Therapy with antimicrobial combinations has been used as long as antimicrobials have been available. Combinations of antibiotics are often used to take advantage of different mechanisms of action and/or toxicity profiles. Well established indications for combination antimicrobial therapy include: (a) empirical treatment of life-threatening infections; (b) treatment of polymicrobial infections; (c) prevention of the emergence of bacterial resistance; and (d) for synergism. Disadvantages of combination therapy include: (a) increased expense; (b) increased risk of adverse effects; (c) antagonism; and (d) superinfection. Combination antimicrobial therapy should be considered for the treatment of serious Gram-negative infections caused by Enterobacter cloacae, Pseudomonas aeruginosa and Serratia marcescens, and certain Gram-positive infections caused by Enterococcus spp. and Staphylococcus spp. Selection of agents should be dependent upon local susceptibility patterns, clinical experience, site of infection, potential toxicities and cost.

We adapted an in vitro pharmacodynamic model of infection to incorporate simulated endocardial vegetations. The bactericidal activities of teicoplanin, vancomycin, gentamicin, and various combinations of these drugs were studied against a strain of methicillin-susceptible Staphylococcus aureus obtained from a patient being treated for endocarditis at Detroit Receiving Hospital. Bacteria were grown overnight, concentrated, and added to a mixture of cryoprecipitate (80%o)

The pharmacodynamic properties of levofloxacin (an optically active isomer of ofloxacin), ofloxacin, and ciprofloxacin, alone and in combination with rifampin, were evaluated over 24 to 48 h against clinical isolates of methicillin-susceptible and -resistant Staphylococcus aureus (MSSA 1199 and MRSA 494, respectively) energy-dependent efilux process mediated by the norA gene appeared to be responsible for the resistance observed. Our data suggest that with levofloxacin there is a more rapid onset of bactericidal activity than with ofloxacin or ciprofloxacin against MSSA 1199 and that the activity of levofloxacin is similar to that of ofloxacin but better than that of ciprofloxacin against MRSA 494. Resistance was noted only after exposure to the low dose of ciprofloxacin. Resistance to ofloxacin did not develop even when the pharmacokinetics of the drug were set to equal those of ciprofloxacin, suggesting that ofloxacin differs from ciprofloxacin irrespective of time of exposure. The resistance to ciprofloxacin that developed in our in vitro model may be mediated by the cfxc-ofr locus, which has been shown to be associated with low-level fluoroquinolone resistance. Overall, levofloxacin demonstrated potent bactericidal activity against S. aureus, without the emergence of resistance in our infection model. Quinolones dosed once daily were more effective than equivalent dosages administered twice daily. The addition of rifampin was not synergistic but prevented the emergence of ciprofloxacin resistance.

No abstract

The pharmacodynamic effects of extended imipenem dosing intervals were studied against two strains of Pseudomonas aeruginosa (ATCC 27853 and an imipenem-resistant mutant, 27853R) in an in vitro model of infection. Imipenem was administered as monotherapy (simulated 1-g bolus every 8 or every 12 h) and in combination with amikacin (7.5-mg/kg bolus every 12 h or a 15-mg/kg bolus once). Monotherapy with imipenem administered every 8 h was equally bactericidal at 24 h compared with regimens combined with amikacin for ATCC 27853. Imipenem administered every 12 h against the sensitive strain and both imipenem monotherapy regimens against the resistant strain demonstrated regrowth at 24 h. Although both amikacin regimens administered as monotherapy resulted in rapid bacterial killing activity with respect to time to a 99.9% reduction in log10 CFU/milliliter, regrowth at 24 h was observed at levels reaching or exceeding the initial inoculum. All combination regimens resulted in no detectable growth by 24 h regardless of dosing interval for either drug or initial susceptibility to imipenem. Results from this study indicate the potential for several novel dosing regimens against P. aeruginosa. Monotherapy with imipenem, 1 g every 8 h, was effective against a sensitive strain of P. aeruginosa. Combination therapy with imipenem and once-daily or twice-daily amikacin resulted in increased killing activity against imipenem-resistant P. aeruginosa. Once-daily or twice-daily amikacin in combination therapy, regardless of P. aeruginosa susceptibility, allowed for extension of imipenem dosing intervals.

The postantibiotic effect (PAE) following three consecutive 2-h exposures to imipenem, temafloxacin, and tobramycin was determined in Pseudomonas aeruginosa. A PAE and a bactericidal effect were consistently observed for imipenem following each cycle of drug exposure and regrowth. In contrast, the PAE increased with repeated exposure with temafloxacin (1.8 to >5 h), but disappeared with tobramycin by the third exposure (0.9 to 0 h). These data show that the in vitro PAE may change within a strain following multiple cycles of drug exposure and bacterial regrowth. Bacterial inocula were prepared from a 20-h overnight growth of three to five colonies picked from an agar plate, diluted 1:1,000 in MHB-S, and incubated for 3 h at 37°C to bring the organisms into a log-linear growth phase. Five to six replicates of antimicrobial agent-exposed cultures were performed for each drug. The starting inoculum for each drug-bacterium combination was adjusted to give a 1 x 106 to 1 x 107 CFU/ml in a total volume of 10 ml of MHB-S. The concentrations of antibiotics were chosen such that a PAE of 1 to 2 h would be produced after the first cycle of drug exposure. Pilot experiments determined that exposure to lOx the MIC (20 Fxg/ml) of imipenem, 4x the MIC (8 ,ug/ml) of temafloxacin, and 2x the MIC (2 ,g/ml) of tobramycin for 2 h at 37°C resulted in PAEs within this range for the test strains. Bacteria were incubated in tubes with or without drugs (control) for 2 h.Following a 2-h incubation of drug plus bacteria, the tubes were centrifuged at 1,200 x g for 15 min, and 9 ml of supernatant was removed. The bacteria contained in the remaining 1-ml aliquot were resuspended in 9 ml of fresh, prewarmed, drug-free MHB-S and were gently vortexed. This procedure was repeated three times. To ensure comparability of growth patterns of the control and the drugexposed cultures, the control-growth tube was diluted in drug-free MHB-S after the washing procedure to bring the bacterial concentration in the control-growth tube to the same level as that in the drug-exposed tube.The effects of a second and a third cycle of drug exposure and removal on the PAE were assessed when drug-exposed cultures had regrown to the level of the initial inoculum prior to the first cycle of drug exposure. One milliliter of a concentrated stock solution of drug was added to the culture tube containing regrowing bacteria at 6 to 8 h (second exposure) or 14 to 16 h (third exposure). The mixture was vortexed and incubated for 2 h. After the 2-h incubation, drug-containing and control cultures were processed and samples were collected as described above following the first exposure.The number of bacteria was quantified by removing 100 pl from the culture at hourly intervals for the first 4 h and after the drug was removed. Samples were serially diluted with cold 0.9% sodium chloride, and 20 pl was plated onto MHA in triplicate. The plates were incubated for 18 to 24 h, and the colonies were counted. When bacterial counts were ex-1723

The pharmacodynamics of once-daily amikacin administered as monotherapy and in combination with aztreonam, ceftazidime, and cefepime against Pseudomonas aeruginosa ATCC 27853 and clinical isolate 16690 (moderately susceptible to ceftazidime) were investigated with an in vitro model of infection over a 24-h period. Monotherapy with aztreonam, ceftazidime, and cefepime and combinations of aztreonam with cefepime or ceftazidime were also studied. MICs and MBCs were determined for viable organisms at 24 h to test for the development of resistance. Once-daily amikacin demonstrated killing activity over the initial 8 h superior to those of all other drugs administered as monotherapy against both strains tested (P < 0.01). Regrowth by 24 h was greatest for the amikacin regimen (P < 0.01) but was apparent for all monotherapy regimens against both strains. No changes in susceptibilities were observed. All combination therapies containing once-daily amikacin achieved 99.9% reductions in log10 CFU/ml by 2.0 h against both strains, with no regrowth of organisms at 24 h. Aztreonam-cefepime and -ceftazidime combinations required approximately 5 to 6 h to achieve a 99.9% reduction in log10 CFU/ml. Although there was no difference in time to the 99.9% kill between the aztreonam-cefepime and -ceftazidime regimens against either strain, the killing activity of these combinations was significantly less than those of regimens containing once-daily amikacin (P < 0.01). Minor differences in the initial susceptibilities of beta-lactams and the monobactam aztreonam against P. aeruginosa may not be important for therapeutic outcomes when used in combination with single-daily aminoglycoside therapy.

Recent controversy surrounding the activity of monoclonal antibodies against endotoxin highlights the necessity of identifying all factors associated with increased mortality, one of which is endotoxin concentrations. Antibiotics may induce different patterns of endotoxin release. We compared the release of free endotoxin (in endotoxin units per milliliter) over 6 h and changes in numbers of CFU of exponentially growing Escherichia coli and Pseudomonas aeruginosa (10(6) to 10(7) CFU/ml) cultured in chemically defined endotoxin-free broth combined with pooled human serum and/or 10 micrograms of E5 immunoglobulin M monoclonal antibody per ml. MICs and MBCs were tested in each medium at the same inoculum. The inoculum was exposed to antibiotics at a single fixed multiple of the MIC for each medium (range, two to eight times the MIC). E5 antibody had no effect on MICs, MBCs, bactericidal activity, or endotoxin release. In the presence of 50% serum, amikacin, ceftazidime, imipenem, and ofloxacin each killed equivalent amounts of E. coli over 6 h; however, ceftazidime induced the highest release of endotoxin. Amikacin and ofloxacin produced the most favorable ratio of endotoxin release to amount of bacterial killing. In the presence of 50% serum, ceftazidime and imipenem reduced the P. aeruginosa inoculum to the greatest extent over 6 h. Although its bactericidal activity was diminished, ofloxacin caused the lowest release of free endotoxin. Imipenem and ofloxacin showed similar low ratios of endotoxin release to bacterial killing. In summary, antibiotic class, presence of serum, and type of organism influenced bactericidal activity and endotoxin release.