
all races |
white |
black |
all races |
white |
black |
all races |
white |
black |
males |
3.0 |
2.9 |
3.1 |
3.5 |
3.8 |
2.6 |
4.7 |
4.0 |
7.5 |
females |
2.1 |
1.7 |
4.4 |
2.4 |
2.3 |
3.6 |
2.9 |
2.8 |
3.0 |
a SEER nine standard registries crude, age-specific rate, 1990-1994. |
Summary of VAO and Update 1996
The strongest evidence for an association between STS and exposure to phenoxy herbicides comes from a series of case-control studies conducted in Sweden (Hardell and Sandstrom, 1979; Eriksson et al., 1981, 1990; Hardell and Eriksson, 1988). The studies, involving a total of 506 cases, show an association between STS and exposure to phenoxy herbicides, chlorophenols, or both. The committee concluded that although these studies have been criticized, there is insufficient justification to discount the consistent pattern of elevated risks and the clearly described and sound methods employed. The methodology and findings from these studies are reviewed in detail in VAO. These findings are supported by a significantly increased risk in a National Institute for Occupational Safety and Health (NIOSH) study (SMR = 9.2, CI 1.9-27.0) for production workers most highly exposed to TCDD (Fingerhut et al., 1991) and a similar increased risk in the IARC cohort (SMR = 6.1, CI 1.7-15.5) for deaths that occurred between 10 and 19 years after first exposure (Saracci et al., 1991; Kogevinas et al., 1992), using a fairly crude exposure classification. These are the two largest, as well as the most highly exposed, occupational cohorts. Some studies in other occupational, environmental, and veteran groups showed an increased risk for STS, but the results were commonly nonsignificant, possibly because of small sample sizes (related to the relative rarity of STS in the population).
A nested case-control study of STS within the IARC multicountry worker cohort (Kogevinas et al., 1995) was done using eleven cases of STS and five controls per case. Detailed exposure reconstruction was performed for all cases of STS and a set of controls by a team of industrial hygienists who did not know case or control status (Kauppinen et al., 1994). The team estimated cumulative exposures to TCDD and numerous phenoxy herbicides and related chemicals. There were associations between STS risk and exposure to ''any phenoxy herbicide," "any dioxin," and several other definitions of exposure. The authors noted that because many of the workers had multiple exposures and few had single exposures, it is difficult to conclude with confidence that the risk is more strongly associated with any specific exposure to the broad class of phenoxy acids and related compounds. There was evidence of increasing risk with increasing cumulative exposure to several agents, including TCDD and 2,4-D, a herbicide that does not contain TCDD.
Collins et al. (1993) point out that "all but one of the confirmed STS cases among more than 5000 workers in 12 plants mentioned in the Fingerhut et al. [(1991)] study occurred among the 754 persons in the [Monsanto] study" and, based on a detailed analysis of the exposure histories of STS cases, argue that TCDD is unlikely to be responsible, but that 4-aminobiphenyl may be.
Several authors reported additional years of follow-up in occupational cohort studies (Bloemen et al., 1993; Lynge, 1993; Asp et al., 1994). Lynge (1993) found the risk of STS similar to that reported in the earlier study of this cohort of Danish herbicide manufacturers (Lynge, 1985). There were 5 cases of STS observed, versus 2.5 expected (RR = 2.0, CI 0.7-4.8). When the definition of exposure was restricted to those with at least one year of work in exposed areas and a ten-year interval was applied between the start of exposure and the time during follow-up when subjects began to be at risk, there were 3 observed cases, compared to 0.5 cases expected (RR = 6.4, CI 1.3-18.7). These workers were engaged in the manufacture of 2,4-D and a related herbicide, 4-chloro-2-methyl-phenoxyacetic acid (MCPA), but not 2,4,5-T.
In the United States, a PCMR study was performed of farmers in 23 states, using occupational information from death certificates (Blair et al., 1993). Based on 98 deaths from STS in white male farmers, the PCMR was 0.9 (CI 0.8-1.1). The numbers of deaths due to STS were small and nonsignificant in the other racial and gender groups: nonwhite males and white and nonwhite females.
The Bertazzi et al. (1993) study of cancer incidence in Seveso, Italy, yielded results similar to those reported in earlier publications of this group (Bertazzi et al., 1989a,b; Pesatori et al., 1992). In the small, most heavily exposed group (zone A), no cases of STS were observed when the class is defined as tumors in ICD 171, "malignant neoplasms of connective and other soft tissues." There were two cases of "soft tissue sarcomas of parenchymal origin," which are not included in ICD 171 but, according to some classifications, belong in the group of tumors under consideration (Mack, 1995). It is difficult to evaluate this finding,
because of the problem of estimating a comparable expected incidence for the same tumors, but the authors noted that 1.4 cases would be expected in this cohort when cancers including ICD 171 and cancers of parenchymal origin are combined. In the larger but less exposed group from zone B, no cases of ICD 171 cancers were observed, whereas about 0.5 was expected. Zone R is the largest group, with considerably lower exposures to TCDD on average. Two cases of STS (ICD 171) were observed in females (RR = 1.6, 95% CI 0.3-7.4). In males, six cases were observed, yielding an RR of 2.8 (CI 1.0-7.3). There appeared to be a trend in increasing risk with increasing duration of residence in zone R.
A PMR study that examined the causes of death among veterans on the state of Michigan's Vietnam era bonus list was recently reported (Visintainer et al., 1995). The mortality rates of 3,364 Vietnam veterans were compared to the mortality rates of 5,229 veterans who served elsewhere. Based on eight deaths from STS, the PMR was 1.1 (CI 0.5-2.2). No data were available to identify whether individual Vietnam veterans had been exposed to herbicides.
Update of the Scientific Literature
There have been no studies since Update 1996 that focused exclusively on the STS-TCDD relationship, although it often receives special attention in studies that look at multiple cancer types. Most of the literature since Update 1996 consists of longer-term follow-up, or expansion, of previously described cohorts. Because STS is such a rare cancer, continued follow-up typically adds few cases to the total, so most of the previously noted suggestive trends persist in the updated reports. No further studies have been done of the original Swedish cohorts that still form the strongest evidence for this connection.
Occupational Studies
As with the other cancer types, the largest occupational cohort to be studied is that in the IARC international study of 21,863 workers in Europe and the United States (Kogevinas et al., 1997). This cohort includes subjects in the 10-country cohort originally described in 1992 (Kogevinas et al., 1992), as well as several German cohorts that had been reported on separately since Update 1996 (Becher et al., 1996; Flesch-Janys et al., 1995). The SMR for STS in all workers exposed to any phenoxy herbicide or chlorphenol was 2.0 (CI 0.9-3.8), based on nine deaths. This corresponds to an increased lifetime risk of roughly 2 per 10,000. The SMR was essentially unchanged when the cohort was restricted to workers exposed to TCDD or higher chlorinated dioxins, although this was based on six deaths (CI 0.8-4.4). Workers not exposed to TCDD had an SMR of 1.35, based on two deaths. When the six deaths were analyzed by duration of exposure, there was an apparent trend, with SMRs of 1.2, 4.8, and 6.5 being reported for the durations 1-4 years, 5-9 years, and 10-19 years. However, the authors of this
study also noted that a diagnosis of STS based on death certificates is known to be inaccurate and that two of the deaths reported were not STS when the slides were reviewed. However, they noted that in several cohorts, a few new cases of STS were also discovered when death records were examined intensively. The authors decided to rely on the death certificate-based numbers but warned that their results for STS should be "interpreted with caution."
In a study of 770 pentachlorophenol workers (Ramlow et al., 1996), there were no deaths from STS (0.2 expected). In Ott and Zober's continued follow-up of workers exposed to TCDD in a 1953 industrial accident (Ott and Zober, 1996), no cases of STS have been observed (0.23 expected). In a study of Canadian sawmill workers (Hertzman et al., 1997), 11 deaths from STS were observed, with an SMR of 1.0 (CI 0.6-1.7). Based on the same number of cases, the SIR was 1.0 (CI 0.6-1.7). No dose-response trend was seen between increasing hours of exposure and either mortality or incidence.
Environmental Studies
All other studies of possible herbicide exposure have involved environmental exposures due either to accidents or occupation (e.g., farmers). In Gambini's study of owners of 1,493 rice-growing farms (Gambini et al., 1997), there was one STS death (0.25 expected). In Bertazzi's continued follow-up of the Seveso accident (Bertazzi et al., 1997), four STS deaths were reported (2.1 expected), all among males in the lowest-exposure zone. In this subgroup, the SMR was 2.1 (CI 0.6-5.4). Svensson's study of cancer incidence and mortality among approximately 11,400 Swedish fishermen (Svensson et al., 1995), with presumably a high intake of organochlorine chemicals in the fatty fish they consume found three STS cases (7.5 expected).
Veterans Studies
Several studies of Vietnam veterans have appeared since Update 1996 . In a study of approximately 2,800 Army Chemical Corps soldiers who went to Vietnam (Dalager and Kang, 1997), STS is not included in the seven types of cancer outcomes reported. With 36 observed cancer deaths, it is highly likely that no STS cases occurred. Watanabe and Kang (1996) did a PMR analysis of more than 60,000 Vietnam era veterans. When Vietnam veterans were compared to non-Vietnam veterans, PMRs near or below 1.0 were observed (based on 59 total deaths). Compared to the U.S. population, a nonstatistically significant PMR of about 1.2 was observed among Vietnam veterans. However, it must be emphasized that this analysis was not of risk of death due to STS, but rather of risk of STS given that death had occurred. A study of 10,716 U.S. Marines who served in Vietnam did not report any STS figures (Watanabe and Kang, 1995). The most recent update of the Ranch Hand study (AFHS, 1996) reported one death from
STS (0.3 expected). Clapp's update of Massachusetts veterans (Clapp, 1997), which previously showed a significant threefold STS excess (Clapp et al., 1991), now reports a nonsignificant OR of 1.6 (CI 0.5-5.4), based on 18 cases.
An Australian study of Vietnam veterans showed no excess in the number of deaths from STS in any service during 1980-1994, with the number of observed deaths ranging from zero to nine in various services and most SMRs <1 (Crane et al., 1997a). A companion study comparing conscripted Australian veterans to military personnel who did not serve in the conflict reported a nonstatistically significant RR of 0.7, based on four deaths among Vietnam veterans and two in the comparison population between 1982 and 1994 (Crane et al., 1997b). Neither of these studies characterized veterans' herbicide exposure.
Synthesis
The bulk of the evidence supporting the STS-dioxin connection still derives from the early Swedish studies. These are somewhat supported by subsequent studies of NIOSH and IARC occupational cohorts, although weaknesses in these studies limit the confidence with which conclusions can be drawn from them. Because soft-tissue sarcomas are rare, it is difficult to discern whether the small increases in the number of cases observed in these studies are due to dioxin
TABLE 7-10 Selected Epidemiologic Studies—All (or unspecified) Skin Cancer Mortality
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
OCCUPATIONAL |
|||
Studies reviewed in VAO |
|||
Fingerhut et al., 1991 |
NIOSH cohort |
4 |
0.8 (0.2-2.1) |
Saracci et al., 1991 |
IARC cohort |
3 |
0.3 (0.1-0.9) |
Alavanja et al., 1988 |
USDA agricultural extension agents |
5 |
1.1 (0.5-2.6) |
Burmeister, 1981 |
Farmers in Iowa |
105 |
1.1 (NS) |
VIETNAM VETERANS |
|||
New Studies |
|||
Dalager and Kang, 1997 |
Army Chemical Corps veterans |
4 |
1.5 (0.3-8.6) |
Watanabe and Kang, 1996 |
Army Vietnam veterans |
234 |
1.0 |
|
Marine Vietnam veterans |
73 |
1.3 (1.0-1.6) |
Studies reviewed in VAO |
|||
Anderson et al., 1986a |
Wisconsin Vietnam veterans |
6 |
0.9 (0.4-2.0) |
Anderson et al., 1986b |
Wisconsin Vietnam veterans |
5 |
1.3 (0.4-3.1) |
NOTE: NS = not significant. a Given when available. |
exposure or are simply random occurrences. The difficulty of diagnosing this condition also increases the concern that low rates could be appreciably affected by a small number of misclassifications among the vast majority of cancer cases that are not STS. This concern was born out in the IARC cohort, in which, upon review of pathology specimens, two of nine cases were judged not to have STS. There is no evidence that Vietnam veterans from the United States or Australia have experienced an elevated rate of this disease.
Because of difficulties in diagnosing this group of tumors, the epidemiologic studies reviewed by the committee were inconsistent with regard to the specific types of tumors included in the analyses. The data available did not permit the committee to determine whether specific forms of STS are or are not associated with TCDD and/or herbicides. Therefore, the committee's findings relate to the class of STS as a whole. Also, because of problems in defining exposure to herbicides in most occupational settings, as well as the difficulty of diagnosing STS, the rarity of the disease, and the restriction of the strongest findings to one country, the committee will continue to closely monitor any substantive new evidence that is relevant to this relationship.
Conclusions
Strength of Evidence in Epidemiologic Studies
No new evidence has been produced to change the committee's earlier judgment that evidence is sufficient to conclude that a positive statistical association exists between exposure to the herbicides (2,4-D, 2,4,5-T and its contaminant TCDD, cacodylic acid, and picloram) and soft-tissue sarcoma.
Biologic Plausibility
A thorough discussion of biologic plausibility with respect to exposure to TCDD or herbicides and soft-tissue sarcoma is contained in Chapter 3; a summary is presented in the conclusion to this chapter.
SKIN CANCERS
Skin cancers are generally divided into two broad categories, those neoplasms that develop from melanocytes (malignant melanoma) and those that do not. The common nonmelanocytic skin cancers, which include basal cell and squamous cell carcinomas, have a far higher incidence rate than malignant melanoma but are considered less aggressive and therefore more treatable. In VAO and Update 1996, all skin cancers were assessed together. However, with this review, the committee has decided to address studies assessing the health risk associated with malignant melanoma (Tables 7-12 and 7-13) separately from those assessing
the risk associated with nonmelanocytic cancers (basal and squamous cell carcinoma) (Tables 7-14 and 7-15). Because nonmelanocytic cancers are highly treatable, studies of these cancers have been divided further into those that discuss mortality and those that discuss incidence. Many studies report results by combining all types of skin cancers or do not specify the type of skin cancers assessed. These are also reported here in the interest of completeness (Tables 7-10 and 7-11).
Background
According to American Cancer Society estimates, 24,300 men and 17,300 women will be diagnosed with melanoma (ICD·9 172.0-172.9) in the United States in 1998, and 4,600 men and 2,700 women will die of this cancer (ACS, 1998). Approximately 1,000,000 cases of nonmelanocytic skin cancers (ICD·9 173.0-173.9), primarily basal cell and squamous cell carcinomas, are diagnosed in the United States each year (ACS, 1998). The American Cancer Society estimates that 1,200 men and 700 women will die from these diseases in 1998.
Skin cancers are far more likely to occur in fair-skinned individuals; the risk for whites is roughly 20 times that for dark-skinned African Americans. Incidence also increases with age, although more strikingly for males than females. Other risk factors for melanoma include the presence of certain moles on the skin, a suppressed immune system, and excessive exposure to ultraviolet (UV) radiation, typically from the sun. A family of history of the disease has been identified
TABLE 7-11 Selected Epidemiologic Studies—All (or unspecified) Skin Cancer Morbidity
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
OCCUPATIONAL |
|||
New Studies |
|||
Ott and Zober, 1996 |
German BASF trichlorophenol production workers |
5 |
1.2 (0.4-2.8) |
Studies reviewed in VAO |
|||
Hansen et al., 1992 |
Danish gardeners |
32 |
1.1 (0.8-1.6) |
Lynge, 1985 |
Danish male production workers |
14 |
0.7 |
Suskind and Hertzberg, 1984 |
Monsanto production workers |
8 |
1.6 |
VIETNAM VETERANS |
|||
Studies reviewed in VAO |
|||
Wolfe et al.. 1990 |
Air Force Ranch Hand veterans |
88 |
1.5 (1.1-2.0) |
CDC, 1988 |
Army enlisted Vietnam veterans |
15 |
0.8 (0.4-1.7) |
NOTE: NS = not significant. a Given when available. |
as a risk factor, but it is unclear whether this is due to genetic factors or to similarities in skin type and sun exposure patterns.
Excessive exposure to UV radiation is the single most important risk factor for nonmelanocytic skin cancers. Certain skin diseases and chemical exposures have also been identified as potential risk factors. SEER incidence data are not available for nonmelanocytic skin cancers.
Average Annual Cancer Incidence (per 100,000 individuals) in the United Statesa Melanomas of the Skin |
|||||||||
|
45-49 years of age |
50-54 years of age |
55-59 years of age |
||||||
|
all races |
white |
black |
all races |
white |
black |
all races |
white |
black |
males |
23 |
25 |
1 |
27 |
31 |
2 |
35 |
40 |
2 |
females |
19 |
22 |
18 |
21 |
0.4 |
20 |
23 |
3 |
|
a SEER nine standard registries crude, age-specific rate, 1990-1994. b Insufficient data to provide a meaningful incidence rate. |
MELANOMA
Summary of VAO and Update 1996
A number of mortality studies of agricultural workers showed either a deficit or no excess in skin cancer mortality (Wiklund, 1983; Blair et al., 1993; Asp et al., 1994). A study of Vietnam veterans with 181 cases of melanoma also showed either a deficit or no excess in skin cancer mortality among Army and Marine veterans (Breslin et al., 1988). One mortality study of environmental exposures with very few cases (N = 3) showed nonsignificant excesses of skin cancer (Bertazzi et al., 1989a).
If most of the deaths reported in Table 7-10, where all skin cancers are combined or where the type of skin cancer is unspecified, are assumed to be due to melanoma, then there are a few additional studies to examine. Two occupational cohorts with TCDD exposures showed deficits in melanoma deaths (Fingerhut et al., 1991; Saracci et al., 1991). Two studies of farm-related occupations showed small (RR = 1.1), nonsignificant increases in risk (Burmeister, 1981; Alavanja et al., 1988). The two Wisconsin Vietnam veteran studies reported, on one hand, a statistically nonsignificant elevation (RR = 1.3) and, on the other hand, a deficit (RR = 0.9) in melanoma based on five to six cases (Anderson et al., 1986a,b).
In terms of cancer incidence, a deficit of skin cancer cases was found among Danish farmers (Ronco et al., 1992). A study of Danish phenoxy herbicide and chlorophenol production workers showed a statistically significant excess of melanoma. This result, based on four cases, was observed in the subgroup of men who had been employed for at least one year, using a ten-year latency period (SIR = 4.3, CI 1.2-10.9) (Lynge, 1985). In addition, Air Force Ranch Hands had
four cases of melanoma, resulting in increased, but nonsignificant risk (SIR = 1.3, CI 0.3-5.2).
Update of the Scientific Literature
Mortality Studies
Since Update 1996, several studies of mortality in relevant cohorts have been published. Some of these are updates of previously reported cohorts. For example, the Seveso, Italy, cohort has been updated (Bertazzi et al., 1997). For zones A, B and R the only increased risk from melanoma was for zone A females based on only case (RR = 9.4, CI 0.1-52.3).
The IARC multinational cohort (Kogevinas et al., 1997) of production workers exposed to phenoxy herbicides, chlorophenols, and dioxins reported no increased risk of malignant melanoma among workers exposed to phenoxy herbicides contaminated with TCDD or higher chlorinated dioxins (SMR = 0.5, CI 0.2-3.2), or among workers with exposure to phenoxy herbicides and chlorophenols but minimal or no exposure to TCDD or higher chlorinated dioxins (SMR = 1.0, CI 0.3-2.4). These subcohorts had four or five cases of melanoma each.
New studies reviewed by the committee include a study of sawmill workers exposed to higher chlorinated PCDD's in chlorophenate wood preservatives. The reported SMR for melanoma in this cohort was 1.4 (CI 0.9-2.0) (Hertzman et al., 1997). A study of Swedish fishermen found a decreased risk of death from melanoma compared to the general Swedish population (Svensson et al., 1995). The fishermen included those from the west coast who reportedly ate lean fish with potentially higher arsenic contamination, and those from the east coat (Baltic) who consumed fatty fish with potentially higher levels of TCDD and other persistent organochlorine compounds. No actual measurements of TCDD or arsenic in the fish or fishermen were made.
The Australian veterans study (Crane et al., 1997a), which compared all Vietnam veterans to the Australian population, reported an elevated melanoma mortality for 1980-1994 (SMR = 1.3, CI 1.0-1.8). When the reference group was non-Vietnam veterans (Crane et al., 1997b), melanoma mortality for 1982-1994 was not elevated (RR = 0.5, CI 0.2-1.3).
Unfortunately, two of the new studies of Vietnam veterans either combined all skin cancer cases or did not specify the type of skin cancer studied. If it is assumed that most of these cancer deaths were due to melanoma, then these studies should be included in this review. A study of the Army Chemical Corps, the group responsible for spraying herbicides by helicopter and around base camp perimeters, found an elevated skin cancer risk of 1.5 (CI 0.3-8.6). This result was based on four cases and used non-Vietnam-based Chemical Corps veterans as the reference group (Dalager and Kang, 1997). When the U.S. male population was used as a reference group, the risk was 2.6 (CI 0.7-6.7).
Watanabe and Kang (1996) conducted a proportional mortality study of Vietnam veterans and found that Marine Vietnam veterans had a significantly higher mortality from all skin cancers compared to Marine non-Vietnam veterans (PMR = 1.3, CI 1.0-1.6), all non-Vietnam veterans (PMR = 1.3, CI 1.0-1.7) and the U.S. male population (PMR = 1.3, CI >1.0). Although control groups were presumably less exposed to Agent Orange and herbicides, no significant differences were observed between Army Vietnam veterans and any control group. Both Army and Marine veterans showed a trend toward higher PMRs with increased latency (i.e., time from last year in Vietnam to year of death).
Incidence
New studies include a study of sawmill workers exposed to higher chlorinated PCDDs that contaminate chlorophenate wood preservatives. The reported SIR for melanoma in this cohort was 1.0 (CI 0.7-1.2) (Hertzman et al., 1997). Also, the study of Swedish fishermen found no elevated incidence of melanoma compared to the general Swedish population (Svensson et al., 1995).
Among Vietnam veterans, there was a follow-up study of cancer incidence among Massachusetts Vietnam veterans. When Massachusetts Vietnam veterans were compared to other Vietnam era veterans, the authors reported an OR of 1.4 (CI 0.7-2.9) for melanoma, based on 21 cases (Clapp, 1997).
Synthesis
Known etiologic agents associated with melanoma include UV radiation or sunlight and polyaromatic hydrocarbons (PAHs). Cytochrome P4501A1 (CYP1A1), a P450 isozyme, under the control of the aryl hydrocarbon receptor (AhR), is expressed in the skin. It has been linked to the development of skin cancer and skin sensitization by PAHs (Gonzalez et al., 1996). Since TCDD toxicity is also mediated by the AhR, it is plausible that exposure to TCDD could be associated with increased risk of melanoma.
The epidemiologic data are not strong in this area. To date, only three studies show statistically significant increases in melanoma mortality. These include a study of phenoxy herbicide production workers, with four cases of melanoma (Lynge, 1985), and two Vietnam veteran studies. The Australian Vietnam veteran study had an SMR of 1.3 (CI 1.0-1.8) (Crane et al., 1997a). This increased risk was not found when the reference group was non-Vietnam veterans (Crane et al., 1997b). In the United States, a study of Marine and Army veterans found that only Marine veterans had an increased risk of melanoma (PMR = 1.3, CI 1.0-1.6) (Watanabe and Kang, 1996). None of these studies controlled for the greatest known risk factor for melanoma, sunlight exposure. Therefore the committee encourages future studies of the melanoma risk of occupational and Vietnam veteran populations to make an effort to control confounding from UV (sunlight) exposures.
Conclusions
Strength of Evidence in Epidemiologic Studies
There is inadequate or insufficient evidence to determine whether an association exists between exposure to the herbicides (2,4-D, 2,4,5-T and its contaminant TCDD, cacodylic acid, and picloram) and melanoma. The evidence regarding association is drawn from occupational and other studies in which subjects were exposed to a variety of herbicides and herbicide components.
TABLE 7-12 Selected Epidemiologic Studies—Melanoma Mortality
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
OCCUPATIONAL |
|||
New Studies |
|||
Hertzman et al., 1997 |
Sawmill workers |
17 |
1.4 (0.9-2.0) |
Kogevinas et al., 1997 |
IARC cohort |
|
|
|
Workers exposed to TCDD (or higher chlorinated dioxins) |
5 |
0.5 (0.2-3.2) |
|
Workers not exposed to TCDD (or higher chlorinated dioxins) |
4 |
1.0 (0.3-2.4) |
Svensson et al., 1995 |
Swedish fisherman |
|
|
|
East coast |
0 |
0.0 (0.0-1.73) |
|
West coast |
6 |
0.7 (0.2-1.5) |
Studies reviewed in Update 1996 |
|||
Blair et al., 1993 |
U.S. farmers in 23 states (white males) |
244 |
1.0 (0.8-1.1) |
Studies reviewed in VAO |
|||
Wigle et al., 1990 |
Saskatchewan farmers |
24 |
1.1 (0.7-1.6) |
Wiklund, 1983 |
Swedish agricultural workers |
268 |
0.8 (0.7-1.0)b |
ENVIRONMENTAL |
|||
New Studies |
|||
Bertazzi et al., 1997 |
Seveso residents |
|
|
|
Males zone R |
3 |
1.1 (0.2-3.2) |
|
Females zone R |
3 |
0.6 (0.1-1.8) |
Studies reviewed in VAO |
|||
Bertazzi et al., 1989a |
Seveso male residents—zones A, B, R |
3 |
3.3 (0.8-13.9) |
VIETNAM VETERANS |
|||
New Studies |
|||
Crane et al., 1997a |
Australian military veterans |
51 |
1.3 (1.0-1.8) |
Crane et al., 1997b |
Australian national service veterans |
16 |
0.5 (0.2-1.3) |
Studies reviewed in VAO |
|||
Breslin et al., 1988 |
Army Vietnam veterans |
145 |
1.0 (0.9-1.1) |
|
Marine Vietnam veterans |
36 |
0.9 (0.6-1.5) |
NOTE: NS = not significant. a Given when available. b 99% CI. |
TABLE 7-13 Selected Epidemiologic Studies—Melanoma Morbidity
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
OCCUPATIONAL |
|||
New Studies |
|||
Hertzman et al., 1997 |
Sawmill workers |
38 |
1.0 (0.7-1.2) |
Svensson et al., 1995 |
Swedish fisherman |
|
|
|
East coast |
0 |
0 (0.0-0.7) |
|
West coast |
20 |
0.8 (0.5-1.2) |
Studies reviewed in Update 1996 |
|||
Lynge, 1993 |
Danish male production workers |
4 |
4.3 (1.2-10.9) |
Studies reviewed in VAO |
|||
Ronco et al., 1992 |
Danish self-employed farmers |
72 |
0.7 (p < .05) |
VIETNAM VETERANS |
|||
New Studies |
|||
Clapp, 1997 |
Massachusetts Vietnam veterans |
21 |
1.4 (0.7-2.9) |
Studies reviewed in VAO |
|||
Wolfe et al., 1990 |
Air Force Ranch Hand veterans |
4 |
1.3 (0.3-5.2) |
NOTE: NS = not significant. a Given when available. |
Biologic Plausibility
Known etiologic agents associated with melanoma include UV radiation or sunlight and PAHs. CYP1A1, a P450 isozyme under the control of the AhR, is expressed in the skin and has been linked to the development of skin cancer and skin sensitization by PAHs (Gonzalez et al., 1996). Since TCDD toxicity is also mediated by the AhR, it is plausible that TCDD exposures could be associated with increased risk of melanoma. A thorough discussion of biologic plausibility with respect to exposure to TCDD or herbicides and melanoma is contained in Chapter 3; a summary is presented in the conclusion to this chapter.
BASAL AND SQUAMOUS CELL (NONMELANOMA) SKIN CANCER
Summary of VAO and Update 1996
Although the mortality from nonmelanoma skin cancer is said to be quite low, because of the highly successful treatment regimes available, several studies have reported cancer risk for the ICD·7 191 or ICD·8/ICD.9 173, the codes for other (nonmelanoma) skin cancers. Two of the mortality studies previously reviewed were of agricultural workers presumed to be exposed to herbicides. A large study of Swedish farmers, with 708 cases, found an SMR of 1.1 (99% CI
1.0-1.2) (Wiklund, 1983). In addition, a study of farmers in 23 U.S. states found a significantly elevated risk of nonmelanoma skin cancers in white male farmers with a PCMR of 1.1 (CI 1.0-1.2) based on 425 deaths (Blair et al., 1993). Although these were large studies, the strength of the association was weak and no verification of exposure status with regard to herbicides was done. Likewise, there was no controlling for confounding from UV (sunlight) exposure.
Since nonmelanocytic skin cancers are highly curable, studies of cancer incidence may be more helpful than mortality studies in evaluating the risk of basal and squamous cell skin cancers. A study of Danish farmers showed significantly reduced risk of these ''other" skin cancers (Ronco et al., 1992). In the Seveso cohort, a small number of cases have been identified, resulting in an increased risk in some analyses but no statistically significant finding (Pesatori et al., 1992; Bertazzi et al., 1993).
The strongest information in VAO and Update 1996 comes from the Air Force Ranch Hand study, which examined Vietnam veterans responsible for aerial herbicide spraying (Wolfe et al., 1990). Ranch Hands were compared to non-Ranch Hand Air Force air cargo personnel serving in Southeast Asia. Wolfe and colleagues found an OR of 1.5 (CI 1.1-2.0) for skin cancer of all types, which they called "sun-exposure neoplasms" (basal and squamous cell carcinomas, melanoma, and malignant epithelial neoplasms not otherwise specified). The Ranch Hand study also reported an OR of 1.5 (CI 1.0-2.1) for basal cell carcinoma alone. Squamous cell carcinoma had an elevated but nonsignificant OR of 1.6 (CI 0.5-5.1) due to the small number of cases (N = 6). To control for confounding by sun exposure, in addition to using a Vietnam-based reference group, these analysis were adjusted for two sun reaction indices and average residential latitude.
A number of studies reported results with all skin cancers combined or did not specify the type of skin cancer (Tables 7-10 and 7-11). If most of these incident cases are assumed to be the more common basal or squamous cell carcinomas, then such studies should be considered in this review. For example, the CDC (1988) studied Army personnel who served in Vietnam compared to non-Vietnam servicemen and found an OR of 0.8 (CI 0.5-1.2) for skin cancer prevalence on exam.
Two of the three occupational studies examining all types of skin cancer combined or unspecified skin cancer reported a nonstatistically significant increase in the incidence of skin cancer. The risk of skin cancer was elevated, but no confidence interval was given for the Nitro, West Virginia, work force involved in manufacturing 2,4,5-T (Suskind and Hertzberg, 1984). Hansen et al. (1992) found increased incidence of skin cancer among Danish gardeners (standardized morbidity ratio [SMbR] = 1.3 for males, CI 0.9-1.8). Studies of two phenoxy herbicide plants in Denmark found no statistically significant increase in skin cancer among exposed production workers (Lynge, 1985).
Update of the Scientific Literature
Mortality Studies
Since Update 1996, several mortality studies of relevant cohorts have been published. The IARC multinational cohort (Kogevinas et al., 1997) of workers exposed to phenoxy herbicides, chlorophenols, and dioxins reported an SMR of 1.2 (CI 0.3-3.2) for nonmelanocytic skin cancer among workers exposed to phenoxy herbicides contaminated with TCDD or higher chlorinated dioxins, based on four cases. There were no cases of skin cancer among those with minimal or no exposure to TCDD or higher chlorinated dioxins.
Svensson and colleagues (1995) reported a study of Swedish fishermen that compared those from the west coast, who were reported to eat lean fish containing potentially higher arsenic contamination, and those from the east coat (Baltic), who consumed fatty fish containing potentially higher levels of TCDD and other organochlorine compounds (Svensson et al., 1995). They report a nonsignificantly elevated risk of death from nonmelanoma skin cancers for west coast fishermen, compared to the general Swedish population (SMR = 3.1, CI 1.0-7.1). No cases were reported for east coast fishermen. No actual measurements of TCDD or arsenic in the fish or fishermen were made.
Cancer Incidence
The study of Swedish fishermen also reported a statistically significant increase in the incidence of nonmelanocytic skin cancer (RR = 2.3, CI 1.4-3.5) among east coast fisherman who ate more fatty fish containing potentially higher levels of TCDD and other organochlorine compounds. West coast fisherman had a nonsignificant elevation in skin cancer incidence (RR = 1.1, CI 0.9-1.4).
Zhong and Rafnsson (1996) reported on the cancer incidence among Icelandic pesticide users. This cohort included individuals who were allowed to buy and handle toxic chemicals including pesticides. Both 2,4,5-T and 2,4-D were used in Icelandic agriculture during the study period. Based on five cases of nonmelanocytic skin cancer, an nonstatistically significant elevation in skin cancer incidence was reported (SIR = 2.8, CI 0.9-6.6).
The strongest recent study in this area was a community case-control study of squamous cell and basal cell (nonmelanocytic) carcinoma, undertaken in Alberta, Canada (Gallagher et al., 1996). The study used a questionnaire to evaluate more than 50 different potential exposures. After adjustment for age, skin and hair color, and mother's ethnic origin, elevated risks for squamous cell carcinoma were seen in subjects ever exposed at work, at home, or by hobby to herbicides (OR = 1.5, CI 1.0-2.3); fungicides or seed treatments (OR = 1.4, CI 0.9-2.10); and insecticides (OR = 1.7, CI 1.1-2.7). Other exposures associated with a statistically significant increase in the risk of squamous cell carcinoma include coal
dust (OR = 1.6 CI 1.0-2.4) and diesel fumes (OR = 1.7 CI 1.1-2.5). No statistically significant associations of herbicide, fungicide, or insecticide exposure with risk of basal cell cancer were found. Exposures associated with a statistically significant increased risk of basal cell carcinoma include dry cleaning agents (OR = 4.6, CI 1.1-19.7), fiberglass dust (OR = 2.0, CI 1.1-3.9), luminous paint (OR = 6.7, CI 1.2-38.0), and asbestos dust (OR = 1.9, CI 1.0-3.5). No multivariate exposure analysis was done in this study.
For studies with a sufficient number of exposed subjects, the total duration of exposure until the time of diagnosis was weighted by the source (intensity) of exposure (direct job, workplace environment, hobby or home) and the duration of exposure (<1, 1-4, 5-19, or > 20 hours per week). This was converted to lifetime hours of exposure to the agent, and exposed subjects were dichotomized into low and high levels of exposure based on this lifetime exposure level. After adjustment for age, skin and hair color, mother's ethnic origin and sunlight exposure in the 10 years prior to diagnosis, statistically significant trends in risk for squamous cell carcinoma were seen with increasing exposure to herbicides, fungicides, seed treatments, and insecticides. For herbicides, the low-exposure group (33 cases) had an OR of 1.9 (CI 1.0-3.6), whereas the high-exposure group (46 cases) had an OR of 3.9 (CI 2.2-6.9) with a trend test p value of <.001. For fungicides and seed treatments, the low-exposure group (40 cases) had an OR of 0.8 (CI 0.4-1.4), and the high-exposure group (56 cases) had an OR of 2.4 (CI 1.4-4.0), with a trend test p value of .003. For insecticides, the low-exposure group (21 cases) had an OR of 0.7 (CI 0.3-1.4), and the high-exposure group (36 cases) had an OR of 2.8 (CI 1.4-5.6), with a trend test p value of .02. In examining the analysis for herbicides, it was noted that the dichotomous evaluation (ever or never) produced an OR lower than that of the low-exposure category in the exposure-response analysis, which also included an adjustment for occupational sunlight exposure. The change in OR between the two analyses suggests an unexpected negative confounding between sunlight and herbicide exposure. This casts some doubt on the strength of the finding and the reliability of the herbicide exposure assessment.
Synthesis
Known etiologic agents associated with basal and squamous cell carcinoma include arsenic and UV radiation or sunlight. Cacodylic acid was an organic arsenic herbicide widely used in Vietnam. CYP1A1, a P450 isozyme, under the control of the AhR is expressed in the skin. It has been linked to the development of skin cancer and skin sensitization by polyaromatic hydrocarbons (Gonzalez et al., 1996). Since TCDD toxicity is also mediated by the AhR, it is plausible that TCDD exposures could be associated with increased risk of skin cancer.
In VAO and Update 1996, an increased risk of mortality from nonmelanoma skin cancers was found among agricultural workers. However, the strength of the association was weak and herbicide exposure was not verified. In addition, these
studies did not control for the most likely source of the slightly elevated risk, exposure to sunlight. The Air Force Ranch Hand study (Wolfe et al., 1990) was also reported in VAO and Update 1996. This study showed a small, but significant, increased incidence of basal cell carcinoma (OR = 1.5, CI 1.0-2.1). The analysis controlled for confounding by sun exposure by using a Vietnam-based reference group (non-Ranch Hand Air Force air cargo crews in Southeast Asia), as well as two personal sun reaction indices and average residential latitude.
The strongest evidence linking herbicide exposure and these other skin cancers comes from a recent community-based case-control study (Gallagher et al., 1996). This study controlled for sun exposure, skin and hair color, and mother's ethnic origin, and found increasing risk of squamous cell carcinoma with increasing lifetime exposure to herbicides. At issue in this study are the large number of different potential exposures examined (>50) and the lack of any multivariate exposure analysis. A number of these exposures might be expected to be statistically significant based on chance alone. However, the exposure-response gradients observed lend credibility to the finding. Other concerns are misclassification and bias introduced through the use of exposure interviews and the attendant exposure algorithm to estimate exposure. Since the associations with herbicides in this study were seen with squamous but not basal cell carcinomas, this argues
TABLE 7-14 Selected Epidemiologic Studies—Other Nonmelanoma (basal and squamous cell) Skin Cancer Mortality
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
OCCUPATIONAL |
|||
New Studies |
|||
Hertzman et al., 1997 |
Sawmill workers |
38 |
1.0 (0.7-1.2) |
Kogevinas et al., 1997 |
IARC cohort |
|
|
|
Workers exposed to TCDD (or higher chlorinated dioxins) |
4 |
1.2 (0.3-3.2) |
|
Workers not exposed to TCDD (or higher chlorinated dioxins) |
0 |
— |
Svensson et al., 1995 |
Swedish fisherman |
|
|
|
East coast |
0 |
0.0 (0-15.4) |
|
West coast |
5 |
3.0 (1.0-7.1) |
Studies reviewed in Update 1996 |
|||
Blair et al., 1993 |
U.S. farmers in 23 states (white males) |
425 |
1.1 (1.0-1.2) |
Studies reviewed in VAO |
|||
Coggon et al., 1986 |
British MCPA chemical workers |
3 |
3.1 (0.6-9.0) |
Wiklund, 1983 |
Swedish agricultural workers |
708 |
1.1 (1.0-1.2)b |
NOTE: NS = not significant. a Given when available. b 99% CI. |
against exposure misclassification or bias. On the other hand, the lack of an association between herbicide exposure and basal or squamous cell cancers casts some doubt on the biologic plausibility of the association. Of note here is the conflicting finding of a significantly elevated risk of basal cell, but not squamous cell, carcinoma in the Ranch Hand study. Thus, although the committee agrees that the Gallagher et al. (1996) study is the best to date, concerns still remained regarding the control of confounding and the adequacy of exposure assessment. Therefore, the committee encourages further study of basal and squamous cell skin cancer incidence among working and Vietnam veteran populations. In any future studies, careful attention should be paid to exposure assessment, as well as to controlling for confounding from UV exposure. In addition, efforts to examine the carcinogenicity of organic arsenicals should be encouraged.
Conclusions
Strength of Evidence in Epidemiologic Studies
There is inadequate/insufficient evidence of an association between exposure to the herbicides (2,4-D, 2,4,5-T and its contaminant TCDD, cacodylic acid, and picloram) and skin cancer. The evidence regarding association is drawn from occupational and other studies in which subjects were exposed to a variety of herbicides and herbicide components.
Biologic Plausibility
Known etiologic agents associated with basal and squamous cell carcinoma include arsenic and UV radiation or sunlight. Cacodylic acid was an organic arsenic herbicide used widely in Vietnam. CYP1A1, a P450 isozyme under the control of the AhR, is expressed in the skin. It has been linked to the development of skin cancer and skin sensitization by PAHs (Gonzalez et al., 1996). Since TCDD toxicity is also mediated by the AhR, it is plausible that TCDD exposures could be associated with increased risk of skin cancer. A thorough discussion of biologic plausibility with respect to exposure to TCDD or herbicides and skin cancer is contained in Chapter 3; a summary is presented in the conclusion to this chapter.
BREAST CANCER
Background
Breast cancer (ICD·9 174.0-174.9 for females) is the single most common cause of cancer among women in the United States, excluding nonmelanocytic skin cancers. The American Cancer Society estimates that 178,700 women will
TABLE 7-15 Selected Epidemiologic Studies—Other Nonmelanoma (basal and squamous cell) Skin Cancer Morbidity
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
OCCUPATIONAL |
|||
New Studies |
|||
Zhong and Rafnsson, 1996 |
Icelandic pesticide users |
5 |
2.8 (0.9-6.6) |
Svensson et al., 1995 |
Swedish fisherman |
|
|
|
East coast |
22 |
2.3 (1.4-3.5) |
|
West coast |
69 |
1.1 (0.9-1.4) |
Studies reviewed in VAO |
|||
Ronco et al., 1992 |
Danish self-employed farmers |
493 |
0.7 (p < .05) |
ENVIRONMENTAL |
|||
New Studies |
|||
Gallagher et al., 1996 |
Alberta, Canada, residents— |
|
|
|
Squamous cell Carcinoma |
|
|
|
All herbicide exposure |
79 |
1.5 (1.0-2.3) |
|
Low herbicide exposure |
33 |
1.9 (1.0-3.6) |
|
High herbicide exposure |
46 |
3.9 (2.2-6.9) |
|
All fungicide exposure |
96 |
1.4 (0.9-2.1) |
|
Low fungicide exposure |
40 |
0.8 (0.4-1.4) |
|
High fungicide exposure |
56 |
2.4 (1.4-4.0) |
|
Alberta, Canada, residents— |
|
|
|
Basal cell Carcinoma |
|
|
|
All herbicide exposure |
70 |
1.1 (0.8-1.7) |
|
All fungicide exposure |
76 |
0.9 (0.6-1.3) |
Studies reviewed in Update 1996 |
|||
Bertazzi et al., 1993 |
Seveso male residents |
|
|
|
zone A |
1 |
2.4 (0.3-17.2) |
|
zone B |
2 |
0.7 (0.2-2.9) |
|
zone R |
20 |
1.0 (0.6-1.6) |
Studies reviewed in VAO |
|||
Pesatori et al., 1992 |
Seveso male residents—zones A and B |
3 |
1.0 (0.3-3.0) |
|
Female residents—zones A and B |
3 |
1.5 (0.5-4.9) |
VIETNAM VETERANS |
|||
Studies reviewed in VAO |
|||
Wolfe et al., 1990 |
Air Force Ranch Hand veterans |
|
|
|
Basal cell carcinoma |
78 |
1.5 (1.0-2.1) |
|
Squamous cell carcinoma |
6 |
1.6 (0.5-5.1) |
NOTE: NS = not significant. a Given when available. |
be diagnosed with breast cancer in the United States in 1998 and that 43,500 will die from the disease (ACS, 1998). Overall, these numbers represent approximately 30 percent of the incidence of new cancer and 16 percent of cancer deaths. Among women aged 40-55, it is the leading cause of cancer death.
Breast cancer incidence generally increases with age. In the age groups that characterize most Vietnam veterans, the incidence for whites is slightly higher than for African Americans. Risk factors other than aging include a personal or family history of breast cancer and reproductive history (specifically, early onset of menarche, late onset of menopause, and either no pregnancies or first full-term pregnancy after 30 years of age). A pooled analysis of six large-scale prospective studies of invasive breast cancer found that alcohol consumption was associated with a linear increase in incidence in women, over the range of consumption reported by most women (Smith-Warner et al., 1998). The potential role of other personal behaviors and environmental factors in breast cancer incidence is being studied extensively.
Average Annual Cancer Incidence (per 100,000 individuals) in the United Statesa Breast Cancer in Females |
|||||||||
|
45-49 years of age |
50-54 years of age |
55-59 years of age |
||||||
|
all races |
white |
black |
all races |
white |
black |
all races |
white |
black |
|
199 |
203 |
203 |
245 |
253 |
217 |
284 |
291 |
267 |
a SEER nine standard registries crude, age-specific rate, 1990-1994. |
Summary of VAO and Update 1996
Data relating herbicide exposure to cancer among women are extremely limited. The committee has attempted to examine cancer among women separately from cancer among men. However, although the data available for men are sparse, data for women are almost nonexistent. Many studies have excluded women from analysis because of their small numbers in the groups under study. For example, in their follow-up of workers from 12 companies, Fingerhut et al. (1991) identified 67 women who were then excluded from the report. Likewise, Moses et al. (1984) excluded three women from their follow-up analysis of workers, and Zack and Suskind (1980) excluded the one woman who was living at the end of this study. Among the studies based on follow-up of workers, women contributed a minor portion of the data.
Occupational Studies
Manz et al. (1991) described a retrospective cohort study of chemical workers employed in an herbicide plant in Hamburg, Germany. The SMR, based on nine breast cancer deaths, was elevated, at 2.2 (CI 1.0-4.1). The small percentage
(7 percent) of women in this study who worked in high-exposure departments precluded separate examination of those with high exposure.
In a study focusing on all persons employed in the manufacture of phenoxy herbicides in Denmark before 1982, Lynge (1985) linked employment records for 1,069 women (contributing 17,624 person-years of follow-up) with the National Cancer Register. Thirteen cases of breast cancer were diagnosed, yielding an SMR of 0.9.
Saracci et al. (1991) reported one breast cancer death among 1,527 women employed in 11 plants with potential exposure to phenoxy herbicides or TCDD in the multicenter IARC occupational study; the mortality rate was lower than among unexposed women in this study. Additional data on morbidity in a subset (N = 701) of these women were described by Kogevinas et al. (1993), who found no excess incidence of breast cancer: seven cases resulted in an SIR of 0.9 (CI 0.4-1.9). Among those female workers exposed to chlorophenoxy herbicides contaminated with TCDD (N = 169), there was an excess overall cancer incidence, based on nine cases (SIR = 2.2, CI 1.0-4.2), with one case of breast cancer (SIR = 0.9, CI 0.0-4.8).
Among women farm workers in Denmark (Ronco et al., 1992), 429 cases of breast cancer were diagnosed, and the SIR of 0.8 was significantly less than unity. The actual level of exposure of these women to herbicides is not defined, however, and it is possible that the reduced incidence of breast cancer reflects general patterns of female cancers seen elsewhere, in which rates are lower for rural than for urban populations. In a similar occupational study based on census data for economically active women in Sweden (Wiklund, 1983), the SIR for breast cancer among farm workers was 0.8. This result is not adjusted for reproductive risk factors for these cancers, and the actual exposures of interest are not defined.
In the United States, a PCMR study was performed using death certificate data for male and female farmers from 23 states (Blair et al., 1993). Occupational and industry data were coded based on the information listed on death certificates. Among white female farmers, 71 deaths from breast cancer yielded a PCMR of 1.0 (CI 0.8-1.3); among nonwhite female farmers, the PCMR, based on 30 breast cancer deaths, was significantly decreased at 0.7 (0.5-1.0).
The 10-year Seveso follow-up (Bertazzi et al., 1989b) provides limited information for women in the high-and medium-exposure groups. Person-years of follow-up were 2,490 in zone A (high exposure); 16,707 in zone B; 114,558 in zone R; and 726,014 in the reference area. Because of the small number of cancer deaths (three) in females in zone A, no conclusions are possible. Among the 14 deaths of zone B women, 5 were due to breast cancer, resulting in a mortality ratio of 0.9 (CI 0.4-2.1). In zone R, the least contaminated area, 28 women died from breast cancer, giving a significantly reduced estimated RR of 0.6 (CI 0.4-0.9).
In another report from Seveso (Bertazzi et al., 1993), cancer incidence during the first 10 years following exposure to TCDD was investigated. In zones A, B and R, the number of breast cancers diagnosed and the corresponding RRs, were respectively, one case in zone A (RR = 0.5, CI 0.1-3.3); 10 in zone B (RR = 0.7, CI 0.4-1.4); and 106 in zone R (RR = 1.1, CI 0.9-1.3), the least contaminated area.
Vietnam Veteran Studies
Thomas et al. (1991) assembled a list of female Vietnam veterans and followed them from 1973 to 1987. Cause-specific estimates of mortality risk among women Vietnam veterans relative to those for Vietnam era veterans who served elsewhere were derived from proportional hazards multivariate models adjusted for rank (officer or enlisted), occupation (nurse or nonnurse), duration of service (at least 10 years), age at entry to follow-up, and race. Of these women, 80 percent were classified as officers or nurses, and the majority served between 3 and 19 years. Slightly more than one-fourth of the cancer deaths among Vietnam veterans were due to breast cancer. The RR was not significantly elevated (RR = 1.2, CI 0.6-2.5) compared to that for other Vietnam era veterans.
Cancer mortality rates for 4,586 female Vietnam veterans were recently compared with rates for 5,325 female veterans who served elsewhere (Dalager et al., 1995a). This study extended the follow-up of Thomas et al. (1991) for four years, through 1991. There were 196 deaths observed among the Vietnam veterans. Based on 26 deaths from breast cancer among Vietnam veterans, the RR was 1.0 (CI 0.6-1.8).
Update of the Scientific Literature
Few new data have been published since Update 1996. The 15-year follow-up of the Seveso, Italy, population indicates no excess of breast cancer: the RRs are 0.6 for the most highly exposed (zone A), and 0.8 for the middle-and low-exposure groups (zones B and R), compared to an unexposed population. Although there was 1 case in zone A, zones B and R had 9 and 67 cases, respectively, leading to fairly stable measures of association. In particular, the 95% CI for the female breast cancer RR of 0.8 in zone R was 0.6-1.0. Thus, evidence from this study suggests a possible protective effect of TCDD exposure.
In contrast, the multinational study conducted by IARC observed an SMR of 2.2 for female breast cancers (95% CI 1.0-4.1) and 2.6 for male breast cancers (CI 0.3-9.3), for workers exposed to TCDD or higher chlorinated dioxins (Kogevinas et al., 1997). The finding for women workers was confined to women from one cohort in Germany, where the SMR was 2.8 (CI 1.3-5.4). This was the only cohort in the IARC study that had a substantial number of female production workers with exposure to TCDD or higher chlorinated dioxins.
Data from Australian Vietnam veterans also indicate an elevation of male breast cancer (Crane et al., 1997a): the SMR was 5.5 (95% CI 1.1-16.1), based on three deaths during 1980-1994. The SRMR comparing male breast cancer deaths to all other deaths was 5.2 (CI 1.04->10.0); compared to other cancer deaths, it was 4.6.
Synthesis
Through Update 1996, there had been a few occupational studies, two environmental studies, and two veteran studies of breast cancer among women exposed to herbicides and/or TCDD (Table 7-16). Most of these studies reported a RR of approximately 1.0 or less, but it is uncertain whether or not female members of the cohorts had substantial chemical exposure. Further follow-up of the Seveso cohort indicates, if anything, a protective effect. However, an expanded multinational occupational study found an increased risk of breast cancer in both males (nonsignificant) and females (significant). This cohort may have been the only occupational one, among those in which breast cancer was analyzed, that had substantial exposure to TCDD. Also notable is the finding, both in this occupational cohort and among Australian Vietnam veterans, of an elevated risk of male breast cancer, despite its rarity. In contrast, TCDD appears to exert a protective effect on the incidence of mammary tumors in experimental animals (see Chapter 3), which is consistent with the tendency for RRs to be less than 1.0. In summary, data continue to be inconclusive regarding whether an association exists between exposure to the herbicides used in Vietnam and the occurrence of breast cancer.
Conclusions
Strength of Evidence in Epidemiologic Studies
There is inadequate or insufficient evidence to determine whether an association exists between exposure to the herbicides (2,4-D, 2,4,5-T and its contaminant TCDD, cacodylic acid, and picloram) and breast cancer. The evidence regarding association is drawn from occupational and other studies in which subjects were exposed to a variety of herbicides and herbicide components.
Biologic Plausibility
In laboratory animals, TCDD has been shown to have a wide range of effects on growth, hormone systems, and other factors associated with the regulation of activities in normal cells. Because animal data suggest that TCDD may act as an antiestrogen, and because it has been shown to inhibit the growth of breast cancer cell lines in tissue culture, the extrapolation to prevention of breast cancers in
TABLE 7-16 Selected Epidemiologic Studies—Breast Cancer
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
OCCUPATIONAL |
|||
New Studies |
|||
Kogevinas et al., 1997 |
IARC cohort, female: identical to Manz et al., 1991 |
9 |
2.2 (1.0-4.1) |
|
IARC cohort, male |
2 |
2.6 (0.3-9.3) |
Studies reviewed in Update 1996 |
|||
Blair et al., 1993 |
Female U.S. farmers in 23 states |
|
|
|
Whites |
71 |
1.0 (0.8-1.3) |
|
Nonwhites |
30 |
0.7 (0.5-1.0) |
Kogevinas et al., 1993 |
Female herbicide spraying and production workers |
7 |
0.9 (0.4-1.9) |
|
Probable exposed to TCDD |
1 |
0.9 (0.0-4.8) |
Studies reviewed in VAO |
|||
Ronco et al., 1992 |
Danish family farm workers |
429 |
0.8 (p < .05) |
Manz et al., 1991 |
German production workers |
9 |
2.2 (1.0-4.1) |
Saracci et al., 1991 |
IARC cohort |
1 |
0.3 (0.0-1.7) |
Lynge, 1985 |
Danish production workers |
13 |
0.9 |
Wiklund, 1983 |
Swedish agricultural workers |
444 |
0.8 (0.7-0.9)b |
ENVIRONMENTAL |
|||
New Studies |
|||
Bertazzi et al., 1997 |
Seveso female residents, mortality |
|
|
|
zone A |
1 |
0.6 (0.0-3.1) |
|
zone B |
9 |
0.9 (0.4-1.5) |
|
zone R |
67 |
0.8 (0.6-1.0) |
Studies reviewed in Update 1996 |
|||
Bertazzi et al., 1993 |
Seveso female residents; follow-up of Bertazzi et al., 1989b |
|
|
|
zone A |
1 |
0.5 (0.1-3.3) |
|
zone B |
10 |
0.7 (0.4-1.4) |
|
zone R |
106 |
1.1 (0.9-1.3) |
Studies reviewed in VAO |
|||
Bertazzi et al., 1989b |
Seveso female residents |
|
|
|
zone B |
5 |
0.9 (0.4-2.1) |
|
zone R |
28 |
0.6 (0.4-0.9) |
VIETNAM VETERANS |
|||
New Studies |
|||
Crane et al., 1997a |
Australian military veterans |
3 |
5.5 (1.1-16.1) |
Studies reviewed in Update 1996 |
|||
Dalager et al., 1995a |
Women Vietnam veterans |
26 |
1.0 (0.6-1.8) |
Studies reviewed in VAO |
|||
Thomas et al., 1991 |
Women Vietnam veterans |
17 |
1.2 (0.6-2.5) |
a Given when available. b 99% CI. |
humans is plausible; however, such an effect has not been clearly demonstrated. Furthermore, a recent occupational study that found a twofold increase in breast cancer in women and two reports of elevated breast cancer mortality in males provide evidence that such extrapolation is inappropriate at this time. A more thorough discussion of TCDD's antiestrogenic properties and of the effect of TCDD or herbicides on breast cells is contained in Chapter 3; a summary discussion of biologic plausibility is presented in the conclusion to this chapter.
Increased Risk of Disease Among Vietnam Veterans
Studies in female Vietnam veterans have not shown an elevated risk of breast cancer, but the length of follow-up may not have been sufficient for an effect, if it exists, to become evident. A report on Australian male Vietnam veterans found three breast cancer deaths, where less than one was expected. Further follow-up among both male and female veterans from the Vietnam era is needed before conclusions can be drawn. A more thorough discussion of the issue of increased risk of disease among Vietnam veterans is included in Chapter 1.
CANCERS OF THE FEMALE REPRODUCTIVE SYSTEM
Background
This section addresses cancers of the cervix (ICD·9 180.0-180.9), endometrium (also referred to as the corpus uteri, ICD-9 182.0-182.1, 182.8), and ovaries (ICD·9 183.0). Statistics for other cancers of the female reproductive system are presented as well. The American Cancer Society estimates the following numbers of new female reproductive system cancer in the United States for 1998 (ACS, 1998):
Site |
New cases |
Deaths |
Cervix |
13,700 |
4,900 |
Endometrium |
36,100 |
6,300 |
Ovary |
25,400 |
14,500 |
Other female genital |
5,200 |
1,400 |
Taken together, these numbers represent roughly 13 percent of new cancer diagnoses and 10 percent of cancer deaths in women.
Incidence patterns and risk factors vary for these diseases. Cervical cancers occur more often in African-American women than in whites, whereas whites are more likely to develop endometrial and ovarian cancers. The incidence of endometrial and ovarian cancer is also dependent on age, with older women at greater risk. Other risk factors for these cancers vary. Human papillomavirus infection is the most important risk factor for cervical cancer. Diet, a family
history of the disease, and breast cancer are among the risk factors for endometrial and ovarian cancers.
Average Annual Cancer Incidence (per 100,000 individuals) in the United Statesa Female Genital System Cancers |
|||||||||
|
45-49 years of age |
50-54 years of age |
55-59 years of age |
||||||
|
all races |
white |
black |
all races |
white |
black |
all races |
white |
black |
cervix |
17 |
16 |
23 |
17 |
15 |
23 |
16 |
13 |
27 |
endometrium |
24 |
25 |
12 |
41 |
44 |
19 |
62 |
65 |
40 |
ovary |
22 |
24 |
13 |
29 |
29 |
26 |
38 |
40 |
21 |
other genital |
3 |
3 |
5 |
4 |
5 |
4 |
6 |
6 |
7 |
overallb |
66 |
67 |
53 |
91 |
92 |
71 |
121 |
125 |
95 |
a SEER nine standard registries crude, age-specific rate, 1990-1994. b Totals may differ from the sum of the column because of independent rounding. |
Summary of VAO and Update 1996
Occupational studies that examined the relationship between exposure to herbicides and uterine or ovarian cancers were extremely limited, largely because most occupational cohorts have included few exposed female workers. In a case-control study specifically designed to address the relation between herbicide exposure and risk of ovarian cancer, Donna et al. (1984) compared exposure histories of 60 women with ovarian cancer to controls (i.e., women with cancers at other sites, including breast, endometrium, cervix, and other organs). Overall, 18 women with ovarian cancer were classified as definitely or probably exposed, compared to 14 controls, giving an OR of 4.4 (CI 1.9-16.1). These findings of elevated risk for ovarian cancer, although striking, have not been replicated in studies conducted since that time. Other studies reporting no association between cancers of female reproductive organs and exposure to herbicides are Wiklund (1983), Saracci et al. (1991), and Ronco et al. (1992). In a cohort of farm workers in Denmark, Ronco et al. (1992) observed SIRs for cervical cancer, uterine cancer, and ovarian cancer, each based on 100 or more cases, that were all significantly less than 1.0.
In the first 10 years following the 1976 Seveso, Italy, accident, Bertazzi et al. (1989b, 1993) observed no increase in cancers of the uterus or cervix among female residents of zones A, B, or R, based on 2, 2, and 21 cases, respectively; the RRs were 2.6, 0.4, and 0.6, respectively. No cases of ovarian cancer were diagnosed among women in zones A or B. Based on 20 cases, the RR for ovarian cancer in zone R, the least contaminated area, was 1.1 (CI 0.7-1.7).
A cohort study of cancer incidence was conducted among employees of two phenoxy herbicide manufacturing facilities in Denmark (Lynge, 1993). This cohort included 1,071 women who were followed for 1947-1987. A statistically significant increase in risk of cervical cancer was found, based on seven cases
(SIR = 3.2, CI 1.3-6.6). An overlapping cohort study of cancer incidence and mortality was conducted among 701 women in the IARC cohort (66 percent were included in the study by Lynge, 1993), who were occupationally exposed to chlorophenoxy herbicides, chlorophenols, and dioxins (Kogevinas et al., 1993). One death was observed from each of the following types of cancer: cervical (SMR = 80), uterus nonspecified (SMR = 192), and ovary (SMR = 74).
In a study of death certificates, Blair et al. (1993) compared occupational groups and found that farmers in 23 states in the United States had an elevated risk for cervical cancer mortality, based on 21 deaths, among nonwhites only (PCMR = 2.0, CI 1.3-3.1). In whites, the PCMR, based on six deaths, was 0.9 (CI 0.3-2.0). The numbers of deaths from cancer of the uterine corpus were small and nonsignificant for both white and nonwhite female farmers.
Cancer mortality rates among 4,586 female Vietnam veterans were recently evaluated, as well as rates among 5,325 female veterans who had served elsewhere (Dalager et al., 1995a). Based on four cases of cancer of the uterine corpus, the RR for Vietnam veterans compared to that of the general U.S. population was 2.1 (CI 0.6-5.4).
Update of the Scientific Literature
Bertazzi et al. (1997) have updated their follow-up of the Seveso population. With 15 years of follow-up of the more than 20,000 exposed women, this cohort provides no evidence that TCDD is associated with deaths from either uterine or ovarian cancer. In zone A, which had the highest exposures, and zone B, with probably the second-highest exposures, the numbers of uterine cancer deaths were lower than expected. In zone R, where exposures were lower, but still greater than those not exposed to the accident, 27 uterine cancers were observed, with 23.7 expected, for a RR of 1.1 (95% CI 0.8-1.7). The numbers of deaths from ovarian cancer were 1 and 0 in zones A and B, respectively, where 0.4 and 2.7 were expected, yielding an elevated but unstable RR for zone A of 2.3 (CI 0.0-12.8). In zone R, the RR was 1.0 (CI 0.6-1.6). If the TCDD did initiate cancers of female reproductive organs, the elapsed time may still be insufficient for these tumors to have come to clinical attention. In particular, women exposed to TCDD during adolescence may be at increased risk for cancers that will not be detected until 20 or more years after the exposure.
The only other new data on female reproductive cancers come from an expanded multinational occupational study conducted by IARC of cohorts from 12 countries (Kogevinas et al., 1997). A total of 36 cohorts were included; 15 of these included females, and 11 of these cohorts had exposures to TCDD or higher chlorinated dioxins. No deaths from cancer of the uterine cervix or the ovary were observed among women workers exposed to TCDD or higher chlorinated dioxins. An SMR of 3.41, based on three cases with exposure to TCDD or higher chlorinated dioxins, was observed for cancer of the endometrium and uterus
(ICD·9 179, 181-182). Two of these three cases occurred in one cohort in Germany, which included most of the TCDD-exposed female production workers.
Synthesis
There has been considerable recent interest in the potential association of exposure to chlorinated hydrocarbons, including TCDD, and female reproductive cancers and other health outcomes in women. For example, teratogenic effects due to maternal exposures to TCDD have been well documented in experimental animals (see Chapter 3). Endometriosis has recently been demonstrated in monkeys exposed to TCDD (Rier et al., 1993), and research on this disease has been proposed for women in the Seveso cohort (Bois and Eskenazi, 1994). Epidemiologic studies regarding female reproductive tract cancers are summarized in Tables 7-17, 7-18, and 7-19. The evidence from these studies remains inconclusive, despite some strong associations with ovarian and uterine cancers, largely because most of the published studies include a small number of cases and/or have poor exposure characterization or too short a follow-up period. The committee concludes that more research is needed on populations of women with documented exposure to herbicides and TCDD.
Conclusions
Strength of Evidence in Epidelniologic Studies
There is inadequate or insufficient evidence to determine whether an association exists between exposure to the herbicides used in Vietnam (2,4-D, 2,4,5-T and its
TABLE 7-17 Selected Epidemiologic Studies—Cancers of the Cervix
Reference |
Study Population |
Exposed Cases |
Estimated Risk (95% CI) |
OCCUPATIONAL |
|||
New Studies |
|||
Kogevinas et al., 1997 |
IARC cohort |
0 |
0 |
Studies reviewed in Update 1996 |
|||
Blair et al., 1993 |
U.S. farmers in 23 states |
|
|
|
Whites |
6 |
0.9 (0.3-2.0) |
|
Nonwhites |
21 |
2.0 (0.3-3.1) |
Studies reviewed in VAO |
|||
Ronco et al., 1992 |
Danish farmers |
|
|
|
Self-employed farmers |
7 |
0.5 |
|
Family workers |
100 |
0.5 |
|
Employees |
12 |
0.8 |
Wiklund, 1983 |
Swedish agricultural workers |
82 |
0.6 |
TABLE 7-18 Selected Epidemiologic Studies—Cancers of the Uterus
Reference |
Study Population |
Exposed Cases |
Estimated Risk (95% CI) |
OCCUPATIONAL |
|||
New Studies |
|||
Kogevinas et al., 1997 |
IARC cohort (includes cancers of the endometrium) |
3 |
3.4 (0.7-10.0) |
Studies reviewed in VAO |
|||
Blair et al., 1993 |
U.S. farmers in 23 states |
|
|
|
Whites |
15 |
1.2 |
|
Nonwhites |
17 |
1.4 |
Ronco et al., 1992 |
Danish farmers |
|
|
|
Self-employed farmers |
8 |
0.6 |
|
Family workers |
103 |
0.8 |
|
Employees |
9 |
0.9 |
Wiklund, 1983 |
Swedish agricultural workers |
135 |
0.9 |
ENVIRONMENTAL |
|||
New Studies |
|||
Bertazzi et al., 1997 |
Seveso residents |
0 |
0.0 |
VIETNAM VETERANS |
|||
Studies reviewed in Update 1996 |
|||
Dalager et al., 1995a |
Women Vietnam veterans |
4 |
2.1 (0.6-5.4) |
TABLE 7-19 Selected Epidemiologic Studies—Ovarian Cancer
Reference |
Study Population |
Exposed Cases |
Estimated Risk (95% CI) |
OCCUPATIONAL |
|||
New Studies |
|||
Kogevinas et al., 1997 |
IARC cohort |
0 |
0.0 |
Studies reviewed in Update 1996 |
|||
Kogevinas et al., 1993 |
IARC cohort |
1 |
0.7 |
Lynge, 1993 |
Danish male production workers |
7 |
3.2 |
Studies reviewed in VAO |
|||
Ronco et al., 1992 |
Danish farmers |
|
|
|
Self-employed farmers |
12 |
0.9 |
|
Family workers |
104 |
0.8 |
|
Employees |
5 |
0.5 |
Donna et al., 1984 |
Female residents near Alessandria, Italy |
18 |
4.4 (1.9-16.1) |
ENVIRONMENTAL |
|||
New Studies |
|||
Bertazzi et al., 1997 |
Seveso residents |
|
|
|
zone A |
1 |
2.3 (0.0-12.8) |
|
zone B |
0 |
0.0 |
|
zone R |
21 |
1.0 (0.6-1.6) |
contaminant TCDD, cacodylic acid, and picloram) and uterine and ovarian cancers. The evidence regarding association is drawn from occupational and other studies in which subjects were exposed to a variety of herbicides and herbicide components.
Biologic Plausibility
A thorough discussion of biologic plausibility with respect to exposure to TCDD or herbicides and uterine and ovarian cancers is contained in Chapter 3; a summary is presented in the conclusion to this chapter.
Increased Risk of Disease Among Vietnam Veterans
Although there are inadequate data at this time to evaluate the possible increased risk of female reproductive cancers in Vietnam veterans, an ongoing study in female veterans of the Vietnam era may shed light on this issue. A more thorough discussion of the issue of increased risk of disease among Vietnam veterans is included in Chapter 1.
PROSTATE CANCER
Background
According to the American Cancer Society estimates, 184,500 new cases of prostate cancer (ICD·9 185) will be diagnosed in the United States in 1998, and 39,200 men will die from the disease (ACS, 1998). This makes prostate cancer the most common cancer among men, excluding nonmelanocytic skin cancers. It is expected to account for approximately 29 percent of new diagnoses and 13 percent of cancer deaths in 1998.
Prostate cancer incidence varies dramatically as a function of age and race. The risk increases fivefold between 45-49 and 50-54 years of age, and nearly triples between 50-54 and 55-59 years of age. As a group, African-American men have the highest recorded incidence of prostate cancer in the world (Miller et al., 1996). Their risk is roughly twice that of whites in the United States, 4 times higher than Alaskan natives, and nearly 7.5 times higher than Korean Americans.
Little is known about the causes of prostate cancer. Other than race and age, risk factors include a family history of the disease and a diet high in fats.
Average Annual Cancer Incidence (per 100,000 individuals) in the United Statesa Prostate Cancer |
|||||||||
|
45-49 years of age |
50-54 years of age |
55-59 years of age |
||||||
|
all races |
white |
black |
all races |
white |
black |
all races |
white |
black |
|
16 |
16 |
26 |
82 |
78 |
143 |
225 |
216 |
404 |
a SEER nine standard registries crude, age-specific rate, 1990-1994. |
Summary of VAO and Update 1996
Occupational Studies
For prostate cancer, several studies have shown elevated risk in agricultural or forestry workers. Mortality was increased in studies of United States Department of Agriculture (USDA) agricultural extension agents (PMR = 1.5, CI 1.1-2.0) and forest and soil conservationists (PMR = 1.6, CI 1.1-2.0) (Alavanja et al., 1988, 1989). However, subsequent case-control analysis of these deaths showed no increased risk of prostate cancer related to being an extension agent (OR = 1.0, CI 0.7-1.5) or a soil conservationist (OR = 1.0, CI 0.6-1.8), although the risk was elevated for forest conservationists (OR = 1.6, CI 0.9-3.0). The risk of prostate cancer was more highly elevated for those whose employment ended prior to 1960 and who had worked for at least 15 years as conservationists (OR = 2.1 for forest workers and 2.9 for soil workers). A case-control study of white male Iowans who died of prostate cancer (Burmeister et al., 1983) found a significant association (OR = 1.2) with farming that was not connected to a specific agricultural exposure. Higher RRs were observed after restricting analysis to those born before 1890 (OR = 1.5) or those age 65 and older (OR = 1.3).
A PCMR study was performed with farmers in 23 states, using occupational information from death certificates (Blair et al., 1993). Based on 3,765 deaths from prostate cancer in white male farmers, out of 119,648 studied, the PCMR was significantly increased, at 1.2 (CI 1.1-1.2). Based on 564 deaths from prostate cancer in nonwhite male farmers, out of 11,446 studied, the PCMR was also significantly increased to 1.1 (CI 1.1-1.2). This increased risk for prostate cancer was observed in 22 of the 23 states studied.
In a large cohort study of Canadian farmers, Morrison et al. (1993) found that an increased risk of prostate cancer was associated with herbicide spraying, and the risk was found to rise with increasing number of acres sprayed. For the entire cohort, the RR for prostate cancer and spraying at least 250 acres was 1.2 (CI 1.0-1.5). Adjustment for potential confounders showed no evidence of confounding for the association. Additional analyses were restricted to a one-third sample of farmers most likely to be exposed to phenoxy or other herbicides (RR = 1.3, CI 1.0-1.8 for >250 acres sprayed). To focus on those farmers most likely exposed to herbicides, additional analyses were restricted to those with no employees (RR = 1.4, CI 1.0-1.9 for > 250 acres sprayed); no customary expenses for assisting in work, which might include spraying (RR = 1.6, CI 1.1-2.2 for > 250 acres sprayed); age between 45 and 69 years (RR = 1.7, CI 1.1-2.8 for > 250 acres sprayed); and a combination of the three restrictions (RR = 2.2, CI 1.3-3.8 for > 250 acres sprayed). In each of these comparisons, a statistical test for trend (increasing risk with increasing number of acres sprayed) was significant.
Although a number of occupational and other studies have examined prostate cancer in relation to potential herbicide exposures, many had very few cases.
These include numerous studies of chemical production workers (Bond et al., 1983; Lynge, 1985; Zober et al., 1990; Manz et al., 1991; Buena de Mesquita et al., 1993; Collins et al., 1993; Becher et al., 1996); pesticide appliers (Blair et al., 1983; Swaen et al., 1992; Asp et al., 1994); and paper and pulp workers (Henneberger et al., 1989; Solet et al., 1989). These small numbers may reflect an insufficient latency period or a cohort that is too young to observe excess prostate cancer. They also may reflect the fact that since prostate cancer tends not to be fatal, studies of mortality tend to have low statistical power to detect an effect if it exists, compared to other cancer sites. Other problems in certain studies are low specificity of exposure definition and, in at least one report, a very short duration of exposure (median duration for 2,4-D and 2,4,5-T was six weeks in Asp et al., 1994).
Environmental Studies
Cancer incidence and mortality were described for the Seveso population in three previous studies (Bertazzi et al., 1989a,b; Pesatori et al., 1992). Cancer incidence for the first 10 years after exposure to TCDD was thoroughly updated for the Seveso, Italy, cohort (Bertazzi et al., 1993). In zones A and B (the more highly exposed areas), four cases of prostate cancer were diagnosed, and the RR was 1.4 (CI 0.5-3.9) (Pesatori et al., 1992). In zone R (the less exposed area), based on 16 cases of prostate cancer (Bertazzi et al., 1993), the RR was 0.9 (CI 0.5-1.5).
Vietnam Veteran Studies
Studies of prostate cancer mortality among Vietnam veterans have not consistently shown an association (Anderson et al., 1986a,b; Breslin et al., 1988). A proportionate mortality study examining causes of death among veterans on the state of Michigan's Vietnam era bonus list (Visintainer et al., 1995) compared cause-specific mortality rates for 3,364 Vietnam veterans with rates for 5,229 age-matched veterans who had served elsewhere. There were 19 deaths from male genital cancers among Vietnam veterans (PMR = 1.1, CI 0.6-1.7). Rates for prostate cancer were not reported separately from testicular cancer.
Update of the Scientific Literature
Several new reports have been published, examining prostate cancer mortality, incidence, or both, in populations with possible or documented exposures to TCDD. Six of these involved occupational exposures (Becher et al., 1996; Ott and Zober 1996; Zhong and Rafusson, 1996; Gambini et al., 1997; Hertzman et al., 1997; Kogevinas et al., 1997): one was an occupational cohort with dietary exposures (Svensson et al., 1995), one was a follow-up of the Seveso cohort, and four were examined deaths among Vietnam veterans (AFHS, 1996; Watanabe
and Kang, 1996; Crane et al., 1997a,b). One occupational study grouped all male genital cancers for examining mortality (Hertzman et al., 1997). None of these provides strong evidence of increased risk for prostate cancer.
The smallest study involved 243 workers exposed during and after a 1953 reactor accident in a West German plant that produced trichlorophenol (Ott and Zober, 1996). Exposures were originally documented on the basis of chloracne, which was observed in nearly half the cohort, although it was several years after the accident that TCDD was identified as the cause of the chloracne. Exposures occurred during three periods: (1) clean-up operations and repair activities that lasted several months; (2) incidental maintenance activities after the completion of the restoration; and (3) demolition of the reactor in 1968-1969. Biomonitoring data were collected later and used to estimate internal doses of dioxin at the time of each worker's peak exposures. In this small cohort, there have been no deaths from prostate cancer, with follow-up through 1992. Four cases were identified during 1960-1992, resulting in a SIR of 1.1 (CI 0.3-2.8). When broken down by dose, there is no clear dose-response relationship: the highest SIR (2.5) occurs in the lowest-dose group (<0.1 μg/kg), and there are no cases in the highest-dose group (> 1 μg/kg). On the other hand, the small size of the cohort coupled with the low number of expected cases would indicate that there is not sufficient power to examine the dose-response relation in this follow-up.
In a reanalysis of the study published by Manz et al. (1991), Becher et al. (1996) report on 2,479 workers exposed to phenoxy herbicides and associated contaminants in four German chemical production plants. The SMRs in the four cohorts ranged from zero to 1.5, based on nine prostate cancer deaths. Combining the data from these plants yields an SMR of 1.3. Again, the small number of expected deaths from prostate cancer suggests a fairly young cohort. These four cohorts were also included in the large IARC study of 36 cohorts in 12 countries (Kogevinas et al., 1997). With 21,863 workers, of whom 13,831 were exposed to TCDD or higher chlorinated dioxins, and a total of 488,482 person-years of follow-up, 43 prostate cancer deaths were observed, yielding an SMR of 1.1 (CI 0.8-1.5). This SMR was similar to the one for workers not exposed to TCDD or higher chlorinated dioxins: 1.1 (CI 0.7-1.6). Biomonitoring data were collected on more than 500 workers, but were not used to categorize individual workers. Production workers did not appear to have considerably higher exposures than sprayers. An overall ''healthy worker effect" was not observed in this study.
The fourth study involving occupational exposures was conducted on rice growers in northern Italy (Gambini et al., 1997). The cohort consisted of 1,487 male workers, with 31,648 person-years of observation, over half of which were at ages 60 or older. Individual exposures were not estimated, but the use of herbicides was documented to begin in the early 1950s and to include 2,4-D (used in the 1960s), 2,4,5-T (1960s and 1970s), MCPA (1960s-1990s), and MCPP (mecoprop, 1960s-1980s). The 19 observed prostate cancer deaths resulted in an SMR of 1.0 (CI 0.6-1.5). No analysis was conducted based on years of work.
A study of 2,449 Icelandic pesticide users, of whom 1,860 were men, took advantage of the country's Cancer Registry to analyze morbidity (Zhong and Rafnsson, 1996). A deficit of prostate cancer cases was observed, with an SIR of 0.7 (CI 0.2-1.2) based on 10 cases. Data on actual work tasks were not available for most of the cohort.
A cohort of sawmill workers from British Columbia, Canada, with chlorophenate exposures was evaluated with respect to both morbidity and mortality (Hertzman et al., 1997). An SMR for prostate cancer was not given, but for all male genital cancers combined it was 1.2 (CI 1.0-1.4). (Prostate cancers generally represent about 90 percent of male genital cancers.) The SIR for prostate cancer was 1.0 (CI 0.9-1.1). Although data were collected on sawmill workers not exposed to chlorophenates, the size of this unexposed group was much smaller than for exposed workers, and data were not presented on its prostate cancer incidence or mortality that would have enabled internal comparisons to be made for prostate cancer.
Svensson et al. (1995) analyzed mortality and morbidity from two cohorts of fisherman: those from the east coast of Sweden, who consume a diet rich in fatty fish, and those from the west coast of Sweden, whose diets involve leaner fish, but who are otherwise socioeconomically quite similar. Plasma toxic equivalents of PCBs and PCDD/F have been documented to be twice as high in east coast as in west coast fishermen or referents. The SMR for prostate cancer among east coast fishermen was 1.0 (CI 0.5-1.8), compared to 1.1 (CI 0.9-1.3) for west coast fishermen. The SIRs for these two groups, respectively, were 1.1 (CI 0.8-1.5) and 1.0 (CI 0.9-1.1).
In the Seveso cohort, the numbers of deaths from prostate cancer among men in zones A, B, and R were 0, 6, and 39, for SMRs of 0, 1.2 (CI 0.5-2.7), and 1.2 (CI 0.8-1.6), respectively.
Finally, two new studies on Vietnam veterans reported data on prostate cancer. Watanabe and Kang (1996) published a study of deaths among Vietnam veterans (N = 33,833) and a comparison group of Vietnam era veterans who served outside Southeast Asia (N = 36,797). This was not a cohort study, since there is no register of those who served in the military during the Vietnam period. Most of these veterans had served in the Army, but about 20 percent had served in the Marine Corps. Proportionate mortality comparisons were made between the two groups of deaths. Results for prostate cancer were: PMR = 1.1 for those who served in the Army in Vietnam and PMR = 1.2 for those who served there as Marines; among those who served outside Southeast Asia, PMRs were 1.2 for Army veterans, and 1.3 for Marine veterans. Those who served in Vietnam were generally similar to those who did not, although slightly more African Americans served in Vietnam and the length of service was substantially longer for those who served in Vietnam compared with those who served elsewhere (e.g., 1.9
percent versus 13.8 percent served one year or less in the Army, and 3.8 percent versus 27.1 percent served 1 year or less in the Marines). In an analysis by years of latency, the highest PMR among those who served in the Army was for the longest latency (at least 16 years from the last year in Vietnam): 1.1.
A recent update on the Ranch Hand cohort found an SMR of 4.0 (two observed deaths) for prostate cancer, when considering only the period after 20 years' latency (AFHS, 1996). A brief report of a case-control study of Massachusetts veterans (Clapp, 1997), which used gastrointestinal cancers as controls, found an OR of 0.8, comparing those who had served in Vietnam (15 cases) with those who had served elsewhere in the same era (118 cases). The latter analysis could be problematic since some studies have shown rectal cancer to be increased in association with herbicide exposures.
Crane et al. (1997a) conducted a detailed analysis of Australian male Vietnam veteran mortality. In the early follow-up period from 1964 to 1979, one death from prostate cancer was observed. However, from 1980 to 1994, there were 36 such deaths, with an SMR of 1.5 (CI 1.1-2.1). By service, the figures for Army, Navy, and Air Force personnel were, respectively, 26 deaths, SMR = 1.6 (CI 1.1-2.4); 8 deaths, SMR = 2.2 (CI 0.9-4.3); and 2 deaths, SMR = 0.5 (CI 0.1-1.9). A companion study comparing conscripted Australian veterans of Vietnam with military personnel who did not serve there reported no deaths in the Vietnam veteran population and one in the comparison population between 1982 and 1994 (Crane et al., 1997b). Although smoking data were not available, prostate cancer is not associated with smoking; therefore, there is no concern about smoking being a confounder of these results.
Synthesis
Results of studies published to date are summarized in Table 7-20. At the time of Update 1996, several of the agricultural studies indicated some elevation in risk of prostate cancer, with some subgroups that had the greatest probability and longest duration of herbicide exposure showing the highest risks (Morrison et al., 1993). Additional agricultural studies published since Update 1996, one on rice growers in northern Italy and the other on Icelandic pesticide users, show no evidence of increased risk of prostate cancer, although it should be noted that exposure information was poor and both cohorts demonstrated a rather strong healthy worker effect.
The earlier major studies of production workers (Fingerhut et al., 1991; Manz et al., 1991; Saracci et al., 1991) showed small, but not statistically significant, elevations in risk. Since Update 1996 , several additional studies, some of which overlap earlier ones (Becher et al., 1996; Kogevinas et al., 1997; and Ott and Zober, 1996), continue to show weak evidence of effects on prostate cancer mortality, with RRs generally <1.5 and mostly in the range of 1.1-1.2.
Two cohort studies reported both morbidity and mortality (Svensson et al.,
1995; Hertzman et al., 1997). Among Swedish fisherman, no association was seen between exposure and prostate cancer mortality (SMR = 1.0) or morbidity (SIR = 1.1); among sawmill workers, no association was seen with morbidity (SIR = 1.0), whereas the association with mortality from male genital tract cancers was significant (SMR = 1.2, CI 1.0-1.4).
In general, when observed associations are this weak, biases could have induced artifactual associations. In this case, it is unclear what those biases may be in such varied populations involving quite different exposure scenarios. In studies using mortality as an outcome, detection bias is unlikely to explain any elevated risks. On the other hand, factors that increase incidence might not be the same as those related to subsequent mortality among those who have the disease. The recent introduction and widespread adoption of PSA (prostate specific antigen) for screening purposes has brought about an apparent increase in incidence in the United States; the pattern of long-term impact on incidence or mortality is difficult to predict for any country or population and will depend on the rapidity with which this screening tool is adopted, its differential across different ages of men, and the aggressiveness of tumors detected early by this test (Gann, 1997). Differences among countries in the rate of use of PSA could produce more variable, less consistent results in studies of exogenous exposures and prostate cancer.
Several new studies have been published on Vietnam veterans. Too few cases were seen among Ranch Hands to provide useful information about this outcome (AFHS, 1996). In a larger study of deaths among U.S. Vietnam veterans, those who had last served in Vietnam more than 16 years earlier had a weakly elevated PMR of 1.1. PMR studies should be viewed cautiously, however, since they include only deaths; in this type of study, findings for any one outcome are strongly influenced by associations between the exposure of interest and other causes of death, particularly the most common ones. In the cohort study of Australian male veterans, which had a much stronger design, an elevated rate of prostate cancer mortality was observed among those who served in Vietnam (SMR = 1.5, CI 1.1-2.1, based on 36 prostate cancer deaths). Unlike results for respiratory cancer, the lack of smoking data for this cohort does not pose a threat to the validity of these findings, since prostate cancer is not associated with smoking. It should also be kept in mind that most Vietnam veterans have not yet reached the age when this cancer tends to appear, and that morbidity may represent a more sensitive outcome than mortality for this cancer site.
Conclusions
Strength of Evidence in Epidemiologic Studies
There is limited/suggestive evidence of an association between exposure to the herbicides (2,4-D, 2,4,5-T and its contaminant TCDD, cacodylic acid, and
TABLE 7-20 Selected Epidemiologic Studies—Prostate Cancer
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
OCCUPATIONAL |
|||
New Studies |
|||
Gambini et al., 1997 |
Italian rice growers |
19 |
1.0 (0.6-1.5) |
Hertzman et al., 1997 |
Canadian sawmill workers |
|
|
|
Mortality |
282 |
1.0 (0.9-1.1) |
|
Morbidity for male genital tract cancers |
116 |
1.2 (1.0-1.4) |
Kogevinas et al., 1997 |
IARC cohort |
43 |
1.1 (0.8-1.5) |
Becher et al., 1996 |
German chemical production workers |
9 |
1.3 |
Ott and Zober, 1996 |
BASF cleanup workers |
4 |
1.1 (0.3-2.8) |
Zhong and Rafusson, 1996 |
Icelandic pesticide users |
10 |
0.7 (0.3-1.2) |
Studies reviewed in Update 1996 |
|||
Asp et al., 1994 |
Finnish herbicide applicators |
5 |
0.8 (0.3-1.8) |
Blair et al., 1993 |
U.S. farmers in 23 states |
|
|
|
Whites |
3,765 |
1.2 (1.1-1.2) |
|
Nonwhites |
564 |
1.1 (1.1-1.2) |
Bueno de Mesquita et al., 1993 |
Dutch production workers |
3 |
2.6 (0.5-7.7) |
Collins et al., 1993 |
Monsanto 2,4-D production workers |
9 |
1.6 (0.7-3.0) |
Studies reviewed in VAO |
|||
Morrison et al., 1993 |
Canadian farmers, age 45-69 years, no employees, or custom workers, sprayed > 250 acres |
20 |
2.2 (1.3-3.8) |
Ronco et al., 1992 |
Danish self-employed farm workers |
399 |
0.9 (p < .05) |
Swaen et al., 1992 |
Dutch herbicide applicators |
1 |
1.3 (0.0-7.3) |
Fingerhut et al., 1991 |
NIOSH cohort |
17 |
1.2 (0.7-2.0) |
|
20 year latency, 1 year exposure |
9 |
1.5 (0.7-2.9) |
Manz et al., 1991 |
German production workers |
7 |
1.4 (0.6-2.9) |
Saracci et al., 1991 |
IARC cohort |
30 |
1.1 (0.8-1.6) |
Zober et al., 1990 |
BASF production workers |
0 |
— (0.0-7.5) |
Alavanja et al., 1989 |
USDA forest conservationists |
|
1.6 (0.9-3.0) |
|
Soil conservationists |
|
1.0 (0.6-1.8) |
Henneberger et al., 1989 |
Paper and pulp workers |
9 |
1.0 (0.7-2.0) |
Solet et al., 1989 |
Paper and pulp workers |
4 |
1.1 (0.3-2.9) |
Alavanja et al., 1988 |
USDA agricultural extension agents |
|
1.0 (0.7-1.5) |
Bond et al., 1988 |
Dow 2,4-D production workers |
1 |
1.0 (0.0-5.8) |
Coggon et al., 1986 |
British MCPA production workers |
18 |
1.3 (0.8-2.1) |
Robinson et al., 1986 |
Paper and pulp workers |
17 |
1.2 (0.7-2.0) |
Lynge, 1985 |
Danish production workers |
9 |
0.8 |
Blair et al., 1983 |
Florida pesticide applicators |
2 |
0.5 |
Burmeister et al., 1983 |
Iowa residents |
|
1.2 (p < .05) |
Wiklund, 1983 |
Swedish agricultural workers |
3,890 |
1.0 (0.9-1.0)b |
Burmeister, 1981 |
Iowa farmers |
1,13 |
1.1 (p < .01) |
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
ENVIRONMENTAL |
|||
New Studies |
|||
Bertazzi et al., 1997 |
Seveso residents |
|
|
|
zone A |
0 |
0 (0.0-5.2) |
|
zone B |
6 |
1.2 (0.5-2.7) |
|
zone R |
39 |
1.2 (0.8-1.6) |
Svensson et al., 1995 |
Swedish fishermen mortality |
12 |
1.0 (0.5-1.8) |
|
Swedish fishermen incidence |
38 |
1.1 (0.8-1.5) |
Studies reviewed in Update 1996 |
|||
Bertazzi et al., 1993 |
Seveso male residents—zone R |
16 |
0.9 (0.5-1.5) |
Studies reviewed in VAO |
|||
Pesatori et al., 1992 |
Seveso male residents—zones A and B |
4 |
1.4 (0.5-3.9) |
Bertazzi et al., 1989a |
Seveso male residents—zones A, B, and R |
19 |
1.6 (1.0-2.7) |
Bertazzi et al., 1989b |
Seveso male residents—zone B |
3 |
2.2 (0.7-6.9) |
VIETNAM VETERANS |
|||
New Studies |
|||
Clapp, 1997 |
Massachusetts Vietnam veterans |
|
|
|
Exposed cancers |
15 |
0.8 (0.4-1.6) |
|
Total cancer |
133 |
|
Crane et al., 1997a |
Australian military veterans |
36 |
1.5 (1.1-2.1) |
|
Army |
26 |
1.6 (1.1-2.4) |
|
Navy |
8 |
2.2 (0.9-4.3) |
|
Air Force |
2 |
0.5 (0.1-1.9) |
AFHS, 1996 |
Ranch Hands |
2 |
4.0 |
Watanabe and Kang, 1996 |
Army Vietnam veterans |
58 |
0.9 |
|
16+ years after discharge |
|
1.1 |
Studies reviewed in Update 1996 |
|||
Visintainer et al., 1995 |
Michigan Vietnam veterans |
19 |
1.1 (0.6-1.7) |
Studies reviewed in VAO |
|||
Breslin et al., 1988 |
Army Vietnam veterans |
30 |
0.9 (0.6-1.2) |
|
Marine Vietnam veterans |
5 |
1.3 (0.2-10.3) |
Anderson et al., 1986b |
Wisconsin Vietnam veterans |
2 |
— |
a Given when available. b 99 Percent CI. |
picloram) and prostate cancer. Although the associations are not large, a number of studies provide evidence that is suggestive of a slight increase in either morbidity or mortality from prostate cancer. The evidence regarding association is drawn from occupational studies in which subjects were exposed to a variety of herbicides and herbicide components and is also based on data from studies of Vietnam veterans. An important consideration is the fact that prostate cancer tends not to be fatal; thus, mortality studies have lower statistical power to detect a comparable effect than a similar-sized morbidity study would have.
Biologic Plausibility
A thorough discussion of biologic plausibility with respect to exposure to TCDD or herbicides and prostate cancer is contained in Chapter 3; a summary is presented in the conclusion to this chapter.
Increased Risk of Disease Among Vietnam Veterans
Studies that have been conducted on Vietnam veterans have had a low likelihood of detecting an increased risk of prostate cancer, if service in Vietnam is associated with this cancer, because of the weak study designs and the relatively young age of Vietnam veterans. The statistically significantly elevated prostate cancer SMR for Australian male Vietnam veterans suggests that U.S. Vietnam veterans may be at increased risk. Further follow-up that includes, in particular, studies of morbidity among living veterans, would help to define the risk. A more thorough discussion of the issue of increased risk of disease in Vietnam veterans is included in Chapter 1.
TESTICULAR CANCER
Background
The American Cancer Society estimates that 7,600 men will be diagnosed with testicular cancer (ICD·9 186.0-186.9) in the United States in 1998 and that 400 will die from the disease (ACS, 1998).
Testicular cancer is far more likely in men younger than 40 than in those who are older. On a lifetime basis, the risk for white men is about four times greater than for African Americans. Cryptorchidism, or undescended testicles, is a major risk factor for testicular cancer. Family history of the disease also appears to play a role. Several other hereditary and environmental factors have been suggested, but research regarding them is inconsistent (Bosl and Motzer, 1997).
Average Annual Cancer Incidence (per 100,000 individuals) in the United Statesa Testicular Cancer |
|||||||||
|
45-49 years of age |
|
50-54 years of age |
|
55-59 years of age |
|
|||
|
all races |
white |
black |
all races |
white |
black |
all races |
white |
black |
|
5.7 |
6.4 |
0.9 |
3.3 |
3.8 |
0.4 |
2.2 |
2.6 |
0.5 |
a SEER nine standard registries crude, age-specific rate, 1990-1994. |
Summary of VAO and Update 1996
A case-control study of 137 testicular cancer cases and 130 hospital controls (Tarone et al., 1991) found an OR of 2.3 (CI 1.0-5.5) for service in Vietnam. Risk for testicular cancer was not significantly elevated by service branch. Another
case-control study among veterans investigated the association between potential Agent Orange exposure and the risk of testicular cancer (Bullman et al., 1994), using subjects chosen from the DVAs Agent Orange Registry. This included 97 veterans with testicular cancer and 311 veterans with no clinical diagnosis recorded on the registry. Risk of testicular cancer was not significantly increased for ground troops, combat duty, service in the III Corps area (a heavily sprayed area), or proximity to other areas where Agent Orange was sprayed. Only Navy veterans had a statistically significant increased risk of testicular cancer (OR = 2.6, CI 1.1-6.2), based on 12 cases among 27 veterans. One of these 27 served in the "brown-water" Navy and may have had Agent Orange exposure due to spraying of riverbanks. In general, other veteran studies and most of the occupational and environmental studies showed no association between exposure and outcome, but the sample sizes of some of these studies were generally small. Results of other studies of testicular cancer have been equivocal. These include studies of chemical production workers in the United States and other countries (Bond et al., 1988; Saracci et al., 1991); agricultural workers (Wiklund, 1983; Ronco et al., 1992; Blair et al., 1993); residents of Seveso (Pesatori et al., 1992; Bertazzi et al., 1993); and Vietnam veterans (Anderson et al., 1986a,b; Boyle et al., 1987; Breslin et al., 1988; Watanabe et al., 1991).
Update of the Scientific Literature
Occupational Studies
The IARC study (Kogevinas et al., 1997) found seven deaths due to testicular cancer, SMR = 1.3 (CI 0.5-2.7). Ramlow et al. (1996) found no cases of death due to testicular cancer, with about 0.2 expected. Hertzman and colleagues' (1997) study of Canadian sawmill workers reported an SIR of 1.0 (CI 0.6-1.4) based on 18 cases. "Male genital cancers" (possibly including prostate) had an SMR of 1.0 (CI 0.8-1.1) with 116 cases.
Environmental Studies
The Seveso follow-up (Bertazzi et al., 1997) did not report testicular cancer separately, but the SMR for genitourinary cancers was less than 1.0. A study of Icelandic pesticide appliers (Zhong and Rafnsson, 1996) reported two cases, with an SIR of 1.2 (CI 0.1-4.3).
Vietnam Veterans' Studies
In studies of Vietnam veterans, Dalager and Kang (1997) found two deaths due to testicular cancer among Army Chemical Corps veterans (SMR = 4.0, CI 0.5-14.5). Watanabe and Kang (1996) reported PMRs of 1.1 and 1.0 for Army
TABLE 7-21 Selected Epidemiologic Studies—Testicular Cancer
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
OCCUPATIONAL |
|||
New Studies |
|||
Hertzman et al., 1997 |
British Columbia sawmill workers |
|
|
|
Mortality |
116 |
1.0 (0.8-1.1) |
|
Incidence |
18 |
1.0 (0.6-1.4) |
Kogevinas et al., 1997 |
IARC cohort |
7 |
1.3 (0.5-2.7) |
Ramlow et al., 1995 |
Pentachlorophenol production workers |
0 |
— |
Studies reviewed in Update 1996 |
|||
Blair et al., 1993 |
U.S. farmers in 23 states |
|
|
|
White males |
32 |
0.8 (0.6-1.2) |
|
Nonwhite males |
6 |
1.3 (0.5-2.9) |
Studies reviewed in VAO |
|||
Ronco et al., 1992 |
Danish self-employed farm workers |
74 |
0.9 |
Saracci et al., 1991 |
IARC cohort |
7 |
2.3 (0.9-4.6) |
Bond et al., 1988 |
Dow 2,4-D production workers |
1 |
4.6 (0.0-25.7) |
Coggon et al., 1986 |
British MCPA production workers |
4 |
2.2 (0.6-5.7) |
Wiklund, 1983 |
Swedish agricultural workers |
101 |
1.0 (0.7-1.2)b |
ENVIRONMENTAL |
|||
New Studies |
|||
Zhong and Rafnsson, 1996 |
Icelandic pesticide users |
2 |
1.2 (0.1-4.3) |
Studies reviewed in Update 1996 |
|||
Bertazzi et al., 1993 |
Seveso residents |
|
|
|
zone A |
0 |
— |
|
zone B |
1 |
— |
|
zone R |
9 |
1.4 (0.7-3.0) |
Studies reviewed in VAO |
|||
Pesatori et al., 1992 |
Seveso residents—zones A and B |
1 |
0.9 (0.1-6.7) |
|
Residents—zone R |
9 |
1.5 (0.7-3.0) |
VIETNAM VETERANS |
|||
New Studies |
|||
Clapp, 1997 |
Massachusetts Vietnam veterans—incidence |
30 |
1.2 (0.4-3.3) |
Crane et al., 1997a |
Australian military veterans |
4 |
NS |
Crane et al., 1997b |
Australian national service veterans |
4 |
1.3 |
Dalager and Kang, 1997 |
Army Chemical Corps veterans |
2 |
4.0 (0.5-14.5) |
Watanabe and Kang, 1996 |
Vietnam service Army |
|
1.1 |
|
Vietnam service Marines |
|
1.0 |
Studies reviewed in Update 1996 |
|||
Bullman et al., 1994 |
Navy veterans |
12 |
2.6 (1.1-6.2) |
Studies reviewed in VAO |
|||
Tarone et al., 1991 |
Patients at three Washington, D.C., area hospitals |
|
2.3 (1.0-5.5) |
Watanabe et al., 1991 |
Army Vietnam veterans |
109 |
1.2 |
|
Marine Vietnam veterans |
28 |
0.8 |
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
Breslin et al., 1988 |
Army Vietnam veterans |
90 |
1.1 (0.8-1.5) |
|
Marine Vietnam veterans |
26 |
1.3 (0.5-3.6) |
Anderson et al., 1986b |
Wisconsin Vietnam veterans |
9 |
1.0 (0.5-1.9) |
Anderson et al., 1986a |
Wisconsin Vietnam veterans |
11 |
1.0 (0.5-1.7) |
a Given when available. b 99% CI. c "Male genital cancers" |
and Marine Vietnam veterans, respectively, based on 142 total deaths. Crane and colleagues report 4 deaths due to testicular cancer among Australian military veterans, with an SMR near unity (1997a). A study comparing conscripted Australian veterans of Vietnam with military personnel who did not serve there reports a RR of 1.3 based on a single death in each of the populations (Crane et al., 1997b). Clapp (1997) reported 30 incident cases in Massachusetts Vietnam veterans, OR = 1.2 (CI 0.4-3.3), compared to veterans who had not served in Vietnam.
Several studies of military working dogs have been reported, which showed abnormal testicular pathology and a moderate excess of seminomas in dogs that died that had worked in Vietnam (Mahaney 1990; Hayes et al., 1990, 1994, 1995a,b). Most ORs were around 2, and the design was that of a proportionate mortality study (i.e., all risks were based on the proportion of dogs that had died from any cause with a given testicular finding at necropsy). There were no measures of exposure to environmental agents.
Synthesis
There is minimal new information on this rare cancer, and what there is provides little evidence supporting a herbicide-testicular cancer connection. The findings in dogs are not felt to carry great weight in the absence of exposure data and without observed excesses in human populations.
Conclusions
Strength of Evidence in Epidemiologic Studies
There is inadequate or insufficient evidence to determine whether an association exists between the herbicides (2,4-D, 2,4,5-T and its contaminant TCDD, cacodylic acid, and picloram) and testicular cancer.
Biologic Plausibility
A thorough discussion of biologic plausibility with respect to exposure to TCDD or herbicides and testicular cancer is contained in Chapter 3; a summary is presented in the conclusion to this chapter.
URINARY BLADDER CANCER
Background
Urinary bladder cancer (ICD-9 188.0-188.9) is the most common of the genitourinary tract cancers. According to American Cancer Society estimates, 39,500 men and 14,900 women will be diagnosed with this cancer in the United States in 1998 and 8,400 men, and 4,100 women will die from the disease (ACS, 1998). In males, where this cancer is about three times more likely to occur than in females, these numbers represent approximately 6 percent of new cancer diagnoses and 3 percent of deaths. Overall, bladder cancer is the fifth most common cancer and the fifth leading cause of cancer death in the United States
Among males in the age groups that characterize most Vietnam veterans, bladder cancer incidence is about twice as high in whites as in African Americans. Rates are slightly higher in white than African-American women. Bladder cancer incidence increases greatly with age for individuals older than 40. For the age groups shown below, the incidence rate in each five-year grouping is roughly double that of the age group before it.
The most important known risk factor for bladder cancer is smoking. About one-half of bladder cancers in men and one-third in women are thought to be due to smoking (Miller et al., 1996). Occupational exposure to aromatic amines (also called arylamines) is also associated with higher incidence. High-fat diets have been implicated as risk factors, along with exposure to the parasite Schistosoma haematobium.
Average Annual Cancer Incidence (per 100,000 individuals) in the United Statesa Urinary Bladder Cancer |
|||||||||
|
45-49 years of age |
50-54 years of age |
55-59 years of age |
||||||
|
all races |
white |
black |
all races |
white |
black |
all races |
white |
black |
males |
14 |
16 |
7 |
29 |
31 |
16 |
52 |
56 |
35 |
females |
4 |
5 |
3 |
9 |
9 |
7 |
17 |
18 |
10 |
a SEER nine standard registries crude, age-specific rate, 1990-1994. |
Summary of VAO and Update 1996
For bladder cancer, Fingerhut et al. (1991) found a small excess mortality in their study of chemical production workers exposed to TCDD. In the total cohort of 5,172 workers, the SMR was 1.6 (CI 0.7-3.0), based on nine cases. In workers
with at least one year of employment and 20 years' latency, there were four cases (SMR = 1.9, CI 0.5-4.8). Other studies of bladder cancer have produced inconclusive results. Occupational studies include chemical production workers in the United States and other countries (Moses et al., 1984; Suskind and Hertzberg, 1984; Bond et al., 1988; Zober et al., 1990; Saracci et al., 1991); agricultural and forestry workers (Burmeister, 1981; Alavanja et al., 1988, 1989; Green, 1991; Ronco et al., 1992; Blair et al., 1993); pesticide appliers (Blair, 1983); and paper and pulp workers (Robinson et al., 1986; Henneberger et al., 1989). Environmental studies of bladder cancer and herbicide or TCDD exposure include the Pesatori et al. (1992) and Bertazzi et al. (1989b, 1993) studies of Seveso residents and the Lampi et al. (1992) study of a Finnish community exposed to chlorophenols. Studies in Vietnam veterans examining bladder cancer include the Breslin et al. (1988) study of Army and Marine Corps Vietnam veterans and a study of veterans in Wisconsin (Anderson et al., 1986a,b). The results of these studies are summarized in Table 7-22.
As a subgroup of the IARC cohort, a cohort of workers was identified who manufactured chlorophenoxy herbicides in two factories in the Netherlands (Bueno de Mesquita et al., 1993). Among 963 exposed male workers, there was one case of bladder cancer (SMR = 1.2, CI 0.0-6.7).
The mortality experience of 754 male production workers at a Monsanto plant was evaluated (Collins et al., 1993). One hundred and twenty-two of these workers developed chloracne as a result of an accidental release of TCDD in 1949. Based on 16 deaths due to bladder cancer, the SMR was 6.8 (CI 3.9-11.1). Eleven of these cases had documented exposure to 4-aminobiphenyl, a known bladder carcinogen; therefore, TCDD exposure was not the primary suspected risk factor. The SMR was not significantly elevated for the other five cases not exposed to 4-aminobiphenyl.
An 18-year follow-up study of cancer incidence and mortality in 1,909 Finnish herbicide appliers was reported (Asp et al., 1994). Employees had previously been identified as being exposed to 2,4-D and 2,4,5-T (Riihimaki et al., 1982). The median total exposure to herbicides was six weeks. Based on 12 cases of bladder cancer, the SIR was 1.6 (CI 0.8-2.8).
Update of Scientific Literature
Occupational Studies
The IARC study (Kogevinas et al., 1997) found an SMR of 1.0 (CI 0.7-1.5) for all workers, with 34 deaths from bladder cancer, and an SMR of 1.4 (CI 0.9-2.1) among workers exposed to TCDD or higher chlorinated dioxins, based on 24 deaths. Hertzman et al.'s (1997) study of Canadian sawmill workers reported an SIR of 1.0 (CI 0.8-1.2) based on 94 cases, and an SMR of 0.9 (CI 0.7-1.2) based on 33 cases.
Environmental Studies
In a study of rice growers, Gambini et al. (1997) found an SMR of 1.0 (CI 0.5-1.8) based on 12 deaths. A follow-up of BASF employees exposed to TCDD as a result of a 1953 chemical reactor accident (Ott and Zober, 1996) found two deaths due to ''bladder or kidney" cancer, from a total of five cases (SIR = 1.4, CI 0.4-3.2). In the Bertazzi et al. (1997) follow-up of Seveso residents, 29 deaths were observed, with SMRs near 1.0 among both women and men in all exposure zones. A study of Swedish fishermen (Svensson et al., 1995) showed SIRs of 0.7 (CI 0.4-1.3) in a presumed elevated-exposure cohort and 0.9 (CI 0.7-1.1) in a comparison cohort, based on 65 cases. SMRs for the two cohorts were 1.3 (CI 0.4-3.1) and 1.0 (CI 0.6-1.6), respectively, based on 25 total deaths.
Vietnam Veterans' Studies
Clapp's (1997) update of his Massachusetts veteran cohort found an OR of 0.6 (CI 0.2-1.3) based on 80 cases. None of the other studies of U.S. Vietnam veterans (Dalager and Kang, 1997; Watanabe and Kang, 1996) published since the release of Update 1996 reported bladder cancer outcomes. Crane and colleagues' (1997a) comparison of all Australian military veterans with the rest of the male Australian population reported 11 deaths due to bladder cancer: SMR = 1.1 (CI 0.6-2.0). A second study of the mortality experience of conscripted Australian veterans relative to military personnel who did not serve in the conflict reported a statistically significant RR of 0.6 based on one death among Vietnam veterans and two in the comparison population (Crane et al., 1997b).
Synthesis
Although there is no evidence that exposure to herbicides alone is related to bladder cancer, RRs in some of the largest cohorts tended to be greater than 1, weakening the committee's prior conclusion that there was positive evidence of no relationship. Coexposure to TCDD and a variety of known bladder carcinogens makes it very difficult to isolate any possible additional effect of herbicides, although little total effect was seen.
Conclusions
Strength of Evidence in Epidemiologic Studies
Based on an evaluation of all of the epidemiologic evidence, the committee felt that the previous conclusion of "limited/suggestive evidence of no association" between exposure to the herbicides (2,4-D, 2,4,5-T and its contaminant TCDD, cacodylic acid, and picloram) and urinary bladder cancer should be
TABLE 7-22 Selected Epidemiologic Studies—Urinary Bladder Cancer
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
OCCUPATIONAL |
|||
New Studies |
|||
Hertzman et al., 1997 |
British Columbia sawmill workers |
|
|
|
Mortality |
33 |
0.9 (0.7-1.2) |
|
Incidence |
94 |
1.0 (0.8-1.2) |
Kogevinas et al., 1997 |
IARC cohort |
|
|
|
Workers exposed to TCDD (or higher chlorinated dioxins) |
24 |
1.4 (0.9-2.1) |
|
Workers exposed to any phenoxy herbicide or chlorophenol |
34 |
1.0 (0.7-1.5) |
Studies reviewed in Update 1996 |
|||
Asp et al., 1994 |
Finnish herbicide applicators—incidence |
12 |
1.6 (0.8-2.8) |
Bueno de Mesquita et al., 1993 |
Dutch production workers |
1 |
1.2 (0.0-6.7) |
Collins et al., 1993 |
Monsanto 2,4-D production workers |
16 |
6.8 (3.9-11.1) |
Studies reviewed in VAO |
|||
Ronco et al., 1992 |
Danish male self-employed farmers |
300 |
0.6 (p < .05) |
Fingerhut et al., 1991 |
NIOSH cohort |
9 |
1.6 (0.7-3.0) |
|
20 year latency |
4 |
1.9 (0.5-4.8) |
Green, 1991 |
Herbicide sprayers in Ontario |
1 |
1.0 (0.01-5.6) |
Saracci et al., 1991 |
IARC cohort |
13 |
0.8 (0.2-1.4) |
Zober et al., 1990 |
BASF production workers |
0 |
— (0.0-15.0) |
Alavanja et al., 1989 |
USDA forest/soil conservationists |
8 |
0.8 (0.3-1.6) |
Henneberger et al., 1989 |
Mortality among paper and pulp workers |
4 |
1.2 (0.3-3.2) |
Alavanja et al., 1988 |
USDA agricultural extension agents |
8 |
0.7 (0.4-1.4) |
Bond et al., 1988 |
Dow 2,4-D production workers |
0 |
— (0-7.2) |
Coggon et al., 1986 |
British MCPA production workers |
8 |
0.9 (0.4-1.7) |
Robinson et al., 1986 |
Paper and pulp workers |
8 |
1.2 (0.6-2.6) |
Lynge, 1985 |
Danish male production workers |
11 |
0.8 |
Blair, 1983 |
Florida pesticide applicators |
3 |
1.6 |
Burmeister, 1981 |
Farmers in Iowa |
274 |
0.9 (NS) |
ENVIRONMENTAL |
|||
New Studies |
|||
Gambini et al., 1997 |
Italian rice growers |
12 |
1.0 (0.5-1.8) |
Ott and Zober, 1996 |
BASF cleanup workers |
2 |
1.4 (0.4-3.2) |
Svensson et al., 1995 |
Swedish fishermen mortality |
|
|
|
East coast |
|
1.3 (0.4-3.1) |
|
West coast |
|
1.0 (0.6-1.6) |
|
Swedish fishermen incidence |
|
|
|
East coast |
|
0.7 (0.4-1.3) |
|
West coast |
|
0.9 (0.7-1.1) |
Studies reviewed in VAO |
|||
Pesatori et al., 1992 |
Seveso male residents—zones A and B |
10 |
1.6 (0.9-3.1) |
|
Female residents—zones A and B |
1 |
0.9 (0.1-6.8) |
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
Lampi et al., 1992 |
Finnish community exposed to chlorophenols |
|
1.0 (0.6-1.9) |
VIETNAM VETERANS |
|||
New Studies |
|||
Clapp, 1997 |
Massachusetts Vietnam veterans |
80 |
0.6 (0.2-1.3) |
Crane et al., 1997a |
Australian military veterans |
11 |
1.1 (0.6-2.0) |
Crane et al., 1997b |
Australian national service veterans |
1 |
0.6 |
Studies reviewed in VAO |
|||
Breslin et al., 1988 |
Army Vietnam veterans |
9 |
0.6 (0.3-1.2) |
|
Marine Vietnam veterans |
4 |
2.4 (0.1-66.4) |
Anderson et al., 1986a |
Wisconsin Vietnam veterans |
0 |
— |
Anderson et al., 1986b |
Wisconsin Vietnam veterans |
1 |
— |
a Given when available. b Many of the employees studied were also exposed to 4-aminobiphenyl, a known bladder carcinogen. |
changed to "inadequate/insufficient evidence to determine whether an association exists."
Biologic Plausibility
A thorough discussion of biologic plausibility with respect to exposure to TCDD or herbicides and bladder cancer is contained in Chapter 3; a summary is presented in the conclusion to this chapter.
RENAL CANCER
Background
Cancers of the kidney (ICD·9 189.0) and renal pelvis (ICD·9 189.1) are often grouped together in epidemiologic studies, although the diseases have different characteristics and may have different risk factors. The American Cancer Society estimates that 17,600 men and 12,300 women will be diagnosed with renal cancers (ICD·9 189.0, 189.1) in the United States in 1998 and that 7,100 men and 4,500 women will die from the disease (ACS, 1998). These figures represent a little more than 2 percent of all new cancer diagnoses and deaths.
Renal cancer is twice is common in men as in women. In the age groups that represent most Vietnam veterans, African-American men have a slightly higher
incidence than white men; African-American and white women have roughly the same rate of the disease. With the exception of Wilm's tumor (which is more likely to occur in children), renal cancer is more common in individuals older than 50 years of age.
Smoking is a well-established risk factor for renal cancer. Phenacetin-containing analgesic abuse has also been implicated. Individuals with certain rare syndromes—notably, von Hippel-Lindau syndrome and tuberous sclerosis—are at higher risk. Other potential factors include diet, weight, and occupational exposure to asbestos and cadmium.
Average Annual Cancer Incidence (per 100,000 individuals) in the United Statesa Kidney and Renal Pelvis Cancer |
|||||||||
|
45-49 years of age |
50-54 years of age |
55-59 years of age |
||||||
|
all races |
white |
black |
all races |
white |
black |
all races |
white |
black |
males |
12 |
12 |
22 |
22 |
22 |
34 |
31 |
31 |
37 |
females |
6 |
6 |
7 |
11 |
11 |
10 |
15 |
16 |
18 |
a SEER nine standard registries crude, age-specific rate, 1990-1994. |
Summary of VAO and Update 1996
Studies of renal cancer have generally produced inconclusive results—in some cases because of small sample sizes, in other cases with SMRs or PMRs near unity. These include studies of chemical production workers in the United States and other countries (Lynge, 1985; Coggon et al., 1986; Bond et al., 1988; Fingerhut et al., 1991; Manz et al., 1991; Saracci et al., 1991; Bueno de Mesquita et al., 1993; Asp et al., 1994); agricultural workers (Burmeister, 1981; Wiklund, 1983; Ronco et al., 1992; Blair et al., 1993); pesticide applicators (Blair, 1983); paper and pulp workers (Robinson et al., 1986; Henneberger et al., 1989); the Seveso population (Pesatori et al., 1992; Bertazzi et al., 1993); and Vietnam veterans (Anderson et al., 1986a,b; Breslin et al., 1988; Kogan and Clapp, 1985, 1988; Clapp et al., 1991). Alavanja et al. (1988, 1989) found excess mortality due to renal cancer in studies of USDA agricultural extension agents (PMR = 2.0, CI 1.2-3.3) and forest and soil conservationists (PMR = 2.1, CI 1.2-3.3). In subsequent case-control studies of these deaths, comparing ever versus never being an extension agent resulted in a RR of 1.7 (CI 0.9-3.3). The RRs for being a soil conservationist and a forest conservationist were 2.4 (CI 1.0-5.9) and 1.7 (CI 0.5-5.5), respectively.
A case-control study of occupational risk factors and renal cell carcinoma included 365 cases identified from the Denmark Cancer Registry and pathology records and 396 controls selected from the country's Central Population Registry (Mellemgaard et al., 1994). Based on 13 cases, the OR for men for herbicide exposure was 1.7 (CI 0.7-4.3). Based on three cases, the OR for herbicide exposure for women was 5.7 (CI 0.6-5.8).
A PMR study examining causes of death among veterans on the state of Michigan's Vietnam era bonus list (Visintainer et al., 1995) compared 3,364 Vietnam veterans with 5,229 age-matched veterans who served elsewhere. Based on 21 cases of renal cancer among Vietnam veterans, the PMR was 1.4 (CI 0.9-2.2).
Update of the Scientific Literature
Occupational Studies
As for other cancer types, the most important new study is the IARC combined cohorts (Kogevinas et al., 1997). This study found 26 cases of kidney cancer and an SMR of 1.6 (CI 1.1-2.4), for workers exposed to TCDD, although there was no trend by duration of exposure or time since exposure, and an SMR of 1.1 (CI 0.7-1.6) for all workers exposed to any phenoxy herbicide or chlorophenol. In a study of 770 pentachlorphenol workers (Ramlow, 1996), three deaths from kidney cancer were found among workers exposed to TCDD, yielding SMRs with broad CIs and high correlation between exposure to PCP and TCDD. Hertzman's study of Canadian sawmill workers reported an SMR of 1.1 (CI 0.9-1.5) and an SIR of 0.9 (CI 0.7-1.1).
Environmental Studies
Renal cancer totals were not reported in Gambini's study of rice growers (Gambini et al., 1997). In Bertazzi's continued Seveso follow-up (Bertazzi et al., 1997), genitourinary tract totals were reported for men, that apparently included bladder cancer, which was also reported separately. SMRs were equal to or less than unity in all zones where numbers were sufficient for analysis. In a study of employees exposed after a reactor accident (Ott and Zober, 1996), five cases of bladder and kidney cancer were observed, with an SIR of 1.4 (CI 0.4-3.2). In a study of Swedish fishermen with unknown exposure to TCDD (Svensson et al., 1995), neither SMRs nor SIRs exceeded unity, with 34 cases observed and 16 deaths. A study of Icelandic pesticide users reported an SIR of 0.52 for cancer of other urinary organs (besides testis and prostate) based on three cases (Zhong and Rafnsson, 1996).
Veteran's Studies
No genitourinary cancer results are reported in either study of Vietnam veterans by Watanabe and Kang (1995, 1996). Clapp's (1997) update of Massachusetts' veterans reported an OR of 1.0 (CI 0.4-2.3) for kidney cancer incidence, based on 55 cases. Crane and colleagues, comparing the mortality of all Australian military veterans of Vietnam with the rest of the male Australian population, reported an SMR of 1.2 (CI 0.8-1.9) based on 22 deaths during 1982-1994 (Crane et al.,
TABLE 7-23 Selected Epidemiologic Studies—Renal Cancer
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
OCCUPATIONAL |
|||
New Studies |
|||
Kogevinas et al., 1997 |
IARC cohort |
|
|
|
Workers exposed to TCDD (or higher chlorinated dioxins) |
26 |
1.6 (1.1-2.4) |
|
Workers exposed to any phenoxy herbicide or chlorophenol |
|
1.1 (0.7-1.6) |
Studies reviewed in Update 1996 |
|||
Mellemgaard et al., 1994 |
Danish Cancer Registry patients |
|
|
|
Occupational herbicide exposure among males |
13 |
1.7 (0.7-4.3) |
|
Occupational herbicide exposure among females |
3 |
5.7 (0.6-5.8) |
Blair et al., 1993 |
U.S. farmers in 23 states |
|
|
|
White males |
522 |
1.1 (1.0-1.2) |
|
Nonwhite males |
30 |
|
|
White females |
6 |
|
|
Nonwhite females |
6 |
|
Studies reviewed in VAO |
|||
Ronco et al., 1992 |
Danish male self-employed farm workers |
141 |
0.6 (p < .05) |
Fingerhut et al., 1991 |
NIOSH cohort |
8 |
1.4 (0.6-2.8) |
Manz et al., 1991 |
German production workers |
3 |
1.6 (0.3-4.6) |
Saracci et al., 1991 |
IARC cohort |
11 |
1.0 (0.5-1.7) |
Alavanja et al., 1989 |
USDA forest conservationists |
|
1.7 (0.5-5.5) |
|
Soil conservationists |
|
2.4 (1.0-5.9) |
Henneberger et al., 1989 |
Paper and pulp workers |
3 |
1.5 (0.3-4.4) |
Alavanja et al., 1988 |
USDA agricultural extension agents |
|
1.7 (0.9-3.3) |
Bond et al., 1988 |
Dow 2,4-D production workers |
0 |
— (0.0-6.2) |
Robinson et al., 1986 |
Paper and pulp workers |
6 |
1.2 (0.5-3.0) |
Coggon et al., 1986 |
British MCPA production workers |
5 |
1.0 (0.3-2.3) |
Lynge, 1985 |
Danish male production workers |
3 |
0.6 |
Wiklund, 1983 |
Swedish agricultural workers |
775 |
0.8 (0.7-0.9)b |
Blair, 1983 |
Florida pesticide applicators |
1 |
0.5 |
Burmeister, 1981 |
Farmers in Iowa |
178 |
1.1 (NS) |
ENVIRONMENTAL |
|||
Studies reviewed in Update 1996 |
|||
Bertazzi et al., 1993 |
Seveso residents |
|
|
|
Males in zone R |
10 |
0.9 (0.4-1.7) |
|
Females in zone R |
7 |
1.2 (0.5-2.7) |
Studies reviewed in VAO |
|||
Pesatori et al., 1992 |
Seveso male residents zones A and B |
0 |
— |
|
Female residents zones A and B |
1 |
1.1 (0.2-8.1) |
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
VIETNAM VETERANS |
|||
New Studies |
|||
Crane et al., 1997a |
Australian military veterans |
22 |
1.2 (0.8-1.9) |
Crane et al., 1997b |
Australian national service veterans |
3 |
3.9 |
Studies reviewed in Update 1996 |
|||
Visintainer et al., 1995 |
Michigan Vietnam veterans |
21 |
1.4 (0.9-2.2) |
Studies reviewed in VAO |
|||
Breslin et al., 1988 |
Army Vietnam veterans |
55 |
0.9 (0.5-1.5) |
|
Marine Vietnam veterans |
13 |
0.9 (0.5-1.5) |
Kogan and Clapp, 1988 |
Massachusetts Vietnam veterans |
9 |
1.8 (1.0-3.5) |
Anderson et al., 1986a |
Wisconsin Vietnam veterans |
1 |
— |
Anderson et al., 1986b |
Wisconsin Vietnam veterans |
2 |
— |
NOTE: NS = not significant. a Given when available. b 99% CI. |
1997a). A second study examining the mortality experience of conscripted Australian veterans relative to military personnel who did not serve in the conflict reported a statistically significant RR of 3.9 based on three deaths among Vietnam veterans and one in the comparison population between 1982 and 1994 (Crane et al., 1997b). Neither Australian study has exposure information.
Synthesis
Since Update 1996, only the study by Kogevinas and colleagues (1997) points to a possible association of herbicides with renal cancer. However, although this result is mildly suggestive, because of the marginal significance, lack of trend data, and heterogeneity of the cohorts it is not strong enough to outweigh the equivocal results from other studies.
Conclusions
Strength of Evidence in Epidemiologic Studies
In the judgment of the committee there is still inadequate or insufficient evidence to determine whether an association exists between exposure to the herbicides (2,4-D, 2,4,5-T and its contaminant TCDD, cacodylic acid, and picloram) and renal cancer.
Biologic Plausibility
A thorough discussion of biologic plausibility with respect to exposure to TCDD or herbicides and renal cancer is contained in Chapter 3; a summary is presented in the conclusion to this chapter.
BRAIN TUMORS
Background
According to the American Cancer Society, approximately 9,800 men and 7,600 women will be diagnosed with new cases of brain and other nervous system cancers (ICD·9 191.0-191.9, 192.0-192.3, 192.8-192.9) in the United States in 1998, and 7,300 men and 6,000 women will die from them (ACS, 1998). These numbers represent approximately 1.5 percent of new cancer diagnoses and 2.5 percent of deaths.
For individuals in the United States age 45-59, brain cancer is slightly more common in males than females and slightly more common in whites than African Americans.
Exposure to ionizing radiation is an established risk factor for brain cancer. Several other potential factors have been examined, but the American Cancer Society notes that the majority of brain cancers are not associated with any known risk factors.
Average Annual Cancer Incidence (per 100,000 individuals) in the United Statesa Brain Cancer |
|||||||||
|
45-49 years of age |
50-54 years of age |
55-59 years of age |
||||||
|
all races |
white |
black |
all races |
white |
black |
all races |
white |
black |
Males |
7 |
7 |
4 |
10 |
11 |
5 |
14 |
14 |
10 |
Females |
5 |
5 |
3 |
7 |
8 |
3 |
8 |
9 |
5 |
a SEER nine standard registries crude, age-specific rate, 1990-1994. |
Summary of VAO and Update 1996
In VAO and Update 1996, the committee reviewed relevant reports on brain cancers, including studies of chemical production workers in the United States and other countries (Lynge, 1985; Coggon et al., 1986; Bond et al., 1988; Fingerhut et al., 1991; Saracci et al., 1991); agricultural workers (Burmeister, 1981; Alavanja et al., 1988; Musicco et al., 1988; Wigle et al., 1990; Morrison et al., 1992; Ronco et al., 1992; Blair et al., 1993; Dean, 1994); pesticide appliers (Blair et al., 1983; Swaen et al., 1992; Asp et al., 1994); paper and pulp workers (Robinson et al., 1986; Henneberger et al., 1989); the Seveso population (Bertazzi et al., 1989a,b; Pesatori et al., 1992; Bertazzi et al., 1993); and Vietnam veterans (Lawrence et al., 1985; Anderson et al., 1986a,b; Boyle et al., 1987; Breslin et al.,
1988; Thomas and Kang, 1990; Dalager et al., 1995a; Visintainer et al., 1995). The majority of the studies found no excess risk of central nervous system tumors.
Update of the Scientific Literature
In an update and expansion of the IARC cohort study, Kogevinas et al. (1997) examined cancer mortality in a cohort of 26,615 male and female workers engaged in the production or application of phenoxy herbicides. These cohorts were assembled from 12 countries, drawn from national studies that followed the same core protocol jointly developed by the participants and coordinated by IARC.
Nonsignificantly decreased risk of death from brain cancer was observed among the group including all workers exposed to any phenoxy herbicide or chlorophenol (SMR = 0.7, CI 0.4-1.0, 22 deaths). When this group was divided into those exposed and unexposed to TCDD or higher chlorinated dioxins, the result was unchanged, and both groups had nonsignificantly decreased risks (exposed SMR = 0.6, CI 0.3-1.1, 12 deaths; unexposed SMR = 0.8, CI 0.4-1.5, 10 deaths). More detailed analysis by exposure variables such as duration and time since first exposure was not conducted for brain cancers.
Although the study included large numbers of workers who were likely to be exposed at levels substantially higher than general population exposures, lack of information about actual exposure limits the investigator's ability to examine exposure-response relationships within the cohort. In addition the inclusion of workers in the exposed group based on ever having worked in a job considered to involve exposure makes it impossible to distinguish heavily exposed workers from those with very minor exposures.
Becher et al. (1996) examined cancer mortality among workers in four German facilities that produced phenoxy herbicides and chlorophenols. The population included workers who had a least one month of employment, resulting in a cohort consisting of 2,479 male workers. The cohort was assembled from four plants, and analysis was conducted on the total cohort divided into four subcohorts corresponding to each plant considered separately.
Based on production information and limited blood dioxin measurements, subcohorts I and II are supposed to have higher TCDD exposures than subcohorts III and IV. Of the four subcohorts, only group I had at least one observed or expected death for brain cancer. An SMR of 2.3 (CI 0.5-6.8) was reported, based on three observed cases.
Ramlow et al. (1996) examined mortality in a cohort of workers exposed to pentachlorophenol, as part of a larger study of Dow chemical manufacturing workers exposed to higher chlorinated dioxins. The study cohort was assembled from company records, starting with a cohort of 2,192 workers ever employed in a department with potential PCDD exposure between 1937 and 1980.
In the study analysis, the U.S. white male death rates (five-year age and calendar specific) and the non-PCP and PCDD male Dow Michigan employees for 1940-1989
were both used as reference values to calculate expected deaths. Four exposure groups were developed for TCDD (1 unit = very low, 1-1.9 = low, 2-2.9 = medium, 3 = high). Calculation of SMRs with exposure lagged by 15 years using both the U.S. and the Dow referent populations found no significant excess mortality for brain cancer (one death observed, one expected). Brain cancer was not included in the more detailed analysis by the four categories of cumulative exposure.
Cancer mortality among a cohort of rice growers in northern Italy was investigated by Gambini et al. (1997). Using a set of registered farm owners consisting of 1,493 males who worked on farms from 1957 to 1992, they examined the cause of death for 958 subjects and compared this with expected numbers calculated from national rates. No direct exposure information was available, so employment on the farm was used as a surrogate for exposure to the range of phenoxy herbicides employed during the study period. A nonsignificant decrease in brain cancer mortality (SMR = 0.9, CI 0.2-2.3, 4 deaths) was reported. Brain cancer was not included in the more detailed analysis with stratification by age at death and duration of exposure (employment as a farmer).
Bertazzi et al. (1997) continued the follow-up of people environmentally exposed to TCDD in Seveso, Italy. The events that led to the exposure and the methods used to study this population have been fully described in earlier reports. This report updates the population after 15 years follow-up. Death from brain cancer was nonsignificantly elevated for men in zone R (SMR = 1.3, CI 0.7-2.3, 12 observed deaths) and for women in zone B (SMR = 3.2, CI 0.6-9.4, 3 deaths) and zone R (SMR = 1.1, CI 0.5-2.2, 8 deaths). More detailed investigation of subjects exposed in zone B did not include brain cancer.
Svensson et al. (1995) studied mortality and cancer incidence in two cohorts of Swedish fishermen. One group (2,896 men) resided on the east coast of Sweden and consumed fish from the Baltic Sea. These fatty fish (particularly salmon and herring) are reported to contain elevated levels of PCB, PCDD, and PCDF. The other group of fishermen (8,477) resided on the west coast of Sweden and were presumed to have a higher intake of lean (and less contaminated) fish, including cod and flat fish. This distinction in exposure by place of residence is reportedly confirmed by the finding that blood levels of dioxin-like compounds were two times higher among east coast than west coast fishermen; however, no data are provided to support this point. East coast fishermen were found to have nonsignificantly decreased mortality and incidence of brain cancer, whereas west coast fishermen had nonsignificantly decreased incidence and nonsignificantly increased mortality (SMR = 1.0, CI 0.6-1.7, 15 deaths observed) compared to Swedish national rates
In a comparison of mortality between Army Chemical Corps Vietnam and non-Vietnam veterans, Dalger and Kang (1997) reported that there was a nonsignificant excess of deaths from brain cancer among Vietnam veterans. The study compared 2,872 Vietnam veterans with 2,737 non-Vietnam veterans (all of whom served in Chemical Corps specialties). All study subjects served at least 18 months' active duty between 1965 and 1973, and vital status ascertainment was
complete for both groups. A nonsignificant increase in death for Vietnam veterans (RR = 1.1, CI 0.1-4.1, 2 observed deaths) and a nonsignificant decrease for non-Vietnam veterans (RR = 0.8, CI 0.0-4.2, 1 death observed) from brain cancer was found compared to U.S. general population rates. When Vietnam and non-Vietnam cohorts were compared directly, the crude rate ratio of brain cancer death was 1.89 (Vietnam versus non-Vietnam). Direct exposure information on the two cohorts was not available, and the presumption that the Vietnam veterans had potentially higher levels of dioxin exposure because of their duties involving Agent Orange and other dioxin-contaminated herbicides (compared to the non-Vietnam Chemical Corps veterans) has not been verified.
Crane et al. (1997a) examined the mortality experience of male Australian Vietnam veterans from 1980 to 1994. The cohort consists of 59,036 male veterans, who were followed from 22 to 32 years. There were 2,067 deaths recorded among this group from 1980 to 1994, and vital status was ascertained for 96.9 percent of the cohort. No excess mortality was observed for brain cancer in the overall military population (SMR = 1.1, CI 0.8-1.5, 39 deaths) or when analyzed separately by branch of service. Study authors have described the strengths and limitations of the Australian veterans cohort study, including virtually complete identification of the study population, a period of follow-up ranging from 22 to 32 years, and vital status ascertainment of 96.9 percent. Among the weaknesses of the study are the possibility of underascertainment of death and the uncertain quality of exposure assessment to a variety of risk factors, including smoking and alcohol consumption, as well as herbicide and dioxin exposure.
Synthesis
As in VAO and Update 1996, the studies reviewed in this report found a small number of cases of brain tumors, and the RRs associated with herbicide exposure are fairly evenly distributed around 1.0, with relatively narrow confidence intervals.
Conclusions
The committee has not changed the conclusion of the earlier reports that there is limited/suggestive evidence of no association between exposure to the herbicides (2,4-D, 2,4,5-T and its contaminant TCDD, cacodylic acid, and picloram) and brain tumors.
Strength of Evidence in Epidemiologic Studies
The evidence regarding association is drawn from occupational, environmental, and veteran studies in which subjects were exposed to a variety of herbicides and herbicide components. These studies reported a small number of cases of brain tumors, with the risks associated with herbicide exposure fairly evenly distributed around 1.0, and with relatively narrow confidence intervals.
TABLE 7-24 Selected Epidemiologic Studies—Brain Tumors
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
OCCUPATIONAL |
|||
New Studies |
|||
Kogevinas et al., 1997 |
IARC cohort |
|
|
|
Workers exposed to TCDD (or higher chlorinated dioxins) |
12 |
0.6 (0.3-1.1) |
|
Workers not exposed to TCDD (or higher chlorinated dioxins) |
10 |
0.8 (0.4-1.5) |
|
Workers exposed to any phenoxy herbicide or chlorophenol |
22 |
0.7 (0.4-1.0) |
Gambini et al., 1997 |
Italian rice growers |
4 |
0.9 (0.2-2.3) |
Becher et al., 1996 |
German chemical production workers |
|
|
|
Subcohort I |
3 |
2.3 (0.5-6.8) |
|
Subcohort II |
0 |
|
|
Subcohort III |
0 |
|
|
Subcohort IV |
0 |
|
Ramlow et al., 1996 |
Pentachlorophenol production workers |
|
|
|
0 year latency |
1 |
|
|
15 year latency |
1 |
|
Studies reviewed in Update 1996 |
|||
Asp et al., 1994 |
Finnish herbicide applicators |
3 |
1.2 (0.3-3.6) |
Dean, 1994 |
Irish farmers and farm workers |
|
|
|
Males |
195 |
|
|
Females |
72 |
|
Blair et al., 1993 |
U.S. farmers in 23 states |
|
|
|
White males |
447 |
1.2 (1.1-1.3) |
|
Nonwhite males |
16 |
1.0 (0.6-1.6) |
|
White females |
9 |
1.1 (0.5-2.1) |
|
Nonwhite females |
1 |
0.4 (0.0-2.1) |
Studies reviewed in VAO |
|||
Morrison et al., 1992 |
Canadian prairie farmers |
|
|
|
250+ acres sprayed with herbicides |
24 |
0.8 (0.5-1.2) |
Ronco et al., 1992 |
Danish male self-employed farm workers |
194 |
1.1 |
Swaen et al., 1992 |
Dutch herbicide applicators |
3 |
3.2 (0.6-9.3) |
Fingerhut et al., 1991 |
NIOSH cohort |
5 |
0.7 (0.2-1.6) |
Saracci et al., 1991 |
IARC cohort |
6 |
0.4 (0.1-0.8) |
Wigle et al., 1990 |
Saskatchewan farmers |
96 |
1.0 (0.8-1.3) |
Alavanja et al., 1989 |
USDA forest/soil conservationists |
6 |
1.7 (0.6-3.7) |
Henneberger et al., 1989 |
Paper and pulp workers |
2 |
1.2 (0.1-4.2) |
Alavanja et al., 1988 |
USDA agricultural extension agents |
|
1.0 (0.4-2.4) |
Bond et al., 1988 |
Dow 2,4-D production workers |
0 |
— (0.0-4.1) |
Musicco et al., 1988 |
Men and women in the Milan, Italy, area |
61 |
1.6 (1.1-2.4) |
Coggon et al., 1986 |
British MCPA production workers |
11 |
1.2 (0.6-2.2) |
Robinson et al., 1986 |
Paper and pulp workers |
4 |
0.6 (0.2-2.1) |
Lynge, 1985 |
Danish male production workers |
4 |
0.7 |
Blair et al., 1983 |
Florida pesticide applicators |
5 |
2.0 |
Burmeister, 1981 |
Farmers in Iowa |
111 |
1.1 (NS) |
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
ENVIRONMENTAL |
|||
New Studies |
|||
Bertazzi et al., 1997 |
Seveso residents |
|
|
|
Males—zone A |
0 |
|
|
Males—zone B |
1 |
0.8 (0.0-4.2) |
|
Males—zone R |
12 |
1.3 (0.7-2.3) |
|
Females—zone A |
0 |
|
|
Females—zone B |
3 |
3.2 (0.6-9.4) |
|
Females—zone R |
8 |
1.1 (0.5-2.2) |
Svensson et al., 1995 |
Swedish fishermen mortality |
|
|
|
East coast |
2 |
0.6 (0.1-2.1) |
|
West coast |
15 |
1.0 (0.6-1.7) |
Swedish fishermen incidence |
|
|
|
|
East coast |
3 |
0.5 (0.1-1.4) |
|
West coast |
24 |
0.9 (0.6-1.4) |
Studies reviewed in Update 1996 |
|||
Bertazzi et al., 1993 |
Seveso male and female residents— |
|
|
|
zones A and B |
0 |
|
|
Male residents—zone R |
6 |
0.6 (0.3-1.4) |
|
Female residents—zone R |
6 |
1.4 (0.6-3.4) |
Studies reviewed in VAO |
|||
Pesatori et al., 1992 |
Seveso male residents—zones A and B |
0 |
— |
|
Female residents—zones A and B |
1 |
1.5 (0.2-11.3) |
Bertazzi et al., 1989a |
Seveso male residents—zones A, B, R |
5 |
1.2 (0.4-3.1) |
|
Female residents—zones A, B, R |
5 |
2.1 (0.8-5.9) |
VIETNAM VETERANS |
|||
New Studies |
|||
Crane et al., 1997a |
Australian military veterans |
39 |
1.1 (0.8-1.5) |
Crane et al., 1997b |
Australian national service veterans |
13 |
1.4 |
Dalager and Kang, 1997 |
Army Chemical Corps veterans |
2 |
1.9b |
Studies reviewed in Update 1996 |
|||
Dalager et al., 1995a |
Women Vietnam veterans |
4 |
1.4 (0.4-3.7) |
Visintainer et al., 1995 |
Michigan Vietnam veterans |
36 |
1.1 (0.8-1.5) |
Boyle et al., 1987 |
Vietnam Experience Study |
3 |
|
Studies reviewed in VAO |
|||
Thomas and Kang, 1990 |
Army Chemical Corps Vietnam veterans |
2 |
5.0 |
Breslin et al., 1988 |
Army Vietnam veterans |
116 |
1.0 (0.3-3.2) |
|
Marine Vietnam veterans |
25 |
1.1 (0.2-7.1) |
Anderson et al., 1986a |
Wisconsin Vietnam veterans |
13 |
1.6 (0.9-2.7) |
Anderson et al., 1986b |
Wisconsin Vietnam veterans |
8 |
0.8 (0.3-1.5) |
Lawrence et al., 1985 |
New York Vietnam veterans |
4 |
0.5 (0.2-1.5) |
NOTE: NS = not significant. a Given when available. b Crude rate ratio of Vietnam to non-Vietnam veterans. |
Biologic Plausibility
Although a possible association between these exposures and brain cancer is considered plausible given current knowledge of ways in which dioxin and herbicides affect human systems, the literature reviewed for this report does not support a change from the previous conclusion of limited/suggestive evidence of no association. A thorough discussion of biologic plausibility with respect to exposure to TCDD or herbicides and brain tumors is contained in Chapter 3; a summary is presented in the conclusion to this chapter.
NON-HODGKIN'S LYMPHOMA
Background
Non-Hodgkin's lymphoma (NHL) (ICD·9 200.0-200.8, 202.0-202.2, 2028-202.9) is the more common of the two primary types of cancer of the lymphatic system. The American Cancer Society estimates that 31,100 men and 24,300 women will be diagnosed with this disease in the United States in 1998 and that 13,000 men and 11,900 women will die from it (ACS, 1998). Collectively, lymphomas (which also include Hodgkin's disease) are the fifth most common form of cancer in the United States and the sixth leading cause of cancer death.
NHL incidence is uniformly higher in males than females and, in most age groups, higher in whites than African Americans. In the cohorts that characterize most Vietnam veterans, rates increase with age for whites and vary inconsistently for African Americans.
The causes of NHL are poorly understood. Individuals with suppressed or compromised immune systems are known to be at higher risk, and some studies show increased incidence in individuals with HIV, human T-cell lymphotropic virus (HTLV), Epstein-Barr virus, and gastric Helicobacter pylori infections. A number of behavioral, occupational, and environmental risk factors have also been proposed (Blair et al., 1997).
Average Annual Cancer Incidence (per 100,000 individuals) in the United Statesa Non-Hodgkin's Lymphoma |
|||||||||
|
45-49 years of age |
50-54 years of age |
55-59 years of age |
||||||
|
all races |
white |
black |
all races |
white |
black |
all races |
white |
black |
males |
24 |
24 |
33 |
30 |
31 |
27 |
36 |
37 |
31 |
females |
12 |
12 |
9 |
18 |
19 |
16 |
26 |
27 |
14 |
a SEER nine standard registries crude, age-specific rate, 1990-1994. |
Summary of VAO and Update 1996
The original VAO committee concluded that a positive association existed between exposure to herbicides and the development of NHL. A large, well-con-
ducted case-control study in Sweden by Hardell (1981) examined NHL and Hodgkin's disease together and found an OR of 6.0 (CI 3.7-9.7) for exposure to phenoxy acids or chlorophenols, based on 105 cases. These results were replicated in further investigations of the validity of exposure assessment and other potential biases (Hardell, 1981). Similar data by Persson et al. (1989) showed an increased risk for NHL in those exposed to phenoxy acids (OR = 4.9, CI 1.0-27.0), based on a logistic regression analysis of 106 cases. Other studies of farmers and agricultural workers were generally positive for an association between NHL and herbicides or TCDD; however, only some were statistically significant. All of the studies of U.S. agricultural workers reviewed showed elevated RRs (although none were statistically significant), and two National Cancer Institute studies of farmers in Kansas and Nebraska (Hoar et al., 1986; Zahm et al., 1990) showed patterns of increased risk linked to use of 2,4-D. The CDC selected-cancers study found an increased risk of NHL in association with service in Vietnam; other studies of veterans, generally with small sample sizes, were consistent with an association. In contrast, studies of production workers—including the largest, most heavily exposed cohorts (Zober et al., 1990; Fingerhut et al., 1991; Manz et al., 1991; Saracci et al., 1991)—indicated no increased risk. Thus, unlike most of the other cancers studied in VAO, where the data did not distinguish between the effects of herbicides and TCDD, these data suggested that the phenoxy herbicides (including 2,4-D) rather than TCDD are associated with NHL.
Occupational studies discussed in Update 1996 included data demonstrating an increased but nonsignificant risk of NHL associated with manufacture of 2,4,5-T or TCDD (Kogevinas et al., 1995) and with herbicide manufacture (Bloemen et al., 1993; Bueno de Mesquita et al., 1993). Similarly, several studies of agricultural and forestry workers also reported increased risks of NHL in farmers using herbicides (Morrison et al., 1994), including the statistically significant increases reported in a large U.S. PCMR study (Blair et al., 1993) and a cancer mortality study from Ireland (Dean, 1994). Environmental studies included the Seveso cohort (Bertazzi et al., 1993), which showed nonsignificant increases in zones A and B for men. Finally, a study of Michigan veterans (Visintainer et al., 1995) showed that the PMR was significantly increased for NHL, although no data were available on herbicide exposure.
The conclusion of Update 1996 was that the more recently published scientific literature continued to support the positive association between exposure to herbicides and NHL.
Update of the Scientific Literature
Occupational Studies
Production Workers Ramlow et al. (1996) evaluated mortality in a cohort of 770 workers with potential PCP exposure for 1940 through 1989, a subset from
a larger cohort of workers with potential exposure to higher chlorinated dioxins. They found an insignificant increase in SMR for lymphopoietic cancer of 1.3 (CI 0.4-3.1). Becher et al. (1996) have also shown a significant increase in SMR for NHL (SMR = 3.3, CI 1.2-7.1, N = 6) in a cohort of 2,479 workers in four plants in Germany with exposure to phenoxy herbicide and contaminants (dioxins and furans). A variety of herbicides was produced, including those known to have been contaminated with TCDD. Mortality from all neoplasms increased with latency and was highest in the largest plant where the highest blood levels of TCDD were recorded. A larger study of 21,863 workers from 12 countries was reported by Kogevinas et al. (1997). Subjects in this updated and expanded multinational study coordinated by IARC were followed from 1939 to 1992. Exposure was reconstructed using job records, company exposure questionnaires, and serum and adipose tissue dioxin levels. Among workers exposed to phenoxy herbicides contaminated with TCDD or higher chlorinated dioxins, mortality increased from all malignant neoplasms (SMR = 1.1, 95% CI 1.0-1.2, 710 deaths), as well as NHL (SMR = 1.4, CI 0.9-2.1, 24 deaths). Risks for all neoplasms and for lymphomas increased with time since first exposure. In workers exposed to phenoxy herbicides with minimal or no contamination by TCDD and higher chlorinated dioxins, mortality from all neoplasms (SMR = 1.0, CI 0.9-1.1,398 deaths), and NHL (SMR = 1.0, 9 deaths), was similar to that expected.
Agricultural Workers In a study reported by Gambini et al. (1997), data were analyzed for a cohort of rice growers that contained 1,493 subjects and for a follow-up that was more than 99 percent complete with regard to both traced subjects and known causes of deaths. A total of 960 subjects (65 percent) died during the observation period (1957-1992). A nonsignificant risk (SMR = 1.3, CI 0.3-3.3) was found for NHL among workers with longer exposure during the period when phenoxy herbicide was in use (1950-1992). Two recent case-control studies (Amadori et al., 1995; Nanni et al., 1996) reported an OR of 1.8 (CI 1.2-2.6) for agricultural farmer-breeders at risk for development of NHL and chronic lymphocetic leukemia combined. A case-control study of NHL and myeloma incidence rates in Italy (Masala et al., 1996) found higher than expected rates of low-grade follicular lymphomas contributed to the excess NHL cases in agricultural areas (known to be exposed to phenoxyacetic acid herbicides), whereas intermediate histocytic lymphomas produced the excess among urban populations known to have higher than expected rates of exposure to organic solvents.
Tatham et al. (1997) examined the relationship between occupational exposures and three subgroups of NHL: small-cell diffuse lymphomas (N = 185), follicular lymphomas (N = 268), and large-cell diffuse lymphomas (N = 526). There were 1,659 controls available for comparison. After controlling for demographic variables and previously identified risk factors for NHL, a significant
positive association was observed for solvent exposure and small-cell diffuse lymphomas (OR = 1.6, CI 1.1-2.2) and a borderline significant association for meat packaging or processing and follicular lymphoma (OR = 1.6, CI 1.0-2.6). However, exposures to herbicides nonsignificantly elevated the risk of only follicular NHL (farming exposure: OR = 1.1, CI 0.8-1.5; non-farming exposure: OR = 1.3, CI 0.8-1.9). No variance from null was observed for the overall NHL cohort or for specific exposure to chlorophenoxy herbicides.
Lastly, a meta-analysis of several studies described above examining the association between NHL and employment as a farmer in the central United States was performed by Keller-Byrne et al. (1997) to test the observation of Blair et al. (1993) that this group is at excess risk of NHL. Six studies were selected for meta-analysis, and the estimated RR was 1.3 (CI = 1.2-1.6).
Environmental Studies
A single new study describing a small German cohort involved in a 1953 reactor accident was reported by Ott and Zober (1996). No cases of NHL have yet been reported among 243 workers followed until 1992, including those whose exposure to TCDD, based on analysis of blood lipids, indicated that they were among a higher-risk group (TCDD doses of > 1 μg/kg body weight). The recently updated Seveso cohort analysis (Bertazzi et al., 1997) also failed to show any increase in NHL deaths among men or women with the exception of men in zone B (RR = 3.3, CI 0.4-11.9).
Vietnam Veteran Studies
In a report by Watanabe and Kang (1996), the mortality experience of 33,833 U.S. Army and Marine Corps Vietnam veterans who died during 1965-1988 was compared with that of 36,797 deceased non-Vietnam veterans using PMRs. Military service information was abstracted from military personnel records and cause of death information from death certificates. This study showed statistically increased risks of NHL among Marine (PMR 1.7, CI 1.2-2.2) but not among Army veterans. Similarly, the Australian Vietnam veterans' report (Crane et al., 1997a) reported the RR of death due to NHL to be 1.3 (CI 0.5-3.5) although this was not statistically significant.
Cytogenetic Studies
To further investigate the possible relationships between agricultural pesticide exposure and increased risk of NHL among farm workers in the north central United States, Garry et al. (1996) performed G-banded chromosome analyses of peripheral blood from workers classified according to primary types of pesticide exposure: herbicides (N = 20), insecticides (N = 18), fumigants (N = 23), and
occupationally unexposed controls (N = 33). Significantly increased rearrangement frequencies were demonstrated in fumigant and insecticide appliers compared to control subjects. At certain chromosome bands, significant excesses of breaks were observed in herbicide appliers, but no breaks were observed in controls. Some of these bands contained genes with potential implications for cancer risk, including oncogenes and genes involved in tumor suppression and apoptosis. Of particular interest with regard to lymphoma risk were the excess rearrangements and breaks involving band 14q32 in fumigant appliers and the excess breaks involving band 18q21 in herbicide appliers; translocations linking 14q32 and 18q21 are the most common rearrangements observed in NHL patients. The potential pathobiological relevance of these cytogenetic events warrants additional investigation at the molecular level.
Synthesis
The recent scientific literature continues to support the conclusion of a positive association between exposure to herbicides and non-Hodgkin's lymphoma and provides a biological rationale that includes the possible interaction between environmental toxins such as phenoxyacid herbicides and oncogene abnormalities of bcl-1 and bcl-2 found in follicular NHL, a tumor that is sharply increasing in incidence in Western countries.
Conclusions
Strength of Evidence in Epidemiologic Studies
Evidence continues to accumulate to conclude of a positive association between exposure to the herbicides (2,4-D, 2,4,5-T and its contaminant TCDD, cacodylic acid, and picloram) and non-Hodgkin's lymphoma. The strength of the evidence regarding association is drawn from occupational and other studies in which subjects were exposed to a variety of herbicides and herbicide components and is increasing in significance, including among studies of Vietnam veterans.
Biologic Plausibility
The biologic plausibility for an etiological cause of follicular NHL, at least, is also strong, given the increasing incidence of these tumors in Western countries and the demonstration of association of high exposure to herbicides and specific chromosome breaks related to NHL oncogenesis. A more thorough discussion of biologic plausibility with respect to exposure to TCDD or herbicides and NHL is contained in Chapter 3; a summary is presented in the conclusion to this chapter.
TABLE 7-25 Selected Epidemiologic Studies—Non-Hodgkin's Lymphoma
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
OCCUPATIONAL |
|||
New Studies |
|||
Gambini et al., 1997 |
Italian rice growers |
|
1.3 (0.3-3.3) |
Kogevinas et al., 1997 |
IARC cohort |
|
|
|
Workers exposed to TCDD (or higher chlorinated dioxins) |
24 |
1.4 (0.9-2.1) |
|
Workers exposed to any phenoxy herbicide or chlorophenol |
9 |
1.0 |
Becher et al., 1996 |
German chemical production workers |
6 |
3.3 (1.2-7.1) |
Keller-Byrne et al., 1997 |
Farmers in the central United States |
|
1.3 (1.2-1.6) |
Nanni et al., 1996 |
Italian farming and animal-breeding workers |
23e |
1.8 (1.2-2.6) |
Ramlow et al., 1996 |
Pentachlorophenol production workers |
1.3 (0.4-3.1) |
|
Amadori et al., 1995 |
Italian farming and animal-breeding workers |
164 |
1.8 (1.2-2.6) |
Studies reviewed in Update 1996 |
|||
Kogevinas et al., 1995 |
IARC cohort diagnosed with NHL |
|
|
|
Exposed to 2,4,5-T |
|
1.9 (0.7-4.8) |
|
Exposed to TCDD |
|
1.9 (0.7-5.1) |
Asp et al., 1994 |
Finnish herbicide applicators |
1 |
0.4 (0.0-2.0) |
Dean, 1994 |
Irish farmers and farm workers |
|
|
|
Males |
244e |
|
|
Females |
84e |
|
Hardell et al., 1994 |
Male residents of northern Sweden |
|
|
|
Exposure to phenoxy herbicides |
25 |
5.5 (2.7-11.0) |
|
Exposure to chlorophenols |
35 |
4.8 (2.7-8.8) |
Morrison et al., 1994 |
Farm operators in 3 Canadian provinces |
|
|
|
All farm operators |
|
0.8 (0.7-0.9) |
|
Highest quartile of herbicides sprayed |
19 |
2.1 (1.1-3.9) |
|
Highest quartile of herbicides sprayed relative to no spraying |
6 |
3.0 (1.1-8.1) |
Blair et al., 1993 |
U.S. farmers in 23 states (white males) |
843 |
1.2 (1.1-1.3) |
Bloemen et al., 1993 |
Dow 2,4-D production workers |
2 |
2.0 (0.2-7.1) |
Bueno de Mesquita et al., 1993 |
Dutch production workers |
|
|
|
Workers exposed to phenoxy herbicides |
2 |
3.0 (0.4-10.8) |
Lynge, 1993 |
Danish male production workers |
10 |
1.7 (0.5-4.5) |
Persson et al., 1993 |
Swedish NHL patients |
|
|
|
Exposure to phenoxy herbicides |
|
2.3 (0.7-7.2) |
|
Occupation as a lumberjack |
|
6.0 (1.1-31.0) |
Zahm et al., 1993 |
Females in eastern Nebraska farms |
|
1.0 (0.7-1.4) |
Kogevinas et al., 1992 |
IARC cohort |
|
|
|
Workers exposed to any phenoxy herbicide or chlorophenol |
11 |
1.0 (0.5-1.7) |
Studies reviewed in VAO |
|||
Hansen et al., 1992 |
Danish gardeners—men and women |
8 |
2.0 (0.9-3.9) |
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
Ronco et al., 1992 |
Danish farm workers—self-employed and employees |
147 |
1.0 |
|
Italian farm workers—self-employed and employees |
14 |
1.3 |
Smith and Christophers 1992 |
Male residents of Australia |
|
|
|
Exposure > 1 day |
15 |
1.5 (0.6-3.7) |
|
Exposure > 30 days |
7 |
2.7 (0.7-9.6) |
Swaen et al., 1992 |
Dutch herbicide applicators |
0 |
— |
Vineis et al., 1991 |
Residents of selected Italian provinces |
|
|
|
Male residents of contaminated areas |
|
2.2 (1.4-3.5) |
Wigle et al., 1990 |
Canadian farmers |
|
|
|
All farmers |
103 |
0.9 (0.8-1.1) |
|
Farmers spraying herbicides on 250+ acres |
10 |
2.2 (1.0-4.6) |
Zahm et al., 1990 |
White male residents of Nebraska |
|
|
|
Ever done farm work |
147 |
0.9 (0.6-1.4) |
|
Ever mixed or applied 2,4-D |
43 |
1.5 (0.9-2.5) |
Alavanja et al., 1989 |
USDA soil conservationists |
|
1.8 (0.7-4.1) |
|
USDA forest conservationists |
|
2.5 (1.0-6.3) |
Corrao et al., 1989 |
Italian farmers licensed to apply pesticides |
|
|
|
Licensed pesticide users and nonusers |
45 |
1.4 (1.0-1.9) |
|
Farmers in arable land areas |
31 |
1.8 (1.2-2.5) |
LaVecchia et al., 1989 |
Residents of the Milan, Italy, area |
|
|
|
Agricultural occupations |
|
2.1 (1.3-3.4) |
Persson et al., 1989 |
Orebro Hospital |
|
|
|
Exposed to phenoxy acids |
6 |
4.9 (1.0-27.0) |
Wiklund et al., 1989b |
Swedish pesticide applicators |
27 |
1.1 (0.7-1.6) |
Alavanja et al., 1988 |
USDA extension agents |
|
1.2 (0.7-2.3) |
Dubrow et al., 1988 |
Ohio residents |
15 |
1.6 (0.8-3.4) |
Olsson and Brandt, 1988 |
Lund Hospital patients |
|
|
|
Exposed to herbicides |
|
1.3 (0.8-2.1) |
|
Exposed to chlorophenols |
|
1.2 (0.7-2.0) |
Wiklund et al., 1988a |
Swedish agricultural and forestry workers |
|
|
|
Workers in land/animal husbandry |
|
1.0 (0.9-1.1) |
|
Timber cutters |
|
0.9 (0.7-1.1) |
Pearce et al., 1987 |
Male residents of New Zealand |
|
|
|
Farming occupations |
|
1.0 (0.7-1.5) |
|
Fencing work |
|
1.4 (0.9-2.2) |
Woods et al., 1987 |
Male residents of Washington State |
|
|
|
Phenoxy herbicide use |
|
1.1 (0.8-1.4) |
|
Chlorophenol use |
|
1.0 (0.8-1.2) |
|
Farming occupations |
|
1.3 (1.0-1.7) |
|
Forestry herbicide applicators |
|
4.8 (1.2-19.4) |
Hoar et al., 1986 |
Kansas residents |
|
|
|
Farmers compared to nonfarmers |
133 |
1.4 (0.9-2.1) |
|
Farmers using herbicides > 20 days/year |
7 |
6.0 (1.9-19.5) |
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
Pearce et al., 1986b |
Male residents of New Zealand |
|
|
|
Agricultural sprayers |
19c |
1.5 (0.7-3.3) |
Pearce et al., 1985 |
Male residents of New Zealand |
|
|
|
Agricultural occupations, ages 20-64 |
|
1.4 (0.9-2.0) |
Burmeister et al., 1983 |
Iowa residents |
|
|
|
Farmers |
|
1.3 |
|
Farmers in 33 counties with highest herbicide use: |
|
|
|
Born before 1890 |
|
3.4 |
|
Born 1890-1900 |
|
2.2 |
|
Born after 1900 |
|
1.3 |
Riihimiki et al., 1983 |
Finnish herbicide applicators |
0 |
— |
Wiklund, 1983 |
Swedish agricultural workers |
|
1.1 (0.9-1.2) |
Cantor, 1982 |
Wisconsin residents |
175 |
1.2 (1.0-1.5) |
Hardell et al., 1980 |
Umea Hospital patients |
|
|
|
Exposed to phenoxy acids |
41 |
4.8 (2.9-8.1)b |
|
Exposed to chlorophenols |
50 |
4.3 (2.7-6.9)b |
ENVIRONMENTAL |
|||
New Studies |
|||
Bertazzi et al., 1997 |
Seveso residents |
|
|
|
Males in zone B |
|
3.3 (0.4-11.9) |
Studies reviewed in Update 1996 |
|||
Bertazzi et al., 1993 |
Seveso residents |
|
|
|
Males in zone B |
3 |
2.3 (0.7-7.4) |
|
Females in zone B |
1 |
0.9 (0.1-6.4) |
|
Males in zone R |
12 |
1.3 (0.7-2.5) |
|
Females in zone R |
10 |
1.2 (0.6-2.3) |
Studies reviewed in VAO |
|||
Lampi et al., 1992 |
Finnish community exposed to chlorophenols |
|
|
|
Compared to two uncontaminated municipalities |
|
2.8 (1.4-5.6) |
|
Compared to cancer control region |
|
2.1 (1.3-3.4) |
Pesatori et al., 1992 |
Seveso residents |
|
|
|
Males in zones A and B |
3 |
|
|
Females in zones A and B |
1 |
|
|
Males in zone R |
13 |
|
|
Females in zone R |
10 |
|
Bertazzi et al., 1989b |
Seveso residents |
|
|
|
Females in zone B |
2 |
|
|
Males in zone R |
3 |
|
|
Females in zone R |
4 |
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
VIETNAM VETERANS |
|||
New Studies |
|||
Crane et al., 1997a |
Australian military veterans |
|
1.3 (0.5-3.5) |
Watanabe and Kang, 1996 |
Marine Vietnam veterans |
|
1.7 (1.2-2.2) |
Studies reviewed in Update 1996 |
|||
Visintainer et al., 1995 |
Michigan Vietnam veterans |
32 |
1.5 (1.0-2.1) |
Studies reviewed in VAO |
|||
Clapp et al., 1991 |
Massachusetts Vietnam veterans |
|
1.2 (0.6-2.4) |
Dalager et al., 1991 |
Vietnam veterans diagnosed with NHL |
100 |
1.0 (0.7-1.8) |
O'Brien et al., 1991 |
Army enlisted Vietnam veterans |
7 |
1.8 |
Thomas et al., 1991 |
Women Vietnam veterans |
3 |
1.3 (0.3-1.8) |
Watanabe et al., 1991 |
Army Vietnam veterans compared to Vietnam-era Army veterans |
140 |
0.8 |
|
Army Vietnam veterans compared to combined Army and Marine Vietnam-era veterans |
140 |
0.9 |
|
Marine Vietnam veterans compared to Vietnam-era veterans |
42 |
1.8 |
|
Marine Vietnam veterans compared to combined Army and Marine Vietnam-era veterans |
42 |
1.2 |
CDC, 1990 |
U.S. men born between 1921 and 1953 |
|
|
|
Vietnam veterans |
99 |
1.5 (1.1-2.0) |
|
Army Vietnam veterans |
45 |
1.2 (0.8-1.8) |
|
Marine Vietnam veterans |
10 |
1.8 (0.8-4.3) |
|
Air Force Vietnam veterans |
12 |
1.0 (0.5-2.2) |
|
Navy Vietnam veterans |
32 |
1.9 (1.1-3.2) |
|
Blue-water Navy Vietnam veterans |
28 |
2.2 (1.2-3.9) |
Michalek et al., 1990 |
Air Force Ranch Hand veterans mortality |
0 |
|
Wolfe et al., 1990 |
Air Force Ranch Hand veterans morbidity |
1 |
|
Breslin et al., 1988 |
Army Vietnam veterans |
108 |
0.8 (0.6-1.0) |
|
Marine Vietnam veterans |
35 |
2.1 (1.2-3.8) |
Garland et al., 1988 |
Navy enlisted personnel 1974-1983 |
|
0.7 |
Burt et al., 1987 |
Army combat Vietnam veterans |
39 |
1.1 (0.7-1.5) |
|
Marine combat Vietnam veterans |
17 |
3.2 (1.4-7.4) |
|
Army Vietnam veterans (service 1967-1969) |
64 |
0.9 (0.7-1.3) |
|
Marine Vietnam veterans (service 1967-1969) |
17 |
2.5 (1.1-5.8) |
Fett et al., 1987b |
Australian Vietnam veterans |
4 |
1.8 (0.4-8.0) |
Anderson et al., 1986a |
Wisconsin Vietnam veterans |
|
|
|
Wisconsin Vietnam veterans compared to Wisconsin nonveterans |
13 |
0.7 |
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
|
Wisconsin Vietnam veterans compared to non-Vietnam-era veterans |
13 |
0.6 |
|
Wisconsin Vietnam veterans compared to Vietnam-era veterans |
13 |
1.0 |
Anderson et al., 1986b |
Wisconsin Vietnam veterans compared to general population |
24 |
0.7 |
|
Wisconsin Vietnam veterans compared to Wisconsin veterans |
24 |
1.1 |
Holmes et al., 1986 |
West Virginia Vietnam veterans compared to West Virginia Vietnam-era veterans |
2 |
1.1 |
Lawrence et al., 1985 |
New York Vietnam veterans |
10 |
1.0 (0.4-2.2) |
a Given when available. b Includes both NHL and Hodgkin's disease. c Only NHL other than lymphosarcoma and reticulosarcoma (ICD 202). d NHL, 4 living cases and 3 deaths listed by Boyle et al., 1987. e Includes NHL and chronic lymphocetic leukemia (CLL) combined. f Includes all lymphomas combined. |
HODGKIN'S DISEASE
Background
Hodgkin's disease (HD) (ICD·9 201.0-201.9) is distinct from NHL in its cell of origin, demographics, and genetics. According to American Cancer Society estimates, 3,700 men and 3,400 women will be diagnosed with the disease in the United States in 1998, and 700 men and an equal number of women will die from it (ACS, 1998).
HD is less common in individuals in the age groups that characterize most Vietnam veterans than in individuals both younger and older. For individuals older than 40, the incidence rate for males generally exceeds that for females and the rate for whites exceeds that for African Americans. However, the very small number of cases indicates that care should be exercised when interpreting the figures.
The potential infectious nature of HD has been a topic of discussion since its earliest description. Increased incidence in individuals with a history of infectious mononucleosis has been observed in some studies, and a link with Epstein-Barr virus has been proposed. In addition to the occupational associations discussed below, higher rates of the disease have been observed in individuals with suppressed or compromised immune systems.
Average Annual Cancer Incidence (per 100,000 individuals) in the United Statesa Hodgkin's Disease |
|||||||||
|
45-49 years of age |
50-54 years of age |
55-59 years of age |
||||||
|
all races |
white |
black |
all races |
white |
black |
all races |
white |
black |
males |
3.4 |
3.5 |
3.4 |
3.6 |
3.7 |
4.3 |
3.4 |
3.3 |
4.3 |
females |
1.7 |
1.7 |
2.5 |
1.5 |
1.7 |
1.5 |
1.7 |
1.9 |
1.7 |
a SEER nine standard registries crude, age-specific rate, 1990-1994. |
Summary of VAO and Update 1996
HD, also a malignant lymphoma, is a neoplastic disease characterized by painless, progressive enlargement of lymph nodes, spleen, and general lymphoid tissues. Fewer studies have been conducted of HD in relation to exposure to herbicides or TCDD than of NHL, but the pattern of results is consistent. The 60 HD cases in the study by Hardell et al. (1981) were examined later by Hardell and Bengtsson (1983), who found ORs of 2.4 (CI 0.9-6.5) for low-grade exposure to chlorophenols and 6.5 (CI 2.7-19.0) for high-grade exposure. The study by Persson et al. (1989) of 54 HD cases showed a large, but not statistically significant, OR of 3.8 (CI 0.5-35.2) for exposure to phenoxy acids. Furthermore, nearly all of the 13 case-control and agricultural worker studies show an increased risk for HD, although only a few of these results are statistically significant for HD. As with NHL, even the largest studies of production workers exposed to TCDD do not indicate an increased risk. The few studies of HD in Vietnam veterans tend to show elevated risks; all but one are statistically significant.
Occupational studies have included those of Kogevinas (1993) showing no increase in HD among 13,898 production workers or sprayers and a negative U.S. PCMR study of farmers in 23 states (Blair, 1993). A smaller study among Finnish herbicide appliers (Asp et al., 1994) showed a nonsignificant increase in SIR, whereas another study (Persson et al., 1993) reported a significant increase in OR for HD among Swedish farmers exposed to phenoxy acid herbicides.
Neither environmental nor veteran studies provided any additional data to strengthen an association, primarily because of the very low incidence of HD among the Seveso cohorts (Bertazzi et al., 1993) and in selected veteran studies (Visintainer et al., 1995). Update 1996 nonetheless concluded that the data continued to support a positive association between exposure to herbicides and HD, based primarily on occupational and environmental studies.
Update of the Scientific Literature
Occupational Studies
Two studies recently reported with positive (although nonsignificant) data for the association between TCDD (Becher et al., 1996) or PCP (Ramlow et al.,
1996) and NHL found no similar association for HD among manufacturing workers. An update of a very large occupational cohort by Kogevinas et al. (1997) also showed no association between HD and phenoxy herbicides or chlorophenols (SMR = 1.0, CI 0.5-1.8) but did show a nonsignificant increase of HD among workers exposed to TCDD or higher chlorinated hydrocarbons (SMR = 1.3, CI 0.6-2.5). Among agricultural workers, Gambini et al. (1997) described no increase in HD (SMR = 0.7, CI 0.1-3.6) among rice growers in northern Italy, again confounded by the extremely low incidence of this malignancy (one case observed among 221 cancer deaths). A prospective epidemiologic study conducted in a rural farming community in Michigan (Waterhouse et al., 1996) demonstrated a significant increase in the combined incidence of lymphopoietic neoplasms, namely NHL, HD, and chronic lymphocytic leukemia (CLL); the combined SIR was 1.4 (95% CI 1.0-1.9; p = .03) and the SIR for HD was 2.9 (CI 1.1-3.4).
Environmental Studies
The Seveso cohort analysis has recently been updated (Bertazzi et al., 1997) and remains one of the highest documented exposures to TCDD but is limited by the relatively small size of the population in zone A. It has now shown an increase in HD in zone B for men (RR = 3.3, CI 0.4-11.9) and for women (RR = 6.5, CI 0.7-23.5), but no increase for either sex in zone R.
Vietnam Veteran Studies
Watanabe and Kang (1996) compared the mortality experience of 33,833 U.S. Army and Marine Corps Vietnam veterans who died during 1965-1988 with that of 36,797 deceased non-Vietnam veterans using PMRs. Service information was abstracted from military personnel records, and cause of death information from death certificates. As with NHL, they demonstrated a significantly increased PMR (1.9, CI 1.2-2.7) among Marine, but not among Army, Vietnam veterans.
Synthesis
The data for HD are more limited than for NHL, due primarily to the lower incidence of this lymphoreticular tumor. Nonetheless, data drawn from agricultural, production, and environmental exposures and, more recently, from Vietnam veterans continues to support the conclusions of a positive association between herbicides and HD. Although not as clearly demonstrated as for NHL, biologic plausibility also exists for a positive association between TCDD and the development of HD due to their common lymphoreticular origin and association with common risk factors.
Conclusions
Strength of Evidence in Epidemiologic Studies
Evidence is sufficient to conclude that there is a positive association between exposure to the herbicides (2,4-D, 2,4,5-T and its contaminant TCDD, cacodylic acid, and picloram) and Hodgkin's disease. The evidence regarding association is drawn from occupational and other studies in which subjects were exposed to a variety of herbicides and herbicide components.
TABLE 7-26 Selected Epidemiologic Studies—Hodgkin's Disease
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
OCCUPATIONAL |
|||
New Studies |
|||
Gambini et al., 1997 |
Italian rice growers |
1 |
0.7 (0.1-3.6) |
Kogevinas et al., 1997 |
IARC cohort |
|
1.0 (0.5-1.8) |
Becher et al., 1996 |
German chemical production workers |
|
NS |
Ramlow et al., 1996 |
Pentachlorophenol production workers |
|
NS |
Waterhouse et al., 1996 |
Residents of Tecumseh, Michigan |
|
2.9 (1.1-3.4) |
Studies reviewed in Update 1996 |
|||
Asp et al., 1994 |
Finnish herbicide applicators |
2 |
1.7 (0.2-6.0) |
Blair et al., 1993 |
U.S. farmers in 23 states—white males |
56 |
1.0 (0.8-1.3) |
Kogevinas et al., 1993 |
IARC cohort—females |
1 |
|
Persson et al., 1993 |
Swedish NHL patients |
|
|
|
Exposure to phenoxy herbicides |
5 |
7,4 (1.4-40.0)d |
Kogevinas et al., 1992 |
IARC cohort |
3 |
0.6 (0.1-1.7) |
Studies reviewed in VAO |
|||
Eriksson et al., 1992 |
Swedish Cancer Registry patients |
|
|
|
Male sawmill workers |
10 |
2.2 |
|
Male farmers |
97 |
1.2 |
|
Male forestry workers |
35 |
1.2 |
|
Male horticulture workers |
11 |
1.2 |
Ronco et al., 1992 |
Danish and Italian farm workers |
|
|
|
Male Danish farmers—self-employed |
27 |
0.6 |
|
Male Italian farmers—self-employed |
10 |
2.9 |
|
Male Italian farmers—employees |
1 |
0.4 |
|
Male Italian farmers—self-employed and employees |
11 |
1.9 |
|
Female Italian farmers—self-employed |
1 |
1.9 |
Swaen et al., 1992 |
Dutch herbicide applicators |
1 |
3.3 |
Fingerhut et al., 1991 |
NIOSH cohort |
3 |
1.2 (0.3-3.5) |
|
20 years latency, 1+ years exposure |
1 |
— |
Green, 1991 |
Herbicide sprayers in Ontario |
0 |
— |
Saracci et al., 1991 |
IARC cohort |
2 |
0.4 (0.1-1.4) |
Zober et al., 1990 |
BASF production workers |
0 |
— |
Alavanja et al., 1989 |
USDA forest/soil conservationists |
4 |
2.2 (0.6-5.6) |
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
LaVecchia et al., 1989 |
Residents of the Milan, Italy, area |
|
|
|
Agricultural occupations |
|
2.1 (1.0-3.8) |
|
Chemical industry occupations |
|
4.3 (1.4-10.2) |
Persson et al., 1989 |
Orebro Hospital patients |
|
|
|
Farming |
6 |
1.2 (0.4-3.5) |
|
Exposed to phenoxy acids |
4 |
3.8 (0.5-35.2) |
Wiklund et al., 1989b |
Swedish pesticide applicators |
15 |
1.5 (0.8-2.4) |
Alavanja et al., 1988 |
USDA agricultural extension agents |
|
|
|
PMR analysis |
6 |
2.7 (1.2-6.3) |
|
Case-control analysis |
6 |
1.1 (0.3-3.5) |
Bond et al., 1988 |
Dow workers with chloracne |
1 |
|
Dubrow et al., 1988 |
Ohio residents |
3 |
2.7 |
Wiklund et al., 1988a |
Swedish agricultural and forestry workers |
|
|
|
Workers in land/animal husbandry |
242 |
1.0 (0.9-1.2) |
|
Workers in silviculture |
15 |
2.3 (1.3-3.7) |
Hoar et al., 1986 |
Kansas residents |
|
|
|
All farmers |
71 |
0.8 (0.5-1.2) |
|
Farm use of herbicides (phenoxy acids and others) |
|
0.9 (0.5-1.5) |
|
Farmers using herbicides >20 days/year |
3 |
1.0 (0.2-4.1) |
|
Farmers using herbicides >15 years |
10 |
1.2 (0.5-2.6) |
Pearce et al., 1985 |
Male residents of New Zealand |
|
|
|
Agricultural occupations, ages 20-64 |
|
1.0 (0.6-2.0) |
Burmeister et al., 1983 |
Iowa residents |
|
1.4 |
Hardell and Bengtsson, 1983 |
Umea Hospital patients |
|
|
|
Exposed to phenoxy acids |
6 |
5.0 (2.4-10.2) |
|
Exposed to high-grade chlorophenols |
9 |
6.5 (2.7-19.0) |
|
Exposed to low-grade chlorophenols |
5 |
2.4 (0.9-6.5) |
Riihimaki et al., 1983 |
Finnish herbicide applicators |
0 |
— |
Wiklund, 1983 |
Swedish agricultural workers |
226 |
1.0 (0.9-1.2)b |
Burmeister, 1981 |
Farmers in Iowa |
|
1.2 |
Hardell et al., 1980 |
Umea Hospital patients |
|
|
|
Exposed to phenoxy acids |
41 |
4.8 (2.9-8.1)c |
|
Exposed to chlorophenols |
50 |
4.3 (2.7-6.9)c |
ENVIRONMENTAL |
|||
New Studies |
|||
Bertazzi et al., 1997 |
Seveso residents |
|
|
|
zone B—male |
2 |
3.3 (0.4-11.9) |
|
zone B—female |
2 |
6.5 (0.7-23.5) |
|
zone R—female |
4 |
1.9 (0.5-4.90 |
Studies reviewed in Update 1996 |
|||
Bertazzi et al., 1993 |
Seveso residents |
|
|
|
zone A—male |
0 |
— |
|
zone A—female |
0 |
— |
|
zone B—male |
1 |
1.7 (0.2-12.8) |
|
zone B—female |
1 |
2.1 (0.3-15.7) |
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
|
zone R—male |
4 |
1.1 (0.4-3.1) |
|
zone R—female |
3 |
1.0 (0.3-3.2) |
VIETNAM VETERANS |
|||
New Studies |
|||
Watanabe and Kang, 1996 |
Marine and Army Vietnam veterans |
|
1.9 (1.2-2.7) |
Studies reviewed in Update 1996 |
|||
Visintainer et al., 1995 |
Michigan Vietnam veterans |
20 |
1.1 (0.7-1.8) |
Studies reviewed in VAO |
|||
Watanabe et al., 1991 |
Army Vietnam veterans compared to Vietnam-era Army veterans |
116 |
1.0 |
|
Marine Vietnam veterans compared to Vietnam-era veterans |
25 |
1.9 |
|
Army Vietnam veterans compared to Vietnam-era veterans |
116 |
1.1 |
|
Marine Vietnam veterans compared to Vietnam-era veterans |
25 |
1.0 |
CDC, 1990 |
U.S. men born between 1921 and 1953 |
|
|
|
Vietnam veterans |
28 |
1.2 (0.7-2.4) |
|
Army Vietnam veterans |
12 |
1.0 (0.5-2.0) |
|
Marine Vietnam veterans |
4 |
1.7 (0.5-5.9) |
|
Air Force Vietnam veterans |
5 |
1.7 (0.6-4.9) |
|
Navy Vietnam veterans |
7 |
1.1 (0.4-2.6) |
Michalek et al., 1990 |
Air Force Ranch Hand veterans mortality |
0 |
— |
Wolfe et al., 1990 |
|
|
|
Breslin et al, 1988 |
Army Vietnam veterans compared to Vietnam-era Army veterans |
92 |
1.2 (0.7-1.9) |
|
Marine Vietnam veterans compared to Marine Vietnam-era veterans |
22 |
1.3 (0.7-2.6) |
Boyle et al., 1987 |
Vietnam Experience Study |
0 |
— |
Fett et al.. 1987 |
Australian Vietnam veterans |
0 |
— |
Anderson et al., 1986a |
Wisconsin Vietnam veterans compared to Wisconsin nonveterans |
6 |
0.5 (0.2-1.2) |
|
Wisconsin Vietnam veterans compared to non-Vietnam-era veterans |
6 |
1.0 (0.4-2.2) |
|
Wisconsin Vietnam veterans compared to Vietnam-era veterans |
6 |
1.0 (0.4-2.1) |
Anderson et al., 1986b |
Wisconsin Vietnam veterans |
4 |
— |
Holmes et al., 1986 |
West Virginia Vietnam veterans compared to West Virginia Vietnam-era veterans |
5 |
8.3 (2.7-19.5) |
Lawrence et al., 1985 |
New York Vietnam veterans compared to New York Vietnam-era veterans |
10c |
1.0 (0.4-2.2) |
a Given when available. b 99% CI. c Includes both non-Hodgkin's lymphoma and Hodgkin's disease. d 90% CI. |
Biologic Plausibility
A thorough discussion of biologic plausibility with respect to exposure to TCDD or herbicides and Hodgkin's disease is contained in Chapter 3; a summary is presented in the conclusion to this chapter.
MULTIPLE MYELOMA
Background
Multiple myeloma (MM) (ICD·9 203.0, 203.2-203.8) is characterized by proliferation of bone marrow stem cells that results in an excess of neoplastic plasma cells and the production of excess abnormal proteins, usually immunoglobulins. The American Cancer Society estimates that 7,200 men and 6,200 women will be diagnosed with this disease in 1998 and that 5,800 men and 5,500 women will die from it (ACS, 1998).
MM incidence is highly age dependent, with a relatively low rate in individuals under 40 and most cases occurring between 55 and 70 years of age. Rates for African Americans are about twice those for whites. Within racial groups, incidence in males is slightly higher than in females.
Increased incidence of MM has been observed in several occupational groups, including farmers and agricultural workers and those with workplace exposure to rubber, leather, paint, and petroleum (Riedel et al., 1991). Individuals with high exposure to ionizing radiation are also at greater risk. Evidence regarding other risk factors is mixed.
Average Annual Cancer Incidence (per 100,000 individuals) in the United Statesa Multiple Myeloma |
|||||||||
|
45-49 years of age |
50-54 years of age |
55-59 years of age |
||||||
|
all races |
white |
black |
all races |
white |
black |
all races |
white |
black |
males |
4 |
3 |
7 |
6 |
5 |
14 |
11 |
10 |
20 |
females |
2 |
2 |
5 |
5 |
4 |
13 |
7 |
6 |
20 |
a SEER nine standard registries crude, age-specific rate, 1990-1994. |
Summary of VAO and Update 1996
Although multiple myeloma has been less extensively studied than the NHL, a consistent pattern of elevated risk was described in most studies evaluated for VAO and Update 1996. Several studies of agricultural and forestry workers provided information on MM risk in relation to herbicide or pesticide exposure. These studies demonstrated an OR or SMR greater than 1.0; several did so at a statistically significant level. The committee determined that the evidence for this association was limited/suggestive, because the individuals in the existing studies (mostly farmers) have, by nature of their occupation, probably been exposed to a
range of potentially carcinogenic agents other than herbicides and TCDD. MM, like NHL and HD, for which there is stronger epidemiologic evidence of an association, is derived from lymphoreticular cells, which adds to the biologic plausibility of an association. The committee concluded that there was continuing accumulation of limited/suggestive evidence of an association between exposure to the herbicides (2,4-D, 2,4,5-T and its contaminant TCDD, cacodylic acid, and picloram) and MM. The evidence regarding this association was drawn from occupational and other studies in which subjects were exposed to a variety of herbicides and herbicide components.
Update of the Scientific Literature
Occupational Studies
Production Workers In an update and expansion of the IARC cohort study, Kogevinas et al. (1997) examined cancer mortality in a cohort of 26,615 male and female workers engaged in the production or application of phenoxy herbicides. These workers were assembled from 12 countries, drawn from national studies that followed the same core protocol developed jointly by the participants and coordinated by IARC. Of the total study population, 21,863 (20,851 men and 1,012 women) were classified as exposed to phenoxy herbicides or chlorophenols based on individual job records and company exposure questionnaires; 4,160 were unexposed; and 592 were classified as unknown exposure status. The great majority of workers were considered exposed if they had ever worked in production or spraying of phenoxy herbicides or chlorophenols (four cohorts were exceptions, with minimum employment periods of 1 to 12 months). The period of follow-up also varied between cohorts; overall it extended from 1939 to 1992 (488,482 person-years at report [PYAR]). A total of 4.4 percent (970 workers) were lost to follow-up. Exposure information varied between cohorts, but in general, exposures were reconstructed from job records. The exposed workers were aggregated into five groups: main production, maintenance, other exposed jobs, unspecified tasks, and sprayers. Based on these categories and information on production processes and the composition of materials used, exposed workers were further classified into three categories: exposed to TCDD or higher chlorinated dioxins; unexposed to the same; and unknown exposure to the same.
Analysis was performed by calculating SMRs and 95% CI, using the World Health Organization mortality data bank to calculate national mortality rates by sex, age (five-year intervals), and calendar period (five year). Within-cohort analysis was also performed using Poisson regression adjusting for time since first exposure, duration of exposure, and employment status. MM mortality was nonsignificantly elevated (SMR = 1.3, CI 0.8-2.1, based on 17 deaths) among the group including all workers exposed to any phenoxy herbicide or chlorophenol. When this group was divided into groups exposed and unexposed to TCDD or
higher chlorinated dioxins, the TCDD-exposed group had a slightly lower risk (SMR = 1.2, CI 0.6-2.3) than the unexposed group (SMR = 1.6, CI 0.7-3.1), although neither achieved statistical significance. More detailed analysis by exposure variables such as duration and time of first exposure was not conducted for multiple myeloma.
Although the study includes large numbers of workers who were likely to be exposed at levels substantially higher than the general population, the lack of information available about actual exposures limits the investigator's ability to examine exposure-response relationships within the cohort. In addition, the inclusion of workers in the exposed group based on ever having worked in a job considered exposed makes it impossible to distinguish heavily exposed workers from those with very minor exposures.
Becher et al. (1996) examined cancer mortality among workers in four German facilities that produced phenoxy herbicides and chlorophenols. These cohorts were also part of the IARC study described above. The population included workers who had at least one month of employment, resulting in a cohort consisting of 2,479 male workers. The cohort was assembled from four plants, and analysis was conducted on the total cohort divided into four subcohorts corresponding to each plant considered separately. The period of follow-up varied between plants, and 100 workers were lost to follow-up. The nature of chemical production varied substantially between plants and over time; some facilities synthesized and formulated a wide range of phenoxy herbicides and chlorophenols (subcohorts III and Iv), whereas others produced primarily 2,4,5-T and/or 2,4,5-TCP (subcohorts I and II). SMRs and 95% CI were calculated using West German mortality rates by five-year age and calendar intervals. Cox regression was performed to evaluate the effect of smoking in the one subcohort for which smoking information was available. Each subcohort was analyzed separately since the exposure pattern was judged to be characteristic for each facility. Based on production information and limited blood dioxin measurements, subcohorts I and II are supposed to have higher TCDD exposures than subcohorts III and IV. Of the four subcohorts, only group I had at least one observed death from MM. In this group, 3 deaths were observed versus 0.6 expected for a SMR of 5.4 (CI 1.1-15.9).
Agricultural Workers Cancer mortality among a cohort of rice growers in northern Italy was investigated by Gambini et al. (1997). Using a set of registered farm owners consisting of 1,493 males who worked on farms from 1957 to 1992, they examined the cause of death for 958 subjects and compared this with expected numbers calculated from national rates. No direct exposure information was available, so employment on the farm was used as a surrogate for exposure to the range of phenoxy herbicides employed during the study period. Cancer mortality was evaluated for MM and observed and expected deaths were 0 and 1.9 for the overall cohort (SMR = 0.0, CI 0.0-2.0). Although the study population is
small, it does describe the experience of a cohort with good follow-up (99 percent) and long latency (37 percent of deaths observed beyond the age of 80). It is limited by crude exposure assessment, however, and the degree to which study subjects were actually exposed to phenoxy herbicides can not be established with any certainty.
Environmental Studies
Bertazzi et al. (1997) continued the follow-up of people environmentally exposed to TCDD in Seveso, Italy. The events that led to the exposure and the methods used to study this population have been fully described in earlier reports. This report updates the population after 15 years' follow-up. Death from multiple myeloma showed significant increases in zone B for women (RR = 6.6, CI = 1.8-16.8), but not for men. There were no increases in zones A and R for either sex.
Vietnam Veteran Studies
Recent veteran studies are limited because of the small number of MM deaths in the few analyses that have been conducted. The only reported data for MM for U.S. veterans come from the Watanabe and Kang study (1996), showing no increase among either Army or Marine veterans (SMR = 0.9 and 0.6, respectively). The Australian Vietnam veteran study (Crane et al., 1997a) found an SMR of 0.6 (CI 0.2-1.4) based on six deaths. A second study examining the mortality experience of conscripted Australian veterans relative to military personnel who did not serve in the conflict reported no MM deaths among Vietnam veterans and one in the comparison population between 1982 and 1994 (Crane et al., 1997b).
Synthesis
The low incidence of MM among the various cohorts that have been studied makes it difficult to draw firm conclusions from every study. Nonetheless, the consistent pattern of elevated risk appears to continue for the few reports evaluating myeloma cases discussed in this report. Two additional production studies and one subset of the Seveso cohort appear to be at increased risk, with SMRs or RRs greater than 1.0 reported. New studies of Vietnam veterans reported lower than expected MM mortality.
The committee has determined that the evidence for this association is limited/suggestive, because individuals in the existing studies (mostly farmers) have, by nature of their occupation, probably been exposed to a range of potentially carcinogenic agents other than herbicides and TCDD. MM—like NHL and HD, for which there is stronger epidemiologic evidence of an association—is derived from lymphoreticular cells, which adds to the biologic plausibility of an associa-
TABLE 7-27 Selected Epidemiologic Studies—Multiple Myeloma
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
OCCUPATIONAL |
|||
New Studies |
|||
Gambini et al., 1997 |
Italian rice growers |
0 |
— |
Kogevinas et al., 1997 |
IARC cohort |
|
|
|
Workers exposed to TCDD (or higher chlorinated dioxins) |
|
1.2 (0.6-2.3) |
|
Workers not exposed to TCDD (or higher chlorinated dioxins) |
|
1.6 (0.7-3.1) |
|
Workers exposed to any phenoxy herbicide or chlorophenol |
17 |
1.3 (0.8-2.1) |
Becher et al., 1996 |
German chemical production workers—Plant I |
3 |
5.4 (1.1-15.9) |
Studies reviewed in Update 1996 |
|||
Asp et al., 1994 |
Finnish herbicide applicators |
3 |
2.6 (0.5-7.7) |
Dean, 1994 |
Irish farmers and farm workers |
170 |
1.0 |
Blair et al., 1993 |
U.S. farmers in 23 states |
|
|
|
White males |
413 |
1.2 (1.0-1.3) |
|
White females |
14 |
1.8 (1.0-3.0) |
|
Nonwhite males |
51 |
0.9 (0.7-1.2) |
|
Nonwhite females |
11 |
1.1 (0.6-2.0) |
|
Farmers in central U.S. states |
|
|
|
White males |
233 |
1.2 |
|
White females |
12 |
2.6 |
Lynge, 1993 |
Danish production workers |
|
|
|
Male |
0 |
— |
|
Female |
2 |
12.5 (1.5-45.1) |
Semenciw et al., 1994 |
Farmers in Canadian prairie provinces |
160 |
0.8 (0.7-1.0) |
Brown et al., 1993 |
Iowa male users of pesticides or herbicides |
111 |
1.2 (0.8-1.7) |
Zahm et al., 1992 |
Eastern Nebraska users of herbicides |
|
|
|
Male |
8 |
0.6 (0.2-1.7) |
|
Female |
10 |
2.3 (0.8-7.0) |
|
Eastern Nebraska users of insecticides |
|
|
|
Male |
11 |
0.6 (0.2-1.4) |
|
Female |
21 |
2.8 (1.1-7.3) |
Studies reviewed in VAO |
|||
Eriksson and Karlsson, 1992 |
Residents of northern Sweden |
20 |
2.2 (1.0-5.7) |
Swaen et al., 1992 |
Dutch herbicide applicators |
3 |
8.2 (1.6-23.8) |
Fingerhut et al., 1991 |
NIOSH cohort |
5 |
1.6 (0.5-3.9) |
|
20 years latency, 1+ years exposure |
3 |
2.6 (0.5-7.7) |
Saracci et al., 1991 |
IARC cohort |
4 |
0.7 (0.2-1.8) |
Alavanja et al., 1989 |
USDA forest/soil conservationists |
|
1.3 (0.5-2.8) |
Boffetta et al., 1989 |
ACS Prevention Study II subjects |
12 |
2.1 (1.0-4.2) |
|
Farmers using herbicides or pesticides |
8 |
4.3 (1.7-10.9) |
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
LaVecchia et al., 1989 |
Residents of the Milan, Italy, area |
|
|
|
Agricultural employment |
|
2.0 (1.1-3.5) |
Morris et al.. 1986 |
Residents of four SEER areas |
|
2.9 (1.5-5.5) |
Pearce et al., 1986 |
Male residents of New Zealand |
|
|
|
Use of agricultural spray |
16 |
1.3 (0.7-2.5) |
|
Likely sprayed 2,4,5-T |
14 |
1.6 (0.8-3.1) |
Cantor and Blair, 1984 |
Wisconsin residents |
|
|
|
Farmers in counties with highest herbicide usage |
|
1.4 (0.8-2.3) |
Burmeister et al., 1983 |
Iowa residents (farmers in counties with highest herbicide usage) |
|
|
|
Born 1890-1900 |
|
2.7 (p < .05) |
|
Born after 1900 |
|
2.4 (p < .05) |
Riihimaki et al., 1983 |
Finnish herbicide applicators |
1 |
2.5 (0.3-14.0) |
ENVIRONMENTAL |
|||
New Studies |
|||
Bertazzi et al., 1997 |
Seveso residents |
|
|
|
zone B—female |
|
6.6 (1.8-16.8) |
Studies reviewed in Update 1996 |
|||
Bertazzi et al., 1993 |
Seveso residents |
|
|
|
zone A—male |
0 |
— |
|
zone A—female |
0 |
— |
|
zone B—male |
2 |
3.2 (0.8-13.3) |
|
zone B—female |
2 |
5.3 (1.2-22.6) |
|
zone R—male |
1 |
0.2 (0.0-1.6) |
|
zone R—female |
2 |
0.6 (0.2-2.8) |
Studies reviewed in VAO |
|||
Pesatori et al., 1992 |
Seveso residents |
|
|
|
Males—zones A and B |
2 |
2.7 (0.6-11.3) |
|
Female—zones A and B |
2 |
4.4 (1.0-18.7) |
|
Males—zone R |
1 |
0.2 (0.0-1.5) |
|
Females—zone R |
3 |
0.9 (0.3-3.1) |
VIETNAM VETERANS |
|||
New Studies |
|||
Crane et al., 1997a |
Australian military veterans |
6 |
0.6 (0.2-1.4) |
Crane et al., 1997a |
Australian military veterans |
0 |
|
Watanabe and Kang, 1996 |
Army Vietnam veterans |
|
0.9 |
|
Marine Vietnam veterans |
|
0.6 |
Studies reviewed in VAO |
|||
Breslin et al., 1988 |
Army Vietnam veterans |
|
0.8 (0.2-2.5) |
|
Marine Vietnam veterans |
2 |
0.5 (0.0-17.1) |
a Given when available. |
tion. The new data available on MM do not change the committee's earlier view that there is a limited/suggestive association between exposure to herbicides and multiple myeloma.
Conclusions
Strength of Evidence in Epidemiologic Studies
There is limited/suggestive evidence of an association between exposure to the herbicides (2,4-D, 2,4,5-T and its contaminant TCDD, cacodylic acid, and picloram) and multiple myeloma. The evidence regarding association is drawn from occupational and other studies in which subjects were exposed to a variety of herbicides and herbicide components.
Biologic Plausibility
A thorough discussion of biologic plausibility with respect to exposure to TCDD or herbicides and multiple myeloma is contained in Chapter 3; a summary is presented in the conclusion to this chapter.
LEUKEMIA
Background
There are four primary types of leukemia (ICD·9 202.4, 203.1, 204.0-204.9, 205.0-205.9, 206.0-206.9, 207.0-207.2, 207.8, 208.0-208.9): the acute and chronic forms of lymphocytic leukemia and the acute and chronic forms of myeloid (or granulocytic) leukemia. According to American Cancer Society estimates, 16,100 men and 12,600 women will be diagnosed with some form of the disease in the United States in 1998, and 12,000 men and 9,600 women will die from it (ACS, 1998). Collectively, leukemias are expected to account for slightly more than 2 percent of all new cancer diagnoses and nearly 4 percent of cancer deaths in 1998.
The different forms of leukemia have different patterns of incidence and, in some cases, different risk factors.
Acute lymphocytic leukemia (ALL) is a disease of the young and of individuals older than 70 years of age, and plays a rather small role in the age groups that characterize most Vietnam veterans. The lifetime incidence of ALL is slightly higher in whites than in African Americans and in males than females. Exposure to high doses of ionizing radiation is a known risk factor for this form of leukemia; evidence for other factors is inconsistent
Acute myeloid leukemia (AML) is the most common leukemia among adults—incidence increasing steadily with age for individuals older than 40. In
the Vietnam veteran age groups, AML accounts for roughly one out of every four leukemias in men and one out of three in women. Overall, this leukemia is slightly more common in males than females. White males have a higher incidence that white females; the lifetime incidence in African-American males and females is roughly equal. Risk factors associated with an increased risk of AML include high doses of ionizing radiation, occupational exposure to benzene, and some medications used in cancer chemotherapy (melphalan, for example). Genetic disorders including Fanconi's anemia and Down's syndrome are associated with an increased risk of AML, and tobacco smoking has been suggested as a risk factor.
Average Annual Cancer Incidence (per 100,000 individuals) in the United Statesa Leukemias |
|||||||||
|
45-49 years of age |
50-54 years of age |
55-59 years of age |
||||||
|
all races |
white |
black |
all races |
white |
black |
all races |
white |
black |
All leukemias |
|||||||||
males |
8 |
8 |
7 |
13 |
13 |
12 |
20 |
21 |
20 |
females |
6 |
5 |
7 |
8 |
8 |
5 |
11 |
11 |
9 |
Acute lymphocytic leukemia |
|||||||||
males |
0.6 |
0.7 |
b |
0.4 |
0.5 |
1.2 |
1.3 |
||
females |
0.6 |
0.6 |
0.6 |
0.5 |
0.5 |
0.5 |
0.3 |
0.3 |
0.4 |
Chronic lymphocytic leukemia |
|||||||||
males |
1.0 |
0.9 |
1.1 |
2.3 |
2.3 |
2.4 |
4.8 |
4.8 |
5.1 |
females |
0.5 |
0.5 |
0.2 |
1.1 |
1.0 |
1.1 |
2.1 |
2.2 |
0.7 |
Acute myeloid leukemia |
|||||||||
males |
2.0 |
2.0 |
2.2 |
2.5 |
2.6 |
2.6 |
4.7 |
4.6 |
4.8 |
females |
2.2 |
2.1 |
2.7 |
2.8 |
2.9 |
1.8 |
3.4 |
3.2 |
3.5 |
Chronic myeloid leukemia |
|||||||||
males |
1.7 |
1.5 |
2.5 |
2.1 |
2.3 |
1.7 |
2.7 |
2.6 |
3.2 |
females |
1.0 |
0.9 |
1.6 |
1.7 |
1.6 |
2.2 |
1.8 |
1.8 |
2.2 |
All other leukemias |
|||||||||
males |
1.8 |
2.1 |
0.3 |
2.5 |
2.6 |
2.6 |
3.8 |
3.9 |
2.7 |
females |
1.0 |
0.8 |
1.1 |
1.2 |
1.3 |
0.4 |
2.0 |
2.0 |
0.4 |
a SEER nine standard registries crude, age-specific rate, 1990-1994. b Insufficient data to provide a meaningful incidence rate. c Includes leukemic reticuloendotheliosis (hairy cell), plasma cell, monocytic, and acute and chronic erythremia and erythroleukemia. |
Chronic lymphocytic leukemia (CLL) is the most common of the four primary types of leukemia for men. It is largely a disease of individuals older than 40, and incidence doubles every five years for individuals in the three age groups that characterize most Vietnam veterans. Over a lifetime, CLL is nearly twice as common in whites than African Americans and more common in men than women. Some occupational groups, notably farmers, appear to have a higher
incidence of CLL than would otherwise be expected. A family history of the disease and a compromised immune system are among additional suspected risk factors. Unlike the other primary forms of leukemia, exposure to ionizing radiation does not appear to be associated with increased incidence of CLL.
The incidence of chronic myeloid leukemia (CML) increases steadily with age for individuals over 30. Lifetime incidence is roughly equal in whites and African Americans and is slightly higher in males than females. For individuals in the age groups that characterize most Vietnam veterans, CML accounts for approximately one in five leukemias. CML is associated with an acquired chromosomal abnormality known as the ''Philadelphia chromosome." Exposure to high doses of ionizing radiation is a known risk factor for this abnormality; other factors are under study.
Little is known about the risk factors associated with other forms of leukemia. However, two human retroviruses have been linked to human leukemias: HTLV-1 appears to cause adult T-cell leukemia or lymphoma, whereas the data linking HTLV-2 to hairy cell leukemia are less definitive.
Summary of VAO and Update 1996
The epidemiologic evidence for an association between exposure to herbicides and leukemia in VAO came primarily from studies of farmers and residents of Seveso, Italy. The observed overall RR for leukemia mortality and incidence in Seveso was elevated, but not significantly. The increase was significant, however, for cases who were in the most highly exposed zone and died five to ten years after the accident. A number of studies of farmers also showed a consistently elevated risk of leukemia, but these results are not necessarily due to herbicide use, because confounding exposures were not controlled for adequately in the analysis. Also, when farmers are stratified by suspected use of herbicide, the incidence of leukemia is generally not elevated. Some studies of chemical workers found an increased risk of leukemia, but the number of cases was small.
The available data on Vietnam veterans are generally not conclusive, because exposure data were inadequate for the cohort being studied. Small sample sizes weaken the studies of the Ranch Hand or Chemical Corps veterans; therefore, excess risks were not likely to be detected.
Since no study has differentiated adequately between exposure solely either to herbicides or to TCDD, or demonstrated a dose-response for any subtype of leukemia, it is not possible to attribute any symptom or subtype of leukemia to result of exposure.
Update 1996 described a number of new studies, each with relatively small numbers of leukemia cases. Three studies of agricultural workers described in the update (Asp et al., 1994; Dean, 1994; Semenciw et al., 1994) described in Update 1996 failed to show any elevation in chronic or acute leukemia risk, as did one production worker study (Kogevinas et al., 1993).
The U.S. PCMR study of farmers in 23 states (Blair et al., 1993) showed an overall significant increase in leukemia deaths (PCMR = 1.3, CI 1.2-1.4), as did the Seveso update for men in zone B only (Bertazzi et al., 1993). However, the single Vietnam veteran study of this period showed no increased leukemia risk (Visintainer et al., 1995).
For Update 1996, the committee concluded that there was inadequate or insufficient evidence to determine whether an association exists between exposure to herbicides and leukemia. The updated studies did not affect the original conclusion of VAO.
Update of the Scientific Literature
Occupational Studies
Production Workers In an update and expansion of the IARC cohort study, Kogevinas et al. (1997) examined cancer mortality in a cohort of 26,615 male and female workers engaged in the production or application of phenoxy herbicides. These cohorts were assembled from 12 countries, drawn from national studies that followed the same core protocol developed jointly by the participants and coordinated by IARC. Of the total study population, 21,863 (20,851 men and 1,012 women) were classified as exposed to phenoxy herbicides or chlorophenols based on individual job records and company exposure questionnaires; 4,160 were unexposed, and 592 were classified as unknown exposure status. The great majority of workers were considered exposed if they had ever worked in production or spraying of phenoxy herbicides or chlorophenols (four cohorts were exceptions, with minimum employment period of 1 to 12 months). The period of follow-up also varied between cohorts; overall it extended from 1939 to 1992 (488,482 PYAR). Overall, 4.4 percent (970 workers) were lost to follow-up. Exposure information varied between cohorts, but in general, exposures were reconstructed from job records.
No significant risk of death from leukemia was observed among the group including all workers exposed to any phenoxy herbicide or chlorophenol. When this group was divided into groups exposed and unexposed to TCDD or higher chlorinated dioxins, the result was unchanged. More detailed analysis by exposure variables such as duration and time since first exposure was not conducted for leukemias.
Becher et al. (1996) examined cancer mortality among workers in four German facilities that produced phenoxy herbicides and chlorophenols. The population included workers who had a least one month of employment, resulting in a cohort consisting of 2,479 male workers. The cohort was assembled from four plants, and the analysis as conducted on the total cohort divided into four subcohorts corresponding to each plant considered separately. The period of follow-up varied between plants, and 100 workers were lost to follow-up. The nature of chemical production varied substantially between plants and over time; some
facilities synthesized and formulated a wide range of phenoxy herbicides and chlorophenols (subcohorts III and IV), whereas others produced primarily 2,4,5-T and/or 2,4,5-TCP (subcohorts I and II). SMRs and 95% CI were calculated using West German mortality rates by five-year age and calendar intervals. Cox regression was performed to evaluate the effect of smoking in the one subcohort for which smoking information was available. Each subcohort was analyzed separately since the exposure pattern was judged to be characteristic for each facility. Based on production information and limited blood dioxin measurements, subcohorts I and II were supposed to have higher TCDD exposures than subcohorts III and IV. Of the four subcohorts, only group I had at least one observed or expected death from leukemia. An SMR of 1.8 (CI 0.5-4.1) was reported, based on four observed cases.
Ramlow et al. (1996) examined mortality in a cohort of workers exposed to pentachlorophenol, as part of a larger study of Dow chemical manufacturing workers exposed to the higher chlorinated dioxins. The study cohort was assembled from company records, starting with a cohort of 2,192 workers ever employed in a department with potential PCDD exposure between 1937 and 1980. From this cohort, 770 workers were identified who were considered to have potential PCP exposure based on work history records. Exposure to PCP was assessed using historical industrial hygiene and process data, resulting in a strategy for ranking jobs by exposure intensity based on a scale of 1 to 3. Exposure to PCDD was analyzed using the process described by Ott et al. (1987), in which semiquantitative, logarithmic exposure intensity scores ranging from 1 to 4 for TCDD and 0 to 2 for H/OCDD were assigned to each job title. Cumulative exposure indices for PCP and dioxin were calculated using these assigned scores. In the study analysis, the U.S. white male death rates (five-year age and calendar specific) and the death rates of non-PCP and PCDD male Dow Michigan employees for 1940 to 1989 were both used as reference values to calculate expected deaths. Leukemia deaths showed no increase (SMR = 1.0, CI 0.1-3.6) in this study. Calculation of SMRs with exposure lagged by 15 years using both the U.S. and the Dow referent populations also found no significant excess mortality for leukemia. Leukemia was not included in the more detailed analysis by the four categories of cumulative exposure.
Agricultural Workers Cancer mortality among a cohort of rice growers in northern Italy was investigated by Gambini et al. (1997). Using a set of registered farm owners consisting of 1,493 males who worked on farms from 1957 to 1992, they examined the cause of death for 958 subjects and compared this with expected numbers calculated from national rates. No direct exposure information was available, so employment on the farm was used as a surrogate for exposure to the range of phenoxy herbicides employed during the study period. A nonsignificant decrease in leukemia mortality (SMR = 0.6, CI 0.2-1.7) was reported. Leukemia was not included in the more detailed analysis with stratification by age at death and duration of exposure (employment as a farmer).
Waterhouse et al. (1996) conducted a survey of total and site-specific cancer incidence among 70,16 male and female adult from 1959 to 1987, and compared the observed number with the expected number, based on age-, sex-, race-, calendar-, and site-specific cancer incidence rates reported by the Connecticut tumor registry. Based on the results of this survey, a hypothesis was advanced concerning the potential risks of exposure to insecticides and herbicides. This was pursued by analyzing for each county in Michigan the comparative annual number of acres and the percentage of acreage treated with agricultural chemicals in 1978 and for 1982-1987. Finally, because of the availability of information on lifestyle risk factors that had been collected in the 1960s on all participants, a nested case-control study was implemented. A significantly increased risk for males and females combined was demonstrated in the incidence of lymphopoietic neoplasms, namely, NHL, HD, and CLL; the combined SIR was 1.4 (95% CI 1.00-1.9; p = .03). Comparison of the Tecumseh, Michigan, cohort (ranked highest in pesticide exposure) with all sites combined was not significantly different from the expected incidence in females (SIR = 1.0, CI 0.9-1.1) and was decreased by more than 10 percent in males (SIR = 0.9, CI 0.8-1.0). However, the SIR for NHL in females was significantly elevated (SIR = 1.9; CI 1.1-3.1; p = .02); the trend for increased risk of lymphoma and leukemia was also evident in males. In the nested case-control study based on risk factor information documented prior to diagnosis, the RR of a family history of lymphoma, leukemia, or MM was significantly increased among patients with lymphoproliferative neoplasms (OR = 3.8, CI 1.5-9.8; p = .005).
Finally, Amadori et al. (1995) conducted a population based case-control study in a highly agricultural area in the northeast of Italy to evaluate the association between farming and animal breeding and the risk of developing NHL and CLL. Occupational histories and other data were collected by personal interview of individuals diagnosed with NHL (N = 164), and CLL (N = 23) between 1988 and 1990, and 977 controls. Estimates of ORs for occupational variables were calculated, after adjustment for sex, age, altitude of municipality, first-degree familiarity, and previous herpes zoster infection. From analysis of the more frequent occupational categories, no occupation had a significantly high risk. When the two job titles of farmers only and farmer-breeders, who are also involved in animal breeding are classified within the extremely varied occupations of animal breeding, agriculture, or fishing, a high risk for NHL and CLL is seen in the farmer-breeders (OR = 1.8, 95% CI 1.2-2.6). Analyses according to histological type show that risks are concentrated in CLL and in low-grade NHL. No effect or trend by period at work or duration of employment in farming and animal breeding was found.
Environmental Studies
Bertazzi et al. (1997) continued the follow-up of people environmentally exposed to TCDD in Seveso, Italy. The events that led to the exposure and the
methods used to study this population have been fully described in earlier reports. This report updates the population after 15-years follow-up. There were no cases of leukemia in zone A. Among men in zone B, there were seven cases of leukemia (RR = 3.1, CI 1.4-6.4). Among women in zone B, there was also one case of leukemia (RR = 0.6, CI 0.0-3.1). In zone R, where residents had much lower potential exposure to TCDD than in zones A and B, there were 12 cases of leukemia in men (RR = 0.8, CI 0.4-1.4) and 12 cases in women (RR = 0.9, CI 0.4-1.5).
Vietnam Veteran Studies
In a comparison of mortality between Army Chemical Corp Vietnam and non-Vietnam veterans, Dalager and Kang (1997) reported that a nonsignificant excess of deaths from leukemia among Vietnam veterans. The study compared 2,872 Vietnam veterans with 2,737 non-Vietnam veterans (all of whom served in Chemical Corps specialties). All study subjects served at least 18 months' active duty between 1965 and 1973, and vital status ascertainment was complete for both groups. A nonsignificant increase in death for Vietnam veterans (RR = 1.0, CI 0.1-3.8, two observed deaths) and a nonsignificant decrease for non-Vietnam veterans (RR = 0.7, CI 0.4-1.0, one death observed) from leukemia was found compared to general U.S. population rates. When Vietnam and non-Vietnam cohorts were compared directly, the crude rate ratio of leukemia death was 1.9 (Vietnam to non-Vietnam veterans). Direct exposure information on the two cohorts was not available, and the presumption that Vietnam veterans had potentially higher levels of dioxin exposure because of their duties involving Agent Orange and other dioxin-contaminated herbicides (compared to the non-Vietnam Chemical Corps veterans) has not been verified. Very similar data have been reported for the Australian Vietnam veteran cohort (Crane et al., 1997b). Nonsignificant decreases in death due to leukemia for Vietnam veterans (SMR = 0.5, CI 0.1-3.0) and for nonveteran deaths (SMR = 0.7, CI 0.1-3.0) were found compared to general Australian population rates. When Vietnam and non-Vietnam cohorts were compared directly, the crude rate ratio of leukemia death was 0.6 (veterans to nonveterans). Again, direct exposure information for dioxin is not available, and the presumption of higher dioxin exposure for Australian Vietnam veterans has not been verified.
Synthesis
As in VAO and Update 1996, the studies reviewed in this report found a small number of cases of leukemia, and it is apparent that the risks associated with herbicide exposure are fairly evenly distributed around the null, with relatively narrow confidence intervals. Some data on agricultural workers suggest an increased risk for all hematopoietic neoplasms (including NHL and MM, as well
as CLL), and very limited data from a small subset of the Seveso cohort suggesting an increased leukemia risk. Biologic plausibility would suggest an association of risk for hematopoietic and lymphoreticular malignancies, as would the etiologic similarity of CLL and NHL; however, the overall incidence and small number of positive studies are inadequate to change the previous classification.
Conclusion
Strength of Evidence in Epidemiologic Studies
There is inadequate or insufficient evidence to determine whether an association exists between exposure to the herbicides (2,4-D, 2,4,5-T and its contaminant TCDD, cacodylic acid, and picloram) and leukemia. The evidence regarding association is drawn from occupational and other studies in which subjects were exposed to a variety of herbicides and herbicide components.
Biologic Plausibility
A thorough discussion of biologic plausibility with respect to exposure to TCDD or herbicides and leukemia is contained in Chapter 3; a summary is presented in the conclusion to this chapter.
SUMMARY
Based on the occupational, environmental, and veteran studies reviewed, the committee has reached one of four standard conclusions about the strength of the evidence regarding association between an exposure to herbicides and/or TCDD and each of the cancers studied. As explained in Chapter 4, these distinctions reflect the committee's judgment that if an association between exposure and an outcome were "real," it would be found in a large, well-designed epidemiologic study in which exposure to herbicides or dioxin was sufficiently high, well characterized, and appropriately measured on an individual basis. Consistent with the charge to the committee by the Secretary of Veterans Affairs in Public Law 102-4 and with accepted standards for scientific reviews, the distinctions between these standard conclusions are based on statistical association, not on causality. The committee used the same criteria to categorize diseases by the strength of the evidence as were used in VAO and Update 1996.
Health Outcomes with Sufficient Evidence of an Association
In VAO and Update 1996, the committee found sufficient evidence of an association between exposure to herbicides and/or TCDD and three cancers: soft-tissue sarcoma, non-Hodgkin's lymphoma, and Hodgkin's disease. The scientific
TABLE 7-28 Selected Epidemiologic Studies—Leukemia
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
OCCUPATIONAL |
|||
New Studies |
|||
Gambini et al., 1997 |
Italian rice growers |
|
0.6 (0.2-1.7) |
Kogevinas et al., 1997 |
IARC cohort |
34 |
1.0 (0.7-1.4) |
Becher et al., 1996 |
German chemical production workers |
|
1.8 (0.5-4.1) |
Ramlow et al., 1996 |
Pentachlorophenol production workers |
|
1.0 (0.1-3.6) |
Waterhouse et al., 1996 |
Residents of Tecumseh, Michigan |
|
1.4 (1.0-1.9) |
Amadori et al., 1995 |
Italian farming and animal-breeding workers |
|
1.8 (1.2-2.6) |
Studies reviewed in Update 1996 |
|||
Asp et al., 1994 |
Finnish herbicide applicators |
2 |
— |
Semenciw et al., 1994 |
Farmers in Canadian prairie provinces |
357 |
0.9 (0.8-1.0) |
Blair et al., 1993 |
U.S. farmers in 23 states |
1,072 |
1.3 (1.2-1.4) |
Kogevinas et al., 1993 |
Female herbicide spraying and production workers |
1 |
— |
Studies reviewed in VAO |
|||
Bueno de Mesquita et al., 1993 |
Dutch production workers |
|
|
|
Workers exposed to phenoxy herbicides |
2 |
2.2 (0.3-7.9) |
Hansen et al., 1992 |
Danish gardeners |
|
|
|
All gardeners—CLL |
6 |
2.5 (0.9-5.5) |
|
All gardeners—all other types of leukemia |
3 |
1.2 (0.3-3.6) |
|
Male gardeners—CLL |
6 |
2.8 (1.0-6.0) |
|
Male gardeners—all other types of leukemia |
3 |
1.4 (0.3-4.2) |
Ronco et al., 1992 |
Danish and Italian farm workers |
|
|
|
Danish self-employed farmers |
|
0.9 |
|
Danish male farmers |
|
1.0 |
|
Italian self-employed farmers |
|
0.7 |
|
Italian male farmers |
|
0.9 |
Fingerhut et al., 1991 |
U.S. chemical workers |
6 |
0.7 (0.2-1.5) |
Saracci et al., 1991 |
Chemical workers |
|
|
|
Exposed |
18 |
|
|
Probably exposed |
0 |
— |
|
Nonexposed |
3 |
0.9 (0.2-2.6) |
|
Unknown exposure |
0 |
— |
Brown et al., 1990 |
Residents of Iowa and Minnesota |
|
|
|
All types of leukemia, ever farmed |
|
1.2 (1.0-1.5) |
|
CLL, ever farmed |
|
1.4 (1.1-1.9) |
|
All types of leukemia, any herbicide use |
|
1.2 (0.9-1.6) |
|
CLL, any herbicide use |
|
1.4 (1.0-2.0) |
|
Herbicide users, phenoxy acid use |
|
1.2 (0.9-1.6) |
|
All types of leukemia, 2,4-D use |
|
1.2 (0.9-1.6) |
|
All types of leukemia, 2,4,5-T use |
|
1.3 (0.7-2.2) |
Reference |
Study Population |
Exposed Casesa |
Estimated Risk (95% CI)a |
Wigle et al., 1990 |
Saskatchewan farmers |
138 |
0.9 (0.7-1.0) |
Zober et al., 1990 |
BASF production workers |
|
|
|
Second additional cohort |
1 |
5.2 (0.4-63.1) |
Alavanja et al., 1988 |
USDA agricultural extension agents |
|
1.9 (1.0-3.5) |
Bond et al., 1988 |
Dow workers with chloracne |
2 |
3.6 (0.4-13.0) |
Blair and White, 1985 |
Residents of Nebraska |
|
|
|
All cases, all leukemia—farming |
|
1.3 |
Burmeister et al., 1982 |
Residents of Iowa |
|
|
|
CLL in white, male farmers |
|
1.9 (1.2-3.1) |
ENVIRONMENTAL |
|||
New Studies |
|||
Bertazzi et al., 1997 |
Seveso residents |
|
|
|
Males—zone B |
7 |
3.1 (1.4-6.4) |
|
Females—zone B |
1 |
0.6 (0.0-3.1) |
Studies reviewed in Update 1996 |
|||
Bertazzi et al., 1993 |
Seveso residents |
|
|
|
Males—zone B |
2 |
1.6 (0.4-6.5) |
|
Females—zone B |
2 |
1.8 (0.4-7.3) |
Studies reviewed in VAO |
|||
Bertazzi et al., 1993 |
Seveso residents |
|
|
|
Males—zones A, B, and R |
4 |
2.1 (0.7-6.9) |
|
Females—zones A, B, and R |
1 |
2.5 (0.2-27.0) |
VIETNAM VETERANS |
|||
New Studies |
|||
Dalager and Kang, 1997 |
Army Chemical Corps veterans |
|
1.0 (0.1-3.8) |
Crane et al., 1997a |
Australian military veterans |
|
0.5 (0.1-3.0) |
Studies reviewed in Update 1996 |
|||
Visintainer et al., 1995 |
Michigan Vietnam veterans |
30 |
1.0 (0.7-1.5) |
a Given when available. |
literature continues to support the classification of these three cancers in the category of sufficient evidence. Based on the literature, there are no additional cancers that satisfy the criteria necessary for this category.
For diseases in this category, a positive association between herbicides and the outcome must be observed in studies in which chance, bias, and confounding can be ruled out with reasonable confidence. The committee also regarded evidence from several small studies that are free from bias and confounding, and show an association that is consistent in magnitude and direction, as sufficient evidence for an association.
Health Outcomes with Limited/Suggestive Evidence of Association
In VAO and Update 1996, the committee found limited/suggestive evidence of an association between herbicide or dioxin exposure and the following cancers: larynx, lung, bronchus (trachea), prostate, and multiple myeloma. The scientific literature continues to support the classification of these diseases in the category of limited/suggestive evidence. Based on the literature, there are no additional cancers that satisfy the criteria necessary for this category.
For outcomes in this category, the evidence must be suggestive of an association between herbicides and the outcome, but may be limited because chance, bias, or confounding could not be ruled out with confidence. Typically, at least one high-quality study indicates a positive association, but the results of other studies may be inconsistent.
Health Outcomes with Inadequate/Insufficient Evidence To Determine Whether an Association Exists
The scientific data for many of the cancers reviewed by the committee were inadequate or insufficient to determine whether an association exists. For these cancers, the available studies are of insufficient quality, consistency, or statistical power to permit a conclusion regarding the presence or absence of an association. For example, studies fail to control for confounding or have inadequate exposure assessment. This category includes hepatobiliary cancers (cancers of the liver and intrahepatic bile duct), nasal and nasopharyngeal cancer, bone cancer, skin cancers (including basal cell carcinoma, squamous cell carcinoma, and nonmelanocytic skin cancers), breast cancer, cancers of the female reproductive system (including cervix, endometrium, and ovaries), testicular cancer, urinary bladder cancer, renal cancer (cancers of the kidney and renal pelvis), and leukemias.
Based on an evaluation of all the epidemiologic evidence, including studies published since the release of Update 1996, the committee felt that the previous conclusion of limited/suggestive evidence of no association between exposure to the herbicides (2,4-D, 2,4,5-T and its contaminant TCDD, cacodylic acid, and picloram) and urinary bladder cancer should be changed to "inadequate/insufficient evidence to determine whether an association exists." Although there is no evidence that exposure to herbicides or dioxin is related to this cancer, RRs in some of the largest cohorts tended to be greater than one, weakening the committee's prior conclusion that there was positive evidence of no relationship. Coexposures to TCDD and a variety of known bladder carcinogens makes it very difficult to isolate the possible additional effect of herbicides, although little total effect was seen.
Health Outcomes with Limited/Suggestive Evidence of No Association
In VAO and Update 1996, the committee found a sufficient number and variety of well-designed studies to conclude that there is limited/suggestive evi-
dence of no association between a small group of cancers and exposure to TCDD or herbicides. This group includes gastrointestinal tumors (colon, rectal, stomach, and pancreatic) and brain tumors. The most recent scientific evidence continues to support the classification of such cancers in this category. Based on an evaluation of the whole of the scientific literature, there are no additional cancers that satisfy the criteria necessary for this category.
As noted above, the committee responsible for this report felt that the previous classification of urinary bladder cancer in this category should be changed to "inadequate/insufficient evidence to determine whether an association exists."
For outcomes in this category, several adequate studies covering the full range of levels of exposure that human beings are known to encounter are mutually consistent in not showing a positive association between exposure to herbicides and the outcome at any level of exposure. These studies have relatively narrow confidence intervals. A conclusion of "no association" is inevitably limited to the conditions, level of exposure, and length of observation covered by the available studies. In addition, the possibility of a very small elevation in risk at the levels of exposure studied can never be excluded.
Biologic Plausibility
Chapter 3 details the committee's evaluation of data from studies with animals and cells regarding the biologic plausibility of a connection between exposure to dioxin or herbicides and various forms of cancer. This section summarizes that evidence. Some of the preceding discussions of cancer outcomes include references to specific relevant papers.
A number of animal species, including strains of rats, mice, and hamsters, have been exposed to TCDD and examined for increases in cancer and tumor incidence. In these studies, TCDD was fed to animals, applied to their skin, injected under their skin, or injected into the abdominal cavity. This research indicates that TCDD can both cause cancers or tumors and enhance the incidence of certain cancers or tumors in the presence of known carcinogens. Increased cancer rates were observed at several different sites in the body, notably the thyroid gland, skin, and lungs. In studies in which liver cancer incidence was enhanced, other adverse changes in the liver were observed. Decreased rates of some cancers—including those of the uterus; pancreas; and pituitary and mammary glands—were also reported. The sites where effects were observed and the exposure levels needed to induce them varied considerably from species to species.
TCDD can act a promoter in the presence of other carcinogens. One study showed that when a single dose of a known carcinogen was applied to the skin of mice followed by multiple doses of TCDD over several months, more skin tumors were seen than would be expected from the single dose of carcinogen alone. Similar results were obtained in rat livers when a single dose of a liver carcinogen was followed by multiple doses of TCDD.
In female rats, enhanced liver tumor formation associated with TCDD exposure is dependent on the presence of intact ovaries, suggesting that complex hormonal interactions are involved in TCDD-induced carcinogenesis.
More generally, TCDD has a wide range of effects on growth regulation, hormone systems, and other factors associated with the regulation of activities in normal cells. These effects may influence tumor formation.
Studies in animals indicate that most TCDD effects are mediated through the AhR, a protein in animal and human cells to which TCDD can bind. It is hypothesized that TCDD exposure induces binding of the AhR to DNA to alter the information obtained from DNA in a way that transforms normal into abnormal cells. Although structural differences in the AhR have been identified, this receptor operates in a similar manner in animals and humans.
As a tumor initiator and promoter, TCDD has been shown to significantly induce CYP1A1 mRNA levels and ethoxyresorufin O-deethylase (EROD) activity in several types of human cancer cells; that is, TCDD affects some forms of cellular metabolism at a very basic level. Experiments involving several strains of mice provide evidence that a functional AhR is required for TCDD induction of CYP1A1 and liver tumor promotion. However, CYP1A1 induction in various mice strains was not directly related to the degree of tumor-promoting capability, suggesting that other undefined genetic factors may play an important role. There is some evidence that certain cells have mechanisms that serve to regulate CYP1A1 tumor promotion.
There are great differences between the susceptibility of different experimental animals to TCDD-induced effects, and the sites at which tumors are induced also varies from species to species. Some mice strains, for example, are differentially responsive to TCDD liver induction due to differences in a protein called cytochrome P450, indicating a genetic susceptibility. Although expression of P450 is highly cell specific, it exhibits a surprisingly similar pattern of hormonal regulation. Studies conducted to compare the AhR in cultured fetal cells and adult liver tumors from TCDD-responsive and less-responsive mice indicate that the responsiveness of fetal cells is likely mediated by the AhR in these cells and is not due to a different allelic form of AhR ligand binding subunit in fetal versus adult cells.
Evidence has also begun to accumulate for non-AhR mediated effects. There is new evidence that the mechanism by which TCDD induces tumor promotion may involve oxygen radicals, since scavengers of hydroxyl radicals or antioxidants hinder the tumor-promoting effects of TCDD in transformed mice fibroblasts. In support of this, other studies have shown that TCDD induces changes in liver cells that lead to a release of oxygen radicals and subsequent oxidative DNA damage. This also suggests that TCDD tumor promotion may be due to interference with gap junctional intercellular communications.
Controversy exists about the TCDD exposure levels required to induce adverse health outcomes.
Limited information is available on health effects of exposure to the herbicides discussed in this report. Several studies of the carcinogenicity of 2,4-D, 2,4,5-T, picloram, and cacodylic acid have been performed in laboratory animals. In general, they produced negative results. However, because some studies do not meet present-day standards for cancer bioassays, and some produced equivocal results, it is not possible to draw confident conclusions at this time.
2,4-D was administered to rats, mice, and dogs in their food, by injecting it under their skin, or placing it directly into their stomachs. All results were negative, except for one study that found an increased rate of brain tumors in male, but not female, rats receiving the highest dose. These tumors also occurred in the control group and thus may have occurred spontaneously and not as a result of 2,4-D exposure. In a recent mutagenicity study, 2,4-D induced significant numbers of mutations in at least one of the cell types tested, either spermatocytes or spermatogonia. Because these results differ from earlier studies, it was hypothesized that different germ cell stages and treatment regimens may account for the observed inconsistencies. Similar results were obtained in a 2,4,5-T mutagenicity study.
2,4,5-T has been administered to rats and mice in their food, in their drinking water, by injecting it under their skin, or by placing it directly into their stomachs. In a recent study, 2,4,5-T exposure increased the formation of DNA adducts by cytochrome P450-derived metabolites of benzo[a]pyrene. The latter effects are particularly interesting since they are strikingly similar to those elicited by dioxin.
Picloram has been tested in rats and mice in their food. Results of all of these studies were negative, with the exception of one study in which liver tumors appeared. These were attributed to the presence of a picloram contaminant, hexachlorobenzene.
A recent study indicates that cacodylic acid (also known as dimethylarsinic acid) may induce DNA modifications that sensitize it to free radical injury, whereas another study concluded that it is a promoter of urinary bladder, kidney, liver, and thyroid gland carcinogenesis in rats. In particular, cacodylic acid may promote rat urinary bladder carcinogenesis by stimulating cell proliferation in the urinary bladder epithelium. An exposure study in mice produced negative results.
The foregoing evidence suggests that a connection between TCDD or herbicide exposure and human health effects is, in general, biologically plausible. However, differences in sensitivity and susceptibility across individual animals, strains, and species; the lack of strong evidence of organ-specific effects across species; and differences in route, dose, duration, and timing of exposure complicate any more definitive conclusions about the presence or absence of a mechanism for the induction of site-specific cancers by TCDD.
Considerable uncertainty remains about how to apply this information to the evaluation of potential health effects of herbicides or dioxin exposure in Vietnam veterans. Scientists disagree over the extent to which information derived from
animals and cellular studies predicts human health outcomes and the extent to which the health effects resulting from high-dose exposure are comparable to those resulting from low-dose exposure. Research on biological mechanisms is burgeoning, and subsequent updates of this report may have more and better information on which to base conclusions.
Increased Risk of Disease Among Vietnam Veterans
Under the Agent Orange Act of 1991, the committee is asked to determine (to the extent that available scientific data permit meaningful determinations) the increased risk of the diseases it studies among those exposed to herbicides during their service in Vietnam. Chapter 1 presents the committee's general findings regarding this charge. Where more specific information about particular health outcomes is available, this information can be found in the preceding discussions of those diseases.
REFERENCES
American Cancer Society (ACS) Cancer Facts and Figures [WWW site]. 1998. URL http://www.cancer.org/statistics/cff98/graphicaldata.html (accessed March 12, 1998).
Air Force Health Study. 1996. An Epidemiologic Investigation of Health Effects in Air Force Personnel Following Exposure to Herbicides. Mortality Update 1996. Brooks AFB, TX: Epidemiologic Research Division. Armstrong Laboratory. AL/AO-TR-1996-0068. 31 pp.
Alavanja MC, Blair A, Merkle S, Teske J, Eaton B. 1988. Mortality among agricultural extension agents. American Journal of Industrial Medicine 14:167-176.
Alavanja MC, Merkle S, Teske J, Eaton B, Reed B. 1989. Mortality among forest and soil conservationists. Archives of Environmental Health 44:94-101.
Amadori D, Nanni O, Falcini F, Saragoni A, Tison V, Callea A, Scarpi E, Ricci M, Riva N, Buiatti E. 1995. Chronic lymphocytic leukaemias and non-Hodgkin's lymphomas by histological type in farming-animal breeding workers: a population case-control study based on job titles. Occupational and Environmental Medicine 52(6):374-379.
Anderson HA, Hanrahan LP, Jensen M, Laurin D, Yick W-Y, Wiegman P. 1986a. Wisconsin Vietnam Veteran Mortality Study: Proportionate Mortality Ratio Study Results. Madison: Wisconsin Division of Health.
Anderson HA, Hanrahan LP, Jensen M, Laurin D, Yick W-Y, Wiegman P. 1986b. Wisconsin Vietnam Veteran Mortality Study: Final Report. Madison: Wisconsin Division of Health.
Asp S, Riihimaki V, Hernberg S, Pukkala E. 1994. Mortality and cancer morbidity of Finnish chlorophenoxy herbicide applicators: an 18-year prospective follow-up. American Journal of Industrial Medicine 26:243-253.
Axelson O, Sundell L, Andersson K, Edling C, Hogstedt C, Kling H. 1980. Herbicide exposure and tumor mortality: an updated epidemiologic investigation on Swedish railroad workers. Scandinavian Journal of Work, Environment, and Health 6:73-79.
Becher H, Flesch-Janys D, Kauppinen T, et al. 1996. Cancer mortality in German male workers exposed to phenoxy herbicides and dioxins [see comments]. Cancer Causes Control. 7:312-321.
Bender AP, Parker DL, Johnson RA, Scharber WK, Williams AN, Marbury MC, Mandel JS. 1989. Minnesota highway maintenance worker study: cancer mortality. American Journal of Industrial Medicine 15:545-556.
Bertazzi PA, Zocchetti C, Pesatori AC. Guercilena S, Sanarico M, Radice L. 1989a. Mortality in an area contaminated by TCDD following an industrial incident. Medicina Del Lavoro 80:316-329.
Bertazzi PA, Zocchetti C, Pesatori AC, Guercilena S, Sanarico M, Radice L. 1989b. Ten-year mortality study of the population involved in the Seveso incident in 1976. American Journal of Epidemiology 129:1187-1200.
Bertazzi A, Pesatori AC, Consonni D, Tironi A, Landi MT, Zocchetti C. 1993. Cancer incidence in a population accidentally exposed to 2,3,7,8-tetrachlorodibenzo-para-dioxin [see comments]. Epidemiology 4:398-406.
Bertazzi PA, Zochetti C, Guercilena S, Consonni D, Tironi A, Landi MT, Pesatori AC. 1997. Dioxin Exposure and Cancer Risk: A 15-Year Mortality Study after the ''Seveso Accident," Epidemiology 8(6):646-652.
Blair A and Kazerouni N. 1997. Reactive chemicals and cancer. Cancer Causes Control 8(3):473-490.
Blair A, Grauman DJ, Lubin JH, Fraumeni JF Jr. 1983. Lung cancer and other causes of death among licensed pesticide applicators. Journal of the National Cancer Institute 71:31-37.
Blair A, Mustafa D, Heineman EF. 1993. Cancer and other causes of death among male and female farmers from twenty-three states. American Journal of Industrial Medicine 23:729-742.
Blair A, Zahm SH, Cantor KP, Ward MH. 1997. Occupational and environmental risk factors for chronic lymphocytic leukemia and non-Hodgkin's lymphoma. In: Marti GE, Vogt RF, Zenger VE, eds. Proceedings of the USPHS Workshop on Laboratory and Epidemiologic Approaches to Determining the Role of Environmental Exposures as Risk Factors for B-Cell Chronic Lymphocytic and Other B-Cell Lymphoproliferative Disorders. U.S. Department of Health and Human Services, Public Health Service.
Bloemen LJ, Mandel JS, Bond GG, Pollock AF, Vitek RP, Cook RR. 1993. An update of mortality among chemical workers potentially exposed to the herbicide 2,4-dichlorophenoxyacetic acid and its derivatives. Journal of Occupational Medicine 35:1208-1212.
Boffetta P, Stellman SD, Garfinkel L. 1989. A case-control study of multiple myeloma nested in the American Cancer Society prospective study. International Journal of Cancer 43:554-559.
Bois FY, Eskenazi B. 1994. Possible risk of endometriosis for Seveso, Italy, residents: an assessment of exposure to dioxin [see comments]. Environmental Health Perspectives 102(5):476-477.
Bond GG, Ott MG, Brenner FE, Cook RR. 1983. Medical and morbidity surveillance findings among employees potentially exposed to TCDD. British Journal of Industrial Medicine 40:318-324.
Bond GG, Wetterstroem NH, Roush GJ, McLaren EA, Lipps TE, Cook RR. 1988. Cause specific mortality among employees engaged in the manufacture, formulation, or packaging of 2,4-dichlorophenoxyacetic acid and related salts. British Journal of Industrial Medicine 45:98-105.
Bosl GJ, Motzer RJ. 1997. Testicular germ-cell cancer. New England Journal of Medicine 337(4): 242-253.
Boyle C, Decoufle P, Delaney R J, DeStefano F, Flock ML, Hunter MI, Joesoef MR, Karon JM, Kirk ML, Layde PM, McGee DL, Moyer LA, Pollock DA, Rhodes P, Scally MJ, Worth RM. 1987. Postservice Mortality Among Vietnam Veterans. Atlanta: Centers for Disease Control. CEH 86-0076. 143 pp.
Breslin P, Kang H, Lee Y, Burt V, Shepard BM. 1988. Proportionate mortality study of U.S. Army and U.S. Marine Corps veterans of the Vietnam War. Journal of Occupational Medicine 30:412-419.
Brown LM, Blair A, Gibson R, Everett GD, Cantor KP, Schuman LM, Burmeister LF, Van Lier SF, Dick F. 1990. Pesticide exposures and other agricultural risk factors for leukemia among men in Iowa and Minnesota. Cancer Research 50:6585-6591.
Brown LM, Burmeister LF, Everett GD, Blair A. 1993. Pesticide exposures and multiple myeloma in Iowa men. Cancer Causes and Control 4:153-156.
Bueno de Mesquita HB, Doornbos G, van der Kuip DA, Kogevinas M, Winkelmann R. 1993. Occupational exposure to phenoxy herbicides and chlorophenols and cancer mortality in the Netherlands. American Journal of Industrial Medicine 23:289-300.
Bullman TA, Watanabe KK, Kang HK. 1994. Risk of testicular cancer associated with surrogate measures of Agent Orange exposure among Vietnam veterans on the Agent Orange Registry. Annals of Epidemiology 4:11-16.
Burmeister LF, Everett GD, Van Lier SF, Isacson P. 1983. Selected cancer mortality and farm practices in Iowa. American Journal of Epidemiology 118:72-77.
Burmeister LF. 1981. Cancer mortality in Iowa farmers: 1971-1978. Journal of the National Cancer Institute 66:461-464.
Cantor KP. 1982. Farming and mortality from non-Hodgkin's lymphoma: a case-control study. International Journal of Cancer 29:239-247.
Cantor KP, Blair A. 1984. Farming and mortality from multiple myeloma: a case-control study with the use of death certificates. Journal of the National Cancer Institute 72:251-255.
Centers for Disease Control. 1988. Health status of Vietnam veterans. II. Physical health. Journal of the American Medical Association 259:2708-2714.
Centers for Disease Control. 1990. The association of selected cancers with service in the U.S. military in Vietnam. III. Hodgkin's disease, nasal cancer, nasopharyngeal cancer, and primary liver cancer. The Selected Cancers Cooperative Study Group [see comments]. Archives of Internal Medicine 150:2495-2505.
Clapp RW, Cupples LA, Colton T, Ozonoff DM. 1991. Cancer surveillance of veterans in Massachusetts, 1982-1988. International Journal of Epidemiology 20:7-12.
Clapp RW. 1997. Update of cancer surveillance of veterans in Massachusetts, USA. International Journal of Epidemiology 26(3):679-681.
Coggon D, Pannett B, Winter PD, Acheson ED, Bonsall J. 1986. Mortality of workers exposed to 2-methyl-4-chlorophenoxyacetic acid. Scandinavian Journal of Work, Environment, and Health 12:448-454.
Coggon D, Pannett B, Winter P. 1991. Mortality and incidence of cancer at four factories making phenoxy herbicides. British Journal of Industrial Medicine 48:173-178.
Collins JJ, Strauss ME, Levinskas GJ, Conner PR. 1993. The mortality experience of workers exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin in a trichlorophenol process accident. Epidemiology 4:7-13.
Cordier S, Le TB, Verger P, Bard D, Le CD, Larouze B, Dazza MC, Hoang TQ, Abenhaim L. 1993. Viral infections and chemical exposures as risk factors for hepatocellular carcinoma in Vietnam. International Journal of Cancer 55:196-201.
Corrao G, Caller M, Carle F, Russo R, Bosia S, Piccioni P. 1989. Cancer risk in a cohort of licensed pesticide users. Scandinavian Journal of Work, Environment, and Health 15:203-209.
Crane PJ, Barnard DL, Horsley KW, Adena MA. 1997a. Mortality of Vietnam Veterans: The Veteran Cohort Study: A Report of the 1996 Retrospective Cohort Study of Australian Vietnam Veterans. Canberra: Department of Veterans' Affairs.
Crane PJ, Barnard DL, Horsley KW, Adena MA. 1997b. Mortality of National Service Vietnam Veterans: A Report of the 1996 Retrospective Cohort Study of Australian Vietnam Veterans. Canberra: Department of Veterans' Affairs.
Dalager NA, Kang HK. 1997. Mortality among Army Chemical Corps Vietnam veterans. American Journal of Industrial Medicine 31:719-726.
Dalager NA, Kang HK, Thomas TL. 1995a. Cancer mortality patterns among women who served in the military: the Vietnam experience. Journal of Occupational and Environmental Medicine 37:298-305.
Dalager NA, Kang HK, Burt VL, Weatherbee L. 1991. Non-Hodgkin's lymphoma among Vietnam veterans. Journal of Occupational Medicine 33:774-779.
Dalager NA, Kang HK, Burt VL, Weatherbee L. 1995b. Hodgkin's disease and Vietnam service [see comments]. Annals of Epidemiology 5:400-406.
Dean G. 1994. Deaths from primary brain cancers, lymphatic and haematopoietic cancers in agricultural workers in the Republic of Ireland. Journal of Epidemiology and Community Health 48:364-368.
Demers PA, Boffetta P, Kogevinas M, Blair A, Miller BA, Robinson CF, Roscoe RJ, Winter PD, Colin D, Matos E, et al. 1995. Pooled reanalysis of cancer mortality among five cohorts of workers in wood-related industries. Scandinavian Journal of Work, Environment and Health 21(3): 179-190.
Donna A, Betta P-G, Robutti F, Crosignani P, Berrino F, Bellingeri D. 1984. Ovarian mesothelial tumors and herbicides: a case-control study. Carcinogenesis 5:941-942.
Dubrow R, Paulson JO, Indian RW. 1988. Farming and malignant lymphoma in Hancock County, Ohio. British Journal of Industrial Medicine 45:25-28.
Eriksson M, Hardell L, Berg NO, Moller T, Axelson O. 1981. Soft-tissue sarcomas and exposure to chemical substances: a case-referent study. British Journal of Industrial Medicine 38:27-33.
Eriksson M, Hardell L, Adami HO. 1990. Exposure to dioxins as a risk factor for soft tissue sarcoma: a population-based case-control study. Journal of the National Cancer Institute 82:486-490.
Eriksson M, Karlsson M. 1992. Occupational and other environmental factors and multiple myeloma: a population based case-control study. British Journal of Industrial Medicine 49:95-103.
Fett MJ, Nairn JR, Cobbin DM, Adena MA. 1987. Mortality among Australian conscripts of the Vietnam conflict era. II. Causes of death. American Journal of Epidemiology 125:878-884.
Fingerhut MA, Halperin WE, Marlow DA, Piacitelli LA, Honchar PA, Sweeney MH, Greife AL, Dill PA, Steenland K, Suruda AJ. 1991. Cancer mortality in workers exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin. New England Journal of Medicine 324:212-218.
Flesch-Janys D, Berger J, Gum P, et al. 1995. Exposure to polychlorinated dioxins and furans (PCDD/F) and mortality in a cohort of workers from a herbicide-producing plant in Hamburg, Federal Republic of Germany. American Journal of Epidemiology 142:1165-1175.
Gallagher RP, Bajdik CD, Fincham S, Hill GB, Keefe AR, Coldman A, McLean DI. 1996. Chemical exposures, medical history, and risk of squamous and basal cell carcinoma of the skin. Cancer Epidemiology, Biomarkers and Prevention 5(6):419-424.
Gambini GF, Mantovani C, Pira E, Piolatto PG, Negri E. 1997. Cancer mortality among rice growers in Novara Province, Northern Italy. American Journal of Industrial Medicine 31:435-441.
Gann PH. 1997. Interpreting recent trends in prostate cancer incidence and mortality. Epidemiology 8(2):117-120.
Garry VF, Kelly JT, Sprafka JM, Edwards S, Griffith J. 1994. Survey of health and use characterization of pesticide appliers in Minnesota. Archives of Environmental Health 49:337-343.
Garry VF, Tarone RE, Long L, Griffith J, Kelly JT, Burroughs B. 1996. Pesticide appliers with mixed pesticide exposure: G-banded analysis and possible relationship to non-Hodgkin's lymphoma, Cancer Epidemiology, Biomarkers and Prevention 5(1):11-16.
Gonzalez FJ, Fernandez-Salguero P, Ward JM. 1996. The role of the aryl hydrocarbon receptor in animal development, physiological homeostasis and toxicity of TCDD. Journal of Toxicological Sciences 21:273-277.
Green LM. 1991. A cohort mortality study of forestry workers exposed to phenoxy acid herbicides . British Journal of Industrial Medicine 48:234-238.
Hansen ES, Hasle H, Lander F. 1992. A cohort study on cancer incidence among Danish gardeners. American Journal of Industrial Medicine 21:651-660.
Hardell L, Bengtsson NO. 1983. Epidemiological study of socioeconomic factors and clinical findings in Hodgkin's disease, and reanalysis of previous data regarding chemical exposure. British Journal of Cancer 48:217-225.
Hardell L, Eriksson M. 1988. The association between soft tissue sarcomas and exposure to phenoxyacetic acids: a new case-referent study. Cancer 62:652-656.
Hardell L, Sandstrom A. 1979. Case-control study: soft-tissue sarcomas and exposure to phenoxyacetic acids or chlorophenols. British Journal of Cancer 39:711-717.
Hardell L, Eriksson M, Lenner P. 1980. Malignant lymphoma and exposure to chemical substances, especially organic solvents, chlorophenols and phenoxy acids. Lakartidningen 77:208-210.
Hardell L, Eriksson M, Lenner P, Lundgren E. 1981. Malignant lymphoma and exposure to chemicals, especially organic solvents, chlorophenols and phenoxy acids: a case-control study. British Journal of Cancer 43:169-176.
Hardell L, Johansson B, Axelson O. 1982. Epidemiological study of nasal and nasopharyngeal cancer and their relation to phenoxy acid or chlorophenol exposure. American Journal of Industrial Medicine 3:247-257.
Hardell L, Bengtsson NO, Jonsson U, Eriksson S, Larsson LG. 1984. Aetiological aspects on primary liver cancer with special regard to alcohol, organic solvents and acute intermittent porphyria: an epidemiological investigation. British Journal of Cancer 50:389-397.
Hardell L. 1981. Relation of soft-tissue sarcoma, malignant lymphoma and colon cancer to phenoxy acids, chlorophenols and other agents. Scandinavian Journal of Work, Environment, and Health 7:119-130.
Hayes HM, Tarone RE, Casey HW, Huxsoll DL. 1990. Excess of seminomas observed in Vietnam service U.S. military working dogs. Journal of the National Cancer Institute 82:1042-1046.
Hayes HM, Tarone RE, Casey HW, Jennings PB Jr, Hildebrandt PK, Reardon MJ. 1994. U.S. military working dogs with Vietnam service: definition and characteristics of the cohort. Military Medicine 159(11):669-675.
Hayes HM, Tarone RE, Cantor KP. 1995a. On the association between canine malignant lymphoma and opportunity for exposure to 2,4-dichlorophenoxyacetic acid. Environmental Research 70:119-125.
Hayes HM, Tarone RE, Casey HW. 1995b. A cohort study of the effects of Vietnam service on testicular pathology of U.S. military working dogs. Military Medicine 160(5):248-255.
Hayes RB. 1997. The carcinogenicity of metals in humans. Cancer Causes Control 8(3):371-385.
Henneberger PK, Ferris BG Jr, Monson RR. 1989. Mortality among pulp and paper workers in Berlin, New Hampshire. British Journal of Industrial Medicine 46:658-664.
Hertzman C, Teschke K, Ostry A, Hershler R, Dimich-Ward H, Kelly S, Spinelli JJ, Gallagher RP, McBride M, Marion SA. 1997. Mortality and cancer incidence among sawmill workers exposed to chlorophenate wood preservatives . American Journal of Public Health 87(1): 71-79.
Hoar SK, Blair A, Holmes FF, Boysen CD, Robel RJ, Hoover R, Fraumeni JF. 1986. Agricultural herbicide use and risk of lymphoma and soft-tissue sarcoma. Journal of the American Medical Association 256:1141-1147.
Hoffman RE, Stehr-Green PA, Webb KB, Evans RG, Knutsen AP, Schramm WF, Staake JL, Gibson BB, Steinberg KK. 1986. Health effects of long-term exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Journal of the American Medical Association 255:2031-2038.
Holmes AP, Bailey C, Baron RC, Bosanac E, Brough J, Conroy C, Haddy L. 1986. West Virginia Department of Health Vietnam-era Veterans Mortality Study: Preliminary Report. Charlestown: West Virginia Health Department.
Kauppinen TP, Pannett B, Marlow DA, Kogevinas M. 1994. Retrospective assessment of exposure through modeling in a study on cancer risks among workers exposed to phenoxy herbicides, chlorophenols and dioxins. Scandinavian Journal of Work, Environment, and Health 20:262-271.
Keller-Byrne JE, Khuder SA, Schaub EA, McAfee O. 1997. A meta-analysis of non-Hodgkin's lymphoma among farmers in the central United States. American Journal of Industrial Medicine 31(4):442-444.
Kogan MD, Clapp RW. 1985. Mortality Among Vietnam Veterans in Massachusetts, 1972-1983. Massachusetts Office of the Commissioner of Veterans Services, Agent Orange Program.
Kogan MD, Clapp RW. 1988. Soft tissue sarcoma mortality among Vietnam veterans in Massachusetts, 1972 to 1983. International Journal of Epidemiology 17:39-43.
Kogevinas M, Saracci R, Bertazzi PA, Bueno de Mesquita BH, Coggon D, Green LM, Kauppinen T, Littorin M, Lynge E, Mathews JD, Neuberger M, Osman J, Pearce N, Winkelmann R. 1992. Cancer mortality from soft-tissue sarcoma and malignant lymphomas in an international cohort of workers exposed to chlorophenoxy herbicides and chlorophenols. Chemosphere 25:1071-1076.
Kogevinas M, Saracci R, Winkelmann R, Johnson ES, Bertazzi PA, Bueno de Mesquita BH, Kauppinen T, Littorin M, Lynge E, Neuberger M. 1993. Cancer incidence and mortality in women occupationally exposed to chlorophenoxy herbicides, chlorophenols, and dioxins. Cancer Causes Control 4:547-553.
Kogevinas M, Kauppinen T, Winkelmann R, Becher H, Bertazzi PA, Bas B, Coggon D, Green L, Johnson E, Littorin M, Lynge E, Marlow DA, Mathews JD, Neuberger M, Benn T, Pannett B, Pearce N, Saracci R. 1995. Soft tissue sarcoma and non-Hodgkin's lymphoma in workers exposed to phenoxy herbicides, chlorophenols and dioxins: two nested case-control studies. Epidemiology 6:396-402.
Kogevinas M, Becher H, Benn T, Bertazzi PA, Boffetta P, Bueno de Mesquita HB, Coggon D, Colin D, Flesch-Janys D, Fingerhut M, Green L, Kauppinen T, Littorin M, Lynge E, Mathews JD, Neuberger M, Pearce N, Saracci R. 1997. Cancer mortality in workers exposed to phenoxy herbicides, chlorophenols, and dioxins. An expanded and updated international cohort study, American Journal of Epidemiology 145(12):1061-1075.
Lampi P, Hakulinen T, Luostarinen T, Pukkala E, Teppo L. 1992. Cancer incidence following chlorophenol exposure in a community in southern Finland. Archives of Environmental Health 47:167-175.
LaVecchia C, Negri E, D'Avanzo B, Franceschi S. 1989. Occupation and lymphoid neoplasms. British Journal of Cancer 60:385-358.
Lawrence CE, Reilly AA, Quickenton P, Greenwald P, Page WF, Kuntz AJ. 1985. Mortality patterns of New York State Vietnam veterans. American Journal of Public Health 75:277-279.
Lynge E. 1985. A follow-up study of cancer incidence among workers in manufacture of phenoxy herbicides in Denmark . British Journal of Cancer 52:259-270.
Lynge E. 1993. Cancer in phenoxy herbicide manufacturing workers in Denmark, 1947-87—an update. Cancer Causes and Control 4:261-272.
Mack TM. 1995. Sarcomas and other malignancies of soft tissue, retroperitoneum, peritoneum, pleura, heart, mediastinum, and spleen. Cancer 75:211-244.
Mahan CM, Bullman TA, Kang HK, Selvin S. 1997. A case-control study of lung cancer among Vietnam veterans. Journal of Occupational and Environmental Medicine 39(8):740-747.
Mahaney FX. 1990. Military working dogs may be a sentinel for human cancer. Journal of the National Cancer Institute 82(12): 1002-1003.
Manz A, Berger J, Dwyer JH, Flesch-Janys D, Nagel S, Waltsgott H. 1991. Cancer mortality among workers in chemical plant contaminated with dioxin. Lancet 338:959-964.
Masala G, Di Lollo S, Picoco C, Crosignani P, Demicheli V, Fontana A , Funto I, Miligi L, Nanni O, Papucci A, Ramazzotti V, Rodella S, Stagnaro E, Tumino R, Vigano C, Vindigni C, Seniori Costantini A, Vineis P. 1996. Incidence rates of leukemias, lymphomas and myelomas in Italy: geographic distribution and NHL histotypes. International Journal of Cancer 68(2): 156-159.
McDuffie HH, Klaassen DJ, Dosman JA. 1990. Is pesticide use related to the risk of primary lung cancer in Saskatchewan? Journal of Occupational Medicine 32:996-1002.
McKinney WP, McIntire DD, Carmody TJ, Joseph A. 1997. Comparing the smoking behavior of veterans and nonveterans. Public Health Reports 112(3):212-217.
Mellemgaard A, Engholm G, McLaughlin JK, Olsen JH. 1994. Occupational risk factors for renal-cell carcinoma in Denmark. Scandinavian Journal of Work, Environment, and Health 20:160-165.
Michalek JE, Wolfe WH, Miner JC. 1990. Health status of Air Force veterans occupationally exposed to herbicides in Vietnam. II. Mortality. Journal of the American Medical Association 264:1832-1836.
Miller BA, Kolonel LN, Bernstein L, Young, Jr. JL, Swanson GM, West D, Key CR, Liff JM, Glover CS, Alexander GA, et al. (eds). 1996. Racial/Ethnic Patterns of Cancer in the United States 1988-1992, National Cancer Institute. NIH Pub. No. 96-4104. Bethesda, MD.
Molinari R, Grande C, Zucali R, Pilotti S, Della Torre G. 1995. Tumours of the nasopharynx. In: Peckham M, Pinedo HM, Veronesi U, eds. Oxford Textbook of Oncology. Oxford, UK: Oxford University Press.
Morris PD, Koepsell TD, Daling JR, Taylor JW, Lyon JL, Swanson GM, Child M, Weiss NS. 1986. Toxic substance exposure and multiple myeloma: a case-control study. Journal of the National Cancer Institute 76:987-994.
Morrison H, Semenciw RM, Morison D, Magwood S, Mao Y. 1992. Brain cancer and farming in western Canada. Neuroepidemiology 11:267-276.
Morrison H, Savitz D, Semenciw RM, Hulka B, Mao Y, Morison D, Wigle D. 1993. Farming and prostate cancer mortality. American Journal of Epidemiology 137:270-280.
Morrison HI, Semenciw RM, Wilkins K, Mao Y, Wigle DT. 1994. Non-Hodgkin's lymphoma and agricultural practices in the prairie provinces of Canada. Scandinavian Journal of Work, Environment, and Health 20:42-47.
Moses M, Lilis R, Crow KD, Thornton J, Fischbein A, Anderson HA, Selikoff IJ. 1984. Health status of workers with past exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin in the manufacture of 2,4,5-trichlorophenoxyacetic acid: comparison of findings with and without chloracne. American Journal of Industrial Medicine 5:161-182.
Mueller N. 1995. Overview: viral agents and cancer. Environmental Health Perspectives 103(Suppl 8):259-261.
Musicco M, Sant M, Molinari S, Filippini G, Gatta G, Berrino F. 1988. A case-control study of brain gliomas and occupational exposure to chemical carcinogens: the risks to farmers. American Journal of Epidemiology 128:778-785.
Nanni O, Amadori D, Lugaresi C, Falcini F, Scarpi E, Saragoni A, Buiatti E. 1996. Chronic lymphocytic leukaemias and non-Hodgkin's lymphomas by histological type in farming-animal breeding workers: a population case-control study based on a priori exposure matrices. Occupational and Environmental Medicine 53(10):652-657.
National Cancer Institute, Surveillance, Epidemiology, and End Results (SEER) Program, Cancer Query System on the Web, 1973-1994 [WWW search engine]. 1998. URL http://wwwseer.ims.nci.nih.gov/ScientificSystems/Canques7394/(accessed March 9, 1998).
National Toxicology Program. 1982a. Carcinogenesis Bioassay of 2,3,7,8-tetrachlorodibenzo-p-dioxin (CAS No. 1746-01-6) in Osborne-Mendel Rats and B6C3F1 Mice (Gavage Study). Research Triangle Park, N.C.: NTP. NTP-80-31; NIH/PUB-82-1765.PC A09/MF A01.
National Toxicology Program. 1982b. Carcinogenesis Bioassay of 2,3,7,8-tetrachlorodibenzo-p-dioxin (CAS No. 1746-01-6) in Swiss Webster Mice (Dermal Study). Research Triangle Park, N.C.: NTP. NTP-80-31; NIH/PUB-82-1765.PC A09/MF A01 .
Olsson H, Brandt L. 1988. Risk of non-Hodgkin's lymphoma among men occupationally exposed to organic solvents. Scandinavian Journal of Work, Environment, and Health 14:246-251.
O'Toole BI, Marshall RP, Grayson DA, et al. 1996. The Australian Vietnam Veterans Health Study: II. self-reported health of veterans compared with the Australian population. International Journal of Epidemiology 25(2):319-330.
Ott MG, and Zober A. 1996. Cause specific mortality and cancer incidence among employees exposed to 2,3,7,8-TCDD after a 1953 reactor accident. Occuptional and Environmental Medicine 53:606-612.
Ott MG, Olson RA, Cook RR, Bond GG. 1987. Cohort mortality study of chemical workers with potential exposure to the higher chlorinated dioxins. Journal of Occupational Medicine 29:422-429.
Pearce NE, Sheppard RA, Smith AH, Teague CA. 1987. Non-Hodgkin's lymphoma and farming: an expanded case-control study. International Journal of Cancer 39:155-161.
Pearce NE, Smith AH, Fisher DO. 1985. Malignant lymphoma and multiple myeloma linked with agricultural occupations in a New Zealand cancer registry-based study. American Journal of Epidemiology 121:225-237.
Pearce NE, Smith AH, Howard JK, Sheppard RA, Giles HJ, Teague CA. 1986. Non-Hodgkin's lymphoma and exposure to phenoxyherbicides, chlorophenols, fencing work, and meat works employment: a case-control study. British Journal of Industrial Medicine 43:75-83.
Percy C, Ries GL, Van Holten VD. 1990. The accuracy of liver cancer as the underlying cause of death on death certificates. Public Health Reports 105:361-368.
Persson B, Dahlander A-M, Fredriksson M, Brage HN, Ohlson C-G, Axelson O. 1989. Malignant lymphomas and occupational exposures. British Journal of Industrial Medicine 46:516-520.
Persson B, Fredriksson M, Olsen K, Boeryd B, Axelson O. 1993. Some occupational exposures as risk factors for malignant lymphomas. Cancer 72:1773-1778.
Pesatori AC, Consonni D, Tironi A, Landi MT, Zocchetti C, Bertazzi PA. 1992. Cancer morbidity in the Seveso area, 1976-1986. Chemosphere 25:209-212.
Ramlow JM, Spadacene NW, Hoag SR, Stafford BA, Cartmill JB, Lerner PJ. 1996. Mortality in a cohort of pentachlorophenol manufacturing workers, 1940-1989. American Journal of Industrial Medicine 30:180-194.
Riedel D, Pottern LM, Blattner WA. 1991. Etiology and epidemiology of multiple myeloma. In: Wiernick PH, Camellos G, Kyle RA, Schiffer CA, eds. Neoplastic Disease of the Blood and Blood Forming Organs. New York: Churchill Livingstone.
Rier SE, Martin DC, Bowman RE, Dmowski WP, Becker JL. 1993. Endometriosis in rhesus monkeys (Macaca mulatta) following chronic exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin. Fundamental and Applied Toxicology 21:433-441.
Ries LAG, Kosary CL, Hankey BF, Miller BA, Harras A, Edwards BK (eds). 1997. SEER Cancer Statistics Review, 1973-1994, National Cancer Institute. NIH Pub. No. 97-2789. Bethesda, MD.
Riihimaki V, Asp S, Hemberg S. 1982. Mortality of 2,4-dichlorophenoxyacetic acid and 2,4,5-trichlorophenoxyacetic acid herbicide applicators in Finland: first report of an ongoing prospective cohort study. Scandinavian Journal of Work, Environment, and Health 8:37-42.
Robinson CF, Waxweiler RJ, Fowler DP. 1986. Mortality among production workers in pulp and paper mills. Scandinavian Journal of Work, Environment, and Health 12:552-560.
Ronco G, Costa G, Lynge E. 1992. Cancer risk among Danish and Italian farmers. British Journal of Industrial Medicine 49:220-225.
Saracci R, Kogevinas M, Bertazzi PA, Bueno De Mesquita BH, Coggon D, Green LM, Kauppinen T, L'Abbe KA, Littorin M, Lynge E, Mathews JD, Neuberger M, Osman J, Pearce N, Winkelmann R. 1991. Cancer mortality in workers exposed to chlorophenoxy herbicides and chlorophenols. Lancet 338:1027-1032.
Semenciw RM, Morrison HI, Morison D, Mao Y. 1994. Leukemia mortality and farming in the prairie provinces of Canada. Canadian Journal of Public Health 85:208-211.
Smith JG, Christophers AJ. 1992. Phenoxy herbicides and chlorophenols: a case control study on soft tissue sarcoma and malignant lymphoma. British Journal of Cancer 65:442-448.
Smith-Warner SA, Spiegelman D, Yaun SS, van den Brandt PA, Folsom AR, Goldbohm RA, Graham S, Holmberg L, Howe GR, Marshall JR, Miller AB, Potter JD, Speizer FE, Willett WC, Wolk A, Hunter DJ. 1998. Alcohol and breast cancer in women: a pooled analysis of cohort studies. Journal of the American Medical Association 279(7):535-540.
Solet D, Zoloth SR, Sullivan C, Jewett J, Michaels DM. 1989. Patterns of mortality in pulp and paper workers. Journal of Occupational Medicine 31:627-630.
Stehr PA, Stein G, Webb K, Schramm W, Gedney WB, Donnell HD, Ayres S, Falk H, Sampson E, Smith SJ. 1986. A pilot epidemiologic study of possible health effects associated with 2,3,7,8-tetrachlorodibenzo-p-dioxin contaminations in Missouri. Archives of Environmental Health 41:16-22.
Stehr-Green P, Hoffman R, Webb K, Evans RG, Knusten A. Schramm W, Staake J, Gibson B, Steinberg K. 1987. Health effects of long-term exposure to 2,3,7.8-tetrachlorodibenzo-p-dioxin. Chemosphere 16:2089-2094.
Suskind RR, Hertzberg VS. 1984. Human health effects of 2,4,5-T and its toxic contaminants. Journal of the American Medical Association 251:2372-2380.
Svensson BG, Mikoczy Z, Stromberg U, Hagmar L. 1995. Mortality and cancer incidence among swedish fishermen with a high dietary intake of persistent organochlorine compounds. Scandinavian Journal of Work, Environmental and Health 21 (2):106-115.
Swaen GMH, van Vliet C, Slangen JJM, Sturmans F. 1992. Cancer mortality among licensed herbicide applicators. Scandinavian Journal of Work, Environment, and Health 18:201-204.
Tarone RE, Hayes HM, Hoover RN, Rosenthal JF, Brown LM, Pottern LM, Javadpour N, O'Connell KJ, Stutzman RE. 1991. Service in Vietnam and risk of testicular cancer. Journal of the National Cancer Institute 83:1497-1499.
Tatham L, Tolbert P, Kjeldsberg C. 1997. Occupational risk factors for subgroups of non-Hodgkin'slymphoma. Epidemiology 8(5):551-558.
Thiess AM, Frentzel-Beyme R, Link R. 1982. Mortality study of persons exposed to dioxin in a trichlorophenol-process accident that occurred in the BASF AG on November 17.1953. American Journal of Industrial Medicine 3:179-189.
Thomas TL, Kang HK. 1990. Mortality and morbidity among Army Chemical Corps Vietnam veterans: a preliminary report. American Journal of Industrial Medicine 18:665-673.
Thomas TL, Kang H, Dalager N. 1991. Mortality among women Vietnam veterans, 1973-1987. American Journal of Epidemiology 134:973-980.
Thomas TL. 1987. Mortality among flavour and fragrance chemical plant workers in the United States. British Journal of Industrial Medicine 44:733-737.
U.S. Bureau of the Census. Statistical Abstract of the United States: 1997 (117th edition). 1997. Washington, DC.
Vineis P, Faggiano F, Tedeschi M, Ciccone G. 1991. Incidence rates of lymphomas and soft-tissue sarcomas and environmental measurements of phenoxy herbicides. Journal of the National Cancer Institute 83:362-363.
Visintainer PF, Barone M, McGee H, Peterson EL. 1995. Proportionate mortality study of Vietnam-era veterans of Michigan. Journal of Occupational and Environmental Medicine 37/4
Watanabe KK, Kang HK. 1995. Military service in Vietnam and the risk of death from trauma and selected cancers. Annals of Epidemiology 5:407-412.
Watanabe KK, Kang HK. 1996. Mortality patterns among Vietnam veterans: a 24-year retrospective analysis. Journal of Occupational and Environmental Medicine 38(3):272-278.
Watanabe KK, Kang HK, Thomas TL. 1991. Mortality among Vietnam veterans: with methodological considerations. Journal of Occupational Medicine 33:780-785.
Waterhouse D, Carman WJ, Schottenfeld D, Gridley G, McLean S. 1996. Cancer incidence in the rural community of Tecumseh, Michigan: a pattern of increased lymphopoietic neoplasms. Cancer 77(4):763-770.
Wigle DT, Semenciw RB, Wilkins K, Riedel D, Ritter L, Morrison HI, Mao Y. 1990. Mortality study of Canadian male farm operators: non-Hodgkin's lymphoma mortality and agricultural practices in Saskatchewan. Journal of the National Cancer Institute 82:575-582.
Wiklund K. 1983. Swedish agricultural workers: a group with a decreased risk of cancer. Cancer 51:566-568.
Wiklund K, Dich J, Holm L-E. 1989. Risk of soft tissue sarcoma, Hodgkin's disease and non-Hodgkin lymphoma among Swedish licensed pesticide applicators. Chemosphere 18:395-400.
Wiklund K, Dich J, Holm L-E, Eklund G. 1989. Risk of cancer in pesticide applicators in Swedish agriculture. British Journal of Industrial Medicine 46:809-814.
Wiklund K, Lindefors BM, Holm L-E. 1988. Risk of malignant lymphoma in Swedish agricultural and forestry workers. British Journal of Industrial Medicine 45:19-24.
Wolfe WH, Michalek JE, Miner JC, Rahe A, Silva J, Thomas WF, Grubbs WD, Lustik MB, Karrison TG, Roegner RH, Williams DE. 1990. Health status of Air Force veterans occupationally exposed to herbicides in Vietnam. I. Physical health. Journal of the American Medical Association 264:1824-1831.
Woods JS, Polissar L, Severson RK, Heuser LS, Kulander BG. 1987. Soft tissue sarcoma and non-Hodgkin's lymphoma in relation to phenoxy herbicide and chlorinated phenol exposure in western Washington. Journal of the National Cancer Institute 78:899-910.
Zack JA, Suskind RR. 1980. The mortality experience of workers exposed to tetrachlorodibenzo-dioxin in a trichlorophenol process accident. Journal of Occupational Medicine 22:11-14.
Zahm SH, Fraumeni JF Jr. 1997. The epidemiology of soft tissue sarcoma. Seminars in Oncology 24(5):504-514.
Zahm SH, Weisenburger DD, Babbitt PA, Saal RC, Vaught JB, Cantor KP, Blair A. 1990. A case-control study of non-Hodgkin's lymphoma and the herbicide 2,4-dichlorophenoxyacetic acid (2,4-D) in eastern Nebraska. Epidemiology 1:349-356.
Zhong Y, Rafnsson V. 1996. Cancer incidence among Icelandic pesticide users, International Journal of Epidemiology 25(6): 1117-1124.
Zhu K, Levine RS, Brann EA, Gnepp DR, Baum MK. 1995. A population-based case-control study of the relationship between cigarette smoking and nasopharyngeal cancer (United States). Cancer Causes Control 6(6):507-512.
Zober A, Messerer P, Huber P. 1990. Thirty-four-year mortality follow-up of BASF employees exposed to 2,3,7,8-TCDD after the 1953 accident. International Archives of Occupational and Environmental Health 62:139-157.