Leukocyte and bacterial interrelationships in experimental meningitis.

Eighty-one rabbits were inoculated with known concentrations of type III pneumococci by cisternal puncture and then started on antibiotic therapy the following day. Aliquots of cerebrospinal fluid were sampled at regular intervals both before and after therapy and then analyzed for bacterial titer and leukocyte count. These data were used to examine the interrelationships of inoculum size, leukocyte count, and bacterial titer to each other and their effects both univariately and multivariately on outcome. Inoculum size was the most important single variable affecting outcome, followed in order by the pretherapeutic bacterial titer and leukocyte count. Higher leukocyte counts early in the course of the disease prior to therapy were associated with a favorable outcome. Continued leukocyte elevation after therapy was associated with a poor prognosis. No correlations were found between inoculum size and subsequent bacterial titers; however, the bacterial titer immediately prior to therapy correlated negatively with the level of the early white cell response. The best prognostic information was obtained from the multivariate analysis of inoculum size and pretherapeutic white cell response.