This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Louie, A.
Right arrow Articles by Drusano, G. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Louie, A.
Right arrow Articles by Drusano, G. L.

 Previous Article  |  Next Article 

Antimicrobial Agents and Chemotherapy, July 2008, p. 2486-2496, Vol. 52, No. 7
0066-4804/08/$08.00+0     doi:10.1128/AAC.01439-07
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

Use of an In Vitro Pharmacodynamic Model To Derive a Linezolid Regimen That Optimizes Bacterial Kill and Prevents Emergence of Resistance in Bacillus anthracis{triangledown}

A. Louie,1* H. S. Heine,2 K. Kim,1 D. L. Brown,1 B. VanScoy,1 W. Liu,1 M. Kinzig-Schippers,3 F. Sörgel,3 and G. L. Drusano1

Ordway Research Institute, Albany, New York,1 U.S. Army Medical Research Institute of Infectious Diseases, Fort Detrick, Maryland,2 Institute for Biomedical and Pharmaceutical Research, Nürnberg-Heroldsberg, Germany3

Received 6 November 2007/ Returned for modification 21 December 2007/ Accepted 26 April 2008


arrow
ABSTRACT
 
Simulating the average non-protein-bound (free) human serum drug concentration-time profiles for linezolid in an in vitro pharmacodynamic model, we characterized the pharmacodynamic parameter(s) of linezolid predictive of kill and for prevention of resistance in Bacillus anthracis. In 10-day dose-ranging studies, the average exposure for ≥700 mg of linezolid given once daily (QD) resulted in >3-log CFU/ml declines in B. anthracis without resistance selection. Linezolid at ≤600 mg QD amplified for resistance. With twice-daily (q12h) dosing, linezolid at ≥500 mg q12 h was required for resistance prevention. In dose fractionation studies, killing of B. anthracis was predicted by the area under the time-concentration curve (AUC)/MIC ratio. However, resistance prevention was linked to the maximum serum drug concentration (Cmax)/MIC ratio. Monte Carlo simulations predicted that linezolid at 1,100 mg QD would produce in 96.7% of human subjects a free 24-h AUC that would match or exceed the average 24-h AUC of 78.5 mg·h/liter generated by linezolid at 700 mg QD while reproducing the shape of the concentration-time profile for this pharmacodynamically optimized regimen. However, linezolid at 700 mg q12h (cumulative daily dose of 1,400 mg) would produce an exposure that would equal or exceed the average free 24-h AUC of 90 mg·h/liter generated by linezolid at 500 mg q12h in 93.8% of human subjects. In conclusion, in our in vitro studies, the QD-administered, pharmacodynamically optimized regimen for linezolid killed drug-susceptible B. anthracis and prevented resistance emergence at lower dosages than q12h regimens. The lower dosage for the pharmacodynamically optimized regimen may decrease drug toxicity. Also, the QD administration schedule may improve patient compliance.


arrow
INTRODUCTION
 
In 2001, Bacillus anthracis was intentionally disseminated in the eastern United States through the national postal system (5, 13). Eleven people developed cutaneous anthrax. An additional 11 individuals developed inhalational anthrax; 5 of them died (4, 15, 16). Over 10,000 persons were prescribed antibiotic prophylaxis for possible exposure to anthrax spores (6).

In response to these events there is a renewed interest in expanding the armamentarium of antibiotics for the treatment of anthrax. As part of this initiative, our laboratory has been examining several agents for the postexposure prophylaxis and/or therapy of B. anthracis (10, 22, 23). To gain insight into the appropriate dose and schedule of administration for these agents, we have employed a hollow fiber (HF) pharmacokinetic/pharmacodynamic (PK/PD) model. This flexible and powerful system has been employed to study a broad range of bacterial and viral pathogens (11, 12, 14, 21-23, 38). The dose and schedule recommendations derived from HF PK/PD models have accurately predicted clinical outcomes in murine and nonhuman primate models of inhalational anthrax (10, 17).

Recently, we showed that the efficacy of the fluoroquinolone levofloxacin was comparable to that of ciprofloxacin for the treatment of B. anthracis in an HF PK/PD model and in a mouse model of inhalational anthrax (10). The evaluation for levofloxacin was instrumental in designing the nonhuman primate study that resulted in levofloxacin gaining an FDA indication for B. anthracis therapy and postexposure prophylaxis (10, 17). Predictions of therapeutic failure when dosed with murine pharmacokinetics were also prospectively validated. In that study, the selection of resistant mutants readily occurred with suboptimal fluoroquinolone dosages (10). Consequently, it is prudent to examine drugs from classes other than the fluoroquinolones for therapeutic intervention.

Linezolid is an oxazolidinone antibiotic that is active against B. anthracis. This drug is available in both oral and intravenous formulations (31), making it attractive for use as postexposure prophylaxis and treatment of B. anthracis in mass casualty scenarios and for treatment of critically ill patients. Linezolid prevents the formation of the 70S ribosomal complex in bacteria, inhibiting protein synthesis (37). B. anthracis elaborates lethal toxin, edema factor, and protective antigen as part of the pathophysiological process (29). While eradicating the pathogen is a necessary part of therapy, it would also be advantageous to have the therapy interrupt toxin production. In vitro studies by Stevens et al. (36) showed linezolid and clindamycin, another protein synthesis inhibitor, reduced the production of toxins by Staphylococcus aureus, including alpha-hemolysin, toxic shock syndrome toxin 1, and Panton-Valentine leukocidin. In contrast, the cell wall-active antibiotic nafcillin increased toxin concentrations in the cell culture medium, perhaps due to release of intracellular toxins with death of the bacterium. Clindamycin was also associated with decreased toxin production compared with the non-protein synthesis inhibitor penicillin and with better survival in mice infected with Clostridium perfringens and in animals and people infected with group A streptococci (34, 35, 39). The effect of linezolid on the production of toxins by B. anthracis has not been examined. However, since it is likely that the protein synthesis inhibitor linezolid will reduce the production of B. anthracis toxins, we felt that it would be important to examine linezolid as an alternative therapy to fluoroquinolone antimicrobials for B. anthracis.


arrow
MATERIALS AND METHODS
 
Bacteria, media, susceptibility testing, and mutation frequency for resistance. The {Delta}-Sterne strain of B. anthracis was evaluated. This strain lacks the pX01 and pX02 virulence plasmids containing the toxin and capsule genes, respectively. Linezolid powder was kindly supplied by Pfizer, Inc. (New York, NY).

MICs of linezolid were determined simultaneously by broth macrodilution and agar dilution methods in Mueller-Hinton II broth (MHB) and Mueller-Hinton II agar (MHA) using the methods described by CLSI (8). MICs were read after 24 h of incubation at 35°C. Trailing end points were observed, in which the turbidity of B. anthracis in test tubes containing antibiotic-supplemented medium gradually decreased with increasing drug concentrations up to and including the highest linezolid concentration examined (32 µg/ml). Thus, after discussion with H. Heine, our coinvestigator at USAMRIID and a member of the CLSI advisory committee, the MIC was defined as the lowest linezolid dilution that resulted in ≥80% reduction in growth compared to the growth controls. Minimum bactericidal concentrations (MBCs) were determined using standard methods (30). To characterize the effect of protein binding on drug activity, MICs were determined simultaneously in MHB, MHB containing 50% complement-inactivated mouse serum, and MHB in 50% complement-inactivated human serum. Mutation frequencies to 1.5x the MIC of linezolid were determined in three trials.

In vitro hollow fiber PK/PD model. The HF PK/PD model described previously (10) was used to study the response of B. anthracis to linezolid exposures, simulating human pharmacokinetics. HF cartridges (FiberCell Systems, Frederick, MD) consist of bundles of HF capillaries encased within a plastic housing. The fibers have numerous pores that permit the passage of nutrients and low-molecular-weight species, such as antibiotics, but exclude bacteria. Approximately 15 ml of extracapillary space lies between the fibers and the cartridge housing. Medium within the central reservoir was continuously pumped through the HFs, and low-molecular-weight compounds rapidly equilibrated across the fibers with the extracapillary space. Thus, microorganisms that were inoculated into the extracapillary space were exposed to conditions approximating those that prevailed within the central reservoir.

Antibiotic was infused over 1 h into the central reservoir of the PK/PD systems at predetermined time points by syringe pumps. Antibiotic-containing medium was isovolumetrically replaced with drug-free medium, simulating a half-life of 5.5 h (31). The rate constant of elimination of antibiotic was the rate of fresh medium infusion divided by the volume of the medium in the total system. The system simulated a single-compartment model with exponential elimination.

For each experiment 15 ml of B. anthracis suspension (107 CFU/ml) was inoculated into the extracapillary space of several HF cartridges, and the experiment was initiated by infusing antibiotic. The PK profiles simulated for linezolid in the HF systems were extrapolated from average human PK values found on the manufacturer's drug package insert (Table 1). Approximately 31% of linezolid is bound to human serum proteins (31). The simulated PK profiles were for the non-protein-bound (free) fraction of linezolid. At predetermined time points an 800-µl sample of bacteria was collected from each HF system. Washed samples were quantitatively cultured onto drug-free MHA (total organisms) and MHA containing 1.3x the MIC of linezolid (resistant organisms). Media samples taken from the central reservoir over the first 48 h were assayed for linezolid concentrations to validate that the desired pharmacokinetic profiles were achieved. The measured drug concentrations were within 10% of the targeted values (data not shown).


View this table:
[in this window]
[in a new window]

  		TABLE 1. Steady-state targeted PK/PD parameters from simulated dose-ranging and dose fractionation studies with the HF PK/PD modela

Linezolid concentration determinations. Samples obtained from each treatment arm were stored at –80°C until they were assayed for their linezolid concentrations. Linezolid concentrations in MHB were measured by high-performance liquid chromatography using the methods described by Li et al. (20). The standard curve ranged from 0.1 mg/liter to 30 mg/liter. UV detection occurred at 253 nm, with storing spectra between 190 and 300 nm. Between-day coefficients of variation were 4.54% at 0.3 mg/liter, 5.46% at 2.5 mg/liter, and 6.47% at 20 mg/liter. The limits of detection and quantification were 0.01 mg/liter and 0.1 mg/liter.

Hollow fiber system experimental design. The linezolid dose-ranging experiment consisted of a no-treatment control arm and eight systems in which the mean serum concentration-time profiles and, hence, the average exposures for regimens of 100 through 800 mg of linezolid given once every 24 h (q24h) were simulated. Bacterial samples were taken from each HF system over the 10-day experiment. The samples were washed twice to prevent drug carryover and were plated onto drug-free MHA and MHA containing 1.3x the MIC of linezolid for enumeration of the total population and the population resistant to ≥1.3x the baseline MIC, respectively.

Two dose fractionation studies were performed to identify the pharmacodynamic parameter(s) linked with kill of the wild-type population and for suppression of resistance. In the first dose fractionation HF study, the free 24-h AUCs for two linezolid doses from the steep part of the dose-response curve derived from the above dose-ranging experiments were fractionated. In the second HF experiment, the free 24-h AUC exposure for a total daily dose that did not select for resistance in the dose-ranging studies was fractionated. Sampling for bacterial counts was as indicated above, except sampling continued for 15 days in order to better characterize the effect of each treatment regimen on prevention of resistance.

Finally, in another dose-ranging HF experiment, the efficacies of regimens which simulated the mean serum drug concentration-time profiles and, hence, average exposures for linezolid at 200, 400, and 500 mg q12h were compared with the simulated regimen for the standard FDA-approved regimen of linezolid at 600 mg given every 12 h. The efficacies of simulated regimens for ciprofloxacin at 500 mg given every 12 h and the pharmacodynamically optimized regimen of linezolid at 700 mg given once every 24 h also were compared.

Modeling of the exposure-response relationship of the total bacterial population and the resistant bacterial subpopulation to drug pressure. To mathematically describe the effects of different doses and schedules of administration of linezolid on the kill of the wild-type B. anthracis population and on the growth of drug-resistant subpopulations, three simultaneous parallel inhomogeneous differential equations, shown below, were used to describe the time course of linezolid concentrations and the total and resistant subpopulations.

The different drug treatment regimens were simultaneously comodeled in a population sense using the population modeling program BigNPAG (18). Bayesian estimates were generated for each regimen.

A mathematical model was constructed to describe exposure-response relationships of a mixed linezolid-sensitive and -resistant bacterial population in the HF PK/PD model under differing amounts of antimicrobial pressure. The mass balance equations (three parallel first-order inhomogeneous differential equations) that describe the linezolid-sensitive and -resistant subpopulations are described in equations 1 to 3.

Formula 1(1)

Formula 2(2)

Formula 3(3)

Formula 4(4)

Formula 5(5)
Equation 1 describes drug pharmacokinetics in the HF system (a standard one-compartment open model with zero-order input and first-order elimination). X1 is the amount of drug in the central compartment. R(1) is a zero-order, time-delimited infusion into the central compartment (in mg/h). SCL (in liters/h) is the rate of clearance of drug from the central compartment; Vc is the volume of the central compartment.

Equations 2 and 3 describe the rates of change of the linezolid-sensitive and -resistant subpopulations, respectively, over time. The model equations for describing the rate of change of the numbers of microorganisms in the sensitive and resistant bacterial subpopulations were developed based on the in vitro observation that bacteria in the HF system are in logarithmic growth phase in the absence of drug and exhibit an exponential density-limited growth rate (equation 4). There is one equation to describe the sensitive population (equation 2) and one for the resistant bacterial subpopulation (equation 3). In each, first-order growth was assumed, up to a density limit. Each population has an independent growth rate constant (sensitive [Kgmax-S] and resistant [Kgmax-R]). As the microorganisms approach maximal bacterial density, they approach stationary phase. This is accomplished by multiplying the first-order growth terms by E (equation 4; a logistic growth term). The maximal bacterial density (POPMAX) is identified as part of the estimation process. Most of the information for identifying this parameter is derived from the bacterial growth in the control group.

Equations 2 and 3 also allow the antibacterial effects of the different drug doses administered to be modeled. For both sensitive and resistant populations, there is an independent effect of the drug dose on the two populations, one mediated through equation 2 (sensitive population) and one through equation 3 (resistant population). There is a maximal kill rate that the drug can induce for each population (Kkmax-S and Kkmax-R). The killing effect of the drug was modeled as a saturable kinetic event (M [equation 5]; M is separate for both sensitive [MS] and resistant [MR] subpopulations) that relates the kill rate to drug concentration, where H is the slope constant and EC50 (mg/liter) is the drug concentration at which the bacterial kill rate is half-maximal. Separate H and EC50 terms are provided for the sensitive and resistant populations. The drug effect observed is a balance between growth and death induced by the drug concentrations achieved. NS and NR are the number of colonies per ml in the sensitive and resistant populations, respectively.

Measured outputs were changes in total population (total = NS + NR) and resistant mutant (NR) densities, where the sensitive (S) and resistant (R) subpopulations were enumerated on drug-free plates and plates containing 1.3x the MIC of the drug, respectively.

Monte Carlo simulations. From dose-ranging and dose fractionation experiments, the lowest dosages of linezolid that optimally killed wild-type B. anthracis and prevented resistance amplification were identified for regimens in which the drug was given on a q24h and a q12h schedule of administration. The mean concentration-time profiles associated with these regimens were used to calculate the mean free 24-h area under the time-concentration curve values (AUCs) of linezolid generated by these exposures. Monte Carlo simulations (n = 9,999 virtual human subjects) were employed to determine the lowest linezolid q24h and q12h administered regimens that would achieve the free 24-h AUCs associated with these regimens in at least 90% of the virtual subjects. The ADAPT II package of programs of D'Argenio and Schumitzky was employed for the simulations (9). The model developed by Meagher and colleagues (25) was employed as the structural model and provided the point estimates of the parameters of the mean parameter vector and their variances. Both normal and log-normal models were evaluated. The distribution was chosen as a function of the fidelity with which the original parameter values and their dispersions were recreated by the simulations.


arrow
RESULTS
 
MIC, MBC, and mutational frequency to resistance. Using a broth microdilution method the MIC of linezolid for the {Delta}-Sterne strain was 1 mg/liter in one trial and 2 mg/liter in four trials. The MIC was 2 mg/liter in the four trials that were conducted using an agar dilution method. Thus, the MIC of 2 mg/liter was used for all the experiments conducted in the current project. The MBC was >32 mg/liter. Addition of 50% mouse or human serum had no effect on the broth MICs. The mutation frequency to 1.5x the MIC of linezolid ranged from –6.46 to –7.24 logs in three trials.

Dose-ranging experiment. B. anthracis in the control HF system grew well, starting at 107 CFU/ml, and approximated 1010 CFU/ml by day 3. A dose-effect response was seen, but regrowth occurred for all regimens except for the linezolid 700 and 800 mg q24h simulated regimens (Fig. 1). The 100, 200, 300, and 400 mg once-daily simulated exposures showed little microbiological activity. Nonetheless, the linezolid 200 through 400 mg q24h simulated regimens generated enough selective pressure to amplify resistant mutants, even in the absence of much overall cell kill (Fig. 2). The linezolid 500 and 600 mg q24h simulated regimens generated considerable antimicrobial kill followed by emergence of resistance between days 4 and 8 of therapy (Fig. 2F and G). The linezolid 700 and 800 mg q24h simulated regimens caused nearly a 3-log10 CFU/ml reduction in bacterial burden. The linezolid 700 mg q24h simulated regimen was the lowest linezolid exposure examined that did not amplify the drug-resistant B. anthracis subpopulation.


Figure 1
View larger version (14K):
[in this window]
[in a new window]

  		FIG. 1. Dose-ranging experiment showing the effects of increasing doses of linezolid on the total B. anthracis population with linezolid administered once every 24 h.


Figure 2
View larger version (16K):
[in this window]
[in a new window]

  		FIG. 2. Effects of the linezolid regimens used in the dose-ranging experiment shown in Fig. 1 on the total B. anthracis populations and mutants with linezolid MICs that were at least 1.3x the MIC for the wild-type isolate. Results for linezolid at 800 mg q24h were similar to those for linezolid at 700 mg q24h (data not shown).

Dose fractionation experiments. In the first dose fractionation study, simulated free 24-h AUCs for the simulated regimens of linezolid at 500 and 600 mg q24h were given as the total 24-h AUC exposure once every 24 h, one-half of the total 24-h AUC exposure every 12 h, or one-sixth of the total 24-h AUC exposure every 4 h. The mean PK/PD exposures targeted for each regimen are shown in Table 1. The results of the experimental arm in which linezolid at 800 mg q24h was simulated confirmed that the fractionated regimens were on the steep part of the dose-response curve. In this dose fractionation experiment, the microbiological effect of linezolid was similar for all schedules of administration evaluated over the first 4 days of this study, indicating that the pharmacodynamic parameter linked with kill of the wild-type population was the AUC/MIC ratio (Fig. 3). Emergence of resistance was not detected during this time period. Thereafter, there was outgrowth for each of the fractionated regimens, with the least-fractionated regimens showing the longest time before emergence of resistance was identified (Fig. 3C to E). The highest dose simulated, linezolid at 800 mg q24h, killed B. anthracis without selecting for resistance. The data suggest that the pharmacodynamic parameter linked with prevention of resistance was the Cmax/MIC ratio.


Figure 3
View larger version (19K):
[in this window]
[in a new window]

  		FIG. 3. Dose fractionation of total daily AUCs generated for linezolid at 500 and 600 mg q24h. (A) Effects of each regimen, when given as one, two, or six equally divided doses each day, on the total B. anthracis population. (B to E) Effects of fractionation of the 600-mg dose of linezolid on total and resistant populations. (F) Effects of linezolid at 800 mg given once daily on these populations. The effects on resistance amplification of fractionating the 24-h AUC exposure generated for linezolid at 500 mg q24h are not shown.

In the second dose fractionation study, the 24-h AUC for the 800-mg/day dose of linezolid was studied in order to evaluate a total daily dose that did not select for resistance as one of the treatment arms. The control bacteria grew well. All the schedules of administration had virtually identical rates of kill for the first 3 to 4 days, before emergence of resistance was observed, again showing that the pharmacodynamic parameter linked with kill of the wild-type population was the AUC/MIC (Fig. 4). Thereafter, the regimen in which the total daily AUC exposure was given as four equally divided doses every 6 h lost its microbiological kill and, because of a rather steep amplification of the resistant mutant subpopulation starting at day 6, diverged from the other schedules of administration. The regimen in which the total daily exposure was given as two equally divided doses every 12 h tracked with the regimen in which the total daily exposure was administered once every 24 h through day 10, when it diverged. There was a slower rate of amplification of the resistant mutant subpopulation (relative to the q6h regimen), starting at day 6, that likely explained the difference. There was no emergence of resistance over the 15 days of this experiment for the linezolid at 800 mg q24h regimen, and a total kill exceeding 3 log10 CFU/ml was obtained (Fig. 4). Thus, the second dose fractionation study also demonstrated that the Cmax/MIC ratio was the pharmacodynamic parameter linked with prevention of resistance.


Figure 4
View larger version (19K):
[in this window]
[in a new window]

  		FIG. 4. Dose fractionation of the total daily AUC generated for linezolid at 800 mg q24h. (A) Effects of all of the schedules of administration evaluated on the total B. anthracis population. (B to E) Effects of individual schedules of administration on the total bacterial population and the subpopulation that has an MIC for linezolid that is at least 1.3x greater than the linezolid MIC of the wild-type B. anthracis strain.

MICs of B. anthracis isolates with decreased susceptibilities to linezolid. The MICs of mutants selected with suboptimal linezolid regimens in the dose-ranging and dose fractionation studies were determined. The MIC of linezolid for the wild-type B. anthracis isolate was 2 mg/liter, while the MICs of the mutants were 3 and 4 mg/liter. Colonies that grew on linezolid-supplemented agar in the mutation frequency experiments had MICs of 3 mg/liter.

Clinical dose experiment. The dose-ranging study reported above, in which different doses of linezolid were administered to HF systems once every 24 h, demonstrated that the simulated regimen of linezolid 700 mg given once every 24 h resulted in maximal kill of the wild-type B. anthracis population. The dose fractionation studies showed that the Cmax/MIC ratio was linked with prevention of resistance. Thus, linezolid at 700 mg given once every 24 h was the minimum pharmacodynamically optimized dosage and schedule of administration examined that maximally killed the total bacterial population and prevented emergence of resistance during therapy. Currently, linezolid is not FDA approved for the treatment of B. anthracis infections. However, linezolid at 600 mg given every 12 h is the FDA-licensed regimen for the treatment of serious staphylococcal infections in humans. Thus, using the in vitro PK/PD model, the efficacies of regimens which simulated the mean serum drug concentration-time profiles for linezolid at 200, 400 and 500 mg q12h were compared with both the clinically prescribed regimen of linezolid at 600 mg q12h and the pharmacodynamically optimized regimen of linezolid at 700 mg q24h. The outcomes of these simulated linezolid regimens were also compared with the efficacy of a simulated regimen for ciprofloxacin at 500 mg given every 12 h, the FDA-approved drug and regimen for the treatment and postexposure prophylaxis of B. anthracis infections in humans. As shown in Fig. 5, the simulated regimens of linezolid at 200 and 400 mg q12h failed due to emergence of resistance during therapy. Linezolid at 500 mg q24h, and the clinically prescribed regimen of linezolid at 600 mg q12h, resulted in amounts of bacterial kill over 15 days similar to that with the pharmacodynamically optimized regimen of linezolid at 700 mg q24h and the standard ciprofloxacin regimen of 500 mg q12h. Emergence of resistance was not seen with any of those regimens (Fig. 5).


Figure 5
View larger version (18K):
[in this window]
[in a new window]

  		FIG. 5. Comparison of the treatment efficacy of the pharmacodynamically derived regimen of linezolid at 700 mg q24h with the simulated regimens for linezolid at 200, 400, and 500 mg q12h in an in vitro PK/PD model. The clinically prescribed regimens of ciprofloxacin at 500 mg given every 12 h and linezolid at 600 mg given every 12 h are also compared. (A) Effects of each regimen on the total B. anthracis population. (B to E) Effects of selected regimens on the total B. anthracis population and on amplification of mutants with decreased susceptibilities to the evaluated drug.

Mathematical modeling of the linezolid 800-mg/day dose fractionation experiment. The three administration schedules from the dose fractionation experiment were modeled in a population sense. The mean parameter values and their interquartile ranges for the model are presented in Table 2. The model fit the data well. The predicted-observed plots for the drug concentration, for the total bacterial population, and for the resistant subpopulation are presented in Fig. 6A to C, respectively. Importantly, the schedule of administration demonstrated its impact on the value for the maximal killing rate for the resistant population (Kkmax-R). For once-daily administration, the post-Bayesian estimates for the 2.5% through 97.5% percentiles was 0.636 to 0.643 h–1 for Kkmax-R. The ranges for the every-12-h and every-6-h administration schedules were 0.475 to 0.482 and 0.489 to 0.496 h–1, respectively. While every-6- and every-12-h administrations were significantly different, but not very different numerically, both were quite different from the once-daily value. This correlates with late emergence of resistance for the fractionated regimens, while no resistance was identified for the once-daily regimen.


View this table:
[in this window]
[in a new window]

  		TABLE 2. Values for the mathematical model describing effects of linezolid concentrations on the total population and resistant subpopulation of B. anthracis


Figure 6
View larger version (17K):
[in this window]
[in a new window]

  		FIG. 6. Predicted-observed plots after the Bayesian step for linezolid concentrations (A), total Bacillus anthracis population (B), or the subpopulation of Bacillus anthracis resistant to linezolid (C).

Monte Carlo simulations. The dose-ranging and dose fractionation studies described above showed that the simulated regimen for linezolid at 700 mg q24h was the lowest drug exposure that was pharmacodynamically optimized to maximize kill of the wild-type B. anthracis strain and to prevent emergence of resistance during therapy. This once-daily administered regimen simulated a mean free 24-h AUC of 78.5 mg·h/liter within the in vitro pharmacodynamic model. Subsequent dose-ranging studies demonstrated that the simulated regimen for linezolid at 500 mg q12h was the lowest twice-daily administered regimen that was able to provide the same outcome benefits. That regimen simulated a mean free 24-h AUC of 90 mg·h/liter. Monte Carlo simulations predicted that linezolid regimens of 1,000 and 1,100 mg q24h would generate free 24-h AUCs that would achieve or exceed the mean free 24-h AUC of 78.5 mg·h/liter that was simulated for the linezolid regimen of 700 mg q24h in 87.1% and 96.7% of human subjects, respectively. For the twice-daily schedule of administration, a simulated regimen of linezolid at 600 mg q12h was predicted to generate a free 24-h AUC that would be equal to or greater than the free 24-h AUC of 90 mg·h/liter that was simulated for the linezolid regimen of 500 mg q12h in 72.9% of virtual human subjects. A linezolid regimen of 700 mg q12h (for a daily cumulative dose of 1,400 mg of linezolid) would achieve or exceed this free 24-h AUC exposure in 93.8% of human subjects.


arrow
DISCUSSION
 
In the fall and winter of 2001, B. anthracis was disseminated in the United States through the postal system. With those events, ciprofloxacin proved to be beneficial for the treatment and prophylaxis of cutaneous and inhalational anthrax (4, 15). However, B. anthracis mutants resistant to fluoroquinolones have been generated in laboratories (10, 22, 23), underscoring the need to identify other antimicrobial agents for the treatment of anthrax.

In this study, the oxazolidinone antibiotic linezolid was selected for study because it is active against B. anthracis. Also, its unique mechanism of action precludes cross-resistance with fluoroquinolones. Unlike the fluoroquinolones, linezolid is a protein synthesis inhibitor (37). Thus, in addition to its antimicrobial effect, it is likely that linezolid would decrease the production of the toxins protective antigen, edema factor, and lethal factor of B. anthracis. This may improve clinical outcomes.

Simulating the mean human drug concentration-time profiles for clinically used regimens of antibiotics in our in vitro PK/PD system, we demonstrated that the clinically prescribed regimen of linezolid at 600 mg administered every 12 h was as effective as ciprofloxacin at 500 mg given every 12 h. This linezolid regimen killed B. anthracis at the same rate as ciprofloxacin, with both antibiotics reducing the bacterial population by 3 to 4 log10 CFU/ml over 10 days. These findings suggest that linezolid is an effective alternative to ciprofloxacin for the treatment of B. anthracis infections.

In our dose-ranging studies, the simulated mean concentration-time profile for linezolid at 700 mg q24h was as effective in killing B. anthracis as the clinical regimen of linezolid at 600 mg q12h. Neither regimen amplified linezolid-resistant mutants. Doses of linezolid of ≤600 mg given once every 24 h selected for linezolid-resistant mutants, ultimately leading to treatment failure.

Previously, in a neutropenic murine thigh infection model, we demonstrated that the AUC/MIC ratio was the pharmacodynamic index linked with killing of Staphylococcus aureus (23). Others have used mouse thigh infection models to show that the AUC/MIC ratio is predictive of kill of Streptococcus pneumoniae for linezolid and that either the AUC/MIC or time above MIC was the pharmacodynamic index predictive of optimized kill of S. aureus (1). The PD index for linezolid linked with prevention of resistance was not examined in these in vivo infection models. However, results derived from an in vitro pharmacodynamic model suggested that prevention of emergence of resistance by several strains of methicillin-resistant S. aureus was concentration dependent. Those studies were not designed to differentiate whether prevention of resistance was linked to the AUC/MIC or Cmax/MIC ratios (3).

The dose fractionation studies that were conducted as part of the current project demonstrated that the AUC/MIC ratio was the pharmacodynamic index for linezolid that was most predictive of kill of wild-type B. anthracis. These dose fractionation studies also clearly showed that the Cmax/MIC ratio was the pharmacodynamic index for linezolid that was linked to prevention of emergence of resistance during therapy.

The pharmacodynamic parameter linked with prevention of resistance was identifiable as much as 13 days after treatment initiation. Traditionally, PD studies last for only 24 h and do not evaluate for emergence of resistance. The current study demonstrates the importance of evaluating kill of both the drug-susceptible and -resistant populations over longer treatment durations in order to better define treatment efficacy. In a mouse model of inhalational anthrax we showed that emergence of resistance that arose between 3 and 5 days after initiation of ciprofloxacin therapy was responsible for treatment failure and death of the infected hosts (10).

The mechanism for reduced susceptibility to linezolid by B. anthracis mutants has not been elucidated. In Staphylococcus aureus, resistance to linezolid is linked to G25765T and T2500A mutations in domain V of the 23S rRNA gene (26, 32). S. aureus has five copies of this gene (32). First-stage mutants of S. aureus have only one to two target genes mutated, resulting in an increase in the MIC for linezolid that is 1.5-fold higher than the MIC for wild-type isolates. S. aureus isolates with mutations in additional 23S rRNA genes have further increases in the MIC to linezolid. A search of the B. anthracis Ames genome (NC_003997) with the NCBI BLAST program using the 23S rRNA gene sequence (X68425) from S. aureus ATCC 12600 indicates that B. anthracis has 11 copies of this gene. In the current study, mutants with MICs that were 1.5- and 2-fold higher than for the parent strain were amplified with suboptimal linezolid exposures, suggesting that the mechanism of resistance to linezolid in B. anthracis is similar to those described for S. aureus.

The CDC recommends a 60-day antibiotic course for treatment of anthrax infections and for postexposure prophylaxis (7, 28), since people may become ill up to 43 days after they are exposed to B. anthracis spores (27). Thrombocytopenia, neutropenia, and anemia have been described with prolonged linezolid therapy (2, 31, 33). Seldomly, peripheral neuropathy and ocular neuritis are seen (19, 31). Importantly, linezolid therapy is rarely terminated because of hematologic abnormalities. Clinical studies demonstrate that linezolid toxicities are linked to duration of therapy, with toxicities rarely seen in people who receive less than 2 weeks of this antibiotic (2, 31). Total daily drug exposure also has a role, since reduction of linezolid from 600 mg given twice daily to 300 mg given twice daily led to improved hematological parameters in patients with linezolid-induced thrombocytopenia or neutropenia (2).

The toxicities of linezolid can be minimized by monitoring the blood cell counts once or twice weekly in patients who have received this drug for more than 2 weeks or by switching the patient to another class of antimicrobial agent 1 to 2 weeks after initiation of linezolid therapy, when antibiotic susceptibilities for the pathogen become available. By that time, the B. anthracis burden and the amount of toxin production by B. anthracis should be substantially reduced in infected hosts. The pharmacodynamic parameter(s) predictive of the toxicities of linezolid is unknown, although toxicity is clearly dependent on total daily dose and duration of therapy.

In our PK/PD model, optimal kill of wild-type B. anthracis without selection of resistance was seen in experimental arms in which the mean serum drug concentration-time profile for linezolid at 700 mg q24h was simulated. For regimens in which twice-daily schedules of administration for linezolid were simulated, linezolid at 500 mg q12h was the lowest dosage examined that achieved these therapeutic end points, representing a 42.8% increase in the total dose of drug administered each day. Monte Carlo simulations predicted that a clinical regimen of linezolid at 1,000 mg q24h would produce in 87.1% of human subjects the average non-protein-bound drug exposure and concentration-time profile simulated for the pharmacodynamically optimized regimen of linezolid at 700 mg given once every 24 h, while a clinical linezolid regimen of 1,100 mg q24h would achieve these PK/PD targets in approximately 96.7% of human recipients. Monte Carlo simulations also predicted that linezolid at 700 mg q12h (for a cumulative daily dose of 1,400 mg) would generate the average drug exposure and concentration-time profile for linezolid at 500 mg q12h in 93.8% of human subjects. Thus, the once-daily administered, pharmacodynamically derived linezolid regimen would optimize the kill of B. anthracis while minimizing the probability for resistance amplification, at a substantially lower cumulative daily dose than a regimen in which linezolid is administered on a 12-hour basis. Hence, the once-daily, pharmacodynamically optimized linezolid regimen may decrease the incidence of drug toxicity and should improve patient adherence to the prescribed antibiotic course. The pharmacodynamically optimized regimen for linezolid warrants further investigation.


arrow
ACKNOWLEDGMENTS
 
This work was supported by grant number P01AI060908 from the National Institute of Allergy and Infectious Diseases.

The content of this report is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute of Allergy and Infectious Diseases, the National Institutes of Health, or USAMRIID.

There are no conflicts of interest to disclose for any author.


arrow
FOOTNOTES
 
* Corresponding author. Mailing address: Ordway Research Institute, 150 New Scotland Avenue, Albany, NY 12208. Phone: (518) 641-6463. Fax: (518) 641-6304. E-mail: alouie{at}ordwayresearch.org Back

{triangledown} Published ahead of print on 5 May 2008. Back


arrow
REFERENCES
 
    1
  1. Andes, D., M. L. van Ogtrop, J. Peng, and W. A. Craig. 2002. In vivo pharmacodynamics of a new oxazolidinone (linezolid). Antimicrob. Agents Chemother. 46:3484-3489.[Abstract/Free Full Text]
    2. 2
  2. Bishop, E., S. Melvani, B. P. Howden, P. G. P. Charles, and M. L. Grayson. 2006. Good clinical outcomes but high rates of adverse reactions during linezolid therapy for serious infections: a proposed protocol for monitoring therapy in complex patients. Antimicrob. Agents Chemother. 50:1599-1602.[Abstract/Free Full Text]
    3. 3
  3. Boak, L. M., J. Li, C. R. Rayner, and R. L. Nation. 2007. Pharmacokinetic/pharmacodynamic factors influencing emergence of resistance to linezolid in an in vitro model. Antimicrob. Agents Chemother. 51:1287-1292.[Abstract/Free Full Text]
    4. 4
  4. Bush, L. M., B. H. Abrams, A. Beall, and C. C. Johnson. 2001. Index case of fatal inhalational anthrax due to bioterrorism in the United States. N. Engl. J. Med. 345:1607-1610.[Free Full Text]
    5. 5
  5. Centers for Disease Control and Prevention. 2001. Evaluation of Bacillus anthracis contamination inside the Brentwood mail processing and distribution center, District of Columbia. MMWR Morb. Mortal. Wkly. Rep. 50:1129-1132.
    6. 6
  6. Centers for Disease Control and Prevention. 2002. Evaluation of postexposure antibiotic prophylaxis to prevent anthrax. MMWR Morb. Mortal. Wkly. Rep. 51:59.[Medline]
    7. 7
  7. Centers for Disease Control and Prevention. 2001. Update: investigation of bioterrorism-related anthrax and interim guidelines for exposure management and antimicrobial therapy, October 2001. MMWR Morb. Mortal. Wkly. Rep. 50:909-919.[Medline]
    8. 8
  8. Clinical Laboratory Standards Institute. 2006. Document M7-A7. Methods for dilutional antimicrobial susceptibility testing for bacteria that grow aerobically; approved standard, 7th ed. Clinical Laboratory Standards Institute, Wayne, PA.
    9. 9
  9. D'Argenio, D. Z., and A. Schumitzk. 1997. ADAPT II. A program for simulation, identification, and optimal experimental design. User manual. Biomedical Simulations Resource, University of Southern California, Los Angeles, CA. http://bmsr.usc.edu/.
    10. 10
  10. Deziel, M. R., H. Heine, A. Louie, M. Kao, W. R. Byrne, J. Basset, L. Miller, K. Bush, M. Kelly, and G. L. Drusano. 2005. Identification of effective antimicrobial regimens for use in humans for the therapy of Bacillus anthracis infections and postexposure prophylaxis. Antimicrob. Agents Chemother. 49:5099-5106.[Abstract/Free Full Text]
    11. 11
  11. Drusano, G. L., J. A. Bilello, S. L. Preston, E. O'Mara, S. Kaul, S. Schnittman, and R. Echols. 2001. Hollow fiber unit evaluation of a new human immunodeficiency virus type 1 protease inhibitor, BMS-232632, for determination of the linked pharmacodynamic variable. J. Infect. Dis. 183:1126-1129.[CrossRef][Medline]
    12. 12
  12. Drusano, G. L., P. A. Bilello, W. T. Symonds, D. S. Stein, J. McDowell, A. Bye, and J. A. Bilello. 2002. Pharmacodynamics of abacavir in an in vitro hollow-fiber model system. Antimicrob. Agents Chemother. 46:464-470.[Abstract/Free Full Text]
    13. 13
  13. Greene, C. M., J. Reefhuuis, C. Tan, A. E. Fiore, S. Goldstein, M. J. Beach, S. C. Redd, D. Valiante, G. Burr, J. Buehler, R. W. Pinner, E. Bresnitz, B. P. Bell, et al. 2002. Epidemiologic investigations of bioterrorism-related anthrax, New Jersey, 2001. Emerg. Infect. Dis. 8:1048-1055.[Medline]
    14. 14
  14. Gumbo, T., A. Louie, W. Liu, D. Brown, P. G. Ambrose, S. M. Bhavnani, and G. L. Drusano. 2007. Isoniazid bactericidal activity and resistance emergence: integrating pharmacodynamics and pharmacogenomics to predict efficacy in different ethnic populations. Antimicrob. Agents Chemother. 51:2329-2336.[Abstract/Free Full Text]
    15. 15
  15. Inglesby, T., T. O'Toole, D. A. Henderson, J. G. Bartlett, M. S. Ascher, E. Eitzen, A. M. Friedlander, J. Gerberding, J. Hauser, J. Hughes, J. McDade, M. T. Osterholm, G. Parker, T. M. Perl, P. K. Russell, K. Tonat, et al. 2002. Anthrax as a biological weapon, 2002: updated recommendations for management. JAMA 287:2236-2252.[Abstract/Free Full Text]
    16. 16
  16. Jernigan, J. A., D. S. Stephen, D. A. Ashford, C. Omenaca, M. S. Topiel, M. Galbraith, M. Tapper, T. L. Fisk, S. Zaki, T. Popovic, R. F. Meyer, C. P. Quinn, S. A. Harper, S. K. Fridkin, J. J. Sejvar, C. W. Shepard, M. McConnell, J. Gaumer, W. J. Shieh, J. M. Malecki, J. L. Gerberding, J. M. Hughes, B. A. Perkins, et al. 2001. Bioterrorism-related inhalational anthrax: the first 10 cases reported in the United States. Emerg. Infect. Dis. 7:933-944.[Medline]
    17. 17
  17. Kao, L. M., K. Bush, R. Barnewell, J. Estep, F. W. Thalacker, P. H. Olson, G. L. Drusano, N., Minton, S. Chien, A. Hemeryck, and M. F. Kelley. 2006. Pharmacokinetic considerations and efficacy of levofloxacin in an inhalational anthrax (postexposure) rhesus monkey model. Antimicrob. Agents Chemother. 50:3535-3542.[Abstract/Free Full Text]
    18. 18
  18. Leary, R., R. Jelliffe, S. Schumitzky, and M. Van Guilder. 2001. An adaptive grid non-parametric approach to pharmacokinetic and dynamic (PK/PD) models, p.389-394. Proc. 14th IEEE Sympos. Computer-Based Med. Syst. IEEE Computer Society, Bethesda, MD.
    19. 19
  19. Lee, E., S. Burger, J. Shah, C. Melton, M. Mullen, F. Warren, and R. Press. 2003. Linezolid-associated toxic optic neutropathy; a report of 2 cases. Clin. Infect. Dis. 37:1389-1391.[CrossRef][Medline]
    20. 20
  20. Li, J., C. R. Rayner, S. Dixson, and R. L. Nation. 2004. Simple method for the assay of linezolid in brain heart infusion broth by high-performance liquid chromatography. Biomed. Chromatogr. 18:1-5.[CrossRef][Medline]
    21. 21
  21. Louie, A., M. R. Deziel, W. Liu, and G. L. Drusano. 2007. Impact of resistance selection and mutant growth fitness on the relative efficacies of streptomycin and levofloxacin for plague therapy. Antimicrob. Agents Chemother. 51:2661-2667.[Abstract/Free Full Text]
    22. 22
  22. Louie, A., K. Kim, D. Brown, B. VanScoy, R. Kulawy, W. Liu, M. Kinzig-Schippers, F. Sorgel, H. S. Heine, and G. L. Drusano. 2007. Dose-ranging and dose-fractionation studies of moxifloxacin against Bacillus anthracis, abstr. A-1442, p.36. Abstr. 47th Intersci. Conf. Antimicrob. Agents Chemother., 2007.
    23. 23
  23. Louie, A., K. Kim, D. Brown, B. VanScoy, R. Kulawy, V. Hsu, C. Fregeau, A. Ardzinski, F. Sorgel, H. S. Heine, and G. L. Drusano. 2007. Impact of spores on the rate of kill of Bacillus anthracis for five different antibiotics in a hollow fiber system, abstr. A-1436, p.34. Abstr. 47th Intersci. Conf. Antimicrob. Agents Chemother., 2007.
    24. 24
  24. Louie, A., W. Liu, M. Deziel, M. Drusano, and G. Drusano. 2004. Pharmacodynamics of linezolid in a neutropenic mouse thigh model of Staphylococcus aureus infection, abstr. A-1865. Abstr. 44th Intersci. Conf. Antimicrob. Agents Chemother., 2004.
    25. 25
  25. Meagher, A. K., A. Forrest, C. R. Rayner, M. C. Birmingham, and J. J. Schentag. 2003. Population pharmacokinetics of linezolid in patients treated in a compassionate-use program. Antimicrob. Agents Chemother. 47:548-553.[Abstract/Free Full Text]
    26. 26
  26. Meka, V. G., S. K. Pillai, G. Sakoulas, C. Wennersten, L. Venkataraman, P. C. DeGirolami, G. M. Eliopoulos, R. C. Moellering, Jr., and H. S. Gold. 2004. Linezolid resistance in sequential Staphylococcus aureus isolates associated with a T2500A mutation in the 23S rRNA gene and loss of a single copy of rRNA. J. Infect. Dis. 190:311-317.[CrossRef][Medline]
    27. 27
  27. Meselson, M., J. Guillemin, M. Hugh-Jones, A. Langmuir, I. Popova, A. Shelokov, and O. Yampolskaya. 1994. The Sverdlovsk anthrax outbreak of 1979. Science 266:1202-1208.[Abstract/Free Full Text]
    28. 28
  28. Meyerhoff, A., and D. Murphy. 2002. Guidelines for treatment of anthrax. JAMA 288:1848.[Free Full Text]
    29. 29
  29. Mikesell, P., B. E. Ivins, J. D. Ristroph, and T. M. Dreier. 1983. Evidence for plasmid-mediated toxin production in Bacillus anthracis. Infect. Immun. 39:371-376.[Abstract/Free Full Text]
    30. 30
  30. National Committee of Clinical and Laboratory Standards. 1999. Document M26-A. Methods for determining bactericidal activity of antimicrobial agents; approved guidelines. National Committee of Clinical and Laboratory Standards, Wayne, PA.
    31. 31
  31. Physicians' Desk Reference. 2007. Physician's desk reference, 61st ed., p.2652-2660. Thomson PDR, Montvale, NJ.
    32. 32
  32. Pillai, S. K., G. Sakoulas, C. Wennersten, G. M. Eliopoulos, R. C. Moellering, Jr., M. J. Ferraro, and H. S. Gold. 2002. Linezolid resistance in Staphylococcus aureus: characterization and stability of resistant phenotype. J. Infect. Dis. 186:1603-1607.[CrossRef][Medline]
    33. 33
  33. Rubinstein, E., R. Isturiz, C. Standiford, G. Smith, T. H. Oliphant, S. Cammarata, B. Hafkin, V. Le, and J. Remington. 2003. Worldwide assessment of linezolid's clinical safety and tolerability: comparator-controlled phase III studies. Antimicrob. Agents Chemother. 40:839-845.
    34. 34
  34. Stevens, D. L., A. E. Bryant, and S. P. Hackett. 1995. Antibiotic effects on bacterial viability, toxin production, and host response. Clin. Infect. Dis. 20(Suppl. 2):S154-S157.[Medline]
    35. 35
  35. Stevens, D. L., A. E. Gibbons, R. Bergstrom, and V. Winn. 1988. The Eagle effect revisited: efficacy of clindamycin, erythromycin and penicillin in the treatment of streptococcal myositis. J. Infect. Dis. 158:23-28.[Medline]
    36. 36
  36. Stevens, D. L., Y. Ma, D. B. Salmi, E. McIndoo, R. J. Wallace, and A. E. Bryant. 2007. Impact of antibiotics on expression of virulence-associated exotoxin genes in methicillin-sensitive and methicillin-resistant Staphylococcus aureus. J. Infect. Dis. 195:202-211.[CrossRef][Medline]
    37. 37
  37. Swaney, S. M., H. Aoki, M. C. Ganoza, and D. L. Shinabarger. 1998. The oxazolidinone linezolid inhibits initiation of protein synthesis in bacteria. Antimicrob. Agents Chemother. 42:3251-3255.[Abstract/Free Full Text]
    38. 38
  38. Tam, V. H., A. Louie, M. R. Deziel, W. Liu, R. Leary, and G. L. Drusano. 2005. Bacterial-population responses to drug-selective pressure: examination of garenoxacin's effect on Pseudomonas aeruginosa. J. Infect. Dis. 192:420-428.[CrossRef][Medline]
    39. 39
  39. Zimbelman, J., A. Palmer, and J. Todd. 1999. Improved outcome of clindamycin compared with beta-lactam antibiotic treatment for invasive Streptococcus pyogenes infection. Pediatr. Infect. Dis. J. 18:1096-1100.[CrossRef][Medline]

Antimicrobial Agents and Chemotherapy, July 2008, p. 2486-2496, Vol. 52, No. 7
0066-4804/08/$08.00+0     doi:10.1128/AAC.01439-07
Copyright © 2008, American Society for Microbiology. All Rights Reserved.



This article has been cited by other articles:

  • Drusano, G. L., Okusanya, O. O., Okusanya, A. O., van Scoy, B., Brown, D. L., Fregeau, C., Kulawy, R., Kinzig, M., Sorgel, F., Heine, H. S., Louie, A. (2009). Impact of Spore Biology on the Rate of Kill and Suppression of Resistance in Bacillus anthracis. Antimicrob. Agents Chemother. 53: 4718-4725 [Abstract] [Full Text]  
  • Yew, W. W., Leung, C. C. (2009). Update in Tuberculosis 2008. Am. J. Respir. Crit. Care Med. 179: 337-343 [Full Text]  

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Louie, A.
Right arrow Articles by Drusano, G. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Louie, A.
Right arrow Articles by Drusano, G. L.

Copyright © 2009 by the American Society for Microbiology.   For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»