We evaluate the power of a recent protocol proposed by the European Union (
EU) for testing the carcinogenicity of inhaled fibers. We assume that every
fiber has oncogenic potential determined by its biopersistence. We use a r
ecently estimated potency for the oncogenic potential of fibers together wi
th experimentally determined "weighted" half-lives (WHL) of a dozen fibers
to generate simulations of long-term bioassays conducted according to the E
U protocol. We analyze these experiments using standard statistical techniq
ues and determine the number of tests that would have yielded significant r
esults. We conclude that the EU protocol will readily detect the carcinogen
ic effect of long-lived fibers, such as amosite with a WHL of more than 450
days, and usually detect the effect of fibers, such as RCF1a, with WHL of
about 40-50 days. However, the EU protocol has very low power to detect eff
ects of short-lived fibers, such as X607,with WHL of about 10 days. (C) 200
1 Academic Press.