REGENERATIVE HYPERPLASIA IS NOT REQUIRED FOR LIVER-TUMOR INDUCTION INFEMALE B6C3F(1) MICE EXPOSED TO TRIHALOMETHANES

Chloroform (TCM), a water disinfection by-product, induced liver tumor s in female mice when administered by gavage in corn oil but not when given in drinking water at comparable daily doses. Because short-term studies showed that the gavage doses also induced liver toxicity, it h as been suggested that the liver tumor response occurs secondary to cy totoxicity and consequent regenerative hyperplasia induced by oxidativ e metabolism of TCM to the toxic dihalocarbonyl intermediate. This stu dy compares dose-response relationships of gavage-administered chlorin ated/ brominated trihalomethanes for hepatotoxicity, replicative DNA s ynthesis, and hepatocarcinogenicity in female B6C3F(1) mice. The liver tumor data were obtained from previously published studies. Because b romine is a better leaving group than chlorine, metabolism of bromodic hloromethane (BDCM) should produce the same intermediates as would be formed from TCM. Hence, the toxicity and carcinogenicity of BDCM was e xpected to be qualitatively similar to that of TCM. Dose responses for liver weight, serum sorbitol dehydrogenase and alanine aminotransfera se (ALT) activities, hepatocyte degeneration, and hepatocyte labeling index (LI, a measure of replicative DNA synthesis) in female mice were similar following 3 weeks of gavage administration (once per day, 5 d ays per week) with TCM, BDCM, or chlorodibromomethane (CDBM). Fits of composite data for these trihalomethanes to a Hill equation model reve aled sigmoidal dose responses for ALT activity and hepatocyte LI and a nearly linear low-dose response for liver tumor incidence. For this f amily of chemicals, the mouse liver tumor response was not associated with an elevated hepatocyte LI at doses of approximately 1 mmol/kg or less. High incidences of liver tumors were observed,vith BDCM and CDBM at doses that had a marginal effect or no effect on the hepatocyte LI . Thus, the carcinogenic effects of trihalomethanes are not simply a c onsequence of cytotoxicity and regenerative hyperplasia. The possible contributions from other activation pathways, including GSH conjugatio n and reductive metabolism, need to be considered in assessments of th e carcinogenicity of the trihalomethanes.