Animal models as predictors of the safety and efficacy of antibiotics.

As opposed to the testing of safety, the testing of the efficacy of antibiotics in animals is not specified in any directives or guidelines and not explicitly required by regulatory authorities. There exists, however, no doubt that in the evaluation of new compounds testing of both safety and efficacy forms an essential link between in vitro tests and clinical trials. It is inconceivable that clinicians would be prepared to conduct a trial in patients without evidence of the efficacy of the antibiotic in question in an appropriate animal model of infection. Both the models for testing safety and those for testing efficacy suffer from a number of shortcomings. If investigators are aware of these deficiencies and take them into account when interpreting the results, the predictive value of the models can be significantly enhanced.