Screening for neurobehavioral toxicity: the need for and examples of validation of testing procedures.

The need for a sensitive and reliable screen to assess environmental agents for potential behavioral and neurological toxicity is discussed. Factors involving strategy, choice of animals and doses, route of administration, duration of study and requirements for the selection of neurobehavioral tests are also evaluated. The primary emphasis concerns the need for standardization and validation of neurobehavioral tests to be used in neurotoxicology. It is suggested that test validation be accomplished by comparing the observed results of known neurotoxicants in animal models which are chosen to predict effects based on reported human symptomatology. As a means of demonstrating how test validation is used in our laboratory, data from a number of experiments concerning the effects of a variety of chemical agents on three measures of motor functioning were discussed. The neurobehavioral effects of acrylamide, and agent known to produce "dying-back" axonopathies, were assessed using separate techniques presumed to measure hindlimb and forelimb functioning and general motor activity. The prediction that acrylamide will first decrease hindlimb functioning, while decreasing forelimb grip strength and motor activity at higher doses, was confirmed. The validity of the hindlimb measurement was supported using a neurotoxicant, carbon disulfide, known to affect motor functioning in a manner similar to acrylamide. The validity of the forelimb technique was shown indirectly using normative data collected from rats of both sexes tested at various ages, i.e., males were stronger than females and grip scores changed as a function of age. The relative sensitivities of the fore- and hindlimb measurements were found to be approximately the same when used to assess the effects of known muscle relaxants, such as phenobarbital and chlordiazepoxide. Finally, it was predicted and confirmed that an environmental agent believed to affect behavior secondarily to effects on other organ systems would affect all measures of motor functioning at approximately the same dose.