Abstract
A simple and rapid ultra-high-performance liquid chromatography (UHPLC) method using UV detection was developed for the simultaneous determination of eight β-lactam antibiotics in human plasma, including four penicillins, amoxicillin (AMX), cloxacillin (CLX), oxacillin (OXA), and piperacillin (PIP), and four cephalosporins, cefazolin (CFZ), cefepime (FEP), cefotaxime (CTX), and ceftazidime (CAZ). One hundred-microliter samples were spiked with thiopental as an internal standard, and proteins were precipitated by acetonitrile containing 0.1% formic acid. Separation was achieved on a pentafluorophenyl (PFP) column with a mobile phase composed of phosphoric acid (10 mM) and acetonitrile in gradient elution mode at a flow rate of 500 μl/min. Detection was performed at 230 nm for AMX, CLX, OXA, and PIP and 260 nm for CFZ, FEP, CTX, and CAZ. The total analysis time did not exceed 13 min. The method was found to be linear at concentrations ranging from 2 to 100 mg/liter for each compound, and all validation parameters fulfilled international requirements. Between- and within-run accuracy errors ranged from −5.2% to 11.4%, and precision was lower than 14.2%. This simple method requires small-volume samples and can easily be implemented in most clinical laboratories to promote the therapeutic drug monitoring of β-lactam antibiotics. The simultaneous determination of several antibiotics considerably reduces the time to results for clinicians, which may improve treatment efficiency, especially in critically ill patients.
INTRODUCTION
Penicillins and cephalosporins form two classes of β-lactam antibiotics. They are extensively prescribed for the treatment of potentially life-threatening infections, including peritonitis, respiratory tract infections, endocarditis, meningitis, and skin and soft tissue infections, which involve a wide range of bacteria (1).
The pharmacokinetic-pharmacodynamic (PK-PD) study of these antibiotics has consistently shown that their antibacterial activity is time dependent; i.e., the time that the plasma concentration remains above the MIC between two administered doses is well correlated with treatment efficiency. According to many authors, the optimal bactericidal activity in critically ill patients is achieved when plasma concentrations are above the MIC or even greater than 4 to 5 times the MIC for 70% to 100% of the dosing interval (2–10). The failure of antibiotic treatment and the development of resistant strains of bacteria may occur in patients with insufficient antibiotic exposure. Conversely, concentrations of β-lactam antibiotics that are too high may induce uncommon but severe adverse reactions, mainly neurotoxicity symptoms, including seizures and encephalopathy (11, 12).
Therapeutic drug monitoring (TDM) of β-lactam antibiotics is therefore clinically relevant to ensure optimal antibiotic exposure, thus improving treatment efficiency while minimizing toxicity and reducing the emergence of resistance. TDM of β-lactam antibiotics is specifically useful in critically ill patients in whom the pharmacokinetics of drugs are altered and variable. Such patients frequently have associated hypoalbuminemia and/or an inflammatory syndrome, resulting in a large increase in the volume of distribution. Some patients may develop organ dysfunction, resulting in antibiotic drug accumulation, whereas other patients with septic shock may exhibit hyperdynamic circulation, leading to increased renal elimination (13, 14).
Several analytical methods for the determination of β-lactam antibiotics in the plasma/serum of patients have been reported (for reviews, see the work of El-Shaboury et al. [15] and Carlier et al. [16]). Many of these methods were developed using conventional high-performance liquid chromatography (HPLC) with UV detection and suffer from a long analysis time (17–21). Other methods, developed by using ultra-high-performance liquid chromatography (UHPLC) coupled with mass spectrometry detection, allow shorter analysis times but require materials that are not readily available in most laboratories (22–27).
In clinical practice, there is a critical need to develop an easy-to-implement method that would cover a wide range of β-lactam antibiotics, thus reducing the time to results for clinicians. Such a method would undoubtedly promote and improve the efficiency of antibiotic TDM. To that end, the simultaneous determination of major β-lactam antibiotics in a single run using UHPLC with UV detection (UHPLC-UV) may be a compromise and has never been reported so far. Specific analytical developments are therefore required to propose a method with simple and rapid sample preparation, adequate separation of the tested compounds, and a sufficient detection sensitivity that would meet the clinician needs for TDM.
The aim of this study was to develop and validate a single UHPLC-UV method for the simultaneous determination in plasma of eight widely prescribed β-lactam antibiotics, including four penicillins, amoxicillin (AMX), piperacillin (PIP), oxacillin (OXA), and cloxacillin (CLX), and four cephalosporins, cefepime (FEP), ceftazidime (CAZ), cefotaxime (CTX), and cefazolin (CFZ).
MATERIALS AND METHODS
Chemicals and reagents.
Amoxicillin sodium, ceftazidime pentahydrate, cefepime dihydrochloride monohydrate, cefotaxime sodium, cefazolin sodium, and piperacillin sodium salt were supplied by Panpharma; oxacillin sodium salt was supplied by Bristol-Myers Squibb; cloxacillin sodium salt monohydrate was supplied by Astellas; and thiopental, used as an internal standard (IS), was supplied by Sigma-Aldrich (Saint-Quentin Fallavier, France). The chemical structures of these compounds are shown in Fig. 1. Phosphoric acid (85%) and chlorhydric acid (1 N) were purchased from VWR (Fontenay-sous-Bois, France). Acetonitrile and methanol (HPLC grade) were obtained from Sigma-Aldrich (Saint-Quentin Fallavier, France). Drug-free plasma was provided by the Etablissement Français du Sang (Rungis, France).
FIG 1.
Chemical structures of amoxicillin (AMX), piperacillin (PIP), oxacillin (OXA), cloxacillin (CLX), cefepime (FEP), cefotaxime (CTX), cefazolin (CFZ), ceftazidime (CAZ), and thiopental (THI; used as the internal standard).
Instrumentation and chromatographic conditions.
The liquid chromatographic system (Ultimate 3000; Thermo-Fisher Scientific, Les Ulis, France) consisted of an UHPLC quaternary pump, a thermostated oven, a thermostated autosampler, and an UV-diode array detector. Data analysis was performed using Chromeleon (version 7.2) software. Separation was achieved at 40°C using a Hypersil Gold pentafluorophenyl (PFP) analytical column (2.1 by 100 mm, 1.9 μm) purchased from Thermo-Fisher Scientific. The mobile phase consisted of 10 mM phosphoric acid (mobile phase A) and acetonitrile (mobile phase B) and was delivered at 500 μl/min in the gradient elution mode. From 0 to 1 min, its composition was 100% mobile phase A; afterward, the percentage of mobile phase B was linearly increased to 55% for 11 min. Next, the mobile phase composition was set back to the initial conditions for 2 min for reequilibration. The autosampler temperature was kept at +4°C, and the injected volume was 10 μl.
From 0 to 4.1 min, the detection wavelength was set at 230 nm. The wavelength was then switched to 260 nm from 4.1 to 9.0 min, and it was reset to 230 nm from 9.0 min to the end of the run. Under these conditions, AMX, PIP, the IS, OXA, and CLX were detected at 230 nm, whereas FEP, CAZ, CTX, and CFZ were detected at 260 nm.
Preparation of solutions.
Stock standard solutions (100 mg/ml in water) of each antibiotic were prepared and stored at −80°C. A standard mix solution containing all antibiotics at 5 mg/ml was then prepared in water and aliquoted, and the aliquots were placed into polypropylene tubes and stored at −80°C. On the day of analysis, spiking solutions (10, 25, 50, 125, 250, 375, 500 mg/liter) were prepared by appropriate dilution of the standard mix solution in water. Calibration standards were prepared by adding 20 μl of each spiking solution to 80 μl of drug-free plasma in order to obtain the following final concentrations: 2, 5, 10, 25, 50, 75, and 100 mg/liter.
Other stock solutions (100 mg/ml in water) of antibiotics were also prepared to spike quality control (QC) samples. Two QC mix solutions containing all antibiotics at 5 mg/ml and 0.5 mg/ml were prepared in water, aliquoted in polypropylene tubes, and stored at −80°C. QC samples with concentrations of 2, 6, 30, and 70 mg/liter (corresponding to a sample with the lower limit of quantitation [LLOQ] and low-, medium-, and high-concentration QC samples, respectively) were prepared by appropriate dilution of the two QC mix solutions in plasma and aliquoted, and the aliquots were placed in polypropylene tubes and stored at −80°C.
A stock solution of the IS at 1 mg/ml was prepared in methanol and stored at −20°C. A working solution was prepared at 50 mg/liter in methanol.
Collection of patient blood samples.
Blood samples were collected from patients as part of their usual follow-up. Samples (5 ml) were collected in heparinized tubes and were centrifuged (2,000 × g, +4°C, 15 min) immediately after receipt. Then, the plasma samples were stored at −80°C until analysis.
Sample preparation.
On the day of analysis, calibration standards were freshly prepared, whereas patient and QC samples were thawed. One hundred microliters of the plasma sample was spiked with 50 μl of the IS working solution at 50 mg/liter, except for the blank sample. A simple preparation requiring protein precipitation was performed by adding 400 μl of acetonitrile containing 0.1% (vol/vol) formic acid.
The mixture was vortexed, left at room temperature for approximately 5 min, and centrifuged at 10,000 × g for 10 min. The supernatant was evaporated to dryness at 40°C under a gentle nitrogen stream. The residue was dissolved in 400 μl of 10 mM phosphoric acid and transferred into a polypropylene vial.
Validation.
Validation of this method was performed according to the European Medicines Agency (EMA) guidelines on bioanalytical method validation (28).
(i) Selectivity and specificity.
Selectivity was assessed by analyzing six different batches of plasma. For each batch, two samples were analyzed: a blank sample and a sample spiked with all tested antibiotics at the LLOQ and with the IS at the working concentration. Specificity was assessed by analyzing different plasma samples containing a high concentration of potentially coadministered drugs to investigate potential interference. No interfering components were considered to be present when the signal was less than 20% of the LLOQ for analytes and 5% for the IS.
(ii) Carryover.
Carryover was assessed by injecting blank samples in triplicate after the highest calibration standard was run. The signal in the blank sample run after the highest calibration standard was run should not be greater than 20% of the LLOQ and 5% for the IS.
(iii) Calibration and linearity.
The linearity study was performed by analyzing calibration standards on 10 separate days at concentrations ranging from 2 and 100 mg/liter, in order to cover a wide range of plasma concentrations. For each antibiotic, calibration curves were established using the ratio of the peak area of the analyte/peak area of the IS (y) versus concentration (x). Slopes, intercepts, and regression coefficients (r2 values) were obtained by linear regression analysis. The best weighting factor was selected in order to balance the increase of variance with concentration and to minimize the error for low concentrations. Backcalculated standard concentrations could not differ from the nominal value by ±15% (±20% was used for the standard with the lowest concentration).
(iv) Accuracy and precision.
A study of within-run accuracy and precision was performed by assaying the QC samples with the four concentrations 12 times in a single analytical run, while a study of the between-run accuracy and precision was performed by assaying the QC samples with the four concentrations in duplicate on 12 separate days. The mean measured concentrations and their standard deviations (SDs) were calculated. Precision was expressed as the coefficient of variation (CV), which was equal to (SD/mean) × 100, and accuracy error was expressed as the bias, which was calculated as ([measured concentration − nominal concentration]/nominal concentration) × 100.
Acceptance criteria were as follows: bias had to be within ±15% of the nominal value (±20% for the LLOQ), and within- and between-run precision had to be less than 15% (20% for the LLOQ).
(v) LLOQ.
The LLOQ was defined as the lowest concentration of analyte which could be determined with acceptable accuracy and precision. The analyte signal of the sample with the LLOQ should be at least 5 times the signal of a blank sample.
(vi) Dilution integrity.
Dilution integrity was assessed by assaying samples with a concentration above the upper limit of quantitation (ULOQ): six replicates of plasma samples with each antibiotic at 150 mg/liter were prepared. Then, these samples were diluted 2-fold with the same matrix in order to obtain a final concentration within the calibration range (75 mg/liter). Accuracy and precision should be within the set criteria (±15%).
(vii) Extraction recoveries.
Extraction recoveries of each antibiotic and the IS were measured at three different levels (corresponding to the QC samples with low, medium, and high concentrations) for six determinations. Extraction recovery was determined by comparing the peak area of the analytes spiked before extraction to the peak area of a standard solution prepared with identical concentrations.
Stability. (i) Stability of stock and mix solutions.
The stability of the stock solutions and the mix solutions held for 4 months at −80°C was assessed. Stability was determined by comparing in triplicate the peak area between solutions stored at −80°C and freshly prepared solutions. The difference between the mean peak areas had to be less than 15%.
(ii) Stability of quality controls.
Stability studies were conducted in triplicate with the QC samples with the three concentrations (low, medium, high). Long-term stability was assessed with QC samples stored for 2 months at −80°C. The freeze-thaw stability of the QC samples was determined over three freeze-thaw cycles. Benchtop stability was conducted after the QC samples were stored for 6 h at room temperature. The stability of QC sample extracts in an autosampler was assessed after 72 h of storage of the sample at +4°C in the autosampler. The stability results for all samples should be within 15% of the nominal concentrations.
RESULTS AND DISCUSSION
Preparation of solutions.
Chemical standards of β-lactam antibiotics as their sodium or chloride salts were selected in order to facilitate their solubility at a high concentration in water. Independent stock solutions of each β-lactam antibiotic were prepared at 100 mg/ml in water. Then, the stock solutions were mixed. This mix solution was concentrated enough to prepare the calibration standard containing the eight antibiotics at the ULOQ.
Chromatographic conditions.
We developed a method for the simultaneous determination of eight β-lactam antibiotics using UHPLC with UV detection. The fundamental principle of UHPLC is to use a column packed with particles smaller than those used for regular HPLC (particle size for UHPLC, <2 μm), in order to provide a shorter run time and an enhanced sensitivity. To gain the full potential of UHPLC, we used a chromatographic system that withstands the high operating pressures generated by UHPLC.
We first performed separation of these compounds using a C18 analytical column in accordance with many other methods (19–27, 29), but the selectivity between AMX and FEP was not optimal. Therefore, we chose a pentafluorophenyl (PFP) column, which offers an alternative selectivity compared to that of C18 toward polar compounds and, thus, a better separation between AMX and FEP.
Most chromatographic methods with UV detection reported in the literature were performed using a mobile phase containing phosphate buffer (17–19, 21, 29). We initially used a mobile phase containing 10 mM phosphate buffer (pH 2) and acetonitrile in a gradient elution mode, but the backpressure dramatically increased, probably because of buffer precipitation, leading to column obstruction. Therefore, we replaced the phosphate buffer with 10 mM phosphoric acid, and the pressure remained suitable and constant.
The mobile phase was delivered in a gradient elution mode (from 0 to 55% acetonitrile from 1 to 11 min) to permit sufficient retention and the adequate separation of AMX and FEP. These conditions provided a short analysis time while maintaining an acceptable resolution between compounds. The total run time did not exceed 13 min, including the reequilibration step, which is shorter than that previously reported using UV detection, which ranged from 21 to 35 min (17–21, 29). Representative chromatograms of a blank sample, a QC sample spiked with the LLOQ, and the QC sample with the highest concentration are shown in Fig. 2.
FIG 2.
Chromatograms of a blank sample (A), a sample spiked with each antibiotic at 2 mg/liter, corresponding to the LLOQ (B), and a sample spiked with each antibiotic at 70 mg/liter, corresponding to the QC sample with a high concentration (C). mAU, milli-absorbance units.
Validation. (i) Selectivity and specificity.
Six different batches of blank plasma were analyzed. For each blank sample, the signal was below 20% of the LLOQ for each antibiotic and 5% for the IS, thus ensuring the selectivity of the method.
Compared to individual methods, simultaneous methods with UV detection are associated with an increased risk of coelution with potentially interfering compounds. No interference was found with plasma samples containing frequently coadministered drugs, including amikacin, gentamicin, tobramycin, vancomycin, teicoplanin, ciprofloxacin, cyclosporine, tacrolimus, everolimus, sirolimus, acetaminophen, fluconazole, voriconazole, posaconazole, and tazobactam, at therapeutic concentrations. We noticed that ofloxacin coeluted with CFZ. This interference could easily be detected by checking the match percentage of the UV spectra, which was 52% between ofloxacin and CFZ. The absence of interference was assessed by use of a match percentage of >90%. Although ofloxacin and CFZ are not likely to be coadministered, the prescriber should be aware that the determination of CFZ in plasma cannot be performed using this method in this specific situation.
(ii) Carryover.
Injection of blank samples in triplicate directly after injection of the calibration standard with the highest concentration showed a signal below 20% of the LLOQ for the three injections for each antibiotic and less than 5% for the IS, thus fulfilling the acceptance criteria.
(iii) Calibration and linearity.
The method was found to be linear over a concentration range of 2 to 100 mg/liter. For each analyte, the best weighting factor was 1/x2. The results for the calibration curves are summarized in Table 1. None of the backcalculated values differed by ±15% from the nominal value for each antibiotic, with a maximum bias of −4.5% being seen for the AMX ULOQ (Table 2).
TABLE 1.
Parameters of linear regression obtained from 10 calibration curves
| Antibiotica | Mean slope ± SD | Intercept | r2 |
|---|---|---|---|
| AMX | 3.36 ± 0.34 | 0.30 | 0.9975 |
| PIP | 3.66 ± 0.24 | −0.07 | 0.9960 |
| OXA | 5.13 ± 0.47 | −0.21 | 0.9957 |
| CLX | 5.19 ± 0.41 | −0.11 | 0.9962 |
| FEP | 7.48 ± 0.41 | 0.20 | 0.9953 |
| CAZ | 6.11 ± 0.72 | −0.02 | 0.9955 |
| CTX | 7.57 ± 0.62 | 1.02 | 0.9952 |
| CFZ | 4.43 ± 0.35 | 0.98 | 0.9953 |
AMX, amoxicillin; PIP, piperacillin; OXA, oxacillin; CLX, cloxacillin; FEP, cefepime; CAZ, ceftazidime; CTX, cefotaxime; CFZ, cefazolin.
TABLE 2.
Mean backcalculated and bias concentrations obtained from 10 calibration curvesa
| Nominal concn (mg/liter) | AMX |
PIP |
OXA |
CLX |
FEP |
CAZ |
CTX |
CFZ |
||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mean calculated concn ± SD (mg/liter) | Bias (%) | Mean calculated concn ± SD (mg/liter) | Bias (%) | Mean calculated concn ± SD (mg/liter) | Bias (%) | Mean calculated concn ± SD (mg/liter) | Bias (%) | Mean calculated concn ± SD (mg/liter) | Bias (%) | Mean calculated concn ± SD (mg/liter) | Bias (%) | Mean calculated concn ± SD (mg/liter) | Bias (%) | Mean calculated concn ± SD (mg/liter) | Bias (%) | |
| 2 | 1.99 ± 0.06 | −0.6 | 2.00 ± 0.06 | 0.2 | 1.99 ± 0.07 | −0.5 | 1.98 ± 0.06 | −0.9 | 1.99 ± 0.07 | −0.7 | 1.99 ± 0.06 | −0.4 | 1.99 ± 0.06 | −0.6 | 1.98 ± 0.07 | −0.9 |
| 5 | 5.07 ± 0.35 | 1.4 | 5.01 ± 0.31 | 0.2 | 5.06 ± 0.34 | 1.3 | 5.11 ± 0.30 | 2.1 | 5.09 ± 0.33 | 1.8 | 5.08 ± 0.31 | 1.5 | 5.08 ± 0.30 | 1.6 | 5.10 ± 0.36 | 1.9 |
| 10 | 9.90 ± 0.47 | −1 | 9.77 ± 0.44 | −2.3 | 9.87 ± 0.56 | −1.3 | 9.93 ± 0.50 | −0.7 | 9.98 ± 0.73 | −0.2 | 9.82 ± 0.54 | −1.8 | 9.91 ± 0.51 | −0.9 | 10.03 ± 0.62 | 0.3 |
| 25 | 25.59 ± 1.34 | 2.3 | 25.36 ± 1.26 | 1.5 | 25.47 ± 1.43 | 1.9 | 25.28 ± 1.32 | 1.1 | 25.23 ± 1.25 | 0.9 | 25.60 ± 1.22 | 2.4 | 25.48 ± 1.25 | 1.9 | 25.30 ± 1.32 | 1.2 |
| 50 | 51.34 ± 1.46 | 2.7 | 50.95 ± 1.06 | 1.9 | 51.06 ± 1.48 | 2.1 | 51.04 ± 1.41 | 2.1 | 50.89 ± 1.65 | 1.8 | 51.25 ± 1.70 | 2.5 | 51.11 ± 1.29 | 2.2 | 51.15 ± 1.19 | 2.3 |
| 75 | 74.78 ± 3.48 | −0.3 | 75.87 ± 3.31 | 1.2 | 74.83 ± 3.34 | −0.2 | 74.97 ± 3.25 | 0 | 74.05 ± 4.54 | −1.3 | 75.21 ± 3.84 | 0.3 | 74.68 ± 3.53 | −0.4 | 74.63 ± 3.58 | −0.5 |
| 100 | 95.46 ± 6.31 | −4.5 | 97.37 ± 5.05 | −2.6 | 96.65 ± 5.82 | −3.3 | 96.20 ± 6.01 | −3.8 | 97.63 ± 5.87 | −2.4 | 95.62 ± 6.77 | −4.4 | 96.16 ± 6.46 | −3.8 | 95.69 ± 6.47 | −4.3 |
AMX, amoxicillin; PIP, piperacillin; OXA, oxacillin; CLX, cloxacillin; FEP, cefepime; CAZ, ceftazidime; CTX, cefotaxime; CFZ, cefazolin.
(iv) Accuracy and precision.
The acceptance criteria for accuracy and precision were met for all compounds. Within-run bias ranged from −5.2% (for AMX) to 11.4% (for CFZ), whereas between-run bias ranged from −0.4% (for AMX) to 7.0% (for CFZ). Within- and between-run precision were lower than 7.2% (for CAZ) and 14.2% (for AMX), respectively (Table 3).
TABLE 3.
Within- and between-run precision (CV) and accuracy (bias)a
| Antibiotic and nominal concn (mg/liter) | Within run (n = 12) |
Between run (n = 24) |
||||
|---|---|---|---|---|---|---|
| Mean measured concn ± SD (mg/liter) | CV (%) | Bias (%) | Mean measured concn ± SD (mg/liter) | CV (%) | Bias (%) | |
| AMX | ||||||
| 2b | 2.03 ± 0.07 | 3.2 | 1.5 | 2.09 ± 0.30 | 14.2 | 4.7 |
| 6 | 5.69 ± 0.14 | 2.5 | −5.2 | 6.15 ± 0.52 | 8.4 | 2.5 |
| 30 | 29.46 ± 1.10 | 3.7 | −1.8 | 29.87 ± 2.21 | 7.4 | −0.4 |
| 70 | 70.40 ± 4.98 | 7.1 | 0.6 | 69.75 ± 6.56 | 9.4 | −0.4 |
| PIP | ||||||
| 2b | 2.11 ± 0.04 | 2.1 | 5.3 | 2.07 ± 0.23 | 11.3 | 3.5 |
| 6 | 5.97 ± 0.15 | 2.4 | −0.4 | 6.20 ± 0.43 | 6.9 | 3.4 |
| 30 | 30.06 ± 1.05 | 3.5 | 0.2 | 30.59 ± 1.95 | 6.4 | 2.0 |
| 70 | 71.54 ± 4.84 | 6.8 | 2.2 | 72.28 ± 5.43 | 7.5 | 3.3 |
| OXA | ||||||
| 2b | 2.12 ± 0.04 | 1.8 | 5.9 | 2.07 ± 0.25 | 12.1 | 3.3 |
| 6 | 5.96 ± 0.13 | 2.1 | −0.6 | 6.29 ± 0.45 | 7.2 | 4.8 |
| 30 | 29.28 ± 1.03 | 3.5 | −2.4 | 30.94 ± 1.78 | 5.8 | 3.1 |
| 70 | 67.75 ± 4.63 | 6.8 | −3.2 | 69.98 ± 5.64 | 8.1 | 0.0 |
| CLX | ||||||
| 2b | 2.12 ± 0.06 | 2.6 | 6.0 | 2.06 ± 0.25 | 12.2 | 2.9 |
| 6 | 5.90 ± 0.13 | 2.2 | −1.6 | 6.37 ± 0.37 | 5.8 | 6.2 |
| 30 | 29.15 ± 1.01 | 3.5 | −2.8 | 31.22 ± 1.58 | 5.1 | 4.1 |
| 70 | 68.72 ± 4.62 | 6.7 | −1.8 | 70.84 ± 5.38 | 7.6 | 1.2 |
| FEP | ||||||
| 2b | 2.12 ± 0.04 | 2.0 | 5.9 | 2.09 ± 0.25 | 12.2 | 4.4 |
| 6 | 6.10 ± 0.12 | 2.0 | 1.6 | 6.13 ± 0.51 | 8.2 | 2.2 |
| 30 | 30.27 ± 1.02 | 3.4 | 0.9 | 30.33 ± 2.38 | 7.9 | 1.1 |
| 70 | 70.93 ± 4.72 | 6.6 | 1.3 | 73.27 ± 4.67 | 6.4 | 4.7 |
| CAZ | ||||||
| 2b | 2.11 ± 0.04 | 2.1 | 5.5 | 2.03 ± 0.26 | 12.9 | 1.5 |
| 6 | 5.94 ± 0.13 | 2.2 | −1.1 | 6.14 ± 0.43 | 7.1 | 2.3 |
| 30 | 32.86 ± 1.45 | 4.4 | 9.5 | 31.07 ± 2.29 | 7.4 | 3.6 |
| 70 | 70.02 ± 5.01 | 7.2 | 0.0 | 71.44 ± 4.44 | 6.2 | 2.1 |
| CTX | ||||||
| 2b | 2.12 ± 0.04 | 2.1 | 6.2 | 1.99 ± 0.26 | 13.2 | −0.4 |
| 6 | 6.11 ± 0.13 | 2.1 | 1.8 | 6.22 ± 0.43 | 7.0 | 3.7 |
| 30 | 30.01 ± 1.03 | 3.4 | 0.0 | 30.98 ± 1.84 | 5.9 | 3.3 |
| 70 | 70.53 ± 4.80 | 6.8 | 0.8 | 70.79 ± 6.31 | 8.9 | 1.1 |
| CFZ | ||||||
| 2b | 2.23 ± 0.05 | 2.2 | 11.4 | 2.09 ± 0.25 | 12.1 | 4.7 |
| 6 | 6.19 ± 0.12 | 2.0 | 3.1 | 6.42 ± 0.47 | 7.4 | 7.0 |
| 30 | 30.04 ± 1.01 | 3.3 | 0.1 | 31.05 ± 1.70 | 5.5 | 3.5 |
| 70 | 70.89 ± 4.79 | 6.8 | 1.3 | 70.68 ± 6.30 | 8.9 | 1.0 |
CV, coefficient of variation; AMX, amoxicillin; PIP, piperacillin; OXA, oxacillin; CLX, cloxacillin; FEP, cefepime; CAZ, ceftazidime; CTX, cefotaxime; CFZ, cefazolin.
Concentration corresponding to the LLOQ.
(v) LLOQ.
The method showed an LLOQ of 2 mg/liter, corresponding to 5 ng injected in the column. The precision and accuracy of LLOQ are presented in Table 3. This LLOQ is similar to that of previously reported HPLC-UV methods (range, 2 to 10 mg/liter) (19–21). Lower LLOQs of HPLC-UV methods were described, but a larger sample volume, up to 500 μl, was required, resulting in significant blood spoliation (17, 18). Of course, mass spectrometry detection offers higher sensitivity but requires a material that is not readily available in most laboratories.
A repeatability study (n = 12) with concentrations lower than the LLOQ was nonetheless conducted and showed that an LLOQ of 1 mg/liter may be acceptable for CLX (CV = 3.0%, bias = −1.3%), FEP (CV = 3.2%, bias = −5.3%), CAZ (CV = 3.2%, bias = −3.7%), and CTX (CV = 3.0%, bias = 5.1%). An LLOQ of 0.5 mg/liter could be considered for AMX (CV = 9.9%, bias = −11.1%), OXA (CV = 10.0%, bias = 7.8%), and CFZ (CV = 3.2%, bias = 4.2%). An LLOQ lower than 2 mg/liter was not validated for PIP. In order to simplify the validation process and implementation of the method in routine practice, the LLOQ was set at 2 mg/liter for all assayed compounds.
To date, there is no consensus regarding the target concentration of β-lactam antibiotics. Several studies have reported that high PK-PD targets are associated with better outcomes in critically ill patients (3, 4, 9, 10, 13). According to many authors, our target was set at a trough concentration above 4 to 5 times the MIC in order to reduce treatment failure and ensure a sufficient tissue exposure (2–5, 7, 10). The breakpoint MICs for problematic pathogens in critically ill patients ranged from 0.5 to 4 mg/liter and from 8 to 32 mg/liter for Staphylococcus aureus and Pseudomonas aeruginosa, respectively (according to EUCAST [30]). Therefore, an LLOQ of 2 mg/liter meets the clinical needs for adequate TDM in most clinical applications.
(vi) Dilution integrity.
The dilution integrity study showed an acceptable accuracy and precision for samples diluted 2-fold. Indeed, bias ranged from −5.2 and −1.0% for all compounds, with CV being <5.3% (n = 6). The dilution of patient plasma samples in which the concentration is found to be higher than the ULOQ in clinical practice is therefore validated.
(vii) Extraction recoveries.
Mean extraction recoveries from the QC samples with the three concentrations are shown in Table 4. The lowest mean recovery was 76.5% for AMX, and the highest was 96.8% for PIP.
TABLE 4.
Recoveries obtained from the QC samples with one of three concentrationsa
| Concn (mg/liter) | n | Mean recovery ± SD (%) |
||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| AMX | PIP | OXA | CLX | FEP | CAZ | CTX | CFZ | IS | ||
| 6 | 6 | 74.8 ± 3.1 | 98.5 ± 2.6 | 90.2 ± 2.3 | 93.6 ± 2.1 | 91.9 ± 2.8 | 74.5 ± 2.8 | 92.8 ± 3.0 | 93.9 ± 2.5 | 88.3 ± 5.5 |
| 30 | 6 | 73.9 ± 3.5 | 92.4 ± 3.4 | 92.5 ± 2.5 | 98.7 ± 2.7 | 87.2 ± 3.0 | 79.4 ± 3.8 | 87.0 ± 3.1 | 87.4 ± 2.9 | 86.1 ± 3.3 |
| 70 | 6 | 80.8 ± 1.3 | 99.5 ± 1.4 | 84.5 ± 1.2 | 91.6 ± 1.4 | 93.4 ± 1.3 | 76.0 ± 1.7 | 93.0 ± 1.5 | 93.5 ± 1.6 | 83.8 ± 7.4 |
| Total | 18 | 76.5 ± 4.9 | 96.8 ± 4.1 | 89.0 ± 4.5 | 94.7 ± 3.9 | 90.8 ± 3.8 | 76.7 ± 3.9 | 90.9 ± 4.0 | 91.6 ± 4.0 | 86.1 ± 5.7 |
n, number of replicates; AMX, amoxicillin; PIP, piperacillin; OXA, oxacillin; CLX, cloxacillin; FEP, cefepime; CAZ, ceftazidime; CTX, cefotaxime; CFZ, cefazolin; IS, internal standard.
Stability.
Seven of the eight stock solutions of each antibiotic, prepared at 100 mg/liter in water, were stable over 4 months at −80°C. The AMX stock solution prepared at 100 mg/ml in water showed significant degradation (−23.3%) and should be prepared extemporaneously. For the other compounds, no significant degradation was found between stock solutions stored at −80°C and freshly prepared solutions (the difference was less than 15%). Aliquots of mix solutions at 5 mg/ml containing each analyte, including AMX, were stable over 4 months at −80°C (Table 5). The concentration-dependent stability of aqueous AMX solutions was previously described in the literature (31, 32).
TABLE 5.
Stability of stock solutions and mix solution
| Antibiotica | Stabilityb |
|
|---|---|---|
| Stock solution, 100 mg/ml in water | Mix solution, 5 mg/ml in water | |
| AMX | −23.3 | −10.5 |
| PIP | −4.7 | −7.7 |
| OXA | −10.8 | −11.9 |
| CLX | −1.1 | −2.2 |
| FEP | 7.4 | 3.3 |
| CAZ | −6.7 | −8.7 |
| CTX | −3.1 | −3.6 |
| CFZ | 3.2 | −1.4 |
| IS | −5.2 | 3.3 |
AMX, amoxicillin; PIP, piperacillin; OXA, oxacillin; CLX, cloxacillin; FEP, cefepime; CAZ, ceftazidime; CTX, cefotaxime; CFZ, cefazolin; IS, internal standard.
Stability data represent the difference in the signal between solutions stored at −80°C for 4 months and the signal of freshly prepared solutions. Stock solutions were prepared independently, and mix solutions were prepared simultaneously.
The results of analysis of the stability of the QC samples under different storage conditions are reported in Table 6. QC samples stored at −80°C in polypropylene tubes were stable for at least 2 months (bias range, −5.1 to 7.8% for all compounds and for the QC samples with the three concentrations). Three consecutive freeze-thaw cycles resulted in no significant degradation of any of the tested compounds (bias range, −0.1 to 14.6%). The QC samples were stable on a benchtop at room temperature for at least 6 h (bias range, −3.0 to 7.8%). QC sample extracts were stable when they were held in an autosampler kept at +4°C for at least 72 h, since the bias ranged from −13.4 to 10.4%. These results are consistent with those of previously reported stability studies. According to the literature, plasma samples kept on a benchtop at room temperature are stable for from 4 h to 12 h (17, 20, 26). The length of time that the QC samples were reported to be stable in a refrigerated autosampler postextraction ranged from 9 h to 24 h (19, 20, 24–27). The long-term stability of QC samples is more controversial. McWhinney et al. reported that QC samples with CAZ, CFZ, and PIP stored at −70°C were stable for 8 months (19). Nemutlu et al. found that QC samples with FEP, CAZ, and CTX stored at −20°C were stable for 2 months (18), whereas Colin et al. reported a stability of shorter than 4 days at −20°C for QC samples with AMX, FEP, and CAZ (24).
TABLE 6.
Stability of QC samples stored under different conditionsa
| Antibiotic and nominal concn (mg/liter) | Stability of QC samples stored at −80°C for over 2 mo (n = 3) |
Stability of QC samples after 3 freeze-thaw cycles (n = 3) |
Stability of QC samples stored at room temp for 6 h (n = 3) |
Stability of QC sample extracts stored at +4°C for 72 h (n = 3) |
||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Mean measured concn ± SD (mg/liter) | CV (%) | Bias (%) | Mean measured concn ± SD (mg/liter) | CV (%) | Bias (%) | Mean measured concn ± SD (mg/liter) | CV (%) | Bias (%) | Mean measured concn ± SD (mg/liter) | CV (%) | Bias (%) | |
| AMX | ||||||||||||
| 6 | 6.42 ± 0.08 | 1.2 | 7.1 | 6.38 ± 0.21 | 3.4 | 6.4 | 6.28 ± 0.11 | 1.7 | 4.7 | 5.52 ± 0.06 | 1.2 | −8.1 |
| 30 | 32.05 ± 0.66 | 2.1 | 6.8 | 34.39 ± 0.52 | 1.5 | 14.6 | 32.33 ± 0.54 | 1.7 | 7.8 | 28.77 ± 1.09 | 3.8 | −4.1 |
| 70 | 73.18 ± 1.04 | 1.4 | 4.5 | 79.11 ± 1.20 | 1.5 | 13.0 | 72.22 ± 4.13 | 5.7 | 3.2 | 67.98 ± 3.23 | 4.8 | −2.9 |
| PIP | ||||||||||||
| 6 | 5.96 ± 0.03 | 0.5 | −0.6 | 6.33 ± 0.14 | 2.2 | 5.4 | 6.02 ± 0.33 | 5.5 | 0.3 | 6.23 ± 0.09 | 1.5 | 3.8 |
| 30 | 29.15 ± 0.72 | 2.5 | −2.8 | 31.25 ± 3.78 | 12.1 | 4.2 | 30.62 ± 0.48 | 1.6 | 2.1 | 30.13 ± 0.94 | 3.1 | 0.4 |
| 70 | 72.34 ± 2.18 | 3.0 | 3.3 | 76.14 ± 4.06 | 5.3 | 8.8 | 71.42 ± 3.24 | 4.5 | 2.0 | 60.61 ± 2.69 | 4.4 | −13.4 |
| OXA | ||||||||||||
| 6 | 6.21 ± 0.21 | 3.4 | 3.4 | 6.51 ± 0.18 | 2.7 | 8.6 | 6.19 ± 0.27 | 4.4 | 3.1 | 5.52 ± 0.07 | 1.3 | −8.1 |
| 30 | 30.17 ± 0.51 | 1.7 | 0.6 | 33.71 ± 0.74 | 2.2 | 12.4 | 31.24 ± 0.56 | 1.8 | 4.1 | 27.02 ± 1.03 | 3.8 | −9.9 |
| 70 | 66.56 ± 1.13 | 1.7 | −4.9 | 77.66 ± 0.22 | 0.3 | 10.9 | 71.38 ± 3.34 | 4.7 | 2.0 | 67.75 ± 4.63 | 6.8 | −13.4 |
| CLX | ||||||||||||
| 6 | 6.20 ± 0.85 | 13.8 | 3.3 | 6.00 ± 0.39 | 6.6 | −0.1 | 6.40 ± 0.07 | 1.1 | 6.7 | 5.68 ± 0.06 | 1.1 | −5.4 |
| 30 | 31.81 ± 0.26 | 0.8 | 6.0 | 33.35 ± 0.57 | 1.7 | 11.2 | 30.62 ± 0.53 | 1.7 | 2.1 | 27.70 ± 1.05 | 3.8 | −7.7 |
| 70 | 70.0 ± 1.28 | 1.8 | 0.0 | 77.13 ± 0.11 | 0.1 | 10.2 | 70.19 ± 3.31 | 4.7 | 0.3 | 62.58 ± 2.86 | 4.6 | −10.6 |
| FEP | ||||||||||||
| 6 | 6.06 ± 0.05 | 0.8 | 1.1 | 6.28 ± 0.14 | 2.2 | 4.7 | 5.92 ± 0.01 | 0.1 | −1.4 | 6.09 ± 0.11 | 1.8 | 1.6 |
| 30 | 30.74 ± 0.55 | 1.8 | 2.5 | 32.61 ± 0.40 | 1.2 | 8.7 | 30.16 ± 0.46 | 1.5 | 0.5 | 30.32 ± 0.93 | 3.1 | 1.1 |
| 70 | 74.48 ± 1.95 | 2.6 | 6.4 | 75.55 ± 0.53 | 0.7 | 7.9 | 67.92 ± 3.22 | 4.7 | −3.0 | 69.34 ± 2.85 | 4.1 | −0.9 |
| CAZ | ||||||||||||
| 6 | 6.03 ± 0.05 | 0.8 | 0.4 | 6.48 ± 0.10 | 1.6 | 8.1 | 6.12 ± 0.02 | 0.3 | 2.1 | 5.95 ± 0.10 | 1.7 | −0.9 |
| 30 | 30.30 ± 0.26 | 0.8 | 1.0 | 33.98 ± 0.32 | 0.9 | 13.3 | 31.54 ± 0.57 | 1.8 | 5.1 | 33.67 ± 1.58 | 4.7 | 10.4 |
| 70 | 72.72 ± 1.70 | 2.3 | 3.9 | 77.64 ± 0.56 | 0.7 | 10.9 | 70.74 ± 4.27 | 6.0 | 1.1 | 68.52 ± 2.89 | 4.2 | −2.1 |
| CTX | ||||||||||||
| 6 | 5.91 ± 0.05 | 0.9 | −1.6 | 6.47 ± 0.16 | 2.4 | 7.8 | 6.14 ± 0.01 | 0.2 | 2.3 | 6.04 ± 0.09 | 1.5 | 0.7 |
| 30 | 29.96 ± 0.27 | 0.9 | −0.1 | 33.17 ± 0.39 | 1.2 | 10.6 | 30.98 ± 0.44 | 1.4 | 3.3 | 29.80 ± 0.91 | 3.1 | −0.7 |
| 70 | 66.43 ± 0.63 | 0.9 | −5.1 | 77.16 ± 0.81 | 1.1 | 10.2 | 71.52 ± 3.33 | 4.7 | 2.0 | 69.01 ± 2.86 | 4.1 | −1.4 |
| CFZ | ||||||||||||
| 6 | 6.42 ± 0.06 | 0.9 | 6.9 | 6.63 ± 0.13 | 2.0 | 10.5 | 6.24 ± 0.01 | 0.1 | 3.9 | 6.39 ± 0.09 | 1.4 | 6.5 |
| 30 | 32.34 ± 0.38 | 1.2 | 7.8 | 34.10 ± 0.34 | 1.0 | 13.7 | 32.34 ± 0.48 | 1.5 | 7.8 | 30.65 ± 0.90 | 2.9 | 2.2 |
| 70 | 71.24 ± 0.62 | 0.9 | 1.8 | 79.76 ± 0.77 | 1.0 | 13.9 | 73.70 ± 3.36 | 4.6 | 5.3 | 69.67 ± 2.91 | 4.2 | −0.5 |
n, number of replicates; CV, coefficient of variation; AMX, amoxicillin; PIP, piperacillin; OXA, oxacillin; CLX, cloxacillin; FEP, cefepime; CAZ, ceftazidime; CTX, cefotaxime; CFZ, cefazolin.
Application to patient samples.
This method is currently performed in our laboratories for TDM purposes. Representative chromatograms obtained from selected patient samples are shown in Fig. 3. Plasma samples were obtained from critically ill patients hospitalized in different departments, including cardiac surgery, nephrology, and intensive care units. So far, the plasma concentrations measured in patient samples have been within the calibration range.
FIG 3.
Chromatograms of plasma samples from patients treated with β-lactam antibiotics. (A) Amoxicillin (AMX) at a concentration of 18.0 mg/liter; (B) piperacillin (PIP) at a concentration of 95.5 mg/liter; (C) cefepime (FEP) at a concentration of 78.3 mg/liter; (D) cefotaxime (CTX) at a concentration of 8.1 mg/liter; (E) cefazolin (CFZ) at a concentration of 32.1 mg/liter; (F) ceftazidime (CAZ) at a concentration of 43.1 mg/liter.
Conclusion.
An UHPLC-UV method with simple sample preparation requiring a small volume of plasma (100 μl) was developed for the simultaneous determination of eight β-lactam antibiotics. All validation parameters meet the required international criteria. This method is very useful in daily clinical practice and allows the accurate and responsive TDM of several β-lactam antibiotics, which may improve the treatment of patients with serious infections.
REFERENCES
- 1.Kohanski MA, Dwyer DJ, Collins JJ. 2010. How antibiotics kill bacteria: from targets to networks. Nat Rev Microbiol 8:423–435. doi: 10.1038/nrmicro2333. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Aubert G, Carricajo A, Coudrot M, Guyomarc'h S, Auboyer C, Zeni F. 2010. Prospective determination of serum ceftazidime concentrations in intensive care units. Ther Drug Monit 32:517–519. doi: 10.1097/FTD.0b013e3181e60ca6. [DOI] [PubMed] [Google Scholar]
- 3.Roberts JA, Ulldemolins M, Roberts MS, McWhinney B, Ungerer J, Paterson DL, Lipman J. 2010. Therapeutic drug monitoring of beta-lactams in critically ill patients: proof of concept. Int J Antimicrob Agents 36:332–339. doi: 10.1016/j.ijantimicag.2010.06.008. [DOI] [PubMed] [Google Scholar]
- 4.Taccone FS, Laterre P-F, Dugernier T, Spapen H, Delattre I, Wittebole X, De Backer D, Layeux B, Wallemacq P, Vincent J-L, Jacobs F. 2010. Insufficient β-lactam concentrations in the early phase of severe sepsis and septic shock. Crit Care 14:R126. doi: 10.1186/cc9091. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Gonçalves-Pereira J, Póvoa P. 2011. Antibiotics in critically ill patients: a systematic review of the pharmacokinetics of β-lactams. Crit Care 15:R206. doi: 10.1186/cc10441. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Urakami T, Aoki Y, Fukuoka M, Magarifuchi H, Nagata M, Nagasawa Z, Nakano Y, Fujito H. 2011. Application of beta-lactam therapeutic drug monitoring in clinical practice using HPLC. J Bioanal Biomed 3:1–5. doi: 10.4155/bio.10.180. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Carlier M, Noë M, De Waele JJ, Stove V, Verstraete AG, Lipman J, Roberts JA. 2013. Population pharmacokinetics and dosing simulations of amoxicillin/clavulanic acid in critically ill patients. J Antimicrob Chemother 68:2600–2608. doi: 10.1093/jac/dkt240. [DOI] [PubMed] [Google Scholar]
- 8.Huttner A, Harbarth S, Hope WW, Lipman J, Roberts JA. 2015. Therapeutic drug monitoring of the β-lactam antibiotics: what is the evidence and which patients should we be using it for? J Antimicrob Chemother 70:3178–3183. doi: 10.1093/jac/dkv201. [DOI] [PubMed] [Google Scholar]
- 9.Chastre J, Luyt C-E. 2015. Continuous β-lactam infusion to optimize antibiotic use for severe sepsis. A knife cutting water? Am J Respir Crit Care Med 192:1266–1268. doi: 10.1164/rccm.201507-1487ED. [DOI] [PubMed] [Google Scholar]
- 10.Martínková J, Malbrain MLNG, Havel E, Šafránek P, Bezouška J, Kaška M. 2016. A pilot study on pharmacokinetic/pharmacodynamic target attainment in critically ill patients receiving piperacillin/tazobactam. Anaesthesiol Intensive Ther 48:23–28. doi: 10.5603/AIT.a2015.0082. [DOI] [PubMed] [Google Scholar]
- 11.Chow KM, Hui AC, Szeto CC. 2005. Neurotoxicity induced by beta-lactam antibiotics: from bench to bedside. Eur J Clin Microbiol Infect Dis 24:649–653. doi: 10.1007/s10096-005-0021-y. [DOI] [PubMed] [Google Scholar]
- 12.Grill MF, Maganti RK. 2011. Neurotoxic effects associated with antibiotic use: management considerations. Br J Clin Pharmacol 72:381–393. doi: 10.1111/j.1365-2125.2011.03991.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Hayashi Y, Lipman J, Udy AA, Ng M, McWhinney B, Ungerer J, Lust K, Roberts JA. 2013. β-Lactam therapeutic drug monitoring in the critically ill: optimising drug exposure in patients with fluctuating renal function and hypoalbuminaemia. Int J Antimicrob Agents 41:162–166. doi: 10.1016/j.ijantimicag.2012.10.002. [DOI] [PubMed] [Google Scholar]
- 14.Roberts JA, Abdul-Aziz MH, Lipman J, Mouton JW, Vinks AA, Felton TW, Hope WW, Farkas A, Neely MN, Schentag JJ, Drusano G, Frey OR, Theuretzbacher U, Kuti JL. 2014. Individualised antibiotic dosing for patients who are critically ill: challenges and potential solutions. Lancet Infect Dis 14:498–509. doi: 10.1016/S1473-3099(14)70036-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.El-Shaboury SR, Saleh GA, Mohamed FA, Rageh AH. 2007. Analysis of cephalosporin antibiotics. J Pharm Biomed Anal 45:1–19. doi: 10.1016/j.jpba.2007.06.002. [DOI] [PubMed] [Google Scholar]
- 16.Carlier M, Stove V, Wallis SC, De Waele JJ, Verstraete AG, Lipman J, Roberts JA. 2015. Assays for therapeutic drug monitoring of β-lactam antibiotics: a structured review. Int J Antimicrob Agents 46:367–375. doi: 10.1016/j.ijantimicag.2015.06.016. [DOI] [PubMed] [Google Scholar]
- 17.Denooz R, Charlier C. 2008. Simultaneous determination of five beta-lactam antibiotics (cefepim, ceftazidim, cefuroxim, meropenem and piperacillin) in human plasma by high-performance liquid chromatography with ultraviolet detection. J Chromatogr B Analyt Technol Biomed Life Sci 864:161–167. doi: 10.1016/j.jchromb.2008.01.037. [DOI] [PubMed] [Google Scholar]
- 18.Nemutlu E, Kir S, Katlan D, Beksaç MS. 2009. Simultaneous multiresponse optimization of an HPLC method to separate seven cephalosporins in plasma and amniotic fluid: application to validation and quantification of cefepime, cefixime and cefoperazone. Talanta 80:117–126. doi: 10.1016/j.talanta.2009.06.034. [DOI] [PubMed] [Google Scholar]
- 19.McWhinney BC, Wallis SC, Hillister T, Roberts JA, Lipman J, Ungerer JPJ. 2010. Analysis of 12 beta-lactam antibiotics in human plasma by HPLC with ultraviolet detection. J Chromatogr B Analyt Technol Biomed Life Sci 878:2039–2043. doi: 10.1016/j.jchromb.2010.05.027. [DOI] [PubMed] [Google Scholar]
- 20.Verdier M-C, Tribut O, Tattevin P, Le Tulzo Y, Michelet C, Bentué-Ferrer D. 2011. Simultaneous determination of 12 beta-lactam antibiotics in human plasma by high-performance liquid chromatography with UV detection: application to therapeutic drug monitoring. Antimicrob Agents Chemother 55:4873–4879. doi: 10.1128/AAC.00533-11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Wolff F, Deprez G, Seyler L, Taccone F, Hites M, Gulbis B, Vincent J-L, Jacobs F, Cotton F. 2013. Rapid quantification of six β-lactams to optimize dosage regimens in severely septic patients. Talanta 103:153–160. doi: 10.1016/j.talanta.2012.10.024. [DOI] [PubMed] [Google Scholar]
- 22.Ohmori T, Suzuki A, Niwa T, Ushikoshi H, Shirai K, Yoshida S, Ogura S, Itoh Y. 2011. Simultaneous determination of eight β-lactam antibiotics in human serum by liquid chromatography-tandem mass spectrometry. J Chromatogr B Analyt Technol Biomed Life Sci 879:1038–1042. doi: 10.1016/j.jchromb.2011.03.001. [DOI] [PubMed] [Google Scholar]
- 23.Carlier M, Stove V, Roberts JA, Van de Velde E, De Waele JJ, Verstraete AG. 2012. Quantification of seven β-lactam antibiotics and two β-lactamase inhibitors in human plasma using a validated UPLC-MS/MS method. Int J Antimicrob Agents 40:416–422. doi: 10.1016/j.ijantimicag.2012.06.022. [DOI] [PubMed] [Google Scholar]
- 24.Colin P, De Bock L, T'jollyn H, Boussery K, Van Bocxlaer J. 2013. Development and validation of a fast and uniform approach to quantify β-lactam antibiotics in human plasma by solid phase extraction-liquid chromatography-electrospray-tandem mass spectrometry. Talanta 103:285–293. doi: 10.1016/j.talanta.2012.10.046. [DOI] [PubMed] [Google Scholar]
- 25.Cazorla-Reyes R, Romero-González R, Frenich AG, Rodríguez Maresca MA, Martínez Vidal JL. 2014. Simultaneous analysis of antibiotics in biological samples by ultra high performance liquid chromatography-tandem mass spectrometry. J Pharm Biomed Anal 89:203–212. doi: 10.1016/j.jpba.2013.11.004. [DOI] [PubMed] [Google Scholar]
- 26.Sime FB, Roberts MS, Roberts JA, Robertson TA. 2014. Simultaneous determination of seven β-lactam antibiotics in human plasma for therapeutic drug monitoring and pharmacokinetic studies. J Chromatogr B Analyt Technol Biomed Life Sci 960:134–144. doi: 10.1016/j.jchromb.2014.04.029. [DOI] [PubMed] [Google Scholar]
- 27.Carlier M, Stove V, De Waele JJ, Verstraete AG. 2015. Ultrafast quantification of β-lactam antibiotics in human plasma using UPLC-MS/MS. J Chromatogr B Analyt Technol Biomed Life Sci 978-979:89–94. doi: 10.1016/j.jchromb.2014.11.034. [DOI] [PubMed] [Google Scholar]
- 28.European Medicines Agency. 2011. Guideline on bioanalytical method validation. European Medicines Agency, London, United Kingdom: http://www.ema.europa.eu/docs/en_GB/document_library/Scientific_guideline/2011/08/WC500109686.pdf. [Google Scholar]
- 29.Briscoe SE, McWhinney BC, Lipman J, Roberts JA, Ungerer JPJ. 2012. A method for determining the free (unbound) concentration of ten beta-lactam antibiotics in human plasma using high performance liquid chromatography with ultraviolet detection. J Chromatogr B Analyt Technol Biomed Life Sci 907:178–184. doi: 10.1016/j.jchromb.2012.09.016. [DOI] [PubMed] [Google Scholar]
- 30.European Committee on Antimicrobial Susceptibility Testing (EUCAST). Clinical breakpoints. http://www.eucast.org/clinical_breakpoints/.
- 31.Cook B, Hill SA, Lynn B. 1982. The stability of amoxycillin sodium in intravenous infusion fluids. J Clin Hosp Pharm 7:245–250. [DOI] [PubMed] [Google Scholar]
- 32.Arlicot N, Marie A, Cade C, Laffon M, Antier D. 2011. Stability of amoxicillin in portable pumps is drug concentration dependent. Pharmazie 66:631–632. [PubMed] [Google Scholar]