TEXT

Top Abstract Text References

Combination therapy is used with the aim of expanding the antimicrobial spectrum, minimizing toxicity, preventing the emergence of resistant mutants during therapy, and obtaining synergistic antimicrobial activity (4). Gatifloxacin (GAT) is a broad-spectrum 8-methoxy fluoroquinolone (5). Quinolones, including GAT, are less active against pseudomonads and other nonfermentative gram-negative bacteria than against members of the family Enterobacteriaceae. While quinolones may be used singly to treat pseudomonal infections, their use with either an aminoglycoside or aztreonam to treat serious Pseudomonas aeruginosa infections is recommended (9). In this study, we examined the killing rate of GAT alone and in combination with other antimicrobial agents against P. aeruginosa and related species.

Thirty clinical strains representing P. aeruginosa, Pseudomonas stutzeri, Burkholderia cepacia, Stenotrophomonas maltophilia, and Acinetobacter spp. were studied. The strains were selected for their different susceptibilities to GAT and included strains at the GAT breakpoints (i.e., MICs of 2 to 8 µg/ml). The nonquinolones, selected for combination studies with GAT, were chosen based on susceptibility testing and treatment recommendations from the NCCLS (12), The Medical Letter (10), and Sanford's Guide to Antimicrobial Therapy (15). GAT was obtained from Kyorin Pharmacutical Co. Ltd, Tochigi, Japan. Cefepime (FEP), aztreonam (ATM), amikacin (AMK), and imipenem (IPM) were prepared at Bristol-Myers Squibb squibb Co., New Brunswick, N.J., or Candiac, Canada. Ceftazidime (CAZ) was from Glaxo Pharmaceuticals, Research Triangle Park, N.C.; trimethoprim-sulfamethoxazole (SXT) was from Hoffmann-La Roche Inc., Nutley, N.J.; ticarcillin-clavulanate (TIM) was from Eli Lilly & Co., Indianapolis, Ind.; piperacillin (PIP), cefoperazone (CFP), chloramphenicol (CHL), and minocycline (MIN) were from Sigma Chemical Co., St. Louis, Mo.

MICs were determined by a broth microdilution method outlined by the NCCLS (12). Time-kill kinetics were performed in Mueller-Hinton broth. All studies involving SXT included 0.2 U of thymidine phosphorylase/ml in the test medium. The 20-ml cultures, grown in 50-ml glass flasks, were incubated at 35°C with shaking. Cells were grown to logarithmic phase with 1 h of preincubation in fresh broth prior to the addition of drugs. In each case, a growth control (i.e., no drug addition) was included. The starting bacterial density was approximately 5 × 10⁵ to 1 × 10⁶ CFU/ml. The viable counts at 0 and 24 h of incubation for each antibiotic tested singly or in combination were determined. Bacterial counts were determined by plating 50-µl samples (directly or 10-fold serial dilutions of the culture sample in saline) onto Mueller-Hinton agar using a spiral plater system (Spiral Biotech, Bethesda, Md.). Results from prior studies indicated that any drug carryover, using the spiral plating system, did not affect the bacterial cell count (6). GAT concentrations tested were at two times the MIC of the strain, not to exceed its NCCLS-approved MIC susceptible breakpoint of 2 µg/ml. Nonquinolones were tested at concentrations equal to their MICs against the test strain, not to exceed their respective NCCLS-approved MIC susceptible breakpoints. The combination included GAT (at two times the GAT MIC) and a nonquinolone (at the nonquinolone MIC); however, in no case did the drug concentration tested exceed the NCCLS-approved susceptible breakpoint of the test drug (12). Synergy was defined as a 2-log₁₀ or more decrease in viable count with the drug combination versus with the more active of the pair at 24 h of drug exposure (8). Conversely, antagonism was defined as a 2-log₁₀ or more increase in viable count with the drug combination compared to the less active of the pair at 24 h of drug exposure (4).

Eight P. aeruginosa strains were evaluated, including two that were intermediately susceptible to GAT (i.e., MIC, 4 µg/ml). Synergistic killing was observed most often (75%) with GAT in combination with IPM, ATM, or PIP (Tables 1 and 2). In addition, ~40 to 60% of the P. aeruginosa strains were synergistically killed with GAT plus CFP or CAZ. In a few cases, synergism occurred with GAT-nonquinolone combinations even though the strain was nonsusceptible to one of the two agents.

While P. stutzeri strains are more susceptible to GAT, synergism was observed in three of the four strains exposed to GAT plus ATM, PIP, or AMK and in two of the four strains treated with GAT plus FEP, CAZ, or CFP (Tables 1 and 2).

Synergistic killing occurred against >80% of the B. cepacia strains with the GAT-CAZ combination and 67% with GAT-ATM (Tables 2 and 3). Many of the B. cepacia strains synergistically killed with these two drug combinations were nonsusceptible to GAT and/or the -lactam.

Similarly, enhanced killing occurred in two-thirds of the S. maltophilia strains when GAT was mixed with CAZ, TIM, or ATM (Tables 2 and 4). This synergy was observed even though most S. maltophilia strains are CAZ and ATM resistant. Synergism was also observed against two of the four Acinetobacter strains with GAT combined with CAZ, ATM, or TIM (Tables 2 and 4).

Antagonism occurred with only one strain. The antagonism observed with P. aeruginosa A22678 exposed to GAT-IPM is likely the result of the development of resistance to IPM. When P. aeruginosa A22678 was exposed to IPM alone or in combination with GAT, the IPM MIC increased from 2 to 8 to 16 µg/ml.

These results show that GAT in combination with a -lactam was often synergistic against nonfermentative bacteria. While IPM was synergistic with GAT against P. aeruginosa, ATM and CAZ were also synergistic with GAT against other nonfermentative bacilli. CAZ was the only extended-spectrum cephalosporin evaluated against all of the strains in this study. However, when the three extended-spectrum cephalosporins were tested against the pseudomonads, CFP, FEP, and CAZ behaved similarly (Tables 1 and 2). PIP and TIM also yielded synergy with GAT against the nonfermentative bacilli. Though the AMK-GAT combination was synergistic against pseudomonads, the combination was less frequently synergistic than a GAT--lactam combination against P. aeruginosa.

The synergism observed with GAT has been reported with other quinolones. Ciprofloxacin was synergistic with IPM or CAZ against ~40% of P. aeruginosa strains (3, 7), against 33% with ATM (11), and against 15% with AMK (7). As summarized in the review papers by Neu (13, 14), fluoroquinolone-aminoglycoside combinations were primarily indifferent compared to a ciprofloxacin-antipseudomonal penicillin or -imipenem combination, which resulted in synergy against 20 to 50% of P. aeruginosa strains. The in vitro synergy of ciprofloxacin plus azlocillin corresponded to the enhanced efficacy of this combination in animal model infections (2, 11). In time-kill studies involving 20 nonfermenters, synergism was noticed when trovafloxacin was combined with CAZ, AMK, or IMP (17). Combinations of levofloxacin or ciprofloxacin with CAZ or CFP were often synergistic against S. maltophilia (16). Though synergism was exhibited against only one of the four Acinetobacter strains in the present study when it was exposed to the GAT-AMK pair, synergism with AMK was observed with levofloxacin, ofloxacin, and ciprofloxacin against all quinolone-susceptible Acinetobacter strains in the Bajaksouzian et. al. study (1). This difference in synergism rates might be due to the differences in the drug concentrations used in the time-kill studies. Bajaksouzian et al. tested AMK at concentrations equal to one-half to one-fourth of the MIC, whereas the present study tested AMK at the MIC.

In summary, GAT is like other quinolones in being synergistic with AMK and -lactams in killing nonfermenters. This synergism was observed against some strains that were nonsusceptible to GAT and/or the nonquinolone.