Association of sperm, vaginal cytology, and reproductive organ weight data with results of continuous breeding reproduction studies in Swiss (CD-1) mice.

In continuous breeding reproduction studies in which an adverse effect on fertility was detected over an 18-week treatment period, a crossover mating trial was then conducted to determine the affected sex. Results of 25 crossover breeding studies conducted using Swiss (CD-1) mice were compared with results of sperm morphology and vaginal cytology examinations (SMVCEs) conducted at the conclusion of the mating trial. SMVCE endpoints include sperm concentration, motility, and morphology, vaginal cytology, and male reproductive organ weights. In most SMVCE studies multiple endpoints were adversely affected. For male reproductive toxicants, sperm motility was decreased in 89% of the studies, and absolute right epididymis and right testis weights were affected less frequently (80% each). Among studies with no detectable reduction in male breeding performance, 87% exhibited no detectable decrease in epididymis weight. Eighty-two percent had no change in cauda epididymis weight and 80% had no significant change in sperm concentration. An increase in female cycle length was associated (100%) with an effect on breeding due to female dysfunction. Overall accuracy, defined as correct identification of toxicants and nontoxicants, was highest for epididymis weight (84%), followed by cauda epididymis weight and sperm motility (79% each), and sperm concentration (76%). Female cycle length was so variable that the overall accuracy of the parameter in 13 studies was 69%. With the variety of chemicals used in this analysis, the association of abnormal sperm morphology with reproductive outcome was 71%. Control data (mean, 95% confidence interval around the mean, median, and statistical sensitivity) for each male endpoint (parent, and offspring at 10 weeks of age following a single breeding) were summarized from each of the two laboratories that conducted the studies. For several endpoints, statistical power was dependent on the laboratory conducting the studies. In general, the statistical sensitivity was relatively high for reproductive organ weights, although it was less for smaller organs such as the prostate. On the basis of both the biological and statistical analyses, it is recommended that multiple SMVCE endpoints, including sperm measures, be included in screens for reproductive toxicants.