A report on the workshop on biological and statistical implications of the ED01 study and related data bases

Summary The following is a summary of the consensus reached on various issues discussed at the Deer Creek State Park Workshop. The key conclusions agreed to by the biologists, statisticians and pathologists during the Workshop were: o 1) The determination of time-to-tumor is important in carcinogenicity studies. The ED 01 study demonstrates that the observed risk is more adequately expressed in a time and dose continuum rather than simply as a function of dose. 2) There is a need for time-to-tumor adjustments in the estimation of tumor rates, particularly in the presence of different mortality patterns due to intercurrent disease. This requires a determination of whether or not the animal died from the tumor of interest, since statistical methodology for fatal tumors differs from that for nonfatal tumors. Although there is disagreement as to whether the cause of an animal's death can be determined, it may be possible to establish whether a particular tumor caused death. 3) Two types of dose responses were observed in the ED 01 study in the two target organs for the test compound 2-AAF. The bladder tumors exhibited a no-observed effect level against a low background rate, while the liver tumors showed much less curvature in the dose-response relation. Older animals showed a high spontaneous rate of hepatic-cell tumors. 4) Even with a study as large as the ED 01 study, statistical uncertainty makes it impossible to establish the true shape of the dose-response curve at low tumor rates. Neither can such studies prove or disprove the existence of thresholds. 5) Although a definitive model for low-dose extrapolation is not currently known, upper limits on risk at low dose can be estimated by the use of linear extrapolation. However, this procedure may be unduly conservative when the actual dose-response curve exhibits thresholdlike behavior. 6) Development of better models for low-dose extrapolation can be aided by the incorporation of more biological information, such as pharmacokinetics, metabolism and comparitive physiology. 7) The ED 01 study is not the only large-scale, whole animal study of carcinogenesis. Other studies of possible interest for further investigation include radiation and skin-painting studies at Oak Ridge National Laboratories, formaldehyde studies by the Chemical Industries Institute of Toxicology, and the nitrosamine study conducted by the British Industrial Biological Research Association.