focused solely on adult animal models, it is important to determine whether in utero exposure affects cancer outcomes and, if so, by what mechanisms.
Conduct epidemiologic studies to evaluate potential health outcomes of phthalate-antiandrogen exposures. Attempt to characterize and evaluate effects in susceptible or highly exposed groups. Confirm and extend current information on the relationship between anogenital distance and infant testosterone concentration.
Conduct toxicity studies of phthalate metabolites to determine potential adverse effects associated with exposure to them.
As discussed in Chapter 5, available data support the appropriateness of cumulative risk assessment of phthalates and other antiandrogen compounds. Research initiatives that would refine such an assessment are outlined below.
Explore combination effects of phthalates, other antiandrogens, and other endocrine-disrupting agents.
Investigate deviations from additivity observed when hypospadias is used as the selected outcome.
Refine estimates of composite scores for disruption of androgen action.
Develop approaches to the epidemiologic assessment of the cumulative effects of phthalates and other antiandrogens.
The committee emphasizes that the quality of results of any risk assessment is based on the data available. The U.S. Environmental Protection Agency (EPA) Integrated Risk Information System (IRIS) is the source of much of the toxicity information used in risk assessment today. Many of the chemical profiles in IRIS need to be updated; the information is no longer relevant or accurate. The phthalate profiles available in IRIS illustrate that point. The committee recognizes that the task of profile review and revision is enormous; however, linking profiles to current literature would be helpful. For example, IRIS profiles of chemicals that also are the subject of interaction profiles produced by the Agency for Toxic Substances and Disease Registry would ideally be linked to the interaction profiles. Furthermore, as EPA moves toward cumulative risk assessment, some consideration should be given to restructuring IRIS so that its process for identifying chemicals for review includes and sets priorities among chemical mixtures, as appropriate, and facilitates cumulative risk assessment conducted by using common adverse outcomes. For example, listing the no-observed-adverse-effect levels or benchmark doses for a variety of effects would facilitate that approach.