Buta-1,3-diene

Epidemiological data

Administrative data

Endpoint:

epidemiological data

Type of information:

other: Observational

Adequacy of study:

key study

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

other: Human workplace investigation, published in peer reviewed literature, minor restrictions in design and/or reporting but otherwise adequate for assessment

Cross-reference

Cross-reference 1

Reason / purpose for cross-reference:

reference to same study

Cross-reference 2

Reason / purpose for cross-reference:

reference to other study

Data source

Reference

Reference 1

Reference Type:

publication

Title:

An updated study of mortality among North American synthetic rubber industry workers

Author:

Sathiakumar N, Graff J, Macaluso M, Maldonado G, Matthews R, Delzell E

Year:

2005

Bibliographic source:

Occup Environ Med; 62, 822-829.

Reference 2

Reference Type:

publication

Title:

Chemical exposures in the synthetic rubber industry and lymphohematopoietic cancer mortality

Author:

Graff JJ, Sathiakumar N, Macaluso M, Maldonado G, Matthews R, Delzell E

Year:

2005

Bibliographic source:

Journal of Occupational and Environmental Medicine 47, 916-932

Reference 3

Reference Type:

publication

Title:

An Updated Study of Mortality Among North American Synthetic Rubber Industry Workers.

Author:

Delzell E, Sathiakumar N, Graff J, Macaluso M, Maldonado G, and Matthew R.

Year:

2006

Bibliographic source:

Health Effects Institute Research Report 132

Materials and methods

Study type:

cohort study (retrospective)

Endpoint addressed:

carcinogenicity

Test guideline

Qualifier:

no guideline followed

Principles of method if other than guideline:

These studies evaluated the association between employment in the styrene-butadiene rubber (SBR) industry, exposure to 1,3-butadiene, mortality from leading causes of death with particular focus on lymphohaematopoietic cancers among workers from 1944-1998. A previous study (Delzell et al. 1996, Macaluso et al. 1996, Sathiakumar et al. 1998) evaluated the mortality experience of these workers through 1991. These studies add a further seven years of follow up.

GLP compliance:

no

Test material

Test material information

Constituent 1

Reference substance name:

1,3-Butadiene

Cas Number:

106-99-0

Molecular formula:

C4H6

IUPAC Name:

1,3-Butadiene

Details on test material:

Butadiene as used in the styrene-butadiene rubber (SBR) industry in North America (7 plants) and Canada (1 plant)

Method

Type of population:

occupational

Ethical approval:

not specified

Details on study design:

The 1944-98 mortality experience of 17,924 workers at 8 North American plants was evaluated in relation to time since hire, work duration and work area (Mortality study). Subjects were men who had worked at the plants for at least 1 year before the end of 1991, and were actively working in or after a specific calendar year between 1943 and 1965, which varied by plant. Workers at 2 of the 8 plants did not have work area /job group information of sufficient quality for quantitative exposure estimation, therefore the association between exposure to 1,3-butadiene, styrene and dimethyldithiocarbamate (DMDTC) and mortality from lymphohematopoietic cancer (LHC) was evaluated among the 16,579 workers at the remaining six plants (LHC study). These studies are an update of a previous mortality investigation where workers were followed from 1944 through 1991. The earlier studies are described in Delzell et al (1996), Macaluso et al (1996) and Sathiakumar et al (1998), the first two references also describe the plants in detail. Updated/refined exposure estimates in Macaluso et al. (2004) were used for the study by Delzell et al. (2001) and in the present investigations. In the present studies, Poisson regression analyses were used to model LHC rates and included as LHC decedents all subjects with LHC as a contributing or underlying cause of death except those whose medical records (where available) confirmed that they did not have LHC. Underlying causes of death rates were compared with those for the general population. Exposure was estimated using work histories.

Exposure assessment:

estimated

Details on exposure:

The previous studies developed time weighted averages (TWA) for butadiene exposure.
The exposure estimation procedure is described in detail in Macaluso et al (2004).

Exposure was estimated using mathematical modeling

Procedures included, for each agent:
1) Identifying at each plant a series of work area/job groups, each of which was homogeneous with respect to its component tasks and exposure potential.
2) Identifying for each plant-specific work arealjob group its component tasks that entailed exposure and documenting historical changes in those tasks.
3) Calculating plant-, work area/job group-, and time-specific average exposure indices and compiling these into job-exposure matrices (JEMs)
4) Linking the time- and work area/job group-specific monomer exposure estimates in the JEMs with each subject's work history to obtain cumulative exposure estimates. The latter computation involved multiplying the calendar year-specific amount of time a worker spent in each work area/job group by the exposure estimate for that work area/job group and calendar year category, and summing over all work area/job title groups and years covered by a subject's employment history.

The cumulative exposure indices for butadiene were:
- Butadiene total ppm-years
- Butadiene ppm-years due to exposures to intensities ≤ 100 ppm,
- Butadiene ppm-years due to exposure to intensities > I00 ppm
- The total number of butadiene peaks (>100 ppm)

Statistical methods:

Poisson regression models to examine lymphohaematopoietic cancer (LHC) rates in relation to chemical exposures, within the six-plant study group and without reference to an external comparison population. These models provided maximum likelihood estimates of the relative rate (RR) for the contrast between categories of one agent, adjusting for other agents and for additional potential confounders. Except where specified, Poisson regression analyses of LHC included all decedents who had LHC mentioned on their death certificates and whose LHC diagnosis was not contradicted by medical records obtained and reviewed by a study pathologist. Poisson regression analyses was performed using the Statistical Analysis System (SAS) procedure GENMOD. External analyses by agent exposure level and factors such as time since hire, work duration and work area were conducted using the standardized mortality ratio (SMR) (observed/expected number of LHC deaths) as the measure of association and using age-, calendar period- and race-specific US male and age- and calendar period-specific Ontario male population rates to compute expected numbers of LHC deaths. In these analyses, a subject counted as an observed event only if he had LHC coded as the underlying cause of death. External analyses were performed using the Occupational Mortality Analysis Program (OCMAP) .

Results and discussion

Results:

MORTALITY STUDY
There were 17,924 subjects (all male). At the end of 1998 the median time since hire was 33 years and the median age was 62 years.
Observed/expected numbers of deaths was 6237/7242 (all causes using standardised mortality ratio [SMR] of 86, 95% Cl 84 - 88).
Observed/expected numbers of deaths for all cancers combined was 1608/1741 (all cancers combined (SMR = 92, 95% Cl 88 - 97).
Lung cancer was the cause of 35% of all deaths (SMR = 91, 95% Cl 84 - 99).
Among the subjects included there were fewer deaths than expected for most forms of cancer with the exception of Hodgkins’s disease, leukemia, colorectal cancer and prostate cancer, none of these achieved statistical significance. A 16% increase in leukemia incidence was mainly attributable to subjects with 20-29 years since hire, 10 years or more employment in the industry and hourly paid (19/7.4, SMR=258, CI 156 - 403), Analyses by work area showed elevated leukemia for polymerisation (18/8.8 SMR=204, CI 121 - 322), coagulation (10/4.3 SMR=231, CI 111 - 425), maintenance labour (15/7.4 SMR=203 , CI 114 - 335) and laboratory operations (14/4.3 SMR=326, CI 178 - 546). The authors concluded that increases in colorectal cancer and prostate cancer did not appear to be related to occupational exposure in the industry.

LHC STUDY
All forms of leukemia in a single agent model adjusting for age and years since hire were positively associated with cumulative exposure to 1,3-butadiene. Macaluso et al. (2004) reported that cumulative exposures to 1,3-butadiene were highly correlated with cumulative exposures to styrene (Spearman ρ = 0.78) and DMDTC (Spearman ρ = 0.60) and cumulative styrene and DMDTC exposure were also positively associated with all leukaemias as were total 1,3-butadiene peaks (> 100 ppm) and total styrene peaks (> 50 ppm). Adjustment for cumulative styrene exposure and DMDTC in a multiple agent model resulted in a weakened association between cumulative 1,3-butadiene exposure and all forms of leukaemia. In single agent models, the strongest association was with 1,3-butadiene total peaks > 100 ppm and the RR was 4.9 (CI = 2.2 – 10.6) for the subgroup with the highest exposure ( ≥ 3136.5 1,3-butadiene total peaks > 100 ppm). Cross classified categories of cumulative exposure to 1,3-butadiene partitioned into ppm-years due to intensities > 100 ppm and ≤ 100 ppm were analysed. No subjects died from leukaemia in the highest category of 1,3-butadiene ppm-years due to intensities > 100 ppm in the subgroup having no or low cumulative 1,3-butadiene exposure due to intensities ≤ 100 ppm. In the subgroup having low or no exposure to 1,3-butadiene due to intensities > 100 ppm, the RR was 1.9 (CI = 0.8 – 4.7) for the category with 7.7 to < 124.7 ppm–years due to intensities ≤ 100 ppm. Sparce data prevented further analysis
Butadiene ppm-years were associated positively with chronic lymphocytic, chronic myelogenous, and residual types of leukemia but not with acute myeloid leukemia. The association with butadiene ppm-years was particularly strong for chronic myelogenous leukemia (RRs for exposure to <33.7, 33.7 to <425.0, and 425.0+ ppm-years, unadjusted for other agents: 1.0, 2.7, and 7.2). No association was found between BD exposure and NHL, no consistent results to indicate multiple myeloma was caused by BD exposure.

Confounding factors:

Subjects were co-exposed to styrene and dimethyldithiocarbamate athough the models used included terms for these exposures. SBR production workers were exposed to many other chemicals and, in the past, some may have been exposed to benzene which was used in small quantities at most plants, primarily in certain laboratory procedures and maintenance operations, as well as in some solution operations. The authors had no information on potential non-occupational confounders or on chemicals other than butadiene, styrene, and DMDTC used in the synthetic rubber industry. Estimates of benzene exposure were used in a previous report on the study by Macaluso et al (1996), but no clear independent effect was reported.

Strengths and weaknesses:

The subjects included constituted a large group with many years of follow-up and reasonably complete vital status ascertainment. This update added considerably to the amount of information available for assessing mortality patterns and used expanded and refined exposure estimates.Problems with historical exposure estimation may have affected the measurement of all relations between the agents and diseases of interest. Work histories sometimes did not fully specify subjects' work areas and were not updated beyond 1991, the end of follow up for the original study. Approximately 4000 workers were still employed in 1991 and did not have complete job histories. The exposure estimation procedures were based on models requiring many assumptions, and the resulting estimates lacked comprehensive validation although Sathiakumar et al. (2007) have validated butadiene estimates at one plant. Errors in exposure measurement could have distorted exposure-response relations and led to residual confounding in analyses that estimated the effect of one agent adjusted for the effects of other agents. However, Graff et al. (2009) have examined the impact of uncertainty in exposure estimation on the dose-response patterns seen for leukemia.High correlation between the agents of interest made estimation of associations difficult. Misclassification of disease was also possible, due to the use of death certificates as the source of cause of death information.

Any other information on results incl. tables

TABLE 1. Data from Mortality Study. Observed/expected number of deaths, standardized mortality ratio (SMR) and 95% CI for selected cancers in all subjects 1944-1998

	All leukaemia	Hodgkin's disease	Multiple myeloma	Non-Hogkins lymphoma	Colorectal cancer	Prostate cancer
Obs/exp SMR (95% Cl)	71/61.1 116 (91-147)	12/10.8 111 (58-195)	26/27.3 95 (62-140)	53/53.3 100 (75-130)	193/177.8 109 (94-125)	154/148.6 104 (88-121)

TABLE 2. Data from LHC Study. Leukaemia Relative Rate (RR) for Cumulative Exposure to Butadiene (Single Agent and Multiple Agent Models)

Butadiene ppm-years	LHC/Person year	Single agent model Relative Risk (95% Cl)	Multiple agent model Relative Risk (95% Cl)
0	10/116,471	1.0	1.0
>0-<33.7	17/154,443	1.4  (0.7-3.1)	1.4  (0.5-3.9)
33.7-<184.7	18/144,109	1.2  (0.6-2.7)	0.9  (0.3-2.6)
184.7-<425.0	18/49,411	2.9  (1.4-6.4)	2.1  (0.7-6.2)
425.0+	18/35,741	3.7  (1.7-8.0)	3.0  (1.0-9.2)

TABLE 3. Data from LHC Study. Leukemia Relative Rate (RR)*, Adjusted for Age and Years Since Hire, for Cross-Classified Categories of Butadiene ppm-years Due to Intensities>100 ppm and Butadiene ppm-years Due to Intensities ≤100 ppm, With Each Classified as 1) No Exposure Plus the First Quartile, 2) the Second and Third Quartiles, and 3) the Fourth Quartile of Leukemia Decedents’ Distributions

		Butadiene ppm-years due to intensities >100 ppm
Butadiene ppm-years due to intensities ≤100 ppm		0-<16.3	16.3-<247.6	247.6+	Marginal butadiene ppm-years due to intensities≤100 ppm
0-<7.7	L/PY RR 95%Cl	21/211,934 1.0 -	6/58,633 0.9 0.4-2.2	No leukemia **	27/242,377 1.0 -
7.7-<124.7	L/PY RR 95%Cl	6/30,336 1.9 0.8-4.7	21/135,827 1.2 0.7-2.2	9/19,125 2.7 1.2-6.0	36/216,788 1.2 0.7-2.0
124.7+	L/PY RR 95%Cl	No leukemia **	9/22,328 3.1 1.4-6.7	9/19,091 2.4 1.1-5.3	18/41,008 2.1 1.2-3.9
Marginal butadiene ppm-years due to intensities >100ppm	L/PY RR 95%Cl	27/273,560 1.0 -	36/185,288 1.6 1.0-2.6	18/41,526 2.8 1.6-5.2	P = 0.44***

L/PY number of leukemias/number of person years.

* All RRs were adjusted for categories of age (<40, 40-49, 50-59, 60-69, 70+) and years since hire (<20, 20-29, 30+) only, except where indicated.

**Category not included in Poisson regression model.

***P value for trend in BD ppm-years due to exposure ≤100 ppm, adjusting for in BD ppm-years due to exposure >100 ppm , age and years since hire.

****P value for trend in BD ppm-years due to exposure >100 ppm, adjusting for in BD ppm-years due to exposure ≤100 ppm , age and years since hire.

Applicant's summary and conclusion

Conclusions:

A positive association was demonstrated between workplace exposure to butadiene in the styrene-butadiene rubber industry and leukemia. Uncertainty remains about the specific agent(s) associated with the excess of leukaemia among synthetic rubber industry workers.

Executive summary:

These studies evaluated approximately 18,000 workers in the synthetic rubber (SBR) industry who were exposed to 1,3 -butadiene, styrene and dimethyldithiocarbamate. The study added 7 years of follow up to a previous investigation so that the workers were followed from 1944 to 1998. Some sub-groups of workers were found to have an excess of mortality from leukemias. There was no clear relationship between employment in the industry and other forms of lymphohaematopoietic cancer. The association between exposure to the specific chemicals and lymphohematopoietic cancer mortality was then investigated. A positive association was found between butadiene and leukemia that was not explained by exposure to other agents examined. Butadiene was most consistently associated with all leukemias combined and with chronic myelogenous leukemia. Butadiene exposure was not associated with non-Hogkins lymphoma and the study provided no persuasive evidence that exposure to BD causes multiple myeloma.