Case-control study of brain and other central nervous system cancer among workers at semiconductor and storage device manufacturing facilities.

OBJECTIVE: This study evaluated the relationship between brain and other central nervous system cancer ('CNS cancer') and exposures at two semiconductor and electronic module manufacturing facilities and at a storage device manufacturing facility. METHODS: The case-control study, nested in a cohort of 126 836 employees, compared 120 CNS cancer cases and 1028 matched controls with respect to employment in 10 process groups and estimated cumulative exposure to 31 known or possible carcinogens. RESULTS: CNS cancer was associated with module manufacturing operations at two facilities. Module manufacturing is a process that begins with production of ceramic substrates followed by attachment of completed semiconductor chips and metal-containing circuitry resulting in a high performing electronic device. Positive associations with the highest tertile of estimated cumulative exposure were found for several chemicals, including 2-butoxyethanol, cyclohexanone, ortho-dichlorobenzene, cadmium, molybdenum, trichloroethylene and vinyl chloride. CONCLUSIONS: Results suggested positive associations between CNS cancer and specific operations and chemicals experienced in the semiconductor and electronic module manufacturing industry. However, lack of external support for these findings precludes a causal interpretation, and the observed associations may have been due to chance.

Retrospective Exposure Assessment for Semiconductor and Storage Device Manufacturing Facilities.

OBJECTIVE: The aim of this study was to develop exposure estimates for a case-control study of central nervous system cancer in semiconductor and storage device manufacturing workers. METHODS: Over 700,000 records on sample measurements were obtained. Mean exposure intensity was estimated for 31 established/suspected carcinogens (agents of interest) in 10 primary exposure groups over three manufacturing eras. We assigned confidence ratings, based on number/type of measurements, to each estimate. RESULTS: Mean exposures decreased by an order of magnitude over the three manufacturing eras and were below applicable occupational exposure limits for 22 of 31 agents of interest. For 25 agents, at least 70% of the exposures were estimated with high confidence. CONCLUSION: This is the largest comprehensive study of exposures in the semiconductor/storage device industry and the first to include cumulative exposure estimates and measures of confidence in the exposure estimates.