In-vitro and in-vivo bacterocodal interactions of clindamycin and ceftazidime, by non-beta-lactamase mechanisms, in experimental Pseudomonas aeruginosa endocarditis caused by a constitutive beta-lactamase overproducing strain.

We studied a ceftazidime-resistant strain of Pseudomonas aeruginosa (PA-48) stably-derepressed for constitutive beta-lactamase overproduction, and its ceftazidime-susceptible parent (PA-96) in order to characterize the ability of clindamycin to: (1) enhance the in-vitro and in-vivo bactericidal activities of ceftazidime for PA-48; and (2) prevent beta-lactamase induction and spontaneous mutation to the derepressed state by the parental strain (PA-96). In vitro, clindamycin synergistically enhanced the bactericidal activity of ceftazidime vs PA-48. In the experimental aortic endocarditis model, the combination of clindamycin with ceftazidime significantly reduced mean intravegetation bacterial densities of PA-48 versus ceftazidime monotherapy and untreated controls at therapy days 6 and 11 (P less than 0.05). Exposure of growing PA-48 cells to clindamycin did not interfere with the hydrolytic function of extracted periplasmic beta-lactamase. Moreover, clindamycin did not suppress cefoxitin-mediated beta-lactamase induction in the parental strain (PA-96). In vitro, clindamycin prevented spontaneous mutation of PA-96 to the stably-derepressed state for beta-lactamase overproduction and also enhanced reversion of derepressed cells of PA-48 to the ceftazidime-susceptible parental phenotype by approximately 2 log10 cfu/ml. This latter effect was mirrored in vivo during clindamycin+ ceftazidime therapy of experimental endocarditis due to strain PA-48, as the proportion of ceftazidime-susceptible cells with vegetations increased by approximately 1-1.5 log10 cfu/g, versus untreated controls or ceftazidime monotherapy recipients. After clindamycin treatment ceased, vegetations contained predominantly ceftazidime-resistant variants. Clindamycin appeared to enhance bactericidal effects of ceftazidime vs PA-48 through non-beta-lactamase mechanisms probably involving promotion and maintenance of spontaneous reversion to the fully-repressed state. However, the concentrations of clindamycin required to achieve these effects are unlikely to be sustained at normal therapeutic dosage.