JAC Advance Access originally published online on August 30, 2006
Journal of Antimicrobial Chemotherapy 2006 58(5):987-993; doi:10.1093/jac/dkl349
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Pharmacokinetic-pharmacodynamic rationale for cefepime dosing regimens in intensive care units
1 School of Pharmacy, University of Queensland Brisbane, QLD 4072, Australia 2 IBMP—Institute for Biomedical and Pharmaceutical Research Paul-Ehrlich-Strasse, 19 D-90562 Heroldsberg, Nürnberg, Germany 3 Anaesthesiology and Critical Care, University of Queensland and Department of Intensive Care Medicine, Royal Brisbane Hospital Brisbane, QLD 4029, Australia
*Corresponding author. Tel: +61-7-33469718; Fax: +61-7-33651688; E-mail: jroos{at}pharmacy.uq.edu.au
Received 25 May 2006; returned 5 July 2006; revised 25 July 2006; accepted 2 August 2006
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionTransparency declarationsReferences |
|---|
Objectives: (i) To develop a population pharmacokinetics (PK) model for cefepime in patients in intensive care units (ICUs). (ii) To assess the pharmacokinetic-pharmacodynamic profile of various cefepime dosing regimens and to assess their expected probability of target attainment (= PTA expectation value) against common ICU pathogens such as Escherichia coli, Klebsiella pneumoniae, Pseudomonas aeruginosa and Acinetobacter baumannii.
Methods: Thirteen ICU patients received cefepime 2 g 12 hourly intravenous (3 min). Twelve blood samples were taken on two occasions: (i) immediately after initial dose; and (ii) between days 3 and 6 after starting therapy. Population PK models were developed using NONMEM. Based on the final covariate model, Monte Carlo simulations were undertaken (n = 1000) to simulate free-drug concentrations of cefepime for two administration methods: (i) intermittent bolus administration (IBA); and (ii) continuous infusion (CI). Concentration–time profiles were evaluated by the probability of achieving free-drug concentration above the MIC for >65% of the dosing interval. Finally, using local MIC distributions of E. coli, K. pneumoniae, P. aeruginosa and A. baumannii the PTA expectation values for each dosing administration method were evaluated.
Results: A three-compartment model with zero-order input best described the concentration–time data. The PTA expectation values for E. coli and K. pneumoniae were >90% in all CI doses but only when administered as 1 g every 6 h and higher daily doses for IBA. For the current treatment protocol, 2 g every 12 h, P. aeruginosa and A. baumannii achieved target concentrations of only 54% and 28%, respectively. For P. aeruginosa, a CI of at least 4 g/day was required to achieve a PTA expectation value >90% while for A. baumannii a 6 g/day CI only achieved a PTA expectation value of 75%.
Conclusions: When given as IBA or CI for E. coli and K. pneumoniae, cefepime should be successful in achieving the bactericidal target. For P. aeruginosa higher doses of cefepime (>4 g/day) are required to achieve the required PTA expectation value. Cefepime fails to achieve the bactericidal target even when administered at high doses, e.g. 6 g/day, for A. baumannii.
Keywords: ß-lactams , critically ill patients , probability of target attainment
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionTransparency declarationsReferences |
|---|
Cefepime is a ‘fourth-generation’ cephalosporin with good activity against Gram-negative microorganisms, i.e. Escherichia coli and Klebsiella pneumoniae, and some activity against Gram-positive microorganisms, i.e. Streptococcus spp.1 Cefepime belongs to the ß-lactam class of antibiotics. The goal of ß-lactam therapy is to achieve a free-drug concentration (indicated by the prefix f)2 above the MIC. This is usually expressed as a fraction of the dosing interval (fT > MIC). It has been reported that the fT > MIC achieved directly impacts on the microbiological killing of ß-lactams.3,4 In vivo animal models of infection5 have demonstrated that for ß-lactams an fT > MIC of about 60–70% is required to achieve near-maximal bacterial killing.
The recommended dosage of cefepime for adults with normal renal function and mild to moderate infections is 1 g every 12 h. This cefepime dosing regimen has been shown to be effective against the majority of Enterobacteriaceae, streptococci and Staphylococcus aureus.6 For critically ill patients such as those treated in intensive care units (ICUs) cefepime’s broad spectrum of activity offers an advantage for empirical antibiotic therapy. In this critically ill patient group, the recommended dose is increased to 2 g every 12 h.7 However, for pathogens such as Pseudomonas aeruginosa and Acinetobacter species, it has been suggested that higher doses of cefepime or different modes of administration may be required to achieve maximal bacterial killing.8
The current dataset has been previously published and analysed using a standard two-stage approach.9,10 This pharmacokinetic (PK) analysis method has some limitations11 and it is now preferred that population PK analysis via non-linear mixed effects model is utilized to provide more accurate estimates of the between-subject variability and therefore should provide more accurate estimation of the probability of target attainment (PTA).
The aims of this study were: (i) to develop a population PK model for cefepime in ICU patients; and (ii) to assess the pharmacokinetic-pharmacodynamic profile of various cefepime dosing regimens and to assess their expected PTA (= PTA expectation value) against common ICU pathogens such as E. coli, K. pneumoniae, P. aeruginosa and Acinetobacter baumannii.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionTransparency declarationsReferences |
|---|
Subjects
Full details on the current dataset have been presented elsewhere.9,10 The study protocol was approved by the Ethics committee of the Royal Brisbane Hospital, Brisbane, Australia. Informed consent was obtained from patients or next of kin.
In summary, 13 ICU patients (11 males) received cefepime 2 g every 12 h as a 3 min intravenous (iv) infusion. Patients were enrolled if they had a serum creatinine concentration of <0.1 mmol/L.
Sampling schedule and determination of cefepime in plasma by HPLC
All blood samples (10 mL) were taken from an in situ arterial line immediately prior to dose administration (time [T] = 0 at the start of the 3 min infusion) and at 5, 10, 20, 30, 60, 90, 120, 240, 360, 480, 600 and 720 min post-start of infusion. The subjects received the same dose of 2 g cefepime twice daily as a 3 min infusion for at least 3–6 days. Occasion 2 started between 60 and 120 h after the first dose. The HPLC assay for measurement of cefepime in plasma was linear from 1 to 200 µg/mL and the intra-day and inter-day imprecision values were under 6%.9,10 For the PK analysis, the values below the limit of quantification (BLQ) of the assay (1 µg/mL) have been substituted by half the quantification limit, as described by Beal.12
Population pharmacokinetics modelling
The concentration versus time data for cefepime in plasma were analysed by a non-linear mixed effects modelling approach13 using NONMEM (Version 5, Level 1.1, GloboMax LLC, Hanover, MD, USA) with double precision with the G77 FORTRAN compiler. The NONMEM runs were executed using Wings for NONMEM (WFN 408b). Data were analysed using the first order conditional estimation (FOCE) method with INTERACTION.
For the population PK analysis, the plasma cefepime concentrations were fitted to one, two or three-compartment models using subroutines from the NONMEM library.13 The concentration–time profile can be described as (Equation 1):
 |
i represents fixed effects parameter of the structural model to be estimated. fij is the function for the prediction of the jth response for the ith subject. Finally, 
ij denotes the jth measurement error for the ith subject. In other words, ij is the difference of the observed concentration from the predicted concentration. It is assumed to be independent and identically distributed with a normal distribution around the mean zero and variance 
2. Between-subject variability (BSV) and between-occasion variability (BOV)
BSV was modelled using an exponential variability model (Equation 2):
 |
i is the value of the parameter for the ith subject, is the typical value of the parameter in the population and finally 
i is a random vector with normal distribution, zero mean and variance–covariance matrix of BSV 
to be estimated.
BOV is the variability of a parameter within a subject during treatment and includes between-occasion variability and within-occasion variability. BOV was assumed to be log normally distributed and modelled over the two PK study occasions (Equation 3):
 |
i,k is the value of the parameter for the ith subject on the kth occasion. Model diagnostics
Statistical comparison of nested models was based on a 
2 test of the difference in the objective function. A decrease in the objective function of 3.84 units (P < 0.05) was considered significant.
Goodness-of-fit was evaluated by visual inspection of diagnostic scatter plots, including observed and predicted concentrations versus time, weighted residual versus time and residual versus predicted concentrations.
Bootstrap
A non-parametric bootstrap method14 (n = 1000) was used to study the uncertainty of all PK parameter estimates. From the bootstrap empirical posterior distribution we have been able to obtain the 95% confidence interval (2.5–97.5% percentile) for the parameters, as described previously.15
Covariate screening
The covariates analysed were age, weight, serum creatinine, creatinine clearance measured by 8 h urine collection, creatinine clearance estimated via C&G equation using total body weight and APACHE II scores. The individual covariates were centred by the median or standard values of occasion one and occasion two. Individual empirical Bayesian (POSTHOC) parameters were plotted against covariate values to assess relationships. If a trend between covariates and PK parameter was observed, then it was considered for inclusion in the population model.
Possible covariates were added in a stepwise fashion into the model. Covariates were kept in the model if there was improvement in the fit over the base model, i.e. decrease in objective function and decrease in the BSV of the parameter.
Visual predictive checks
Using the final covariate model a visual predictive check was performed by simulating 10 000 subjects to assess the predictive performance of the model. The visual predictive checks were generated using a Perl Script (version 1e).16 The visual checks and representative percentiles [10th, 50th (median) and 90th percentile] were visually assessed using Prism® 2005 (Version 4.03).
Dosing simulations
Five intermittent bolus administration (IBA) and three continuous infusion (CI) dosing regimens were simulated using Monte Carlo simulations. The five bolus dose regimens evaluated were 2 g every 12 h (same treatment regimen as in this study protocol), 2 g every 8 h, 1 g every 12 h, 1 g every 6 h and 1 g every 4 h, while the three CI regimens evaluated were 2, 4 or 6 g over 24 h with a loading dose of 0.5 g. Each Monte Carlo simulation generated free-concentration time profiles for 1000 subjects per dosing regimen using the parameters from the final covariate model. A value of 10% protein binding was used in all simulations.17,18 From this data the fT > MIC was calculated for each simulated subject using linear interpolation. The PTA was obtained by counting the subjects who achieved free-cefepime concentrations for at least 65% of the dosing interval.3,19
MIC distributions
MIC distributions were derived from cefepime MIC50, MIC90 and range collected from Australian laboratories provided by the Queensland Health Pathology Service (QHPS) for E. coli, K. pneumoniae, P. aeruginosa and A. baumannii isolates. The MIC distributions were estimated from 2794 strains of E. coli, 896 strains of K. pneumoniae, 1853 strains of P. aeruginosa and 234 strains of A. baumannii.
The PTA expectation values were calculated by multiplying the PTA at each MIC by the fraction of organisms susceptible at that concentration of the respective MIC distribution. The sum of those individual products is the PTA expectation value for the respective MIC distribution. The PTA expectation value can be interpreted as the probability of successful treatment of infections caused by bacteria with a specific susceptibility pattern (MIC distribution) in the studied patient population. These calculations were performed for the first 12 h of treatment (occasion 1). Calculations were performed on the first 12 h as it represents the worst case scenario of PTA expectation as the free-concentrations approach steady-state with the different dosing regimens and that likely onset of sepsis treatment can be ascertained.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionTransparency declarationsReferences |
|---|
Subjects
The patients' age ranged from 34 to 75 years (median, 60 years); estimated total body weight ranged from 56 to 128 kg (median, 75 kg); APACHE II scores ranged from 4 to 24 (median, 11); and the 8 h urine collection resulted in a creatinine clearance that ranged from 2.3 to 11.7 L/h (median, 7.1 L/h). The dataset comprises a total of 307 quantifiable samples. Five measurements were below the limit of quantification.
Model building
The best base model based on the model building criteria consisted of a three-compartment model with a full variance-covariance matrix between clearance (CL) and central volume of distribution (V1), a diagonal BSV for peripheral volume of distribution (V2) and a combined residual unknown variability (RUV). The model supported between-occasion variability on CL and V1. The values of inter-compartmental clearance between the third and first compartment (Q3) and the peripheral volume of distribution (V3) were fixed. The final objective function for this model was 1227.391.
Figure 1 shows the plot of the observed cefepime concentrations versus time overlaid by the predicted typical cefepime concentrations versus time. The values of the parameters for the final base model are given in Table 1. Table 1 presents the 95% confidence interval for the parameters computed from all bootstrap runs.
|
|
Creatinine clearance measured by 8 h urine collection was the only covariate to describe cefepime clearance. The final model was represented by Equation 4:
 |
|
Figure 2 (a and b) shows a plot of visual predictive check with the final covariate model for occasion 1 and occasion 2. These plots show that the final PK model describes the measured cefepime concentrations adequately on both occasions. All subsequent cefepime Monte Carlo simulations were then based on this model.
|
Dosing simulations
Intermittent bolus administration.. Figure 3(a) shows the PTA versus MIC profiles for the different intermittent short-term infusion regimens. The recommended dosing regimen for patients with mild to moderate infections, 1 g every 12 h, appears to provide a high PTA up to an MIC of 0.25 mg/L (inclusive). However, the recommended ICU treatment protocol, 2 g every 12 h, provides a high PTA up to and including an MIC of 0.5 mg/L. In addition, the dosing regimens of 1 g every 4 h or 2 g every 8 h provide very similar and robust (>90%) PTA up to and including an MIC of 2 mg/L.
|
Continuous infusion.. The PTA versus MIC profiles for the different CI dosing regimens with a loading dose is shown in Figure 3(b). The low-dose CI of 2 g cefepime per day provides a robust (>90%) PTA up to an MIC of 2 mg/L. However, a high-dose of cefepime (6 g CI) showed a robust (>90%) PTA up to an MIC of 8 mg/L (inclusive).
PTA expectation values
The assessment of PTA expectation value versus the MIC distributions for the first occasion is shown in Table 3. When cefepime is administered as 1 g every 4 h, 1 g every 6 h or 2 g every 8 h, the population PTA expectation value is >90% for E. coli and K. pneumoniae. For the current treatment protocol of 2 g every 12 h, cefepime achieves a 54% and 28% PTA expectation value for P. aeruginosa and A. baumannii, respectively. This is further reduced for 1 g every 12 h, which would provide PTA expectation values of 35.5% and 11.6%, respectively.
|
The PTA expectation values for E. coli and K. pneumoniae when cefepime is given as a CI was >90% for all dose groups, i.e. 2, 4 or 6 g/day. To achieve a >90% PTA expectation value for P. aeruginosa a dose of at least 4 g/day of cefepime as a CI is required. However, at the maximum recommended cefepime dose of 6 g/day administered as a CI, the PTA expectation value for A. baumannii is at best only 75%.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionTransparency declarationsReferences |
|---|
The current study presents a population PK model for cefepime in ICU patients who had serum creatinine concentrations below the upper limit of normal. It includes stochastic simulations (often called Monte Carlo simulations) under various dosing regimens to assess the PTA for common ICU pathogens.
It is necessary to appreciate that the PTA for maximal bacterial cell killing of ß-lactams following the administration of a fixed dose will depend on the between-subject variability of PK parameters, in particular clearance and volume of distribution.20 This must be then incorporated with different MIC distributions for specific pathogens in various parts of the world. The current study used susceptibility patterns obtained from the QHPS. Therefore, our PTA expectation values apply for Australian resistance patterns, whereas our PTA versus MIC profiles apply for ICU patients worldwide. However, it is important to note that by using the PTA profiles given in Figure 3 (a and b), the PTA expectation values can be obtained for any given MIC distribution.
It should be noted that the inclusion criterion of the initial study9 was a ‘normal’ serum creatinine and some of these patients had very high creatinine clearances.9,10 As creatinine clearance was a predictor of cefepime clearance, the patients with the high creatinine clearance will result in low trough cefepime concentration. It is in these patients that our data show the need for higher than normal doses of cefepime either using IBA or CI to cover all PTA expectation values.
An important finding from the model building was that BSV was greater than BOV. This supports the concept that cefepime could be dose-individualized as there are only small changes in PK parameters from day to day. This could be achieved empirically by using creatinine clearance to predict the likely dose or via blood sampling and a target concentration intervention approach.15
More sensitive assays for cefepime have been published in the past years.21 However, the area under the curve from time zero to the last quantifiable concentration was at least 91% of the area under the curve from time zero to infinity in our study. Therefore, our assay was sensitive enough for our objectives. The number of data points below the quantification limit was small (<2%). Therefore, we could not show that our BLQ handling method provided less bias in the model parameters as suggested previously.12,22–24
There is convincing data for penicillins from animal experiments that only the non-protein bound concentration is microbiologically active.3,5,25 Also, the duration of unbound concentrations above the MIC is the key determinant to achieve optimum therapy for this class of antibiotics. The present study has shown that administration via CI with a loading dose offers significant advantage when compared with IBA. This is based on achieving a higher PTA for the same daily dose per 12 h. Furthermore, the results of the present study suggest that CIs offer an advantage when treating P. aeruginosa.
However, against A. baumannii, at a dose of 6 g per 24 h, the CI achieved a PTA value of 75%, whereas intermittent regimens at best could only achieve a PTA expectation of 61% (see Table 3). What this means is that for this pathogen cefepime doses greater than 6 g per day via CIs may need to be considered to optimize therapy, as well as providing antimicrobial cover with other synergic agents, e.g. fluoroquinolones.26
While the use of CI therapy in ICU patients could possibly have disadvantages, for example, the extra iv line may be associated with a higher probability of a line infection, increasing costs and morbidity27 and some drugs may be unstable at room temperature or incompatible with other simultaneously administered drugs, requiring the placement of a separate line, in our experience these potential problems are of nuisance value only. However, reconstituted solution of cefepime is stable for up to 24 h at room temperature or in a refrigerator (<5°C) for up to 7 days.7 Thus, cefepime is ideally suited to CI administration.
Simulation of longer infusion times, e.g. 30 min, 3 or 5 h infusion, instead of CI, have been suggested to minimize the problems discussed above. For this mode of administration the optimal length of infusion is about the fT > MIC target multiplied by the dosing interval. Thus the duration of infusion that achieves the highest PTAs is 
7.8 h (12 h x 65%) for every 12 h dosing and 5.2 h (8 h x 65%) for every 8 h dosing. 27
Nevertheless, CI clearly remains the optimal mode of administration if higher targets of an fT > MIC above 70% of the dosing interval are required. However, further clinical studies on the exact target in critically ill sepsis patients are required. Furthermore, more clinical data about the effectiveness of continuous versus prolonged infusion should be collected in future clinical studies.
Conclusions
Creatinine clearance measured by 8 h urine collection to assess ICU patients’ renal function appears to be a useful predictor for cefepime clearance and potentially could be used to individualize cefepime therapy. Cefepime when administered as 1 g every 6 h has a >90% PTA expectation value for killing E. coli and K. pneumoniae. However for P. aeruginosa, a daily dose of 4 g/day of cefepime administered as CI is required to achieve a PTA expectation value of >90%, while for A. baumannii even a CI of 6 g/day cefepime only achieved a PTA expectation value of 75%.
|
Transparency declarations |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionTransparency declarationsReferences |
|---|
None to declare.
|
Acknowledgements |
|---|
We would like to thank the Queensland Health Pathology Service (QHPS) for providing the MIC50, MIC90 and the range for the pathogens of interest.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionTransparency declarationsReferences |
|---|
1 Barradell LB and Bryson HM. (1994) Cefepime. A review of its antibacterial activity, pharmacokinetic properties and therapeutic use. Drugs 47:471–505.[Web of Science][Medline]
2
Mouton JW, Dudley MN, Cars O, et al. (2005) Standardization of pharmacokinetic/pharmacodynamic (PK/PD) terminology for anti-infective drugs: an update. J Antimicrob Chemother 55:601–7.
3 Craig WA. (1995) Interrelationship between pharmacokinetics and pharmacodynamics in determining dosage regimens for broad-spectrum cephalosporins. Diagn Microbiol Infect Dis 22:89–96.[CrossRef][Web of Science][Medline]
4 Jacobs MR. (2001) Optimisation of antimicrobial therapy using pharmacokinetic and pharmacodynamic parameters. Clin Microbiol Infect 7:589–96.[CrossRef][Web of Science][Medline]
5 Craig WA. (1998) Pharmacokinetic/pharmacodynamic parameters: rationale for antibacterial dosing of mice and men. Clin Infect Dis 26:1–10.[Web of Science][Medline]
6 Kovarik J, Rozenberg-Arska M, Visser M, et al. (1990) Pharmacodynamics of cefepime. Scand J Infect Dis Suppl 74:270–3.[Medline]
7 Product Information: Maxipime®, cefepime hydrochloride for injection. (Revised December 2003). Princeton, NJ: Bristol-Myers Squibb Company.
8 Sader HS and Jones RN. (2005) Comprehensive in vitro evaluation of cefepime combined with aztreonam or ampicillin/sulbactam against multi-drug resistant Pseudomonas aeruginosa and Acinetobacter spp. Int J Antimicrob Agents 25:380–4.[CrossRef][Web of Science][Medline]
9
Lipman J, Wallis SC, Rickard C. (1999) Low plasma cefepime levels in critically ill septic patients: pharmacokinetic modeling indicates improved troughs with revised dosing. Antimicrob Agents Chemother 43:2559–61.
10
Lipman J, Wallis SC, Boots RJ. (2003) Cefepime versus cefpirome: the importance of creatinine clearance. Anesth Analg 97:1149–54.
11 Steimer JL, Mallet A, Golmard JL, et al. (1984) Alternative approaches to estimation of population pharmacokinetic parameters: comparison with the nonlinear mixed-effect model. Drug Metab Rev 15:265–92.[Web of Science][Medline]
12 Beal SL. (2001) Ways to fit a PK model with some data below the quantification limit. J Pharmacokinet Pharmacodyn 28:481–504.[CrossRef][Web of Science][Medline]
13 Beal SL and Sheiner LB. (1998) NONMEM User Guides (I-VIII)University of California at San Francisco, San Francisco: University of California NONMEM Project Group.
14 Parke J, Holford NH, Charles BG. (1999) A procedure for generating bootstrap samples for the validation of nonlinear mixed-effects population models. Comput Methods Programs Biomed 59:19–29.[CrossRef][Web of Science][Medline]
15 Matthews I, Kirkpatrick C, Holford N. (2004) Quantitative justification for target concentration intervention—parameter variability and predictive performance using population pharmacokinetic models for aminoglycosides. Br J Clin Pharmacol 58:8–19.[CrossRef][Web of Science][Medline]
16 Bulitta J and Holford NH. (2005) Assessment of predictive performance of pharmacokinetic models based on plasma and urine data. PAGANZ Population Approach Group in Australia & New Zealand, Brisbane, Australia.
17
Oster S, Edelstein H, Cassano K, et al. (1990) Open trial of cefepime (BMY 28142) for infections in hospitalized patients. Antimicrob Agents Chemother 34:954–7.
18 Ismail MM. (2005) Disposition kinetics, bioavailability and renal clearance of cefepime in calves. Vet Res Commun 29:69–79.[CrossRef][Web of Science][Medline]
19 Drusano GL. (2004) Antimicrobial pharmacodynamics: critical interactions of ‘bug and drug’. Nat Rev Microbiol 2:289–300.[CrossRef][Web of Science][Medline]
20 Bradley JS, Dudley MN, Drusano GL. (2003) Predicting efficacy of antiinfectives with pharmacodynamics and Monte Carlo simulation. Pediatr Infect Dis J 22:982–92.[Web of Science][Medline]
21 Cherti N, Kinowski JM, Lefrant JY, et al. (2001) High-performance liquid chromatographic determination of cefepime in human plasma and in urine and dialysis fluid using a column-switching technique. J Chromatogr B Biomed Sci Appl 754:377–86.[CrossRef][Medline]
22 Duval V and Karlsson MO. (2002) Impact of omission or replacement of data below the limit of quantification on parameter estimates in a two-compartment model. Pharm Res 19:1835–40.[CrossRef][Web of Science][Medline]
23 Hennig S, Waterhouse TH, Wainwright CE, et al. (2006) A D-optimal designed population pharmacokinetic study of itraconazole capsules and solution in adults with cystic fibrosis. PAGE Population Approach Group in Europe, Belgium.
24 NMusers. Lower Limit of Quantification http://huxley.phor.com/nonmem/nm/99jul142005.html, 2005. (July 2006, date last accessed).
25
Merrikin DJ, Briant J, Rolinson GN. (1983) Effect of protein binding on antibiotic activity in vivo. J Antimicrob Chemother 11:233–8.
26
Drago L, De Vecchi E, Nicola L, et al. (2005) In vitro selection of resistance in Pseudomonas aeruginosa and Acinetobacter spp. by levofloxacin and ciprofloxacin alone and in combination with ß-lactams and amikacin. J Antimicrob Chemother 56:353–9.
27
Lomaestro BM and Drusano GL. (2005) Pharmacodynamic evaluation of extending the administration time of meropenem using a Monte Carlo simulation. Antimicrob Agents Chemother 49:461–3.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. M. Nicasio, R. E. Ariano, S. A. Zelenitsky, A. Kim, J. L. Crandon, J. L. Kuti, and D. P. Nicolau Population Pharmacokinetics of High-Dose, Prolonged-Infusion Cefepime in Adult Critically Ill Patients with Ventilator-Associated Pneumonia Antimicrob. Agents Chemother., April 1, 2009; 53(4): 1476 - 1481. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. S. Chin, W. Y. Seo, D. L. Paterson, B. A. Potoski, and A. Y. Peleg Cefepime MIC Breakpoint Resettlement in Gram-Negative Bacteria Antimicrob. Agents Chemother., January 1, 2009; 53(1): 337 - 338. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||