Estimation of Epidermal Carcinogenic Potency

Estimation of Epidermal Carcinogenic Potency. MCKEE, R. H.,NICOUCH, M. J., SCALA, R. A., AND LEWIS, S. C. (1990). Fundam.Appl. Toxicol. 15, 320–328. This report compared two statisticalmethods of estimating tumor latency, the Weibull distributionmodel and the Kaplan-Meier method. Parallelism of dose-responsecurves of different materials and quantitative reproducibilityof dermal carcinogenesis data were also examined. The Weibullmethod has the advantage of producing parameter estimates, evenwhen tumor yield is low. The Kaplan-Meier method, on the otherhand, is free of distribution assumptions. Overall, since thecomparisons of potency are made on the basis of parameters fromthe same assumed distribution on the same strain of animal,the Weibull estimates are favored. A comparison of dose-responsedata for benzo[a]pyrene and catalytically cracked clarifiedoil indicated that the slopes of the two dose-response curveswere significantly different. Thus the relative carcinogenicpotencies of different materials vary with dose, and potencycomparisons must necessarily be dose-specific. The quantitativereproducibility of dermal carcinogenesis bioassays was alsoassessed. The dose-response curves from the three studies ofone material had significantly different slopes. Thus the resultssuggested that there were sources of biological variabilitywhich could contribute to experimental error