Garenoxacin Treatment of Experimental Endocarditis Caused by Viridans Group Streptococci
http://www.100md.com
《抗菌试剂及化学方法》
1.Division of Infectious Diseases, Department of Medicine,2.Division of Clinical Microbiology, Department of Laboratory Medicine and Pathology, Mayo Clinic College of Medicine, Rochester, Minnesota
ABSTRACT
The activity of garenoxacin was compared to that of levofloxacin or penicillin in a rabbit model of Streptococcus mitis group (penicillin MIC, 0.125 μg/ml) and Streptococcus sanguinis group (penicillin MIC, 0.25 μg/ml) endocarditis. Garenoxacin and levofloxacin had MICs of 0.125 and 0.5 μg/ml, respectively, for both study isolates. Rabbits with catheter-induced aortic valve endocarditis were given no treatment, penicillin at 1.2 x 106 IU/8 h intramuscularly, garenoxacin at 20 mg/kg of body weight/12 h intravenously, or levofloxacin at 40 mg/kg/12 h intravenously. For both isolates tested, garenoxacin area under the curve (AUC)/MIC and maximum concentration of drug in serum (Cmax)/MIC ratios were 368 and 91, respectively. Rabbits were sacrificed after 3 days of treatment; cardiac valve vegetations were aseptically removed and quantitatively cultured. For S. mitis group experimental endocarditis, all studied antimicrobial agents were more active than no treatment (P < 0.001), whereas for S. sanguinis group endocarditis, no studied antimicrobial agents were more active than no treatment. We conclude that AUC/MIC and Cmax/MIC ratios may not predict activity of some quinolones in experimental viridans group endocarditis and that garenoxacin and levofloxacin may not be ideal choices for serious infections caused by some quinolone-susceptible viridans group streptococci.
INTRODUCTION
Viridans group streptococci are commensal bacterial flora that are common causes of native valve endocarditis. Viridans group streptococcal endocarditis is a severe infection associated with high morbidity and mortality. Its treatment includes administration of a prolonged course of a bactericidal antimicrobial regimen, typically administered intravenously (1). The long-term care and costs associated with such therapy and the emergence of viridans group streptococci relatively resistant to beta-lactams (8, 25) have led to a search for new antimicrobials for the treatment (and prophylaxis) of viridans group streptococcal endocarditis. Combination therapy including an aminoglycoside has been advocated by the American Heart Association to treat penicillin-intermediate and -resistant viridans group streptococcal strains causing infective endocarditis (1) and to shorten therapy for endocarditis caused by penicillin-susceptible viridans group streptococcal strains (26). However, aminoglycoside side effects limit the use of such approaches (16).
Quinolones are generally considered bactericidal antimicrobial agents with high oral bioavailability and good penetration into valvular vegetations (12), making them theoretically suitable for treatment of infective endocarditis. Quinolones in combination with rifampin have been used to treat right-sided Staphylococcus aureus endocarditis (10). However, in the case of viridans group streptococci, quinolones have, in some cases, failed as treatments for experimental endocarditis (4, 22), and emergence of quinolone-resistant viridans group streptococci in bone marrow transplant patients receiving prophylaxis with levofloxacin has been reported (24). Garenoxacin is a new quinolone with enhanced potency against gram-positive bacteria, including penicillin-resistant Streptococcus spp. and Staphylococcus spp. (3, 17), that could potentially address the above-mentioned limitations of quinolones.
Experimental infective endocarditis is a rigorous test of bactericidal activity of antimicrobial agents (9). We designed a rabbit model of infective endocarditis to determine the in vivo activity of garenoxacin in the treatment of viridans group streptococcal endocarditis in comparison to penicillin and levofloxacin.
(This work was presented in part at the 43rd [14 to 17 September 2003, Chicago, Ill.] and 44th [30 October to 2 November 2004, Washington, D.C.] Interscience Conference on Antimicrobial Agents and Chemotherapy.)
MATERIALS AND METHODS
Microorganisms. Thirty viridans group streptococci (nine Streptococcus sanguinis group isolates, nine Streptococcus mitis group isolates, seven Streptococcus mutans group isolates, three Streptococcus salivarius group isolates, and two S. anginosus group isolates) were studied in vitro; all isolates were associated with human infective endocarditis. Organisms were reidentified (23) and classified according to Facklam criteria (15). Two isolates were selected for in vivo studies: IDRL-4600, a penicillin-susceptible S. mitis group isolate, and IDRL-4954, a penicillin-intermediate S. sanguinis group isolate. Both were ciprofloxacin susceptible according to Clinical and Laboratory Standard Institute (CLSI) interpretive standards (7), and both were previously studied in experimental rabbit endocarditis in our laboratory (22).
Antimicrobial agents. Ceftriaxone, levofloxacin (for in vitro studies), and vancomycin were purchased from Sigma-Aldrich, Inc. (St. Louis, MO). Ciprofloxacin was purchased from USP (Rockville, MD). Garenoxacin was provided by Bristol-Myers Squibb (Syracuse, N.Y.). Levofloxacin (for in vivo studies) was purchased from Ortho-McNeil Pharmaceutical, Inc. (Raritan, N.J.). Penicillin was purchased from Monarch Pharmaceuticals, Inc. (Bristol, TN).
Antimicrobial susceptibility testing. MICs and minimum bactericidal concentrations (MBCs) were determined by broth microdilution following CLSI guidelines (5, 6).
Time-kill studies. Time-kill studies were performed by inoculating an overnight culture of bacteria (final concentration, 105 to 106 CFU/ml) into flasks containing Mueller Hinton broth supplemented with 2% lysed horse blood in a final volume of 25 ml. Immediately after inoculation, the antimicrobial agent was added in doubling concentrations to each flask at final concentrations ranging from 0.125 to 16 μg/ml. Samples were removed from the flasks just before the addition of antimicrobial agent and after 6 and 24 h of incubation at 37°C in 5% CO2. One-hundred microliters of serial dilutions was plated on blood agar plates for quantitative culture. A bactericidal effect was defined as a 3 log10 CFU/ml reduction of the initial inoculum.
Vegetation induction and colonization. Aortic valve vegetations were induced in New Zealand White rabbits by insertion of a polyethylene catheter across the aortic valve via the right carotid artery, as previously described (22). Bacterial endocarditis was induced 24 h after catheter placement by intravenous bacterial challenge with 1 ml of 106 to 108 CFU of viridans group streptococci. Twenty-four hours later, a blood culture was obtained to confirm the presence of endocarditis. Animals with negative blood cultures were excluded from the study.
Pharmacokinetic studies. Serum antimicrobial concentrations were determined in five healthy animals at 30 min and 1, 2, 4, 6, 12, and 24 h after administration of a single dose of the study antimicrobial agents. Serum concentrations were determined in triplicate by microbiologic assay. Bioassays were performed on Mueller-Hinton agar seeded with Klebsiella pneumoniae ATCC 10031 as the indicator organism for garenoxacin and levofloxacin and with Micrococcus luteus ATCC 9341 for penicillin. Plates were incubated for 20 h at 37°C in room air for quinolones and at 30°C in room air for penicillin. The standard curve was determined using pooled rabbit serum. Using this method, the unbound fraction of the antimicrobial agent was measured. Pharmacokinetic parameters were calculated with Origin version 6.0 (Microcal Software, Inc., Northhampton, MA).
Treatment groups. One-hundred twenty-four rabbits with viridans group streptococcal endocarditis were assigned to one of four different treatment regimens (16 per group): (i) Control (no treatment); (ii) garenoxacin at 20 mg/kg of body weight/12 h intravenously; (iii) levofloxacin at 40 mg/kg/12 h intravenously; and (iv) penicillin at 1.2 x 106 IU/8 h intramuscularly. Treatment was started 24 h after bacterial challenge and lasted for 3 days. Therapeutic monitoring of animals treated with garenoxacin, levofloxacin, or penicillin was performed on blood collected 30 min after administration of the first dose of antimicrobial agent on the second day of treatment.
Twelve hours after administration of the last dose of the antimicrobial agent, the animals were sacrificed with 100 mg/kg of intravenously administered pentobarbital. The chest cavity was opened, the heart was excised, and vegetations were aseptically removed, weighed, and homogenized in 2 ml of Todd-Hewitt broth in a Stomacher 80 (Tekman Co., Cincinnati, Ohio). The homogenate was quantitatively cultured. Serial 10-fold dilutions in Todd-Hewitt broth were plated on blood agar plates (0.1-ml aliquot per plate) and incubated for 48 h at 37°C in 5% CO2. Results of quantitative cultures were expressed as log10 CFU per gram of valve vegetation. Animals that died before completing 3 days of treatment were excluded from statistical analysis.
Detection of emergence of antimicrobial resistance. MIC testing with the same antimicrobial agent administered for treatment was performed on bacteria recovered from aortic valve vegetations.
Identification of quinolone resistance-determining region (QRDR) mutations. Isolates studied in vivo were tested (before treatment) for QRDR mutations. DNA was prepared for amplification from cells grown in Todd-Hewitt broth overnight by alkaline wash as previously described (18). PCR amplification of gyrA and parC was performed in a total reaction volume of 50 μl, as previously described (19). DNA sequencing was performed by ABI PRISM Big Dye Terminator cycle sequencing with the ABI PRISM 377 automated DNA sequencer (Applied Biosystems). Sequence data were analyzed using Sequencher 3.1.1 (GeneCodes Inc., Ann Arbor, MI). Codon positions 81 and 85 of gyrA and 79 and 83 of parC were analyzed for mutations in comparison with previously published QRDR sequences of viridans group streptococci.
Statistical analysis. Differences in means of log10 CFU per gram of vegetation for all groups were analyzed by one-way analysis of variance.
A Tukey-Kramer post hoc range test was used to determine which means differed. Differences in mortality between treatment groups were compared using the Fisher's exact test. MICs between treatments were compared by a Wilcoxon signed-rank test. All tests were two tailed. Significance was defined as P < 0.05.
RESULTS
In vitro susceptibility testing and time-kill studies. The 90% minimum inhibitory concentrations (MIC90) were 0.125, 1, 0.25, and 2 μg/ml for garenoxacin, levofloxacin, penicillin, and vancomycin, respectively; the 90% minimum bactericidal concentrations (MBC90) were 2, >128, 1, and >128 μg/ml for garenoxacin, levofloxacin, penicillin, and vancomycin, respectively, for the 30 study isolates.
Among the isolates studied, garenoxacin showed lower MIC90, MBC90, MIC, and MBC ranges than levofloxacin (P < 0.0001). Antimicrobial susceptibilities of the isolates studied in vivo are summarized in Table 1. A bactericidal effect was observed in time-kill studies with 0.5 μg/ml garenoxacin (4 times the MIC) for both strains; 2 μg/ml levofloxacin (4 times the MIC) for both strains; 4 μg/ml (8 times the MIC) and 8 μg/ml (16 times the MIC) ciprofloxacin for the S. mitis group and the S. sanguinis group isolates, respectively; and 0.25 μg/ml (2 times the MIC) and 16 μg/ml (64 times the MIC) penicillin for the S. mitis group and the S. sanguinis group isolates, respectively (Fig. 1 and 2).
Presence of QRDR mutations. No QRDR mutations were detected in the isolates studied in vivo.
Pharmacokinetic studies. Table 2 summarizes pharmacokinetic parameters and the pharmacodynamic index for both strains studied in vivo. Time above MIC was 100% for all antimicrobial groups studied. Whereas levofloxacin showed greater area under the curve (AUC) and maximum concentration of drug in serum (Cmax) values than garenoxacin, garenoxacin AUC/MIC, AUC/MBC, Cmax/MIC, and Cmax/MBC ratios were greater than the respective values for levofloxacin.
In vivo results. Results of treatment are summarized in Table 3 and Fig. 3. Mean 30-min antimicrobial serum concentrations monitored during treatment were 25, 19, and 41 μg/ml, for penicillin, garenoxacin, and levofloxacin, respectively, in the group infected with the S. mitis group isolate and 26, 18, and 37 μg/ml for penicillin, garenoxacin, and levofloxacin, respectively, in the group infected with the S. sanguinis group isolate. Antimicrobial concentrations measured in infected animals were higher than those assessed in healthy animals after a single dose. No differences were found in antimicrobial levels between the two groups (P = 0.7 for penicillin and levofloxacin; P = 0.1 for garenoxacin). Six animals died before the end of treatment, four with levofloxacin-treated S. mitis group endocarditis and one each with garenoxacin- and levofloxacin-treated S. sanguinis group endocarditis (P = 0.1). For S. mitis group experimental endocarditis, all three antimicrobial agents studied were more active than was no treatment (P < 0.001). There were no significant differences between results of treatment with garenoxacin and levofloxacin. For S. sanguinis group experimental endocarditis, no antimicrobial agent tested was more active than no treatment. No emergence of resistance was detected.
DISCUSSION
In our experimental rabbit model, trovafloxacin had previously failed as a treatment of viridans group streptococcal endocarditis (22). Furthermore, levofloxacin was relatively ineffective in a rabbit model of viridans group streptococcal endocarditis (4). Given the reported improved activity of garenoxacin against viridans group streptococci both in vitro and in an experimental rat model of experimental endocarditis (13), we designed the present study to test garenoxacin in our model of ciprofloxacin-susceptible viridans group streptococcal endocarditis. While garenoxacin did show in vitro activity against the isolates studied, in vivo activity was demonstrated in endocarditis caused by a penicillin-susceptible S. mitis group isolate but not a penicillin-intermediate S. sanguinis group isolate. These results are similar to our prior results with trovafloxacin.
Although newer quinolones have been reported to be active in the treatment of endocarditis (11, 13, 14, 21), based on our present and prior studies (22) quinolone activity in viridans group streptococcal endocarditis appears to depend on characteristics of the strain causing the disease beyond MICs. Entenza et al. proposed that antimicrobial activity in endocarditis may depend on the ability of the agent to achieve bactericidal concentrations inside vegetations. In their study, sparfloxacin was slower than ceftriaxone to achieve bactericidal concentrations in vegetations, potentially explaining differences between the activity of quinolones and beta-lactams in the 3-day treatment model (11). To the best of our knowledge, cardiac valvular penetration of levofloxacin and garenoxacin has not been defined.
In our study, quinolone activity in viridans group streptococcal endocarditis was isolate dependent and independent of MICs, suggesting that factors other than cardiac valvular penetration may be important. We hypothesize that lack of activity of study antimicrobial agents in S. sanguinis group endocarditis is related to the isolate being relatively tolerant to these agents, as suggested by the high MBC/MIC ratios. On the other hand, other strain-specific factors such as glycocalyx formation may influence antimicrobial penetration into vegetations and consequently activity in this model.
The time-kill studies demonstrate that a bactericidal effect occurred with penicillin concentrations twice the MIC for the S. mitis group strain and 64 times the MIC for the S. sanguinis group strain. Time-kill studies may predict penicillin treatment failure of S. sanguinis group endocarditis. Some differences between strains were observed in ciprofloxacin time-kill studies where bactericidal concentrations were 8 and 16 times the MICs for S. mitis group and S. sanguinis group, respectively; however, no differences were found for the other study quinolones. In vitro tests were performed with a lower inoculum than occurs in vivo in vegetations (20). The presence of high titers of organisms in vegetations may explain why conventional susceptibility tests do not predict in vivo activity. When we performed garenoxacin and penicillin time-kill studies with an inoculum of 107 CFU/ml, no bactericidal effect was observed with antimicrobial concentrations ranging from 0.125 to 16 μg/ml for the S. sanguinis group isolate and a bactericidal effect occurred at 8 μg/ml for the S. mitis group isolate (data not shown).
Microbiological assay was the method chosen to measure the free fraction of the antimicrobial agents studied. Pooled rabbit serum was used to prepare standard concentrations. Other methods may not yield directly comparable results. Nevertheless, the pharmacodynamic indices for garenoxacin and levofloxacin were adequate based on cutoffs proposed defined by Andes et al. for successful treatment of experimental endocarditis (i.e., Cmax/MIC greater than 8, AUC/MIC greater than 100, and time above MIC being 100%, depending on the compound) (2). Based on failure of these criteria in our study, we speculate that pharmacodynamic indices predictive of in vivo activity in strains with high MBC/MIC ratios may be those related to MBC and not to MIC (i.e., AUC/MBC, Cmax/MBC, and/or time above MBC, depending on the compound). This hypothesis would also explain findings reported in previous studies with trovafloxacin (22). In our opinion, this issue deserves further study, since in our study 6 out of 30 viridans group streptococci isolated from patients with endocarditis had garenoxacin MBCs at or above that of the S. sanguinis group isolate studied in vivo.
Our findings, along with the emergence of viridans group streptococci resistant to quinolones during quinolone prophylaxis (24), suggest that levofloxacin and garenoxacin may not be ideal for treatment or prophylaxis of viridans group streptococcal infections.
In summary, the present study shows in vitro activity of garenoxacin against clinical viridans group streptococcal isolates. However, MICs did not predict in vivo quinolone activity in experimental S. sanguinis group endocarditis. Possibly, AUC/MBC and/or Cmax/MBC are better predictors of in vivo quinolone activity in viridans group streptococcal endocarditis, since they may better reflect antimicrobial tolerance.
ACKNOWLEDGMENTS
We acknowledge financial support provided by Brystol-Myers Squibb Company Pharmaceutical Research Institute and a grant from the "Instituto de Salud Carlos III," Spanish Ministry of Health (to P.A.-A.).
REFERENCES
American College of Cardiology/American Heart Association, Task Force on Practice Guidelines (Committee on Management of Patients with Valvular Heart Disease). 1998. ACC/AHA guidelines for the management of patients with valvular heart disease. J. Am. Coll. Cardiol. 32:1486-1588.
Andes, D. R., and W. A. Craig. 1998. Pharmacodynamics of fluoroquinolones in experimental models of endocarditis. Clin. Infect. Dis. 27:47-50.
Bassetti, M., L. M. Dembry, P. A. Farrel, D. A. Callan, and V. T. Andriole. 2002. Antimicrobial activities of BMS-284756 compared with those of fluoroquinolones and beta-lactams against gram-positive clinical isolates. Antimicrob. Agents Chemother. 46:234-238.
Chambers, H. F., Q. Xiang, L. Liu, L. Chow, and C. Hackbarth. 1999. Efficacy of levofloxacin for experimental aortic-valve endocarditis in rabbits infected with viridans group streptococcus or Staphylococcus aureus. Antimicrob. Agents Chemother. 43:2742-2746.
Clinical and Laboratory Standards Institute. 1999. Methods for determining bactericidal activity of antimicrobial agents, 1st ed. Approved standard M26-A. Clinical and Laboratory Standards Institute, Wayne, Pa.
Clinical and Laboratory Standards Institute. 2003. Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically, 6th ed. Approved standard M7-A6. Clinical and Laboratory Standards Institute, Wayne, Pa.
Clinical and Laboratory Standards Institute. 2004. MIC testing. Supplemental tables M100-S14 (M7). Clinical and Laboratory Standards Institute, Wayne, Pa.
Doern, G. V., M. J. Ferraro, A. B. Brueggemann, and K. L. Ruoff. 1996. Emergence of high rates of antimicrobial resistance among viridans group streptococci in the United States. Antimicrob. Agents Chemother. 40:891-894.
Durack, D. T., and R. G. Petersdorf. 1973. Chemotherapy of experimental streptococcal endocarditis. I. Comparison of commonly recommended prophylactic regimens. J. Clin. Investig. 52:592-598.
Dworkin, R. J., B. L. Lee, M. A. Sande, and H. F. Chambers. 1989. Treatment of right-sided Staphylococcus aureus endocarditis in intravenous drug users with ciprofloxacin and rifampicin. Lancet ii:1071-1073.
Entenza, J. M., M. Blatter, M. P. Glauser, and P. Moreillon. 1994. Parenteral sparfloxacin compared with ceftriaxone in treatment of experimental endocarditis due to penicillin-susceptible and -resistant streptococci. Antimicrob. Agents Chemother. 38:2683-2688.
Entenza, J. M., I. Caldelari, M. P. Glauser, and P. Moreillon. 1999. Efficacy of levofloxacin in the treatment of experimental endocarditis caused by viridans group streptococci. J. Antimicrob. Chemother. 44:775-786.
Entenza, J. M., J. Vouillamoz, M. P. Glauser, and P. Moreillon. 2004. Efficacy of garenoxacin in treatment of experimental endocarditis due to Staphylococcus aureus or viridans group streptococci. Antimicrob. Agents Chemother. 48:86-92.
Entenza, J. M., J. Vouillamoz, M. P. Glauser, and P. Moreillon. 1997. Levofloxacin versus ciprofloxacin, flucloxacillin, or vancomycin for treatment of experimental endocarditis due to methicillin-susceptible or -resistant Staphylococcus aureus. Antimicrob. Agents Chemother. 41:1662-1667.
Facklam, R. 2002. What happened to the streptococci: overview of taxonomic and nomenclature changes. Clin. Microbiol. Rev. 15:613-630.
Francioli, P., W. Ruch, and D. Stamboulian. 1995. Treatment of streptococcal endocarditis with a single daily dose of ceftriaxone and netilmicin for 14 days: a prospective multicenter study. Clin. Infect. Dis. 21:1406-1410.
Fung-Tomc, J. C., B. Minassian, B. Kolek, E. Huczko, L. Aleksunes, T. Stickle, T. Washo, E. Gradelski, L. Valera, and D. P. Bonner. 2000. Antibacterial spectrum of a novel des-fluoro(6) quinolone, BMS-284756. Antimicrob. Agents Chemother. 44:3351-3356.
Hall, L., K. A. Doerr, S. L. Wohlfiel, and G. D. Roberts. 2003. Evaluation of the MicroSeq system for identification of mycobacteria by 16S ribosomal DNA sequencing and its integration into a routine clinical mycobacteriology laboratory. J. Clin. Microbiol. 41:1447-1453.
Janoir, C., V. Zeller, M. D. Kitzis, N. J. Moreau, and L. Gutmann. 1996. High-level fluoroquinolone resistance in Streptococcus pneumoniae requires mutations in parC and gyrA. Antimicrob. Agents Chemother. 40:2760-2764.
Le, T., and A. S. Bayer. 2003. Combination antibiotic therapy for infective endocarditis. Clin. Infect. Dis. 36:615-621.
Levine, D. P., H. P. Holley, I. Eiseman, P. Willcox, and K. Tack. 2004. Clinafloxacin for the treatment of bacterial endocarditis. Clin. Infect. Dis. 38:620-631.
Piper, K. E., M. S. Rouse, K. L. Ronningen, J. M. Steckelberg, W. R. Wilson, and R. Patel. 2000. Trovafloxacin treatment of viridans group streptococcus experimental endocarditis. Antimicrob. Agents Chemother. 44:2554-2556.
Prabhu, R. M., K. E. Piper, L. M. Baddour, J. M. Steckelberg, W. R. Wilson, and R. Patel. 2004. Antimicrobial susceptibility patterns among viridans group streptococcal isolates from infective endocarditis patients from 1971 to 1986 and 1994 to 2002. Antimicrob. Agents Chemother. 48:4463-4465.
Razonable, R. R., M. R. Litzow, Y. Khaliq, K. E. Piper, M. S. Rouse, and R. Patel. 2002. Bacteremia due to viridans group streptococci with diminished susceptibility to levofloxacin among neutropenic patients receiving levofloxacin prophylaxis. Clin. Infect. Dis. 34:1469-1474.
Tuohy, M., and J. A. Washington. 1997. Antimicrobial susceptibility of viridans group streptococci. Diagn. Microbiol. Infect. Dis. 29:277-280.
Wilson, W. R., R. L. Thompson, C. J. Wilkowske, J. A. Washington, Jr., E. R. Giuliani, and J. E. Geraci. 1981. Short-term therapy for streptococcal infective endocarditis. Combined intramuscular administration of penicillin and streptomycin. JAMA 245:360-363.(Paloma Anguita-Alonso, Ma)
ABSTRACT
The activity of garenoxacin was compared to that of levofloxacin or penicillin in a rabbit model of Streptococcus mitis group (penicillin MIC, 0.125 μg/ml) and Streptococcus sanguinis group (penicillin MIC, 0.25 μg/ml) endocarditis. Garenoxacin and levofloxacin had MICs of 0.125 and 0.5 μg/ml, respectively, for both study isolates. Rabbits with catheter-induced aortic valve endocarditis were given no treatment, penicillin at 1.2 x 106 IU/8 h intramuscularly, garenoxacin at 20 mg/kg of body weight/12 h intravenously, or levofloxacin at 40 mg/kg/12 h intravenously. For both isolates tested, garenoxacin area under the curve (AUC)/MIC and maximum concentration of drug in serum (Cmax)/MIC ratios were 368 and 91, respectively. Rabbits were sacrificed after 3 days of treatment; cardiac valve vegetations were aseptically removed and quantitatively cultured. For S. mitis group experimental endocarditis, all studied antimicrobial agents were more active than no treatment (P < 0.001), whereas for S. sanguinis group endocarditis, no studied antimicrobial agents were more active than no treatment. We conclude that AUC/MIC and Cmax/MIC ratios may not predict activity of some quinolones in experimental viridans group endocarditis and that garenoxacin and levofloxacin may not be ideal choices for serious infections caused by some quinolone-susceptible viridans group streptococci.
INTRODUCTION
Viridans group streptococci are commensal bacterial flora that are common causes of native valve endocarditis. Viridans group streptococcal endocarditis is a severe infection associated with high morbidity and mortality. Its treatment includes administration of a prolonged course of a bactericidal antimicrobial regimen, typically administered intravenously (1). The long-term care and costs associated with such therapy and the emergence of viridans group streptococci relatively resistant to beta-lactams (8, 25) have led to a search for new antimicrobials for the treatment (and prophylaxis) of viridans group streptococcal endocarditis. Combination therapy including an aminoglycoside has been advocated by the American Heart Association to treat penicillin-intermediate and -resistant viridans group streptococcal strains causing infective endocarditis (1) and to shorten therapy for endocarditis caused by penicillin-susceptible viridans group streptococcal strains (26). However, aminoglycoside side effects limit the use of such approaches (16).
Quinolones are generally considered bactericidal antimicrobial agents with high oral bioavailability and good penetration into valvular vegetations (12), making them theoretically suitable for treatment of infective endocarditis. Quinolones in combination with rifampin have been used to treat right-sided Staphylococcus aureus endocarditis (10). However, in the case of viridans group streptococci, quinolones have, in some cases, failed as treatments for experimental endocarditis (4, 22), and emergence of quinolone-resistant viridans group streptococci in bone marrow transplant patients receiving prophylaxis with levofloxacin has been reported (24). Garenoxacin is a new quinolone with enhanced potency against gram-positive bacteria, including penicillin-resistant Streptococcus spp. and Staphylococcus spp. (3, 17), that could potentially address the above-mentioned limitations of quinolones.
Experimental infective endocarditis is a rigorous test of bactericidal activity of antimicrobial agents (9). We designed a rabbit model of infective endocarditis to determine the in vivo activity of garenoxacin in the treatment of viridans group streptococcal endocarditis in comparison to penicillin and levofloxacin.
(This work was presented in part at the 43rd [14 to 17 September 2003, Chicago, Ill.] and 44th [30 October to 2 November 2004, Washington, D.C.] Interscience Conference on Antimicrobial Agents and Chemotherapy.)
MATERIALS AND METHODS
Microorganisms. Thirty viridans group streptococci (nine Streptococcus sanguinis group isolates, nine Streptococcus mitis group isolates, seven Streptococcus mutans group isolates, three Streptococcus salivarius group isolates, and two S. anginosus group isolates) were studied in vitro; all isolates were associated with human infective endocarditis. Organisms were reidentified (23) and classified according to Facklam criteria (15). Two isolates were selected for in vivo studies: IDRL-4600, a penicillin-susceptible S. mitis group isolate, and IDRL-4954, a penicillin-intermediate S. sanguinis group isolate. Both were ciprofloxacin susceptible according to Clinical and Laboratory Standard Institute (CLSI) interpretive standards (7), and both were previously studied in experimental rabbit endocarditis in our laboratory (22).
Antimicrobial agents. Ceftriaxone, levofloxacin (for in vitro studies), and vancomycin were purchased from Sigma-Aldrich, Inc. (St. Louis, MO). Ciprofloxacin was purchased from USP (Rockville, MD). Garenoxacin was provided by Bristol-Myers Squibb (Syracuse, N.Y.). Levofloxacin (for in vivo studies) was purchased from Ortho-McNeil Pharmaceutical, Inc. (Raritan, N.J.). Penicillin was purchased from Monarch Pharmaceuticals, Inc. (Bristol, TN).
Antimicrobial susceptibility testing. MICs and minimum bactericidal concentrations (MBCs) were determined by broth microdilution following CLSI guidelines (5, 6).
Time-kill studies. Time-kill studies were performed by inoculating an overnight culture of bacteria (final concentration, 105 to 106 CFU/ml) into flasks containing Mueller Hinton broth supplemented with 2% lysed horse blood in a final volume of 25 ml. Immediately after inoculation, the antimicrobial agent was added in doubling concentrations to each flask at final concentrations ranging from 0.125 to 16 μg/ml. Samples were removed from the flasks just before the addition of antimicrobial agent and after 6 and 24 h of incubation at 37°C in 5% CO2. One-hundred microliters of serial dilutions was plated on blood agar plates for quantitative culture. A bactericidal effect was defined as a 3 log10 CFU/ml reduction of the initial inoculum.
Vegetation induction and colonization. Aortic valve vegetations were induced in New Zealand White rabbits by insertion of a polyethylene catheter across the aortic valve via the right carotid artery, as previously described (22). Bacterial endocarditis was induced 24 h after catheter placement by intravenous bacterial challenge with 1 ml of 106 to 108 CFU of viridans group streptococci. Twenty-four hours later, a blood culture was obtained to confirm the presence of endocarditis. Animals with negative blood cultures were excluded from the study.
Pharmacokinetic studies. Serum antimicrobial concentrations were determined in five healthy animals at 30 min and 1, 2, 4, 6, 12, and 24 h after administration of a single dose of the study antimicrobial agents. Serum concentrations were determined in triplicate by microbiologic assay. Bioassays were performed on Mueller-Hinton agar seeded with Klebsiella pneumoniae ATCC 10031 as the indicator organism for garenoxacin and levofloxacin and with Micrococcus luteus ATCC 9341 for penicillin. Plates were incubated for 20 h at 37°C in room air for quinolones and at 30°C in room air for penicillin. The standard curve was determined using pooled rabbit serum. Using this method, the unbound fraction of the antimicrobial agent was measured. Pharmacokinetic parameters were calculated with Origin version 6.0 (Microcal Software, Inc., Northhampton, MA).
Treatment groups. One-hundred twenty-four rabbits with viridans group streptococcal endocarditis were assigned to one of four different treatment regimens (16 per group): (i) Control (no treatment); (ii) garenoxacin at 20 mg/kg of body weight/12 h intravenously; (iii) levofloxacin at 40 mg/kg/12 h intravenously; and (iv) penicillin at 1.2 x 106 IU/8 h intramuscularly. Treatment was started 24 h after bacterial challenge and lasted for 3 days. Therapeutic monitoring of animals treated with garenoxacin, levofloxacin, or penicillin was performed on blood collected 30 min after administration of the first dose of antimicrobial agent on the second day of treatment.
Twelve hours after administration of the last dose of the antimicrobial agent, the animals were sacrificed with 100 mg/kg of intravenously administered pentobarbital. The chest cavity was opened, the heart was excised, and vegetations were aseptically removed, weighed, and homogenized in 2 ml of Todd-Hewitt broth in a Stomacher 80 (Tekman Co., Cincinnati, Ohio). The homogenate was quantitatively cultured. Serial 10-fold dilutions in Todd-Hewitt broth were plated on blood agar plates (0.1-ml aliquot per plate) and incubated for 48 h at 37°C in 5% CO2. Results of quantitative cultures were expressed as log10 CFU per gram of valve vegetation. Animals that died before completing 3 days of treatment were excluded from statistical analysis.
Detection of emergence of antimicrobial resistance. MIC testing with the same antimicrobial agent administered for treatment was performed on bacteria recovered from aortic valve vegetations.
Identification of quinolone resistance-determining region (QRDR) mutations. Isolates studied in vivo were tested (before treatment) for QRDR mutations. DNA was prepared for amplification from cells grown in Todd-Hewitt broth overnight by alkaline wash as previously described (18). PCR amplification of gyrA and parC was performed in a total reaction volume of 50 μl, as previously described (19). DNA sequencing was performed by ABI PRISM Big Dye Terminator cycle sequencing with the ABI PRISM 377 automated DNA sequencer (Applied Biosystems). Sequence data were analyzed using Sequencher 3.1.1 (GeneCodes Inc., Ann Arbor, MI). Codon positions 81 and 85 of gyrA and 79 and 83 of parC were analyzed for mutations in comparison with previously published QRDR sequences of viridans group streptococci.
Statistical analysis. Differences in means of log10 CFU per gram of vegetation for all groups were analyzed by one-way analysis of variance.
A Tukey-Kramer post hoc range test was used to determine which means differed. Differences in mortality between treatment groups were compared using the Fisher's exact test. MICs between treatments were compared by a Wilcoxon signed-rank test. All tests were two tailed. Significance was defined as P < 0.05.
RESULTS
In vitro susceptibility testing and time-kill studies. The 90% minimum inhibitory concentrations (MIC90) were 0.125, 1, 0.25, and 2 μg/ml for garenoxacin, levofloxacin, penicillin, and vancomycin, respectively; the 90% minimum bactericidal concentrations (MBC90) were 2, >128, 1, and >128 μg/ml for garenoxacin, levofloxacin, penicillin, and vancomycin, respectively, for the 30 study isolates.
Among the isolates studied, garenoxacin showed lower MIC90, MBC90, MIC, and MBC ranges than levofloxacin (P < 0.0001). Antimicrobial susceptibilities of the isolates studied in vivo are summarized in Table 1. A bactericidal effect was observed in time-kill studies with 0.5 μg/ml garenoxacin (4 times the MIC) for both strains; 2 μg/ml levofloxacin (4 times the MIC) for both strains; 4 μg/ml (8 times the MIC) and 8 μg/ml (16 times the MIC) ciprofloxacin for the S. mitis group and the S. sanguinis group isolates, respectively; and 0.25 μg/ml (2 times the MIC) and 16 μg/ml (64 times the MIC) penicillin for the S. mitis group and the S. sanguinis group isolates, respectively (Fig. 1 and 2).
Presence of QRDR mutations. No QRDR mutations were detected in the isolates studied in vivo.
Pharmacokinetic studies. Table 2 summarizes pharmacokinetic parameters and the pharmacodynamic index for both strains studied in vivo. Time above MIC was 100% for all antimicrobial groups studied. Whereas levofloxacin showed greater area under the curve (AUC) and maximum concentration of drug in serum (Cmax) values than garenoxacin, garenoxacin AUC/MIC, AUC/MBC, Cmax/MIC, and Cmax/MBC ratios were greater than the respective values for levofloxacin.
In vivo results. Results of treatment are summarized in Table 3 and Fig. 3. Mean 30-min antimicrobial serum concentrations monitored during treatment were 25, 19, and 41 μg/ml, for penicillin, garenoxacin, and levofloxacin, respectively, in the group infected with the S. mitis group isolate and 26, 18, and 37 μg/ml for penicillin, garenoxacin, and levofloxacin, respectively, in the group infected with the S. sanguinis group isolate. Antimicrobial concentrations measured in infected animals were higher than those assessed in healthy animals after a single dose. No differences were found in antimicrobial levels between the two groups (P = 0.7 for penicillin and levofloxacin; P = 0.1 for garenoxacin). Six animals died before the end of treatment, four with levofloxacin-treated S. mitis group endocarditis and one each with garenoxacin- and levofloxacin-treated S. sanguinis group endocarditis (P = 0.1). For S. mitis group experimental endocarditis, all three antimicrobial agents studied were more active than was no treatment (P < 0.001). There were no significant differences between results of treatment with garenoxacin and levofloxacin. For S. sanguinis group experimental endocarditis, no antimicrobial agent tested was more active than no treatment. No emergence of resistance was detected.
DISCUSSION
In our experimental rabbit model, trovafloxacin had previously failed as a treatment of viridans group streptococcal endocarditis (22). Furthermore, levofloxacin was relatively ineffective in a rabbit model of viridans group streptococcal endocarditis (4). Given the reported improved activity of garenoxacin against viridans group streptococci both in vitro and in an experimental rat model of experimental endocarditis (13), we designed the present study to test garenoxacin in our model of ciprofloxacin-susceptible viridans group streptococcal endocarditis. While garenoxacin did show in vitro activity against the isolates studied, in vivo activity was demonstrated in endocarditis caused by a penicillin-susceptible S. mitis group isolate but not a penicillin-intermediate S. sanguinis group isolate. These results are similar to our prior results with trovafloxacin.
Although newer quinolones have been reported to be active in the treatment of endocarditis (11, 13, 14, 21), based on our present and prior studies (22) quinolone activity in viridans group streptococcal endocarditis appears to depend on characteristics of the strain causing the disease beyond MICs. Entenza et al. proposed that antimicrobial activity in endocarditis may depend on the ability of the agent to achieve bactericidal concentrations inside vegetations. In their study, sparfloxacin was slower than ceftriaxone to achieve bactericidal concentrations in vegetations, potentially explaining differences between the activity of quinolones and beta-lactams in the 3-day treatment model (11). To the best of our knowledge, cardiac valvular penetration of levofloxacin and garenoxacin has not been defined.
In our study, quinolone activity in viridans group streptococcal endocarditis was isolate dependent and independent of MICs, suggesting that factors other than cardiac valvular penetration may be important. We hypothesize that lack of activity of study antimicrobial agents in S. sanguinis group endocarditis is related to the isolate being relatively tolerant to these agents, as suggested by the high MBC/MIC ratios. On the other hand, other strain-specific factors such as glycocalyx formation may influence antimicrobial penetration into vegetations and consequently activity in this model.
The time-kill studies demonstrate that a bactericidal effect occurred with penicillin concentrations twice the MIC for the S. mitis group strain and 64 times the MIC for the S. sanguinis group strain. Time-kill studies may predict penicillin treatment failure of S. sanguinis group endocarditis. Some differences between strains were observed in ciprofloxacin time-kill studies where bactericidal concentrations were 8 and 16 times the MICs for S. mitis group and S. sanguinis group, respectively; however, no differences were found for the other study quinolones. In vitro tests were performed with a lower inoculum than occurs in vivo in vegetations (20). The presence of high titers of organisms in vegetations may explain why conventional susceptibility tests do not predict in vivo activity. When we performed garenoxacin and penicillin time-kill studies with an inoculum of 107 CFU/ml, no bactericidal effect was observed with antimicrobial concentrations ranging from 0.125 to 16 μg/ml for the S. sanguinis group isolate and a bactericidal effect occurred at 8 μg/ml for the S. mitis group isolate (data not shown).
Microbiological assay was the method chosen to measure the free fraction of the antimicrobial agents studied. Pooled rabbit serum was used to prepare standard concentrations. Other methods may not yield directly comparable results. Nevertheless, the pharmacodynamic indices for garenoxacin and levofloxacin were adequate based on cutoffs proposed defined by Andes et al. for successful treatment of experimental endocarditis (i.e., Cmax/MIC greater than 8, AUC/MIC greater than 100, and time above MIC being 100%, depending on the compound) (2). Based on failure of these criteria in our study, we speculate that pharmacodynamic indices predictive of in vivo activity in strains with high MBC/MIC ratios may be those related to MBC and not to MIC (i.e., AUC/MBC, Cmax/MBC, and/or time above MBC, depending on the compound). This hypothesis would also explain findings reported in previous studies with trovafloxacin (22). In our opinion, this issue deserves further study, since in our study 6 out of 30 viridans group streptococci isolated from patients with endocarditis had garenoxacin MBCs at or above that of the S. sanguinis group isolate studied in vivo.
Our findings, along with the emergence of viridans group streptococci resistant to quinolones during quinolone prophylaxis (24), suggest that levofloxacin and garenoxacin may not be ideal for treatment or prophylaxis of viridans group streptococcal infections.
In summary, the present study shows in vitro activity of garenoxacin against clinical viridans group streptococcal isolates. However, MICs did not predict in vivo quinolone activity in experimental S. sanguinis group endocarditis. Possibly, AUC/MBC and/or Cmax/MBC are better predictors of in vivo quinolone activity in viridans group streptococcal endocarditis, since they may better reflect antimicrobial tolerance.
ACKNOWLEDGMENTS
We acknowledge financial support provided by Brystol-Myers Squibb Company Pharmaceutical Research Institute and a grant from the "Instituto de Salud Carlos III," Spanish Ministry of Health (to P.A.-A.).
REFERENCES
American College of Cardiology/American Heart Association, Task Force on Practice Guidelines (Committee on Management of Patients with Valvular Heart Disease). 1998. ACC/AHA guidelines for the management of patients with valvular heart disease. J. Am. Coll. Cardiol. 32:1486-1588.
Andes, D. R., and W. A. Craig. 1998. Pharmacodynamics of fluoroquinolones in experimental models of endocarditis. Clin. Infect. Dis. 27:47-50.
Bassetti, M., L. M. Dembry, P. A. Farrel, D. A. Callan, and V. T. Andriole. 2002. Antimicrobial activities of BMS-284756 compared with those of fluoroquinolones and beta-lactams against gram-positive clinical isolates. Antimicrob. Agents Chemother. 46:234-238.
Chambers, H. F., Q. Xiang, L. Liu, L. Chow, and C. Hackbarth. 1999. Efficacy of levofloxacin for experimental aortic-valve endocarditis in rabbits infected with viridans group streptococcus or Staphylococcus aureus. Antimicrob. Agents Chemother. 43:2742-2746.
Clinical and Laboratory Standards Institute. 1999. Methods for determining bactericidal activity of antimicrobial agents, 1st ed. Approved standard M26-A. Clinical and Laboratory Standards Institute, Wayne, Pa.
Clinical and Laboratory Standards Institute. 2003. Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically, 6th ed. Approved standard M7-A6. Clinical and Laboratory Standards Institute, Wayne, Pa.
Clinical and Laboratory Standards Institute. 2004. MIC testing. Supplemental tables M100-S14 (M7). Clinical and Laboratory Standards Institute, Wayne, Pa.
Doern, G. V., M. J. Ferraro, A. B. Brueggemann, and K. L. Ruoff. 1996. Emergence of high rates of antimicrobial resistance among viridans group streptococci in the United States. Antimicrob. Agents Chemother. 40:891-894.
Durack, D. T., and R. G. Petersdorf. 1973. Chemotherapy of experimental streptococcal endocarditis. I. Comparison of commonly recommended prophylactic regimens. J. Clin. Investig. 52:592-598.
Dworkin, R. J., B. L. Lee, M. A. Sande, and H. F. Chambers. 1989. Treatment of right-sided Staphylococcus aureus endocarditis in intravenous drug users with ciprofloxacin and rifampicin. Lancet ii:1071-1073.
Entenza, J. M., M. Blatter, M. P. Glauser, and P. Moreillon. 1994. Parenteral sparfloxacin compared with ceftriaxone in treatment of experimental endocarditis due to penicillin-susceptible and -resistant streptococci. Antimicrob. Agents Chemother. 38:2683-2688.
Entenza, J. M., I. Caldelari, M. P. Glauser, and P. Moreillon. 1999. Efficacy of levofloxacin in the treatment of experimental endocarditis caused by viridans group streptococci. J. Antimicrob. Chemother. 44:775-786.
Entenza, J. M., J. Vouillamoz, M. P. Glauser, and P. Moreillon. 2004. Efficacy of garenoxacin in treatment of experimental endocarditis due to Staphylococcus aureus or viridans group streptococci. Antimicrob. Agents Chemother. 48:86-92.
Entenza, J. M., J. Vouillamoz, M. P. Glauser, and P. Moreillon. 1997. Levofloxacin versus ciprofloxacin, flucloxacillin, or vancomycin for treatment of experimental endocarditis due to methicillin-susceptible or -resistant Staphylococcus aureus. Antimicrob. Agents Chemother. 41:1662-1667.
Facklam, R. 2002. What happened to the streptococci: overview of taxonomic and nomenclature changes. Clin. Microbiol. Rev. 15:613-630.
Francioli, P., W. Ruch, and D. Stamboulian. 1995. Treatment of streptococcal endocarditis with a single daily dose of ceftriaxone and netilmicin for 14 days: a prospective multicenter study. Clin. Infect. Dis. 21:1406-1410.
Fung-Tomc, J. C., B. Minassian, B. Kolek, E. Huczko, L. Aleksunes, T. Stickle, T. Washo, E. Gradelski, L. Valera, and D. P. Bonner. 2000. Antibacterial spectrum of a novel des-fluoro(6) quinolone, BMS-284756. Antimicrob. Agents Chemother. 44:3351-3356.
Hall, L., K. A. Doerr, S. L. Wohlfiel, and G. D. Roberts. 2003. Evaluation of the MicroSeq system for identification of mycobacteria by 16S ribosomal DNA sequencing and its integration into a routine clinical mycobacteriology laboratory. J. Clin. Microbiol. 41:1447-1453.
Janoir, C., V. Zeller, M. D. Kitzis, N. J. Moreau, and L. Gutmann. 1996. High-level fluoroquinolone resistance in Streptococcus pneumoniae requires mutations in parC and gyrA. Antimicrob. Agents Chemother. 40:2760-2764.
Le, T., and A. S. Bayer. 2003. Combination antibiotic therapy for infective endocarditis. Clin. Infect. Dis. 36:615-621.
Levine, D. P., H. P. Holley, I. Eiseman, P. Willcox, and K. Tack. 2004. Clinafloxacin for the treatment of bacterial endocarditis. Clin. Infect. Dis. 38:620-631.
Piper, K. E., M. S. Rouse, K. L. Ronningen, J. M. Steckelberg, W. R. Wilson, and R. Patel. 2000. Trovafloxacin treatment of viridans group streptococcus experimental endocarditis. Antimicrob. Agents Chemother. 44:2554-2556.
Prabhu, R. M., K. E. Piper, L. M. Baddour, J. M. Steckelberg, W. R. Wilson, and R. Patel. 2004. Antimicrobial susceptibility patterns among viridans group streptococcal isolates from infective endocarditis patients from 1971 to 1986 and 1994 to 2002. Antimicrob. Agents Chemother. 48:4463-4465.
Razonable, R. R., M. R. Litzow, Y. Khaliq, K. E. Piper, M. S. Rouse, and R. Patel. 2002. Bacteremia due to viridans group streptococci with diminished susceptibility to levofloxacin among neutropenic patients receiving levofloxacin prophylaxis. Clin. Infect. Dis. 34:1469-1474.
Tuohy, M., and J. A. Washington. 1997. Antimicrobial susceptibility of viridans group streptococci. Diagn. Microbiol. Infect. Dis. 29:277-280.
Wilson, W. R., R. L. Thompson, C. J. Wilkowske, J. A. Washington, Jr., E. R. Giuliani, and J. E. Geraci. 1981. Short-term therapy for streptococcal infective endocarditis. Combined intramuscular administration of penicillin and streptomycin. JAMA 245:360-363.(Paloma Anguita-Alonso, Ma)