81
EXPOSURE ASSESSMENT FOR AN EPIDEMIOLOGIC STUDY OF BRAIN TUMORS AMONG
PETROCHEMICAL RESEARCH WORKERS.
R. Kalmes, G. Brorby, M. Kelsh, Exponent, Oakland, CA; P. Buffler, University of California, Berkeley, Berkeley, CA.
Some of the epidemiologic studies of employees in the petrochemical industry have reported higher risks of brain cancer; however, no specific agent has been consistently associated with brain cancer risk across these studies or in studies of similar workers. Interpreting the results of these studies has been further complicated because they have relied on proxies of exposure (e.g., job title, length of employment, and work location) in the absence of direct measurements of exposure (e.g., air-monitoring data or biological measurements). Although surrogates of exposure are useful in this context, the power of an occupational epidemiologic study to evaluate and quantify exposure-effect relationships depends, in part, on the resolution of meaningful differences in exposure within the study population. As part of a cohort mortality and nested case control study of petrochemical workers at a Southern California facility, a concerted effort was made to examine a wide variety of possible sources of quantitative and qualitative exposure information, including systematic file reviews, facility closure and feasibility reports, and key informant interviews. This case study presents the exposure methodologies that were used in the mortality study and how these methodologies were enhanced to provide more detailed classifications for use in the case-control study with regard to potential solvent, mercury, radiation, and electric and magnetic field exposure. This case study will also describe the use of key informants and expert panels throughout the study and their importance not only in terms of recalled experience and accuracy, but also as a key component in establishing employee confidence in the validity of the study methods and, ultimately, the results.
82
RETROSPECTIVE EXPOSURE ASSESSMENT FOR A CASE-CONTROL STUDY OF BRAIN CANCER AMONG
EMPLOYEES AT A CHEMICAL RESEARCH FACILITY.
A. Kreft, Rohm and Haas Company, Spring House, PA.
Agent-specific exposure data were assembled in a case-control type epidemiological investigation at a Rohm and Haas Company Research Facility. The study involved 12 individuals who had developed primary malignant brain cancers, three individuals who had developed benign tumors, and four matched controls for each case. The purpose of the exposure assessment was to assemble a list of chemical and physical agents that a study participant may have been exposed to while working at the facility, and the intensity and duration of such exposures.
For the detailed exposure assessment, 2016 research documents (e.g., business and technical reports and research notebooks ) associated with work done by all study participants were identified. A list of chemical and physical agents was assembled from each of the documents for a total of 20,653 agent names. For purposes of epidemiological analysis the resulting agents were then assigned a unique identifier (e.g., CAS Reg. No.) and assigned to 36 different categories. The CAS Reg. No. was used to identify specific chemicals with mention in a study associated with occurrence of brain cancer. The categories were based on the following: agent known to cause brain tumors; agent known to cause cancer; agent used frequently at Rohm and Haas; and agent with specific exposure concern. Lastly, the list was merged with the work histories and epidemiological analyses were conducted.
The study findings showed no statistically significant association with any of the workplace related risk factors with the occurrence of brain cancer or benign brain tumors. In the analysis of building location, several of the individuals with brain cancer worked on similar projects and in the same building at one time during their careers, but did not show a statistical significance compared with the control population.
83
THE CHARACTERISTIC CHANGES IN EXPOSURE LEVELS IN CHLOROPRENE MONOMER AND POLYMER
PRODUCTION.
N. Esmen, University of Illinois at Chicago, Chicago, IL; T. Hall, M. Phillips, P. Jones, H. Basara, University of Oklahoma, Oklahoma City, OK; G. Marsh, University of Pittsburgh, Pittsburgh, PA.
In an epidemiologic study, the characteristics of exposure to chloroprene were investigated through the examination of company engineering documents and interviews of knowledgeable personnel on processes, their development, work practices, and changes in both. These observed relationships were summarized by relative changes in the exposure levels by mathematical modeling all task-specific exposure potentials. The results of the models showed good agreement with the extensive exposure data available (41,657 personal samples from four facilities located in three countries).
After the process was introduced (1931), it stayed small until a rapid expansion in 1940. 19401950 was the learning phase, in which almost all processes took place under sparsely controlled conditions with many upsets and mistakes from which to be learned. High exposures occurred frequently at many process points during the upsets and normal operations.
The evolutionary improvement-innovation phases of the production (19501975) exhibited rapid changes in the production procedures and equipment culminating in the introduction of a new process. Although, responsible for impressive exposure reduction or elimination, these improvements were implemented for many reasons such as efficiency, product refinement, and cost, all unrelated to health concerns. After the maturation of chloroprene production, the hygienic considerations, personal air monitoring became integrated to the process management with further reduction of exposures.
The exposure trends show that the characteristic exposure levels at different epochs were declining sporadically and unevenly: Introduction 7050 (normalized units), Learning 5030, Improvement 3020, Process innovations 205, Maturation 53, Exposure controls 30.1. There was about a 30-fold process-related reduction before the recognition and introduction of exposure-specific controls. The results demonstrated the importance of process analysis and when they were compared to trends observed in several other industries, they showed remarkable similarities between unrelated industries, suggesting that an epochal analysis might be an important tool in exposure reconstruction.
84
EXPOSURES TO CHLOROPRENE MONOMER AND OTHER PROCESS CHEMICALS DURING
POLYMERIZATION PROCESSES AT FOUR POLYCHLOROPRENE SYNTHESIS FACILITIES.
P. Jones, T. Hall, H. Basara, M. Phillips, The University of Oklahoma, Oklahoma City, OK; N. Esmen, University of Illinois at Chicago, Chicago, IL; G. Marsh, J. Buchanich, University of Pittsburgh, Pittsburgh, PA; R. Leonard, E.I. du Pont de Nemours Inc., Wilimington, DE.
Early studies of chloroprene workers conducted by both NIOSH and the major manufacturer in the U.S. did not demonstrate an increase of cancer at any site. In the 1990s however, a large scale rodent bioassay conducted under the NTP indicated a potential increased lung cancer risk in chronically-exposed rodents. This observation spurred a re-evaluation of previous studies of chloroprene workers. An international epidemiologic study of four major polychloroprene production facilities encompassing production facilities in four countries was begun in 1997. During the mid 1970s, this process was enclosed, thereby significantly reducing the potential for worker exposure. In the area known as Polymer Operations, exposures to chloroprene monomer now occur primarily during process sampling, cleaning process screens/filters, and during leaks and/or spills. At one production facility, historical exposure estimates were dominated by exposures that occurred during the actual polymerization process when process samples were collected from open top reaction vessels. Exposure data consisted of more than 40,000 personal exposure measurements for both a- and b-chloroprene along with other process-related chemicals. Air sampling in polymer operations began in the early 1970s with a majority of the air samples (90%) collected between 1975 and 1991. These data indicate a clear decreasing trend in polymer operator exposures to both chloroprene isomers. In the 1970s, maximum exposure estimates for polymer operators ranged between 1 and > 250 ppm. The median exposures for these same occupations were typically less than 3 ppm. Maximum exposures for these same occupations today have been reduced to maximum of < 6 ppm with the median for the occupations in this production area of < 0.9 ppm. It should be noted that the actual human exposures occurring since the mid-1980s to present are significantly mitigated through the use of respiratory protection devices that are required where monomer exposure can occur.
85
EXPOSURES TO CHLOROPRENE MONOMER (α and β) AND OTHER PROCESS CHEMICALS DURING
MONOMER PRODUCTION AT FOUR POLYCHLOROPRENE SYNTHESIS FACILITIES.
T. Hall, P. Jones, H. Basara, M. Phillips, University of Oklahoma, Oklahoma City, OK; N. Esmen, University of Illinois at Chicago, Chicago, IL; R. Leonard, I.E. du Pont de Nemours Inc., Willimington, DE; M. Gary, J. Buchanich, University of Pittsburgh, Pittsburgh, PA
Polychloroprene production on an industrial scale began in the United States during the 1930s to replace natural rubber. Early studies of chloroprene workers conducted by both NIOSH and the major U.S. manufacturer did not demonstrate an increased risk of cancer at any site. In the 1990s, a rodent bioassay (NTP) indicated a potential increased lung cancer risk in chronically-exposed rodents. Polychloroprene production is divided into two major subdivisions that are classified as monomer and polymer. Potential exposures to workers differ within these production subclassifications due to the different production processes in use. Originally, chloroprene was synthesized using acetylene as a basic building block. During the 1970s, the synthesis process for the monomer was altered to use butadiene as the basic building block for the monomer. In the area known as Monomer Operations, exposures to chloroprene monomer could occur during process sampling, or during leaks and/or spills. An international epidemiologic study of four major polychloroprene production facilities encompassing production facilities in three countries was begun in 1997. Exposure data was gathered from each facility for use in reconstructing historical exposures. Exposure data consisted of more than 40,000 personal exposure measurements for both α- and β-chloroprene along with other process-related chemicals. Air sampling in monomer operations began as early as 1967 with a majority of the air samples (90%) collected between 1975 and 1991. These data indicate a clear decreasing trend in monomer operator exposures to both isomers. In the 1970s, exposure estimates for monomer operators ranged between 1 and 20 ppm, while in the 1990s these same occupations experienced exposures ranging from < 0.001 to 2.6 ppm. It should be noted that exposures occurring in the mid 1970s to present are significantly mitigated through the use of respiratory protection devices that are required where monomer exposure can occur.
86
MORTALITY PATTERNS AMONG INDUSTRIAL WORKERS EXPOSED TO CHLOROPRENE AND OTHER
SUBSTANCES.
G. Marsh, A. Youk, J. Buchanich, University of Pittsburgh, Pittsburgh, PA; N. Esmen, University of Illinois at Chicago, Chicago, IL; T. Hall, University of Oklahoma, Oklahoma City, OK.
We conducted an historical cohort study to investigate the total and cause-specific mortality experience of industrial workers potentially exposed to chloroprene (CD) and other substances, including vinyl chloride monomer (VCM), with emphasis on cancers of the liver and lung. CD is a monomer used almost exclusively for the production of polychloroprene elastomers and latexes. In 1999, The International Agency for Research on Cancer (IARC) classified CD as a possible carcinogen (Group 2B) based on sufficient animal and inadequate human evidence. VCM was classified by IARC in 1987 as a known human carcinogen (Group 1) based on sufficient animal and human evidence. Subjects were 12,500 workers ever employed at one of two U.S. industrial sites (Louisville, Ky., and Pontchartrain, La.) or two European sites (Maydown, Northern Ireland, and Grenoble, France), with earliest CD production dates ranging from 1942 to 1969. Two sites (Louisville and Maydown) synthesized CD with the acetylene process that produced VCM exposures. We determined vital status through 2000 for 95% of subjects and cause of death for 95% of the deaths. Historical exposures for individual workers were estimated quantitatively for CD and VCM. We computed standardized mortality ratios (SMRs; using national and regional mortality comparisons) and internal cohort relative risks in relation to demographic and work history/exposure factors with adjustment for potential confounding factors. For the 5506 subjects from the larger U.S. site (Louisville) that began CD production in 1945 (44% of cohort), cumulative CD and VCM exposures ranged from 0.001 to 1351.5 and 0.001 to 58.0 ppm-years, respectively. Regional rate-based SMRs for all cancers combined, respiratory system cancer, and liver cancer were, respectively, 0.73 (p < .01), 0.74 (p < .01), and 0.84. Similar SMRs were found among the 4727 subjects exposed to CD. Among 2603 subjects exposed to VCM, corresponding SMRs were 0.59 (p < .01), 0.35 (p < .01), and 0.38.
Posted May 30, 2005