Objective: Despite the high rate of therapeutic failures in ventilator-associated pneumonia, up to now there has been no animal model specifically designed for antimicrobial evaluation. A rabbit model of ventilator-associated pneumonia is described for the first time in this study.
Design: Prospective, randomized experimental study.
Setting: An animal research laboratory.
Subjects: Male New Zealand healthy rabbits (n = 44).
Interventions: After oral intubation and an hour of mechanical ventilation, animals in the ventilator-associated pneumonia group (n = 22) were infected intrabronchially with a calibrated inoculum of Enterobacter aerogenes. The nonventilated pneumonia group (n = 22) was composed of animals that received the same inoculum in the absence of mechanical ventilation. Rabbits from both groups were randomly killed 3, 6, 12, 24, or 48 hrs after inoculation. Pneumonia evaluation was based on histologic (macroscopic and microscopic score) and bacteriologic (bacterial count) findings.
Main results: Infected animals undergoing mechanical ventilation rapidly developed a progressive bilateral and multifocal pneumonia. Lung bacterial mean (sd) concentration was 6.48 (0.71) log10 colony-forming units (cfu) per gram of tissue at the 48th hour, whereas bacteremia occurred in most cases. In the nonventilated pneumonia group, pneumonia was less severe in terms of bacterial count (3.18 [1.86] log10 cfu/g;p < .05), and spleen cultures remained negative. In addition, microscopic examination revealed noninfectious lung injury in the ventilator-associated pneumonia group, especially hyaline membrane filling alveolar spaces. Of note, these features were never observed in the nonventilated pneumonia group.
Conclusions: An animal model of ventilator-associated pneumonia was obtained in immunocompetent rabbits. Histopathologic and bacteriologic features were similar to those found in humans. Obviously, pneumonia was more severe when animals underwent mechanical ventilation, especially in terms of systemic spread. Noninfectious lung injury corresponding to ventilation-induced lung injury may explain the difference. This model emphasizes the strong impact of both mechanical ventilation and infection on lung because they seem to act synergistically when causing alveolar damage. Moreover, it seems well suited to testing antimicrobial effectiveness.