Comparative killing rates of gatifloxacin and ciprofloxacin against 14 clinical isolates: impact of bacterial strain and antibiotic concentration

Diagnostic Microbiology and Infectious Disease 44 (2002) 59 – 61

www.elsevier.com/locate/diagmicrobio

Note

Comparative killing rates of gatifloxacin and ciprofloxacin against 14 clinical isolates: impact of bacterial strain and antibiotic concentration Susan L. Pendlanda,*, Melinda M. Neuhauserb, Kevin W. Gareyb, Jennifer L. Prausea, Rose Jungc a

The University of Illinois at Chicago, Department of Pharmacy Practice, Microbiology Research Laboratory, Chicago, IL, USA b University of Houston, Department of Clinical Sciences and Administration, Houston, TX, USA c The University of Colorado Health Sciences Center, Department of Pharmacy Practice, Denver, CO, USA Received 21 January 2002; accepted 27 May 2002

Abstract The influence of bacterial strain and antibiotic concentration on the time to achieve in vitro bactericidal activity was determined for gatifloxacin and ciprofloxacin using time-kill methodology. Killing rates were significantly affected by bacterial strain, antibiotic concentration, and type of fluoroquinolone. The most rapid bactericidal activity was seen against members of the Enterobacteriaceae and with fluoroquinolone concentrations of 8 –16X MIC. In general, gatifloxacin demonstrated faster killing against Acinetobacter baumanii, Legionella pneumophila, Staphylococcus aureus, and Streptococcus pneumoniae. © 2002 Elsevier Science Inc. All rights reserved.

Selection of an appropriate antibiotic is multi-faceted and includes knowledge of antibiotic activity, bacterial susceptibility patterns, and site of infection. An antibiotic with bactericidal activity is considered desirable; however, this activity is influenced by several factors including type of organism and achievable antibiotic concentrations. Although not as well recognized, the rate at which bactericidal activity is achieved may also play an important role, especially for patients with compromised immune function or infections of the central nervous system. Fluoroquinolones have been shown to demonstrate faster bactericidal activity than vancomycin and various ␤-lactam antibiotics against staphylococci, streptococci, enterococci, and members of the Enterobacteriaceae (Fung-Tomc et al., 2000). They exhibit concentration-dependent killing kinetics, with an optimum bactericidal concentration (OBC) of approximately 8 –10⫻ MIC (Morrissey, 1997; Gradelski et al., 2002). The OBC of the newer fluoroquinolones has been determined predominately against Gram-positive organisms. The purpose of this study was to examine the influence of bacterial strain, fluoroquinolone selection, and antimicrobial concentration on the time to achieve bactericidal activity. A variety of Gram-positive and Gram-negative * Corresponding author. Tel.: ⫹1-312-996-8639; fax: ⫹1-312-4131797. E-mail address: [email protected] (S.L. Pendland).

pathogens were evaluated to characterize differences in bacterial strains. Gatifloxacin, a newer, broad-spectrum fluoroquinolone was compared to the first-generation agent, ciprofloxacin. Concentrations above, below, and equal to the OBC were tested. Multiple early time-points were utilized to assess the impact of these variables on killing rates. MICs and time-kill assays were performed on 14 clinical isolates obtained from the Microbiology Laboratory at the University of Illinois Medical Center (Chicago, IL). The organisms tested included 2 strains each of Escherichia coli, Klebsiella pneumoniae, Acinetobacter baumanii, Staphylococcus aureus (MSSA), Streptococcus pneumoniae (1 penicillin/erythromycin resistant), Enterococcus faecalis, and Legionella pneumophila. Appropriate control strains per NCCLS guidelines were utilized for validation of the MIC results (NCCLS, 2001). Each inoculum was prepared from log phase growth organisms and adjusted with sterile saline (distilled water for Legionella) until the turbidity matched a 0.5 McFarland standard using a spectrophotometer at 625 nm. Each suspension was further diluted in an appropriate broth to obtain a final inoculum of approximately 5 ⫻ 105 CFU/mL. The exact inoculum size was determined via colony counts. Gatifloxacin (Bristol-Myers Squibb, Wallingford, CT) and ciprofloxacin (United States Pharmacopeia, Rockville, MD) powders were prepared according to the manufacturer’s recommendations or NCCLS guidelines (NCCLS,

0732-8893/02/$ – see front matter © 2002 Elsevier Science Inc. All rights reserved. PII: S 0 7 3 2 - 8 8 9 3 ( 0 2 ) 0 0 4 2 2 - 4

60

S.L. Pendland et al. / Diagnostic Microbiology and Infectious Disease 44 (2002) 59 – 61

Table 1 MICs and killing rates of gatifloxacin and ciprofloxacin against 14 clinical isolates Organism

E. coli

Strain

AA6 LB572

K. pneumoniae

DL1 PS65

A. baumanii

JI81 X32825

L. pneumoniae

UIC1 UIC9

S. aureus

SM301 JR333

S. pneumoniae

JH1 B23-61

E. faecalis

M54624 S35435

a

Antibiotic

Gatifloxacin Ciprofloxacin Gatifloxacin Ciprofloxacin Gatifloxacin Ciprofloxacin Gatifloxacin Ciprofloxacin Gatifloxacin Ciprofloxacin Gatifloxacin Ciprofloxacin Gatifloxacin Ciprofloxacin Gatifloxacin Ciprofloxacin Gatifloxacin Ciprofloxacin Gatifloxacin Ciprofloxacin Gatifloxacin Ciprofloxacin Gatifloxacin Ciprofloxacin Gatifloxacin Ciprofloxacin Gatifloxacin Ciprofloxacin

MIC (␮g/mL)

0.03 0.015 4 8 0.06 0.03 0.06 0.06 0.03 0.25 0.06 0.5 0.008 0.015 0.008 0.015 0.12 0.5 0.25 0.5 0.5 1 0.5 1 0.25 0.5 0.25 0.5

Time (hours) to 3 log10 decrease in inoculum 4X MIC

8X MIC

16X MIC

1.64 1.53 3.13 3.51 1.53 1.50 1.70 2.05 5.16 12.7 19.59 27.34a 13.70 10.90 3.99 4.19 19.42 19.52 9.43 13.19 7.54 6.83 9.64 13.84 16.28 15.70 16.84 20.07

1.49 1.12 3.18 3.67 1.11 1.15 0.92 1.78 5.13 9.85 15.32 26.19a 3.88 6.85 1.72 3.02 19.11 20.53 4.00 12.72 7.09 10.09 9.58 13.59 10.91 16.38 7.97 7.76

0.92 0.87 3.47 3.97 0.81 0.99 1.0 1.15 4.81 14.16 14.32 20.31 1.96 5.15 1.11 1.71 14.88 19.09 2.93 13.03 5.87 5.84 10.91 13.02 14.92 15.69 12.36 9.67

projected time

2000). Cation-adjusted Mueller-Hinton broth (Difco, Detroit, MI) was used to test all organisms except S. pneumoniae and L. pneumophila. Cation-adjusted Mueller-Hinton broth ⫹ 3% lysed horse blood (Remel, Lenexa, KS) was used in all assays with S. pneumoniae, while buffered yeast extract (BYE) broth (Ristroph et al., 1980) was used for L. pneumophila. MICs were performed in duplicate using the microbroth dilution method (NCCLS, 2000). The bactericidal activity of the antimicrobial agents was determined in duplicate using the time-kill method (NCCLS, 1999). Antimicrobial concentrations tested were 4⫻, 8⫻ and 16⫻ MIC. Test tubes with broth media and known concentrations of the antibiotics were inoculated with the organism. Control tubes were utilized that contained no antimicrobial agent. The final inoculum was confirmed at time 0; subsequent viable counts were determined at 0.25, 0.5, 0.75, 1, 2, 4, 8, and 24 h. Antibiotic carryover was prevented by saline (distilled water for Legionella) dilutions. The diluted samples were plated (WASP Spiral Plater, Microbiology International, Frederick, MD) on appropriate agar media (blood agar or BCYE). Colony counts were read (Protos Colony Counter, Microbiology International) after 24 – 48 h of incubation in humidified air (5% CO2 for S. pneumoniae) at 35°C. The rate and extent of killing were determined by plotting viable counts (log10

CFU/mL) against time (hours). Bactericidal activity was defined as a ⱖ3 log10 decrease in CFU/mL. The lower limit of detection was 1.3 log10 CFU/mL. Time to achieve bactericidal activity was extrapolated from the time-kill curve using log-transformed regression analysis. Statistical analysis was performed using Systat Version 9.01 (SPSS, Inc, Chicago, IL). The univariable effect of bacterial strain, fluoroquinolone selection, and antibiotic concentration was performed by paired sample t test and repeated measures ANOVA. Multivariate analysis was performed using a general linear estimate model. A p-value of ⬍0.05 was considered significant. The MICs and time to achieve bactericidal activity are listed in Table 1. The MICs of gatifloxacin and ciprofloxacin were within one or two dilutions against clinical isolates of E. coli, K. pneumoniae, S. pneumoniae, E. faecalis, L. pneumophila and S. aureus. Gatifloxacin was threefold more potent against A. baumanii than ciprofloxacin. In the multivariate analysis, type of organism, antibiotic concentration, and type of fluoroquinolone independently impacted the time to achieve bactericidal activity. The rate of bactericidal activity was significantly influenced by different organisms (p ⬍ 0.001). With both fluoroquinolones, the time to achieve bactericidal activity averaged ⬍5 h for E. coli, K. pneumoniae, and L. pneumoniae, 5–10 h for S.

S.L. Pendland et al. / Diagnostic Microbiology and Infectious Disease 44 (2002) 59 – 61

pneumoniae, and ⬎ 10 h for A. baumanii, E. faecalis, and S. aureus. Mean bactericidal activity against the 14 strains was observed at 7.4 and 9.8 h for gatifloxacin and ciprofloxacin, respectively (p ⬍ 0.05). Differences in the time to reach bactericidal activity were observed with S. aureus, S. pneumoniae, and L. pneumophila, and were especially pronounced for A. baumanii. Antibiotic concentration significantly impacted the rate of killing activity (p ⫽ 0.004). Time to achieve bactericidal activity with either agent averaged 10 h at 4⫻ MIC and approximately 7.6 and 8 h at 8⫻ and 16⫻ MIC, respectively. Few studies have compared the bactericidal rates of different fluoroquinolones against Gram-positive and Gramnegative organisms (Fung-Tomc et al., 2000; Hoellman et al., 1999). Hoellman et al. performed time-kill studies with gatifloxacin and ciprofloxacin against 12 pneumococcal strains (Hoellman et al., 1999). At 4 h, bactericidal activity was achieved with gatifloxacin in 4 strains at 4⫻ MIC and 6 strains at 8⫻ MIC compared to only one strain for ciprofloxacin at each concentration. In our study, the time to achieve bactericidal activity averaged 8 and 11 h for gatifloxacin and ciprofloxacin for S. pneumoniae (n ⫽ 2), respectively, and was not achieved by 4 h with either agent at any concentration. Differences in the killing rates of organisms in these studies may be a result of strain variability. In a study by Fung-Tomc et al., gatifloxacin and levofloxacin achieved bactericidal activity against two strains of MSSA within 2 h when tested at concentrations of 10⫻ MIC. In our study, time to achieve bactericidal activity was considerably higher with both gatifloxacin and ciprofloxacin against MSSA (n ⫽ 2) when tested at concentrations of 4⫻, 8⫻, and 16⫻ MIC. However, gatifloxacin demonstrated considerably faster killing against one of the MSSA strains (JR333), with bactericidal activity observed at 3– 4 h compared to 13 h for ciprofloxacin when tested at 8 –16⫻ MIC. Compared to the Gram-positive organisms, the time to achieve bactericidal activity was faster for both gatifloxacin and ciprofloxacin in Gram-negative isolates. Other investigators have also reported a more rapid killing of Gramnegative bacteria by fluoroquinolones (Fung-Tomc et al., 2000; Debbia et al., 1987; Lewin et al., 1997). Similar to other studies (Fung-Tomc et al., 2000), gatifloxacin and ciprofloxacin achieved bactericidal activity by approximately 2 h against sensitive strains of E. coli and K. pneumoniae. Such rapid killing may not be observed with organisms demonstrating higher fluoroquinolone MICs, as the killing rates for a resistant strain of E. coli were double the rates observed with the susceptible strain. Of note, gatifloxacin achieved bactericidal activity faster than ciprofloxacin against A. baumanii and L. pneumonphila (8⫻ and 16⫻ MIC). While additional in vitro and clinical data are needed, our results suggest enhanced killing rates with gatifloxacin against these problematic pathogens.

61

Fluoroquinolones are known to exhibit concentrationdependent killing. Optimal bactericidal activity has been shown to occur at fluoroquinolone concentrations at approximately 10⫻ MIC. This study supports and extends these observations by concluding that the time to achieve bactericidal activity may also be optimized at similar concentrations. Time to achieve bactericidal activity appeared to plateau as concentrations were increased from 8⫻ to 16⫻ MIC, further supporting the evidence that concentrations of 10⫻ MIC may be optimal for fluoroquinolones. In conclusion, the time to achieve in vitro bactericidal activity was significantly influenced by organism, fluoroquinolone agent, and antimicrobial concentration. Additional studies are needed to further characterize the differences in bactericidal activities of first-, second-, and thirdgeneration fluoroquinolones; however, in critically ill patients, every effort should be made to achieve optimal concentrations of fluoroquinolones of at least 10⫻ MIC of the infectious organism. Acknowledgments This work was supported by a grant from Bristol-Myers Squibb. References Debbia, E., Schito, G. C., Nicoletti, G., & Speciale, A. (1987). In vitro activity of pefloxacin against Gram-negative and Gram-positive bacteria in comparison with other antibiotics. Chemioterapia, 4, 319 –323. Fung-Tomc, J. C., Gradelski, E., Valera, L., Kolek, B., & Bonner, D. P. (2000). Comparative killing rates of fluoroquinolones and cell wallactive agents. Antimicrobial Agents and Chemotherapy, 44, 1377– 1380. Gradelski, E., Benjamin, K, Bonner, D., & Fung-Tomc, J. (2002). Bactericidal mechanism of gatifloxacin compared with other fluoroquinolones. Journal of Antimicrobial Chemotherapy, 49, 185–188. Hoellman, D. B., Lin, G., Jacobs, M. R., & Appelbaum, P. C. (1999). Anti-pneumococcal activity of gatifloxacin compared with other quinolone and non-quinolone agents. Journal of Antimicrobial Chemotherapy, 43, 645– 649. Lewin, C. S. (1992). Antibacterial activity of a 1,8-naphthyridine quinolone, PD 131628. Journal of Medical Microbiology, 36, 353–357. Morrissey, I. (1997). Bactericidal index: a new way to assess quinolone bactericidal activity in vitro. Journal of Antimicrobial Chemotherapy, 39, 713–717. National Committee for Clinical Laboratory Standards. (2001). Performance standards for antimicrobial susceptibility testing; eleventh informational supplement. NCCLS document M100 –S11. Wayne, PA. National Committee for Clinical Laboratory Standards. (2000). Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically, 5th ed. Approved standard M7–A5. NCCLS. Wayne, PA. National Committee for Clinical Laboratory Standards. (1999). Methods for determining bactericidal activity of antimicrobial agents. Approved Guideline M26-A. NCCLS. Wayne, PA. Ristroph, J. D., Hedlund, K. W., & Allen, R. G. (1980). Liquid medium for growth of Legionella pneumophila. Journal of Clinical Microbiology, 11, 19 –21.