Related Experiment Videos
Evaluation of historical control data in carcinogenicity studies
H Greim1, H P Gelbke, U Reuter
1Institute of Toxicology and Environmental Hygiene, Technical University of Munich, D-85354 Freising-Weihenstephan, Germany. helmut.greim@lrz.tum.de
Human & Experimental Toxicology
|December 6, 2003
Summary
Statistical analysis of animal carcinogenicity studies requires careful use of historical control data. This paper outlines criteria for valid historical control data comparison to ensure reliable study results.
Area of Science:
- Toxicology
- Carcinogenicity Studies
- Biostatistics
Background:
- Carcinogenicity studies in animals are crucial for assessing chemical health hazards.
- Concurrent control groups are standard, but historical controls help account for spontaneous tumor variations.
- Variability in historical control data necessitates strict evaluation criteria.
Purpose of the Study:
- To discuss sources of variability in spontaneous tumor incidences in untreated animals.
- To present databanks for historical control data.
- To detail criteria for the valid application of historical control data in carcinogenicity study evaluations.
Main Methods:
- Statistical comparison of treated groups with concurrent and historical control groups.
- Identification and analysis of factors influencing tumor incidences in untreated animals.
- Review of biostatistical data evaluation methods for carcinogenicity studies.
Main Results:
- Historical control data can be valuable but require rigorous validation.
- Key criteria for using historical data include same species/strain, breeder, lab, study design, methods, assessment criteria, and a limited time window.
- Unpublished data may be used if criteria are met and data are sufficiently detailed.
Conclusions:
- Proper application of historical control data enhances the reliability of carcinogenicity study interpretations.
- Adherence to established criteria ensures the scientific validity of using historical control data.
- The DFG Commission's criteria provide a framework for evaluating the usability of historical control data.