Hypothesis-driven weight-of-evidence analysis for the endocrine disruption potential of benzene.

Exposure to benzene has many sources, from gasoline refueling to tobacco combustion. Although the toxicology of benzene is well studied, the potential for environmental exposure and a heightened interest in identifying substances that may cause toxicity by interacting with the endocrine systems of humans and wildlife resulted in benzene being placed on the second list of chemicals for possible screening under the USEPA's Endocrine Disruptor Screening Program. Therefore, we conducted a thorough, systematic literature search and used a weight-of-evidence methodology to test hypotheses regarding the potential for benzene to act via estrogen, androgen, thyroid, and steroidogenic pathways. The methodology included an assessment of data quality and a semi-quantitative weighting of endocrine-responsive endpoints measured in various types of studies according to their relevance for evaluating each hypothesis. This maximized use of all relevant and reliable literature on benzene and enabled a transparent comparison of evidence supporting and opposing each hypothesized mode of action. While benzene affected reproductive organ weights and histopathology in a few studies, there was no consistent pattern of effects suggestive of an estrogen, androgen, thyroid or steroidogenic mode of action. Based on data from multiple animal species, benzene appears to lack endocrine activity by these pathways.