LB11058, a New Cephalosporin with High Penicillin-Binding Protein 2a Affinity and Activity in Experimental Endocarditis Due to Homogeneously Methicillin-Resistant Staphylococcus aureus

Microorganisms and growth conditions.Three isolates expressing homogeneous resistance to methicillin and extensively characterized in experimental endocarditis were used (10, 15, 30): (i) the penicillinase-negative reference MRSA strain COL, (ii) a penicillinase-producing transformant of strain COL (strain COL-Bla+) carrying the penicillinase-encoding plasmid pI524 (30), and (iii) the penicillinase-producing homogeneously methicillin-resistant strain P8-Hom, selected by serial passage of parent strain P8 on methicillin-containing medium (10). Unless stated otherwise, the bacteria were grown at 35В°C in tryptic soy broth (Difco Laboratories, Detroit, Mich.) with aeration or on tryptic soy agar (Difco) supplemented with 2% of NaCl (for methicillin). Bacterial stocks were kept at в€’70В°C in tryptic soy broth supplemented with 10% (vol/vol) glycerol.

Antibiotics and reagents.LB11058 was provided by LG Life Sciences, Ltd., Daejeon, Korea). Methicillin, flucloxacillin, amoxicillin, clavulanic acid, and amoxicillin-clavulanate (5:1 [wt/wt] ratio) were purchased from GlaxoSmithKline AG (Thoerishaus, Switzerland). Vancomycin was purchased from Eli Lilly (Vernier, Switzerland). All other chemicals were reagent-grade commercially available products.

Susceptibility testing, population analysis profiles, and time-kill curves.The MICs of the drugs for the tests bacteria were determined by a previously described broth macrodilution method (27) with a final inoculum of 10⁵ to 10⁶ CFU/ml. For MIC determinations of the combination of amoxicillin-clavulanate, the drugs were used in a ratio of 2:1 (wt/wt). MBCs were measured by subculturing 0.01-ml samples from each MIC tube showing no turbidity onto agar plates, followed by another 48 h of incubation at 35В°C. The MBC was defined as the lowest antibiotic concentration resulting in ≥99.9% killing of the original inoculum. The phenotypic expression of beta-lactam resistance was evaluated by population analysis profiles. Large bacterial inocula (10⁹ CFU), as well as smaller inocula (10⁶, 10⁵, and 10³ CFU) were spread onto tryptic soy agar plates containing twofold serial dilutions of antibiotics (15, 30). The numbers of colonies growing on the plates were determined after 48 h of incubation at 35В°C. Population analysis profile curves were generated by plotting the number of CFU growing on the plates against the concentrations of antibiotic in the plates.

For time-kill curves, series of flasks containing fresh prewarmed medium were inoculated with ca. 10⁶ CFU/ml (final concentration) from an overnight culture of bacteria and further incubated at 35В°C. Immediately after inoculation, antibiotics were added to the flasks at final concentrations approximating either steady-state levels of free drug (for LB11058) or peak levels (for vancomycin and ceftriaxone) produced in the serum in humans after treatment with the selected standard doses and mimicked in our rat model (see Results). These concentrations were 5 mg/liter for LB11058, 40 mg/liter for vancomycin, and 200 mg/liter for ceftriaxone.

Samples were removed from the flasks just before and at various times after antibiotic addition, serially diluted, and plated onto blood agar for colony counts. Antibiotic carryover was minimized both by appropriate sample dilution for all of the drugs and for beta-lactams by supplementation of the plates with 0.2% (vol/vol) of Bacillus cereus 569/H9 beta-lactamase (penicillin amino-beta-lactam hydrolase; Genzyme Diagnostics, Kent, England). Each time-kill experiment was performed on two independent occasions.

Production of endocarditis and installation of the infusion pump.Sterile aortic vegetations were produced in rats as previously described (19). An intravenous (i.v.) line was inserted via the jugular vein into the superior vena cava and connected to a programmable infusion pump (Pump 44; Harvard Apparatus, Inc., South Natick, Mass.) to deliver the antibiotics (14). The pump was set to release a volume of 0.1 ml of saline per h to keep the catheter patent until the onset of therapy. No i.v. lines were placed in the control animals. This did not affect the incidence of infection.

Bacterial endocarditis was induced 24 h after catheterization by i.v. challenge of the animals with 0.5 ml of saline containing 10⁵ CFU of either strain COL-Bla+ or strain P8-Hom. This inoculum size was 10 times larger than the minimum inoculum producing endocarditis in 90% of untreated rats irrespective of the infectious organisms.

Antibiotic treatment of experimental endocarditis.Therapy was started 12 h after inoculation and lasted 5 days. Antibiotics were administered at changing flow rates in order to produce either constant drug concentrations in serum of 5, 10, and 250 mg of LB11058/liter or to simulate drug kinetics in human serum during treatment with 1 g of vancomycin i.v. every 12 h (13) or 2 g of ceftriaxone i.v. every 24 h (8). This required total amounts of antibiotics (in milligrams per kilogram of body weight) of 14.3 mg, 30.1 mg, and 1.584 g of LB11058 per 24 h. Note that the concentrations of LB11058 in the serum and the amounts of drug delivered were not proportional. This was most likely due to a high level of LB11058 protein-binding, which became saturable over 250/liter (see Results). Above this threshold, free drug was likely to become distributed or secreted, thus necessitating more compound to maintain the steady-state concentration. Amounts for other drugs were 23.2 mg of vancomycin per 12 h and 1.06 g of ceftriaxone per 24 h (18). Before administration, LB11058 was diluted out to concentrations of up to 19.8 mg/ml in a 0.051 N NaOH solution containing 15% of PEG400. LB11058 was stable under these conditions and lost <5% of activity per 6 h at room temperature (H. Youn, LG Life Sciences [unpublished data]).

Pharmacokinetic and pharmacodynamic studies.Concentrations of antibiotic in serum were determined on day 2 of therapy in groups of three to six uninfected or infected rats per experiments. Drug levels in the infected animals were derived from the internal controls of therapeutic experiments, in which adequate drug delivery was tested routinely. Blood was drawn by puncturing the periorbital sinuses of the animals (one puncture per animal) at several time points during and after antibiotic administration. Antibiotic concentrations were determined by an agar diffusion assay with antibiotic medium 1 (Difco) by using Bacillus subtilis ATCC 6633 as the indicator organism for LB11058 and vancomycin and Escherichia coli ATCC 25922 as the indicator organism for ceftriaxone. The diluent was pooled rat serum. The limits of detection of the assays were 2 mg/liter for LB11058, 0.6 mg/liter for vancomycin, and 3 mg/liter for ceftriaxone. The linearity of the standard curves was assessed by a regression coefficient of ≥0.996, and the intraplate and interplate variations were ≤10%.

In certain experiments, antibiotic binding to rat and human serum proteins was determined by using an ultrafiltration technique reported earlier (5, 12). LB11058 was diluted at increasing concentrations in either in 5% glucose or pooled (decomplemented) rat or human serum, let stand for 30 min at room temperature, and filtered by using the Amicon Centrifree System (Amicon, Inc., Beverly, Mass.). The drug concentrations in the original samples and in the ultrafiltrates were determined as described above. Of note, ca. 9% (at the concentrations tested) of LB11058 remained bound to the Centrifree cones. These values were subtracted to total drug concentrations for the calculation of protein binding. Serum binding of vancomycin (40%) and ceftriaxone (80%) were extrapolated from previous literature (18, 21).

Evaluation of infection.Infected control rats were killed at the time of treatment onset, i.e., 12 h after inoculation, in order to determine both the frequency and the severity of valve infection at the start of therapy. Treated rats that completed therapy were sacrificed 12 h after the trough level of the last antibiotic dose of any of the test drugs was achieved when no residual antibiotic could be detected in the blood. Vegetations, blood, and spleens were cultured as previously described (10, 13, 30). Few animals died before or within the first 24 h of therapy. Only rats that received a minimum of 24 h of the treatment were subjected to vegetation bacterial counts. Bacterial densities in valves and vegetations were expressed as log₁₀ CFU per gram of tissue. The minimum detection level was ≥2 log₁₀ CFU/g of vegetation. For statistical comparisons, culture-negative vegetations were considered to contain 2 log₁₀ CFU/g.

Statistical analysis.The incidences of valve infection were compared by the Fisher exact test. The Mann-Whitney rank sum test for comparison of differences between residual bacterial titers in the vegetations was used in certain analyses. Differences were considered significant when the P value was <0.05 as determined by using two-tailed significance levels.

Pharmacokinetics, pharmacodynamics, and drug delivery.LB11058 binding to rat and human serum proteins was determined by ultrafiltration by using the Amicon Centrifree System. LB11058 was shown to be highly bound (96 to 98%) to serum proteins in rats (H. Y. Joo, J. E., Shin, L. H. Choi, D. H. Park, S. H. Kim, S. H. Lee, and H. Youn, Abstr. 42nd Intersci. Conf. Antimicrob. Agents Chemother., abstr. F-331, p. 174, 2002). Hence, it was important to set up a drug delivery system providing enough total drug in serum and tissues to ensure a minimum free fraction equal to or greater than the MIC for the test microorganisms. Figure 3 depicts the profile of free LB11058 when diluted into pooled rat or human serum in vitro. It can be seen that the drug was highly bound up to concentrations of 250 mg/liter for rats and 200 mg/liter for humans. However, binding became saturable and free drug substantially increased above these cutoff concentrations. By comparison, vancomycin was bound in rat serum at 40% (21), and ceftriaxone was bound in rat serum at 80% (18). Thus, we inferred that, based on an MIC of 1 mg of LB11058/liter for MRSA, the concentrations of LB11058 in the serum of rats should be close to 250 mg/liter or more in order to be effective.

• Open in new tab
• Download powerpoint

FIG. 3.

Free fraction of LB11058 in rat and human sera as determined by ultrafiltration, followed by antibiotic activity measurement by bioassay (5, 12). Pooled rat serum (▪) and human serum (▵) were spiked with increasing concentrations of LB11058 and let stand at room temperature, and then the free fractions were recovered by ultrafiltration as described previously (5, 12). The percentage of free over total fraction is plotted on the graph. Values of <2 mg/liter could not be accurately determined due to the limit of detection of the dosing method (see Materials and Methods). Control experiments with amoxicillin in rat sera disclosed a free fraction of ≥80% at any of the tested concentrations (not depicted on the graph). Results are the means of ≥2 individual experiments with a variation of <15%.

To test this hypothesis in rats, three treatment protocols were used. In the first protocol the drug was administered to produce a constant concentration of 5 mg/liter of LB11058 in the serum (expected free fraction of 0.1 to 0.2 mg/liter; 0% time over the MIC). In the second protocol we targeted a constant concentration of 10 mg/liter (expected free fraction of 0.2 to 0.4 mg/liter; 0% time over the MIC). In the third protocol we targeted a constant concentration of ca. 250 mg/liter (expected free fraction of 5 to 10 mg/liter; 100% time over the MIC). Positive and negative controls with vancomycin and ceftriaxone ensured 100 and 0% time over the MIC, respectively.

Treatment of experimental endocarditis.Table 2 indicates that low dosages to LB11058 failed to cure the infected animals. This was presumably due to insufficient amounts of free drug at the sites of infection. However, after increasing drug dosage to ensure free serum concentrations of 5 to 10 times the MIC, LB11058 proved highly effective. Control treatment with vancomycin was also effective but resulted in a greater number of treatment failures. Few negative controls treated with ceftriaxone failed to respond to therapy (Table 2).

The overall mortality ranged between 10 and 20% in all groups. It occurred within first 24 h of therapy and was most likely due to postsurgery complications. Mortality tended to be lower in the high-dose LB11058 group, but this was not statistically significant. Spleens were infected (10 to >1,000 CFU/g of tissue) in all rats with positive valve cultures, except for the single treatment failure in the high-dose LB11058 group infected with MRSA P8. This rat had a positive valve culture but a negative spleen culture. Thus, LB11058 dosage producing free drug concentrations as low as 5 to 10 times the MIC proved highly bactericidal against experimental endocarditis due to homogeneously methicillin-resistant MRSA.