Abstract
Background
Since severe infections frequently cause acute kidney injury (AKI), continuous renal replacement therapy (CRRT) is often initiated for regulation of inflammatory mediators and renal support. Thus, it is necessary to decide the antibiotic dosage considering the CRRT clearance in addition to residual renal function. Some of the hemofilters used in CRRT are known to adsorb antibiotics, and clearance of antibiotics may differ depending on the adsorptive characteristics of hemofilters. Although assay systems for blood and CRRT filtrate concentrations are required, no method for measuring antibiotics concentrations in filtrate has been reported. We developed a UHPLC–MS/MS method for simultaneous quantification of antibiotics commonly used in ICU, comprising carbapenems [doripenem (DRPM) and meropenem (MEPM)], quinolones [ciprofloxacin (CPFX), levofloxacin (LVFX) and pazufloxacin (PZFX)] and anti‐MRSA agents [linezolid (LZD), and tedizolid (TZD)] in CRRT filtrate samples.
Methods
Filtrate samples were pretreated by protein precipitation. The analytes were separated with an ACQUITY UHPLC CSH C18 column under a gradient mobile phase consisting of water and acetonitrile containing 0.1% formic acid and 2 mM ammonium formate.
Results
The method showed good linearity over wide ranges. Within‐batch and batch‐to‐batch accuracy and precision for each drug fulfilled the criteria of the US Food and Drug Administration guidance. The recovery rate was more than 87.20%. Matrix effect ranged from 99.57% to 115.60%. Recovery rate and matrix effect did not differ remarkably between quality control samples at different concentrations.
Conclusion
This is the first report of a simultaneous quantification method of multiple antibiotics in filtrate of CRRT circuit.
Keywords: antibiotics, continuous renal replacement therapy, filtrate, hemofilter, intensive care unit, ultra‐high‐performance liquid chromatography
Chromatograms for measurement of doripenem (DRPM), meropenem (MEPM), linezolid (LZD), tedizolid (TZD), ciprofloxacin (CPFX), levofloxacin (LVFX), and pazufloxacin (PZFX) in lower limit of quantitation (LLOQ) filtrate sample (left) compared with blank filtrate (right).
1. INTRODUCTION
For the treatment of severe infections in intensive care unit (ICU), administration of broad‐spectrum antibiotics selected based on the assumed focus of infection and causative bacteria is recommended. 1 Sepsis often causes acute kidney injury (AKI) due to factors such as shock, decreased circulating blood flow due to systemic inflammatory response syndrome, inflammation in the kidney, and tubular damage. 2 According to a cohort study of 1177 patients with sepsis in ICU conducted at 198 institutions in 24 European countries, 51% of the patients developed AKI with mortality of 41%, 3 indicating that septic AKI in the ICU is one of the important complications. Continuous renal replacement therapy (CRRT) is often initiated in patients with septic AKI for the regulation of inflammatory mediators and renal support. However, since drugs smaller than the membrane pores of hemofilters are removed by filtration and diffusion, drugs having properties such as small molecular weight, small volume of distribution, hydrophilicity, and low protein binding rate are susceptible to removal. 4 Thus, it is necessary to decide the antibiotic dosage considering the CRRT clearance in addition to the residual renal function.
Hemofilters used in CRRT are classified into two types based on their cytokine adsorption properties: non‐adsorptive membranes such as asymmetric cellulose triacetate (ATA) and polyethersulfone (PES) membranes, and adsorptive membranes such as polymethyl methacrylate (PMMA) and AN69ST membranes. 5 , 6 These hemofilters have been reported to adsorb some antimicrobials. Adsorptive membranes adsorb some drugs by their hydrophobic and ionic bonding properties. 7 , 8 , 9 , 10 , 11 On the contrary, non‐adsorptive membranes also adsorb some drugs, although the mode of binding remains unclear. 12 , 13 , 14 , 15 In vitro studies showed 90% adsorption of gentamicin on polyacrylonitrile copolymer (PAN) membranes, 14 90% adsorption of tigecycline on polysulfone (PS) membranes, 13 61.4% adsorption of teicoplanin on PS membranes, 75.6% on PMMA membranes, 9 and 22% adsorption of linezolid on PMMA membranes. 7 In clinical practice, blood teicoplanin concentration was lower than expected when administered under CRRT using PMMA membranes, indicating adsorption of teicoplanin on PMMA membranes. 16 These suggest the need to adjust the dose according to the type of membrane. However, because the dosage of antibiotics in patients under CRRT is generally set considering filtration and diffusion only, without taking into account the differences in adsorption properties between hemofilters, 17 some combinations of hemofilters and antibiotics may fail to achieve pharmacokinetic and pharmacodynamics goals, resulting in treatment failure or emergence of resistant bacteria.
In addition, only in vitro or ex vivo studies on the adsorption of antibiotics to hemofilter have been reported, and those findings have not been sufficiently verified in vivo. Given the difficulties to develop an experimental system that simulates human blood and the pathological condition of AKI, the clinical significance of the in vitro and ex vivo data remains unclear. Therefore, it is very important to evaluate the adsorption of antibiotics on hemofilter in vivo. However, to examine the adsorptive properties of hemofilters in vivo, it is necessary to measure the drug concentrations in blood before and after passing through the hemofilter and drug concentration in the filtrate through the hemofilter. 18 Thus, a measurement system that can measure both blood and filtrate concentrations is required. Although there are many methods that measure the concentrations of antibiotics in blood using high‐performance liquid chromatography and ultra‐high‐performance liquid chromatography coupled with tandem mass spectrometry (UHPLC–MS/MS), 19 , 20 , 21 , 22 , 23 no method that measures the antibiotic concentrations in filtrate has been documented.
With this background, we aimed to establish a system for evaluating the antibiotic adsorption properties of various hemofilters by developing a UHPLC–MS/MS assay for CRRT filtrate, which can simultaneously quantify carbapenems [doripenem (DRPM) and meropenem (MEPM)], quinolones [ciprofloxacin (CPFX), levofloxacin (LVFX), and pazufloxacin (PZFX)] and anti‐MRSA agents [linezolid (LZD) and tedizolid (TZD)] that are commonly used in ICU patients.
2. MATERIALS AND METHODS
2.1. Materials
Chemicals are shown with purity in parentheses. Standard substances of DRPM (98%), MEPM (98%), LZD (98%), CPFX (98%), and LVFX (98%) were purchased from Sigma‐Aldrich (St. Louis), and those of TZD (98%) and PZFX (98%) from Toronto Research Chemicals. For the isotopically labeled internal standard (IS), MEPM‐d6 (97%), LZD‐d3 (98%), TZD‐d3 (98%) CPFX‐d8 (98%), and LVFX‐d8 (98%) were purchased from Toronto Research Chemicals; DRPM‐d4 (98%) from Alsachim; and PZFX‐d4 (98%) from Santa Cruz Biotechnology. Acetonitrile, methanol (MeOH), ultrapure water (H2O), 2‐propanol, ammonium formate, formic acid (all chromatographical grade), and dimethyl sulfoxide (analytical grade) were purchased from Wako Pure Chemical Ind. Ltd.
2.2. CRRT modalities and hemofilters
Continuous renal replacement therapy was performed using a roller pump (TR‐55X; Toray) with one of the following dialyzers: ATA filters with surface area of 2.1 m2 (AUT‐21eco; Nipro), PES filters with surface area of 2.1 m2 (PUT‐21eco; Nipro), PMMA filters with surface area of 1.8 m2 (CH‐1.8 W; Toray) and AN69ST filters with surface area of 1.5 m2 (Sepxiris150; Baxter). The initial blood flow rate was set at 100 ml/min, and CRRT dose was set at 600 ml/h (continuous venovenous hemofiltration), which was adjusted according to clinical need. The post‐dilution technique was applied, and fluid to substitute ultrafiltrate production was reinfused after hemofiltration.
2.3. Patients and sample collection
Patients who were treated with intravenous DRPM, MEPM, LZD, TZD, CPFX, LVFX, or PZFX in ICU at Oita University Hospital were recruited. Filtrate samples were collected from a filtrate drainage line into tubes containing ethylenediaminetetraacetic acid. At more than 3 days after starting one or more of the above antibiotics, filtrate sample was collected at 1 h (Cmax) after infusion of the antibiotics. The collected filtrate sample was stored at −40°C until measurement. The study was conducted in accordance with the ethical standards of our institute and with the Helsinki Declaration of 1975, as revised in 2013. The protocol for this study was approved by the Ethics Committee of Oita University Faculty of Medicine (review reference number: 1557) before the study was started. Written informed consent was obtained from the patients participating in this study or their legally authorized representatives.
2.4. Preparation of stock and working solutions
The concentrations and solvents used to prepare stock solutions of the antibiotics were determined by referring to previous reports. 19 , 20 , 21 , 22 , 23 , 24 , 25 Polypropylene tubes were used in this study, because our previous study found no adsorption of any drug to polypropylene tubes. 26 The stock solutions of standards and IS were prepared at the following concentrations using the solvent indicated: DRPM (1 mg/ml in 25% MeOH), DPRM‐d4 (0.1 mg/ml in 25% MeOH), MEPM (1 mg/ml in 25% MeOH), MEPM‐d6 (0.1 mg/ml in 25% MeOH), LZD (2.5 mg/ml in 25% MeOH), LZD‐d3 (0.1 mg/ml in 25% MeOH), TZD (0.2 mg/ml in dimethyl sulfoxide), TZD‐d3 (0.04 mg/ml in MeOH), CPFX (1 mg/ml in 25% MeOH + 20 mM acetic acid), CPFX‐d8 (0.25 mg/ml in 50% MeOH + 20 mM acetic acid), LVFX (1 mg/ml in 20% MeOH), LVFX‐d8 (0.1 mg/ml in 20% MeOH), PZFX [1 mg/ml in MeOH + HCl (99.5:0.5, v/v)], and PZFX‐d4 [0.1 mg/ml in MeOH + HCl (99.5:0.5, v/v)]. These stock solutions were stored at −40°C.
The working standard solutions were prepared by diluting the respective stock solutions with 50% MeOH to the following concentrations: DRPM 100 μg/ml, MEPM 250 μg/ml, LZD 100 μg/ml, TZD 10 μg/ml, CPFX 50 μg/ml, LVFX 100 μg/ml, and PZFX 100 μg/ml. These concentrations corresponded to the upper limit of the calibration ranges. The working IS solutions were prepared by diluting the respective stock solutions with 50% MeOH to the following concentrations: DRPM‐d4 10 μg/ml, MEPM‐d6 5 μg/ml, LZD‐d3 10 μg/ml, TZD‐d3 1 μg/ml, CPFX‐d8 1 μg/ml, LVFX‐d8 2 μg/ml, and PZFX‐d4 4 μg/ml. Calibrators were prepared by diluting the working solutions (upper limit of calibration range) 500‐, 200‐, 100‐, 50‐, 20‐, 5‐, and 2‐fold using 50% MeOH.
Stock solutions and working solutions for quality controls (QCs) had the same concentrations as those for the calibrators, but were prepared separately. Lower limit of quantification (LLOQ), low QC (QC A), middle QC (QC B), and high QC (QC C) samples were prepared by diluting the working solutions for QC 20/10000, 60/10000, 600/10000, and 7500/10000, respectively, using 50% MeOH.
2.5. Sample preparation
Filtrate samples were pretreated by protein precipitation with 50% MeOH and acetonitrile. Blank filtrate was prepared using plasma obtained from healthy volunteers. Plasma sample was centrifuged in an Amicon® Ultra‐15 centrifugal filter device (Merk Millipore Ltd.) at 20,600 × g at 4°C until almost all the plasma was filtered, and the supernatant was collected. For calibrator and QC samples, 50 μl of blank filtrate was transferred to a 2‐ml polypropylene tube, and 50 μl of each calibrator or QC sample in 50% MeOH, 25 μl of IS working solution in 50% MeOH, and 100 μl of acetonitrile were added in that order. For patient filtrate and blank filtrate samples, 50 μl of filtrate sample, 25 μl of IS solution in 50% MeOH, 100 μl of acetonitrile, and 50 μl of 50% methanol were added in that order into a polypropylene tube. The mixtures were vortexed for 1 min, and centrifuged at 20,600 × g at 4°C for 10 min. The supernatants were collected and transferred to a 96‐well collection plate (Waters Corp.). The sample plate was sealed with a sealing cap (Waters Corp.) and kept at 4°C until assay. Twenty microliters of sample was injected into the UHPLC.
2.6. Liquid chromatography condition and instrumental analysis parameter
The chromatographic conditions of the instrumental parameter of UHPLC–MS/MS system were set according to the report by Kai et al., 26 In brief, an Acquity UPLC® I‐Class System (Waters Corp.) with a triple‐stage quadrupole mass spectrometer (Xevo TQ‐D) was used with an Acquity CSH C18 column (1.7 μm, 2.1 × 50 mm) and an Acquity CSH C18 Van‐Guard pre‐column (1.7 μm, 2.1 × 5 mm) (both Waters Corp.). The column temperature was 40°C, and sample temperature was 4°C. The mobile phase contained 0.1% formic acid and 2 mM ammonium formate in ultrapure water‐acetonitrile (98:2, v/v for solution A and 2/98, v/v for solution B). The gradient profile was 100% A 0.5 min, changing linearly to 100% B 0.5–1.5 min, holding 1.5 min, returning to 100% A, and holding 4 min. The flow rate was 0.4 ml/min. The total analysis time was 7 min.
The ionization parameters were as follows: cone voltage 28 V, electrospray voltage 2.5 kV, source temperature 150°C, cone gas flow (N2) 50 L/h, desolvation gas flow (N 2) 800 L/h, and desolvation temperature 600°C. The mass spectrometer was tuned automatically to each drug and IS, using the MassLynx V4.1 system software package (Waters Corp.) by the IntelliStart standard optimization procedures. Multiple reaction monitoring (MRM) analysis was performed using argon as collision gas, and the MS/MS transitions monitored in the positive ion mode are listed in Table 1. The dwell time for each transition was 8 ms.
TABLE 1.
MS transition data of DRPM, MEPM, CPFX, LVFX, PZFX, LZD, and TZD.
| Compound | Parent ion (m/z) | Cone voltage (V) | Product ion (m/z) | Collision energy (V) | Ion mode |
|---|---|---|---|---|---|
| DRPM | 421.15 | 26 | 274.04 | 18 | ES+ |
| DRPM‐d4 | 424.98 | 20 | 346.12 | 14 | ES+ |
| MEPM | 384.16 | 24 | 140.91 | 18 | ES+ |
| MEPM‐d6 | 390.21 | 24 | 147.01 | 16 | ES+ |
| CPFX | 332.14 | 28 | 231 | 36 | ES+ |
| CPFX‐d8 | 340.2 | 38 | 253.03 | 42 | ES+ |
| LVFX | 362.14 | 30 | 261.1 | 26 | ES+ |
| LVFX‐d8 | 370.2 | 36 | 265.19 | 32 | ES+ |
| PZFX | 319.1 | 22 | 284.05 | 18 | ES+ |
| PZFX‐d4 | 322.97 | 20 | 288.07 | 20 | ES+ |
| LZD | 338.01 | 30 | 235.02 | 20 | ES+ |
| LZD‐d3 | 341.2 | 34 | 235.09 | 20 | ES+ |
| TZD | 371 | 28 | 343.06 | 18 | ES+ |
| TZD‐d3 | 374.07 | 30 | 346.12 | 18 | ES+ |
Abbreviations: CPFX, ciprofloxacin; DRPM, doripenem; LVFX, levofloxacin; LZD, linezolid; MEPM, meropenem; PZFX, pazufloxacin; TZD, tedizolid.
2.7. Full validation
The assay was fully validated according to the guidelines for bioanalytical method validation of the US Food and Drug Administration (FDA). 27
2.7.1. Accuracy and precision
Accuracy and precision were calculated and expressed as relative error (RE%) and coefficient of variation (CV%), respectively, for each analytical batch (within‐batch) and for three validation batches (batch‐to‐batch). Each validation batch contained eight calibrators and 6 replicates of QC samples at four levels (LLOQ, QC A, QC B, and QC C). The acceptance criterion for precision was below 15% CV for QCs (but 20% for LLOQ), and that for accuracy was between −15 and 15% for QCs (but between −20% and 20% for LLOQ).
2.7.2. Recovery rate and matrix effect
Recovery rates were evaluated by comparing the peak areas of QC A, QC B, and QC C with those of deproteinized blank filtrate samples prepared from plasma samples of three different healthy volunteers spiked at the respective QC levels.
Matrix effects were evaluated by comparing the peak areas of deproteinized blank filtrate samples of three different healthy volunteers spiked at QC levels of A, B, and C with those of matrix‐free solutions containing 100% standard and IS. Matrix effect was IS‐compensated.
2.7.3. Selectivity
Selectivity was analyzed using blank filtrate samples prepared from plasma samples of six different volunteers, without adding any standard solution or IS solution. The baseline signals at the expected retention time of each drug were evaluated as interfering peaks.
2.7.4. Stability
Freeze–thaw stability was evaluated using QC A and QC B in triplicate measurement. Fifty microliters of blank filtrate was spiked with 5 μl of a QC sample (QC A or QC B) at 10 × concentration. Three cycles of freezing at −40°C for 2 h and thawing at room temperature were performed. Then, 45 μl of 50% MeOH, 25 μl of IS, and 100 μl of acetonitrile were added. Stability was evaluated by calculating the precision and accuracy of the QC samples.
Autosampler stability was evaluated by calculating the precision and accuracy of QC A, QC B, and QC C samples analyzed after storage in an autosampler at 4°C for 24 or 72 h.
2.7.5. Calculation and linearity
The calibration curve was constructed using analyte‐specific MRM quantifier transitions. The analysis response was calculated as a ratio of the peak area of individual analyte to that of the corresponding IS. Using the TargetLynx V4.1 software package (Waters), weighted linear regression (1/x) for each concentration and analytical response was calculated for each analytical batch. Linearity was confirmed by calculating the concentrations of calibrators back‐calculated from the calibration curves, and RE%.
2.8. Calculation of adsorption on hemofilter
The adsorption (Ads) rate on hemofilter was calculated using the following equations, as reported previously 18 :
Ads = (Ci × Qpli) – [(Co × Qplo) – (Co × Qplo) + Cuf × Quf].
Ads (%) = Ads (μg/min)/(Ci × Qbi) × 100.
Qpli = Qbi × (1 – Ht/100).
Qplo = Qpli – Quf = Qbi × (1 – Ht/100) – Quf.
where Ci and Co are the plasma concentrations on the filter inlet side and filter outlet side, respectively; Cuf is the filtrate concentration; Qbi is the blood flow rate on the filter inlet side; Quf is the ultrafiltration flow rate (ml/min) on the filter inlet side; and Qpli and Qplo are the plasma flow rates on the filter inlet side and filter outlet side, respectively. The plasma concentrations of DRPM, MEPM, LZD, TZD, CPFX, LVFX, and PZFX were measured by the UHPLC–MS/MS method described by Kai et al., 26
3. RESULTS AND DISCUSSION
3.1. Mass spectrometric and chromatographic characteristics
During MS tuning, [M + H]+ signals in positive ion mode with the highest intensity were selected. The mass spectra of precursor ions and product ions of the seven drugs are shown in Supplemental Figure S1. The precursor ion was automatically selected by MRM analysis, and the product ion with the highest sensitivity and specificity was selected. Figure 1 shows the chromatograms of blank filtrate and LLOQ, and Supplemental Figure S2 shows the chromatograms of QC C and patient samples. The retention time was 1.86 min for DRPM, 1.93 min for MEPM, 1.98 min for CPFX, 1.97 min for LVFX, 1.98 min for PZFX, 2.18 min for LZD, and 2.23 min for TZD. The total measurement time per sample was 7 min. The peaks in LLOQ were single and sharp. Signal‐to‐noise ratio (S/N) was ≥165, which was higher than that recommended in the FDA guidance (S/N ≥ 10). No interference peak was observed at the retention times of all the drugs in all the blank filtrate samples prepared from six different volunteers, confirming good selectivity.
FIGURE 1.
Chromatograms for measurement of doripenem (DRPM), meropenem (MEPM), linezolid (LZD), tedizolid (TZD), ciprofloxacin (CPFX), levofloxacin (LVFX), and pazufloxacin (PZFX) in lower limit of quantitation (LLOQ) filtrate sample (left) compared with blank filtrate (right).
3.2. Full validation
Supplemental Figure S3 shows the linearity of the seven drugs. The correlation coefficients (R 2) for the drugs calculated by linear regression were as follows: R 2 ≥ 0.9981 for DRPM (calibration range: 0.2–100 μg/ml), R 2 ≥ 0.9934 for MEPM (calibration range: 0.5–250 μg/ml), R 2 ≥ 0.9953 for CPFX (calibration range: 0.1–50 μg/ml), R 2 ≥ 0.9967 for LVFX (calibration range: 0.2–100 μg/ml), R 2 ≥ 0.9971 for PZFX (calibration range: 0.2–100 μg/ml), R 2 ≥ 0.9990 for LZD (calibration range: 0.2–100 μg/ml), and R 2 ≥ 0.9994 for TZD (calibration range: 0.02–10 μg/ml). The relative errors calculated from the calibration curves were <15% for all the drugs, indicating good linearity. Carryover was evaluated by injecting blank samples three times after measuring the highest concentration in the calibration range. All drugs fulfilled the validation criteria of FDA guidance for carryover (20% or less of the analyte and 5% or less of the internal standard at LLOQ) (Data not shown).
Table 2 shows the validation results for precision and accuracy of the seven drugs. For all the drugs, the results of precision and accuracy fulfilled the acceptance criteria of FDA guidance: for LLOQ, precision was ≤11.06% CV (within‐batch) and ≤ 10.65% CV (batch‐to‐batch), and accuracy was ≤|9.01%| (within‐batch) and ≤ |6.67%| (within‐batch); for QCs, precision was ≤8.55% CV (within‐batch) and ≤6.87% CV (within‐batch), and accuracy was ≤|11.03%| (within‐batch) and ≤ |7.33%| (within‐batch).
TABLE 2.
Validation results (accuracy and precision) for measuring filtrate concentrations of (A) DRPM, MEPM, CPFX, and LVFX. (B) PZFX, LZD, and TZD.
| (A) | Nominal DRPM concentrations (μg/ml) | Nominal MEPM concentrations (μg/ml) | |||||||
|---|---|---|---|---|---|---|---|---|---|
| LLOQ | QC A | QC B | QC C | LLOQ | QC A | QC B | QC C | ||
| 0.2 | 0.6 | 6 | 75 | 0.5 | 1.5 | 15 | 187.5 | ||
| Within‐batch | |||||||||
| 1 | Mean (ng/ml) | 0.19 | 0.57 | 6.41 | 72.73 | 0.49 | 1.53 | 16.65 | 177.25 |
| Precision (% CV) | 10.98 | 7.07 | 6.90 | 3.59 | 9.91 | 4.70 | 1.85 | 6.57 | |
| Accuracy (%) | −5.40 | −5.71 | 6.89 | −3.03 | −2.10 | 1.70 | 11.03 | −5.47 | |
| 2 | Mean (ng/ml) | 0.21 | 0.57 | 6.20 | 71.01 | 0.51 | 1.49 | 16.44 | 194.50 |
| Precision (% CV) | 10.23 | 8.55 | 5.27 | 4.90 | 5.65 | 6.06 | 2.25 | 3.13 | |
| Accuracy (%) | 3.60 | −5.54 | 3.37 | −5.32 | 1.05 | −0.71 | 9.61 | 3.73 | |
| 3 | Mean (ng/ml) | 0.19 | 0.58 | 6.21 | 75.94 | 0.50 | 1.42 | 14.87 | 181.55 |
| Precision (% CV) | 11.06 | 5.94 | 4.83 | 3.95 | 4.06 | 3.02 | 3.26 | 6.03 | |
| Accuracy (%) | −4.70 | −2.52 | 3.50 | 1.26 | −0.01 | −5.44 | −0.89 | −3.17 | |
| Batch‐to‐batch | Mean (ng/ml) | 0.20 | 0.57 | 6.28 | 73.34 | 0.50 | 1.47 | 15.87 | 185.33 |
| Precision (% CV) | 10.65 | 6.87 | 5.47 | 4.67 | 6.09 | 5.37 | 5.87 | 6.30 | |
| Accuracy (%) | −2.17 | −4.51 | 4.59 | −2.21 | −0.21 | −1.96 | 5.79 | −1.16 | |
| Nominal CPFX concentrations (μg/ml) | Nominal LVFX concentrations (μg/ml) | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| LLOQ | QC A | QC B | QC C | LLOQ | QC A | QC B | QC C | ||
| 0.1 | 0.3 | 3 | 37.5 | 0.2 | 0.6 | 6 | 75 | ||
| Within‐batch | |||||||||
| 1 | Mean (ng/ml) | 0.10 | 0.28 | 2.90 | 36.25 | 0.20 | 0.55 | 5.57 | 74.48 |
| Precision (% CV) | 4.76 | 2.89 | 3.45 | 3.60 | 6.62 | 3.49 | 1.39 | 4.25 | |
| Accuracy (%) | 1.32 | −7.23 | −3.23 | −3.32 | −0.35 | −7.72 | −7.16 | −0.69 | |
| 2 | Mean (ng/ml) | 0.10 | 0.29 | 3.03 | 37.37 | 0.20 | 0.57 | 5.54 | 72.80 |
| Precision (% CV) | 7.15 | 4.12 | 6.37 | 8.41 | 4.85 | 1.42 | 2.95 | 4.90 | |
| Accuracy (%) | −4.08 | −4.57 | 1.13 | −0.33 | 1.24 | −5.72 | −7.71 | −2.93 | |
| 3 | Mean (ng/ml) | 0.10 | 0.29 | 3.01 | 37.00 | 0.20 | 0.55 | 5.64 | 76.14 |
| Precision (% CV) | 6.64 | 1.88 | 2.09 | 5.28 | 4.63 | 1.96 | 2.36 | 3.16 | |
| Accuracy (%) | −0.25 | −3.34 | 0.33 | −1.34 | −0.45 | −8.81 | −6.05 | 1.52 | |
| Batch‐to‐batch | Mean (ng/ml) | 0.10 | 0.28 | 2.98 | 36.91 | 0.20 | 0.56 | 5.58 | 74.47 |
| Precision (% CV) | 6.37 | 3.40 | 4.67 | 6.01 | 5.17 | 2.71 | 2.31 | 4.32 | |
| Accuracy (%) | −1.14 | −5.05 | −0.64 | −1.57 | 0.15 | −7.33 | −6.97 | −0.70 | |
| (B) | Nominal PZFX concentrations (μg/ml) | Nominal LZD concentrations (μg/ml) | Nominal TZD concentrations (μg/ml) | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| LLOQ | QC A | QC B | QC C | LLOQ | QC A | QC B | QC C | LLOQ | QC A | QC B | QC C | ||
| 0.2 | 0.6 | 6 | 75 | 0.2 | 0.6 | 6 | 75 | 0.02 | 0.06 | 0.6 | 7.5 | ||
| Within‐batch | |||||||||||||
| 1 | Mean (ng/ml) | 0.21 | 0.64 | 6.34 | 79.86 | 0.19 | 0.55 | 6.00 | 70.36 | 0.02 | 0.06 | 0.60 | 7.28 |
| Precision (% CV) | 7.58 | 5.22 | 2.18 | 5.95 | 6.24 | 4.19 | 2.05 | 2.31 | 2.73 | 2.80 | 1.65 | 3.63 | |
| Accuracy (%) | 2.80 | 6.48 | 5.70 | 6.48 | −6.10 | −7.85 | 0.07 | −6.19 | −6.65 | −6.24 | −0.55 | −2.93 | |
| 2 | Mean (ng/ml) | 0.20 | 0.62 | 6.40 | 77.97 | 0.18 | 0.60 | 6.11 | 77.48 | 0.02 | 0.06 | 0.59 | 7.33 |
| Precision (% CV) | 4.15 | 2.27 | 4.52 | 3.70 | 3.52 | 4.01 | 2.83 | 2.36 | 4.10 | 1.58 | 4.15 | 2.71 | |
| Accuracy (%) | 2.17 | 3.03 | 6.74 | 3.96 | −9.01 | 0.12 | 1.87 | 3.30 | −8.00 | −5.83 | −2.40 | −2.30 | |
| 3 | Mean (ng/ml) | 0.21 | 0.62 | 6.40 | 80.42 | 0.19 | 0.55 | 5.99 | 73.73 | 0.02 | 0.06 | 0.59 | 7.28 |
| Precision (% CV) | 4.51 | 4.05 | 2.49 | 1.69 | 4.42 | 1.39 | 2.43 | 2.76 | 5.01 | 3.07 | 1.58 | 2.18 | |
| Accuracy (%) | 6.21 | 3.58 | 6.64 | 7.22 | −5.49 | −8.29 | −0.10 | −1.69 | −4.50 | −8.28 | −1.18 | −2.99 | |
| Batch‐to‐batch | Mean (ng/ml) | 0.21 | 0.63 | 6.38 | 79.42 | 0.19 | 0.57 | 6.04 | 73.85 | 0.02 | 0.06 | 0.59 | 7.29 |
| Precision (% CV) | 5.41 | 4.17 | 3.13 | 4.14 | 4.98 | 5.41 | 2.49 | 4.68 | 4.14 | 2.68 | 2.67 | 2.75 | |
| Accuracy (%) | 3.78 | 4.44 | 6.39 | 5.89 | −6.67 | −5.17 | 0.61 | −1.53 | −6.38 | −6.78 | −1.38 | −2.74 | |
Abbreviations: CPFX, ciprofloxacin; DRPM, doripenem; LVFX, levofloxacin; LZD, linezolid; MEPM, meropenem; PZFX, pazufloxacin; TZD, tedizolid.
Table 3 shows the recovery rate, matrix effect, as well as freeze–thaw and autosampler stability for the seven drugs. Recovery rate was ≥87.20% for DRPM, ≥94.68% for MEPM, ≥104.61% for CPFX, ≥96.18% for LVFX, ≥97.15% for PZFX, ≥95.58% for LZD and ≥96.93% for TZD. Matrix effect was 102.91%–115.60% for DRPM, 108.59%–114.50% for MEPM, 101.63%–105.93% for CPFX, 108.58%–114.46% for LVFX, 99.57–107.08% for PZFX, 100.00%–104.53% for LZD, and 103.83%–109.83% for TZD. Recovery rate and matrix effect were not remarkably different between QCs for all the drugs. Our assay system was designed to simultaneously measure several drugs with different properties. However, it was impossible to extract all the drugs efficiently at the same time using solid‐phase extraction. Therefore, deproteinization was selected as the extraction method. In general, the deproteinization method is selected in simultaneous measurement systems because the extraction step is simple. 19 , 22 , 23 , 25 However, extraction by the deproteinization method may result in interference of ionization such as ion suppression and ion enhancement. 28 In addition, since the cone is more easily contaminated when using deproteinization compared with solid‐phase extraction, frequent maintenance is required. On the contrary, since the CRRT filtrate contains only the components after being filtered by the hemofilter, it is considered that deproteinized filtrate samples may not contain a large quantity of biological substances, unlike plasma that contains a large amount of matrix. In fact, ion suppression and ion enhancement were slight, resulting in good linearity and sharp peaks even at LLOQ. When QCA and QC B were subjected to three freeze–thaw cycles, precision was ≤6.60% CV and accuracy was ≤|8.70%|. After placing the three QC samples in autosampler at 4°C for 24 h, precision was ≤6.14% CV and accuracy was ≤|12.67%|, and those for 72 h were ≤5.60% and ≤|14.56%|, respectively.
TABLE 3.
Validation results (stability, recovery rate, and matrix effect) for measuring filtrate concentrations of (A) DRPM, MEPM, CPFX, and LVFX. (B) PZFX, LZD, and TZD.
| (A) | Nominal DRPM concentrations (μg/ml) | Nominal MEPM concentrations (μg/ml) | ||||
|---|---|---|---|---|---|---|
| QC A | QC B | QC C | QC A | QC B | QC C | |
| 0.6 | 6 | 75 | 1.5 | 15 | 187.5 | |
| Freeze–thaw stability | ||||||
| Mean (ng/ml) | 0.61 | 6.35 | 1.60 | 16.30 | ||
| Precision (% CV) | 1.89 | 4.56 | 5.78 | 0.63 | ||
| Accuracy (%) | 2.35 | 5.86 | 6.37 | 8.70 | ||
| Autosampler stability (24 h) | ||||||
| Mean (ng/ml) | 0.61 | 6.18 | 71.26 | 1.49 | 16.68 | 211.25 |
| Precision (% CV) | 3.85 | 2.10 | 4.23 | 6.14 | 1.13 | 3.05 |
| Accuracy (%) | 1.25 | 2.98 | −4.98 | −0.98 | 11.18 | 12.67 |
| Autosampler stability (72 h) | ||||||
| Mean (ng/ml) | 0.60 | 6.87 | 71.04 | 1.51 | 16.71 | 194.39 |
| Precision (% CV) | 4.27 | 2.39 | 1.56 | 5.65 | 1.11 | 5.02 |
| Accuracy (%) | −0.27 | 14.56 | −5.28 | 0.49 | 11.39 | 3.67 |
| Recovery rate (%) | 87.20 | 97.96 | 105.20 | 94.68 | 99.06 | 97.49 |
| Recovery rate (% CV) | 10.01 | 1.77 | 5.15 | 0.39 | 4.98 | 2.25 |
| Matrix effect (%) | 102.91 | 115.60 | 103.33 | 114.50 | 108.59 | 115.71 |
| Matrix effect (%CV) | 11.40 | 2.06 | 2.20 | 7.74 | 5.52 | 1.17 |
| Nominal CPFX concentrations (μg/ml) | Nominal LVFX concentrations (μg/ml) | |||||
|---|---|---|---|---|---|---|
| QC A | QC B | QC C | QC A | QC B | QC C | |
| 0.3 | 3 | 37.5 | 0.6 | 6 | 75 | |
| Freeze‐thaw stability | ||||||
| Mean (ng/ml) | 0.30 | 3.13 | 0.57 | 6.20 | ||
| Precision (% CV) | 1.13 | 3.54 | 2.09 | 6.60 | ||
| Accuracy (%) | −1.60 | 4.30 | −5.52 | 3.26 | ||
| Autosampler stability (24 h) | ||||||
| Mean (ng/ml) | 0.32 | 3.05 | 38.24 | 0.60 | 6.10 | 74.92 |
| Precision (% CV) | 2.74 | 4.53 | 3.75 | 4.25 | 4.23 | 2.72 |
| Accuracy (%) | 6.50 | 1.80 | 1.98 | −0.30 | 1.71 | −0.11 |
| Autosampler stability (72 h) | ||||||
| Mean (ng/ml) | 0.29 | 3.09 | 38.18 | 0.57 | 5.57 | 71.23 |
| Precision (% CV) | 1.86 | 2.74 | 2.43 | 3.18 | 2.35 | 2.73 |
| Accuracy (%) | −1.77 | 2.96 | 1.80 | −4.45 | −7.11 | −5.02 |
| Recovery rate (%) | 106.51 | 104.61 | 109.17 | 100.96 | 96.18 | 92.72 |
| Recovery rate (% CV) | 1.64 | 2.97 | 6.74 | 2.75 | 2.42 | 5.10 |
| Matrix effect (%) | 105.93 | 101.63 | 102.29 | 114.46 | 108.58 | 112.04 |
| Matrix effect (% CV) | 8.43 | 0.83 | 4.12 | 3.98 | 1.70 | 2.41 |
| (B) | Nominal PZFX concentrations (μg/ml) | Nominal LZD concentrations (μg/ml) | Nominal TZD concentrations (μg/ml) | ||||||
|---|---|---|---|---|---|---|---|---|---|
| QC A | QC B | QC C | QC A | QC B | QC C | QC A | QC B | QC C | |
| 0.6 | 6 | 75 | 0.6 | 6 | 75 | 0.06 | 0.6 | 7.5 | |
| Freeze‐thaw stability | |||||||||
| Mean (ng/ml) | 0.63 | 6.38 | 0.62 | 6.41 | 0.06 | 0.61 | |||
| Precision (%CV) | 1.15 | 2.06 | 1.53 | 0.40 | 2.96 | 1.56 | |||
| Accuracy (%) | 4.44 | 6.41 | 3.77 | 6.78 | −0.83 | 1.45 | |||
| Autosampler stability (24 h) | |||||||||
| Mean (ng/ml) | 0.63 | 5.98 | 73.64 | 0.59 | 6.09 | 78.08 | 0.06 | 0.60 | 7.67 |
| Precision (%CV) | 3.95 | 3.24 | 2.76 | 2.94 | 4.62 | 3.05 | 0.69 | 1.58 | 1.61 |
| Accuracy (%) | 4.34 | −0.30 | −1.81 | −1.89 | 1.52 | 4.10 | −0.98 | 0.19 | 2.30 |
| Autosampler stability (72 h) | |||||||||
| Mean (ng/ml) | 0.63 | 6.26 | 77.38 | 0.59 | 5.81 | 72.83 | 0.06 | 0.60 | 7.59 |
| Precision (%CV) | 5.60 | 4.50 | 1.25 | 5.11 | 4.65 | 4.00 | 4.85 | 0.83 | 2.36 |
| Accuracy (%) | 4.50 | 4.42 | 3.18 | −2.03 | −3.17 | −2.89 | 2.54 | 0.47 | 1.24 |
| Recovery rate (%) | 97.15 | 97.67 | 101.54 | 100.83 | 95.58 | 100.26 | 97.67 | 96.93 | 99.17 |
| Recovery rate (%CV) | 6.41 | 1.83 | 1.84 | 1.43 | 1.01 | 2.83 | 0.69 | 2.74 | 0.62 |
| Matrix effect (%) | 104.64 | 107.08 | 99.75 | 100.00 | 104.53 | 102.81 | 109.83 | 103.83 | 107.04 |
| Matrix effect (%CV) | 1.10 | 0.88 | 1.05 | 13.23 | 1.71 | 0.75 | 1.35 | 1.18 | 0.65 |
Abbreviations: CPFX, ciprofloxacin; DRPM, doripenem; LVFX, levofloxacin; LZD, linezolid; MEPM, meropenem; PZFX, pazufloxacin; TZD, tedizolid.
3.3. Comparison with reported methods
Because this is the first report of simultaneous quantification of multiple antibiotics in filtrate of CRRT circuit, comparison with other assay methods for filtrate is not possible. For quantification of plasma concentrations, we used a similar UHPLC–MS/MS assay that we previously reported for measuring multiple antimicrobials in plasma. 26 For this assay, the calibration ranges were 0.2–100 μg/ml for DRPM, 0.1–50 μg/ml for MEPM, 0.02–10 μg/ml for CPFX, 0.04–20 μg/ml for LVFX, 0.04–20 μg/ml for PZFX, 0.1–50 μg/ml for LZD, and 0.01–5 μg/ml for TZD. 26 Meanwhile, the ranges of calibration curves for measuring plasma concentrations using other previously reported methods were 0.05–100 μg/ml for DRPM, 20 0.1–100 μg/ml for MEPM, 23 0.025–30 μg/ml for CPFX, 19 0.09–4.53 μg/ml for LVFX, 25 0.02–0.5 μg/ml for PZFX, 22 1–100 μg/ml for LZD, 24 and 0.00074–1.5 μg/ml for TZD. 21 Comparing our results with other reports, the concentration ranges of our calibration curves are relatively broad. Our pretreatment method is more straightforward and can extract multiple drugs in the same batch, whereas more complicated methods such as solid‐phase extraction and ethanol precipitation extraction were used in other studies. 19 , 20 , 21 , 22 , 23 , 24 , 25 The run time of our method is also shorter (7 min 26 ) than previous reports (8–22 min). 19 , 20 , 21 , 22 , 23 , 24 , 25 Our UHPLC–MS/MS assay yields sharper peaks 26 compared with some methods that used high‐performance liquid chromatography (HPLC) with ultraviolet (UV) detection. 20 , 22 , 23
3.4. Clinical application
To evaluate the clinical applicability of the developed method, drug concentrations in CRRT filtrates sampled 1 h after intravenous administration of antibiotics at more than 3 days from treatment initiation were measured. Table 4 shows the patient profile, filtrate concentrations, plasma concentrations of antibiotics, and adsorption rates of antibiotics on hemofilters. The concentrations in the filtrate were within the calibration curve ranges for all the drugs: 12.32–38.28 μg/ml (calibration range: 0.2–100 μg/ml) for DRPM, 19.38–53.99 μg/ml (calibration range: 0.5–250 μg/ml) for MEPM, 6.81–13.54 μg/ml (calibration range: 0.1–50 μg/ml) for CPFX, 7.50–10.52 μg/ml (calibration range: 0.2–100 μg/ml) for LVFX, 29.90–57.64 μg/ml (calibration range: 0.2–100 μg/ml) for PZFX, 12.29–25.83 μg/ml (calibration range: 0.2–100 μg/ml) for LZD, and 0.35–0.87 μg/ml (calibration range: 0.02–10 μg/ml) for TZD. Although it was difficult to identify the drugs adsorbed on the membranes because of the small samples (Table 4), the adsorption rates differed among the membranes, suggesting the need to adjust the dosage depending on the type of membrane. Negative adsorption rates were observed in some membranes. Previous reports have shown that levofloxacin is released after adsorption on PAN membranes, 29 suggesting that a similar phenomenon may have occurred.
TABLE 4.
Patient characteristics, filtrate concentrations, plasma concentrations, and hemofilter adsorption rates of DRPM, MEPM, CPFX, LVFX, PZFX, LZD, and TZD.
| Substance | Gender | Age (years) | Weight (kg) | Height (cm) | Dosage (mg) | Dosing period (days) | Membrane | Filtrate concentration (μg/ml) | Plasma concentration on filter inlet side (μg/ml) | Plasma concentration on filter outlet side (μg/ml) | Adsorption (%) |
|---|---|---|---|---|---|---|---|---|---|---|---|
| DRPM | Female | 52 | 43.2 | 151.7 | 500 (every 8 h) | 3 | AN69ST | 38.28 | 25.68 | 30.12 | −8.46 |
| Male | 51 | 89.0 | 182.0 | 500 (every 8 h) | 3 | ATA | 12.32 | 19.59 | 26.74 | −31.60 | |
| Male | 71 | 60.1 | 171.0 | 1000 (every 8 h) | 3 | AN69ST | 27.91 | 18.08 | 27.37 | 11.85 | |
| MEPM | Female | 51 | 51.7 | 160.6 | 500 (every 6 h) | 3 | AN69ST | 53.99 | 59.22 | 55.86 | 11.92 |
| Female | 74 | 61.0 | 152.1 | 500 (every 6 h) | 3 | ATA | 34.81 | 42.57 | 36.82 | 19.25 | |
| Female | 56 | 60.8 | 144.0 | 500 (every 6 h) | 9 | PES | 19.38 | 28.27 | 30.29 | −68.75 | |
| CPFX | Female | 74 | 61.0 | 152.1 | 800 (once‐daily) | 3 | PMMA | 13.54 | 18.19 | 18.48 | −7.72 |
| Male | 71 | 87.4 | 168.0 | 800 (once‐daily) | 3 | ATA | 6.94 | 9.08 | 8.12 | 6.17 | |
| Male | 81 | 51.0 | 144.5 | 800 (once‐daily) | 3 | AN69ST | 6.81 | 10.88 | 13.03 | 11.21 | |
| LVFX | Male | 51 | 89.0 | 182.0 | 500 (once‐daily) | 3 | ATA | 7.50 | 8.82 | 8.70 | 8.25 |
| Male | 71 | 60.1 | 171.0 | 500 (once‐daily) | 3 | ATA | 10.32 | 5.04 | 6.59 | −35.65 | |
| Female | 65 | 51.7 | 160.6 | 500 (once‐daily) | 3 | PES | 10.52 | 9.69 | 10.18 | 0.30 | |
| PZFX | Male | 90 | 38.0 | 149.6 | 1000 (once‐daily) | 3 | PMMA | 44.88 | 48.73 | 49.65 | 3.10 |
| Male | 78 | 61.8 | 159.8 | 1000 (once‐daily) | 3 | ATA | 29.90 | 35.54 | 31.67 | 9.35 | |
| Female | 65 | 51.7 | 160.6 | 1000 (once‐daily) | 3 | AN69ST | 57.64 | 64.25 | 73.20 | −3.64 | |
| LZD | Male | 71 | 60.1 | 171.0 | 600 (every 12 h) | 3 | AN69ST | 14.06 | 7.49 | 10.61 | −42.89 |
| Female | 74 | 61.0 | 152.1 | 600 (every 12 h) | 3 | PMMA | 25.83 | 29.04 | 30.16 | 2.29 | |
| Male | 72 | 87.4 | 168.0 | 600 (every 12 h) | 3 | ATA | 12.29 | 9.08 | 11.76 | −24.70 | |
| TZD | Female | 65 | 51.7 | 160.6 | 200 (once‐daily) | 3 | PMMA | 0.87 | 5.52 | 5.88 | 6.38 |
| Male | 71 | 60.1 | 171.0 | 200 (once‐daily) | 3 | ATA | 0.35 | 1.96 | 1.30 | 46.17 | |
| Male | 72 | 60.4 | 164.0 | 200 (once‐daily) | 3 | AN69ST | 0.39 | 2.77 | 2.77 | 14.15 |
Abbreviations: ATA, asymmetric cellulose triacetate; CPFX, ciprofloxacin; DRPM, doripenem; LVFX, levofloxacin; LZD, linezolid; MEPM, meropenem; PES, polyethersulfone; PMMA, polymethyl methacrylate; PZFX, pazufloxacin; TZD, tedizolid.
As mentioned in Introduction, several antimicrobial agents have been known to adsorb on hemofilter, but no method that allows accurate determination of the amount of adsorption was available. 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 We report for the first time a UHPLC–MS/MS assay for simultaneous measurement of CRRT filtrate levels of seven antibiotics frequently used in ICU. In a previous study, we have established and validated a simultaneous quantification method for plasma concentrations of 12 antimicrobial agents, 26 which included the seven drugs investigated in the present study. Since the method for measuring CRRT filtrate concentrations and that for assaying plasma concentrations use the same extraction method and MS conditions, it would be possible to simultaneously measure drug concentrations in the filtrate and in the plasma easily. The measured blood and filtrate drug concentrations can then be used to calculate the amount of drug adsorption on hemofilters, which will provide important information on whether dose adjustment is necessary.
4. CONCLUDING REMARKS
We report for the first time a UHPLC–MS/MS method for simultaneous quantification of CCRT filtrate concentrations of carbapenems (DRPM and MEPM), quinolones (CPFX, LVFX, and PZFX), and anti‐MRSA agents (LZD and TZD) that are frequently used in ICU patients. This assay system is simple, rapid, and fulfills the criteria of FDA guidance. Antibiotic concentration in CRRT filtrate together with blood concentrations can be used to calculate the amount of drug adsorption on hemofilters, providing important information on whether dose adjustment is required.
CONFLICT OF INTEREST
The authors declare no conflict of interest.
Supporting information
Appendix S1
ACKNOWLEDGMENTS
This work was supported in part by a Grant‐in‐Aid for Encouragement of Scientists from the JSPS (21H04202) and a Grant‐in‐Aid for Early‐Career Scientists from the JSPS (20K16048).
Kai M, Tanaka R, Suzuki Y, et al. UHPLC–MS/MS method for simultaneous quantification of doripenem, meropenem, ciprofloxacin, levofloxacin, pazufloxacin, linezolid, and tedizolid in filtrate during continuous renal replacement therapy. J Clin Lab Anal. 2023;37:e24815. doi: 10.1002/jcla.24815
DATA AVAILABILITY STATEMENT
Raw data were generated at Oita University Hospital. Derived data supporting the findings of this study are available from the corresponding author Kai M on request.
REFERENCES
- 1. Seymour CW, Gesten F, Prescott HC, et al. Time to treatment and mortality during mandated emergency Care for Sepsis. N Engl J Med. 2017;376(23):2235‐2244. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Bellomo R, Kellum JA, Ronco C, et al. Acute kidney injury in sepsis. Intensive Care Med. 2017;43(6):816‐828. [DOI] [PubMed] [Google Scholar]
- 3. Vincent JL, Sakr Y, Sprung CL, et al. Sepsis in European intensive care units: results of the SOAP study. Crit Care Med. 2006;34(2):344‐353. [DOI] [PubMed] [Google Scholar]
- 4. Pistolesi V, Morabito S, Di Mario F, Regolisti G, Cantarelli C, Fiaccadori E. A guide to understanding antimicrobial drug dosing in critically ill patients on renal replacement therapy. Antimicrob Agents Chemother. 2019;63(8):e00583‐19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Honore PM, Jacobs R, Joannes‐Boyau O, et al. Newly designed CRRT membranes for sepsis and SIRS–a pragmatic approach for bedside intensivists summarizing the more recent advances: a systematic structured review. ASAIO J. 2013;59(2):99‐106. [DOI] [PubMed] [Google Scholar]
- 6. Rimmelé T, Kellum JA. High‐volume hemofiltration in the intensive care unit: a blood purification therapy. Anesthesiology. 2012;116(6):1377‐1387. [DOI] [PubMed] [Google Scholar]
- 7. Hiraiwa T, Moriyama K, Matsumoto K, et al. In vitro evaluation of linezolid and Doripenem clearance with different Hemofilters. Blood Purif. 2020;49(3):295‐301. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Reiter K, Bordoni V, Dall'Olio G, et al. In vitro removal of therapeutic drugs with a novel adsorbent system. Blood Purif. 2002;20(4):380‐388. [DOI] [PubMed] [Google Scholar]
- 9. Shiraishi Y, Okajima M, Sai Y, Miyamoto K, Inaba H. Elimination of teicoplanin by adsorption to the filter membrane during haemodiafiltration: screening experiments for linezolid, teicoplanin and vancomycin followed by in vitro haemodiafiltration models for teicoplanin. Anaesth Intensive Care. 2012;40(3):442‐449. [DOI] [PubMed] [Google Scholar]
- 10. Sime FB, Pandey S, Karamujic N, et al. Ex vivo characterization of effects of renal replacement therapy modalities and settings on pharmacokinetics of Meropenem and Vaborbactam. Antimicrob Agents Chemother. 2018;62(10):e01306‐18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Tian Q, Gomersall CD, Ip M, Tan PE, Joynt GM, Choi GY. Adsorption of amikacin, a significant mechanism of elimination by hemofiltration. Antimicrob Agents Chemother. 2008;52(3):1009‐1013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Onichimowski D, Nosek K, Ziółkowski H, Jaroszewski J, Pawlos A, Czuczwar M. Adsorption of vancomycin, gentamycin, ciprofloxacin and tygecycline on the filters in continuous renal replacement therapy circuits: in full blood in vitro study. J Artif Organs. 2021;24(1):65‐73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Onichimowski D, Ziółkowski H, Nosek K, Jaroszewski J, Rypulak E, Czuczwar M. Comparison of adsorption of selected antibiotics on the filters in continuous renal replacement therapy circuits: in vitro studies. J Artif Organs. 2020;23(2):163‐170. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. PKN L, Tian Q, Ip M, Gomersall CD. In vitro adsorption of gentamicin and netilmicin by polyacrylonitrile and polyamide hemofiltration filters. Antimicrob Agents Chemother. 2010;54(2):963‐965. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Tian Q, Gomersall CD, Wong A, et al. Effect of drug concentration on adsorption of levofloxacin by polyacrylonitrile haemofilters. Int J Antimicrob Agents. 2006;28(2):147‐150. [DOI] [PubMed] [Google Scholar]
- 16. Nakamura A, Takasu O, Sakai Y, et al. Development of a teicoplanin loading regimen that rapidly achieves target serum concentrations in critically ill patients with severe infections. J Infect Chemother. 2015;21(6):449‐455. [DOI] [PubMed] [Google Scholar]
- 17. Hoff BM, Maker JH, Dager WE, Heintz BH. Antibiotic dosing for critically ill adult patients receiving intermittent hemodialysis, prolonged intermittent renal replacement therapy, and continuous renal replacement therapy: an update. Ann Pharmacother. 2020;54(1):43‐55. [DOI] [PubMed] [Google Scholar]
- 18. Ohchi Y, Goto K, Yasuda N, Koga H, Kitano T. High efficiency removal of cytokines and HMGB‐1 by continuous hemofiltration with a dual layered Polyethersulfone membrane: an ex vivo study. Ther Apher Dial. 2019;23(2):173‐179. [DOI] [PubMed] [Google Scholar]
- 19. Grondin C, Zhao W, Fakhoury M, Jacqz‐Aigrain E. Determination of ciprofloxacin in plasma by micro‐liquid chromatography‐mass spectrometry: an adapted method for neonates. Biomed Chromatogr. 2011;25(7):827‐832. [DOI] [PubMed] [Google Scholar]
- 20. Ikeda K, Ikawa K, Morikawa N, et al. Quantification of doripenem in human plasma and peritoneal fluid by high‐performance liquid chromatography with ultraviolet detection. J Chromatogr B Analyt Technol Biomed Life Sci. 2008;867(1):20‐25. [DOI] [PubMed] [Google Scholar]
- 21. Iqbal M. A highly sensitive and efficient UPLC‐MS/MS assay for rapid analysis of tedizolid (a novel oxazolidinone antibiotic) in plasma sample. Biomed Chromatogr. 2016;30(11):1750‐1756. [DOI] [PubMed] [Google Scholar]
- 22. Sousa J, Alves G, Campos G, Fortuna A, Falcao A. First liquid chromatography method for the simultaneous determination of levofloxacin, pazufloxacin, gatifloxacin, moxifloxacin and trovafloxacin in human plasma. J Chromatogr B Analyt Technol Biomed Life Sci. 2013;930:104‐111. [DOI] [PubMed] [Google Scholar]
- 23. Zou L, Meng F, Hu L, Huang Q, Liu M, Yin T. A novel reversed‐phase high‐performance liquid chromatographic assay for the simultaneous determination of imipenem and meropenem in human plasma and its application in TDM. J Pharm Biomed Anal. 2019;169:142‐150. [DOI] [PubMed] [Google Scholar]
- 24. la Marca G, Villanelli F, Malvagia S, et al. Rapid and sensitive LC‐MS/MS method for the analysis of antibiotic linezolid on dried blood spot. J Pharm Biomed Anal. 2012;67–68:86‐91. [DOI] [PubMed] [Google Scholar]
- 25. Fang PF, Cai HL, Li HD, et al. Simultaneous determination of isoniazid, rifampicin, levofloxacin in mouse tissues and plasma by high performance liquid chromatography‐tandem mass spectrometry. J Chromatogr B Analyt Technol Biomed Life Sci. 2010;878(24):2286‐2291. [DOI] [PubMed] [Google Scholar]
- 26. Kai M, Tanaka R, Suzuki Y, et al. Simultaneous quantification of plasma levels of 12 antimicrobial agents including carbapenem, anti‐methicillin‐resistant Staphylococcus aureus agent, quinolone and azole used in intensive care unit using UHPLC‐MS/MS method. Clin Biochem. 2021;90:40‐49. [DOI] [PubMed] [Google Scholar]
- 27. US Department of Health and Human Services, Food and Drug Administration, Guidance for Industry, Bioanalytical Method Validation. 2018; Accessed September 1st, 2021.https://www.fda.gov/files/drugs/published/Bioanalytical‐Method‐Validation‐Guidance‐for‐Industry.pdf
- 28. Taylor PJ. Matrix effects: the Achilles heel of quantitative high‐performance liquid chromatography‐electrospray‐tandem mass spectrometry. Clin Biochem. 2005;38(4):328‐334. [DOI] [PubMed] [Google Scholar]
- 29. Utrup TR, Mueller EW, Healy DP, Callcut RA, Peterson JD, Hurford WE. High‐dose ciprofloxacin for serious gram‐negative infection in an obese, critically ill patient receiving continuous venovenous hemodiafiltration. Ann Pharmacother. 2010;44(10):1660‐1664. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Appendix S1
Data Availability Statement
Raw data were generated at Oita University Hospital. Derived data supporting the findings of this study are available from the corresponding author Kai M on request.