Rodent 2-year cancer bioassays and in vitro and in vivo genotoxicity tests insufficiently predict risk or model development of human carcinomas

Abstract

In 2018, the National Institutes of Health estimated that 1,735,350 new cancer cases were diagnosed in the United States and 609,640 patients died from various forms of cancer. The vast majority of those cancer deaths occurred from cancers derived from epithelial cells, that is, carcinomas. The most lethal cancers in American women based upon estimated deaths occurred in the lung and bronchus (352,209), breast (205,675), colon and rectum (123,617), pancreas (96,571), and ovary (71,141). The largest cancer killers in men occurred in the lung and bronchus (427,587), prostate (140,086), colon and rectum (135,542), pancreas (100,599), and liver (80,526). Histologically different carcinomas are initiated, clonally expanded, and progressed in association with particular changes in genes and also epigenetic alterations. Currently employed genotoxicity testing protocols emphasize testing for the initiating (mutational) potential of the test agent. While 2-year chronic rodent cancer bioassays test for the entire spectrum of carcinogenic transformation and development, the high doses used in these assays induce cytotoxicity leading to increased cellular proliferation rates and high false-positive rates of tumor induction in non-genotoxic chemicals. The low cancer induction from high radiation exposures experienced by atomic bomb survivors in Hiroshima and Nagasaki, Japan, and the epidemiological evidence showing that cigarette smoking duration and not intensity is associated with lung cancer risk both support a more important role for tumor promotion rather than initiation in the clinical presentation of human carcinomas. Cancer hazard assessment testing protocols and weight-of-the-evidence analysis of agent-specific cancer risk should be better aligned with the pathogenesis of human carcinoma.