Group studies risk-assessment for indoor air pollution

January 24, 1992

By Diane LaMacchia

Benzene is a common indoor air pollutant known to be carcinogenic to humans. In an important step toward gauging the health risks of benzene and other indoor pollutants, scientists at LBL and UC Berkeley are developing computer models that integrate information about how benzene reacts in the body with data on how cancer progresses.

The models will make it possible to use data from laboratory experiments and industrial settings -- in which exposures to toxic chemicals can be very high -- to predict the health risks from the typically lower doses of pollutants encountered in an indoor environment.

Benzene was chosen for the study because it is a widely-used industrial solvent whose effects have been studied for a hundred years. There is a wealth of data on occupational exposure to benzene, a compound contained in degreasers and gasoline, as well as tobacco smoke.

The team of researchers includes Frederic Bois and Penelope Compton from LBL and Martyn Smith and Robert Spear from UC Berkeley's School of Public Health.

Bois, a member of Energy and Environment's indoor environment group, says the disposition, transport, and metabolism of benzene is complex. When benzene enters the body, it is transformed enzymatically into other compounds -- called metabolites -- such as benzene oxide, phenol, hydroquinone, and catechol. These compounds are more toxic than benzene itself. Ultimately, the metabolites are excreted, but in the process these compounds have had their effect on the body.

Bois says the formation of metabolites from benzene involve non- linear enzymatic reactions. They are formed in the liver and transported to the bone marrow where they cause cancer.

"Benzene by itself has minimal effects in in vitro tests, while many of its metabolites have cell or gene toxicity, or both," Bois says.

The scientists are using the model to look at why benzene causes cancer while one of its metabolites, phenol, does not. According to the model, after benzene itself is introduced, the metabolites form differently compared to when phenol is administered. One of the model's conclusions is that metabolites such as phenol and catechol reinforce each other's toxic effects, and these effects are much larger when the process starts with the administration of benzene itself.

The model is also being used to predict the effects of benzene exposure in the workplace. "It had been assumed that if you are exposed to 300 parts per million of benzene in the air for 10 minutes, you'd have the same exposure as at three parts per million for 1,000 minutes," Bois says, "but this is not true. These two conditions won't create the same exposure."

In the workplace, average exposure is monitored and is not considered harmful if it does not exceed 10 ppm. But this average can be the result of widely fluctuating exposures. Occupational Safety and Health Administration legislation is currently in the works to limit peak exposure, but there has never been a scientific rationale for doing so. "Now we are able to provide a justification," Bois says.

While the benzene model works on the cellular and extracellular level, the cancer model operates at the level of the cell nucleus. One question the model can address is whether the effects from carcinogens administered at high doses are due mostly to genetic toxicity or to the activation of cell toxicity and subsequent cell proliferation. It appears that some carcinogenic compounds do not alter DNA but rather increase cell proliferation, according to Bois. If this is the case, then it may be possible to establish a particular dose threshold for benzene, rather than banning exposure completely. However, Bois says, there is a good indication that benzene acts, in part, by damaging DNA.

The cancer model will also be used to investigate the differences among various types of genetic damage and learn which of them is more likely to lead to cancer. Site specificity can also be explored. For example, why does cancer appear more often in the liver than in muscle tissue or the brain?

Hopefully, this model will lead to an understanding of the crucial factors in carcinogenesis, Bois says. This knowledge could help scientists determine where to direct their research and lead them to create effective prevention strategies for environmentally related cancer.