Abstract
OBJECTIVE: The Organisation for Economic Co-operation and Development (OECD) has completed phase 2 of an international program to validate the rodent Hershberger bioassay. DESIGN: The Hershberger bioassay is designed to identify suspected androgens and antiandrogens based on changes in the weights of five androgen-responsive tissues (ventral prostate, paired seminal vesicles and coagulating glands, the levator ani and bulbocavernosus muscles, the glans penis, and paired Cowper's or bulbourethral glands). Protocol sensitivity and reproducibility were tested using two androgen agonists (17alpha-methyl testosterone and 17beta-trenbolone), four antagonists [procymi-done, vinclozolin, linuron, and 1,1-dichoro-2,2-bis-(p-chlorophenyl)ethylene (p,p'-DDE)], and a 5alpha-reductase inhibitor (finasteride). Sixteen laboratories from seven countries participated in phase 2. RESULTS: In 40 of 41 studies, the laboratories successfully detected substance-related weight changes in one or more tissues. The one exception was with the weakest antiandrogen, linuron, in a laboratory with reduced sensitivity because of high coefficients of variation in all tissue weights. The protocols performed well under different experimental conditions (e.g., strain, diet, housing protocol, bedding, vehicle). There was good agreement and reproducibility among laboratories with regard to the lowest dose inducing significant effects on tissue weights. CONCLUSIONS: The results show that the OECD Hershberger bioassay protocol is reproducible and transferable across laboratories with androgen agonists, weak androgen antagonists, and a 5alpha-reductase inhibitor. The next validation phase will employ coded test substances, including positive substances and negative substances having no androgenic or antiandrogenic activity.