Neoplastic conversion induced in rat liver by diethylnitrosamine (DEN) was quantified by measuring preneoplastic and neoplastic lesions over a 34 week period in the beginning of which the carcinogen was given at three dose levels and two dose rates for the first 10 weeks, after which animals were maintained for 24 weeks with either no further exposure or were fed phenobarbital (PB) to promote neoplastic development of cells converted by DEN. DEN was injected s.c. in male F344 rats at weekly or biweekly intervals for total doses of 1, 2 or 4 mmol/kg body wt and then the rats were maintained on basal diet alone or diet containing 0.05% PB. At the end of exposure, DEN had produced a dose-related decrease in centrilobular glutamine synthetase-expressing (GS+) hepatocytes which is indicative of mild cytotoxicity. All doses induced foci that were gamma-glutamyltranspeptidase-positive and iron storage-deficient. The multiplicity of foci in the middle dose exceeded that in the low dose by about a factor of two and, in the high dose, was > 10-fold greater. A few GS+ foci were found in the high dose group only. At 34 weeks, neoplasms were present in the middle and high dose groups. Administration of PB after DEN increased the multiplicity of foci in all dose groups, most substantially in the low dose group. The effect of PB on liver neoplasm yield was marginal in the low non-carcinogenic dose, whereas it enhanced the multiplicity in the weakly carcinogenic middle dose by approximately 10-fold. Four principal findings were made: (i) even at the low doses used, a mild cytotoxic response not evidenced by morphological changes in conventional histopathology was manifested in the GS+ centrilobular subpopulation of hepatocytes; (ii) the dose response over a 4-fold dose range of DEN alone and when followed by PB was non-linear; (iii) the precursor role of foci in the evolution of liver neoplasms was evident and was most conspicuous in the case of GS+ foci; and (iv) a high level of foci induction was required for the evolution of neoplasms, even with PB promotion. The finding of non-linearity with increasing doses of DEN raises questions about the assumption that effects of carcinogens at high doses can be quantitatively extrapolated to low doses.
