Alasdair P. MacGowan*, Mandy Wootton and H. Alan Holt

Journal of Antimicrobial Chemotherapy (1999) 43, 345 349 JAC The antibacterial efficacy of levofloxacin and ciprofloxacin against Pseudomonas aeruginosa assessed by combining antibiotic exposure and bacterial susceptibility Alasdair P. MacGowan*, Mandy Wootton and H. Alan Holt Bristol Centre for Antimicrobial Research and Evaluation, Southmead Health Services NHS Trust and University of Bristol, Department of Medical Microbiology, Southmead Hospital, Westbury-on-Trym, Bristol BS10 5NB, UK Ciprofloxacin has a four-fold greater in-vitro activity than levofloxacin against Pseudomonas aeruginosa, but levofloxacin has a four-fold higher area under the serum concentration time curve (AUC) for an equivalent dose. It has been proposed that the AUC/MIC ratio is a general predictor of antibacterial efficacy for quinolones. Using an in-vitro kill curve technique, performed in quadruplicate, with nine antibiotic concentrations and three strains of P. aeruginosa with varying quinolone susceptibility, we constructed sigmoidal dose response curves for AUC 0 6.5 /MIC and area under the bacterial kill curve (AUBKC) or AUC 0 24 /MIC and log change in viable count at 24 h ( 24). For levofloxacin the log AUC 0 6.5 /MIC ratio to produce 50% of the maximal effect was 0.74 ± 0.13 (r 2 = 0.9435) for levofloxacin and 0.82 ± 0.06 (r 2 = 0.7935) for ciprofloxacin. The log AUC 0 24 /MIC ratio to produce 50% maximal effect was 1.58 ± 0.13 (r 2 = 0.7788) for levofloxacin and 1.37 ± 0.12 (r 2 = 0.7207) for ciprofloxacin. An AUC 0 24 /MIC ratio of 125 produced 85.4% of the maximal response with levofloxacin and 81.5% with ciprofloxacin. These data suggest that levofloxacin and ciprofloxacin have equivalent activity against P. aeruginosa at equivalent AUC/MIC ratios. Introduction The main pharmacodynamic predictor of antibacterial efficacy for ciprofloxacin is the ratio of drug exposure as defined by concentration and time divided by the ciprofloxacin susceptibility of the pathogen as indicated by the MIC. The use of in-vitro pharmacodynamic models has indicated that the area under the serum concentration time curve (AUC)/MIC ratio can be related to bacterial killing for ciprofloxacin, ofloxacin, levofloxacin and moxifloxacin (BAY 12-8039). 1 4 The parameter used to measure bacterial killing is of critical importance in these calculations and the area under the bacterial kill curve (AUBKC) or a derivative has gained some acceptance though other parameters are used. 2,4,6,7 In addition there are data to show that it is not crucial whether ciprofloxacin, levofloxacin and ofloxacin are dosed once or twice a day; their antibacterial effects are similar provided they have a similar AUC/MIC ratio. 1,8 However, some in-vitro and animal data favour maximum serum concentration (C max )/MIC ratios as determining outcome but this effect may be of secondary importance to the AUC/ MIC. 9 11 Furthermore, human data from ITU pneumonia and COPD exacerbations indicate that an AUC/MIC ratio of 125 is related to pathogen clearance and clinical outcome. 12,13 From these data it can be speculated that, provided quinolones have the same AUC/MIC ratio, they will be equally effective in laboratory models or clinical practice. However, this assumes that all quinolones kill different bacterial species in an equivalent way. This is not the case with pneumococci even if the MICs are similar. 14 The purpose of this study was to compare the bactericidal activity of levofloxacin and ciprofloxacin against Pseudo - monas aeruginosa by using a sigmoid exposure response model, to determine the 50% AUC/MIC ratios for lethal effect for each agent. *Corresponding author. Tel: 44-117-9595652; Fax: 44-117-9593154. 1999 The British Society for Antimicrobial Chemotherapy 345

A. P. MacGowan et al. Materials and methods Three bacterial strains were used with varying quinolone susceptibilities, P. aeruginosa strain 5761 being the most susceptible, strain 11683 of intermediate susceptibility, and strain 8545 the least susceptible. MICs were determined for levofloxacin and ciprofloxacin in a macrobroth dilution method using Isosensitest broth and increases in drug concentration of 0.1 mg/l. Levofloxacin (Hoechst Marion Roussel, Uxbridge, UK) and ciprofloxacin (Bayer PLC, Newbury, UK) were used. In the time kill curve experiments they were used at concentrations of 0.25, 0.5, 1, 2, 3, 4, 5, 6 and 10 mg/l. The kill curves were performed by a modification of established technique. 15 Twentymillilitre Isosensitest broths (Unipath, Basingstoke, UK), were inoculated to give a final inoculum of 10 6 cfu/ml. For each antimicrobial a set of ten broths was incorporated with the required antibiotic concentration plus a growth control. These were sampled and diluted as necessary, and viable counts were performed using a spiral plater (Spiral System; Don Whitley, Shipley, UK) at time 0, then every 30 min up to 2.5 h, then hourly up to 6.5 h, and finally at 24 h. Counts were obtained on nutrient agar plates containing 0.1% magnesium chloride which inactivated any residual antibiotic. Viable counts were read manually after incubation at 37 C for 24 h. Each kill curve was performed in quadruplicate. Data analysis Time kill curves were drawn by plotting log cfu/ml against time in hours. The AUBKC for 0 6.5 h was calculated using the Graph Pad Prism software package (Graph Pad Software Inc., San Diego, CA, USA). The AUBKC was determined by the trapezoid rule. Drug exposure was calculated by multiplying the quinolone concentrations by 6.5 h or 24 h then dividing by the MIC for the strain to give an AUC 0 6.5 /MIC ratio or AUC 0 24 /MIC ratio. A sigmoidal dose response with HILLSLOPE 1.0 was fitted to the data for log AUC 0 6.5 /MIC and AUBKC or log AUC 0 24 /MIC and log change in viable count at 24 h ( 24). The minimum and maximum values, log AUC/MIC producing the 50% response between the maximum and minimum responses, and r 2 were calculated. The percentage of maximal response produced by an AUC 0 24 /MIC of 125 was read off the dose response curve for each drug. Results The three strains of P. aeruginosa used had levofloxacin MICs of 0.8 mg/l (strain 5761), 0.12 mg/l (strain 11683) and 4 mg/l (strain 8545); the equivalent MICs of ciprofloxacin were 0.09 mg/l (5761), 0.6 mg/l (11683) and 1.2 mg/l (8545). As might be expected because of lower ciprofloxacin MIC values, the degree of killing at equivalent quinolone concentrations in the kill curves was usually significantly superior with ciprofloxacin than with levofloxacin (Table). With strain 5761, the most quinolone-susceptible, this was only apparent at concentrations of 0.5 mg/l as at greater concentrations the strain was rapidly killed by both drugs. For strain 11683, which was of intermediate susceptibility, and strain 8545, the least susceptible, both agents were equivalent at low concentration as little killing occurred Table. Area under the bacterial time kill curve (AUBKC) for three strains of P. aeruginosa exposed to levofloxacin or ciprofloxacin at fixed concentrations for 6.5 h AUBKC (log cfu/ml.h) strain 5761 strain 11683 strain 8545 Concentration (mg/l) levofloxacin ciprofloxacin levofloxacin ciprofloxacin levofloxacin ciprofloxacin 0.25 26.1 0.5 6.5 3.3 a 32.2 1.4 34.5 0.7 36.8 1.2 34.4 1.0 0.5 10.6 0.9 3.8 0.7 a 29.6 0.7 29.5 1.3 34.8 0.8 32.6 1.0 1 4.2 0.6 2.9 1.0 27.7 1.1 10.7 2.5 a 33.0 0.9 26.8 2.6 a 2 2.6 0.3 2.2 0.5 21.6 0.8 3.2 0.6 a 31.8 0.9 15.9 2.1 a 3 2.1 0.1 1.9 0.5 14.4 0.8 2.2 0.4 a 29.8 1.6 9.5 3.3 a 4 2.1 0.1 1.7 0.4 11.9 4.4 2.0 0.7 a 24.7 1.4 6.5 1.4 a 5 1.8 0.3 1.7 0.3 8.7 3.6 1.4 0.1 a 20.3 1.5 5.1 1.0 a 6 1.7 0.2 1.4 0.1 5.5 2.6 1.4 0.1 a 18.0 2.6 3.7 1.0 a 10 1.5 0.2 1.4 0.1 4.0 1.0 1.4 0.1 a 7.2 3.6 2.4 0.3 a a P 0.05, Mann Whitney U test. Respective levofloxacin and ciprofloxacin MICs were as follows: strain 5761, 0.8 and 0.09 mg/l; strain 11683, 1.2 and 0.6 mg/l; strain 8545, 4 and 1.2 mg/l. 346

Levofloxacin vs ciprofloxacin against P. aeruginosa with either drug, but at higher concentrations ciprofloxacin was more bactericidal than levofloxacin as measured by AUBKC (Table). A dose response curve was fitted to the data using AUC 0 6.5 /MIC ratio and AUBKC. This allowed the drugs bactericidal properties to be compared taking into account ciprofloxacin s lower MIC. The log AUC 0 6.6 /MIC to produce 50% of the maximal effect was 0.74 0.13 (95% CI, 0.49 0.99; r 2 0.9435) for levofloxacin and 0.82 0.06 (95% CI, 0.69 0.94; r 2 0.7935) for ciprofloxacin. The maximum AUBKC for the levofloxacin AUC 0 6.5 /MIC responses was 41.3 3.8 (95% CI, 33.9 48.8) and the minimum 1.6 1.8 (95% CI, 5.2 to 2.0). Equivalent values for ciprofloxacin were: maximum, 47.4 2.3 (95% CI, 42.8 51.9); minimum 4.8 1.1 (95% CI, 6.9 to 2.6) (Figure 1). A similar log AUC 0 24 /MIC response curve was calculated using 24. The log AUC 0 24 /MIC ratio required to produce 50% of the maximum growth was 1.58 0.13 (95% CI, 1.33 1.84, r 2 0.7788) for levofloxacin and 1.37 0.12 (95% CI, 1.13 1.61, r 2 0.7207) for ciprofloxacin. The maximum 24 was 3.8 0.6 (95% CI, 2.6 5.0) for levofloxacin and 5.0 1.0 (95% CI, 3.0 7.0) for ciprofloxacin. Minimum values were 6.3 0.7 (95% CI, 7.7 5.0) for levofloxacin and 4.8 0.3 (95% CI, 4.4 5.5) for ciprofloxacin (Figure 2). An AUC 0 24 /MIC of 125 produced 85.4% of the maximal response for levofloxacin and 81.5% for ciprofloxacin. Discussion It is already known that levofloxacin has higher MIC values than ciprofloxacin and is less bactericidal at equivalent concentrations against P. aeruginosa. However, the main determinant of outcome with quinolones is the drug exposure modified by the pathogen s susceptibility. Pharmacokinetic data indicate that a single 500 mg dose of levofloxacin produces an AUC of 45.6 mg/l/h, while an equivalent dose of ciprofloxacin produces one of 9.9 mg/l/h. 16,17 Multiple doses of 500 mg levofloxacin and ciprofloxacin also produce AUC 0 24 values which are approximately four times greater for levofloxacin than for ciprofloxacin. 18,19 Ciprofloxacin is about four times more active than levofloxacin in terms of mean MIC, MIC 50 and MIC 90 for P. aeruginosa, including strains isolated in the UK (Fish & Chow; 18 D. Felmingham, personal communication), hence, it is likely that, in the treatment of P. aeruginosa, levofloxacin and ciprofloxacin will have similar AUC/MIC ratios. Data from in-vitro models using P. aeruginosa indicate that for ofloxacin and ciprofloxacin the AUC/MIC thresholds required to produce a maximum bacterial effect as measured by AUBKC are the same. 4 However, when ofloxacin and ciprofloxacin were modelled Figure 1. (a) Levofloxacin and (b) ciprofloxacin against three strains of P. aeruginosa: relationship between AUC 0 6.5 /MIC and AUBKC. Figure 2. (a) Levofloxacin and (b) ciprofloxacin against three strains of P. aeruginosa: relationship between AUC 0 24 /MIC and 24; an AUC/MIC ratio of 125 is marked by the vertical broken line. 347

A. P. MacGowan et al. simulating pharmacologically achievable concentrations, then ciprofloxacin was superior to ofloxacin, but AUC/ MIC ratios were not comparable and favoured ciprofloxacin. 5 Data using Staphylococcus aureus would also support the concept that, provided the AUC/MIC ratios were similar, levofloxacin and ciprofloxacin have similar bactericidal activity in an in-vitro model. 8 A sigmoid E max model has been used previously to study the bactericidal activity of ciprofloxacin but the rate of kill was compared with the concentration/mic ratio so the data are not comparable to our own. 20 These data indicate that, provided the AUC/MIC ratios are equivalent, then the bactericidal activities of levofloxacin and ciprofloxacin are equivalent, hence an AUC 24 /MIC ratio of 125 produced an 80 85% maximal response with both agents and the log E 50 are equivalent. This may not be true of all pathogen groups as it has been suggested that levofloxacin is more bactericidal than ciprofloxacin against Streptococcus pneumoniae. 14 Our data, in conjunction with many other findings, suggest that in the therapy of P. aeruginosa, ciprofloxacin and levofloxacin will be equally active, provided the drug exposure/mic ratios are the same. This would seem likely given the pharmacokinetics and in-vitro susceptibilities of these agents. Acknowledgement We would like to thank Mr M. Barlow of Hoechst Marion Roussel for his support in performing this study. References 1. MacGowan, A. P. & Bowker, K. E. (1998). Sequential antimicrobial therapy, pharmacokinetic and pharmacodynamic considerations in sequential therapy. Journal of Infection (in press). 2. Bowker, K. E., Wootton, M., Holt, H. A., Reeves, D. S. & MacGowan, A. P. (1997). Bactericidal activity of Bay 12-8039 against Streptococcus pneumoniae explored using an in vitro continuous bacterial culture model. In Program and Abstracts of the Thirty-Seventh Interscience Conference on Antimicrobial Agents and Chemotherapy, Toronto, 1997. Abstract F-134, p. 169. American Society for Microbiology, Washington, DC. 3. Madaras-Kelly, K. J., Larsson, A. J. & Rotschafer, J. C. (1996). A pharmacodynamic evaluation of ciprofloxacin and ofloxacin against two strains of Pseudomonas aeruginosa. Journal of Antimicrobial Chemotherapy 37. 703 10. 4. Madaras-Kelly, K. J., Ostergaard, B. E., Hovde, L. B. & Rotschafer, J. C. (1996). Twenty-four-hour area under the concentration time curve/mic ratio as a generic predictor of fluoroquinolone antimicrobial effect by using three strains of Pseudomonas aeruginosa and an in vitro pharmacodynamic model. Antimicrobial Agents and Chemotherapy 40, 627 32. 5. Firsov, A. A., Vostrov, S. N., Shevchenko, A. A. & Cornaglia, G. (1997). Parameters of bacterial killing and regrowth kinetics and antimicrobial effect examined in terms of area under the concentration time curve relationships: action of ciprofloxacin against Escherichia coli in an in vitro dynamic model. Antimicrobial Agents and Chemotherapy 41, 1281 7. 6. MacGowan, A. P., Wootton, M., Hedges, A. J., Bowker, K. E., Holt, H. A. & Reeves, D. S. (1996). A new time kill method of assessing the relative efficacy of antimicrobial agents alone and in combination developed using a representative β-lactam, aminoglycoside and fluoroquinolone. Journal of Antimicrobial Chemotherapy 38, 193 203. 7. Bowker, K. E., Holt, H. A., Reeves, D. S. & MacGowan, A. P. (1996). Pharmacodynamics of meropenem as explored by use of a single compartment in vitro continuous bacterial culture model to simulate once, twice and three times a day dosing. In Program and Abstracts of the Thirty-Sixth Interscience Conference on Antimicrobial Agents and Chemotherapy, New Orleans, LA, 1996. Abstract 46, p. 9. American Society for Microbiology, Washington, DC. 8. Kang, S. L., Rybak, M. J., McGrath, B. J., Kaatz, G. W. & Seo, S. M. (1994). Pharmacodynamics of levofloxacin, ofloxacin and ciprofloxacin, alone and in combination with rifampin, against methicillin susceptible and resistant Staphylococcus aureus in an in vitro infection model. Antimicrobial Agents and Chemotherapy 38, 2702 9. 9. Blaser, J., Stone, B. B., Groner, M. C. & Zinner, S. H. (1987). Comparative study with enoxacin and netilmicin in a pharmacodynamic model to determine importance of ratio of antibiotic peak concentration to MIC for bactericidal activity and emergence of resistance. Antimicrobial Agents and Chemotherapy 31, 1054 60. 10. Drusano, G. L., Johnson, D. E., Rosen, M. & Standford, H. C. (1993). Pharmacodynamics of a fluoroquinolone antimicrobial agent in a neutropenic rat model of Pseudomonas sepsis. Antimicrobial Agents and Chemotherapy 37, 483 90. 11. Sullivan, M. C., Cooper, B. W., Nightingale, C. H., Quintiliani, R. & Lawlor, M. T. (1993). Evaluation of the efficacy of ciprofloxacin against Streptococcus pneumoniae by using a mouse protection model. Antimicrobial Agents and Chemotherapy 37, 234 9. 12. Forrest, A., Nix, D. E., Ballow, C. H., Goss, T. F., Birmingham, M. C. & Schentag, J. J. (1993). Pharmacodynamics of intravenous ciprofloxacin in seriously ill patients. Antimicrobial Agents and Chemotherapy 37, 1073 81. 13. Forrest, A., Amantea, M., Collins, D. A., Chodosh, S. & Schentag, J. J. (1993). Pharmacodynamics of oral OPC 17116 in patients with acute bacterial exacerbations of chronic bronchitis. In Program and Abstracts of the Thirty-Third Interscience Conference on Antimicrobial Agents and Chemotherapy, New Orleans, LA, 1993. Abstract 81, p. 134. American Society for Microbiology, Washington, DC. 14. George, J. & Morrissey, I. (1997). The bactericidal activity of levofloxacin and other fluoroquinolones against Streptococcus pneumoniae. Clinical Microbiology and Infection 3, Suppl. 2, 293. 15. Krogstad, D. J. & Moellering, R. C. (1986). Antimicrobial combinations. In Antibiotics in Laboratory Medicine, 2nd edn (Lorian, V., Ed.), pp. 537 95, William & Wilkins, Baltimore, MD. 16. Holland, M. L., Chien, S.-C., Corrado. M. L. et al. (1994). The pharmacokinetic profile of levofloxacin following once-or twice-daily 500 mg oral administration of levofloxacin hemihydrate. In Abstracts of the Fifth International Symposium on New Quinolones, Singapore, 1994. Abstract 25. 348

Levofloxacin vs ciprofloxacin against P. aeruginosa 17. Bergan, T., Thorsteinsson, S. B., Solberg, R., Bjornskau, L., Kolstad, I. M. & Johnsen, S. (1987). Pharmacokinetics of ciprofloxacin: intravenous and increasing oral doses. American Journal of Medicine, 82, Suppl. 4A, 97 102. 18. Fish, D. N. & Chow, A. T. (1997). The clinical pharmacokinetics of levofloxacin. Clinical Pharmacokinetics 32, 101 19. 19. Wilson, A. P. R. & Grüneberg, R. N. (Eds) (1997). C i p r o flo x a c i n: 10 years of clinical experience, pp. 67 87. Maxim Medical, Oxford. 20. Hyatt, J. M., Nix, D. E., Stratton, C. W. & Schentag, J. J. (1995). In vitro pharmacodynamics of piperacillin, piperacillin tazobactam, and ciprofloxacin alone and in combination against Staphylococcus aureus, Klebsiella pneumoniae, Enterobacter cloacae and Pseudomonas aeruginosa. Antimicrobial Agents and Chemotherapy 39, 1711 6. Received 3 February 1998; returned 23 March 1998; revised 7 April 1998; accepted 7 May 1998 349