Activity of Meropenem with and without Ciprofloxacin and Colistin against Pseudomonas aeruginosa and Acinetobacter baumannii

Abstract

Time-kill synergy studies showed that at 24 h, subinhibitory meropenem and ciprofloxacin concentrations of 0.06 to 128 and 0.03 to 32 μg/ml, respectively, showed synergy against 34/51 Pseudomonas aeruginosa strains; subinhibitory concentrations of meropenem (0.06 to 8 μg/ml) and colistin (0.12 to 1 μg/ml) showed synergy against 13 isolates. Subinhibitory meropenem and ciprofloxacin concentrations of 0.25 to 2 and 0.12 to 16 μg/ml, respectively, showed synergy against 18/52 Acinetobacter baumannii strains at 24 h. Subinhibitory meropenem and colistin concentrations of 0.03 to 64 and 0.06 to 8 μg/ml, respectively, showed synergy against 49 strains at 24 h.

Systemic infections, sometimes of a life-threatening nature, caused by gram-negative nonfermentative bacilli are increasingly encountered, especially in immunocompromised and debilitated hosts. The problem is exacerbated by the increasing drug resistance of these organisms. While carbapenems remain the drugs of choice for the treatment of infections caused by most of these organisms (with the exception of Stenotrophomonas maltophilia and some Chryseobacterium species), carbapenem resistance has developed in these organisms (4, 9, 14, 16-19, 22, 23).

Pseudomonas aeruginosa and Acinetobacter baumannii are two of the most important gram-negative nonfermentative human pathogens. Carbapenem resistance occurs commonly in P. aeruginosa, but most A. baumannii strains remain susceptible to this group of drugs (4, 9, 11-14, 16-19, 21-23). If these organisms become panresistant to all available antibacterials (including carbapenems), the only currently available drug is colistin, a compound whose use was discontinued several decades ago because of potential toxicity.

In the absence of new compounds to treat these panresistant organisms, other therapeutic alternatives must be sought. Previous studies have demonstrated synergistic activity between β-lactams and quinolones against various species of gram-negative and gram-positive bacilli by time-kill experiments (1-3, 5, 7, 25-27). The present study investigated possible synergy between meropenem, a commonly used carbapenem with a wide spectrum of activity, and ciprofloxacin (the usual quinolone chosen for the treatment of P. aeruginosa and other gram-negative infections) or colistin against P. aeruginosa and A. baumannii by synergy time-kill experiments. Checkerboard titrations were not used because we and others have found this method to be not as discriminatory as time-kill studies (1-3).

The bacteria tested comprised 51 P. aeruginosa isolates. Fifty of these were recently isolated from clinical specimens at Hershey Medical Center (some from cystic fibrosis and otherwise critically ill patients) and identified by standard methodology (20). One additional VIM-2 β-lactamase-producing P. aeruginosa isolate was included and was obtained from Ronald Jones (JMI Laboratories, North Liberty, IA).

A total of 52 A. baumannii strains were studied. Fifty of these were isolated from clinical specimens at the University of Cologne, Cologne, Germany. Two additional meropenem-resistant A. baumannii strains were studied. One of these strains produced OXA-24 and OXA-51 β-lactamases (JMI Laboratories). The resistance mechanism in the second strain, which was obtained from Kenneth Thomson (Creighton University Medical School, Omaha, NE), was not identified. All 103 organisms were isolated from different patients, avoiding replicates from the same patient.

Identification of the A. baumannii strains to the species level was confirmed by amplified ribosomal RNA gene restriction analysis (24). To exclude copy strains, A. baumannii strains were included on the basis of exhibiting a unique randomly amplified polymorphic DNA analysis profile (15). The latter technique was not applied to P. aeruginosa isolates. Organisms were frozen at −70°C in double-strength milk (Difco Laboratories, Detroit, MI) until testing. Meropenem and ciprofloxacin powder were obtained from their respective manufacturers, and colistin sulfate was obtained from Sigma, St. Louis, MO.

The Clinical and Laboratory Standards Institute-approved microdilution method (6) was used to test the MICs of all of the drugs alone against each of the 103 organisms. All of the compounds were tested alone and in combination at concentrations up to four times above and four times below the MICs of the drugs tested alone. Inoculum concentrations were >5 × 10⁵ CFU/ml. In our experience, relatively high inoculum concentrations (up to 5 × 10⁶ CFU/ml) are sometimes necessary for time-kill testing of gram-negative bacilli. The inoculum used conforms to normal usage and has been used by our group in many previous studies (1, 2, 5, 7, 25-27). For this study, synergy was defined as a ≥2-log₁₀ decrease in the number of CFU per milliliter between the combination and its more active component after 3, 6, 12, and 24 h, with the number of surviving organisms in the presence of the combination being ≥2 log₁₀ CFU/ml below the starting inoculum. At least one of the drugs was present at a concentration which did not significantly affect the growth curve of the organism when used alone (8). Addition was defined as the effect of a second drug being similar to that of the single more effective compound, and antagonism was defined as the combination yielding colony counts higher than those seen with the more active single drug alone (8). The minimum countable number of CFU per milliliter was approximately 300, and drug carryover was addressed by dilution, as we have previously described (1-3, 5, 7, 25-27). Synergy time-kill analysis was not done when the MIC of ciprofloxacin or colistin was ≥128 μg/ml. The chi-square test was used to compare proportions of killing synergy between groups by using Yates continuity correction in two-by-two tables (10).

For the MICs and synergy time-kill data obtained with the 51 P. aeruginosa isolates, see Table S1 in the supplemental material. The MICs (micrograms per milliliter) of drugs alone were as follows: meropenem, 0.12 to 256; ciprofloxacin, 0.06 to 64; colistin, 0.25 to 4. Meropenem plus ciprofloxacin, at 3 h, yielded synergy against 7 isolates at subinhibitory concentrations (micrograms per milliliter) of meropenem of 0.03 to 4 and ciprofloxacin of 0.03 to 0.25; at 6 h, the drugs showed synergy against15 isolates with subinhibitory meropenem concentrations of 0.03 to 8 μg/ml and ciprofloxacin concentrations (mostly sub-MICs) of 0.03 to 2 μg/ml. At 12 h, the drugs showed synergy against 38 isolates at subinhibitory meropenem concentrations of 0.06 to 128 μg/ml and ciprofloxacin concentrations (mostly sub-MICs) of 0.03 to 32 μg/ml. At 24 h, the drugs showed synergy against 34 isolates at subinhibitory meropenem and ciprofloxacin concentrations of 0.06 to 128 and 0.03 to 32 μg/ml, respectively. When meropenem was combined with colistin, they showed synergy against 26 isolates at 3 h with subinhibitory meropenem and colistin concentrations of 0.03 to 64 and 0.12 to 1 μg/ml, respectively. At 6 h, the drugs showed synergy against 23 isolates with subinhibitory concentrations of meropenem of 0.06 to 64 μg/ml and colistin of 0.12 to 1 μg/ml. After 12 h, the drugs showed synergy against 24 isolates with subinhibitory meropenem and colistin concentrations of 0.03 to 4 and 0.12 to 1 μg/ml, respectively. After 24 h, the drugs showed synergy against 13 isolates with subinhibitory concentrations of meropenem of 0.06 to 8 μg/ml and colistin of 0.12 to 1 μg/ml.

There were six P. aeruginosa isolates for which the meropenem MICs were ≥16 μg/ml. Meropenem and ciprofloxacin showed synergy against two of them between 3 and 24 h, and meropenem and colistin showed synergy against four of them at meropenem MICs of 4 to 8 μg/ml. The meropenem MIC was 256 μg/ml for the carbapenem-resistant, VIM-2-producing isolate, and ciprofloxacin and colistin showed synergy against it at a meropenem MIC of ≥64 μg/ml.

For the MIC and time-kill data obtained with 52 A. baumannii strains, see Table S2 in the supplemental material. The meropenem-ciprofloxacin combination was tested against 40 strains, and the meropenem-colistin combination was tested against 51 strains. The MICs (micrograms per milliliter) of the drugs alone were as follows: meropenem, 0.12 to 256; ciprofloxacin, 0.06 to 256; colistin, 0.12 to 128. Meropenem plus ciprofloxacin, at 3 h, yielded synergy at subinhibitory concentrations (micrograms per milliliter) of meropenem (0.25) and ciprofloxacin (0.12 to 0.25) for 2 strains; at 6 h, the drugs showed synergy against 10 strains at subinhibitory meropenem concentrations of 0.03 to 4 μg/ml and ciprofloxacin MICs (mostly sub-MICs) of 0.03 to 8 μg/ml. At 12 h, the drugs showed synergy against 26 strains at subinhibitory meropenem concentrations (0.03 to 4 μg/ml) and ciprofloxacin concentrations (mostly subinhibitory) of 0.03 to 16 μg/ml. At 24 h, the drugs showed synergy against 18 strains at subinhibitory meropenem and ciprofloxacin concentrations of 0.25 to 2 and 0.12 to 16 μg/ml, respectively. When meropenem was combined with colistin, the drugs showed synergy against 20 strains at 3 h at subinhibitory meropenem and colistin concentrations of 0.03 to 64 and 0.03 to 16 μg/ml, respectively. At 6 h, the drugs showed synergy against 39 strains at subinhibitory concentrations of meropenem of 0.03 to 64 μg/ml and colistin of 0.06 to 8 μg/ml. After 12 h, the drugs showed synergy against 50 strains at subinhibitory meropenem and colistin concentrations of 0.03 to 64 and 0.06 to 16 μg/ml, respectively. After 24 h, the drugs showed synergy against 49 strains at subinhibitory concentrations of meropenem of 0.03 to 64 μg/ml and colistin of 0.06 to 8 μg/ml.

For six A. baumannii strains, the meropenem MICs were ≥16 μg/ml. For two of these, for which the meropenem MIC was 32 μg/ml, the meropenem-colistin combination at 8.0 and ≤2.0 μg/ml, respectively, produced synergistic killing after 3 h. For the remaining four strains, against which synergy was observed, the meropenem or colistin MICs in the combination were too high to be clinically achievable.

Figure 1 depicts the percentage of organisms against which the antibiotic combinations showed additive or synergistic effects over time. Figure 1 shows that the combination of meropenem and ciprofloxacin gave more synergistic killing of P. aeruginosa isolates than did the meropenem-colistin combination by 12 and 24 h (P < 0.001). The meropenem-ciprofloxacin combination was synergistic against a significant percentage of the strains of A. baumannii at 12 h (P < 0.001), and this effect then declined by 24 h. The combination of meropenem and colistin showed increased synergic killing of A. baumannii compared to meropenem-ciprofloxacin at 6, 12, and 24 h (P < 0.001).

FIG. 1.

Time-kill synergy studies of meropenem combined with ciprofloxacin (A) and colistin (B) against 51 isolates of P. aeruginosa and of meropenem combined with ciprofloxacin (C) and colistin (D) against 41 and 51 strains of A. baumannii, respectively. In each pair of bars, the one on the left represents an additive effect and the one on the right represents synergy. **, Significantly higher percentage of organisms showing synergy (P < 0.001).

In vitro detection of synergy between two antibacterials has not been properly standardized. It has been shown by our group and others (1-3, 26, 27) that unless synergy is very clear, as in the case of quinolones and cephalosporins against S. maltophilia, time-kill synergy testing is more discriminatory (but more time-consuming) than is checkerboard analysis. Bajaksouzian et al. (2) have shown synergy between levofloxacin and amikacin against Acinetobacter species by time-kill analysis, while Visalli and coworkers (26) have shown synergy between levofloxacin and cefpirome, ceftazidime, gentamicin, and meropenem against P. aeruginosa by time-kill analysis. In none of the above studies (including the present one) has the mechanism of synergy been elucidated; this work will be the subject of future studies. We do not know the clinical significance of synergistic combinations when one or both MICs are not easily clinically achievable.

Our results confirm and extend the use of time-kill analysis to detect synergy between meropenem and ciprofloxacin or colistin at subinhibitory concentrations. In an era when the developmental pipeline for new antibacterials active against gram-negative bacilli is, for all practical purposes, dry, combination therapy is the only therapeutic option for resistant organisms until new compounds are discovered and developed.