TEXT

Top Abstract Text References

Trovafloxacin is a new quinolone with significant activity against enterococcal species (2, 3). In studies of Enterococcus spp. we found bactericidal antagonism when trovafloxacin was combined with rifampin (10) and considered that it was important to examine this combination in an in vivo model. Pyelonephritis caused by Enterococcus faecalis GK, a clinical isolate, in the rat is a reproducible model of systemic enterococcal infection (4, 9). Trovafloxacin was obtained from Pfizer Inc., Groton, Conn. The standard powder of trovafloxacin was used for the in vitro studies, and a preparation for intravenous injection (5 mg/ml) was used for intramuscular injection into the rats. Rifampin was obtained from Hoechst Marion Roussel, Cincinnati, Ohio.

Pyelonephritis was produced in 100 randomly bred female Sprague-Dawley rats (weight, 120 to 140 g) by intravenous injection of 1 ml of an overnight culture of E. faecalis GK in brain heart infusion (BHI) broth (Difco) containing 1.3 × 10⁹ CFU per ml. Seven days after infection, before starting therapy, 10 rats were killed to confirm the presence of infection in the kidneys. The kidneys were removed, weighed, and homogenized. Serial 10-fold dilutions were made in BHI broth. Aliquots were placed in molten BHI agar, and colony counts were determined after 48 h of incubation. The lower limit of detection of bacteria in the kidneys was 10 CFU/g of kidney. The rats were randomly divided into four groups of 20 rats each defined by the antibiotics that they were to receive: group 1, control group (no antibiotics); group 2, trovafloxacin at 10 mg/kg of body weight; group 3, rifampin at 10 mg/kg; and group 4, trovafloxacin and rifampin at the same doses but injected at different sites. The antibiotics were injected intramuscularly in the hind limb twice daily (alternative hind limbs were used for the two daily injections). The last antibiotic injection was given at least 18 h before harvest. The numbers of bacteria in the kidneys of the rats were quantitated after 1 and 4 weeks of treatment with antibiotics.

Antibiotic levels in the serum and urine of rats that were not infected were determined. After 2 days of therapy, four rats from each group receiving antibiotics were killed at 1 and 16 h after administration. Urine was aspirated from the bladder, and heart blood samples were collected at the time of killing. The levels of trovafloxacin and rifampin in serum were measured by the disk diffusion method (1). The levels of trovafloxacin in urine were determined by high-pressure liquid chromatography (11) in the laboratory of the Central Research Division of Pfizer Inc.

The MICs and minimal bactericidal concentrations of the antibiotics for E. faecalis GK determined by the macrotube dilution technique (8) were as follows: trovafloxacin, 0.25 and 0.5 µg/ml, respectively; rifampin, 2 and >16 µg/ml, respectively.

In the first week after treatment started there were five deaths in groups 1 and 2 and three deaths in groups 3 and 4. There were no significant differences in the numbers of deaths between the groups. Postmortem examination of these animals showed macroscopic abscesses in the kidneys, and no further studies were done with these kidneys. At 4 weeks the mean log₁₀ CFU of E. faecalis per gram of kidney from each of the groups treated with antibiotics was significantly less than the mean log₁₀ CFU of bacteria per gram of kidney from the control group (Table 1). The value for the group treated with the combination was lower than the value for the groups treated with either drug alone, but the differences were not statistically significant. When the results obtained after 1 and 4 weeks of therapy were combined, there was no detectable infection in 7 of 34 kidneys in animals in group 4, 6 of 34 kidneys in animals in group 3, and 3 of 30 kidneys in animals in group 2. All but 1 of 28 kidneys in animals in the control group (group 1) had detectable kidney infection. The levels of trovafloxacin and rifampin in serum presented in Table 2 approximated those observed clinically in humans. However, the reported half-life of trovafloxacin for rats (0.9 and 2.2 h) and the low renal clearance (11) with low levels in urine may account for the relative ineffectiveness of trovafloxacin as a single agent in this model of enterococcal infection.

Despite the antagonism of the two drugs in vitro, there was no evidence of antagonism when the two drugs were used in combination to treat the rat model of pyelonephritis caused by E. faecalis GK. The in vitro antagonism of trovafloxacin and rifampin may be compared with the in vitro evidence of antagonism of ciprofloxacin and rifampin that was described with Staphylococcus aureus. Despite the in vitro results, the use of rifampin in combination with fluoroquinolones has been shown to be of benefit for staphylococcal osteomyelitis in animals (5).

Although fluoroquinolones and rifampin have been used in combination to treat clinical infections caused by Enterococcus spp. (6, 7), few studies have supported this practice. In the present study, treatment of enterococcal pyelonephritis with trovafloxacin in combination with rifampin was not as effective as treatment with the combination of ampicillin and an aminoglycoside reported previously (9). In those studies with E. faecalis GK, 4 weeks of therapy with ampicillin in combination with streptomycin resulted in a mean log₁₀ CFU of <1.0 per g of kidney, with only 2 of 20 kidneys being infected. For the treatment of severe infections caused by susceptible strains of Enterococcus spp., the antibiotics of choice remain the combination of ampicillin or vancomycin and an aminoglycoside.

Our results suggest that despite evidence of in vitro antagonism, trovafloxacin and rifampin may be a useful combination in vivo. These results support further studies of the clinical use of the combination of trovafloxacin and rifampin for the treatment of enterococcal infections.