home edit2 whole TSD November 2002 PDF format - OEHHA

derived as follows,
∞
∫
q = f(q )⋅ q dq
bar 1 1 1
0
with f(q 1 ) as the frequency distribution whose log-likelihood function follows a chi-square distribution.
The selection of a cancer potency value should be made on the basis of the most sensitive site, species,
and study, in the absence of evidence that such a value is not representative. The hamster studies of
Pietra and Shubik (1960) and Toth et al. (1961a, 1969) indicate the hamster may not be as sensitive as
the mouse when individual tumor sites are compared. Although the limited data available do not rule out
the possibility that the hamster may be the more sensitive species, data are not available to make
quantitative comparisons of hamsters and mice. This, coupled with the fact that there are extensive
studies on mice, suggests the mouse urethane studies are more appropriate than hamster studies in
developing a cancer potency value. Only two studies in the rat are useful for deriving cancer potency
values. The Tannebaum et al. (1962) study showing development of Zymbal’s gland tumors is of
limited use since there is no supporting evidence that this site of tumor development is the most sensitive.
The Schmähl et al. (1977) study is also of questionable value because of incomplete reporting of tumor
incidence in untreated animals, in spite of showing sensitive induction of mammary tumors in female rats.
In light of these limits on the studies in hamster and rats, the body of data showing tumor induction in
mouse has been deemed most appropriate for the development of a cancer potency value.
Calculation of the geometric mean of q human and q bar values from the most sensitive sites of malignancy
development in the oral mouse studies resulted in values of 0.5 and 1.4 (mg/kg-day) -1 , respectively.
The geometric mean of the q bar values provides an estimate of the upper 95% confidence limit on the
distribution of values. Calculation of the geometric mean of q human and q bar values from mouse studies
where the lung was the most sensitive site of malignancy development resulted in values of 0.8 and 1.9
(mg/kg-day) -1 , respectively. These mean values, coupled with the q human values from the sensitive
multiple dose study by Schmähl et al. (1977), indicate the most plausible estimate of cancer potency for
urethane falls in the range of 0.6 to 3.0 (mg/kg-day) -1 . Therefore, as a reasonable estimate to the
cancer potency, 1.0 E+0 (mg/kg-day) -1 has been adopted as a cancer potency value.
A unit risk value based upon air concentrations was derived by OEHHA/ATES using an assumed
human breathing rate of 20 m 3 /day, 70 kg human body weight, and 100% fractional absorption after
inhalation exposure. The calculated unit risk value is 2.9 E-4 (µg/m 3 ) -1 .
543