National Academies Press: OpenBook

Gulf War and Health: Volume 2: Insecticides and Solvents (2003)

Chapter: 6. Cancer and Exposure to Solvents

« Previous: 5. Cancer and Exposure to Insecticides
Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

6
CANCER AND EXPOSURE TO SOLVENTS

INTRODUCTION

The associations between exposure to organic solvents and the incidence of and mortality from cancer have been investigated extensively in a number of industries and occupations, including dry-cleaning, painting, printing, and rubber and shoe manufacturing. As a result, the body of evidence on exposure to organic solvents and cancer reviewed by the committee is quite large compared with that on other health effects. To help the reader become more familiar with the studies on exposure to solvents and cancer, this chapter is organized differently from Chapter 5.

Following this general introduction is a description of the major occupational cohort studies that are cited throughout the chapter; these studies provide findings for multiple cancer outcomes. The committee describes the essential study design characteristics and pertinent information for each of these cohorts organized by the type of solvent exposure.

The occupational cohort studies examined populations with known or suspected exposure to the solvents under review. Many of them have been updated and expanded to include more cohort members, longer periods of assessment, and other estimates of exposure. All of the various studies that follow a particular cohort, such as the NIOSH Pliofilm cohort, are described together in Table 6.1. The committee reviewed all the papers related to each major cohort in drawing its conclusions of association, but it is usually the findings from the most recent papers that are provided in the data-analysis tables at the end of the section on each cancer site. In some cases, results from the earlier papers were never reproduced, so the committee used the earlier results in its analysis.

A description of key case-control studies at the beginning of each section is followed by a table that outlines the studies’ characteristics and design elements and is similar to the table of cohort studies at the beginning of the chapter. Discussions of strengths and limitations of the studies that formed the basis of the committee’s conclusions are presented in the sections on the specific outcomes.

Background information on the types of cancer or cancer in general is provided in Chapter 5, and the reader is referred to those sections for that information. A review of the pertinent toxicology and findings from other organizations charged with evaluating the carcinogenicity of organic solvents is provided at the end of this introduction.

As in Chapter 5, the cancer outcomes are presented in the order of the ninth revision of the International Classification of Disease (ICD-9).

The Literature on Exposure to Organic Solvents

The literature on exposure to organic solvents and cancer outcomes provides information on specific solvent exposures (for example, benzene, trichloroethylene, and

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

tetrachloroethylene), on mixtures of specific organic solvents, or on mixtures of unspecified solvents. In many studies, exposure to solvents was not assessed specifically; rather, surrogates of exposure were used, such as job title, industry type, or occupation.

As is discussed in Chapter 2, a study’s ability to determine exposure accurately and specifically is critical in evaluating its overall quality. For the purposes of this report, the committee used studies that assessed exposure to specific organic solvents or to solvent mixtures as the primary evidence for its conclusions. The committee also included surrogates of exposure in drawing its conclusions, but it viewed those studies as supportive or supplemental evidence. Those studies included exposure of painters, printers, aircraft maintenance and manufacturing workers, service-station attendants, and shoe and boot manufacturers. All those studies are included in the data-analysis tables that accompany discussions of each cancer outcome.

The committee found most of the cancer literature focused on the following compounds: benzene, trichloroethylene, tetrachloroethylene, methylene chloride, and mixtures of unspecified organic solvents. Therefore, most of the committee’s conclusions on cancer outcomes are focused on exposure to those compounds. A smaller number of studies analyzed associations between cancer outcomes and toluene, xylene, isopropyl alcohol, methyl ethyl ketone, phenol, and other individual solvents, but for most agents, there was insufficient evidence for the committee to draw conclusions.

For exposure to tetrachloroethylene, the committee included studies of dry-cleaning and laundry workers as part of the primary evidence in drawing conclusions of associations. As a result, the conclusions related to exposure to tetrachloroethylene are also related to exposure to “dry-cleaning solvents.” The committee acknowledges that dry cleaners and launderers are likely exposed to other organic solvents and chemicals, including naphtha, Stoddard solvent, carbon tetrachloride, trichloroethylene, and 1,1,1-trichloroethane (IARC, 1995). As a result, the committee decided to consider studies of both tetrachloroethylene and dry-cleaning solvents in forming their conclusions, thereby including the possibility that one of the other solvents used in that industry contributed to the risks observed in some of the studies on dry cleaners and launderers.

The committee based its review of cancer outcomes only on studies of humans that had a comparison or control group (cohort and case-control studies). Case reports, case series, review articles, and meta-analyses related to cancer were excluded from the committee’s review. Although the committee reviewed ecologic, cross-sectional, proportionate mortality ratio (PMR), and mortality odds ratio studies, it did not consider them critical to its decision and excluded them from the discussions. Chapter 2 describes their specific limitations.

Toxicity and Determination of Carcinogenicity

Excess incidence of cancer has been observed in animals exposed to the specific organic solvents reviewed by the committee. Benzene, perhaps the most thoroughly investigated solvent, is a well-established carcinogen and has repeatedly been shown to induce hematopoietic cancers and cancers of the ovaries, mammary glands, pancreas, and liver (ATSDR, 1997a). The International Agency for Research on Cancer (IARC) has determined that benzene is “carcinogenic to humans” as determined in studies of both humans and animals. IARC bases its determination of benzene’s carcinogenicity on evidence from human studies that is considered “sufficient,” whereas the available animal

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

data are considered “limited.” Most of the human studies cited by IARC involve the increased risk of leukemia and other lymphatic and hematopoietic cancers (IARC, 1987). The National Toxicology Program (NTP) has also classified benzene as “known to be a human carcinogen” in its most recent report on carcinogens on the basis of animal and human studies (NTP, 2001).

On the basis of animal studies, trichloroethylene has been associated with liver cancer in one strain of one species (mouse) (ATSDR, 1997b). Liver and renal cell cancers and mononuclear cell leukemia have typically been seen after exposure to tetrachloroethylene. According to the Agency for Toxic Substance and Disease Registry (ATSDR, 1997c), the relevance to humans of rodent toxicology studies on trichloroethylene and tetrachloroethylene is unclear, given that some mechanisms of action differ. However, a great deal of research has been conducted over the last decade, and some mechanisms of action appear to be similar in rodents and humans such as genotoxic and cytotoxic actions of mercapturic acid derivatives of both trichloroethylene and tetrachloroethylene in the kidney (see Chapter 4 for more information). IARC has also reviewed trichloroethylene and tetrachloroethylene and determined that both are “probably carcinogenic to humans.” The evidence from animal studies is stronger and considered to be “sufficient,” whereas the evidence from human studies is considered “limited” (IARC, 1995). In addition, the NTP has identified both trichloroethylene and tetrachloroethylene as “reasonably anticipated to be human carcinogens” (NTP, 2001).

Exposure to methylene chloride in some rodent species has consistently produced excess numbers of cancers of the liver and lung and benign mammary tumors (ATSDR, 2000). IARC has determined that exposure to methylene chloride is “possibly carcinogenic to humans,” and the NTP concluded that it is “reasonably anticipated to be a human carcinogen” (IARC, 1999; NTP, 2001). IARC has determined that toluene and xylene are “not classifiable as to their carcinogenicity to humans” in that there was inadequate evidence from studies of humans and animals (IARC, 1999).

Chloroform has produced liver and kidney tumors in a strain-, sex-, species-, and dose-dependent manner and, on the basis of sufficient evidence from animal studies, is “reasonably anticipated to be a human carcinogen” according to the NTP (ATSDR, 1997d; NTP, 2001). Chloroform was once used as an anesthetic, but its association with cancer in nonmedical exposures in humans has not been investigated extensively. The committee did not review studies on the efficacy of solvents as therapeutic agents (see Chapter 2). Chapter 4 provides details on the adverse effects of chloroform as observed in experimental studies.

In addition to evaluating the carcinogenicity of specific chemical agents, IARC has analyzed whether particular occupations pose a greater risk for exposure to carcinogenic agents. In fact, IARC has determined that working in the rubber industry and in boot and shoe manufacturing and repair pose such a risk (IARC, 1987), and it determined that there is “sufficient evidence for the carcinogenicity of occupational exposure as a painter” (IARC, 1989). Although IARC identifies exposures of concerns and specific cancer outcomes that demonstrate an increased risk, its overall charge is to determine whether a specific agent or occupation is carcinogenic, not whether an agent causes a specific cancer outcome. It is important to distinguish the objectives of IARC’s program and the charge of the present committee. The purpose of the IARC program is to determine whether agents or occupational exposures are carcinogenic, whereas this committee is charged with determining whether there is an association between exposure to a specific agent and

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

chronic human illnesses. As discussed earlier in this report, the committee uses experimental evidence only when it is required by the definitions of the categories of association. Only the category of “Sufficient Evidence of a Causal Association” requires support from experimental evidence. For each conclusion of causality, the animal data that provides a plausible mechanism for the outcome being discussed are described—as in the section on chronic exposure to benzene and acute leukemia. Additional information on the toxicology and available experimental data on a number of solvents reviewed in this report can be found in Chapter 4.

DESCRIPTION OF THE COHORT STUDIES

In reviewing the published epidemiologic literature on exposure to organic solvents, the committee examined a number of occupational cohort studies that provided information on the association between cancer mortality or incidence and exposure to specific organic solvents or to mixtures of organic solvents. Deaths or incident cases of cancer in the cohort studies were recorded, exposed populations were followed over time, and the relationship between rates of cancer and exposure was assessed with statistical methods. Because the cohort studies played an important role in the committee’s conclusions and are referred to throughout this chapter, they are described here according to the solvents they examined. Table 6.1 presents for each study a description of the population, the followup period, the number of subjects, the relevant exposures, the methods used to assess exposure to organic solvents, the statistical methods, and the adjustment for potential confounding variables. Similar tables for case-control studies are found in the sections on each type of cancer.

Studies of Workers Exposed to Benzene

Benzene is used in chemical processes often as an intermediate in the manufacture of other chemicals and end products. Occupational exposure to benzene has been studied primarily in industrial workers, including rubber, chemical, and petroleum and gasoline workers. On the basis of human studies of those occupational groups and animal studies over the last 60 years, the allowable occupational health standard for benzene has steadily decreased in the United States. In 1987, the permissible exposure limit (PEL) set by the Occupational Safety and Health Administration (OSHA) was reduced from 10 parts per million (ppm) to a time weighted average over 8 hours (8-hr TWA) of 1 ppm (NIOSH, 1997).

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

TABLE 6.1 Description of Cohort Studies Related to Exposure to Organic Solvents

Reference

Description

Study Group (N)

Comparison Group (N)

Exposure Assessment and Other Relevant Exposures

Analysis and Adjustment for Potential Confounders

Benzene

NIOSH Pliofilm Cohort

Infante et al., 1977

Mortality experience (1940–1975) of white male Pliofilm workers (at least 1 day in 1940–1949) at three Goodyear facilities in Ohio

748

(1) US white male population

Employment in a benzene-exposed occupation as verified through historical air exposure measurements

SMR

Age, time period

(2) 1447 white, male fibrous-glass construction workers in Ohio

Rinsky et al., 1981

Mortality experience (1940–1975) of

 

US white male population

Employment in a benzene-exposed occupation as verified through historical air exposure measurements

SMR

Age, sex, time period

(1) the original cohort and

(1) 748

(2) a second group of white male Pliofilm workers (at least 1 day in 1950–1959)

(2) 258

at three Goodyear facilities in Ohio

Rinsky et al., 1987

Mortality experience (1940–1981) of white male Pliofilm workers (at least 1 day in 1940–1965) at three Goodyear facilities in Ohio

1165

US white male population

Employment in a benzene-exposed occupation as verified through historical air exposure measurements, with cumulative exposure indexes

SMR

Age, time period

Paxton et al., 1994a, 1996

Mortality experience (1940–1987) of white male Pliofilm workers (at least 1 day in 1940–1965) at three Goodyear facilities in Ohio

1212

US white male population

Employment in a benzene-exposed occupation as verified through modified historical air exposure measurements, with cumulative exposure indexes

SMR

Age, time period

Paxton et al., 1994b

Mortality experience (1940–1987) of white male Pliofilm workers (at least 1 day in 1940–1965) at three Goodyear facilities in Ohio

1868

US white male population

Employment in a benzene-exposed occupation as verified through modified historical air exposure measurements, with cumulative exposure indexes

SMR, Cox proportional hazards model

Age, sex, location, time of first Pliofilm employment

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Description

Study Group (N)

Comparison Group (N)

Exposure Assessment and Other Relevant Exposures

Analysis and Adjustment for Potential Confounders

Crump, 1994, 1996

Mortality experience (1940–1987) of white male Pliofilm workers (at least 1 day in 1940–1965) at three Goodyear facilities in Ohio

1717

US white male population

Employment in a benzene-exposed occupation as verified through modified historical air exposure measurements, with cumulative exposure indexes

Life-table analysis

Age, sex

Wong, 1995

Mortality experience (1940–1987) of white male Pliofilm workers (at least 1 day in 1940–1965) at three Goodyear facilities in Ohio

1868

US general population

Employment in a benzene-exposed occupation as verified through historical air exposure measurements, with cumulative exposure indexes

SMR

Age

Chinese Workers Cohort

Yin et al., 1987

Mortality experience (1972–1981) of benzene-exposed workers (at least 6 months) in China

28,460 total

15,643 men

12,817 women

28,257 unexposed

Employment in a benzene-exposed occupation as verified through historical air exposure measurements from factory records

RR, SMR

Age, sex

Yin et al., 1989

Mortality experience (1972–1981) of benzene-exposed workers (at least 6 months) in China

28,460 total

15,643 men

12,817 women

28,257 unexposed

Employment in a benzene-exposed occupation as verified through historical air exposure measurements from factory records

RR, SMR

Age, sex, smoking

Yin et al., 1994

Incidence and mortality experience (1972–1981) of benzene-exposed workers (at least 6 months) in China

28,460 total

15,643 men

12,817 women

28,257 unexposed

Employment in a benzene-exposed occupation as verified through historical air exposure measurements from factory records

SMR, Poisson

Age, sex, time of first employment

Li et al., 1994

Incidence and mortality experience (1972–1987) of benzene-exposed workers (at least 1 day) in China

74,828 total

38,833 men

35,995 women

35,805 unexposed

Employment in a benzene-exposed occupation

RR

Sex

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Description

Study Group (N)

Comparison Group (N)

Exposure Assessment and Other Relevant Exposures

Analysis and Adjustment for Potential Confounders

Yin et al., 1996a,b

Incidence and mortality experience (1972–1987) of benzene-exposed workers (at least 1 day) in China

74,828 total

38,833 men

35,995 women

35,805 unexposed

Employment in a benzene-exposed occupation as verified through historical air exposure measurements from factory records

RR

Age, sex

Hayes et al., 1996

Mortality experience (1972–1987) of benzene-exposed workers (at least 1 day) in China

74,828 total

38,833 men

35,995 women

35,805 unexposed

Employment in a benzene-exposed occupation with cumulative exposure assigned by industrial hygienist from historical records

RR (Poisson), trend analysis

Age, sex

Hayes et al., 1997

Incidence (1972–1987) in benzene-exposed workers (at least 1 day) in China

74,828 total

38,833 men

35,995 women

35,805 unexposed

Employment in a benzene-exposed occupation with cumulative and average exposure assigned by industrial hygienist from historical records

RR (Poisson)

Age, sex

Other Cohort Studies

McMichael et al., 1976

Mortality experience (1964–1973) of male rubber workers (at least 1 day) in four plants in Ohio and Wisconsin

18,903

1968 US male population

Employment at one of four rubber-manufacturing plants

SMR

Age, race

Wilcosky et al., 1984

Cases, age 40–84 years, selected retrospectively from a cohort of active and retired male rubber workers in a plant in Akron, Ohio, in 1964–1973; an age-stratified, 20% random sample from the original cohort served as the control group

NA

1336 (20% of 6678)

Linkage of worker histories to plant solvent-use records; work in process area with known solvent use equates to exposure

Race-specific

ORs

Age

 

Other exposures: trichloroethylene, tetrachloroethylene, toluene, xylenes, naphthas, ethanol, acetone, phenol

 

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Description

Study Group (N)

Comparison Group (N)

Exposure Assessment and Other Relevant Exposures

Analysis and Adjustment for Potential Confounders

Pippard and Acheson, 1985

Mortality experience (1939–1982) of male boot and shoe manufacturers (in 1939) in three tow ns in Great Britain

5017

County general populations

Job title

SMR

Age, time period

 

Other exposures: trichloroethylene, solvents

Wong, 1987a

Mortality experience (1946–1977) of male chemical workers (at least 6 months) in seven US plants

7676

US general population

Job title and employment duration

SMR, Mantel-Haenszel RR

Age, race

Wong, 1987b

Mortality experience (1946–1977) of male chemical workers (at least 6 months) in seven US plants

7676

US general population

Job title and employment duration

SMR, Mantel-Haenszel RR

Age, race

Paci et al., 1989

Mortality experience (1939–1984) of shoe workers (at least 1 day) in Florence, Italy

2013 total

1008 men

1005 women

Italy general population

Plant production records and work histories

SMR

Age, sex, calendar year

Walker et al., 1993

Mortality experience (1940–1982) of shoe-manufacturing workers (at least 1 month in 1940–1979) in Ohio

7814 total

2529 men

5285 women

US general population

Employment at one of two plants

SMR

Age, sex, race, time period

 

Other exposures: MEK, acetone, naphtha, isopropyl alcohol, methanol, ethylene glycol monoethyl ether, xylene

 

Greenland et al., 1994

White, male cases of cancer (multiple sites; died in 1969–1984) and controls in a cohort of transformer-assembly workers in Massachusetts

1821 cases

1202 controls

Internal comparison

Job titles rated for exposure by industrial hygienist

Logistic OR (nested case-control)

 

Other exposures: trichloroethylene, solvents

Age, death year, covariates that altered an estimate >20%

Lagorio et al., 1994

Mortality experience (1981–1992) of self-employed gas-station attendants (in 1980) in Italy

2665 total

2308 men

357 women

Latium region, Italy general population

Environmental survey and duration of employment

SMR

Age, sex

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Description

Study Group (N)

Comparison Group (N)

Exposure Assessment and Other Relevant Exposures

Analysis and Adjustment for Potential Confounders

Heineman et al., 1995

Brain tumor incidence in women (1980–1984) in Shanghai, China

276

Shanghai general population

Job title

SIR

Age

Fu et al., 1996

Mortality experience (1939–1991) of shoe-manufacturing workers in England (1939) and Italy (1950–1984)

6223 total

5220 men

1003 women

England and Italy general populations

Job title

SMR

Age, sex, time period

Schnatter et al., 1996a,b

Cases of lymphohematopoietic cancers (died in 1964–1983) and controls in a cohort of Canadian petroleum-distribution workers

29 cases, matched 1:4

Internal comparison

Industrial hygienist review based on work histories, site characterizations, surveys

Mantel-Haenszel OR (nested case-control)

Smoking, family cancer history, x-ray history

Ireland et al., 1997

Mortality experience (1940–1991) of male US chemical-plant workers (at least 1 day in 1940–1977) in Monsanto company plant in Sauget, IL

4172

Illinois general population

Industrial hygienist exposure estimates based on work records

SMR

Lynge et al., 1997

Incidence experience (1970–1991) in service-station workers (1970) in Denmark and Scandinavia

18,969 total

16,524 men

2445 women

Nation general populations

Job title

SMR

Age, sex

Rushton and Romaniuk, 1997

Cases of leukemia and controls in a cohort of petroleum-distribution workers (1975–1992) in UK

91 cases, matched 1:4

Internal comparison

Measurements factored in occupational hygiene estimates, work histories, job descriptions, fuel compositions

OR (nested case-control), logistic regression

Age, smoking, date of hire, employment duration, socioeconomic status

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Description

Study Group (N)

Comparison Group (N)

Exposure Assessment and Other Relevant Exposures

Analysis and Adjustment for Potential Confounders

Trichloroethylene

Aircraft and Aerospace Workers

Garabrant et al., 1988

Mortality experience (1958–1982) of aircraft-manufacturing workers (at least 1 day) at an aircraft-manufacturing facility in San Diego County, California (with at least 4 years of cumulative company employment)

14,067 total

11,898 men

2169 women

US general population

Employment determined through company work records and interviews

SMR

Age, sex, race, calendar year, duration of employment, year of death

Spirtas et al., 1991

Mortality experience (1952–1982) of aircraft-maintenance workers (at least 1 year in 1952–1956) at Hill Air Force Base in Utah

14,457 total

10,730 men

3727 women

Utah white population

Industrial hygienist assessment from interviews, surveys, hygiene files, position descriptions

SMR, trend analysis

Age, sex, calendar period

 

Other exposures: Stoddard solvent, isopropyl alcohol, trichloroethane, acetone, toluene, MEK, methylene chloride

 

Blair et al., 1998

Incidence and mortality experience (1952–1990) of aircraft-maintenance workers (at least 1 year in 1952–1956) at Hill Air Force Base in Utah

14,457 total

10,730 men

3727 women

Utah white population

Industrial hygienist assessment from interviews, surveys, hygiene files, position descriptions

SMR, RR (Poisson)

Age, sex, calendar period

 

Other exposures: Stoddard solvent, isopropyl alcohol, trichloroethane, acetone, toluene, MEK, methylene chloride

 

Morgan et al., 1998

Mortality experience (1950–1993) of aerospace workers (at least 6 months) Hughes Aircraft plant in Arizona

20,508 total

(4733 exposed)

13,742 men

6766 women

US general population

Exposure matrixes generated by employees and industrial hygienists

SMR, Cox proportional hazards model

Age, sex

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Description

Study Group (N)

Comparison Group (N)

Exposure Assessment and Other Relevant Exposures

Analysis and Adjustment for Potential Confounders

Boice et al., 1999

Mortality experience (1960–1996) of aircraft-manufacturing workers (at least 1 year) Lockheed Martin facility in California

77,965 total

62,477 men

15,488 women

General California population of white workers

Abstracted from walkthrough surveys, hygiene files, job descriptions

SMR, RR (Poisson)

Age, sex, race, dates of first and last employment

 

Other exposures: tetrachloroethylene, solvents

Other Cohort Studies

Axelson et al., 1978

Mortality experience (1955–1975) of Swedish men occupationally exposed during the 1950s and 1960s

518

Sweden general population

Biologic monitoring for U-TCA

RR

Age

Axelson et al., 1994

Mortality experience (1955–1986) of Swedish workers occupationally exposed during the 1950s and 1960s

1670 total

1421 men

249 women

Sweden general population

Biologic monitoring for U-TCA

SMR, SIR (Poisson)

Age, sex, time period

Anttila et al., 1995

Incidence experience (1967–1992) of workers biologically monitored for occupational exposure to halogenated solvents (1965–1983) at the Finnish Institute of Occupational Health

3974 total

2050 men

1924 women

Finland general population

Biologic monitoring for U-TCA, and blood metabolites of tetrachloroethylene and trichloroethane

SIR

Age, sex, time period

 

Other exposures: trichloroethane, tetrachloroethylene

 

Ritz, 1999

Mortality experience (1951–1989) of male uranium-processing plant workers (at least 3 years, with first hire in 1951–1972) in Ohio

3814

(1) External comparison with US general population

(2) Internal comparison among workers monitored for exposure

Exposure matrixes generated by employees and industrial hygienists

SMR, RR (conditional logistic regression)

Age, calendar year, time since first hired, pay type, radiation dose

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Description

Study Group (N)

Comparison Group (N)

Exposure Assessment and Other Relevant Exposures

Analysis and Adjustment for Potential Confounders

Hansen et al., 2001

Incidence experience (1968–1996) in Danish workers (1947–1989) occupationally exposed

803 total

658 men

145 women

Denmark general population

Biologic monitoring for U-TCA

SIR

Age, sex, calendar year, period of first employment, employment duration

Tetrachloroethylene

Dry-cleaning Cohorts

Brown and Kaplan, 1987

Mortality experience (1960–1982) of dry cleaners (at least 1 year, before 1960) in four US labor unions

1690

US general population

Employment in dry-cleaning shops using tetrachloroethylene or other solvents

SMR

Age, time period

Ruder et al., 1994

Mortality experience (1960–1990) of dry cleaners (at least 1 year, before 1960) in four US labor unions

1701 total

592 men

1109 women

SMRs calculated with modified life-table analysis system of NIOSH

Employment in dry-cleaning shops using tetrachloroethylene or other solvents

SMR

Age, sex, time period

Ruder et al., 2001

Mortality experience (1960–1996) of dry cleaners (at least 1 year, before 1960) in four US labor unions

1701 total

592 men

1109 women

SMRs calculated with modified life-table analysis system of NIOSH

Employment in dry-cleaning shops using tetrachloroethylene or other solvents

SMR

Age, sex, time period

Blair et al., 1990

Mortality experience (1948–1978) of members of a dry-cleaning union in St. Louis

5365 total

1319 men

4046 women

US general population

Exposure index created from job title and length of union membership

SMR, trend analysis

Age, sex, calendar year, race

Other Cohort Studies

Lynge and Thygesen, 1990

Incidence experience (1970–1980) of Danish laundry and dry-cleaning workers (in 1970)

10,600 total

2033 men

8567 women

Denmark general population

Dry-cleaning, job title

SIR

Age

 

Other exposures: trichloroethylene, 1,1,2-trichloro-1,2,2-trifluoroethane

 

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Description

Study Group (N)

Comparison Group (N)

Exposure Assessment and Other Relevant Exposures

Analysis and Adjustment for Potential Confounders

Anttila et al., 1995

Incidence experience (1967–1992) of workers biologically monitored for occupational exposure to halogenated solvents (1965–1983) at the Finnish Institute of Occupational Health

3974 total

2050 men

1924 women

Finland general population

Biologic monitoring

SIR

Age, sex, time period

 

Other exposures: trichloroethane, trichloroethylene

 

Boice et al., 1999

Mortality experience (1960–1996) of aircraft-manufacturing workers (at least 1 year) at Lockheed Martin facility in California

77,965 total

62,477 men

15,488 women

General California population of white workers

Abstracted from walkthrough surveys, hygiene files, job descriptions

SMR, RR (Poisson)

Age, sex, race, dates of first and last employment

 

Other exposures: trichloroethylene, solvents

Methylene Chloride

Kodak Park Cohort

Friedlander et al., 1978

PMR of former or current exposed workers (1956–1976) at Kodak Park

334

Deaths of former or current unexposed workers (1956–1976) at Kodak Park

Employment in methylene chloride area

PMR, SMR

Age, sex

Mortality experience (1964–1976) of hourly-wage male workers (in 1964) at Kodak Park

751

Hourly-wage male workers at Kodak Park and in New York state (excluding New York City) men

 

Hearne and Friedlander, 1981

Mortality experience (1964–1980) of hourly-wage male workers (in 1964) at Kodak Park

750

Hourly-wage male workers at Kodak Park and in New York state (excluding New York City) men

Employment in methylene chloride area

SMR

Age, sex

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Description

Study Group (N)

Comparison Group (N)

Exposure Assessment and Other Relevant Exposures

Analysis and Adjustment for Potential Confounders

Hearne et al., 1987

Mortality experience (1964–1984) of hourly-wage male workers (in 1964) at Kodak Park

1013

(1) New York state (excluding New York City) men (1945–1990)

(2) Over 40,000 Rochester-based Kodak workers

Employment in roll-coating division, cumulative exposure assigned by industrial hygienist from historical records

SMR

Age, sex, time period

Hearne et al., 1990

Mortality experience (1964–1988) of hourly-wage male workers (in 1964) at Kodak Park

1013

(1) New York state (excluding New York City) men (1945–1990)

(2) Over 40,000 Rochester-based Kodak workers

Employment in roll-coating division, cumulative exposure assigned by industrial hygienist from historical records

SMR, trend analysis

Age, sex, time period

Hearne and Pifer, 1999

Mortality experience (1946–1994) of two overlapping cohorts of exposed male workers (at least 1 year in 1946–1970; any employment in 1964–1970) at Kodak Park

(1) 1311

(2) 1013

New York state (excluding New York City) men (1945–1990)

Employment in methylene chloride area, cumulative exposure assigned by industrial hygienist from historical records

SMR, trend analysis

Age, sex, time period

Cellulose-Fiber Production Cohort

Ott et al., 1983

Mortality experience (1954–1977) of cellulose-fiber production plant workers (at least 3 months) in Rock Hill, SC

1271 total

551 men

720 women

York County, SC, general population

Employment in plant, comprehensive industrial hygienist survey

SMR, conditional risk, Cox regression

Age, sex, race, year of first exposure

 

Other exposures: acetone, methanol

Lanes et al., 1990

Mortality experience (1954–1986) of cellulose-fiber production plant workers (at least 3 months in 1954–1977) in Rock Hill, SC

1271 total

551 men

720 women

York County, SC, general population

Employment in plant, comprehensive industrial hygienist survey

SMR

Age, sex, race

 

Other exposures: acetone, methanol

 

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Description

Study Group (N)

Comparison Group (N)

Exposure Assessment and Other Relevant Exposures

Analysis and Adjustment for Potential Confounders

Lanes et al., 1993

Mortality experience (1954–1990) of cellulose-fiber production plant workers (at least 3 months in 1954–1977) in Rock Hill, SC

1271 total

551 men

720 women

York County, SC, general population

Employment in plant and comprehensive industrial hygienist survey

SMR

Age, sex, race

 

Other exposures: acetone, methanol

 

Other Cohort Studies

Gibbs et al., 1996

Mortality experience (1970–1989) of cellulosefiber production workers (at least 3 months) in Cumberland, MD

3211 total

2187 men

1024 women

Allegheny County, MD, general population

Workplace monitoring data, job title

SMR

Age, sex, race, time period

Tomenson et al., 1997

Mortality experience (1946–1994) of male cellulose triacetate film-base workers (any employment in 1946–1988) in Brantham, UK

1785

England and Wales mortality rates

Workplace monitoring

SMR, trend analysis

Age, time period

Toluene and Xylene

Swedish Paint Industry Cohort

Lundberg, 1986

Incidence and mortality experience (1955–1981) of male, Swedish paint-industry workers (at least 5 years in 1955–1975) with long-term exposure to organic solvents

416

Sweden general population

Industry employment, historical air exposure measurements

SMR

Sex, time period

Lundberg and MilatouSmith, 1998

Incidence and mortality experience (1955–1994) of male, Swedish paint-industry workers (at least 5 years in 1955–1975) with long-term exposure to organic solvents

411

Sweden general population

Industry employment, historical air exposure measurements

SMR, SIR

Age, sex, time period

Other Cohort Studies

Anttila et al., 1998

Incidence experience (1973–1992) in workers biologically monitored for occupational exposure to aromatic hydrocarbons (1973–1983) at the Finnish Institute of Occupational Health

5301 total

3922 men

1379 women

Finland general population

Biologic monitoring

SIR

Age, sex, time period

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Description

Study Group (N)

Comparison Group (N)

Exposure Assessment and Other Relevant Exposures

Analysis and Adjustment for Potential Confounders

Svensson et al., 1990

Incidence and mortality experience (1925–1985) of Swedish male rotogravure workers (at least 3 months)

1020

Population-specific rates for geographic area around factory

Exposures evaluated through plant visits, biologic monitoring, workplace measurements, historical documents, interviews

SMR, SIR

Age, sex, calendar year, location

 

Other exposures: benzene, solvents

 

Solvents

UK Rubber Worker Cohort

Parkes et al., 1982

Mortality experience (1946–1975) of UK male rubber workers (at least 1 year in 1946–1960)

33,815

UK general population

Industry employment

SMR

Age, sex

Sorahan et al., 1986

Mortality experience (1946–1980) of UK male rubber workers (at least 1 year in 1946–1960)

36,445

UK general population

Industry employment

SMR, regression models and life tables

Age, sex, age at hire, entry cohort, location, work sector

Sorahan and Cathcart, 1989

Mortality experience (1946–1985) of UK male rubber workers (at least 1 year in 1946–1960)

36,691

UK general population

Industry employment

SMR, regression models and life tables

Age at hire, entry cohort, location, work sector, duration of employment

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Description

Study Group (N)

Comparison Group (N)

Exposure Assessment and Other Relevant Exposures

Analysis and Adjustment for Potential Confounders

Other Cohort Studies

Costantini et al., 1989

Mortality experience (1950–1983) of male workers at tanneries (at least 6 months) in Tuscany, Italy

2926

Italy general population

6 months of employment

SMR

Age, time period

Guberan et al., 1989

Incidence and mortality experience (1970–1984) of painters and electricians (in 1970) in the Canton of Geneva

1916 painters

1948 electricians

Switzerland regional population

Job title

SMR, SIR

Age

Acquavella et al., 1993

Mortality experience (1950–1987) of workers hired at a metal components manufacturing facility (at least 6 months in 1950–1967) in Iowa

3630 total

2664 men

966 women

Iowa general population

Occupational titles, departments

SMR, RR

Age, sex, time period

Berlin et al., 1995

Incidence and mortality experience (1967–1987) of Swedish workers occupationally exposed to solvents

5791 total

5283 men

508 women

Sweden general population

Patients with solvent-related disorders

SMR, SIR

Age, sex

Lynge et al., 1995

Incidence experience (1970–1987) in Danish printing workers (in 1970)

19,127 total

15,534 men

3593 women

Economically active people in Denmark

Job title

SIR

Age, alcohol and tobacco use

Steenland and Palu, 1999

Mortality experience (through 1994) of members of US painters unions (at least 1 year of membership; born before 1940) in four states

42,170

US general population, nonpainter cohort

Union membership

SMR, SRR

Age, time period

Other Specific Solvents

Isopropyl Alcohol and Methyl Ethyl Ketone

Alderson and Rattan, 1980

Mortality experience (1935–1975) of workers in Shell MEK dewaxing or isopropyl alcohol plants (at least 1 year) in Britain

262 IAP

446 MEK

US general population

Employment in one of two plants

SMR

Age, time period

Phenol

Dosemeci et al., 1991

Mortality experience (1966–1979) of white male workers (employed before 1966) employed at five facilities producing or using phenol and formaldehyde

14,861

US general population

Employment at one of five facilities

SMR

Age, time period

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Description

Study Group (N)

Comparison Group (N)

Exposure Assessment and Other Relevant Exposures

Analysis and Adjustment for Potential Confounders

Ethanol and Isopropyl Alcohol

Teta et al., 1992

Mortality experience (1940–1983) of male isopropanol and ethanol production workers (1940–1978) at two facilities in South Charleston, WV, and Texas City, TX

1031

(1) White subjects and South Charleston cohort compared with general US white male population

(2) Nonwhite subjects from Texas compared with US nonwhite male population

Employment in one of two plants

SMR

Age, time period, duration of assignment, time since first assignment, year of first assignment, type of assignment

NOTE: SMR=standardized mortality ratio; RR=relative risk; SIR=standardized incidence ratio; OR=odds ratio; PMR=proportional mortality ratio; SRR=standardized relative risk; NIOSH=National Institute for Occupational Safety and Health; IAP=isopropyl alcohol; MEK=methyl ethyl ketone; U-TCA=urinary metabolite of trichloroethylene.

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Although the association between benzene and cancer was assessed in a number of cohort studies, two studies provide the most comprehensive data. The first was designed to investigate mortality in three Ohio rubber-manufacturing plants, referred to as the “Pliofilm” cohort (Crump, 1994, 1996; Infante et al., 1977; Paxton, 1996; Paxton et al., 1994a,b; Rinsky et al., 1987; Wong, 1995). Benzene was used in the production of rubber hydrochloride, a natural rubber cast film used primarily for wrapping foods and marketed under the trade name Pliofilm. The plants were chosen because of the relatively high exposure to benzene and the lack of other toxic chemicals in use. In 1975, after a report of several leukemia cases, the National Institute for Occupational Safety and Health conducted a retrospective study of 748 Pliofilm workers who were exposed to benzene (Infante et al., 1977). Rinsky and colleagues (1981, 1987) expanded on the work of Infante and colleagues by increasing the size of the cohort to 1006 workers, extending the years of observation, and collecting additional exposure data from the plants’ processes, company records, and air-sampling data (Rinsky et al., 1981), thereby providing estimates of exposure for each job in the various areas of the plant (Rinsky et al., 1987). The retrospective exposure assessment was further modified by Crump and Allen (1984) and Paustenbach and colleagues (1992). Crump and Allen (1984) developed an exposure matrix based on the concept that the benzene levels in the workplace may have improved over time, whereas the exposure matrix developed by Paustenbach and colleagues (1992) incorporated more detailed information from monitoring devices, the changing length of the workweek over the years, the impact of World War II on production, engineering controls, and other available and experimental data to assess exposure to benzene. Considerable controversy surrounds the assessments of exposure (Wong, 1995). Different authors have used the different exposure assessments in their analyses: the Rinsky exposure assessment (Paxton et al., 1994a,b; Rinsky et al., 1981, 1987), the Crump assessment (Paxton et al., 1994a,b), and the Paustenbach assessment (Crump, 1994; Paxton et al., 1994a,b). The differences in the exposure assessments lead to differences in the estimates of relative risks for various sites of cancer according to exposure to benzene. The differences in relative risk are important in setting regulatory standards but were not sufficiently different to affect the committee’s determination of the magnitude of association.

The second key study of occupational exposure to benzene was conducted in China. After conducting a nationwide benzene-monitoring survey in China during 1979–1981, Yin and colleagues at the Chinese Academy of Preventive Medicine identified a cohort of about 30,000 workers who were occupationally exposed to benzene or mixtures containing benzene (Yin et al., 1987). Subjects were selected from painting, shoe-making, rubber synthesis, leather, and adhesive and organic-chemical synthesis factories. A sample of 28,257 workers employed in machine production, textile, and cloth factories was taken to represent an unexposed comparison population. Later studies expanded the original cohort to 74,828 benzene-exposed and 35,805 nonexposed workers and included a detailed assessment of exposure to benzene. Those studies were conducted in collaboration with the US National Cancer Institute (Hayes et al., 1996, 1997; Li et al., 1994; Yin et al., 1987, 1989, 1994, 1996a,b). Like the Pliofilm study, studies of the cohort yielded valuable information regarding the risk of developing or dying from cancer in relation to exposure to benzene.

Other important cohort studies of benzene-exposed workers include those of American chemical workers (Ireland et al., 1997; Wong, 1987a,b), female workers in China

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

(Heineman et al., 1995), other rubber-plant workers (McMichael et al., 1976; Wilcosky et al., 1984), shoe-manufacturing workers (Fu et al., 1996; Paci et al., 1989; Pippard and Acheson, 1985; Walker et al., 1993), transformer-assembly workers (Greenland et al., 1994), filling and service-station attendants (Lagorio et al., 1994, Lynge et al., 1997), and petroleum distributors (Rushton and Romaniuk, 1997; Schnatter et al., 1996a,b). These studies differed substantially from the two preceding studies in that exposure was much lower. For example, levels in the Pliofilm cohort ranged from 7.2 to 24.9 ppm (Rinsky et al., 1981, 1987), whereas levels in the cohort of petroleum workers ranged from 0.01 to 6.2 ppm (Ireland et al., 1997; McMichael et al., 1976; Rushton and Romaniuk, 1997; Schnatter et al., 1996a,b).

Studies of Workers Exposed to Trichloroethylene

The most important use of trichloroethylene has been in the removal of greases, tars, and oils from metal parts. It has also been used by the textile industry to scour cotton, wool, and other fabrics and as a solvent in waterless dying and finishing operations (ATSDR, 1997b). The regulatory limit set by OSHA is a 8-hr TWA of 100 ppm (NIOSH, 1997).

Mortality in relation to exposure to trichloroethylene has been examined in four large cohort studies of aircraft and aerospace manufacturing and maintenance workers (Blair et al., 1998; Boice et al., 1999; Garabrant et al., 1988; Morgan et al., 1998). In general, industrial hygienists reviewed information obtained from walkthrough surveys, interviews of long-term employees, and historical information on job titles and tasks, operations, and worksites to classify workers by duration and intensity of exposure. The first study was conducted to evaluate mortality rates among 14,457 aircraft maintenance workers at Hill Air Force Base, Utah, in response to concerns expressed by workers in the middle-1970s about potential health effects of chemical exposure (Spirtas et al., 1991). Trichloroethylene was used as a vapor degreaser until 1978, when it was replaced with 1,1,1-trichloroethane. It was also used to clean small electric parts at work benches until 1968 (Blair et al., 1998). Other solvents used at the base were primarily other chlorinated hydrocarbons, aromatic hydrocarbons, and carbon tetrachloride. Blair and co-workers (1998) extended the followup of the cohort assembled by Spirtas and colleagues (1991) by 8 years. The estimates of exposure developed for the initial cohort study (Stewart et al., 1991) were also used in the extended followup study.

Two additional large cohort studies evaluated mortality in aircraft manufacturing facilities where trichloroethylene was commonly used as a degreaser. Boice and colleagues (1999) investigated 77,965 workers at Lockheed Martin’s Burbank California factories, and Morgan and co-workers (1998) reported on 20,508 employees at a Hughes Aircraft manufacturing facility in Arizona.

Other important but smaller cohort studies of workers exposed to trichloroethylene are those of Swedish trichloroethylene production workers (Axelson et al., 1978, 1994), US uranium processing-plant workers (Ritz, 1999), and other workers occupationally exposed to trichloroethylene in Finland (Anttila et al., 1995) and Denmark (Hansen et al., 2001). Exposure-response analyses in the Scandinavian studies were based on biologic monitoring of urinary-trichloroacetic acid (U-TCA, a metabolite of trichloroethylene) measured in urine samples from workers (Anttila et al., 1995; Axelson et al., 1978, 1994; Hansen et al., 2001). Hansen and colleagues (2001) also used data on levels of trichloroethylene in the breathing

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

zone of workers. The US study used semiquantitative exposure estimates based on expert review (Ritz, 1999).

Studies of Workers Exposed to Tetrachloroethylene and Dry-cleaning Solvents

Tetrachloroethylene has been used for metal cleaning and vapor degreasing and for dry-cleaning and textile processing. The PEL set by OSHA is a 8-hr TWA of 100 ppm (NIOSH, 1997).

Occupational exposure to tetrachloroethylene has been studied primarily in dry-cleaning workers because of its widespread use. Dry-cleaning workers are extensively exposed to organic solvents, which are integral to the dry-cleaning process. The evolution of the dry-cleaning process has seen the use of a variety of solvents. Most of the early dry-cleaning solvents were petroleum-based and included naphtha and Stoddard solvent. The petroleum-based solvents were replaced in the 1930s largely with carbon tetrachloride, a less expensive alternative (IARC, 1995). Information about the toxicity and corrosiveness of carbon tetrachloride led to its replacement in the 1950s with chlorinated hydrocarbons. Today, tetrachloroethylene is the most commonly used dry-cleaning solvent in the United States. Other solvents and chemicals used in dry-cleaning include 1,1,2-trichloro-1,2,2-trifluoroethane, and 1,1,1-trichloroethane (IARC, 1995).

An important study examined a cohort of 1708 US dry-cleaning workers drawn from four labor unions, first reported by Brown and Kaplan (1987), and updated by Ruder and colleagues (1994, 2001). The original study investigated mortality through 1982, the first update extended the followup through 1990 (Ruder et al., 1994), and the most recent study updated mortality through 1996 (Ruder et al., 2001). In the updates, two subcohorts were evaluated on the basis of employment in shops where tetrachloroethylene was the cleaning solvent (625 workers) or in shops where tetrachloroethylene use could not be confirmed or another solvent was used as the cleaning solvent (1083 workers).

Another study, of 5365 members of a dry-cleaning union in Missouri, assessed mortality in relation to estimated levels of exposure to dry-cleaning solvents (Blair et al., 1990). Exposure indexes were based on job title and type of establishment. Information on the type of solvent used was not available so workers who specifically used tetrachloroethylene could not be identified.

Other studies of exposure to tetrachloroethylene and cancer include a US study of aircraft manufacturing workers (Boice et al., 1999) and a Finnish study of workers occupationally exposed to three halogenated hydrocarbons, including tetrachloroethylene (Anttila et al., 1995). The exposure assessment of the US study was based on expert review of walkthrough surveys and historical documents and other approaches (Boice et al., 1999), whereas the Finnish study estimated level of exposure from biologic monitoring (Anttila et al., 1995). Another study examined Danish laundry and dry-cleaning workers whose chemical exposure was inferred from their occupations, as specified by census industry codes (Lynge and Thygesen, 1990).

Studies of Workers Exposed to Methylene Chloride

Methylene chloride (dichloromethane) has been used in degreasing, in paint stripping, as an aerosol propellant, and in the manufacture of textiles, plastics, and photographic film. A large proportion of workers exposed to methylene chloride are

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

involved in metal cleaning, industrial paint stripping, and using ink solvents (ATSDR, 2000). The regulatory limits have decreased as information on toxicity has accumulated (ATSDR, 2000). The current PEL set by OSHA is a 8-hr TWA of 25 ppm (NIOSH, 1997).

The key occupational cohort study of exposure to methylene chloride was an incidence and mortality study of workers in an Eastman Kodak plant (Friedlander et al., 1978; Hearne and Friedlander, 1981; Hearne and Pifer, 1999; Hearne et al., 1987, 1990). The workers, ranging in number from 750–1311, were exposed chronically to methylene chloride in the manufacturing of cellulose triacetate, a photographic film base, as confirmed by study personnel who used air sampling and gas chromatography. In the most recent update, Hearne and Pifer (1999) followed the mortality experience through 1994 for two groups of workers: 1311 workers who first worked in film-support manufacturing and related operations in 1946–1970 and the Roll Coating Cohort (1964–1970) of 1013 employees that was previously studied (Hearne et al., 1987, 1990). Exposure to methylene chloride was estimated by combining air-monitoring data with information on work histories.

Other important studies include a study of workers employed at a plant that produced cellulose triacetate film base in the UK (Tomenson et al., 1997), and studies of workers in the production of cellulose fiber at a Hoechst manufacturing plant in South Carolina (Lanes et al., 1990, 1993; Ott et al., 1983) and a Hoechst plant in Maryland (Gibbs et al., 1996). Estimates of exposure in the UK cohort were derived from area-monitoring results, work histories, and historical information on production processes (Tomenson et al., 1997). Some exposure-monitoring data were obtained on the South Carolina cohort in the 1970s (Ott et al., 1983), but exposure estimates were unavailable for most of the study period. Exposure measurements were not used in the mortality analysis of the South Carolina cohort (Lanes et al., 1990, 1993), but Gibbs and co-workers (1996) used the available exposure data to determine high and low exposure ranges.

Studies of Workers Exposed to Other Specific Solvents

Three studies of solvent-production plants were used to evaluate mortality in relation to exposure to specific solvents: isopropanol (Alderson and Rattan, 1980; Teta et al., 1992), methyl ethyl ketone (Alderson and Rattan, 1980), phenol (Dosemeci et al., 1991), and toluene, xylene, and styrene (Anttila et al., 1998). The main uses of isopropanol are as a chemical intermediate and in applications in medicine and industry (Logsdon and Loke, 1996). The PEL is a 8-hr TWA of 400 ppm (NIOSH, 1997). Methyl ethyl ketone is used primarily as a solvent in industry. The regulatory limit set by OSHA is 200 ppm (NIOSH, 1997). Phenol is commonly used in the production of epoxy resins and polycarbonates, phenolic resins and molding compounds, caprolactam, aniline alkylphenols, and xylenols; as a fungicide or disinfectant; and in a variety of medications (ATSDR, 1998; Wallace, 1996). The occupational exposure limit set by OSHA is a 8-hr TWA of 5 ppm (NIOSH, 1997). Toluene and xylene are used in the manufacture of a variety of chemicals and as solvents for paints, lacquers, gums, printing inks, and resins. Styrene is used in the production of polystyrene plastics and resins and as an intermediate in the production of such copolymers as styrene-acrylonitrile, acrylonitrile-butadiene-styrene, and styrene-butadiene rubber. The occupational regulatory limit for toluene is a 8-hr TWA of 200 ppm (NIOSH, 1997), and the PEL for xylene and styrene is a 8-hr TWA of 100 ppm (NIOSH, 1997).

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

The key studies for evaluating risks posed by those solvents include the following. Alderson and Rattan (1980) evaluated mortality in 262 male workers employed in isopropanol plant and 446 male employees of two methyl ethyl ketone dewaxing plants, using the type of plant as an indicator of exposure. Teta and colleagues (1992) conducted a cohort mortality study of 1031 workers employed at two facilities that produced ethanol and isopropanol. Employment in an isopropanol strong-acid production unit was used as an exposure surrogate for isopropanol. Dosemeci and co-workers (1991) conducted a mortality followup of 14,861 workers employed in five plants that manufactured or used phenol and formaldehyde. Estimates of exposure to phenol were derived from expert review of information obtained from walkthrough survey reports, historical monitoring results, and other workplace information. Anttila and colleagues (1998) investigated 3922 male and 1379 female Finnish workers occupationally exposed to toluene, xylene, and styrene; level of exposure was determined from biologic monitoring (Anttila et al., 1995).

Studies of Workers Exposed to Unspecified Mixtures of Organic Solvents

Solvents are used in numerous occupations, so the committee examined cancer mortality and incidence in workers in a number of occupations that may have involved exposure to organic solvents. It is important to note that workers in the occupations in question have potential exposure to numerous chemicals in addition to solvents.

Painters have the potential for frequent and high level exposure to many types of organic solvents. Organic solvents, such as, toluene, xylene, glycols, and methylene chloride, have been used over the years in the composition of paint, paint thinners, cleaners, and strippers. Painters are exposed to numerous other chemical and environmental agents, including pigments, dusts, resins, and silicates. In most instances, it was not possible to identify which solvents were used, and the committee referred to them as unspecified mixtures of organic solvents. Lundberg and Milatou-Smith (1998) evaluated cancer mortality and incidence in a cohort of 411 male workers who had been employed for more than 5 years in 1955–1975 in the Swedish paint-manufacturing industry (followup of Lundberg, 1986). Guberan and colleagues (1989) studied cancer mortality and incidence in 1916 painters in Geneva, Switzerland, who were identified from the 1970 census. The largest cohort study of painters was conducted by Steenland and Palu (1999), who evaluated mortality patterns in a cohort of 42,170 painters who were members of the Painters Union for 1 year or more before 1979.

Other cohorts exposed to unspecified chemical mixtures involve printers (Lynge et al., 1995; Svensson et al., 1990); workers in solvent-production plants (Berlin et al., 1995); metal workers (Acquavella et al., 1993); rubber workers (Parkes et al., 1982; Sorahan and Cathcart, 1989; Sorahan et al., 1986); workers in tanneries (Costantini et al., 1989); and shoemakers (Fu et al., 1996; Paci et al., 1989; Pippard and Acheson, 1985; Walker et al., 1993). In the past, some shoemaking cohorts would have had considerable exposure to benzene (e.g., Fu et al., 1996; Paci et al., 1989). However, because the composition of glues has changed, benzene being replaced with other solvents, and because there were no precise estimates of exposure, the committee classified those cohorts as being exposed to unspecified mixtures of organic solvents.

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

ORAL, NASAL, AND LARYNGEAL CANCER

Description of Case-Control Studies

Two case-control studies reviewed by the committee that included oral, nasal, or laryngeal cancers are described in Table 6.2. One population-based case-control study examined the risk of nasal and nasopharyngeal cancer associated with occupational exposure to organic solvents (Hardell et al., 1982), and a second examined the risk of oral cavity and laryngeal cancers associated with work in the dry-cleaning industry (Vaughan et al., 1997). Both included interviews with study subjects concerning occupational history. In the former study, exposure to organic solvents was self-reported; in the latter, levels of exposure to tetrachloroethylene in cleaning jobs were assigned by an industrial hygienist.

Epidemiologic Studies of Exposure to Organic Solvents and Oral Cancer

Three cohort studies and one case-control study failed to provide strong evidence of an association between tetrachloroethylene and dry-cleaning solvents and oral cancer. Ruder and colleagues (2001) found a strong, increased risk of cancer of the tongue (standardized mortality ratio [SMR]=9.03, 95% confidence interval [CI]=1.86–26.39) in the subcohort exposed only to tetrachloroethylene. In another cohort of dry cleaners, Blair and colleagues (1990) found no increased risk of cancers of the buccal cavity and pharynx (SMR=1.0, 95% CI=0.3–2.2).

The case-control study undertaken by Vaughan and colleagues (1997) found little evidence of an increased risk of oral cancer in dry-cleaning workers (odds ratio [OR]possible exposure=1.2, 95% CI=0.3–4.6; ORprobable exposure=1.5, 95% CI=0.2–9.5). The relative risks increased with increasing cumulative exposure but not with duration of employment, although there was considerable statistical uncertainty in the trend.

In only one study was the risk of cancers of the mouth and throat evaluated among workers exposed to phenol (Dosemeci et al., 1991); no increased risk was found (SMR=0.8, 95% CI=0.4–1.5).

The risk of oral cancer in industries exposed to solvents was estimated in cohorts of workers in cellulose-fiber production (Lanes et al., 1990), shoe manufacture (Walker et al., 1993), methyl ethyl ketone dewaxing (Alderson and Rattan, 1980), and ethanol and isopropanol production (Teta et al., 1992). Although specific solvents were used in most of those occupations, many other solvents were also used, and specific solvents were not evaluated in any of the studies. There was no evidence of positive associations in any of these studies.

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

TABLE 6.2 Description of Case-Control Studies of Oral, Nasal, and Laryngeal Cancer and Exposure to Organic Solvents

Reference

Description of Study Population

Number of Cases

Number of Controls

Relevant Exposure(s)

Determination of Exposure

Analysis

Adjustment for Confounding

Hardell et al., 1982

Male cases, age 28–85 years, reported to the Swedish Cancer Registry in 1970–1979 from the three northernmost counties of Sweden; controls from authors’ previous studies on soft tissue sarcoma and malignant lymphoma were used; controls identified from the Swedish National Population Registry and the National Registry for Causes of Death

44 nasal

27 nasopharyngeal

541

Organic solvents

Occupational histories (job titles) and exposures (self-reported) obtained through questionnaire and supplemented by telephone interview (direct or proxy)

Exposure frequency

None

Vaughan et al., 1997

Cases, age 20–74 years when diagnosed, identified from a cancer surveillance system covering 13 counties in western Washington; cases lived in one of the three largest counties and were diagnosed in 1983–1990; population-based controls, frequency-matched by age and sex, identified through RDD

491 oral cavity

235 laryngeal

724

Tetrachloroethylene

Dry-cleaning work

In-person interviews to assess employment and duration in dry-cleaning occupations; probability of exposure from decade of employment; cumulative exposure measurements calculated by duration and occupation-specific time-weighted averages

Conditional logistic regression

Age, sex, education, study period, alcohol consumption, cigarette smoking, race

Response rates: 85.2% of oral cavity cancer cases, 80.8% of laryngeal cancer cases, 80.3% of controls

NOTE: RD=random-digit dialing.

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Summary and Conclusion

There was little consistent evidence of an association between oral cancer and exposure to tetrachloroethylene and dry-cleaning solvents. Although there were several positive studies, most were based on small numbers of exposed cases and did not have sufficient statistical power. For exposure to phenol, only one study was identified and a risk of oral cancer was not found. No risk was also found among the occupational studies on solvent mixtures. Table 6.3 identifies the studies reviewed on oral cancer. Unless indicated in the table, the study populations include both men and women.

The committee concludes, from its assessment of the epidemiologic literature, that there is inadequate/insufficient evidence to determine whether an association exists between chronic exposure to solvents under review and oral cancer.

TABLE 6.3 Selected Epidemiologic Studies—Oral Cancer and Exposure to Organic Solvents

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Specific and Unspecified Mixtures of Organic Solvents

Cohort Studies—Mortality

Ruder et al., 2001

Dry-cleaning union workers exposed to tetrachloroethylene

 

 

Cancer of the tongue

3

9.03 (1.86–26.39)

Walker et al., 1993

Ohio shoe-manufacturing employees

 

 

Males and females

5

0.67 (0.22–1.59)

Teta et al., 1992

Male workers at ethanol/isopropanol production plants

 

 

South Charleston

2

1.3 (0.2–4.8)a

 

Texas City

1

1.4 (0.0–8.4)a

Dosemeci et al., 1991

Male industrial workers exposed to phenol

11

0.8 (0.4–1.5)

Blair et al., 1990

St. Louis, MO, dry-cleaning workers

5

1.0 (0.3–2.2)

Lanes et al., 1990

Cellulose-fiber production workers

2

2.31 (0.39–7.60)

Alderson and Rattan, 1980

Male British workers at two methyl ethyl ketone dewaxing plants

2

15.38 (1.86–55.54)a

Case-Control Study

Vaughan et al., 1997

Oral cancer among dry-cleaning workers

 

 

Possible exposure to tetrachloroethylene

7

1.2 (0.3–4.6)

 

Probable exposure to tetrachloroethylene

4

1.5 (0.2–9.5)

 

Cumulative exposure to tetrachloroethylene (ppm-years)

 

 

1–29 ppm-years

3

1.0 (0.1–7.0)

 

30+ ppm-years

4

1.4 (0.2–8.7)

 

Duration of employment

 

 

1–9 years

6

1.4 (0.3–5.7)

 

10+ years

1

0.4 (0.0–31.6)

aRisk estimate and 95% CI calculated by the committee using standard methods from the observed and expected numbers presented in the original study.

Epidemiologic Studies of Exposure to Organic Solvents and Nasal Cancer

Few studies with sufficient numbers of cases to assess the relationship between exposure to benzene and oral cancer were available. The cohort study by Yin and colleagues (1996a) of

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

benzene-exposed workers in China showed an increased risk of nasopharyngeal cancer among male workers (relative risk [RR]=2.1, 95% CI=0.7–9.3). The RR was 2.4 (95% CI=0.8–10.5).

A cancer mortality study of shoe-manufacturing workers by Fu and colleagues (1996) analyzed the risk associated with exposure to solvents (found mostly in glues) and leather dusts in two cohorts of shoemakers in England and Florence, Italy. Risk of nasal cancer from occupational exposures was strongly increased (SMR=7.41, 95% CI=3.83–12.94) in the English cohort (only one death from nasal cancer was found in the Florence cohort). Exposure was assessed by using job titles, and no specific exposures were identified. It is not clear whether solvents, leather dust, or other agents contributed to the increased risk of nasal cancer.

Hardell and colleagues (1982) conducted a case-control study of nasal and nasopharyngeal cancers and exposure to various agents, including solvents. Although they provided no estimates of relative risk, the committee calculated relative risks and CIs from the raw data provided and found weak associations with exposure to high-grade organic solvents (nasal: OR=1.24, 95% CI=0.51–2.91; nasopharyngeal: OR=1.27, 95% CI=0.47–3.68).

Summary and Conclusion

The only study on exposure to benzene and risk of nasal cancer had a highly imprecise estimate of effect. Other studies are needed to determine whether an association exists. For exposure to solvent mixtures, the English shoemaker study showed a strong association. However, the Swedish case-control study (Hardell et al., 1982) did not corroborate those findings. Table 6.4 identifies the studies reviewed by the committee on nasal cancer. Unless indicated in the table, the study populations include both men and women.

The committee concludes, from its assessment of the epidemiologic literature, that there is inadequate/insufficient evidence to determine whether an association exists between chronic exposure to solvents under review and nasal cancer.

TABLE 6.4 Selected Epidemiologic Studies—Nasal Cancer and Exposure to Organic Solvents

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Specific and Unspecified Mixtures of Organic Solvents

Cohort Studies—Mortality

Yin et al., 1996a

Chinese workers exposed to benzene

 

 

Male

12

2.1 (0.7–9.3)

 

Total

14

2.4 (0.8–10.5)

Fu et al., 1996

Shoemakers in England and Florence

 

 

English cohort

12

7.41 (3.83–12.94)

 

Florence cohort

1

9.09 (0.23–50.65)

Alderson and Rattan, 1980

Male British workers at an isopropyl alcohol plant

1

50.0 (1.3–278.5)a

Case-Control Study

Hardell et al., 1982

Male cases exposed to high-grade organic solvents

 

 

Nasal

8

1.24 (0.51–2.91)a

 

Nasopharyngeal

5

1.27 (0.47–3.68)a

aRisk estimate and 95% CI calculated by the committee using standard methods from the observed and expected numbers presented in the original study.

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Epidemiologic Studies of Exposure to Organic Solvents and Laryngeal Cancer

In a cohort of dry cleaners, Blair and colleagues (1990) found 60% excess mortality from laryngeal cancers (SMR=1.6, 95% CI=0.3–4.7). In a case-control study of laryngeal cancer, Vaughan and colleagues (1997) found an association with possible exposure to tetrachloroethylene among dry cleaners (OR=2.3; 95% CI=0.5–10.2) but not among those probably exposed (OR=0.9, 95% CI=0.1–12.9). Risk increased with duration of employment in the dry-cleaning industry but not with increasing cumulative exposure.

Cohorts of workers in ethanol and isopropanol production (Teta et al., 1992) and shoe manufacture (Walker et al., 1993) were evaluated for their cancer mortality, and there was little evidence of an association (total of three exposed deaths).

Summary and Conclusion

For exposure to tetrachloroethylene and dry-cleaning solvents and risk of laryngeal cancer, both studies’ findings were based on very few exposed cases, and this resulted in imprecise estimates of relative risk. Similarly, the studies on solvent mixtures were limited by the small number of exposed cases and lack of positive findings. As a result, there was insufficient evidence to conclude that there were associations between exposure to specific solvents or solvent mixtures and laryngeal cancer. The studies reviewed by the committee are identified in Table 6.5. Unless indicated in the table, the study populations include both men and women.

The committee concludes, from its assessment of the epidemiologic literature, that there is inadequate/insufficient evidence to determine whether an association exists between chronic exposure to solvents under review and laryngeal cancer.

TABLE 6.5 Selected Epidemiologic Studies—Laryngeal Cancer and Exposure to Organic Solvents

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Specific and Unspecified Mixtures of Organic Solvents

Cohort Studies—Mortality

Walker et al., 1993

Ohio shoe manufacturing employees

 

 

Females

2

3.34 (0.40–12.09)

Teta et al., 1992

Male workers at ethanol/isopropanol production plants

 

 

South Charleston

1

1.4 (0.0–8.0)a

 

Texas City

1

3.3 (0.1–18.6)a

Blair et al., 1990

St. Louis, MO, dry-cleaning workers

3

1.6 (0.3–4.7)

Case-Control Study

Vaughan et al., 1997

Laryngeal cancer among dry-cleaning workers

 

 

Possible exposure to tetrachloroethylene

4

2.3 (0.5–10.2)

 

Probable exposure to tetrachloroethylene

1

0.9 (0.1–12.9)

 

Cumulative exposure to tetrachloroethylene (ppm-years)

 

 

1–29 ppm-years

2

2.0 (0.2–17.9)

 

30+ ppm-years

2

2.5 (0.3–19.1)

 

Duration of employment

 

 

1–9 years

3

1.9 (0.3–10.2)

 

10+ years

2

5.5 (0.4–75.0)

aRisk estimate and 95% CI calculated by the committee with standard methods from the observed and expected numbers presented in the original study.

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

GASTROINTESTINAL TRACT TUMORS

Description of Case-Control Studies

Several case-control studies were used to evaluate the risks of cancer at gastrointestinal sites in relation to occupational exposures (Table 6.6), and one study was used to assess the risk of colorectal and pancreatic cancers posed by exposure to tetrachloroethylene in drinking water (Paulu et al., 1999).

All the studies but one used interviews with subjects to assess occupational history and in some cases occupational exposures; one study used self-administered questionnaires of next of kin (Kauppinen et al., 1995). The response rates in the latter study were 50% or less, so the results were unlikely to be representative of the entire study population. There were four reports (Dumas et al., 2000; Gérin et al., 1998; Goldberg et al., 2001; Parent et al., 2000) of one multisite case-control study conducted in Montreal in the 1980s. The novel features of the study include use of a mixed control population (cancer and population controls), in-depth interviews to obtain details of each job of each subject, translation of the interviews by a team of industrial hygienists and chemists into semiquantitative indexes of exposure to about 300 physical and chemical agents, and good information on potential confounding factors. Ekstrom and colleagues (1999) investigated gastric cancer and also used experts to attribute exposure on the basis of questionnaires. The study by Paulu and colleagues (1999) of colorectal and pancreatic cancers used estimates of exposure to tetrachloroethylene in drinking water.

Alcohol is a risk factor for esophageal cancer, and this requires consideration in evaluating the association between solvent exposure and esophageal cancer. Two studies (Parent et al., 2000; Vaughan et al., 1997) considered this confounding variable, and one did not (Gérin et al., 1998). Risk factors for other gastrointestinal cancers are less well defined.

Epidemiologic Studies of Exposure to Organic Solvents and Esophageal Cancer

The risk of esophageal cancer was increased in Danish workers who were biologically monitored for a urinary metabolite of trichloroethylene (standardized incidence ratio [SIR]=4.2, 95% CI=1.5–9.2) (Hansen et al., 2001). There was no gradient in risk with cumulative exposure although fairly high relative risks were found for long duration of employment (SIR=6.6, 95% CI=1.8–17). An equivalent risk estimate and 95% CI were also found among workers with high cumulative exposure. Blair and colleagues (1998) reported an excess risk of esophageal cancer mortality among white male aircraft-maintenance workers exposed to trichloroethylene (SMR=5.6, 95% CI=0.7–44.5). In a cohort of aircraft-manufacturing workers in California, Boice and colleagues (1999) found no association between esophageal cancer and potential exposure to trichloroethylene (SMR=0.83, 95% CI=0.34–1.72).

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

TABLE 6.6 Description of Case-Control Studies of Gastrointestinal Tract Tumors and Exposure to Organic Solvents

Reference

Description of Study Population

Number of Cases

Number of Controls

Relevant Exposure(s)

Determination of Exposure

Analysis

Adjustment for Confounding

Canadian studies

Gérin et al., 1998

Male cases and controls, age 35–70 years, diagnosed in 19 large Montreal-area hospitals in 1979–1985 and histologically confirmed for one of 19 anatomic cancer sites; frequency matched by approximate age, population-based controls were also chosen from electoral lists and with random-digit dialing (see also Dumas et al., 2000; Goldberg et al., 2001; Parent et al., 2000)

99 esophageal

251 stomach

497 colon

257 rectal

116 pancreas

1,066 subjects for each site, consisting of 533 population controls and 533 randomly selected subjects from the eligible cancer control group

Benzene

Toluene

Xylene

In-person interviews with specific questions on details of each job subject had; analyzed and coded by a team of chemists and industrial hygienists (about 300 exposures) on semiquantitative scales

Unconditional logistic regression

Age, family income, ethnicity, cigarette smoking, respondent status

Response rates: 82% of all cases, 71% of population controls

Dumas et al., 2000

Same as above (see also Gérin et al., 1998; Goldberg et al., 2001; Parent et al., 2000)

257 rectal

1,295 cancer

533 population

Toluene

Xylene

Methylene chloride

Trichloroethylene

Acetone

See above

See above

Age, education respondent status, cigarette smoking, beer consumption body mass index

Response rates: 84.5% of cases, 72% of population

 

 

Parent et al., 2000

Same as above (see also Dumas et al., 2000; Gérin et al., 1998; Goldberg et al., 2001)

99 esophageal

2,299 cancer

533 poulation

Toluene

Solvents

See above

See above

Age, respondent status, birthplace, educational level, beer consumption, spirits consumption, β-carotene index, cigarette smoking (length and pattern)

Response rates: 75% of cases, 71% of population controls

 

Goldberg et al., 2001

Same as above (see also Dumas et al., 2000; Gérin et al., 1998; Parent et al., 2000)

497 colon

1,514 cancer

533 population

Benzene

Xylene

Toluene

See above

See above

Age, respondent status, ethnicity, nonoccupational factors (such as cigarette smoking, alcohol consumption)

Response rates: 82% of cases, 72% of population controls

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Description of Study Population

Number of Cases

Number of Controls

Relevant Exposure(s)

Determination of Exposure

Analysis

Adjustment for Confounding

US studies

Vaughan et al., 1997

Cases, age 20–74 years when diagnosed, identified from a cancer-surveillance system covering 13 counties in western Washington; cases lived in one of the three largest counties and were diagnosed in 1983–1990; population-based controls, frequency-matched by age and sex, identified through RDD

404 esophageal and gastric cardia

724

Tetrachloroethylene

Dry-cleaning work

In-person interviews on occupational history (job titles; including duration, exposure probability, cumulative exposure calculations)

Conditional logistic regression

Age, sex, education, study period, alcohol consumption, cigarette smoking, race

Response rates: 82.9% of cases, 80.3% of controls

Paulu et al., 1999

Cases reported to the Massachusetts Cancer Registry, diagnosed in 1983–1986 among residents of five upper Cape Cod towns; living controls were selected from the records of HCFA and through RDD; deceased controls identified by the state Department of Vital Statistics and Research files

311 colon-rectum

36 pancreas

1,158 (colon-rectum)

622 (pancreas)

Tetrachloroethylene

Calculated relative delivered dose accounting for location and years of residence, water flow, pipe characteristics

Multiple logistic regression

Age at diagnosis, vital status, sex, occupational exposure to solvents; specific cancer risk factors controlled in respective analyses

Response rates: 79% of cases, 76% of HCFA controls, 74% of RDD controls, 79% of next of kin of deceased controls

European studies

Fredriksson et al., 1989

Cases age 30–75 years identified through the Swedish Cancer Registry among patients diagnosed in 1980–1983; cases residents of the Umea region and alive during the study’s data collection; randomly selected population controls from the National Population Register were frequency-matched on age and sex

329 colon

658

Trichloroethylene Organic solvents

Dry-cleaning work Painter

Mailed questionnaire assessing occupational history (job titles); telephone interviews followed if necessary; solvent exposures independently coded by two physicians and one hygienist

Mantel-Haenszel

Age, sex, physical activity

Kauppinen et al., 1995

Deceased cases as of April 1990, age 40–74 years at diagnosis in 1984–1987; identified cases and controls from the Finnish Cancer Registry; controls of similar age and period of diagnosis selected from deceased cases of stomach, colon, or rectal cancer

595 pancreatic

1,622

Solvents

Mailed questionnaire to next of kin assessing lifetime work history (job titles); assignment of exposures by industrial hygienist and use of a job-exposure matrix

Unconditional logistic regression

Age, sex, tobacco smoking, diabetes mellitus, alcohol consumption

Response rates: 47% of cases, 50% of controls

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Description of Study Population

Number of Cases

Number of Controls

Relevant Exposure(s)

Determination of Exposure

Analysis

Adjustment for Confounding

Ekstrom et al., 1999

Cases, age 40–79 years, residing in one of the five counties, born in Sweden, and diagnosed in 1989–1995, identified and histologically confirmed by participating clinicians from all hospitals in the study area; control subjects randomly selected from the population register

565 gastric

1,164

Organic solvents

In-person interview with professional interviewer; occupational epidemiologists to assess type of exposure and duration from self-reports of exposure and job titles

Unconditional logistic regression

Age, sex

Response rates: 62.4% of cases, 75.9% of controls

Kaerlev et al., 2000

Cases, age 35–69 years, in 10 European countries in 1995–1997 identified from hospital and pathology departments and regional and national cancer registers and histologically confirmed; population-based controls randomly selected from each study base except in Spain, where hospital-based colon cancer cases were used as controls

79 small bowel adenocarcinoma

579 colon cancer

2070 population

Dry-cleaning work

Standard questionnaire administered in person or over the telephone to determine occupational exposures (job or industry titles, specific work tasks), lifestyle factors; occupation and industry codes used to categorize exposure

Unconditional logistic regression

Sex, country, year of birth

Response rates: 74% of cases, 64% of controls

Chinese study

Ji et al., 1999

Cases, age 30–74 years, identified through the Shanghai Cancer Registry among patients to diagnosed from October 1990 to June 1993; cases confirmed through histopathology, gross pathology, or tomography; randomly selected population controls from Shanghai were frequency matched on age and sex

451 pancreatic

1,552

Chemical and rubber work

Rubber work

Printing

In-person interview with professional interviewers to assess occupational history (job titles); job titles coded by the authors using a scheme developed for use in the Third National Census in 1982

Unconditional logistic regression

Age, education, income, cigarette smoking, other occupations

Response rates: 78.2% of cases, 84.5% of controls

NOTE: HCFA=Health Care Financing Administration; RDD=random-digit dialing.

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Three cohort studies of workers in the dry-cleaning industry and in aircraft manufacture reported positive associations with esophageal cancer. In a cohort of dry-cleaning union members, Ruder and colleagues (2001) observed associations between esophageal cancer and exposure to tetrachloroethylene (SMR=2.47, 95% CI=1.35–4.14) and between esophageal cancer and long-term exposure to tetrachloroethylene (SMR=5.03, 95% CI=2.41–9.47). The risk of esophageal cancer observed in workers exposed solely to tetrachloroethylene (SMR=2.65, 95% CI=0.85–6.20) was similar to the risk observed in workers exposed to tetrachloroethylene and other dry-cleaning solvents (SMR=2.40, 95% CI=1.10–4.56). Blair and colleagues (1990) found an increased risk of esophageal cancer (SMR=2.1, 95% CI=1.1–3.6) in another cohort of dry-cleaning union workers; the risk among those with high exposure to dry-cleaning solvents was slightly higher (SRR=1.3) than the risk in the referent group with medium exposure. An increased risk associated with potential exposure to tetrachloroethylene was found in a cohort of aircraft-manufacturing workers (SMR=1.47, 95% CI=0.54–3.21) (Boice et al., 1999). However, no exposure-response pattern was apparent.

Vaughan and colleagues (1997) identified cases of several types of cancer, including two morphologic types of esophageal cancer, in examining the risks from occupational exposure. The risk of esophageal squamous cell carcinoma was increased for possible exposure to tetrachloroethylene (OR=3.6, 95% CI=0.5–27.0) and probable exposure (OR=6.4, 95% CI=0.6–68.9). Increases in risk of esophageal adenocarcinoma were found to be associated with possible exposure to tetrachloroethylene but not with probable exposure.

Gérin and colleagues (1998) reported no association between medium or high exposure to benzene and risk of esophageal cancer (OR=0.9, 95% CI=0.3–2.4). In the large cohort of Chinese benzene-exposed factory workers, increased rates of mortality from esophageal cancer were found (RR=1.8, 95% CI=0.8–4.5) (Yin et al., 1996a). The cohort was examined further, including information on cumulative exposure to benzene (Hayes et al., 1996). However, when the cumulative exposure data were categorized, the relative risks did not increase with increasing exposure. The analyses did not adjust for alcohol use, an important risk factor for esophageal cancer and a potential confounder.

Gérin and colleagues (1998) found an increased risk of esophageal cancer with exposure to xylene (OR=1.4, 95% CI=0.5–3.8) or toluene (OR=1.9, 95% CI=0.9–4.2) in the medium or high exposure category. The risk estimates in the low exposure category for those substances were around unity. In a more comprehensive analysis of the same study, Parent and colleagues (2000) found a similar risk associated with “substantial” exposure to toluene (OR=1.5, 95% CI =0.6–3.7). The risk was further increased when the analysis was restricted to cases with squamous cell carcinoma (OR=2.4, 95% CI=0.9–6.4).

The only study of methylene chloride was the comprehensive cohort study of Kodak employees (Hearne and Pifer, 1999; Hearne et al., 1987, 1990). The most recent followup of the cohort included two cases of esophageal cancer in the group exposed to methylene chloride and no excess risk was observed.

A cohort study of workers in US chemical plants evaluated the association between exposure to phenol and esophageal cancer risk (Dosemeci et al., 1991). Although a slightly increased risk was associated with “any” exposure to phenol (SMR=1.6, 95% CI=0.9–2.6), other studies of unspecified mixtures of solvents yielded no increased risks of esophageal cancer (Anttila et al., 1995: SIR=0.41, 95% CI=0.01–2.29; Garabrant et al., 1988: SIR=1.14, 95% CI=0.62–1.92; Parent et al., 2000: OR=1.1, 95% CI=0.7–1.7).

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Summary and Conclusion

Although almost all the studies of esophageal cancer and exposure to tetrachloroethylene and dry-cleaning solvents showed positive associations, the small number of studies (four) and the lack of increased risk with increased exposure led some committee members to favor the inadequate/insufficient category of association. In addition, some committee members expressed concern over the lack of control or adjustment for tobacco and alcohol use (known risk factors for esophageal cancer) (see Chapter 2 and Appendix E for information on smoking as a potential confounder), whereas others believed that the lack of increased risk of lung and bladder cancer in the same studies constituted evidence that confounding alone could not account for the observed increase in esophageal cancer. As a result, several committee members believed that the evidence was inadequate/insufficient to determine whether an association exists between esophageal cancer and exposure to tetrachloroethylene or dry-cleaning solvents, and others felt that the evidence was limited/suggestive of an association. After extensive discussion and deliberation, the committee decided that it could not reach a consensus on the association. Future committees may re-examine this literature and any new studies that are conducted in the interim to clarify the association between exposure to tetrachloroethylene or dry-cleaning solvents and the risk of esophageal cancer.

In cohort studies of workers exposed to trichloroethylene, most risk estimates for esophageal cancer were increased but highly variable (because there were few exposed subjects), and the estimates of risk were not adjusted for known risk factors, including alcohol consumption. For exposure to benzene, xylene, toluene, methylene chloride, phenol, and solvent mixtures, some of the studies provided positive findings while others did not. Most were not statistically precise. The key studies reviewed by the committee on esophageal cancer are identified in Table 6.7. Unless indicated in the table, the study populations include both men and women.

The committee concludes, from its assessment of the epidemiologic literature, that there is inadequate/insufficient evidence to determine whether an association exists between chronic exposure to solvents under review, other than tetrachloroethylene and dry-cleanings solvents, and esophageal cancer.

TABLE 6.7 Selected Epidemiologic Studies—Esophageal Cancer and Exposure to Organic Solvents

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Trichloroethylene

Cohort Study—Incidence

Hansen et al., 2001

Biologically monitored Danish workers

 

 

Any exposure

6

4.2 (1.5–9.2)

 

>75 months

4

6.6 (1.8–17)

 

Low cumulative exposure

3

6.5 (1.3–19)

 

High cumulative exposure

3

4.2 (1.5–9.2)

 

Low mean exposure

5

8.0 (2.6–19)

 

High mean exposure

1

1.3 (0.02–7.0)

 

Low employment duration

2

4.4 (0.5–16)

 

High employment duration

4

6.6 (1.8–17)

Cohort Studies—Mortality

Boice et al., 1999

Aircraft-manufacturing workers in California, potential routine exposure

7

0.83 (0.34–1.72)

Blair et al., 1998

Aircraft-maintenance workers in Utah, ever exposed

10

5.6 (0.7–44.5)

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Tetrachloroethylene and Dry-cleaning Solvents

Cohort Studies—Mortality

Ruder et al., 2001

Dry-cleaning union workers

14

2.47 (1.35–4.14)

 

Long-term exposurea

10

5.03 (2.41–9.47)

 

Tetrachloroethylene-only

5

2.65 (0.85–6.20)

 

Tetrachloroethylene-plus other solvents

9

2.40 (1.10–4.56)

Boice et al., 1999

Aircraft-manufacturing workers in California

 

 

Potential routine exposure

6

1.47 (0.54–3.21)

 

≥5 years routine or intermittent

3

0.91 (0.13–1.60)

Blair et al., 1990

Dry-cleaning union members in St. Louis, MO

 

 

Dry-cleaning solvents

13

2.1 (1.1–3.6)b

 

Medium exposure (white males)

1

2.9 (0.1–18.6)b

 

Medium exposure (black males)

7

3.6 (1.5–7.6)b

 

High exposure (white males)

0

 

High exposure (black males)

1

2.0 (0.1–11.1)b

Case-Control Study

Vaughan et al., 1997

Cases from the dry-cleaning industry

 

Squamous cell (possible exposure)

2

3.6 (0.5–27.0)

 

Squamous cell (probable)

2

6.4 (0.6–68.9)

 

Adenocarcinoma (possible)

2

1.1 (0.2–5.7)

 

Adenocarcinoma (probable)

1

0.9 (0.1–10.0)

Benzene

Cohort Studies—Mortality

Gérin et al., 1998

Male residents of Montreal, Canada

 

 

Medium or high exposure

5

0.9 (0.3–2.4)

Yin et al., 1996a

Chinese factory workers ever exposed to benzene

27

1.8 (0.8–4.5)

 

Males

25

2.0 (0.9–5.4)

 

Females

2

0.8 (0.1–16.7)

Hayes et al., 1996

Chinese factory workers (cumulative exposure to benzene)

 

 

None

7

1.0

 

<10 ppm-years

5

3.5

 

10–39 ppm-years

1

0.5

 

40–99 ppm-years

3

1.3

 

100–400 ppm-years

5

1.1

 

>400 ppm-years

13

3.2

 

 

 

p-trend=0.09

Xylene and Toluene

Case-Control Studies

Parent et al., 2000

Male residents of Montreal, Canada

 

 

Esophageal

 

 

Any toluene

16

1.2 (0.7–2.2)

 

Substantial toluene

7

1.5 (0.6–3.7)

 

Squamous cell

 

 

Any toluene

15

2.0 (1.0–3.9)

 

Substantial toluene

6

2.4 (0.9–6.4)

Gérin et al., 1998

Male residents of Montreal, Canada

 

 

Medium or high exposure, xylene

5

1.4 (0.5–3.8)

 

Medium or high exposure, toluene

9

1.9 (0.9–4.2)

Methylene Chloride

Cohort Study—Mortality

Hearne and Pifer, 1999

Male Kodak workers in New York state, employed >1 year

 

Methylene chloride cohort

2

0.63 (0.07–2.28)

 

Roll-coating division (external control)

4

1.42 (0.38–3.65)

 

Roll-coating division (internal control)

4

1.40 (0.38–3.58)

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Phenol

Cohort Study—Mortality

Dosemeci et al., 1991

Male workers in five US chemical plants

 

Any exposure

15

1.6 (0.9–2.6)

 

No exposure

4

1.0

 

Low exposure

11

0.9

 

Medium exposure

10

2.3

 

High exposure

1

1.1

Unspecified Mixtures of Organic Solvents

Cohort Studies—Incidence

Anttila et al., 1995

Finnish workers biologically monitored for exposure to halogenated hydrocarbons (trichloroethylene, tetrachloroethylene, 1,1,1-trichloroethane)

1

0.41 (0.01–2.29)

Garabrant et al., 1988

Aircraft-manufacturing workers in California, employed >4 years

14

1.14 (0.62–1.92)

Case-Control Study

Parent et al., 2000

Male residents of Montreal, Canada

 

 

Esophageal

39

1.1 (0.7–1.7)

 

Squamous cell

30

1.4 (0.8–2.5)

aLong-term exposure designates 5 years or more of exposure with at least 20 years of latency time.

b95% CI calculated by the committee with standard methods from the observed and expected numbers presented iu the original study.

Epidemiologic Studies of Exposure to Organic Solvents and Stomach Cancer

The risk of stomach cancer was not increased in a cohort of Danish workers biologically monitored for a metabolite of trichloroethylene (Hansen et al., 2001). Another Scandinavian cohort, also of workers biologically monitored for trichloroethylene metabolites, showed an increased risk of stomach cancer (SIR=1.28, 95% CI=0.75–2.04), but higher concentrations of the metabolite were not associated with greater risks (Anttila et al., 1995). Blair and colleagues (1998) reported increased stomach cancer incidence with exposure greater than 5 years (SIR=3.1, 95% CI=0.8–12.1 for 5–25 unit-years of exposure), but the risks did not increase with increasing unit-years of exposure. No increased risk of stomach cancer mortality was reported in the study (SMR=0.9, 95% CI=0.4–1.9). A cohort study of aircraft-manufacturing workers in California had increased mortality (SMR=1.32, 95% CI=0.77–2.12), but an exposure-response analysis was not presented (Boice et al., 1999). A nested case-control study of rubber workers in Ohio showed no increased risks (Wilcosky et al., 1984).

Two cohort studies of workers in the dry-cleaning industry suggested no association between occupational dry-cleaning solvent exposure and stomach cancer (Blair et al., 1990; Ruder et al., 1994). Boice and co-workers (1999) found an increased risk of stomach cancer in the California cohort of aircraft- manufacturing workers (SMR=1.42, 95% CI=0.57–2.93).

No associations were found in the large cohort of Chinese benzene-exposed workers (Hayes et al., 1996; Yin et al., 1996b). In the Montreal case-control study (Gérin et al., 1998), the risk of stomach cancer was associated with medium exposure to benzene (OR=1.5, 95% CI =0.8–3.2) and high exposure (OR=1.3, 95% CI=0.5–3.2). A slightly increased risk was seen in rubber workers potentially exposed to benzene (Wilcosky et al., 1984).

In a cohort study of rotogravure workers exposed primarily to toluene, Svensson and colleagues (1990) reported an increased risk of stomach cancer mortality (SMR=2.72, 95% CI

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

=1.09–5.61). Similar risk estimates were reported among subjects with 5 years of exposure or more and at least a 10-year latency period. The cancer incidence results were similar. In the Montreal case-control study (Gérin et al., 1998), the risk of stomach cancer was increased with high exposure to toluene (OR=1.7, 95% CI=0.6–4.8), and a similar risk estimate was reported for exposure to xylene (OR=1.8, 95% CI=0.3–9.5).

In the Finnish study of workers biologically monitored for aromatic hydrocarbon exposures (xylene, toluene, and styrene), Anttila and colleagues (1998) found increased stomach cancer incidence (SIR=1.18, 95% CI=0.54–2.23). In a study of rubber workers potentially exposed to xylene or toluene, no excess risks of stomach cancer were found (Wilcosky et al., 1984).

Two cohort studies of workers exposed to methylene chloride found no persuasive evidence of associations between stomach cancer and exposure (Hearne and Pifer, 1999; Tomenson et al., 1997). Stomach cancer risk was somewhat increased in the Kodak cohort (Hearne and Pifer, 1999), but Tomenson and co-workers (1997) found no association in the cohort of cellulose triacetate workers they followed.

Dosemeci and colleagues (1991) examined stomach cancer associated with exposure to phenol in a cohort of workers at five US chemical plants. No increased risk was found to be associated with “any” exposure (SMR=0.8, 95% CI=0.5–1.3) or with categories of increasing exposure. The case-control study conducted among rubber workers included results for several specific solvents (naphthas, ethanol, acetone, isopropanol, and toluene mixture) and stomach cancer risk (Table 6.8). Risk was slightly increased with some exposures but not others (Wilcosky et al., 1984).

In several cohort studies (Acquavella et al., 1993; Anttila et al., 1995; Berlin et al., 1995; Costantini et al., 1989; Fu et al., 1996; Garabrant et al., 1988) and one case-control study (Ekstrom et al., 1999), the association between stomach cancer and exposure to mixed solvents was examined. Except for the Florence cohort of shoemakers (Fu et al., 1996), the studies showed no association with stomach cancer risk. Fu and colleagues (1996) reported a 92% excess risk of stomach cancer associated with solvents used by shoemakers.

Summary and Conclusion

With only one study showing a highly variable positive association, the committee concluded that the data were insufficient to determine whether an association exists between the risk of stomach cancer and exposure to trichloroethylene. For exposure to tetrachloroethylene and dry-cleaning solvents, benzene, xylene, toluene, methylene chloride, phenol, other specific solvents, and solvent mixtures, the results were mixed. Table 6.8 identifies the data points considered by the committee in making its conclusion regarding association for stomach cancer. Unless indicated in the table, the populations cited include both men and women.

The committee concludes, from its assessment of the epidemiologic literature, that there is inadequate/insufficient evidence to determine whether an association exists between chronic exposure to solvents under review and stomach cancer.

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

TABLE 6.8 Selected Epidemiologic Studies—Stomach Cancer and Exposure to Organic Solvents

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Trichloroethylene

Cohort Studies—Incidence

Hansen et al., 2001

Biologically monitored Danish workers

3

0.8 (0.2–2.3)

Blair et al., 1998

Aircraft-maintenance workers in Utah

 

 

No trichloroethylene exposure

6

1.5 (0.4–6.0)

 

<5 unit-years

1

0.3 (0.1–2.6)

 

5–25 unit-years

7

3.1 (0.8–12.1)

 

>25 unit-years

6

2.0 (0.5–8.1)

Anttila et al., 1995

Finnish workers biologically monitored for exposure

 

 

Years since first measurement

 

 

0–9 years

6

1.32 (0.48–2.87)

 

10–19 years

4

0.63 (0.17–1.60)

 

20+ years

7

2.98 (1.20–6.13)

 

Whole period

17

1.28 (0.75–2.04)

 

Mean personal U-TCA level

 

 

<100 µmol/L

12

1.65 (0.98–1.39)

 

100+ µmol/L

4

0.91 (0.25–2.32)

Cohort Studies—Mortality

Boice et al., 1999

Aircraft-manufacturing workers in California, potential routine exposure

17

1.32 (0.77–2.12)

Blair et al., 1998

Aircraft-maintenance workers in Utah, employed >1 year

23

0.9 (0.4–1.9)

Wilcosky et al., 1984

Male rubber workers in Ohio, exposed >1 year

5

1.0

Tetrachloroethylene and Dry-cleaning Solvents

Cohort Studies—Mortality

Boice et al., 1999

Aircraft-manufacturing workers in California, potential routine exposure

7

1.42 (0.57–2.93)

Ruder et al., 1994

Dry-cleaning labor-union workers

5

0.61 (0.20–1.43)

 

Males

2

0.43 (0.05–1.54)

 

Females

3

0.86 (0.18–2.53)

Blair et al., 1990

Dry-cleaning union members in St. Louis, MO

 

 

Dry-cleaning solvents

11

0.8 (0.4–1.4)

Benzene

Cohort Studies—Incidence

Yin et al., 1996a

Chinese factory workers ever exposed to benzene

85

0.9 (0.7–1.4)

 

Males

71

0.9 (0.6–1.4)

 

Females

14

1.0 (0.4–2.8)

Hayes et al., 1996

Chinese factory workers (cumulative exposure to benzene)

 

 

None

43

1.0

 

<10 ppm-years

6

0.6

 

10–39 ppm-years

13

1.0

 

40–99 ppm-years

12

0.9

 

100–400 ppm-years

25

1.0

 

400+ ppm-years

27

1.2

 

 

 

p-trend=0.63

Case-Control Study

Gérin et al., 1998

Male residents of Montreal, Canada

 

 

Benzene, medium exposure

11

1.5 (0.8–3.2)

 

Benzene, high exposure

6

1.3 (0.5–3.2)

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Study Population

Exposed Cases

Estimated Relative Ris (95% CI)

Xylene and Toluene

Cohort Studies—Incidence

Anttila et al., 1998

Finnish workers biologically monitored for exposure to aromatic hydrocarbons (styrene, toluene, xylene)

9

1.18 (0.54–2.23)

Svensson et al., 1990

Male rotogravure workers in Sweden

7

2.34 (0.94–4.82)

≥5 years exposure with ≥10 years latency (primarily to toluene)

5

2.18 (0.71–5.09)

Cohort Studies—Mortality

Svensson et al., 1990

Male rotogravure workers in Sweden

7

2.72 (1.09–5.61)

≥5 years exposure with ≥10 years latency (primarily to toluene)

5

2.53 (0.82–5.91)

Wilcosky et al., 1984

Male rubber workers in Ohio, exposed >1 year

 

Xylenes

3

0.53

 

Toluene

1

 

Benzene

12

1.3

Case-Control Study

Gérin et al., 1998

Male residents of Montreal, Canada

 

 

Xylene, medium exposure

7

1.0 (0.4–2.3)

 

Xylene, high exposure

2

1.8 (0.3–9.5)

 

Toluene, medium exposure

7

1.0 (0.4–2.2)

 

Toluene, high exposure

5

1.7 (0.6–4.8)

Methylene Chloride

Cohort Studies—Mortality

Hearne and Pifer, 1999

Male Kodak workers in New York state, employed >1 year

 

Methylene chloride cohort

6

1.40 (0.51–3.04)

 

Roll-coating division (external control)

5

1.25 (0.40–2.91)

 

Roll-coating division (internal control)

5

1.26 (0.41–2.93)

Tomenson et al., 1997

Male cellulose triacetate film workers in Brantham, UK, ever employed

6

0.63 (0.23–1.37)

Other Organic Solvents

Cohort Studies—Mortality

Dosemeci et al., 1991

Male workers in five US chemical plants

 

Phenol

 

 

No exposure

10

1.1

 

Any exposure

18

0.8 (0.5–1.3)

 

Low exposure

11

1.0

 

Medium exposure

5

0.5

 

High exposure

2

1.1

Wilcosky et al., 1984

Male rubber workers in Ohio, exposed >1 year

 

Specialty naphthas

18

1.1

 

Ethanol

8

1.1

 

Acetone

1

 

Isopropanol

14

1.4

 

Phenol

6

1.4

 

VM&P naphtha

3

1.1

 

Solvent “A” (toluene mixture)

15

1.4

Unspecified Mixtures of Organic Solvents

Cohort Studies—Incidence

Anttila et al., 1995

Finnish workers biologically monitored for exposure to halogenated hydrocarbons (trichloroethylene, tetrachloroethylene, 1,1,1-trichloroethane)

19

1.28 (0.77–1.99)

Berlin et al., 1995

Swedish patients with acute solvent-related disorders

2

0.5 (0.1–1.9)

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Cohort Studies—Mortality

Fu et al., 1996

Shoemakers in England and Florence

 

 

English cohort

 

 

Probable solvents

29

0.88 (0.59–1.27)

 

High solvent

5

1.72 (0.56–4.02)

1.92 (1.02–3.29)

 

Florence cohort

 

1.93 (1.00–3.36)

 

Probable solvents

13

 

 

High solvent

12

 

Acquavella et al., 1993

Metal-components manufacturing workers

 

Solvents, ever exposed

0

0.0 (0.0–17.6)

Costantini et al., 1989

Male leather workers in Tuscany, Italy, employed >6 months

6

0.43 (0.16–0.94)

Garabrant et al., 1988

Aircraft manufacturing workers in California, employed >4 months

9

0.40 (0.18–0.76)

Case-Control Study

Ekstrom et al., 1999

Residents of two regions in Sweden

 

Organic solvents, ever exposed

232

1.08 (0.87–1.33)

NOTE: U-TCA=urinary metabolite of trichloroethylene.

Epidemiologic Studies of Exposure to Organic Solvents and Colon Cancer

In a study conducted in Sweden (Fredriksson et al., 1989), a 640% excess risk of colon cancer (OR=7.4, 95% CI=1.1–47.0) was found with exposure to trichloroethylene among dry cleaners. Two cohort studies of aircraft workers did not show increased mortality from colon cancer (Blair et al., 1998; Boice et al., 1999). However, the incidence of colon cancer was increased in the cohort of aircraft-maintenance workers in Utah (Blair et al., 1998). Incidence increased with increasing unit-years of exposure (Table 6.9). Colon cancer risk was not increased in two cohorts biologically monitored for exposure to trichloroethylene (Anttila et al., 1995; Hansen et al., 2001). Anttila and co-workers found no increased risk associated with number of years since first measurement, which represents an approach to account for latency.

A case-control study of colon cancer in Sweden conducted by Fredriksson and colleagues (1989) reported an increased risk of colon cancer among female dry cleaners (OR=2.0, 95% CI=0.5–7.1). Paulu and colleagues (1999) conducted a case-control study of residents of Cape Cod that showed an increased risk of colorectal cancers with tetrachloroethylene exposure in drinking water (OR11 year latency=2.0, 95% CI=0.6–5.8) for exposures less than the 50th percentile, whereas the risk was somewhat lower when exposure equal to or greater than the 50th percentile was considered (OR11 year latency=1.5, 95% CI=0.4–4.4). A cohort of dry-cleaning workers experienced an increased risk of intestinal (excluding rectal) cancer (SMR=1.48, 95% CI=1.01–2.09), and findings were similar in a subcohort exposed to both tetrachloroethylene and other solvents; however, the risk was not similarly increased among those exposed only to tetrachloroethylene (Ruder et al., 2001). Because the analysis was based on cases of “intestinal” cancer, the findings are difficult to interpret; different classes of intestinal tract cancers, including cancers of the small and large intestines, are derived from distinct cells and may have different etiology. Real effects may be masked when diseases with different

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

etiology are investigated as one disease. Furthermore, when exposure was restricted to tetrachloroethylene, no increase risk was apparent.

Goldberg and colleagues (2001) reported an increased colon cancer risk with exposure to benzene at “substantial,” “medium,” and “high” levels; and risks increased by about 10% for every 10 years of exposure. In the large cohort of Chinese benzene-exposed workers, no increased risks of colorectal cancer were found (RR=0.9, 95% CI=0.5–1.7) (Yin et al., 1996a). Increasing cumulative exposure did not appear to be associated with increased risk (Hayes et al., 1996).

In detailed followup analyses of the population-based case-control study of occupational exposure and cancer in Montreal, the risk of colon cancer was increased with “substantial” exposure to xylene (OR=1.5, 95% CI=0.6–3.7) and toluene (OR=1.5, 95% CI=0.8–2.5) (Goldberg et al., 2001). Risk estimates increased with increasing concentration of xylene and toluene, and risk increased by 20% for every 10 years of exposure. Levels were assessed on the basis of responses to interviews and structured questionnaires that were coded by a team of chemists and industrial hygienists. Low levels were assumed if a person had been exposed peripherally or at normal levels, and high levels were assumed if a person directly handled a product that contained one of the chemicals of concern. Medium concentration fell between those two.

Colon cancer risk was not increased among workers monitored for hydrocarbons, which included styrene, toluene, and xylene (Anttila et al., 1998). In a cohort study of cancer incidence and mortality in toluene-exposed rotogravure workers, Svensson and colleagues (1990) reported an increased risk of combined colon and rectal cancer mortality (SMR=2.18, 95% CI=0.88–4.49) and incidence (SIR=1.49, 95% CI=0.68–2.84). Similar risk estimates were reported among subjects with prolonged exposure.

Two cohort studies of workers exposed to methylene chloride reported no increased colon cancer risk in the roll-coating division at Kodak (Hearne and Pifer, 1999) or in cellulose triacetate film workers in the UK (Tomenson et al., 1997). The committee concluded that these results did not indicate excess risk of colon cancer posed by exposure to methylene chloride.

One study reported no increased colon cancer risk associated with exposure to phenol (Dosemeci et al., 1991). Risk of colon cancer was not associated with exposure to solvents in a cohort of patients with solvent-related disorders (Berlin et al., 1995) or among workers biologically monitored for halogenated hydrocarbons (Anttila et al., 1995). Self-reported occupation as a painter and occupational exposure to solvents were each associated with colon cancer in a study conducted in Sweden (Fredriksson et al., 1989).

Summary and Conclusion

Because the increased risk and exposure-response pattern support an association between colon cancer and exposure to trichloroethylene, several committee members believed that the evidence was limited/suggestive of an association. Other members felt that the positive associations were balanced by the negative findings in other cohort studies and in studies with biologic monitoring of metabolites of exposure. Therefore, the committee decided not to have a consensus conclusion. Additional research will help to clarify the relationship between exposure to trichloroethylene and the risk of colon cancer.

Results of the three studies on tetrachloroethylene and dry-cleaning solvents are insufficient to determine whether an association exists for colon cancer, because of the described

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

limitations of the cohort study and because the estimates of risk in the case-control studies are imprecise, being based on few exposed cases.

The body of evidence on colon cancer and exposure to benzene and mixtures of toluene and xylene was rather small (five studies), including one high quality case-control study (Goldberg et al., 2001) and two cohort studies that included exposures other than benzene, toluene, or xylene in their analyses (Anttila et al., 1998; Svensson et al., 1990). Anttila and colleagues (1998) assessed the association between colon cancer and hydrocarbons that included styrene in addition to toluene and xylene; and Svensson and colleagues (1990), who examined a cohort of Swedish rotogravure printers, focused primarily on toluene but acknowledged the presence of other solvents and chemical agents. Although one study showed an association with exposure to benzene, toluene, and xylene (Goldberg et al., 2001), the large Chinese factory-worker study (Hayes et al., 1996; Yin et al., 1996b), which combined colon and rectal cancer, did not. The strengths of the Goldberg and colleagues study (2001) included adjustment for most known risk factors and occupational exposures; use of incident, histologically-confirmed cases; and an independent assessment of exposure by experts. The strength of the Chinese cohort study (Hayes et al., 1996) was the relatively accurate estimates of exposure, but its limitations included use of mortality instead of incidence and lack of assessment of confounding factors. As a result, some committee members concluded that the evidence was limited/suggestive of an association, and others concluded that it was insufficient to determine whether an association exists. After much deliberation, the committee decided that it could not reach a consensus on an association of colon cancer and exposure to benzene and mixtures of toluene and xylene. Further studies on these exposure-outcome relationships may provide evidence as to whether an association exists.

For exposure to methylene chloride, phenol, and mixtures of solvents, the studies did not provide any evidence of an association between exposure and risk for colon cancer. All of the studies reviewed by the committee are identified below in Table 6.9, and unless indicated, the study populations include both men and women.

The committee concludes, from its assessment of the epidemiologic literature, that there is inadequate/insufficient evidence to determine whether an association exists between chronic exposure to solvents under review, other than trichloroethylene, benzene, toluene, and xylene, and colon cancer.

TABLE 6.9 Selected Epidemiologic Studies—Colon Cancer and Exposure to Organic Solvents

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Trichloroethylenee

Cohort Studies-Incidence

Hansen et al., 2001

Biologically monitored Danish workers

 

 

Males

5

0.7 (0.2–1.6)

 

Females

1

0.7 (0.01–4.0)

Blair et al., 1998

Aircraft-maintenance workers in Utah

 

 

No trichloroethylene exposure

22

4.1 (1.4–11.8)

 

≤5 unit-years

15

2.9 (1.0–8.9)

 

5–25 unit-years

14

4.3 (1.4–13.0)

 

>25 unit-years

23

5.7 (2.0–16.7)

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Study Population

Exposed Cases

Estimated Relative Ris (95% CI)

Anttila et al., 1995

Biologically monitored Finnish workers

 

 

Years since first measurement:

 

 

0–9 years

3

1.23 (0.25–3.59)

 

10–19 years

3

0.62 (0.13–1.80)

 

20+ years

2

0.92 (0.11–3.31)

 

Whole period

8

0.84 (0.36–1.66)

Cohort Studies—Mortality

Boice et al., 1999

Aircraft-manufacturing workers in California, potential routine exposure

30

1.07 (0.72–1.52)

Blair et al., 1998

Aircraft-maintenance workers in Utah

54

1.4 (0.8–2.4)

 

Males

 

 

No trichloroethylene exposure

21

1.5 (0.7–3.3)

 

<5 unit-years

19

1.5 (0.7–3.3)

 

5–25 unit-years

12

1.5 (0.7–3.6)

 

>25 unit-years

15

1.5 (0.7–3.3)

Case-Control Study

Fredriksson et al., 1989

Residents of Sweden

 

Trichloroethylene

NA

1.5 (0.4–5.7)

 

Trichloroethylene among dry cleaners

NA

7.4 (1.1–47.0)

Tetrachloroethylene and Dry-cleaning Solvents

Cohort Study—Mortality

Ruder et al., 2001

Dry-cleaning labor-union workers (intestine)

32

1.48 (1.01–2.09)

 

Long-term exposurea

13

1.48 (0.79–2.58)

 

Tetrachloroethylene only

8

1.18 (0.51–2.33)

Case-Control Studies

Paulu et al., 1999

Residents of Cape Cod, MA

 

 

Colon-rectum (11-year latency)

 

 

≤Median

6

2.0 (0.6–5.8)

 

>Median

5

1.5 (0.4–4.4)

Fredriksson et al., 1989

Residents of Sweden

 

Female dry cleaners

5

2.0 (0.5–7.1)

Benzene

Cohort Studies—Mortality

Yin et al., 1996a

Chinese factory workers ever exposed to benzene (colon-rectum)

 

 

Total

34

0.9 (0.5–1.7)

 

Males

24

1.1 (0.5–2.3)

 

Females

10

0.7 (0.3–2.0)

Hayes et al., 1996

Chinese factory workers (colorectal; benzene cumulative exposure)

 

 

None

17

1.0

 

<10 ppm-years

7

1.5

 

10–39 ppm-years

4

0.7

 

40–99 ppm-years

3

0.5

 

100–400 ppm-years

8

0.8

 

400+ ppm-years

12

1.4

 

 

 

p-trend=0.91

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Case-Control Study

Goldberg et al., 2001

Male residents of Montreal, Canada

 

 

Substantial exposure

21

1.6 (0.9–2.8)

 

Low concentration

39

0.8 (0.5–1.2)

 

Medium concentration

26

1.5 (0.9–2.4)

 

High concentration

6

3.4 (1.0–11.2)

 

Duration (10 years)

71

1.1 (0.9–1.2)

Xylene and Toluene

Cohort Studies—Incidence

Anttila et al., 1998

Finnish workers biologically monitored for exposure to aromatic hydrocarbons (styrene, toluene, and xylene)

2

0.34 (0.04–1.21)

Svensson et al., 1990

Male rotogravure workers in Sweden—colon-rectum

9

1.49 (0.68–2.84)

 

≥5 yrs exposure with >10 yrs latency (primarily toluene)

8

1.74 (0.75–3.43)

Cohort Study—Mortality

Svensson et al., 1990

Male rotogravure workers in Sweden—colon-rectum

7

2.18 (0.88–4.49)

 

≥5 yrs exposure with >10 yrs latency (primarily to toluene)

6

2.41 (0.89–5.25)

Case-Control Study

Goldberg et al., 2001

Male residents of Montreal, Canada

 

 

Xylene

 

 

Substantial exposure

10

1.5 (0.6–3.7)

 

Low concentration

44

1.4 (0.9–2.0)

 

Medium concentration

11

1.7 (0.8–3.5)

 

High concentration

5

4.0 (1.1–15.1)

 

Duration (10 years)

60

1.2 (1.0–1.4)

 

Toluene

 

 

Substantial exposure

27

1.5 (0.8–2.5)

 

Low concentration

31

1.3 (0.8–2.0)

 

Medium concentration

31

1.3 (0.8–2.1)

 

High concentration

9

2.4 (1.0–5.7)

 

Duration (10 years)

71

1.2 (1.0–1.4)

Methylene Chloride

Cohort Studies—Mortality

Hearne and Pifer, 1999

Male Kodak workers in New York state (colon-rectum), employed >1 year

 

 

Methylene chloride cohort

15

1.15 (0.64–1.90)

 

Roll-coating division (external control)

10

0.75 (0.36–1.37)

 

Roll-coating division (internal control)

10

0.87 (0.42–1.60)

Tomenson et al., 1997

Male cellulose triacetate film workers in Brantham, UK, ever employed

6

0.90 (0.33–1.96)

Phenol

Cohort Study—Mortality

Dosemeci et al., 1991

Male workers in five US chemical plants

 

Phenol, any exposure

33

0.9 (0.6–1.3)

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Study Population

Exposed Cases

Estimated Relative Ris (95% CI)

Unspecified Mixtures of Organic Solvents

Cohort Studies—Incidence

Anttila et al., 1995

Finnish workers biologically monitored for exposure to halogenated hydrocarbons (trichloroethylene, tetrachloroethylene, 1,1,1-trichloroethane)

8

0.74 (0.32–1.44)

Berlin et al., 1995

Swedish patients with acute solvent-related disorders

2

0.6 (0.1–2.2)

Case-Control Study

Fredriksson et al., 1989

Residents of Sweden

 

Organic solvents

 

 

High-grade

NA

2.1 (0.8–5.8)

 

Low-grade

NA

1.3 (0.8–2.0)

 

Painters

7

3.0 (0.9–9.2)

NOTE: NA=not available.

aLong-term exposure designates 5 years or more of exposure with at least 20 years of latency.

Epidemiologic Studies of Exposure to Organic Solvents and Rectal Cancer

The two studies of workers biologically monitored for a metabolite of trichloroethylene showed imprecise associations (Anttila et al., 1995; Hansen et al., 2001). Risks did not increase with increasing mean urinary trichloroacetic acid, a metabolite of trichloroethylene (SIR=0.85, 95% CI=0.10–3.07) (Anttila et al., 1995).

No associations with rectal cancer mortality were found in the studies of aircraft and aerospace workers. Blair and colleagues (1998) found no increased risk of rectal cancer among aircraft maintenance workers in Utah (SMR=0.4, 95% CI=0.1–1.5). Among a trichloroethylene-exposed subcohort of aerospace workers, the SMR for “high” exposure was 1.38 (95% CI=0.45–3.21) (Morgan et al., 1998). In the cohort of aircraft-manufacturing workers in California, the SMR was 1.29 (95% CI=0.59–2.45) (Boice et al., 1999). Rectal cancer risk was found to be increased in the Montreal case-control study with “any” exposure to trichloroethylene (OR=2.0, 95% CI=1.0–3.9), but it was much lower with “substantial” exposure to trichloroethylene (OR=0.9, 95% CI=0.3–3.2) (Dumas et al., 2000).

The cohort of dry-cleaning union members experienced an excess risk of rectal cancer mortality after exposure to tetrachloroethylene and other dry-cleaning solvents (SMR=2.16, 95% CI=0.86–4.45) (Ruder et al., 2001). Paulu and colleagues (1999) conducted a case-control study of residents of Cape Cod that showed an increased risk of colorectal cancers with exposure to tetrachloroethylene from drinking water (OR11 year latency=2.0, 95% CI=0.6–5.8) when exposure was less than the 50th percentile; the risk was somewhat lower when exposure was equal to or greater than the 50th percentile (OR11 year latency=1.5, 95% CI=0.4–4.4). Because the analysis was based on cases of “intestinal” cancer, the findings are difficult to interpret; different classes of intestinal tract cancers, including cancers of the small and large intestines, are derived from different cells and may have different etiology. Real effects may be masked when diseases with different etiology are studied as one disease. No cases were found in workers exposed only to tetrachloroethylene.

In the large cohort of Chinese benzene-exposed workers, no increased risk of colorectal cancer was found (RR=0.9, 95% CI=0.5–1.7) (Yin et al., 1996a). Increasing cumulative exposure did not appear to be associated with increased risk (Hayes et al., 1996). Only in the population-based case-control study of occupational exposure and cancer in Montreal was

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

exposure to xylene, toluene, and benzene assessed. Gérin and colleagues (1998) found no association with “high” exposure to benzene (OR=0.8, 95% CI=0.3–2.5). In additional analyses of the data, Dumas and colleagues (2000) found increased rectal cancer risk with “substantial” exposure to xylene (OR=2.9, 95% CI=1.1–7.3) and “substantial” exposure to toluene (OR=1.7, 95% CI=1.0–3.0).

In the cohort study of Finnish workers biologically monitored for exposure to solvents, rectal cancer incidence was associated with exposure to styrene, xylene, and toluene (aromatic hydrocarbons) (SIR=1.88, 95% CI 0.81–3.71) (Anttila et al., 1998).

The cohort study of cellulose triacetate film workers in the UK showed no increased risk of rectal cancer (SMR=0.44, 95% CI=0.05–1.57) (Tomenson et al., 1997). Dumas and colleagues (2000) reported an increased risk of rectal cancer with “any” exposure to methylene chloride (OR=1.2, 95% CI=0.5–2.8); the risk was higher with “substantial” exposure (OR=3.8, 95% CI=1.1–12.9).

The association between rectal cancer and exposure to phenol was assessed in one cohort study (Dosemeci et al., 1991); there was no strong evidence of an association (SMR=1.4, 95% CI=0.8–2.2). Dumas and colleagues (2000) found an increased risk with “any” exposure to acetone (OR=2.3, 95% CI=1.1–4.7), which increased to an OR of 4.8 (95% CI=1.8–13.0) with “substantial” exposure.

There were three cohort studies of rectal cancer and exposure to unspecified mixtures of solvents. An early study of aircraft-manufacturing workers (Garabrant et al., 1988) and a study of Swedish patients with solvent-related disorders (Berlin et al., 1995) showed no associations with rectal cancer (SIR=1.04 and 0.99, respectively). Anttila and colleagues (1995) reported an increased risk of rectal cancer among workers biologically monitored for halogenated hydrocarbon exposure (SIR=1.63, 95% CI=0.87–2.78).

Summary and Conclusion

In summary, there was inconclusive evidence of an association between trichloroethylene and rectal cancer. The main limitation of the studies was the small number of exposed cases; this limits the precision of the estimates and the statistical power to detect associations. For exposure to tetrachloroethylene and dry-cleaning solvents, the committee concluded that the results did not indicate an excess risk of rectal cancer. Despite the suggestive findings on toluene and xylene, only two studies specifically examined rectal cancer risk. To determine whether an association exists, other high-quality studies are required.

Because there were only two studies to draw inferences from, the evidence for methylene chloride was inadequate. Furthermore, the committee could not draw conclusions from single studies of each exposure although there was suggestive evidence from the high-quality study of Dumas and colleagues regarding exposure to acetone. For exposure to solvent mixtures, the findings were mixed with no persuasive evidence of an association.

Table 6.10 identifies the studies and data points considered by the committee in making its conclusion regarding association. Unless indicated, the study populations identified in the table include both men and women.

The committee concludes, from its assessment of the epidemiologic literature, that there is inadequate/insufficient evidence to determine whether an association exists between chronic exposure to solvents under review and rectal cancer.

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

TABLE 6.10 Selected Epidemiologic Studies—Rectal Cancer and Exposure to Organic Solvents

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Trichloroethylene

Cohort Studies—Incidence

Hansen et al., 2001

Biologically monitored Danish workers

7

1.3 (0.5–2.7)

Anttila et al., 1995

Biologically monitored Finnish workers

 

 

Entire period since first measurement

12

1.71 (0.88–2.98)

 

0–9 years

3

1.59 (0.33–4.64)

 

10–19 years

8

2.22 (0.96–4.36)

 

20+ years

1

0.67 (0.02–3.72)

 

Mean personal U-TCA level:

 

 

<100 µmol/L

9

2.34 (1.07–4.44)

 

100+ µmol/L

2

0.85 (0.10–3.07)

Cohort Studies—Mortality

Boice et al., 1999

Aircraft-manufacturing workers in California, potential routine exposure

9

1.29 (0.59–2.45)

Blair et al., 1998

Aircraft-maintenance workers in Utah, employed >1 year

5

0.4 (0.1–1.5)

Morgan et al., 1998

Aerospace workers in Arizona

 

Trichloroethylene-exposed subcohort

6

1.06 (0.39–2.31)

 

High trichloroethylene-exposure

5

1.38 (0.45–3.21)

Case-Control Study

Dumas et al., 2000

Male residents of Montreal, Canada

 

 

Any exposure

12

2.0 (1.0–3.9)

 

Substantial exposure

3

0.9 (0.3–3.2)

Tetrachloroethylene and Dry-cleaning Solvents

Cohort Study—Mortality

Ruder et al., 2001

Dry-cleaning labor-union workers (intestine)

 

 

Tetrachloroethylene plus other solvents

7

2.16 (0.86–4.45)

Case-Control Study

Paulu et al., 1999

Residents of Cape Cod, MA

 

 

Colon-rectum (11-year latency)

 

 

≤Median

6

2.0 (0.6–5.8)

 

>Median

5

1.5 (0.4–4.4)

Benzene

Cohort Studies—Mortality

Yin et al., 1996a

Chinese factory workers ever exposed to benzene (colon-rectum)

 

 

Total

34

0.9 (0.5–1.7)

 

Males

24

1.1 (0.5–2.3)

 

Females

10

0.7 (0.3–2.0)

Hayes et al., 1996

Chinese factory workers (colorectal; benzene cumulative exposure)

 

 

None

17

1.0

 

<10 ppm-years

7

1.5

 

10–39 ppm-years

4

0.7

 

40–99 ppm-years

3

0.5

 

100–400 ppm-years

8

0.8

 

400+ ppm-years

12

1.4

 

 

 

p-trend=0.91

Case-Control Study

Gérin et al., 1998

Male residents of Montreal, Canada

 

 

High exposure

4

0.8 (0.3–2.5)

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Xylene and Toluene

Cohort Study—Incidence

Anttila et al., 1998

Finnish workers biologically monitored for exposure to aromatic hydrocarbons (styrene, toluene, xylene)

8

1.88 (0.81–3.71)

Case-Control Study

Dumas et al., 2000

Male residents of Montreal, Canada

 

 

Xylene

 

 

Any exposure

39

1.3 (0.9–1.9)

 

Substantial exposure

7

2.9 (1.1–7.3)

 

Toluene

 

 

Any exposure

50

1.4 (1.0–2.0)

 

Substantial exposure

17

1.7 (1.0–3.0)

Methylene Chloride

Cohort Study—Mortality

Tomenson et al., 1997

Male cellulose triacetate film workers in Brantham, UK, ever employed

2

0.44 (0.05–1.57)

Case-Control Study

Dumas et al., 2000

Male residents of Montreal, Canada

 

 

Any exposure

7

1.2 (0.5–2.8)

 

Substantial exposure

5

3.8 (1.1–12.9)

Phenol and Acetone

Cohort Study—Mortality

Dosemeci et al., 1991

Male workers in five US chemical plants

 

Phenol

 

 

No exposure

6

1.1

 

Any exposure

18

1.4 (0.8–2.2)

 

Low exposure

9

1.4

 

Medium exposure

9

1.5

 

High exposure

0

0

Case-Control Study

Dumas et al., 2000

Male residents of Montreal, Canada

 

 

Acetone

 

 

Any exposure

11

2.3 (1.1–4.7)

 

Substantial exposure

8

4.8 (1.8–13.0)

Unspecified Mixtures of Organic Solvents

Cohort Studies—Incidence

Anttila et al., 1995

Finnish workers biologically monitored for exposure to halogenated hydrocarbons (trichloroethylene, tetrachloroethylene, 1,1,1-trichloroethane)

13

1.63 (0.87–2.78)

Berlin et al., 1995

Swedish patients with acute solvent-related disorders

2

0.99 (0.1–3.6)

Cohort Study—Mortality

Garabrant et al., 1988

Aircraft-manufacturing workers in California, employed >4 years

15

1.04 (0.59–1.73)

NOTE: U-TCA=urinary metabolite of trichloroethylene.

Epidemiologic Studies of Exposure to Organic Solvents and Pancreatic Cancer

Several cohort studies of trichloroethylene-exposed workers showed no increased risk of pancreatic cancer. In the cohort of aircraft-manufacturing workers examined by Boice and

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

colleagues (1999), risk was shown to decrease (SMR=0.41, 95% CI=0.17–0.85) with potential exposure to trichloroethylene. Blair and co-workers (1998) also found no association between pancreatic cancer incidence and all categories of “unit-years” of exposure. Pancreatic cancer mortality was weakly increased in the same study (SMR=1.2, 95% CI=0.6–2.3). A third cohort study of aerospace workers in Arizona also found no risk of pancreatic cancer posed by exposure to high levels of trichloroethylene (SMR=0.66, 95% CI=0.24–1.43) (Morgan et al., 1998).

Two cohorts of workers biologically monitored for metabolites of trichloroethylene reported mixed findings. Hansen and colleagues (2001) found no association (SIR=1.0, 95% CI =0.2–3.0) between the trichloroethylene metabolite and pancreatic cancer, whereas Anttila and colleagues (1995) found an increased risk (SIR=1.61, 95% CI=0.81–2.88). However, no exposure-response relationship was indicated as the mean concentration of the metabolite increased.

Ruder and colleagues (2001) found an association between pancreatic cancer and exposure to tetrachloroethylene and other solvents (SMR=1.89, 95% CI=1.06–3.11), but no increase in risk was found in the subcohort of workers exposed only to tetrachloroethylene (SMR =0.80, 95% CI=0.17–2.35). An increased risk of pancreatic cancer was found in the cohort of workers monitored for solvents, including tetrachloroethylene (SIR=3.08, 95% CI=0.63–8.99) (Anttila et al., 1995).

The case-control study of multiple cancer sites performed by Gérin and colleagues (1998) indicated no association between pancreatic cancer and medium or high exposure to toluene (OR=0.6, 95% CI=0.2–2.2), xylene (OR=1.1, 95% CI=0.4–3.3), or benzene (OR=0.4, 95% CI=0.1–1.4). In the cohort study of Finnish workers exposed to xylene, toluene, and styrene, Anttila and colleagues (1998) found increased risks of pancreatic cancer (SIR=1.26, 95% CI=0.41–2.93). A case-control study of pancreatic cancer by Ji and colleagues (1999) conducted in Shanghai showed increased risks in various occupational groups, especially among male painters (OR=5.2, 95% CI=1.1–25.0).

The comprehensive cohort study of methylene chloride-exposed workers at Kodak has been followed for a number of years (Hearne et al., 1987, 1990). In the initial publication, an excess of pancreatic cancer was observed (SMR=3.1); however, in the second study, after 4 additional years of followup, there was no increase in pancreatic cancer mortality. In a study examining the same cohort and another cohort of Kodak workers, Hearne and Pifer (1999) showed an increased risk of pancreatic cancer associated with career exposure to methylene chloride of over 800 ppm-years on the basis of three cases (SMR=2.34, compared with internal controls). Among workers who were employed in 1964–1970 in the roll coating division, the highest risk was found in the lowest cumulative-exposure category of less than 400 ppm (SMR=2.58, compared with internal controls). Other cohorts of methylene chloride-exposed workers had very few cases of pancreatic cancer and reported no increased risk (Gibbs et al., 1996; Lanes et al., 1990, 1993; Tomenson et al., 1997).

The study of workers in five chemical plants found no increased risk of pancreatic cancer associated with exposure to phenol (SMR=0.6, 95% CI=0.4–1.1) (Dosemeci et al., 1991).

Several studies reported associations between pancreatic cancer and unspecified mixtures of organic solvents (Table 6.11). The studies showing positive associations included male leather workers in Italy (SMR=1.46, 95% CI=0.39–3.73) (Costantini et al., 1989), aircraft-manufacturing workers (SMR=1.19, 95% CI=0.83–1.67) (Garabrant et al., 1988), and the

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

case-control study in Finland of high exposure to solvents (OR=2.01, 95% CI=0.98–4.10) (Kauppinen et al., 1995).

Summary and Conclusion

For exposure to trichloroethylene, tetrachloroethylene and dry-cleaning solvents, xylene, toluene, benzene, methylene chloride, phenol, and solvent mixtures and the risk of pancreatic cancer, the evidence was limited by mixed results, the lack of exposure-response relationships, and imprecise estimates of risk. Table 6.11 identifies the studies reviewed by the committee in making its conclusion regarding association. All of the study populations include both men and women unless stated otherwise.

The committee concludes, from its assessment of the epidemiologic literature, that there is inadequate/insufficient evidence to determine whether an association exists between chronic exposure to solvents under review and pancreatic cancer.

TABLE 6.11 Selected Epidemiologic Studies—Pancreatic Cancer and Exposure to Organic Solvents

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Trichloroethylene

Cohort Studies—Incidence

Hansen et al., 2001

Biologically monitored Danish workers

3

1.0 (0.2–3.0)

Blair et al., 1998

Aircraft maintenance workers in Utah

 

 

No trichloroethylene exposure

6

0.7 (0.2–2.3)

 

<5 unit-years

6

0.7 (0.2–2.1)

 

5–25 unit-years

2

0.4 (0.1–1.8)

 

>25 unit-years

51

0.7 (0.2–2.4)

Anttila et al., 1995

Biologically monitored Finnish workers

 

 

Entire period since first measurement:

11

1.61 (0.81–2.88)

 

0–9 years

1

0.56 (0.01–3.10)

 

10–19 years

8

2.30 (0.99–4.52)

 

20+ years

2

1.31 (0.16–4.74)

 

Mean personal U-TCA level:

 

 

<100 µmol/L

6

1.61 (0.59–3.50)

 

100+ µmol/L

3

1.31 (0.27–3.82)

Cohort Studies—Mortality

Boice et al., 1999

Aircraft-manufacturing workers in California, potential routine exposure

7

0.41 (0.17–0.85)

Blair et al., 1998

Aircraft-maintenance workers in Utah, employed >1 year

33

1.2 (0.6–2.3)

Morgan et al., 1998

Aerospace workers in Arizona

 

Trichloroethylene-exposed subcohort

11

0.76 (0.38–1.37)

 

High trichloroethylene-exposure

6

0.66 (0.24–1.43)

Tetrachloroethylene and Dry-cleaning Solvents

Cohort Study—Incidence

Anttila et al., 1995

Biologically monitored Finnish workers

3

3.08 (0.63–8.99)

Cohort Study—Mortality

Ruder et al., 2001

Dry-cleaning labor-union workers

18

1.53 (0.91–2.42)

 

Tetrachloroethylene-only

3

0.80 (0.17–2.35)

 

Tetrachloroethylene-plus other solvents

15

1.89 (1.06–3.11)

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Xylene, Toluene, Benzene

Cohort Study—Incidence

Anttila et al., 1998

Finnish workers biologically monitored for exposure to aromatic hydrocarbons (styrene, toluene, xylene)

5

1.26 (0.41–2.93)

Case-Control Studies

Ji et al., 1999

Residents of Shanghai, China

 

 

Chemical and rubber workers (female)

5

1.4 (0.4–4.7)

 

Rubber workers (female)

5

1.7 (0.5–5.8)

 

Printers (male)

4

5.2 (1.1–25.0)

Gérin et al., 1998

Male residents of Montreal, Canada

 

 

Xylene, medium/high exposure

4

1.1 (0.4–3.3)

 

Toluene, medium/high exposure

3

0.6 (0.2–2.2)

 

Benzene, medium/high exposure

3

0.4 (0.1–1.4)

Methylene Chloride

Cohort Studies—Mortality

Hearne and Pifer, 1999

Male Kodak workers in New York state, employed >1 year

 

Methylene chloride cohort

5

0.92 (0.30–2.14)

 

Internal comparison, ≥800 ppm-years

3

2.34

 

Roll-coating division (New York state control)

8

1.51 (0.65–2.98)

 

Roll-coating division (Kodak Rochester control)

8

1.55 (0.67–3.06)

Tomenson et al., 1997

Male cellulose triacetate film workers, ever employed

3

0.68 (0.14–1.99)

Gibbs et al., 1996

Cellulose-fiber production workers

 

 

High exposure, males

1

0.35 (0.01–1.92)

 

High exposure, females

0

0

 

Low exposure, males

2

0.89 (0.11–3.22)

 

Low exposure, females

1

0.58 (0.01–3.23)

Lanes et al., 1993

Cellulose-fiber production plant workers, employed >3 months

2

0.83 (0.10–2.99)

Phenol

Cohort Study—Mortality

Dosemeci et al., 1991

Male workers in five US chemical plants

 

Phenol, any exposure

14

0.6 (0.4–1.1)

Unspecified Mixtures of Organic Solvents

Cohort Study—Incident

Anttila et al., 1995

Finnish workers biologically monitored for exposure to halogenated hydrocarbons (trichloroethylene, tetrachloroethylene, 1,1,1-trichloroethane)

12

1.56 (0.81–2.72)

Cohort Studies—Mortality

Fu et al., 1996

Shoemakers in England and Florence

 

 

English cohort

25

0.70 (0.45–1.04)

 

Florence cohort

2

0.54 (0.07–1.95)

Acquavella et al., 1993

Metal-components manufacturing workers, ever exposed

1

2.9 (0.1–16.0)

Costantini et al., 1989

Male leather workers in Tuscany, Italy, employed >6 months

4

1.46 (0.39–3.73)

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Garabrant et al., 1988

Aircraft-manufacturing workers in California, employed >4 years

34

1.19 (0.83–1.67)

Pippard and Acheson, 1985

Male boot and shoe makers in three English towns

 

Rushden

21

0.96 (0.59–1.46)

 

Street

2

0.40 (0.05–1.46)

 

Stafford

6

0.96 (0.35–2.11)

McMichael et al., 1976

Male rubber workers in Ohio and Wisconsin

 

Age 40–64

17

0.95 (0.55–1.52)a

 

Age 65–84

40

1.08 (0.77–1.47)a

 

Age 40–84

57

1.03 (0.78–1.33)a

Case-Control Study

Kauppinen et al., 1995

Residents of Finland

 

Solvents (all)

20

1.22 (0.73–2.07)

 

Solvents (high)

14

2.01 (0.98–4.10)

NOTE: U-TCA=urinary metabolite of trichloroethylene.

a95% CI calculated by the committee with standard methods from the observed and expected numbers presented in the original study.

HEPATOBILIARY CANCERS

Description of Case-Control Studies

Two population-based (Hardell et al., 1984; Heinemann et al., 2000) and two hospital-based (Hernberg et al., 1988; Stemhagen et al., 1983) case-control studies examined risk of liver cancer associated with occupational exposure to solvents (Table 6.12). Self-administered questionnaires or interviews were used to obtain occupational history information in each study. The study by Stemhagen and colleagues (1983) used job titles as surrogates of exposure. In two of the other studies, exposures were inferred by industrial hygienists (Heinemann et al., 2000; Hernberg et al., 1988). In the two hospital-based studies, there were no adjustments for potentially confounding variables.

Epidemiologic Studies of Exposure to Organic Solvents and Hepatobiliary Cancers

Anttila and colleagues (1995) found an increased risk (SIR=2.27, 95% CI=0.74–5.29) of liver cancer in the cohort of workers biologically monitored for a metabolite of trichloroethylene. Likewise, Hansen and colleagues (2001) reported an increased risk of cancer of the liver and biliary passages in men in a Danish cohort of biologically monitored workers (SIR=2.6, 95% CI=0.8–6.0).

In a large cohort study of aircraft-manufacturing workers in California, Boice and colleagues (1999) found no association between liver cancer and exposure to trichloroethylene (SMR=0.54, 95% CI=0.15–1.38). Another cohort of aircraft manufacturers (Morgan et al., 1998) showed no increased mortality in trichloroethylene-exposed workers, and there was no evidence of a trend in mortality with cumulative exposure.

In the cohort of workers at Hill Air Force Base in Utah, Blair and colleagues (1998) did not find excess risk of liver cancer mortality or incidence, nor was there any apparent increase with increasing cumulative exposure to trichloroethylene.

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

TABLE 6.12 Description of Case-Control Studies of Liver Cancer and Exposure to Organic Solvents

Reference

Description of Study Population

Number of Cases

Number of Controls

Relevant Exposure(s)

Determination of Exposure

Analysis

Adjustment for Confounding

Stemhagen et al., 1983

Cases identified through diagnosis in New Jersey hospitals in 1975–1980 or from death certificates in 1975–1979, all with histologic confirmation; controls selected from hospital records and death certificates and matched for age, race, sex, county of residence

265

530

Laundering, cleaning, other garment service work

In-person interview (direct or proxy) to assess occupational history (job titles)

OR

None

Response rates: 89.5% of cases, 77% of controls

Hardell et al., 1984

Deceased male cases, age 25–80 years at diagnosis, reported to the Swedish Cancer Registry and diagnosed in 1974–1981, with histologic confirmation; controls selected from the National Population Register, matched on sex, age, year of death, municipality

98

200

Organic solvents

Mailed questionnaires to next of kin, assessing work history (job titles) and occupational or leisure-time exposure to specific chemicals (self-reports)

Mantel-Haenszel rate ratio, Miettinen 95% CI

Age

Hernberg et al., 1988

Deceased cases, reported to the Finnish Cancer Register in 1976–1978 and 1981; controls selected from deceased stomach-cancer cases and coronary-infarction patients reported in 1977

344

476 stomach

385 coronary infarction

Solvents

Questionnaire mailed to next of kin to elicit list of occupations and employers; hygienist assigned exposure classification based on occupation

Likelihood-based ORs

None

Response rates: 71.7% of cases, 72.8% of stomach-cancer controls, 69.0% of infarction controls

Heinemann et al., 2000

Female cases, under 65 years old, identified at 64 clinics in six European countries in 1990–1994 (prevalent cases) and 1994–1996 (incident cases); hospital controls selected from respective clinics and matched on age; population controls selected from citizen registers

317

1,779

Dry-cleaning work Solvents

In-person interview (direct or proxy) with questionnaire assessing occupational history (industry titles) and specific agent exposures (self-reports)

Logistic regression

Age, center, smoking, alcohol, oral contraceptive use, hepatitis infection

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Ritz (1999) found associations between exposure to trichloroethylene and liver cancer among male workers in the uranium-processing industry (SMR=1.66, 95% CI=0.71–3.26). The association is based on a small number of cases and may be confounded by other potential exposures involved in uranium processing. Axelson and colleagues (1994) also found an increased incidence in male Swedish workers exposed to trichloroethylene (SIR=1.41, 95% CI =0.38–3.60). A cohort study of transformer-assembly workers (Greenland et al., 1994) showed no positive associations between occupational exposure to trichloroethylene and cancers of the liver, gallbladder, and biliary tract combined (OR=0.54, 95% CI=0.11–2.63).

In the cohort of aircraft-manufacturing workers (Boice et al., 1999), an association between exposure to tetrachloroethylene and liver cancer was observed (SMR=2.05, 95% CI=0.83–4.23) but no increase was observed with increasing duration of exposure. Bond and colleagues (1990) found an association between mortality and liver cancer in a cohort of chemical workers (OR=1.8, 95% CI=0.8–4.3).

A cohort study of US dry cleaners showed no association to liver cancer (SMR=0.7, 95% CI=0.2–1.7) (Blair et al., 1990). Ruder and colleagues (2001) detected only one case of liver cancer among the subcohort of workers exposed to tetrachloroethylene and other solvents (SMR=0.16, 95% CI=0.00–1.32).

In a case-control study of women in the Multicentre International Liver Tumour Study, three women with hepatocellular cancer reported working as dry cleaners (OR=0.65; 95% CI=0.12–3.44) (Heinemann et al., 2000). Another study of occupational risk factors for liver cancer found an association among men employed in laundering, cleaning, and other garment services (RR=2.5, 95% CI=1.02–6.14) (Stemhagen et al., 1983). Further investigation by the authors showed that the cases were concentrated among people who processed clothes and potentially had exposure to other chemicals.

Friedlander and colleagues (1978) established a cohort of workers in one department at Kodak where methylene chloride was the primary solvent exposure for more than 30 years. In the most recent followup (Hearne and Pifer, 1999), one death from liver cancer was observed (SMR=0.42, 95% CI=0.01–2.36).

A cohort study of workers producing cellulose triacetate fibers at a Hoechst Celanese plant in South Carolina showed an excess risk of biliary and liver cancers (SMR=2.98, 95% CI =0.81–7.63) (Lanes et al., 1993). In another Hoechst Celanese facility manufacturing cellulose triacetate, Gibbs and colleagues (1996) found one death from liver cancer (SMR=0.81, 95% CI =0.02–4.49).

No cases of liver cancer were found in a cohort of male workers producing cellulose triacetate film base at a plant in the UK (Tomenson et al., 1997). A nested case-control study of liver and biliary tract cancer cases identified among male hourly employees of the Dow Chemical Company did not show an excess risk in workers exposed to methylene chloride (RR=0.8, 95% CI=0.2–3.6) (Bond et al., 1990).

Yin and colleagues (1996a,b) followed a cohort of factory workers with known exposure to benzene. Increased relative risks of cancers of the liver and gallbladder combined were reported for men (RR=1.3, 95% CI=0.9–1.9) but not for women (RR=0.4, 95% CI=0.2–1.3). This cohort was further examined by Hayes and colleagues (1996), who evaluated the relative risks according to cumulative exposure to benzene. There was some suggestion that the relative risks for liver and gallbladder cancer increased with increasing cumulative exposure, although chance could not be ruled out confidently (p value for linear trend=0.16).

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

A nested case-control study by Greenland and colleagues (1994) of transformer-assembly workers exposed to benzene showed a positive association between occupational exposure to benzene and cancers of the liver, gallbladder, and biliary tract combined (OR=2.76, 95% CI=0.68–11.2).

A cohort of toluene-exposed German rotogravure workers studied by Wiebelt and Becker (1999) experienced a higher risk of liver cancer mortality than the population of West Germany (SMR=1.98, 95% CI=0.34–7.16). Dosemeci and colleagues (1991) conducted a study of US industrial workers that showed no increase in liver cancer mortality among those occupationally exposed to phenol (SMR=1.0, 95% CI=0.4–1.9). The association with phenol did not increase markedly from low to medium to high exposure.

Three population-based case-control studies (Hardell et al., 1984; Heinemann et al., 2000; Hernberg et al., 1988) examined the association of liver cancer in relation to the broad category of solvents or mixed solvents. Heinemann and colleagues (2000) found no associations (OR=1.05, 95% CI=0.52–2.09). Hernberg and colleagues (1988) did not find increased risks among males (OR=0.6, 90% CI 0.3–1.3), but increased risks were found among women (OR=3.4, 90% CI 1.3–8.6). Hardell and colleagues (1984) also found an association (OR=1.8, 95% CI=0.99–3.4).

In the cohort studies, no associations were found in aircraft-manufacturing workers (SMR =0.92) (Boice et al., 1999), but the three cohort studies of painters all showed excess risks of liver cancer: Steenland and Palu (1999) found an SMR of 1.25 (95% CI=1.03–1.50), Matanoski and colleagues (1986) an SMR of 1.47 (95% CI=0.98–2.13), and Morgan (1981) an SMR of 1.93.

Summary and Conclusion

Although most studies examined the risk for liver cancer broadly, others combined liver cancer with gallbladder and biliary tract cancers. For exposure to trichloroethylene, tetrachloroethylene and dry-cleaning solvents, toluene, and phenol, most studies did not find an increased risk. Although one study on methylene chloride found a positive association, the small number of cases of liver cancer and the lack of corroborating studies were limitations of the literature. The studies on exposure to benzene were also limited with only three studies of benzene-exposed workers (two of which were on the same cohort), and only weak evidence of excess risks. For exposure to unspecified mixtures of organic solvents, although some of the estimates of risk were positive, others were not, and the role of confounding by other exposures or risk factors was a limitation. Table 6.13 identifies the key studies and the relevant data points reviewed by the committee in drawing its conclusion. Unless indicated in the table, the study populations include both men and women.

The committee concludes, from its assessment of the epidemiologic literature, that there is inadequate/insufficient evidence to determine whether an association exists between chronic exposure to solvents under review and hepatobiliary cancers.

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

TABLE 6.13 Selected Epidemiologic Studies—Hepatobiliary Cancers and Exposure to Organic Solvents

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Trichloroethylene

Cohort Studies—Incidence

Hansen et al., 2001

Biologically monitored Danish workers

 

 

Males

5

2.6 (0.8–6.0)

 

Females

0

Blair et al., 1998

Aircraft-maintenance workers in Utah

 

 

Males:

 

 

No exposure

1

0.8 (0.1–12.0)

 

<5 unit-years

2

1.2 (0.1–13.8)

 

5–25 unit-years

1

1.0 (0.1–16.0)

 

>25 unit-years

3

2.6 (0.3–25.0)

Anttila et al., 1995

Biologically monitored workers in Finland

 

 

Entire period since first measurement:

5

2.27 (0.74–5.29)

 

0–9 years

0

—(0.0–6.59)

 

10–19 years

2

1.74 (0.21–6.29)

 

≥20 years

3

6.07 (1.25–17.7)

 

Mean personal U-TCA level

 

 

<100 µmol/L

2

1.64 (0.20–5.92)

 

100+ µmol/L

2

2.74 (0.33–9.88)

Axelson et al., 1994

Biologically monitored Swedish workers

4

1.41 (0.38–3.60)

Cohort Studies—Mortality

Boice et al., 1999

Aircraft-manufacturing workers in California

 

 

All exposed factory workers

4

0.54 (0.15–1.38)

 

Duration of potential exposure (routine or intermittent)

 

 

<1 year

4

0.53 (0.18–1.60)

 

1–4 years

3

0.52 (0.15–1.79)

 

≥5 years

6

0.94 (0.36–2.46)

Ritz, 1999

White male US uranium-processing plant workers

 

 

Trichloroethylene, cutting fluids, or kerosene

8

1.66 (0.71–3.26)

 

Trichloroethylene—light exposure

 

 

>2 years, no latency

3

0.93 (0.19–4.53)

 

>2 years, 15-year latency

3

1.16 (0.24–5.60)

 

>5 years, no latency

3

1.90 (0.35–10.3)

 

>5 years, 15-year latency

3

2.86 (0.48–17.3)

 

Trichloroethylene—moderate exposure

 

 

>2 years, no latency

1

4.97 (0.48–51.1)

 

>2 years, 15-year latency

1

5.53 (0.54–56.9)

 

>5 years, no latency

1

8.82 (0.79–98.6)

 

>5 years, 15-year latency

1

12.1 (1.03–144)

Blair et al., 1998

Aircraft-maintenance workers in Utah

 

 

Males

 

 

No exposure

3

0.5 (0.1–2.4)

 

<5 unit-years

6

1.1 (0.3–4.1)

 

5–25 unit-years

3

0.9 (0.2–4.3)

 

>25 unit-years

3

0.7 (0.2–3.2)

 

Females:

 

 

No exposure

3

4.2 (0.7–25.0)

 

<5 unit-years

1

1.6 (0.2–18.2)

 

5–25 unit-years

0

 

>25 unit-years

2

2.3 (0.3–16.7)

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Morgan et al., 1998

Aerospace workers in Arizona

 

 

Trichloroethylene-exposed subcohort:

6

0.98 (0.36–2.13)

 

Low cumulative exposure

3

1.32 (0.27–3.85)

 

High cumulative exposure

3

0.78 (0.16–2.28)

 

Peak and cumulative exposure:a

 

 

Peak: medium and high vs low and no exposure

3

0.98 (0.29–3.35)

 

Cumulative (low)

3

2.12 (0.59–7.66)

 

Cumulative (high)

3

1.19 (0.34–4.16)

Greenland et al., 1994

White male transformer-assembly workers, ever exposed

NA

0.54 (0.11–2.63)

Tetrachloroethylene and Dry-cleaning Solvents

Cohort Studies—Mortality

Ruder et al., 2001

US dry-cleaning workers in four labor unions

1

0.16 (0.00–1.32)

Boice et al., 1999

Aircraft- manufacturing workers

 

 

All exposed factory workers

7

2.05 (0.83–4.23)

 

Duration of potential exposure (routine or intermittent)

 

 

<1 year

3

1.38 (0.40–4.69)

 

1–4 years

4

0.39 (0.39–3.47)

 

≥5 years

5

1.29 (0.46–3.65)

Blair et al., 1990

Dry-cleaning union members in Missouri

5

0.7 (0.2–1.7)

Bond et al., 1990

Male chemical-company workers in Michigan, ever exposed

6

1.8 (0.8–4.3)

Case-Control Studies

Heinemann et al., 2000

Females in the Multicentre International Liver Tumour Study

 

 

Dry-cleaning, ever employed

3

0.65 (0.12–3.44)

Stemhagen et al., 1983

Laundering, cleaning, other garment-services workers in New Jersey, employed >6 months

10

2.50 (1.02–6.14)

Methylene Chloride

Cohort Studies—Mortality

Hearne and Pifer, 1999

Male cellulose triacetate photographic-film base workers in Kodak Park, employed >1 year

1

0.42 (0.01–2.36)

Tomenson et al., 1997

Male cellulose triacetate-fiber film base workers in the UK

 

 

Never exposed

0

 

All exposed

0

Gibbs et al., 1996

Male cellulose triacetate-fiber production workers in Maryland

 

 

No exposure

0

 

Low exposure

1

0.75 (0.029–4.20)

 

High exposure

1

0.81 (0.020–4.49)

Lanes et al., 1993

Cellulose triacetate-fiber production workers in South Carolina, employed >3 months

4

2.98 (0.81–7.63)

Bond et al., 1990

Male chemical-company workers in Michigan, ever exposed

2

0.8 (0.2–3.6)

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Benzene

Cohort Studies—Mortality

Yin et al., 1996a

Chinese factory workers, ever exposed

 

 

Total cohort

109

1.2 (0.8–1.6)

 

Men

101

1.3 (0.9–1.9)

 

Women

8

0.4 (0.2–1.3)

Hayes et al., 1996

Chinese factory workers

 

 

<10 ppm-years

12

1.1

 

10–39 ppm-years

12

0.8

 

40–99 ppm-years

9

0.6

 

100–400 ppm-years

44

1.6

 

400+ ppm-years

28

1.2

 

 

 

p-trend=0.16

Greenland et al., 1994

White male transformer-assembly workers, ever exposed

NA

2.76 (0.68–11.2)

Other Specific Organic Solvents

Cohort Studies—Mortality

Wiebelt and Becker, 1999

Male German rotogravure printing-plant workers, employed >1 year—toluene

3

1.98 (0.34–7.16)

Dosemeci et al., 1991

US white male industrial workers—phenol

 

 

Any exposure

8

1.0 (0.4–1.9)

 

Level of cumulative exposure:

 

 

None

4

1.2

 

Low exposure

1

0.3

 

Medium exposure

6

1.6

 

High exposure

1

1.4

Unspecified Mixtures of Organic Solvents

Cohort Studies—Mortality

Boice et al., 1999

Aircraft-manufacturing workers

 

 

All exposed factory workers

17

0.92 (0.54–1.47)

 

Duration of potential exposure (routine or intermittent)

 

 

<1 year

10

1.35 (0.63–2.87)

 

1–4 years

14

0.75 (0.38–1.47)

 

≥5 years

31

0.97 (0.54–1.72)

Steenland and Palu, 1999

Members of US painters’ unions

 

 

Total cohort

119

1.25 (1.03–1.50)

 

20 years since first union membership

90

1.17 (0.95–1.44)

Greenland et al., 1994

White male transformer-assembly workers, ever exposed to solvents

NA

0.69 (0.18–2.60)

Matanoski et al., 1986

US painters and allied tradesmen

28

1.47 (0.98–2.13)

Morgan et al., 1981

Male US paint or varnish manufacturing workers, employed >1 year

6

1.93

Case-Control Studies

Heinemann et al., 2000

Women in Multicentre International Liver Tumor Study, ever exposed to solvents

18

1.05 (0.52–2.09)

Hernberg et al., 1988

Finnish cases and deceased controls

 

 

Solvent exposure, 10-year latency

 

 

Males

7

0.6 (0.3–1.3)b

 

Females

7

3.4 (1.3–8.6)b

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Hardell et al., 1984

Male residents of Sweden

 

 

High-grade exposure to organic solvents

 

 

Primary liver cancer

22

1.8 (0.99–3.4)

 

Hepatocellular carcinoma

20

2.1 (1.1–4.0)

NOTE: NA=not available; U-TCA=urinary metabolite of trichloroethylene.

aInternal cohort analyses for peak and cumulative exposure to trichloroethylene classifications used Cox proportional-hazards models.

b90% CI.

LUNG CANCER

Description of Case-Control Studies

Three population-based case-control studies (Table 6.14) reported the risk of lung cancer associated with job title and self-reported exposure to specific chemicals (Brownson et al., 1993; Pohlabeln et al., 2000) or to tetrachloroethylene in drinking water (Paulu et al., 1999). Smoking is a known risk factor for lung cancer and was accounted for in each of the studies (see Chapter 2 and Appendix E for more information on smoking).

Epidemiologic Studies of Exposure to Organic Solvents and Lung Cancer

No association was found between incidence of lung cancer and concentrations of the biologic marker of exposure to trichloroethylene (SIR=0.92, 95% CI=0.59–1.35) (Anttila et al., 1995). A cohort of aircraft-manufacturing workers with potential exposure to trichloroethylene was followed for nearly 30 years, but no association between lung cancer and trichloroethylene was reported (SMR=0.76, 95% CI=0.60–0.95) (Boice et al., 1999). The authors noted that most workers were exposed to a variety of substances routinely or intermittently. A mortality study of civilian aircraft-maintenance workers at Hill Air Force Base in Utah that included an extensive assessment of exposure to trichloroethylene did not show an increase in lung cancer mortality or incidence in men (over 25 unit-years: SMR=1.1, 95% CI=0.7–1.8) or women (over 25 unit-years: SMR=0.4, 95% CI=0.1–1.8) (Blair et al., 1998). The cohort study of aircraft manufacturers in Arizona (Morgan et al., 1998) did not show increased mortality from lung cancer in the trichloroethylene-exposed subcohort (SMR=1.10, 95% CI=0.89–1.34).

In addition to those studies, no associations were found for rubber-industry workers exposed to trichloroethylene (OR=0.64) (Wilcosky et al., 1984) or to transformer-assembly workers (OR=1.01, 95% CI=0.69–1.47) (Greenland et al., 1994).

A cohort study of dry-cleaning workers exposed to tetrachloroethylene and other solvents showed increased mortality from lung cancer (SMR=1.46, 95% CI=1.07–1.95) (Ruder et al., 2001). In the subcohort of workers exposed only to tetrachloroethylene, the relative risks were not as great (SMR=1.17, 95% CI=0.71–1.83). Blair and colleagues (1990) reported a comparable relative risk of lung cancer in another cohort of dry cleaners (SMR=1.3, 95% CI=0.9–1.7).

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

TABLE 6.14 Description of Case-Control Studies of Lung Cancer and Exposure to Organic Solvents

Reference

Description of Study Population

Number of Cases

Number of Controls

Relevant Exposure(s)

Determination of Exposure

Analysis

Adjustment for Confounding

Brownson et al., 1993

White, female cases, age 30–84 years, identified through the Missouri Cancer Registry as diagnosed in 1986–1991, with 77% histologic confirmation; controls younger than 65 years selected from state driver’s license files; controls older than 65 years were selected from the records of HCFA

429

1,021

Dry-cleaning work

Printing-industry work

Trained interviewers conducted telephone and in-person interviews (direct or proxy) to assess specific occupational (job titles) and exposure history (self-reported)

Multiple logistic regression

Age, history of lung disease, active smoking

Response rates: 66% of cases, 67% of controls

Paulu et al., 1999

Cases reported to the Massachusetts Cancer Registry, diagnosed in 1983–1986 among residents of five upper Cape Cod towns; living controls were selected from the records of HCFA and through RDD; deceased controls identified by the state Department of Vital Statistics and Research files

243

1,206

Tetrachloroethylene

Exposure dose estimated in areas of contaminated drinking water, accounting for location and years of residence, water flow, pipe characteristics

Multiple logistic regression

Age at diagnosis, vital status, sex, occupational exposure to solvents; specific cancer risk factors controlled in respective analyses, such as smoking

Response rates: 79% of cases, 76% of HCFA controls, 74% of RDD controls, 79% of next of kin of deceased controls

Pohlabeln et al., 2000

Nonsmoking cases identified in 12 study centers from seven European countries between 1988–1994 (96.5% were histologically confirmed); community-based controls were selected in six centers, hospital-based controls in five centers, and both community and hospital-based controls in one center

650

1,542

Laundry and dry cleaners

Interviewers used a common questionnaire to record lifelong occupational histories which were coded by job titles

Unconditional logistic regression

Sex, occupation, age, center, occasional smoking (ever smoked occasionally, but fewer than 400 cigarettes), residence, diet, and environmental tobacco smoke

Response rates: across centers the rate varied between 55% and 95%; in three centers, the response rate among control subjects was lower than 50%

NOTE: HCFA=Health Care Financing Administration; RDD=random-digit dialing.

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

An increased risk of lung cancer was found in workers exposed to tetrachloroethylene (SIR=1.92, 95% CI=0.62–4.48) (Anttila et al., 1995). Among the cohort of aircraft manufacturers, no association was found between exposure to tetrachloroethylene and lung cancer (SMR=1.08, 95% CI=0.79–1.44) (Boice et al., 1999). The authors indicated that concentrations of tetrachloroethylene in the air samples were relatively low and that exposures were well below permissible concentrations. Wilcosky and colleagues (1984), in a study of rubber-industry workers, reported no increased risk of lung cancer with exposure to tetrachloroethylene.

The risk of developing lung cancer in relation to exposure to tetrachloroethylene was evaluated in the two case-control studies, and these were adjusted for smoking. A risk associated with self-reported exposure to tetrachloroethylene was found among lifetime nonsmokers (OR=2.1, 95% CI=1.2–3.7) (Brownson et al., 1993). A case-control study of residents of upper Cape Cod (Paulu et al., 1999) showed excess risks of lung cancer with increasing level of estimated exposure to tetrachloroethylene in drinking water. A third case-control study of female, nonsmoking laundry and dry cleaners also found an increased risk of lung cancer (OR=1.83, 95% CI=0.98–3.40) (Pohlabeln et al., 2000).

Several cohort studies of workers exposed to methylene chloride provided little support for an association between exposure and lung cancer (Gibbs et al., 1996; Hearne and Pifer, 1999; Hearne and Friedlander, 1981; Hearne et al., 1987, 1990; Lanes et al., 1990, 1993; Tomenson et al., 1997). Those large cohort studies followed methylene chloride-exposed workers for many years, some with repeated followup, and examined the association between exposure and cancer mortality, but they reveal no excess of lung cancer associated with exposure.

Yin and colleagues (1996a,b) reported increased relative risks of cancers of the trachea, bronchi, and lung combined in benzene-exposed males (RR=1.5, 95% CI=1.0–2.2) and in the entire cohort of exposed men and women (RR=1.4, 95% CI=1.0–2.0); no association was found in benzene-exposed women (RR=1.0, 95% CI=0.4–2.9). Hayes and colleagues (1996) assessed the cumulative exposure to benzene in the overall cohort and found that the relative risks of tracheal, bronchial, and lung cancers combined increased with increasing exposure (p trend=0.01).

Two nested case-control studies, one of transformer-assembly workers (Greenland et al., 1994) and another of rubber-industry workers (Wilcosky et al., 1984), did not show any associations between occupational exposure to benzene and lung cancer. No data were presented on increasing levels of exposure.

Three studies provided evidence on the association between toluene and lung cancer. A cohort of toluene-exposed German rotogravure workers showed increased lung cancer mortality when compared with the mortality in West Germany (SMR=1.23, 95% CI=0.81–1.92) (Wiebelt and Becker, 1999). Job-specific subcohorts with different levels of exposure to toluene demonstrated a range of lung cancer risk, from no risk in printing-cylinder preparation occupations to an SMR of 1.77 (95% CI=0.77–4.39) in finishing workers who had the lowest level of exposure over the entire observation period.

A Swedish cohort of rotogravure printers showed similar increases in respiratory tract cancer mortality from exposure to toluene (SMR=1.76, 95% CI=1.03–2.91) (Svensson et al., 1990); however, no gradient was found with duration of exposure.

Factory workers who were considered to have high exposure in a plant manufacturing chlorinated toluenes (benzyl chloride, benzal chloride, benzotrichloride, and benzoyl chloride) experienced increased lung cancer mortality (SMR=3.31, 95% CI=1.59–6.09) (Sorahan and

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Cathcart, 1989), and there was evidence that risk increased with increasing exposure. In a nested case-control study of 26 cases of lung cancer from the cohort of chlorinated toluene production workers, RR of lung cancer associated with benzotrichloride and “other chlorinated toluenes” was 1.36 (95% CI=0.43–24) and 1.12 (95% CI=0.30–4.22), respectively, per 10 years of exposed employment. In a study of rubber workers, Wilcosky and colleagues (1984) did not find an association (OR=0.55).

There were three studies of the association between exposure to phenol and lung cancer. A cohort study with many exposed cases showed no association (Dosemeci et al., 1991), and no association was reported for workers in the rubber industry (Wilcosky et al., 1984). Kauppinen and colleagues (1993) reported a four-fold excess risk (SMR=4.04, 95% CI=1.83–8.89) after adjusting for smoking among those exposed for at least 1 month.

Several other solvents used in the Gulf War—including naphtha, ethanol, xylenes, isopropanol, ethyl acetate, and acetone (Wilcosky et al., 1984)—were investigated in relation to lung cancer, and no associations with these solvents were found. Anttila and colleagues (1995) found an increase in lung and bronchial cancer risk with exposure to 1,1,1-trichloroethane (SIR=1.31, 95% CI=0.16–4.71).

Several studies, including some described previously, examined the association of lung cancer with unspecified mixtures of solvents. Most studies did not identify “solvents” or “organic solvents” as the exposure being evaluated but instead defined exposure by various occupational titles or groups, such as painters (Engholm and Englund, 1982; Englund 1980; Matanoski et al., 1986; Morgan et al., 1981; Steenland and Palu, 1999; Stockwell and Matanoski, 1985), printers (Malker and Gemne, 1987), workers in transformer assembly (Greenland et al., 1994), ethanol and isopropanol production (Teta et al., 1992), isopropanol and methyl ethyl ketone production (Alderson and Rattan, 1980), shoe manufacturing (Walker et al., 1993), chemical manufacturing (Waxweiler et al., 1981), aircraft manufacturing (Boice et al., 1999), or metal-component manufacturing (Acquavella et al., 1993). As shown in Table 6.15, many of the studies showed positive associations between occupational exposure and lung cancer.

Generally, workers in the studies were exposed to solvents and other chemical agents not reviewed by this committee. More important, although smoking is associated with lung cancer, no adjustments for it were made in most cohort studies. (That is not unusual; obtaining smoking histories, especially in mortality studies, is difficult.) Furthermore, one study that specifically examined lung cancer risk among nonsmokers in the printing industry found no association with occupational exposure (Brownson et al., 1993). Unlike farm workers, who have some of the lowest smoking rates in the United States, painters, truck drivers, construction workers, carpenters, and auto mechanics have some of the highest rates. For example, in a National Center for Health Statistics survey of men 20–64 years old in 1978–1980, 55.1% of painters surveyed reported that they smoked (US Surgeon General, 1985).

To understand the impact of smoking on risk of lung cancer, the committee examined the risk estimates for bladder cancer and cardiovascular disease—diseases for which smoking is also a known risk factor. Almost all the studies reported rates of bladder cancer and cardiovascular disease similar to those of lung cancer. Almost all the studies also stated that asbestos was present in the workplace and probably contributed to the slightly increased observed risks of lung cancer. As a result, the committee determined that exposure to solvents alone was an unlikely explanation for the increased risks of lung cancer and that confounding by smoking was possibly biasing the results.

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Summary and Conclusion

Although there are different types of respiratory cancers, most studies assessed exposure in relation to lung cancer. Exposures were typically defined by occupations in which the solvents were known to be present; and confounding factors, especially smoking, were not consistently controlled for in the analyses. In reviewing the literature, the committee found that there was no evidence from any of the studies of a positive association between exposure to trichloroethylene, methylene chloride, benzene, phenol, and other specific solvents and risk of lung cancer.

For exposure to tetrachloroethylene and dry-cleaning solvents, some committee members believed that the overall evidence was limited by the possibility of confounding, and that most findings were based on small numbers of cases exposed by different routes. The cohort studies did not control for other occupational exposures or smoking, an important potential confounder for lung cancer. However, other committee members believed that the consistently positive findings and evidence of a dose-response relationship in the case-control study by Paulu and colleagues were supportive of a conclusion of limited/suggestive evidence. Both case-control studies adjusted for smoking and still found relatively high relative risks of lung cancer. As a result, the committee decided not to state a formal consensus conclusion. Additional studies that control for smoking and address other concerns related to misclassification of exposure are needed before a more definitive conclusion as to exposure to tetrachloroethylene and dry-cleaning solvents and the risk of lung cancer can be reached.

Although several cohort studies reported increased risk of lung cancer associated with exposure to toluene, estimates were weak. Workers were probably exposed to other compounds that were not controlled for in the analyses. In addition, information on smoking was not available. A unique relationship between solvent exposure and lung cancer was not found, and the committee concluded that the evidence was inadequate/insufficient to support an association between exposure to unspecified mixtures of organic solvents and lung cancer.

Several studies on specific organic solvents and solvent mixtures found positive associations between exposure and the risk of lung cancer. However, most studies on specific solvents were too small and inconsistent in their findings to support conclusions. Some studies showed positive associations, but they were limited by lack of information on smoking habits among cohort members.

With respect to exposure to solvent mixtures, many studies reported positive findings; most, however, were based on occupational titles or industries and lacked specific analyses of “solvents” or “organic solvents.” Although their results suggested a possible relationship, the lack of smoking data, the lack of exposure specificity, and the potential for confounding by other occupational exposures (such as to asbestos) limited their utility. Future research with sufficient power would help to clarify whether an association between exposure to solvent mixtures or the interactions of various solvents and lung cancer exists, as is indicated in some of the studies reviewed by the committee. Table 6.15 identifies the key studies and the relevant data points evaluated by the committee in drawing its conclusion. Unless indicated in the tables, the study populations include both men and women.

The committee concludes, from its assessment of the epidemiologic literature, that there is inadequate/insufficient evidence to determine whether an association exists between chronic exposure to solvents under review, other than tetrachloroethylene and dry-cleaning solvents, and lung cancer.

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

TABLE 6.15 Selected Epidemiologic Studies—Lung Cancer and Exposure to Organic Solvents

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Trichloroethylene

Cohort Studies—Incidence

Blair et al., 1998

Aircraft-maintenance workers in Utah

 

 

Males

 

 

No exposure

22

1.0 (0.5–1.9)

 

<5 unit-years

24

1.0 (0.6–2.0)

 

5–25 unit-years

11

0.8 (0.4–1.6)

 

>25 unit-years

15

0.8 (0.4–1.7)

 

Females

 

 

No exposure

0

 

<5 unit-years

1

0.6 (0.1–5.3)

 

5–25 unit-years

0

 

>25 unit-years

0

Anttila et al., 1995

Biologically monitored workers in Finland

 

 

Entire period since first measurement

25

0.92 (0.59–1.35)

 

0–9 years

11

1.19 (0.59–2.13)

 

10–19 years

9

0.67 (0.30–1.26)

 

≥20 years

5

1.11 (0.36–2.58)

 

Mean personal U-TCA level

 

 

<100 µmol/L

16

1.02 (0.58–1.66)

 

100+ µmol/L

7

0.83 (0.33–1.71)

Cohort Studies—Mortality

Boice et al., 1999

Aircraft-manufacturing workers in California

 

 

All exposed factory workers

78

0.76 (0.60–0.95)

 

Duration of potential exposure (routine or intermittent)

 

 

<1 year

66

0.85 (0.65–1.13)

 

1–4 years

63

0.98 (0.74–1.30)

 

≥5 years

44

0.64 (0.46–0.89)

Blair et al., 1998

Aircraft-maintenance workers in Utah

 

 

Males

 

 

No exposure

51

1.0 (0.7–1.6)

 

<5 unit-years

43

1.0 (0.6–1.6)

 

5–25 unit-years

23

0.9 (0.5–1.6)

 

>25 unit-years

38

1.1 (0.7–1.8)

 

Females

 

 

No exposure

2

0.4 (0.1–1.6)

 

<5 unit-years

2

0.6 (0.1–2.4)

 

5–25 unit-years

11

0.6 (0.1–4.7)

 

>25 unit-years

2

0.4 (0.1–1.8)

Morgan et al., 1998

Aerospace workers in Arizona

 

 

Entire trichloroethylene-exposed cohort

97

1.10 (0.89–1.34)

 

Cumulative

 

 

Low

45

1.49 (1.09–1.99)

 

High

52

0.90 (0.67–1.20)

 

Peak: medium and high vs low and no exposure

64

1.07 (0.82–1.40)

Greenland et al., 1994

White male transformer-assembly workers, ever exposed

NA

1.01 (0.69–1.47)

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Wilcosky et al., 1984

Rubber-industry workers in Ohio

 

 

Cumulative exposure of more than 1 year

 

 

White males

11

0.64

Tetrachloroethylene and Dry-cleaning Solvents

Cohort Studies—Mortality

Ruder et al., 2001

US dry-cleaning workers in four labor unions

 

 

Males, nonwhite

25

1.52 (1.05–2.39)

 

Females, nonwhite

16

1.88 (1.07–3.05)

 

Exposed to tetrachloroethylene only

19

1.17 (0.71–1.83)

 

Exposed to tetrachloroethylene and other dry-cleaning solvents

46

1.46 (1.07–1.95)

Boice et al., 1999

Aircraft manufacturing workers in California

 

 

All exposed factory workers

46

1.08 (0.79–1.44)

 

Duration of potential exposure (routine or intermittent)

 

 

<1 year

33

1.15 (0.80–1.66)

 

1–4 years

51

1.09 (0.80–1.48)

 

≥5 years

36

0.71 (0.49–1.02)

Anttila et al., 1995

Biologically monitored workers in Finland

5

1.92 (0.62–4.48)

Blair et al., 1990

Dry-cleaning union members in Missouri

47

1.3 (0.9–1.7)

Wilcosky et al., 1984

Rubber-industry workers in Ohio

 

 

Cumulative exposure of more than 1 year

 

 

White males

2

0.26

Case-Control Studies

Pohlabeln et al., 2000

Occupational exposure among nonsmoking females in Europe

 

 

Laundry and dry cleaners for at least 6 months

19

1.83 (0.98–3.40)

Paulu et al., 1999

Residents in upper Cape Cod

 

 

>75th percentile tetrachloroethylene-water exposure

 

 

0-year latent period

11

1.8 (0.8–3.9)

 

5-year latent period

6

1.7 (0.6–4.5)

 

7-year latent period

5

1.6 (0.5–4.4)

 

9-year latent period

4

1.8 (0.5–6.0)

 

>90th percentile tetrachloroethylene-water exposure

 

 

0-year latent period

5

3.7 (1.0–11.7)

 

5-year latent period

3

3.3 (0.6–13.4)

 

7-year latent period

3

6.2 (1.1–31.6)

 

9-year latent period

3

19.3 (2.5–141.7)

Brownson et al., 1993

Occupational exposure among females in Missouri—Dry-cleaning

 

 

Lifetime nonsmokers

23

2.1 (1.2–3.7)

 

Exposure range

 

 

Low: ≤1.125 years

NA

0.6 (0.2–1.7)

 

High: ≥1.125 years

NA

2.9 (1.5–5.4)

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Methylene Chloride

Cohort Studies—Mortality

Hearne et al., 1999

Male cellulose triacetate photographic-film base workers in Kodak Park, employed >1 year

 

 

Methylene chloride cohort

27

0.75 (0.49–1.09)

 

Roll-coating cohort

 

 

New York state external control

28

0.82 (0.55–1.19)

 

Kodak Rochester external control

28

0.89 (0.59–1.29)

Tomensen et al., 1997

Male cellulose triacetate-fiber production workers in the UK, ever employed

 

 

Never exposed

1

0.16 (0–0.88)

 

All exposed

19

0.48 (0.29–0.75)

 

Cumulative exposure (nonsmokers)

 

 

<400 ppm-years

6

0.32

 

400–799 ppm-years

2

0.51

 

≥800 ppm-years

1

0.37

 

Unassigned exposure

10

0.68

Gibbs et al., 1996

Cellulose triacetate-fiber production workers in Maryland

 

 

Males

 

 

No exposure

6

0.59 (0.22–1.29)

 

Low exposure

20

0.78 (0.48–1.20)

 

High exposure

15

0.55 (0.31–0.91)

 

Females

 

 

No exposure

0

NA (0.0–4.92)

 

Low exposure

9

1.09 (0.50–2.07)

 

High exposure

2

2.29 (0.28–8.29)

Lanes et al., 1993

Cellulose triacetate-fiber production workers in South Carolina (cohort), employed >3 months

13

0.80 (0.43–1.37)

Benzene

Cohort Studies—Mortality

Yin et al., 1996a

Chinese factory workers, ever exposed

 

 

Total cohort

125

1.4 (1.0–2.0)

 

Males

109

1.5 (1.0–2.2)

 

Females

16

1.0 (0.4–2.9)

Hayes et al., 1996

Chinese factory workers

 

 

Cumulative exposure

 

 

No exposure

41

1.0

 

<10 ppm-years

10

1.2

 

10–39 ppm-years

13

1.0

 

40–99 ppm-years

19

1.4

 

100–400 ppm-years

38

1.4

 

400+ ppm-years

41

1.7

 

 

 

p-trend=0.01

Greenland et al., 1994

White male transformer-assembly workers, ever exposed

NA

0.58 (0.31–1.07)

Wilcosky et al., 1984

Rubber-industry workers in Ohio

 

 

Cumulative exposure of more than 1 year

 

 

White males

23

0.69

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Toluene

Cohort Studies—Mortality

Wiebelt and Becker, 1999

Male German rotogravure printing-plant workers, employed >1 year

 

 

Total cohort

44

1.23 (0.81–1.92)

 

Printing-cylinder preparation workers

7

0.83 (0.20–2.77)

 

Printing/proof printing workers

25

1.30 (0.72–2.49)

 

Finishing workers

13

1.77 (0.77–4.39)

Svensson et al., 1990

Male rotogravure printing-plant workers in Sweden

 

 

Total cohort

16

1.76 (1.03–2.91)

 

≥5 year exposure, >10 year latency

9

1.26 (0.57–2.38)

Sorahan and Cathcart, 1989

Male chemical-factory workers

 

 

Low exposure to chlorinated toluenes

16

1.39 (0.80–2.27)a

 

High exposure to chlorinated toluenes

10

3.31 (1.59–6.09)a

 

Benzotrichloride

NA

1.36 (0.43–4.24)

 

Other chlorinated toluenes

NA

1.12 (0.30–4.22)

Wilcosky et al., 1984

White, male rubber-industry workers in Ohio

 

 

Cumulative exposure for more than 1 year

3

0.55

Phenol

Cohort Studies—Mortality

Kauppinen et al., 1993

Finnish woodworkers

 

 

Any exposure (>1 month)

5

4.04 (1.83–8.89)

 

Duration >5 years

6

3.08 (0.70–13.6)

Dosemeci et al., 1991

US white male industrial workers

 

 

Any exposure

146

1.1 (0.9–1.3)

 

Level of cumulative exposure

 

 

None

70

1.2

 

Low exposure

68

1.2

 

Medium exposure

60

1.1

 

High exposure

18

1.4

Wilcosky et al., 1984

White male rubber-industry workers in Ohio

 

 

Cumulative exposure for more than 1 year

 

 

White males

13

0.95

 

Black males

2

0.91

Other Organic Solvents

Cohort Studies—Mortality

Anttila et al., 1995

Biologically monitored workers in Finland

 

 

1,1,1-trichloroethane ever measured in urine

2

1.31 (0.16–4.71)

Wilcosky et al., 1984

Rubber-industry workers in Ohio

 

 

Cumulative exposure for more than 1 year

 

 

Specialty naphthas (white males)

43

0.70

 

Specialty naphthas (black males)

2

0.39

 

Ethanol (white males)

21

1.0

 

Xylenes (white males)

10

0.61

 

Ethyl acetate (white males)

6

0.84

 

Acetone (white males)

5

0.86

 

Isopropanol (white males)

27

0.64

 

Isopropanol (black males)

2

0.56

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Unspecified Mixtures of Organic Solvents

Cohort Studies—Incidence

Malker and Gemne, 1987

Swedish printing-industry workers, employed in 1960

 

 

Males

190

1.5 (1.2–1.7)b

 

Females

9

1.3

Engholm and England, 1982

Male members of the Swedish Painters Union

 

 

Years since entry into the union

 

 

≥0

81

1.28 (p<0.05)

 

≥5

75

1.24 (p<0.05)

 

≥10

74

1.31 (p<0.05)

 

≥15

66

1.28 (p<0.05)

 

≥20

58

1.26 (p<0.05)

 

≥25

51

1.32 (p<0.05)

Cohort Studies—Mortality

Boice et al., 1999

Aircraft-manufacturing workers in California

 

 

Potential routine exposure to mixed solvents

221

0.88 (0.77–1.01)

Steenland and Palu, 1999

Members of US painters unions

 

 

Total cohort

1746

1.23 (1.17–1.29)

 

20 years since first union membership

1360

1.24 (1.18–1.31)

Greenland et al., 1994

White male transformer-assembly workers, ever exposed

NA

1.57 (1.08–2.27)

Acquavella et al., 1993

Metal-components manufacturing workers

 

 

Solvents, ever exposed

4

1.9 (0.5–4.9)

Walker et al., 1993

Shoe-manufacturing workers in Ohio, employed >1 month

 

 

Total cohort

99

1.47 (1.20–1.80)

 

Males

68

1.56 (1.22–1.99)

 

Females

31

1.30 (0.89–1.86)

Teta et al., 1992

Male ethanol and isopropanol production workers

 

 

South Charleston SC plant

 

 

All workers, ever employed

14

0.87 (0.5–1.5)

 

Workers in exposed unit ≥10 years

2

0.56

 

Texas City, TX plant

 

 

All workers, ever employed

8

1.10 (0.5–2.2)

 

Workers in exposed unit ≥10 years

1

NA

Matanoski et al., 1986

US painters and allied tradesmen

 

 

Total cohort

448

1.06 (0.96–1.16)

Stockwell and Matanoski, 1985

Male construction and maintenance painters in New York

 

Usual occupation of painter

51

2.75 (1.45–5.21)

Engholm and England, 1982

Male members of the Swedish Painters Union

 

 

Years since entry into the union:

 

 

≥0

124

1.27 (p<0.05)

 

≥5

118

1.25 (p<0.05)

 

≥10

114

1.28 (p<0.05)

 

≥15

103

1.27 (p<0.05)

 

≥20

92

1.26 (p<0.05)

 

≥25

80

1.27 (p<0.05)

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Morgan et al., 1981

Male US paint or varnish manufacturing workers

 

 

Solvents excluding lacquer, 1+ years of exposure

51

1.14

Waxweiler et al., 1981

Male synthetic-chemical plant workers, ever employed

42

1.49 (1.08–2.03)a

Alderson and Rattan, 1980

Male workers at dewaxing plants in the UK, employed >1 year

 

 

Workers in isopropanol alcohol plant

2

0.78 (0.09–2.81)a

 

Workers in methyl ethyl ketone plant

1

0.17 (0.00–0.93)a

Englund, 1980

Male Swedish painters, ever certified or union member

124

1.27 (1.06–1.52)a

Case-Control Study

Brownson et al., 1993

Occupational exposure among females in Missouri—Printing industry

 

 

Lifetime nonsmokers

6

0.8 (0.3–2.0)

 

Exposure range

 

 

Low: ≤8 years

NA

0.6 (0.2–2.2)

 

High: >8 years

NA

1.3 (0.5–3.7)

NOTE: NA=not available; U-TCA=urinary metabolite of trichloroethylene.

a95% CI calculated by the committee with standard methods from the observed and expected numbers presented in the original study.

b99% confidence limits.

BONE CANCER

Epidemiologic Studies of Exposure to Organic Solvents and Bone Cancer

Blair and colleagues (1998) extended the followup of a cohort of aircraft-manufacturing workers (Spirtas et al., 1991) that used a detailed exposure-assessment method. An increased bone cancer risk was reported after adjustment for age, calendar time, and sex (RR=2.1, 95% CI =0.2–18.8) in workers exposed to trichloroethylene.

One study assessed whether an association existed between exposure to benzene and bone cancer. To evaluate the specific relationship between exposure to benzene and cancer risk, Wong (1987a) examined mortality among workers employed in seven chemical-manufacturing plants. An increased relative risk of bone cancer was found in the exposed group of workers (SMR=3.17, 95% CI=0.38–11.46). Wong (1987b) also estimated exposure to benzene in terms of 8-hour TWAs and peak levels of exposure and found that the relative risk of bone cancer increased with duration of exposure (SMR=6.63; one exposed case). Those results were inconclusive for an association because of the very small number of exposed cases, which resulted in highly variable risk estimates.

Fu and colleagues (1996) examined two historical cohorts of shoe workers in England and Florence, Italy, and used job titles to assess cancer mortality in relation to exposure to leather dusts and solvents. A slight increase in bone cancer was observed in the English cohort among those with probable solvent exposure (SMR=1.12, 95% CI=0.03–6.26), and no cases of bone cancer were reported in the Florence cohort. The committee reviewed the study by Nielsen and colleagues (1996) in which the risk of bone cancer in a cohort of lithographers was examined. Only one exposed case was observed (SIR=11.4, 95% CI=0.6–56.0).

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Summary and Conclusion

The relationship between exposure to organic solvents and bone cancer was reported in only four cohort studies, representing three different solvent exposures. Each study had low power, and exposure assessment relied primarily on job titles as surrogates of exposure. No case-control studies of bone cancer were identified. Studies with larger numbers of exposed cases (increased power) and more precise exposure assessment are needed for the committee to evaluate the relationship between solvent exposures and risk of bone cancer. Table 6.16 identifies the key studies and relevant data points reviewed by the committee in drawing its conclusion. Unless indicated in the table, the study populations include both men and women.

The committee concludes, from its assessment of the epidemiologic literature, that there is inadequate/insufficient evidence to determine whether an association exists between chronic exposure to solvents under review and bone cancer.

TABLE 6.16 Selected Epidemiologic Studies—Bone Cancer and Exposure to Organic Solvents

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Trichloroethylene

Cohort Study—Mortality

Blair et al., 1998

Aircraft maintenance workers in Utah, employed in exposed area >1 year

5

2.1 (0.2–18.8)

Benzene

Cohort Study—Mortality

Wong, 1987a,b

Male Chemical Manufacturers Association workers

 

 

Continuous exposure

2

3.17 (0.38–11.46)

 

Duration of exposure

 

 

<5 years

1

2.63

 

5–14 years

1

6.63

 

≥15 years

0

Unspecified Mixtures of Organic Solvents

Cohort Study—Incidence

Nielsen et al., 1996

Danish lithographers, ever employed

1

11.4 (0.6–56.0)

Cohort Study—Mortality

Fu et al., 1996

Shoe-manufacturing workers

 

 

English cohort, employed in 1939

6

2.08 (0.76–4.52)

 

Probable solvent based on work area

1

1.12 (0.03–6.26)

 

Florence cohort, ever employed

0

0 (0.0–3.45)

SOFT TISSUE SARCOMA

Because of the lack of available studies on the relationship between exposure to organic solvents and soft tissue sarcomas (STS), a conclusion regarding association could not be drawn. Only one study (Serraino et al., 1992) identified by the committee analyzed the relationship between relevant exposures reviewed in this report (“benzene/solvents” and “dyes/paints”) and STS. Although the population-based case-control study observed an increased risk of STS among men exposed to “benzene/solvents” for more than 10 years (OR=2.2, 95% CI=0.9–5.5), the study was limited by the use of self-reported exposures. Additional studies are needed to support the relationship before a conclusion regarding association can be drawn.

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

SKIN CANCER

Description of Case-Control Studies

Table 6.17 identifies the study characteristics of two papers from the Montreal multisite cancer case-control study on the association between exposure to specific organic solvents and solvent mixtures and the risk of melanoma. The study was designed so that in-person interviews were used mostly, and telephone interviews or self-administered questionnaires were limited to next of kin or hard-to-interview subjects. The interviews included a job-specific module to obtain detailed information on each job that a subject held in the entire working history, such as dates of employment, the employer’s activities and products, job tasks, and work environment. Using the job-history information, a team of industrial hygienists and chemists estimated exposures to about 300 of the most common occupational agents. The study population was used for exposure-specific and cancer-specific studies. Fritschi and Siemiatycki (1996a) evaluated the relationship between melanoma and exposure to 85 chemical substances, 13 occupations, and 20 industries. The same set of cases was studied by Gérin and colleagues (1998) in relation to occupational exposure to the hydrocarbons benzene, toluene, xylene, and styrene.

Epidemiologic Studies of Exposure to Organic Solvents

No increase in melanoma mortality was found among two cohorts of aircraft-maintenance workers (Blair et al., 1998; Boice et al., 1999), a cohort of workers monitored for a metabolite of trichloroethylene (Hansen et al., 2001), or a cohort of uranium-processing workers (Ritz, 1999). An increased risk of melanoma (for “any exposure,” OR=3.6, 95% CI=1.5–9.1) was found in a case-control study in Montreal (Fritschi and Siemiatycki, 1996a). Two studies of workers monitored for exposure to trichloroethylene found mixed results for nonmelanoma skin cancers (Axelson et al., 1994; Hansen et al., 2001). The studies were based on small numbers of exposed cases and did not control for exposure to sunlight, an important confounding variable.

Two studies of tetrachloroethylene-exposed workers, one in aircraft manufacturing (Boice et al., 1999) and one of dry-cleaning workers (Blair et al., 1990), showed no association between exposure and skin cancer risk. Boice and colleagues examined melanoma specifically, and Blair and colleagues looked at all skin cancers combined.

Several other solvents thought to have been used in the Gulf War—including methylene chloride, benzene, toluene, xylene, and phenol—were investigated in relation to melanoma. The committee identified only one relevant epidemiologic study for each substance. A cohort study of cellulose-fiber production workers showed an increased risk of melanoma mortality (SMR=1.94 95% CI=0.24–7.00) (Lanes et al., 1993). No increased risks were reported for melanoma and exposure to benzene, toluene, or xylene (Gérin et al., 1998) or for any type of skin cancer and exposure to phenol (Dosemeci et al., 1991).

Melanoma was not found to be associated with exposure to unspecified mixtures of solvents, as reported in several occupational studies (Anttila et al., 1995; Berlin et al., 1995; Boice et al., 1999; Bourguet et al., 1987; Fritschi and Siemiatycki, 1996a). The findings on risk of nonmelanoma skin cancer and unspecified mixtures of solvents were inconsistent but tended to be negative. Studies that found positive associations were variable and based on few exposed cases, and sunlight exposure was not controlled for.

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

TABLE 6.17 Description of Case-Control Studies of Melanoma Skin Cancers and Exposure to Organic Solvents

Reference

Description of Study Population

Number of Cases

Number of Controls

Relevant Exposure(s)

Determination of Exposure

Analysis

Adjustment for Confounding

Fritschi and Siemiatycki, 1996a

Male cases and controls, age 35–70 years, diagnosed in 19 large Montreal-area hospitals in 1979–1985 and histologically confirmed for one of 19 anatomic cancer sites; age-matched, population-based controls were also chosen from electoral lists and RDD (see also Gérin et al., 1998)

103 melanoma

1,066 subjects for each site, consisting of

533 population controls and

533 randomly selected subjects from the eligible cancer control group

Trichloroethylene

Solvents

In-person interviews with segments on work histories (job titles); exposures attributed by a team of chemists and industrial hygienists

Unconditional logistic regression

Age, years of schooling, ethnicity

Response rates: 83% of all cases, 71% of population controls

Gérin et al., 1998

Male cases and controls, age 35–70 years, diagnosed in 19 large Montreal-area hospitals in 1979–1985 and histologically confirmed for one of 19 anatomic cancer sites; age-matched, population-based controls were also chosen from electoral lists and RDD (see also Fritschi and Siemiatycki, 1996a)

103 melanoma

1,066 subjects for each site, consisting of

533 population controls and

533 randomly selected subjects from the eligible cancer control group

Benzene

Toluene

Xylene

In-person interviews with segments on work histories (job titles); exposures attributed by a team of chemists and industrial hygienists

Unconditional logistic regression

Age, family income, ethnicity, cigarette smoking, respondent status

Response rates: 82% of all cases, 71% of population controls

NOTE: RDD=random-digit dialing

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Summary and Conclusion

Several studies of specific and mixed solvents examined the role of exposure and the risk of skin cancer, specifically melanoma and nonmelanoma. However, small numbers of exposed cases and the lack of validated exposure assessment were limitations of the studies. Almost all studies on specific organic solvents and unspecified mixtures of organic solvents found no association between exposure and incidence or mortality from melanoma, nonmelanoma skin cancer, or skin cancer in general. Several studies used biologic monitoring to assess exposure (e.g., Anttila et al., 1995; Axelson et al., 1994; Hansen et al., 2001), but the inability of most studies to control for well-established risk factors of skin cancer—such as age, ethnicity, geography, presence of nevi, and time spent in the sun—limits the validity of their findings. Tables 6.1819 identify the key studies and relevant data points reviewed by the committee in drawing its conclusions. Unless indicated in the tables, the study populations include both men and women.

The committee concludes, from its assessment of the epidemiologic literature, that there is inadequate/insufficient evidence to determine whether an association exists between chronic exposure solvents under review and melanoma or nonmelanoma skin cancer.

TABLE 6.18 Selected Epidemiologic Studies—Melanoma Skin Cancers and Exposure to Organic Solvents

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Trichloroethylene

Cohort Study—Incidence

Hansen et al., 2001

Biologically monitored male Danish workers

2

0.9 (0.1–3.4)

Cohort Studies—Mortality

Boice et al., 1999

Aircraft-manufacturing workers, ever exposed

2

0.46 (0.06–1.67)

Ritz, 1999

White male uranium-processing workers, ever exposed

 

 

Melanoma and nonmelanoma

4

0.64 (0.17–1.63)

Blair et al., 1998

Aircraft-maintenance workers, ever exposed

9

1.0 (0.3–3.1)

Case-Control Study

Fritschi and Siemiatycki, 1996a

Male residents of Montreal, Canada

 

Any exposure

8

3.6 (1.5–9.1)

 

Substantial exposure

4

3.4 (1.0–12.3)

Tetrachloroethylene and Dry-cleaning Solvents

Cohort Studies—Mortality

Boice et al., 1999

Aircraft-manufacturing workers, ever exposed

2

0.95 (0.12–3.43)

Blair et al., 1990

Dry-cleaning union members in Missouri, ever employed

 

 

Melanoma and nonmelanoma

2

0.8 (0.1–2.8)

Other Specific Organic Solvents

Cohort Studies—Mortality

Lanes et al., 1993

Cellulose-fiber production workers in South Carolina, exposed to methylene chloride

2

1.94 (0.24–7.00)

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Dosemeci et al., 1991

Male US industrial workers, ever exposed to phenol

 

 

Melanoma and nonmelanoma

7

0.9 (0.4–1.8)

Case-Control Study

Gérin et al., 1998

Male residents of Montreal, Canada, any exposure

 

 

Benzene

11

0.6 (0.3–1.2)

 

Toluene

5

0.4 (0.1–0.9)

 

Xylene

3

0.3 (0.1–0.8)

Unspecified Mixtures of Organic Solvents

Cohort Studies—Incidence

Anttila et al., 1995

Workers in Finland biologically monitored for halogenated hydrocarbons

5

0.71 (0.23–1.66)

Berlin et al., 1995

Patients with solvent-related disorders, ever exposed

3

0.7 (0.1–2.0)

Cohort Studies—Mortality

Boice et al., 1999

Aircraft-manufacturing workers, ever exposed

 

 

Mixed solvents

10

0.87 (0.42–1.60)

Morgan et al., 1981

Male paint and coatings manufacturing workers, ever employed

 

 

Melanoma and nonmelanoma

4

1.48

Case-Control Study

Fritschi and Siemiatycki, 1996a

Male residents of Montreal, Canada, ever exposed

 

Solvents

33

0.8 (0.5–1.3)

TABLE 6.19 Selected Epidemiologic Studies—Nonmelanoma Skin Cancers and Exposure to Organic Solvents

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Trichloroethylene

Cohort Study—Incidence

Hansen et al., 2001

Biologically monitored male Danish workers

15

1.0 (0.6–1.6)

Cohort Studies—Mortality

Ritz, 1999

White male uranium-processing workers, ever exposed

 

 

Melanoma and nonmelanoma

4

0.64 (0.17–1.63)

Axelson et al., 1994

Biologically monitored male Swedish workers

 

 

Nonmelanoma

8

2.36 (1.02–4.65)

Tetrachloroethylene and Dry-cleaning Solvents

Cohort Study—Mortality

Blair et al., 1990

Dry-cleaning union members in Missouri, ever employed

 

 

Melanoma and nonmelanoma

2

0.8 (0.1–2.8)

Other Specific Organic Solvents

Cohort Study—Mortality

Dosemeci et al., 1991

Male US industrial workers, ever exposed to phenol

 

 

Melanoma and nonmelanoma

7

0.9 (0.4–1.8)

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Unspecified Mixtures of Organic Solvents

Cohort Studies—Incidence

Anttila et al., 1995

Workers in Finland biologically monitored for halogenated hydrocarbons

2

0.46 (0.06–1.67)

Berlin et al., 1995

Patients with solvent-related disorders, ever exposed

4

1.5 (0.4–4.0)

Bourguet et al., 1987

Male tire and rubber manufacturing workers, ever exposed

 

 

Low solvent exposure

15

0.6

 

Medium solvent exposure

7

1.1

 

High solvent exposure

34

1.1

Cohort Study—Mortality

Morgan et al., 1981

Male paint and coatings manufacturing workers, ever employed

 

 

Melanoma and nonmelanoma

4

1.48

BREAST CANCER

Description of Case-Control Studies

All breast cancer case-control studies were population-based and are identified in Table 6.20 below. Aschengrau and colleagues (1998) evaluated the relationship between the risk of breast cancer and exposure to tetrachloroethylene in drinking water, which was estimated on the basis of an algorithm that accounted for residential history, water flow, and pipe characteristics, as established by Webler and Brown (1993). Three other case-control studies evaluated breast cancer risk and occupational exposure (Band et al., 2000; Hansen, 1999; Petralia et al., 1999). Each of the studies ascertained exposure differently: through use of occupational titles (Band et al., 2000), through linkage of pension-fund occupational-history information with solvent use in industries (Hansen, 1999), and through interviews with subjects to obtain occupational histories, which were linked with job-exposure matrixes to assign cumulative exposure measures (Petralia et al., 1999). Potential confounding variables were handled adequately in three of the four studies (Aschengrau et al., 1998; Band et al., 2000; Petralia et al., 1999).

Epidemiologic Studies of Exposure to Organic Solvents and Breast Cancer

Risk of breast cancer was not increased in two Scandinavian studies of biologically monitored workers exposed to trichloroethylene (Anttila et al., 1995; Hansen et al., 2001). Among women exposed to trichloroethylene as aircraft-maintenance workers, Blair and colleagues (1998) found a risk of breast cancer associated with any exposure to trichloroethylene (SMR=1.8, 95% CI=0.9–3.3) and more than a 3-fold risk associated with continuous low exposure to trichloroethylene (SMR=3.4, 95% CI=1.4–8.0). Other cohort studies of aircraft workers in which trichloroethylene was considered a predominant solvent did not report increased mortality rates from breast cancer (male and female combined) (Boice et al., 1999; Morgan et al., 1998).

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

TABLE 6.20 Description of Case-Control Studies of Breast Cancer and Exposure to Organic Solvents

Reference

Description of Study Population

Number of Cases

Number of Controls

Relevant Exposure(s)

Determination of Exposure

Analysis

Adjustment for Confounding

Aschengrau et al., 1998

Female cases reported to the Massachusetts Cancer Registry, diagnosed in 1983–1986 among residents of five upper Cape Cod towns; living controls were selected from the records of HCFA and through RDD; deceased controls identified by the state Department of Vital Statistics and Research files

258

686

Tetrachloroethylene

Relative delivered dose estimated in model accounting for location and years of residence, water flow, pipe characteristics

Multiple logistic regression

Age at diagnosis, vital status, family and personal history of breast cancer or disease, age at first birth, occupational exposure to solvents

Response rates: 79% of cases, 76% of HCFA controls, 74% of RDD controls, 79% of next of kin of deceased controls

Petralia et al., 1999

Female cases, age 40 years or more, identified through major hospitals in two New York counties in 1986–1991, with histologic confirmation; controls randomly selected from lists of the NY State Department of Motor Vehicles, matched for age and county

301

316

Benzene

In-person interviews to assess lifetime occupational history; occupations and industries coded; assigned potential exposures to polycyclic aromatic hydrocarbons through use of a job-exposure matrix

Unconditional logistic regression

Age, years of education, age at first birth, age at menarche, history of benign breast disease, family breast cancer history, Quetelet index, months of lactation

Participation rates: 66% of cases, 62% of controls

Hansen, 1999

Female cases, identified through the Danish Cancer Registry, born in 1934–1969 with diagnosis in 1970–1989; controls randomly selected from the central population register, matched for year of birth and sex

7,802

7,802

Industries with extensive solvent use

Past employment determined through linkage to the national pension fund files; occupations from five industrial groupings (except administrative jobs) classified as exposed to solvents

Conditional logistic regression

Age, social class, age at first child, number of children

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Description of Study Population

Number of Cases

Number of Controls

Relevant Exposure(s)

Determination of Exposure

Analysis

Adjustment for Confounding

Band et al., 2000

Female cases, under 75 years old, identified through the British Columbia Cancer Registry as Canadian citizens residing in British Columbia and diagnosed in 1988–1989, with histologic confirmation; controls selected from voter list, matched for age

995

1020

Laundering and dry-cleaning (occupation and industry)

Mailed questionnaire to assess lifetime job history with occupational and industry coding

Conditional logistic regression

Ethnicity, age at menarche and menopause, smoking, marital status, education, alcohol consumption, other medical factors

Response rates: 74.7% of cases, 76.1% of controls

NOTE: HCFA=Health Care Financing Administration; RDD=random-digit dialing..

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Some epidemiologic studies have either specifically addressed exposure to tetrachloroethylene or examined the effect of dry-cleaning work in relation to breast cancer risk. Aschengrau and colleagues (1998) conducted a case-control study in the Cape Cod area where drinking water was contaminated with tetrachloroethylene. Increased risks of breast cancer were found with longer latency. Excluding exposures occurring within the 9 years before diagnosis, the adjusted relative risks of breast cancer increased with increasing exposure (exposure greater than the 90th percentile: OR=7.8, 95% CI=0.9–167.0). A study of aircraft-manufacturing workers showed a slight increase in mortality from breast cancer with potential routine exposure to tetrachloroethylene (SMR=1.16, 95% CI=0.32–2.97) (Boice et al., 1999). A cohort of dry-cleaning union members showed no increase in breast cancer mortality (SMR=0.91, 95% CI=0.55–1.40) or in workers exposed only to tetrachloroethylene (SMR=0.78, 95% CI=0.28–1.69) (Ruder et al., 2001).

Band and colleagues (2000) examined the potential risk of premenopausal and postmenopausal breast cancer in multiple occupations in British Columbia. Increased risk of breast cancer was observed in women reporting any or usual work in laundry and dry-cleaning. Exposure to tetrachloroethylene and other solvents was based on an occupational title of “laundering/dry-cleaning” for premenopausal women and dry-cleaning occupational and industry titles for postmenopausal women. Among numerous positive associations, postmenopausal women experienced an almost 5-fold risk if their usual occupation was laundry and dry-cleaning (OR=4.85, 95% CI=1.26–18.7). The other cohort studies of dry cleaners did not show positive associations between this employment and breast cancer risk in women (Blair et al., 1990).

The association between exposure to benzene and the risk of breast cancer was assessed in two cohort studies and one case-control study. Petralia and colleagues (1999) reported an increased adjusted risk of premenopausal breast cancer with exposure to benzene and that risk increased with probability and duration of exposure (duration at least 4 years: OR=3.38, 95% CI=1.25–9.17). The large cohort study of Chinese benzene-exposed workers did not show an increase in breast cancer mortality (RR=0.9, 95% CI=0.3–3.2) (Yin et al., 1996a). The Danish cohort study, in which exposure was assessed using pension-fund records of job history, showed an increased incidence of breast cancer in men (OR=2.2, 95% CI=1.4–3.6) (Hansen, 2000).

The cohort of aircraft-maintenance workers studied by Blair and colleagues experienced an increased risk of breast cancer with exposure to methylene chloride (RR=3.0, 95% CI=1.0–8.8). The study also produced increased relative risks of breast cancer with exposure to several specific solvents; positive associations with 1,1,1-trichloroethane, acetone, isopropyl alcohol, toluene, and methyl ethyl ketone were found.

Employees of a cellulose-fiber production plant with heavy methylene chloride use did not experience a rate of breast cancer higher than that in the local county population (SMR=0.54, 95% CI=0.11–1.57) (Lanes et al., 1993).

Several studies examined the potential relationship between breast cancer risk and exposure to mixtures of solvents (Anttila et al., 1995, 1998; Berlin et al., 1995; Blair et al., 1998; Cocco et al., 1998; Hansen, 1999, Shannon et al., 1988; Weiderpass et al., 1999). The studies characterized the exposure in general terms, such as “organic solvents” or “mixed solvents.” In two instances (Anttila et al., 1995, 1998), specific constituent solvents were mentioned, but separate analyses were not performed. Hansen (1999) found increased risks of breast cancer (predominantly premenopausal) associated with occupations and industries with heavy solvent use (over 10 years of employment: OR=1.31, 95% CI=1.01–1.75). In a cohort study of lamp manufacturers, Shannon and colleagues (1998) found a 2-fold risk of breast cancer in solvent-

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

exposed workers in coiling and wire drawing. An association with any exposure to solvents was found among aircraft-maintenance workers (SMR=1.6, 95% CI=0.9–2.8) (Blair et al., 1998). No associations were found in studies of workers exposed to aromatic or halogenated solvents (Anttila et al., 1995, 1998), in patients with solvent-related disorders (Berlin et al., 1995), and in aircraft-manufacturing workers (Garabrant et al., 1988).

Summary and Conclusion

In most occupational settings, multiple solvent exposures occurred, so exposures to specific solvents may be highly correlated. Because many studies used occupational titles as exposure surrogates, the ability to assess an association between specific solvents and breast cancer risk was compromised.

A number of studies assessed breast cancer risk and solvent exposure in general, and others provided exposure estimates for specific individual solvents, such as trichloroethylene, tetrachloroethylene, dry-cleaning solvents, benzene, and methylene chloride. The evidence was limited by nonspecific exposure assessments and a reliance on mortality from breast cancer. Nondifferential misclassification of exposure, poor control for confounding, and low statistical power due to small numbers were additional limitations. Table 6.21 identifies the key studies and relevant data points reviewed by the committee in drawing its conclusion. Unless indicated in the tables, the study populations include both men and women.

The committee concludes, from its assessment of the epidemiologic literature, that there is inadequate/insufficient evidence to determine whether an association exists between chronic exposure to solvents under review and breast cancer.

TABLE 6.21 Selected Epidemiologic Studies—Breast Cancer and Exposure to Organic Solvents

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Trichloroethylene

Cohort Study—Incidence

Hansen et al., 2001

Biologically monitored workers in Denmark

 

 

Females, ever exposed

4

0.9 (0.2–2.3)

Cohort Studies—Mortality

Boice et al., 1999

Aircraft-manufacturing workers in California

 

 

Potential routine exposure

7

1.31 (0.53–2.69)

Blair et al., 1998

Aircraft-maintenance workers in Utah (females)

 

 

Any exposure

20

1.8 (0.9–3.3)

 

Low, continuous

8

3.4 (1.4–8.0)

 

Frequent peaks

10

1.4 (0.7–3.2)

Morgan et al., 1998

Aerospace workers in Arizona

 

Ever exposed

16

0.75 (0.43–1.22)

 

Low exposure

11

1.03 (0.51–1.84)

 

High exposure

5

0.47 (0.15–1.11)

Anttila et al., 1995

Finnish workers biologically monitored for exposure to halogenated hydrocarbons

34

0.85 (0.59–1.18)

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Tetrachloroethylene and Dry-cleaning Solvents

Cohort Studies—Mortality

Ruder et al., 2001

US dry-cleaning workers

 

 

Dry-cleaning, employed >1 year

20

0.91 (0.55–1.40)

 

Tetrachloroethylene only

6

0.78 (0.28–1.69)

Boice et al., 1999

Aircraft-manufacturing workers in California

 

 

Potential routine exposure

4

1.16 (0.32–2.97)

Blair et al., 1990

Dry-cleaning union members in Missouri (females)

36

1.0 (0.7–1.4)

Case-Control Studies

Band et al., 2000

Female cases from the British Columbia Cancer Registry

 

 

Premenopausal

 

 

Laundering or dry-cleaning (occupation)

 

 

Usual exposure

1

 

Ever exposed

4

1.77 (0.41–7.72)

 

Postmenopausal

 

 

Laundry or dry-cleaning (occupation)

 

 

Usual exposure

8

4.85 (1.26–18.7)

 

Ever exposed

8

1.33 (0.55–3.19)

 

Laundering and dry-cleaning (industry)

 

 

Usual exposure

9

5.24 (1.41–19.5)

 

Ever exposed

12

1.42 (0.68–2.99)

 

Power laundries or dry cleaners (industry)

 

 

Usual exposure

9

2.00 (0.78–5.13)

 

Ever exposed

21

1.67 (0.89–3.13)

Aschengrau et al., 1998

Residents of upper Cape Cod, MA

 

Postmenopausal women

 

 

75th percentile (9-year latency)

NA

3.4 (0.7–19.1)

 

90th percentile (9-year latency)

NA

7.8 (0.9–167.0)

Benzene

Cohort Study—Incidence

Hansen, 2000

Male members of the national pension fund in Denmark

 

 

No lag time

19

2.2 (1.4–3.6)

 

>10 years lag

12

2.5 (1.3–4.5)

Cohort Study—Mortality

Yin et al., 1996a

Chinese factory workers (females), ever exposed

8

0.9 (0.3–3.2)

Case-Control Study

Petralia et al., 1999

Female residents of New York state

 

 

Any exposure

56

1.91 (1.18–3.08)

 

Duration <4 years

8

0.80 (0.30–2.16)

 

Duration ≥4 years

16

3.38 (1.25–9.17)

 

Low probability

8

1.22 (0.42–3.56)

 

Medium or high probability

16

2.14 (0.89–5.12)

 

Low intensity

16

2.38 (0.97–5.87)

 

Medium or high intensity

8

1.07 (0.37–3.07)

 

Low cumulative

13

1.43 (0.59–3.47)

 

Medium or high cumulative

11

2.21 (0.77–6.36)

 

10 to 19 year latency

5

1.23 (0.34–4.46)

 

≥20 year latency

16

2.09 (0.85–5.14)

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Methylene Chloride

Cohort Studies—Mortality

Blair et al., 1998

Aircraft-maintenance workers in Utah, ever exposed (females)

4

3.0 (1.0–8.8)

Lanes et al., 1993

Cellulose-fiber production-plant workers in South Carolina, employed >3 months

3

0.54 (0.11–1.57)

Other Specific Organic Solvents

Cohort Study—Mortality

Blair et al., 1998

Aircraft-maintenance workers in Utah, employed >1 year (females)

 

 

1,1,1-Trichloroethane

3

3.3 (1.0–11.2)

 

Acetone

7

1.9 (0.8–4.6)

 

Isopropyl alcohol

8

3.7 (1.6–8.4)

 

Methyl ethyl ketone

8

2.1 (0.9–4.7)

 

Stoddard solvent

15

1.2 (0.6–2.4)

 

Toluene

10

2.0 (0.9–4.2)

Unspecified Mixtures of Organic Solvents

Cohort Studies—Incidence

Anttila et al., 1998

Finnish workers biologically monitored for exposure to aromatic hydrocarbons

15

0.79 (0.44–1.30)

 

Latency 0–9 years

8

0.61 (0.26–1.21)

 

Latency 10+ years

7

1.18 (0.47–2.42)

Anttila et al., 1995

Finnish workers biologically monitored for exposure to halogenated hydrocarbons

34

0.85 (0.6–1.2)

 

Latency 0–9 years

12

0.84 (0.4–1.5)

 

Latency 10+ years

22

0.85 (0.5–1.3)

Berlin et al., 1995

Swedish patients with solvent-related disorders (females)

3

1.1 (0.2–3.2)

Garabrant et al., 1988

Aircraft-manufacturing workers in California, employed >4 years

16

0.91 (0.52–1.48)

Shannon et al., 1988

Lamp manufacturing workers, employed >6 months

 

Coiling or wire drawing

8

2.04 (0.88–4.02)

Cohort Study—Mortality

Blair et al., 1998

Aircraft-maintenance workers in Utah (females)

 

 

Organic solvents, ever exposed

28

1.6 (0.9–2.8)

Case-Control Studies

Hansen, 1999

Danish women employed in industries with heavy solvent use

 

 

Employed ≥1 year, no lag

743

1.27 (1.13–1.43)

 

Employed ≥1 year, ≥15 years lag

472

1.43 (1.24–1.67)

 

Employed >10 years, no lag

113

1.31 (1.01–1.75)

 

Employed >10 years, ≥15 years lag

97

1.97 (1.39–2.79)

Cocco et al. 1998

Male breast cancer cases from National Mortality Follow-Back Survey exposed to organic solvents

 

 

Low probability

26

0.8 (0.5–1.3)

 

Medium probability

15

0.9 (0.5–1.6)

 

High probability

1

0.5 (0.1–4.3)

 

Low intensity

18

0.8 (0.4–1.4)

 

Medium intensity

20

0.8 (0.5–1.5)

 

High intensity

4

0.7 (0.2–1.2)

NOTE: NA=not available.

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

FEMALE REPRODUCTIVE CANCERS

Epidemiologic Studies of Exposure to Organic Solvents and Cervical Cancer

Anttila and colleagues (1995) conducted a followup study of cancer incidence among Finnish workers who were biologically monitored for exposure to trichloroethylene and other halogenated hydrocarbons. An increased risk of cervical cancer was observed among women with any exposure to trichloroethylene (SIR=2.42, 95% CI=1.05–4.77), and the relative risks increased with increasing exposure (RRlow=1.86, 95% CI=0.38–5.45; RRhigh=4.35, 95% CI=1.41–10.1). A similar biomonitoring study in Denmark showed a 3.8-fold risk of cervical cancer (SIR=3.8, 95% CI=1.0–9.8) (Hansen et al., 2001), but no exposure-response pattern was reported.

A retrospective cohort study by Blair and colleagues (1998) of US civilians employed in aircraft maintenance showed increased mortality from cervical cancer (SMR=3.0, 95% CI=0.5–6.5); the relative risk was increased among women with high cumulative exposure (SMR=3.0, 95% CI=0.8–11.7). Two studies of aerospace and aircraft-manufacturing workers (Boice et al., 1999; Morgan et al., 1998) found no deaths from cervical cancer.

Only two studies specifically examined the association between exposure to tetrachloroethylene and cervical cancer. However, given that tetrachloroethylene is often used in dry-cleaning work, studies on laundry and dry-cleaning workers were also reviewed.

Boice and colleagues (1999) did not observe any cases of cervical cancer among women exposed to tetrachloroethylene. The biomonitoring study of tetrachloroethylene-exposed workers by Anttila and colleagues (1995) found an increased risk of cervical cancer (SIR=3.20, 95% CI =0.39–11.6).

Ruder and colleagues (2001) found an increased risk of cervical cancer (SMR=1.95, 95% CI=1.00–3.40) among members of a dry-cleaning union representing four areas in the United States, and Blair and colleagues (1990) estimated an SMR of 1.7 (95% CI=1.0–2.0) in a study of dry-cleaning union members in Missouri. Ruder and colleagues (2001) found that the risk of cervical cancer increased further among women who were exposed for more than 5 years (SMR=2.78 for less than 20 years of latency and 2.40 for 20 years or more of latency), but Blair and colleagues (1990) found no increases in risk with increasing exposure.

The available evidence concerning an association between exposure to methylene chloride and risk of cervical cancer was sparse. Gibbs and colleagues (1996) examined the risk of cervical cancer among women employed in cellulose-fiber production. They found an increased risk of cervical cancer among workers with low or high exposure (SMRlow=2.96, 95% CI=0.96–6.92; SMRhigh=5.40, 95% CI=0.14–30.1), but it was also increased in workers with no measured exposure to methylene chloride (SMR=7.02, 95% CI=0.18–39.1). Shannon and colleagues (1988) studied lamp manufacturing workers who were exposed to methylene chloride and other solvents and substances used during the manufacturing process, particularly coiling and wire drawing, and found no increased risk of cervical cancer.

The study of aircraft-manufacturing workers by Boice and colleagues (1999) did not show any cases of cervical cancer in women exposed to mixtures of solvents. In 1995, Berlin and colleagues examined cancer incidence and mortality patterns among patients with suspected solvent-related disorders and found that the incidence of cervical cancer was high (SIR=3.7, 95% CI=2.2–6.2).

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Summary and Conclusion

A limited number of papers were found that reported the risk of cervical cancer in connection with exposure to specific solvents. There were no case-control studies, and most of the cohort studies had very few cases. The studies of cervical cancer and specific solvent exposures did not provide evidence of an association between most of the specific solvents or solvent mixtures except for trichloroethylene.

For exposure to trichloroethylene and cervical cancer, three cohort studies showed an increased risk of cervical cancer with exposure to trichloroethylene, and two other studies did not have sufficient numbers or followup to find any deaths. An exposure-response relationship for the highest exposure was reported in two of the biologic monitoring studies (Anttila et al., 1995; Blair et al., 1998). Some committee members believed that the evidence of an association between cervical cancer and exposure to trichloroethylene should be classified as limited/suggestive. However, some committee members were concerned about confounding by socioeconomic status and the increased risk of exposure to the human papilloma virus (HPV), which is associated with the development of cervical cancer (NCI, 2002). The studies compared the risk of cervical cancer among unskilled workers of low socioeconomic status in Scandinavian countries with that of the general population. The positive associations could have been attributed to the lack of control for socioeconomic status or HPV infection. Moreover, no trends were seen with duration of employment or cumulative exposure in the other biologically monitored study of Danish workers (Hansen et al., 2001). There was also a concern that the numbers of studies and exposed cases were too small to support a conclusion that the evidence was limited/suggestive. In addition, three studies (Blair et al., 1998; Boice et al., 1999; Morgan et al., 1998) had no exposed cases, although cases of cervical cancer were expected. Thus, some committee members concluded that the evidence was inadequate/insufficient to determine whether an association exists.

As a result, after extensive discussion, the committee could not reach a consensus as to whether the evidence was limited/suggestive of an association or was inadequate/insufficient to determine whether an association exists between cervical cancer and exposure to trichloroethylene. Future studies that control for socioeconomic status are needed to determine whether there is an association between exposure to trichloroethylene and the risk of cervical cancer.

For exposure to tetrachloroethylene and dry-cleaning solvents, although several studies were positive, exposure-response patterns were absent and limited the strength of the evidence. Only one study on the exposure to unspecified mixtures of organic solvents and risk of cervical cancer was identified. Although the finding was particularly strong, no exposure-relationship pattern was reported. Additional corroborating studies are needed before a determination can be made that an association exists between exposure to solvent mixtures and cervical cancer. Table 6.22 identifies the studies reviewed by the committee in making its conclusion regarding association.

The committee concludes, from its assessment of the epidemiologic literature, that there is inadequate/insufficient evidence to determine whether an association exists between chronic exposure to solvents under review, other than trichloroethylene, and cervical cancer.

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

TABLE 6.22 Selected Epidemiologic Studies—Cervical Cancer and Exposure to Organic Solvents

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Trichloroethylene

Cohort Studies—Incidence

Hansen et al., 2001

Biologically monitored workers in Denmark

4

3.8 (1.0–9.8)

Anttila et al., 1995

Biologically monitored workers in Finland

 

 

Whole period of followup (mean individual urinew level)

8

2.42 (1.05–4.77)

 

<100 µmol/L

3

1.86 (0.38–5.45)

 

100+ µmol/L

5

4.35 (1.41–10.1)

Cohort Studies—Mortality

Boice et al., 1999

Aircraft-manufacturing workers in California, potential routine exposure

0

Blair et al., 1998

Aircraft-maintenance workers in Utah, ever exposed

5

1.8 (0.5–6.5)

 

<5 unit-years

1

0.9 (0.1–8.3)

 

5–25 unit-years

0

 

>25 unit-years

4

3.0 (0.8–11.7)

Morgan et al., 1998

Aerospace workers in Arizona, employed >6 months

0

Tetrachloroethylene and Dry-cleaning Solvents

Cohort Study—Incidence

Anttila et al., 1995

Biologically monitored workers in Finland

2

3.20 (0.39–11.6)

Cohort Studies—Mortality

Ruder et al., 2001

US dry-cleaning workers

 

 

Employed >1 year

12

1.95 (1.00–3.40)

 

Employed 5+ years, <20 years of latency

4

2.78 (0.75–7.71)

 

Employed 5+ years, ≥20 years of latency

3

2.40 (0.48–7.86)

Boice et al., 1999

Aircraft-manufacturing workers in California, potential routine exposure

0

Blair et al., 1990

Dry-cleaning union members in Missouri

21

1.7 (1.0–2.0)

Methylene Chloride

Cohort Studies—Mortality

Gibbs et al., 1996

Cellulose triacetate-fiber workers, employed >3 months

 

 

Methylene chloride, no exposure

1

7.02 (0.18–39.1)

 

Methylene chloride, low probability

5

2.96 (0.96–6.92)

 

Methylene chloride, high probability

1

5.40 (0.14–30.1)

Shannon et al., 1988

Lamp-manufacturing workers (primary exposure to methylene chloride), employed >6 months

1

1.05 (0.03–5.86)

Unspecified Mixtures of Organic Solvents

Cohort Studies—Mortality

Boice et al., 1999

Aircraft-manufacturing workers in California

 

 

Mixed solvents, potential routine exposure

0

Berlin et al., 1995

Swedish patients with solvent-related disorders

14

3.7 (2.2–6.2)

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Epidemiologic Studies of Exposure to Organic Solvents and Ovarian and Uterine Cancer

There was a paucity of studies regarding exposure to specific solvents and ovarian or uterine cancer. No studies showed meaningful increases in the risk of uterine or ovarian cancer in relation to exposure to trichloroethylene. The number of exposed subjects was extremely small in these studies, so the data are not informative in drawing a conclusion regarding association. There were no reports of an association between risk of uterine or ovarian cancer and exposure to methylene chloride or of an association between ovarian cancer and exposure to unspecified mixtures of organic solvents (Boice et al., 1999).

Summary and Conclusion

A limited body of evidence was available for the committee to review concerning specific and unspecified solvent exposure and risk of ovarian or uterine cancer. Very few studies had sufficient power to permit meaningful analyses. Tables 6.23 and 6.24 identify the key studies reviewed by the committee in drawing its conclusions regarding various solvent exposures and ovarian or uterine cancer.

The committee concludes, from its assessment of the epidemiologic literature, that there is inadequate/insufficient evidence to determine whether an association exists between chronic exposure to solvents under review and ovarian or uterine cancer.

TABLE 6.23 Selected Epidemiologic Studies—Ovarian Cancer and Exposure to Organic Solvents

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Trichloroethylene

Cohort Study—Incidence

Hansen et al., 2001

Biologically monitored workers in Denmark

2

2.1 (0.2–7.6)

Cohort Studies—Mortality

Boice et al., 1999

Aircraft-manufacturing workers in California, potential routine exposure

1

0.58 (0.01–3.22)

Morgan et al., 1988

Aerospace workers in Arizona, employed >6 months

 

 

Ever exposed

8

1.21 (0.52–2.38)

 

Low exposure

2

0.61 (0.07–2.21)

 

High exposure

6

1.79 (0.66–3.88)

Tetrachloroethylene and Dry-cleaning Solvents

Cohort Study—Mortality

Boice et al., 1999

Aircraft-manufacturing workers in California, potential routine exposure

0

Methylene Chloride

Cohort Study—Mortality

Shannon et al., 1988

Lamp-manufacturing workers (primary exposure to methylene chloride), employed >6 months

1

1.47 (0.04–8.19)

Unspecified Mixtures of Organic Solvents

Cohort Study—Mortality

Boice et al., 1999

Aircraft-manufacturing workers in California

 

 

Mixed solvents, potential routine exposure

2

0.57 (0.07–2.07)

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

TABLE 6.24 Selected Epidemiologic Studies—Uterine and Endometrial Cancer and Exposure to Organic Solvents

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Trichloroethylene—Uterine

Cohort Study—Incidence

Hansen et al., 2001

Biologically monitored workers in Denmark

1

1.0 (0.01–5.4)

Cohort Studies—Mortality

Boice et al., 1999

Aircraft-manufacturing workers in California, potential routine exposure

1

0.64 (0.02–3.57)

Morgan et al., 1988

Aerospace workers in Arizona, ever exposed

1

0.16 (0.00–0.91)

Tetrachloroethylene and Dry-cleaning Solvents—Uterine

Cohort Studies—Mortality

Boice et al., 1999

Aircraft-manufacturing workers in California, potential routine exposure

0

Blair et al., 1990

Dry-cleaning union members in Missouri

8

1.0 (0.4–2.0)

Methylene Chloride—Uterine/Endometrial

Cohort Studies—Mortality

Shannon et al., 1988

Lamp-manufacturing workers (primary exposure to methylene chloride), employed >6 months

2

2.14 (0.26–7.60)

Unspecified Mixtures of Organic Solvents—Uterine

Cohort Study—Mortality

Boice et al., 1999

Aircraft-manufacturing workers in California

 

 

Mixed solvents, potential routine exposure

1

0.31 (0.01–1.71)

UROLOGIC CANCERS

Description of Case-Control Studies

The characteristics of the case-control studies considered by the committee in drawing its conclusions of association are described below for each cancer site. The principal strengths and limitations of the studies are discussed below by cancer site.

Epidemiologic Studies of Exposure to Organic Solvents and Prostate Cancer

Gérin and co-workers (1998) evaluated 15 cancer risks, including the risk of prostate cancer, related to such occupational exposures as the hydrocarbons benzene, toluene, xylene, and styrene. The study had excellent information on exposures, as assessed by in-depth interviews that were coded blindly by a team of chemists and industrial hygienists, and sufficient information on most risk factors (see Table 6.25).

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

TABLE 6.25 Description of Case-Control Study of Prostate Cancer and Exposure to Organic Solvents

Reference

Description of Study Population

Number of Cases

Number of Controls

Relevant Exposure(s)

Determination of Exposure

Analysis

Adjustment for Confounding

Gérin et al., 1998

Male cases, age 35–75 years, diagnosed in one of 19 large Montreal-area hospitals in 1979–1985 and histologically confirmed; controls identified concurrently at 18 other cancer sites; age-matched, population-based controls were also chosen from electoral lists and random-digit dialing

449

1066, consisting of 533 population controls and 533 randomly selected subjects from other cases of cancer

Benzene

Toluene

Xylene

In-person interviews (direct or proxy) with segments on work histories (job titles and self-reported exposures); analyzed and coded by a team of chemists and industrial hygienists (about 300 exposures on semiquantitative scales)

Logistic regression

Age, family income, ethnicity, cigarette smoking, respondent status

Response rates: 82% of all cases, 71% of population controls

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Prostate cancer mortality and incidence were not associated with exposure to trichloroethylene in the cohort studies conducted in the aircraft industry (Blair et al., 1998: SMR=1.0, 95% CI=0.5–2.1 for 5–25 unit-years of exposure; Boice et al., 1999: SMR=1.03, 95% CI=0.70–1.45). Morgan and colleagues (1998) found an excess risk of 18% (SMR=1.18, 95% CI=0.73–1.80), and the relative risks did not increase with increasing exposure to trichloroethylene. Other studies of occupationally exposed workers also found no association with exposure to trichloroethylene, including studies by Greenland and colleagues (1994) (OR=0.82, 95% CI=0.46–1.46), Hansen and colleagues (2001) (SIR=0.6, 95% CI=0.2–1.3), and Wilcosky and colleagues (1984) (OR=0.62 [CI not provided by the authors, and the committee was unable to calculate it]). A cohort study of US uranium-processing workers found no increase in prostate cancer when the subjects were exposed to “light” amounts of trichloroethylene with increasing years of exposure latency (SMR ranged from 0.78–1.04). After “moderate” exposure, risk increased; the SMR ranged from 1.35 (95% CI=0.17–10.4) after more than 2 years of exposure and no latency to 1.96 (95% CI=0.25–15.6) after more than 5 years of exposure and a 15-year latency (Ritz, 1999).

Axelson and co-workers (1994) reported an increased risk of prostate cancer among Swedish men occupationally exposed to trichloroethylene (RR=1.25, 95% CI=0.84–1.84). Anttila and colleagues (1995) reported an increased risk of prostate cancer after exposure to trichloroethylene (RR=1.38, 95% CI=0.73–2.35).

Two cohort studies of dry-cleaning workers did not show a positive association between tetrachloroethylene and risk of prostate cancer (Blair et al., 1990: SMR=0.7, 95% CI 0.2–1.7; Ruder et al., 1994: SMR=0.82, 95% CI 0.33–1.69).

A German study of rotogravure printers showed no association between prostate cancer and exposure to toluene (SMR=0.67, 95% CI=0.13–2.66) (Wiebelt and Becker, 1999). No increased risk of prostate cancer was found in workers in Montreal who reported exposure to toluene (ORhigh=0.4, 95% CI=0.1–1.4) (Gérin et al., 1998). A study of white, male rubber workers in Ohio found an association between exposure to toluene and prostate cancer (OR=2.6) (Wilcosky et al., 1984).

Both the Montreal study and the rubber-workers study evaluated prostate cancer risk and exposure to xylene: Gérin and co-workers (1998) reported an imprecise estimate of effect of “high” exposure to xylene (OR=1.4, 95% CI=0.5–4.0), and Wilcosky and colleagues (1984) reported an OR of 1.5.

Gérin and colleagues also examined exposure to benzene and risk of prostate cancer. An association was found with “medium” exposure to benzene (OR=1.7, 95% CI=0.9–3.0) but not “high” exposure (OR=0.9, 95% CI=0.4–2.1).

Only one cohort study of US cellulose-fiber production workers reported exposure to methylene chloride. An increased risk of prostate cancer was found (SMRhigh=1.79, 95% CI=0.95–3.06; SMRlow=1.40, 95% CI=0.64–2.66; SMRno exposure=1.04, 95% CI=0.22–3.05) (Gibbs et al., 1996). The magnitude of the risk increased in the high exposure group, with 20 years or more of latency (SMR=2.08; p<0.05) and with 20 years or more of exposure with 20 years of latency (SMR=2.91; p<0.05).

No increased risk of prostate cancer with exposure to unspecified mixtures of solvents was apparent in the studies the committee reviewed except one (Anttila et al., 1995). Studies that showed no increase include cohort studies by Boice and colleagues (1999) (SMR=1.0, 95% CI 0.78–1.26), Garabrant and colleagues (1988) (SMR=0.93,

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

95% CI=0.60–1.37), Matanoski and colleagues (1986) (SMR=0.99, 95% CI=0.82–1.18), Morgan and colleagues (1981) (SMR=0.84 [CI not provided by the study authors, and the committee was unable to calculate it]), and Greenland and colleagues (1994) (OR=0.84, 95% CI=0.49–1.42). The cohort study by Anttila and colleagues (1995) of Finnish workers monitored for exposure to halogenated hydrocarbons showed an SIR of 1.38 (95% CI=0.76–2.32).

Summary and Conclusion

Results of several large cohort studies of trichloroethylene exposed workers did not support an association between exposure and risk of prostate cancer, nor did the cohort studies of dry-cleaning workers. Although one positive study was identified for exposure to toluene, xylene, and benzene individually, other studies did not find an association, or the studies lacked any evidence of an exposure-response relationship. For exposure to methylene chloride, one study provided evidence for increased risk of prostate cancer with increasing years of exposure and latency, but other corroborating studies were not found. All but one of the studies on solvent mixtures found a positive association, therefore the committee was not able to determine whether an association exists between exposure and prostate cancer. Table 6.26 provides the key data points for each exposure reviewed by the committee in drawing its conclusion.

The committee concludes, from its assessment of the epidemiologic literature, that there is inadequate/insufficient evidence to determine whether an association exists between chronic exposure to solvents under review and prostate cancer.

TABLE 6.26 Selected Epidemiologic Studies—Prostate Cancer and Exposure to Organic Solvents

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Trichloroethylene

Cohort Studies—Incidence

Hansen et al., 2001

Biologically monitored Danish workers, ever exposed

6

0.6 (0.2–1.3)

Blair et al., 1998

Aircraft-maintenance workers in Utah

 

 

No exposure

61

1.0 (0.7–1.4)

 

<5 unit-years

64

1.1 (0.7–1.6)

 

5–25 unit-years

38

1.0 (0.6–1.6)

Anttila et al., 1995

Finnish workers monitored for exposure

 

 

Entire period since first measurement

13

1.38 (0.73–2.35)

 

0–9 years

2

1.09 (0.13–3.91)

 

10–19 years

3

0.56 (0.12–1.64)

 

20+ years

8

3.57 (1.54–7.02)

Axelson et al., 1994

Swedish men occupationally exposed, trichloroethylene

26

1.25 (0.84–1.84)

Cohort Studies—Mortality

Ritz, 1999

White male US uranium-processing workers

 

 

Duration of exposure, exposure lag

 

 

Trichloroethylene, light exposure

 

 

>2 years, no lag

10

0.78 (0.33–1.85)

 

>2 years, 15-year lag

10

0.91 (0.38–2.18)

 

>5 years, no lag

8

0.83 (0.33–2.09)

 

>5 years, 15-year lag

8

1.04 (0.40–2.70)

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

 

Trichloroethylene, moderate exposure

 

 

>2 years, no lag

1

1.35 (0.17–10.4)

 

>2 years, 15-year lag

1

1.44 (0.19–11.4)

 

>5 years, no lag

1

1.58 (0.20–12.5)

 

>5 years, 15-year lag

1

1.96 (0.25–15.6)

Boice et al., 1999

Aircraft-manufacturing workers in California, routine exposure

32

1.03 (0.70–1.45)

Blair et al., 1998

Aircraft-maintenance workers in Utah

 

 

No trichloroethylene exposure

33

1.2 (0.7–2.1)

 

<5 unit-years

19

0.9 (0.5–1.8)

 

5–25 unit-years

13

1.0 (0.5–2.1)

Morgan et al., 1998

Aerospace workers in Arizona

 

 

Any exposure

21

1.18 (0.73–1.80)

 

Low exposure

7

1.29 (0.52–2.66)

 

High exposure

14

1.13 (0.62–1.89)

Greenland et al., 1994

White male US transformer manufacturers, ever exposed

NA

0.82 (0.46–1.46)

Wilcosky et al., 1984

White male rubber workers in Ohio, exposed >1 year

3

0.62

Tetrachloroethylene and Dry-Cleaning Solvents

Cohort Studies—Mortality

Ruder et al., 1994

Dry-cleaning labor-union workers

7

0.82 (0.33–1.69)

Blair et al., 1990

Members of a dry-cleaning union in St. Louis, MO

5

0.7 (0.2–1.7)

Toluene

Cohort Studies—Mortality

Wiebelt and Becker, 1999

Male German rotogravure printers, Employed >1 year

2

0.67 (0.13–2.66)

Wilcosky et al., 1984

White male rubber workers in Ohio, exposed >1 year

3

2.6

Case-Control Study

Gérin et al., 1998

Male residents of Montreal, Canada

 

 

Low exposure

51

1.0 (0.7–1.5)

 

Medium exposure

17

1.3 (0.7–2.5)

 

High exposure

3

0.4 (0.1–1.4)

Xylene

Cohort Study—Mortality

Wilcosky et al., 1984

White male rubber workers in Ohio, exposed >1 year

8

1.5

Case-Control Study

Gérin et al., 1998

Male residents of Montreal, Canada

 

 

Low exposure

46

1.2 (0.8–1.8)

 

Medium exposure

11

0.8 (0.4–1.7)

 

High exposure

6

1.4 (0.5–4.0)

Benzene

Cohort Study—Mortality

Wilcosky et al., 1984

White male rubber workers in Ohio, exposed >1 year

11

0.73

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Study Population

Exposed Cases

Estimated Relative Risk (95% CI)

Case-Control Study

Gérin et al., 1998

Male residents of Montreal, Canada

 

 

Low exposure

64

1.1 (0.8–1.5)

 

Medium exposure

22

1.7 (0.9–3.0)

 

High exposure

9

0.9 (0.4–2.1)

Methylene Chloride

Cohort Study—Mortality

Gibbs et al., 1996

Cellulose-fiber production workers

 

 

High exposure

13

1.79 (0.95–3.06)

 

≥20-year latency

NA

2.08 (p<0.05)

 

≥20-year latency and ≥20-year duration

NA

2.91 (p≤0.05)

 

Low exposure

9

1.40 (0.64–2.66)

 

No exposure

3

1.04 (0.22–3.05)

Unspecified Mixtures of Organic Solvents

Cohort Study—Incidence

Anttila et al., 1995

Finnish workers monitored for exposure

 

 

Halogenated hydrocarbons

14

1.38 (0.76–2.32)

Cohort Studies—Mortality

Boice et al., 1999

Aircraft-manufacturing workers in California

 

 

Mixed solvents, routine exposure

70

1.0 (0.78–1.26)

 

Years exposed

 

 

<1

31

0.99 (0.65–1.49)

 

1–4

64

0.81 (0.59–1.13)

 

≥5

139

0.77 (0.58–1.02)

 

 

 

p-trend=0.06

Greenland et al., 1994

White male US transformer-assembly workers

 

Solvents, ever exposed

NA

0.84 (0.49–1.42)

Garabrant et al., 1988

Aircraft-manufacturing workers in California

25

0.93 (0.60–1.37)

Matanoski et al., 1986

US painters and allied tradesmen union members

117

0.99 (0.82–1.18)

Morgan et al., 1981

Male US paint and coatings manufacturers, employed >1 year

29

0.84

NOTE: NA=not available.

Epidemiologic Studies of Exposure to Organic Solvents and Bladder Cancer

All but one (Aschengrau et al., 1993) of the case-control studies of bladder cancer reviewed by the committee used occupational history to assess exposure, and in some studies information on specific chemical exposures was also obtained (Gérin et al., 1998; Pesch et al., 2000a). The study by Aschengrau and colleagues assessed exposure on the basis of estimated doses of tetrachloroethylene found in public drinking water in five towns of Cape Cod, Massachusetts. Two studies included interviews with proxies if subjects were too ill to be interviewed (Morrison et al., 1985; Teschke et al., 1997). Most studies simply grouped exposure defined broadly on the basis of occupation, including work in the painting industry (Cordier et al., 1993; Jensen et al., 1987; La Vecchia et al., 1990; Morrison et al., 1985; Vineis and Magnani, 1985), in laundry and dry-cleaning services (Silverman et al., 1989a,b; Smith et al., 1985), and in both fields (Schoenberg et al., 1984; Teschke et al.,

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

1997). Very low response rates were found in the study by Risch and colleagues (1988) of exposure to paints. A number of case-control studies used self-reported information on exposures that were not otherwise validated (Jensen et al., 1987; La Vecchia et al., 1990; Risch et al., 1988; Schoenberg et al., 1984; Smith et al., 1985; Vineis and Magnani, 1985) and thus may be subject to recall bias if the controls were not ill.

The main accepted risk factor for bladder cancer is cigarette smoking, and all but one (Smith et al., 1985) of the case-control studies of bladder cancer and exposure to specific and unspecified organic solvents reviewed by the committee accounted for smoking in some way. In most studies, subjects were asked whether they currently smoked; others inquired about the number of cigarettes smoked per day (Aschengrau et al., 1993), lifetime cigarette consumption (Risch et al., 1988), or duration of cigarette smoking (Schoenberg et al., 1984).

Case-control studies of bladder cancer and exposure to organic solvents that had reasonably good assessments of exposure, adequate control for confounding, and histologic confirmation of outcome include those by: Aschengrau et al., 1993; Cordier et al., 1993; Gérin et al., 1998; Pesch et al., 2000a; and Silverman et al., 1989a,b (see Table 6.27).

In addition to the case-control study by Pesch and colleagues (2000a) described above, there were several cohort studies of biologically monitored workers, transformer manufacturers, and aircraft and aerospace workers. One study of US aircraft-maintenance workers showed an RRany exposure of 1.2 (95% CI=0.5–2.9) for bladder cancer (Blair et al., 1998); another showed no association (RR=0.55, 95% CI=0.18–1.28) (Boice et al., 1999). An increased risk of bladder cancer also was observed in a cohort of US aerospace-manufacturing workers (SMR=1.36, 95% CI=0.59–2.68) (Morgan et al., 1998). Three other cohort studies of workers biologically monitored for exposure to trichloroethylene, as shown by the presence of a urinary metabolite, showed no association (relative risks ranged from 0.61 to 1.1) between this marker and bladder cancer (Anttila et al., 1995; Axelson et al., 1994; Hansen et al., 2001). A nested case-control study within a cohort of transformer-assembly facility workers did not find an association between risk of bladder cancer and exposure to trichloroethylene (OR=0.85, 95% CI=0.32–3.32) (Greenland et al., 1994).

The case-control study by Pesch and colleagues (2000a) suggested a positive association with urothelial carcinoma (a cancer of the urinary tract that affects mostly the bladder). Using a job-exposure matrix, the authors found a 10% excess risk (OR=1.1; 95% CI=0.9–1.4) among men with “high” exposure and a 60% excess risk among women with “high” exposure (OR=1.6, 95% CI=1.0–2.5). On the basis of the job task-exposure matrix, the magnitude of the OR for men with “high” exposure was increased to 1.3 (95% CI=0.9–1.7). The job task-exposure matrix was not used to evaluate exposure of women.

The committee used two cohort studies and several case-control studies of dry-cleaning workers and other workers exposed to tetrachloroethylene in determining whether there is an association between exposure and bladder cancer. The two dry-cleaner cohort studies showed increased mortality from bladder cancer: Blair and colleagues (1990) reported an increased SMR of 1.7 (95% CI=0.7–3.3), and Ruder and colleagues (2001) found a 122% increase in mortality from bladder cancer (SMR=2.22, 95% CI=1.06–4.08). The magnitude of the association was higher (SMR=4.31, 95% CI=1.85–8.76) among workers employed more than 5 years and in whom 20 years had passed since the first exposure.

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

TABLE 6.27 Description of Case-Control Studies of Bladder Cancer and Exposure to Organic Solvents

Reference

Description of Study Population

Number of Cases

Number of Controls

Relevant Exposure(s)

Determination of Exposure

Analysis

Adjustment for Confounding

Schoenberg et al., 1984

Male cases, age 21–84 years, with histologically confirmed diagnosis in New Jersey in February 1978–1979; controls identified through RDD (age 21–64 years) and HCFA records (age 65–85 years), stratified for age

658

1258

Painting or artistic work

Dry-cleaning work

In-person interviews with questionnaires assessing lifetime occupational history (job titles)

Logistic regression

Age, duration of cigarette smoking, other occupations

Response rates: 89.7% of cases, 86.6% of controls

Morrison et al., 1985

Cases, age 21–89 years, identified at hospitals in Boston; Greater Manchester County, UK; and Nagoya, Japan, in 1976–1978; controls selected from respective areas’ electoral registers, matched for age and sex

430 Boston

399 UK

226 Japan

397 Boston

493 UK

443 Japan

Painting work

In-person interview (direct or proxy) assessing occupational history (job titles); job titles were coded

Logistic regression

Age, cigarette smoking

No response rates provided

Smith et al., 1985

Cases and controls, age 21–84 years, residing in the nine SEER population-based areas and New Jersey, who participated in the NCI National Bladder Cancer Study; cases with histologically confirmed cancer; controls frequency matched for age and sex

7748 total participants with and without bladder cancer, classified as: worked in dry-cleaning operations (N=103), experienced related exposures (N=5776), or neither (unexposed; N=1869)

Laundry and dry-cleaning work

In-person interview with structured questionnaire regarding occupational history (job or industry titles)

Logistic regression

Age, sex

No response rates provided

Vineis and Magnani, 1985

Cases, age less than 70 years, identified from the Main Hospital in Torino, Italy, in 1978–1983; controls from same hospital diagnosed with benign urologic conditions, matched for age

512

596

Painting work

In-person interviews regarding lifetime occupational history (job or industry titles) obtained in hospital and blindly coded

Mantel-Haenszel

Age, smoking

No response rates provided

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Description of Study Population

Number of Cases

Number of Controls

Relevant Exposure(s)

Determination of Exposure

Analysis

Adjustment for Confounding

Jensen et al., 1987

Cases reported to the Danish Cancer Registry from Copenhagen area in 1979–1981, with 99% histologic verification; controls selected from the residents of the municipalities from which cases arose, matched for sex and age

371

771

Painting work

In-person interviews with questionnaire assessing occupational history (job titles and self-reported exposures)

Logistic regression

Age, sex, smoking

Response rates: 94.4% of cases, 75.1% of controls

Risch et al., 1988

Cases, age 35–79 years, identified through a combination of cancer registry reporting and hospital record review in four cities in Canada in 1979–1982 with histologic confirmation; controls selected randomly from population listings and matched on birth year, sex, and residence area

826

792

Organic solvents

Paints

In-person interview with questionnaire assessing specific occupational exposures (self-reports)

Conditional logistic regression

Matching variables, lifetime cigarette consumption

Response rates: 67% of cases, 53% of controls

Silverman et al., 1989a

White cases, age 21–84 years, in 10 US areas in 1977–1978 with histologic confirmation; controls identified through RDD (age 21–64 years) and HCFA records (age 65–84 years), matched for age and geographic area

2100

3874

Dry-cleaning, ironing, and pressing work

Questionnaire administered by in-person interview (job or industry titles); industries and job titles coded by study authors and grouped by potential exposures

Logistic regression

Smoking, age

No response rates provided

Silverman et al., 1989b

Nonwhite cases, same study as above (Silverman et al., 1989a)

126

383

See above

See above

See above

See above

La Vecchia et al., 1990

Cases, age less than 75 years, admitted to NCI or clinics and hospitals in Milan, Italy, in 1985–1988 with histologic confirmation; controls admitted to the same network of hospitals for acute nonneoplastic conditions

263

287

Painting work

Chemical-industry work

Structured questionnaire to assess lifetime employment in 19 industries or occupations and 14 specific agents (job or industry titles and self-reported exposures)

Mantel-Haenszel

Age, sex, smoking

Response rate: >97% of cases and controls

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×

Reference

Description of Study Population

Number of Cases

Number of Controls

Relevant Exposure(s)

Determination of Exposure

Analysis

Adjustment for Confounding

Aschengrau et al., 1993

Cases reported to the Massachusetts Cancer Registry, diagnosed in 1983–1986 among residents of five upper Cape Cod towns; living controls were selected from HCFA records and through RDD; deceased controls identified by the state Department of Vital Statistics and Research files

61

852

Tetrachlorid e-ethylene

Exposure dose estimated in areas of contaminated drinking water, accounting for location and years of residence, water flow, pipe characteristics

Logistic regression

Sex, age at diagnosis, vital status, educational level, usual number of cigarettes smoked, occupational exposure to solvents, specific cancer risk factors controlled for in respective analyses

Response rates: 80.6% of cases, 75.9% of HCFA controls, 73.9% of RDD controls, 78.8% of next of kin of deceased controls

Cordier et al., 1993

Cases, under age 80 years, from seven French hospitals in 1984–1987, with histologic confirmation; controls selected from the same hospitals from patients admitted for causes other than cancer, respiratory disease, or symptoms related to bladder cancer, matched for sex, age, ethnicity, and residence

765

765

Solvents

Painting work

In-person interviews with segments on work histories (job titles); analyzed and coded by a team of experts in industrial hygiene

Logistic regression

Smoking status, hospital, age, place of residence

No response rates provided

Teschke et al., 1997

Cases, age 19 years and over, registered with the British Columbia Cancer Agency in 1990–1991 with histologic confirmation; controls selected from the provincial voters list and matched on age and sex

105

139

Laundry personnel

Painters

Occupational histories (job titles) and self-reported exposures obtained (direct or proxy) through standardized questionnaire (in-person or telephone interview)

Adjusted ORs

Sex, age, cigarette smoking

Response rates: 88.2% of cases, 80.3% of controls

Suggested Citation:"6. Cancer and Exposure to Solvents." Institute of Medicine. 2003. Gulf War and Health: Volume 2: Insecticides and Solvents. Washington, DC: The National Academies Press. doi: 10.17226/10628.
×