Assessment of learning, memory, and attention in developmental neurotoxicity regulatory studies: synthesis, commentary, and recommendations.

Cognitive tests of learning and memory (L&M) have been required by U.S. Environmental Protection Agency (EPA) developmental neurotoxicity test (DNT) guidelines for more than two decades. To evaluate the utility of these guidelines, the EPA reviewed 69 pesticide DNT studies. This review found that the DNT provided or could provide the point-of-departure for risk assessment by showing the Lowest Observable Adverse Effect Level (LOAEL) in 28 of these studies in relation to other reported end points. Among the behavioral tests, locomotor activity and auditory/acoustic startle provided the most LOAELs, and tests of cognitive function and the Functional Observational Battery (FOB) the fewest. Two issues arose from the review: (1) what is the relative utility of cognitive tests versus tests of unconditioned behavior, and (2) how might cognitive tests be improved? The EPA sponsored a symposium to address this. Bushnell reviewed studies in which both screening (locomotor activity, FOB, reflex ontogeny, etc.) and complex tests (those requiring training) were used within the same study; he found relatively little evidence that complex tests provided a LOAEL lower than screening tests (with exceptions). Levin reviewed reasons for including cognitive tests in regulatory studies and methods and evidence for the radial arm maze and its place in developmental neurotoxicity assessments. Driscoll and Strupp reviewed the value of serial reaction time operant methods for assessing executive function in developmental neurotoxicity studies. Vorhees and Williams reviewed the value of allocentric (spatial) and egocentric cognitive tests and presented methods for using the Morris water maze for spatial and the Cincinnati water maze for egocentric cognitive assessment. They also reviewed the possible use of water radial mazes. The relatively lower impact of cognitive tests in previous DNT studies in the face of the frequency of human complaints of chemical-induced cognitive dysfunction indicates that animal cognitive tests need improvement. The contributors to this symposium suggest that if the guidelines are updated, they be made more specific by recommending preferred tests and providing greater detail on key characteristics of such tests. Additionally, it is recommended that guidance be developed to address important issues with cognitive tests and to provide the information needed to improve the design, conduct, and interpretation of tests of higher function within a regulatory context. These steps will maximize the value of cognitive tests for use in hazard evaluation and risk assessment.