Invasiveness, metastasis, and transplantability of mouse liver nodules.

Data are presented which show that various mouse liver nodular lesions, such as hyperplastic nodules, hepatocellular adenomas, and hepatocellular carcinomas, differ not only morphologically but also biologically regarding their behavior in the primary and secondary hosts. The hyperplastic nodules grew slowly by expansion without affecting the integrity of the lobular structure. The primary hepatocellular adenomas also grew slowly by expansion but eliminated the histoid liver structure. Neither lesions metastasized to the regional lymph nodes and distant organs. Upon isogeneic transplantation, the adenomas required a long latent period before manifesting an expansive type of growth without invasion and metastasis. The primary hepatocellular carcinomas grew fast, destroyed surrounding parenchyma, and those showing moderately to poorly differentiated cellular morphology metastasized to regional lymph nodes and distant organs. Upon transplantation hepatocellular carcinomas grew fast, invaded locally, metastasized distantly, and killed the recipients within a few weeks. Thus hepatocellular adenomas and carcinomas represent two distinct entities both morphologically and biologically. Thus the agents should be classified according to the type of nodular liver lesions they "induce" in the mouse liver. The least innocuous agents would be those resulting in the development of hyperplastic nodules. They should not be labelled as hepatocarcinogens. On the other side of the spectrum should be the agents "inducing" the hepatocellular carcinomas and they should be classified as mouse liver carcinogens. The agents capable of affecting significantly the hepatocellular adenomas but not the hepatocellular carcinomas should be considered as tumorigens, since they are causally related only with the benign liver lesions. Combining both benign and malignant liver tumors for statistical purposes is not justified unless each type of tumor has been affected in a statistically significant incidence. If this were not the case, classification of the agent should be in accordance with the type of lesions induced significantly. In order to use the induction of hepatocellular carcinomas in rodents as an indication of carcinogenic risk in humans appropriately, one has to verify within the limits of feasibility that the identical carcinogenic effects and the subsequent events occur both in rodents and humans.(ABSTRACT TRUNCATED AT 400 WORDS)