A critique of animal models in antibiotic research.

The interactions between microbes and antimicrobial agents in vitro are often markedly different from those in vivo, presumably because of pharmacological factors and host defenses. Although animal models have long been used to evaluate the efficacy of chemotherapeutic agents, there are few guidelines for the conduct of these studies. Therefore, the results of these investigations must be assessed critically in terms of the balance between the infectious challenge and the therapeutic response, and in terms of their relevance to human disease. It has been clearly demonstrated that the "freshness" of the infecting organism, the interval between infection and therapy, the order of administration of drug combinations, and the presence of foreign bodies, may have a crucial impact upon the outcome of animal trials. Less well studied have been the influence of the quantity and growth phase of the infecting inoculum, the use of single- as opposed to multiple-dose therapy, and the anatomical and pharmacological attributes of the particular animal species. Greater understanding of these factors would resolve some of the apparently contradictory results of various investigations, and would permit more rational design of these studies in the future.