How to solve the problem of spontaneous bacterial clearance when testing new antibiotic treatment: results on experimental pneumonia due to a derepressed cephalosporinase-producing Enterobacter cloacae.

Because the magnitude of spontaneous bacterial clearance can be similar or even higher than treatment effect, depending upon experimental model and bacterial strain used, this work investigated the value of rendering rats immunosuppressed to facilitate bacterial implantation and reduce spontaneous bacterial clearance. In a first step, rats received a single intravenous cyclophosphamide dose 4 days before infection. Three different doses were tested: 10, 20, and 40 mg/kg. After modeling with NONMEM V, the cyclophosphamide dose required to maintain white blood cell count <1.0 × 10(3)/μL from day 4 to day 5 was 30 mg/kg. In a second step, influence of immunosuppression on lung bacterial titers was characterized. Rats were given one of the three intravenous cyclophosphamide doses (0, 10, 30 mg/kg), and after 4 days, they were infected by tracheal injection of 8.9 ± 0.1 log10 cfu Enterobacter cloacae before being sacrificed at different times. Bacteria in homogenized lungs were quantitatively cultured on Drigalski agar. Bacterial lung count was closely influenced by the grade of induced leukopenia. A single intravenous 30 mg/kg cyclophosphamide dose 4 days before infection suppressed the spontaneous clearance of E. cloacae for at least 30 h without significantly increasing animal mortality; this result seems to be linked to a white blood cell count maintained lower than 1.0 × 10(3)/μL for all the time. This modified animal model could be contributive in the evaluation of antibacterial agents, especially to simulate the behavior of intensive care unit immunocompromised patients.