Reproduction and teratogenesis: current methods and suggested improvements.

The risks to human reproduction, particularly in regard to possible induction of birth defects, from present levels of environmental chemicals are thought to be low but in need of closer scrutiny than is now the case. Methods currently used to estimate the hazards to reproduction of new chemicals and drugs are reviewed and some shortcomings and potential sources of error are noted. Specific suggestions are made for improving existing methodology. In addition to the widely used treatment span encompassing all or most of the period of organogenesis, it is proposed that some exposure periods of shorter duration be added to avert possibly adaptative changes in maternal homeostatic systems. Greater diversity of test animals, beyond the rodent-rabbit species largely used today, is urged with the objective of better matching metabolic pathways in a test animal to those of man. The demonstration of an embryotoxicity effect-level must be regarded as essential in all tests, and once found this level should be used as the basis of extrapolating downward to find an acceptable tolerance level of dosage. More attention should be focused on postnatal functional evaluation after prenatal and perinatal exposure of test animals to environmental chemicals, especially those that might interfere with the maturation of the central nervous system, endocrine glands, and immunological mechanisms. The use of nonhuman primates should be reserved for the testing of drugs essential for use during human pregnancy or early pregnancy, because of the scarcity of such animals and the high costs of doing meaningful tests with them.