Comparative Activities of Ciprofloxacin and Levofloxacin against Streptococcus pneumoniae in an In Vitro Dynamic Model

Abstract

The activities of levofloxacin (500 mg every 24 h) and ciprofloxacin (750 mg every 12 h) against six pneumococcal isolates in an in vitro dynamic model were compared. For one strain, levofloxacin reduced the inoculum by over 4 log CFU/ml and ciprofloxacin reduced the inoculum by over 2 log CFU/ml. For four isolates, both drugs reduced inocula by 4 log CFU/ml within 6 h, suggesting that this dose of ciprofloxacin should be as effective as levofloxacin against these pneumococci.

Penicillin resistance among clinical isolates of Streptococcus pneumoniae is becoming increasingly common in the United States as well as in most countries of the world (1). Many of these organisms are also resistant to erythromycin, azithromycin, and clarithromycin and for eradication require higher concentrations of most cephalosporins. One of the advantages of the extended-spectrum fluoroquinolone antibiotics is their enhanced activity against pneumococci, including penicillin-resistant strains (K. P. Klugman and T. Capper, Abstr. 35th Intersci. Conf. Antimicrob. Agents Chemother., abstr. E9, p. 87, 1995). Levofloxacin, the l isomer of ofloxacin, when given once daily, achieves good results in the treatment of respiratory infections caused by these organisms (3). Ciprofloxacin, which has been less well accepted for use in the treatment of pneumococcal infections, is given in twice-daily doses.

In this study, the pharmacodynamics of levofloxacin and ciprofloxacin were compared, using an in vitro dynamic model that mimics the pharmacokinetics and dosing of the antibiotics in humans (2). The efficacy of a simulated ciprofloxacin oral dose of 750 mg every 12 h against six clinical isolates of S. pneumoniae with differing susceptibilities to penicillin and macrolides was compared with that of a simulated levofloxacin oral dose of 500 mg every 24 h.

The MICs and minimum bactericidal concentrations of ciprofloxacin and levofloxacin for the six pneumococcal strains (kindly provided by John Lonks) are presented below (see Fig. 2). Two strains (3190 and 5056A) were resistant to both penicillin and erythromycin, two (6691 and 2309A) were penicillin resistant and erythromycin susceptible, and two (18032 and 9742) were susceptible to both antibiotics. These organisms were introduced into a two-compartment artificial capillary model that simulates human pharmacokinetics in vitro and exposes bacteria to changing concentrations of antibiotics (2).

FIG. 2.

Means of duplicate bacterial counts for six isolates of S. pneumoniae in an in vitro model. (a) Simulated levofloxacin dosing at 500 mg every 24 h; (b) simulated ciprofloxacin dosing at 750 mg every 12 h. MBC, minimum bactericidal concentration.

The model consists of a central compartment, which mimics levels of antimicrobial agents in serum, and six bioreactors, artificial capillary chambers (Unisyn Fibertech Corporation, San Diego, Calif.) connected in series, which constitute the peripheral compartment. Growth-phase bacteria were introduced and incubated to a density of 2 × 10⁵ to 5 × 10⁶, at which time the antibiotic was introduced into the central compartment (time zero). At the end of a 60-min infusion, the drug concentration reached a maximum, analogous to the C_max that is reached after oral administration. Antibiotic-free, sterile Mueller-Hinton broth was infused and eliminated at a rate selected to mimic the drugs' half-life. The antibiotic, whose concentration was changing, was pumped rapidly to the six bioreactors, where it diffused through the capillary walls into the bacterium-containing medium.

The central compartment remained sterile during the experiments. Its contents were sampled hourly for the first 8 h and then every 2 h. Drug concentrations were determined in the samples by bioassay, using well plates seeded with Escherichia coli ATCC 25922. The peripheral compartments were sampled at 2-h intervals for 24 h (except for h 16 or 18) to determine drug concentrations and bacterial counts. Colonies were counted by plating these samples as well as by filtration plating.

Central-compartment concentrations of the two antibiotics over time are shown in Fig. 1. In the absence of antibiotics, bacterial counts increased from a mean value of 6.10 log CFU/ml to 7.08 at 2 h, 8.33 at 4 h, and 0.04 at 8 h (data not shown). Figure 2 shows the results of antibiotic-induced killing with levofloxacin (Fig. 2a) and ciprofloxacin (Fig. 2b). Strain 18032 (resistant to both ciprofloxacin and levofloxacin; Pen^s Ery^s) was not killed by either drug. For the other five strains, levofloxacin produced a reduction of at least 4 log CFU/ml by 6 h. Ciprofloxacin reduced the bacterial counts of four strains by at least 4 log CFU/ml by 6 h, but for one strain (5056A; Pen^r Ery^r) it reduced the inoculum by only 2.7 log CFU/ml by the same time point. Levofloxacin reduced this organism's count by more than 4 log CFU/ml by 6 h. Strain 6691 (Pen^r Ery^s) was initially killed by ciprofloxacin but showed regrowth to 2 to 3 log CFU by 24 h. For five of the six organisms, the MICs of ciprofloxacin did not change after drug exposure. One strain (6691) did show an increase in MIC to 8 to 16 mg/liter, but on subsequent passages on blood agar plates there was a quick reversion to the original MIC.

FIG. 1.

Time course of central-compartment antibiotic concentrations. (a) Levofloxacin, simulated at 500 mg every 24 h; (b) ciprofloxacin, simulated at 750 mg every 12 h.

Levofloxacin has been introduced for the treatment of pneumococcal infections of the respiratory tract (3). Although ciprofloxacin was not initially approved for the treatment of pneumococcal pneumonia, it has been approved for treatment of upper respiratory tract infections caused by these organisms. In the in vitro model described here, ciprofloxacin given at a simulated dose of 750 mg every 12 h produced a level of killing similar to that caused by levofloxacin administered at 500 mg every 24 h for four of six tested pneumococcal isolates (at least for the first 6 to 8 h). These data suggest that the two drugs should produce similar clinical effects for most strains of S. pneumoniae. The clinical activities of these drugs in the doses studied here should be compared in clinical studies.