Correlation of in vitro chemopreventive efficacy data from the human epidermal cell assay with animal efficacy data and clinical trial plasma levels

Abstract

The human epidermal cell (HEC) assay, which uses carcinogen exposed normal skin keratinocytes to screen for cancer prevention efficacy, was used to screen possible preventive agents. The endpoints measured were inhibition of carcinogen-induced growth and induction of involucrin, an early marker of differentiation. Sixteen of twenty agents (apigenin, apomine, budesonide, N-(2-carboxyphenyl)retinamide, ellagic acid, ibuprofen, indomethacin, melatonin, (-)-2-oxo-4-thiazolidine carboxylic acid, polyphenon E, resveratrol, beta-sitosterol, sulfasalazine, vitamin E acetate, and zileuton) were positive in at least one of the two assay endpoints. Four agents (4-methoxyphenol, naringenin, palmitoylcarnitine chloride, and silymarin) were negative in the assay. Nine of the sixteen agents were positive for both endpoints. Agents that showed the greatest response included: ellagic acid > budesonide, ibuprofen > apigenin, and quinicrine dihydrochloride. Fifty-eight of sixty-five agents that have been evaluated in the HEC assay have also been evaluated in one or more rodent bioassays for cancer prevention and several are in clinical trials for cancer prevention. The assay has an overall predictive accuracy of approximately 91.4% for efficacy in rodent cancer prevention irrespective of the species used, the tissue model, or the carcinogen used. Comparison of the efficacious concentrations in vitro to plasma levels in clinical trials show that concentrations that produced efficacy in the HEC assay were achieved in clinical studies for 31 of 33 agents for which plasma levels and/or C(max) levels were available. For two agents, 9-cis-retinoic acid (RA) and dehydroepiandrosterone (DHEA), the plasma levels greatly exceeded the highest concentration (HC) found to have efficacy in vitro. Thus, the HEC assay has an excellent predictive potential for animal efficacy and is responsive at clinically achievable concentrations.