188 • Alternatives to Animal Use in Research, Testing, and Education 
Table 8-1.— The Response of Known Human Carcinogens to 
Rodent Carcinogenicity and Bacterial Mutagenicity Assays 
Chemical 
Rat 
bioassay 
Mouse 
bioassay 
Ames 
test 
4-Aminobiphenyl 
+ 
+ 
+ 
Arsenic 
— 
— 
— 
Asbestos 
+ 
+ 
— 
Benzene 
— 
+ 
— 
Benzidine 
+ 
+ 
+ 
Bis(chloromethyl)ether 
+ 
+ 
+ 
Chromium; some chromium compounds. . . . 
+ 
— 
+ 
Cyclophosphamide 
-1- 
+ 
+ 
Diethylstilbestrol 
+ 
+ 
- 
Melphalan 
+ 
+ 
+ 
Mustard gas 
n.d. 
+ 
+ 
2-Naphthylamine 
- 
+ 
+ 
Soot, tars 
— 
+ 
+ 
Vinyl chloride 
+ 
+ 
+ 
KEY: + = Positive results (carcinogenic to rodents or mutagenic to bacteria). 
— = Negative results (not carcinogenic or not mutagenic), 
n.d. = No data. 
SOURCES: From H. Bartsch, L. Tomatis, and C. Malaveille, “Mutagenicity and Carcinogenicity of Environmental Chemicals,'' 
Regul. Toxicol. Pharmacol. 2:94-105, 1982; D. Brusick, “Evaluation of Chronic Rodent Bioassays and Ames Assay 
Tests as Accurate Models for Predicting Human Carcinogens,” Application of Biological Markets to Carcinogen 
Testing, H. Milman and S. Sell (eds.) (New York: Plenum Press, 1983); B.D. Goldstein, C.A. Snyder, S. Laskin et 
al., “Myelogenous Leukemia in Rodents Inhaling Benzene,” Toxicol. Lett. 13:169-173, 1982; and J.V. Soderman (ed.), 
Handbook of Identified Carcinogens and Noncarcinogens, Vols. I and II (Boca Raton, FL: CRC Press, 1982). 
substances have not been shown to be noncarcino- 
genic, and many authorities maintain that the in- 
formation is insufficient to make any statement 
about the proportion of noncarcinogens that are 
also nonmutagens in the Ames test (4,116). 
Use of the Ames Test in 
a Battery of Tests 
The predictive value of the Ames test, or other 
mutagenicity tests, can be improved by combin- 
ing it with additional short-term assays to form 
a test battery . Although no U .S . regulatory agency 
has yet recommended a specific combination, most 
authorities recommend that an appropriate bat- 
tery should include information from a minimum 
of three types of tests: 
• gene mutation (Ames test, mouse lymphoma 
test); 
• chromosomal mutation (in vivo Chinese ham- 
ster ovary cell cytogenetics); and 
• DNA damage (sister chromatid exchange, 
unscheduled DNA repair). 
At least one test should include a mammalian in 
vitro cell, tissue, or organ culture assay (4). 
In a recent study, 18 Ames tests averaged 66 per- 
cent “accuracy” (number of chemicals correctly 
identified/number of chemicals tested). Compara- 
tive results from six batteries of short-term tests 
that included the Ames test increased the accuracy 
to 82 to 90 percent (58,111). 
CURRENT TRENDS 
As long as toxicological data continue to be re- 
quired by regulators and by the courts to protect 
human health, animal testing will continue for the 
foreseeable future . Even major progress in the de- 
velopment and implementation of alternatives will 
not necessarily eliminate whole-animal tests . Fur- 
thermore, there are several impediments to devel- 
opment and implementation: 
• A large number of scientists have been trained 
to solve health problems and to invent new 
products using animal models. 
