This research has two main goals: (1) Develop and compare statistical methods for testing the carcinogenic potential of drugs and other chemical substances used by humans. (2) Investigate modified experimental designs to reduce the cost of drug development. To achieve these goals, several studies will be conducted. A nonparametric procedure will be developed for estimating distributions of time to onset or and time to death from occult tumors in the absence of cause-of-death information. A method for imputing the number of fatal tumors in an experiment that lacks cause-of-death data will be developed to modify the International Agency for Research on Cancer cause-of-death test. This procedure will estimate the lag time between onset of and death from an occult tumor, when cause-of-death data are unavailable. Using this procedure, the difference in the development of a certain disease between caloric restricted groups and ad libitum groups will be examined. The results of this study can be used for prediction of the human dietary effect. Software will be developed for a wide range of potential users. Using Monte Carlo simulations, numerical comparisons of power will be performed among different procedures under various conditions in long-term animal experiments. By comparing the IARC cause-of-death test with the trend tests that do not use cause-of-death information, the possible loss of efficiency in the absence of cause-of-death information will be examined. Most animal carcinogenicity studies are designed with a single terminal interface. In this project, age-adjusted trend tests for the tumor incidence rate for single-sacrifice experiments will be investigated. A method will be developed to identify the tumor incidence rate for time intervals preceding the final time interval. To reduce the cost of experiments, optimal designs with reduced sample size and shortened study duration will be investigated through a power study. A more rapid experiment will expedite the marketing of potential drugs. Therefore, finding an optimal design with a shortened study duration will be of great economic and medical value.
| Moon, Hojin; Ahn, Hongshik; Kodell, Ralph L (2005) An age-adjusted bootstrap-based Poly-k test. Stat Med 24:1233-44 |
| Mancuso, Jessica Y; Ahn, Hongshik; Chen, James J et al. (2002) Age-adjusted exact trend tests in the event of rare occurrences. Biometrics 58:403-12 |
| Mancuso, J Y; Ahn, H; Chen, J J (2001) Order-restricted dose-related trend tests. Stat Med 20:2305-18 |
