The methods used to evaluate the toxicological effects of PCBs in animals have been reviewed. 2. The data show that Toxic Equivalency Factors (TEFs) could be developed to assess the potential toxicity of PCB mixtures for certain specific target organ effects (such as the liver and immune system) but would be inappropriate for other effects (e.g. thyroid function and neurochemical effects). More data on a wider range of individual PCB congeners and a method for systematically balancing toxicodynamic and toxicokinetic data are required before the TEF approach can be fully evaluated. 3. With the exception of the teratogenic effects seen in mice and the anti-oestrogenic effects reported in in vitro studies, there are insufficient data on individual PCB congeners to evaluate the structure-activity relationships for the effects of PCBs on reproduction. The data also show that individual PCBs may have opposing effects on a particular aspect of reproduction (for example individual PCB congeners may have either oestrogenic or anti-oestrogenic effects). Studies with individual PCB congeners have shown both enhancement and antagonism of the teratogenic effects of 2, 3, 7, 8-tetrachloro dibenzo-p-dioxin (TCDD) in the mouse. It is not possible to use TEFs to evaluate the reproductive effects of PCBs. 4. The mechanism(s) responsible for the effects of PCBs on postnatal neurobehavioural development in rodents and monkeys have not been elucidated. At least two groups of PCBs which might be responsible for the observed effects have been identified in this review, one affecting the dopaminergic system and the other group affecting thyroid hormone levels. Considerably more research would be required before the TEF approach could be applied to the effects of PCBs on postnatal neurobehavioural development. This would include research on an appropriate animal model to determine whether the critical toxicological mechanism is mediated through the Ah receptor. 5. The reproductive toxicity of complex PCB mixtures such as those found in foods will depend on the identifies and relative proportions of individual PCB congeners in the mixture. It is not possible to give an accurate estimate of a NOAEL or LOAEL from the reproduction studies using commercial PCB mixtures which could be readily applied to the safety assessment of PCBs present as contaminants in food. 6. It is concluded that the data presented in this paper support the hypothesis that there is no satisfactory method derived from the available studies in laboratory animals for evaluating the potential risk of adverse effects on reproduction posed by contamination of foods with PCBs.
