The staircase test: some evidence of nonspecificity for anxiolytics.

In the staircase test, a naive mouse is placed in a Plexiglas chamber containing a five-step staircase, and the number of rearings and steps climbed are recorded for 3 min. A claim for drug-class specificity has been made because conventional anxiolytics reduced rearings at doses that did not reduce steps climbed, while non-anxiolytics affected both measures in parallel. In the present study chlordiazepoxide, meprobamate, and ethanol registered the expected true positive effect by reducing rearings at doses that did not reduce steps climbed. Nicotine, which has some clinical anxiolytic action, registered a small true positive. The benzodiazepine anxiolytic alprazolam reduced both measures, a false negative, although it reduced rearings more than steps climbed. The putative novel anxiolytics CGS 9896, ketanserine, and tracazolate registered negatives, as did the known clinical anxiolytic buspirone. The non-anxiolytics phencyclidine and phenacetin registered true negatives, but morphine registered a clear false positive. The anxiogenics FG 7142 and pentylenetetrazol produced no significant effects. Because of the equivocal false negative for alprazolam, the clear false negative for buspirone, and the clear false positive for morphine, we concluded that the test lacks the degree of therapeutic-class specificity previously proposed but may still be useful in basic research.