Pre-Marketing Sources of Data Regarding Reproductive and Developmental Safety of Prenatal Drug Exposures

PRE-MARKETING SOURCES OF DATA REGARDING REPRODUCTIVE AND DEVELOPMENTAL SAFETY OF PRENATAL DRUG EXPOSURES

The traditional methods for evaluating drug safety in the pre-marketing phases of drug development, i.e. animal reproductive and developmental toxicity studies and randomized clinical trials, have limited application with respect to human pregnancy.

Reproductive and developmental toxicity studies conducted in selected animal species provide the first source of information about potential human risks for a variety of pregnancy outcomes. Results of these experiments are considered in the context of existing knowledge about the reproductive or developmental effects of similar chemical entities and the presence or absence of any theoretical concerns due to the drug’s mechanisms of action or pharmacologic properties. On the basis of this overall evaluation, a new drug can be marketed with reassurances that the animal data do not raise concerns about human pregnancy exposure, or conversely, with the recommendation that until human data are available, pregnancy should be avoided (Moore et al., 1995). However, there can be differences in the sensitivity and human compara-bility of the various animal species that are selected for toxicity testing; there may be differences in the dose, route of administration and metabolism in the animal model relative to usual human clinical use; and maternal toxic effects in the test species may play a role. For these and other reasons, there are limitations to the predictive value of these pre-clinical studies for human pregnancy exposures and outcomes (Brent, 1986; Scialli et al., 2004). Thus, human pregnancy data are ultimately necessary to establish human preg-nancy drug safety.

Clinical trials are the second traditional method of evaluating drug safety. For obvious ethical reasons, pregnant women typically are not recruited for trials during any phase of drug development. If and when unintended pregnancies occur during the course of a trial or post-marketing study, pregnancy outcomes can provide useful preliminary information regard-ing the risks of exposure (O’Quinn et al., 1999). However, these data usually involve a small number of subjects. There is a trend to include larger numbers of women of childbearing age in clinical trials, and this will undoubtedly result in a larger number of exposed pregnancies in such trials. Nevertheless, these numbers are likely to be too small to provide mean-ingful information.