Animal to Human Non-Cancer Dose Extrapolation for Toluene

Compare and contrast the methods and values used for toluene with those used for the noncancer effects of arsenic. We have sufficient information from arsenic contamination of drinking water in Taiwan and Chile to develop a protective dose for humans ingesting arsenic in drinking water therefore we do not need to extrapolate results from another animal species. Both a LOAEL and NOAEL are available from the drinking water studies from Taiwan. A NOAEL of 0.0008 mg/kg-day and a LOAEL of 0.014 mg/kg-day were reported. Based on these doses, it is clear that arsenic is more toxic to humans than toluene because the arsenic doses needed to cause toxic effects are much lower than those causing toxic effects by toluene. 

The only potentially relevant uncertainty factor for this study is for extrapolation to sensitive members of a population. Although more than 28,000 people were exposed to arsenic in the Taiwanese studies, these were from a specific area and did not necessarily represent the range of sensitivities to other ethnic groups (e.g., Caucasians). As a result, the U.S. ERA used a factor of 3 for this variable. This lower factor was used instead of the typical factor of 10 because data were available for 

Overall, this should indicate that there is much uncertainty when toxicity data are used to establish safe levels for human exposure. The goal is to ensure that the toxicity of a chemical is not underestimated in humcms so that we will have a margin of safety in setting regulatory limits for exposure. As a result, in most cases the established doses are more restrictive than necesscu y to protect human health. 

One main concern in extrapolating data involves estimating effects at low levels from the very high levels tested. At sufficiently low levels, the effects are so minor or rcu e that they likely Ccm t be measured. At the other end is death, which is easily mecisured. In most situations, humcms will be exposed to chemical concentrations closer to the low levels, where effects cire often minor, than to those causing death. However, most laboratory studies use levels just below those at which death occurs. Therefore, the degree of toxic effect at low levels needs to be estimated from effects at higher levels. 

The problems inherent in such extrapolations are illustrated in figure 7.2. We typically collect data at a few dose levels, represented by the dots. Using mathematical equations, we then draw a straight line that bisects these data points. Once we get to the lowest dose tested ( a in figure 7.2), the line ends because we don t know if the relationship between dose and response seen at high concentrations is the same at low concentrations. 

---

Table 7.1. Animal to Human Non-Cancer Dose Extrapolation for Toluene...

---