USE OF HISTORICAL CONTROLS IN TIME-ADJUSTED TREND TESTS FOR CARCINOGENICITY

We develop a method for incorporating historical control information i nto time-adjusted tests for dose effects in carcinogenicity studies. A fter discretizing the time scaler we use a multinomial distribution to model the number of animals dying with tumor in each interval and spe cify a Dirichlet prior distribution on the control tumor death rates i n each interval. Data ti om past studies are used to estimate the para meters characterizing the prior. A score test derived from the resulti ng Dirichlet-multinomial generalizes the test of Tarone (1982, Biometr ics 38, 215-220) and reduces, in the limit, to the log-rank test in th e case of a diffuse prior. The methodology is illustrated with data fr om a study of the fire retardant 2,2-Bis(bromomethyl)-1,3-propanediol.