A validated LC-ESI-MS/MS assay for simultaneous plasma quantification of 11 antimicrobials to support personalized dosing in critically ill children

Introduction: Critically ill patients with severe infections demonstrate profound alterations in pharmacokinetic behavior. Therapeutic drug monitoring (TDM)-informed antimicrobial dose optimization is thus essential in intensive care settings to ensure maximal bactericidal activity while mitigating toxicity, creating an urgent need for accessible analytical methodologies. Although various studies exist in this domain, the inherent technical constraints of current methodologies persistently prevent the full resolution of these unmet clinical requirements.

Methods: A rapid and sensitive LC-ESI-MS/MS method was developed and validated for the simultaneous quantification of 11 antimicrobials in human plasma. Sample preparation was performed by a simple one-step protein precipitation using methanol containing 0.1% formic acid. The analytes were separated on a Kinetex C18 column with a “corner-folded cleaver-shaped” gradient elution program and detected using multiple reaction monitoring (MRM) in positive ionization mode.

Results and Discussion: The gradient elution program achieved baseline separation of all tested analytes within 8 min, with symmetrical peak shapes and no endogenous interference. The developed method was proven to be free of matrix effects, excellent linearity (R² > 0.99 for all analytes), and met international bioanalytical validation criteria across clinically relevant concentration ranges. Notably, this validated method was successfully applied to children receiving mono- or combination therapy for infections. The assay meets requirements for clinical TDM implementation, providing pediatricians with reliable antibiotic concentration measurements within a clinically relevant timeframe.

1 Introduction

Microbial infections are a significant cause of health loss globally and are becoming the second leading cause of death worldwide (Collaborators, 2022). In reality, managing infections in critically ill patients presents a distinct and complex challenge. This complexity arises because the prevailing treatment frameworks are largely derived from infection models and clinical data that fail to adequately address the unique pharmacokinetic (PK) profiles and extreme illness severity characteristic of these individuals. Critically ill patients often exhibit profound alterations in key PK parameters—such as volume of distribution, clearance, protein binding, and tissue penetration—differing markedly from other patient populations. This mismatch frequently culminates in inadequate antimicrobial levels at both the bloodstream and the actual site of infection (Rawson et al., 2021; Tangden et al., 2017). Moreover, the rise of antibiotic resistance in various bacteria further complicates treatment and increases mortality rates, highlighting the importance of optimizing the use of antimicrobial agents to maximize therapeutic success and prolong the clinical lifespan of currently available drugs by limiting the emergence of resistance (Roberts et al., 2014). Indeed, suboptimal antimicrobial dosing frequently results in either insufficient drug exposure (leading to therapeutic failure and/or antimicrobial resistance) or excessive exposure that elevates toxicity risk (Sumi et al., 2019).

In this context, therapeutic drug monitoring (TDM), as an individualized treatment strategy, is emerging as a key tool for optimizing antimicrobial therapy. By incorporating PK monitoring, this strategy effectively enhances target concentration achievement and pharmacodynamic (PD) optimization. Notably, the IATDMCT Infectious Diseases Group formally recommended antibiotic TDM in a prior position paper, covering glycopeptides, aminoglycosides, beta-lactams (e.g., penicillins, cephalosporins, monobactams, and carbapenems), and oxazolidinones (e.g., linezolid) (Abdul-Aziz et al., 2020).

In particular, the clinical imperative for implementing antimicrobial TDM in pediatric populations stems from the intersection of developmental pharmacology and clinical complexity (Kearns et al., 2003). Children exhibit distinct PK/PD profiles due to ongoing maturation of drug-metabolizing enzymes, dynamic fluctuations in glomerular filtration rates, and age-dependent body composition changes. These physiological singularities are compounded by antimicrobial agents with nonlinear PK characteristics (e.g., voriconazole), high interindividual variability (e.g., posaconazole), and significant drug-drug interaction potential. In addition, critically ill pediatric patients with hepatic or renal impairment face further risks, as impaired drug metabolism and excretion pathways may lead to dangerous accumulation or subtherapeutic exposure. Moreover, concentration-dependent toxicities—such as imipenem-associated nephrotoxicity—necessitate rigorous dosage individualization to sustain therapeutic efficacy while minimizing adverse effects (Bruns and Dohna-Schwake, 2022; Rashed et al., 2019; Martson et al., 2020).

Suitable and accessible analytical methods are a prerequisite for implementing TDM. Indeed, literature has reported various methods to support routine TDM for antimicrobial agents. Early, Immunoassays (RIA, FIA, and ELISA) were used to measure the concentration of antimicrobial agents (Crossley et al., 1980; Hu et al., 1990; Schwenzer et al., 1983). It should be noted that the immunoassay is occasionally affected by endogenous cross-reactive substances, resulting in a false increase in measurement, and the simultaneous determination of multiple drugs is not possible (Singer et al., 2020). Afterwards, high-performance liquid chromatography (HPLC) coupled to ultraviolet detection or an evaporative light-scattering detector have been reported (Legrand et al., 2016; Verdier et al., 2011). Notably, liquid chromatography-tandem mass spectrometry (LC-MS/MS) has served as an efficient tool with high specificity and accuracy in clinical laboratories over the last 10–15 years (Shipkova and Svinarov, 2016; Li et al., 2017). Recently, various methods based on LC-MS/MS have been developed for the simultaneous determination of antimicrobial agents (Neef et al., 2024; Radovanovic et al., 2022). Still, almost all of them have a large plasma sample consumption (Lefeuvre et al., 2017; Sun et al., 2022) and complex sample clean-up methods, such as liquid-liquid extraction (Ferrari et al., 2019) and solid phase extraction (Colin et al., 2013; Ohmori et al., 2011) (Table 1), which limit the daily clinical application of the developed methods to a certain extent. The shortcomings of existing analytical methods underscore the urgent need to develop novel approaches that feature rapid sample processing, minimal sample consumption, and enhanced analytical efficiency.

Table 1

Table 1. Comparison of this study with various previously published analytical methods for antimicrobial agents using the LC-MS/MS technique.

Therefore, this study aimed to develop and validate a LC-electrospray (ESI)-MS/MS method for simultaneous analysis of 11 antimicrobial agents (ATBs; i.e., caspofungin, imipenem, linezolid, meropenem, latamoxef, piperacillin, voriconazole, posaconazole, cefmetazole, tigecycline, and cefazolin) in human plasma and to investigate its applicability for routine TDM in clinical practice in children with microbial infections. Of note, this study successfully determined the 11 ATBs simultaneously with minimum run time (6.5 min) and small-volume plasma consumption (10 μL). Impressively, our method minimized medical costs by optimizing extraction processing and enabled the analysis with a simple one-step protein precipitation (0.1% FA in MeOH) for sample pretreatment. The fast and efficient method provides valuable concentration data for patients and pediatricians.

2 Experimental methods

2.1 Materials

Latamoxef (MOX) and voriconazole (VCZ) were purchased from the National Institutes for Food and Drug Control (Beijing, China). Caspofungin (CAS), cefmetazole (CMZ), linezolid (LZD), meropenem (MEM), imipenem (IPM), tigecycline (TIG), linezolid-d3 (LZD-d3), piperacillin-d5 (PRL-d5), voriconazole-d3 (VCZ-d3), meropenem-d6 (MEM-d6), and caspofungin acetate-d4 (CAS-d4) were obtained from Toronto Research Chemicals Inc. (Toronto, Canada). Cefazolin (CFZ) was purchased from Anpel-Trace Standard Technical Services (Shanghai) Co., Ltd. Piperacillin (PRL) was supplied by TMstandard® (Beijing, China). Posaconazole (POS) was purchased from LGC Standards (Molsheim, France). HPLC-grade methanol (MeOH) and acetonitrile (ACN) were purchased from Merck KGaA (Darmstadt, Germany). Isopropanol (IPA) of HPLC grade was supplied by Fisher Scientific (Fairlawn, USA). LC-MS grade formic acid (FA) was obtained from TCI (Shanghai, China) Development Co., Ltd. ACS-grade ammonium acetate (NH₄AC) was obtained from Sigma-Aldrich Co. LLC (Wilmington, United States). Ultrapure water was prepared using a Milli-Q water purification system (Millipore, Bedford, USA).

Chromatographic columns including Kinetex C8 100 Å (2.1 mm × 50 mm, 5 μm), Kinetex C18 100 Å (2.1 mm × 50 mm, 1.7 μm), Kinetex C18 100 Å (2.1 mm × 50 mm, 2.6 μm), Kinetex C18 100 Å (2.1 mm × 50 mm, 5 μm), Kinetex C18 100 Å (2.1 mm × 100 mm, 2.6 μm), Luna C18 (2) 100 Å (2 mm × 50 mm, 5 μm), Gemini C18 110 Å (2 mm × 50 mm, 3 μm), and Synergi™ Fusion-RP 80 Å (2 mm × 50 mm, 4 μm) were purchased from the Phenomenex Inc. (Torrance, CA, USA). The Waters Symmetry C18 column (4.6 mm × 75 mm, 3.5 μm) was obtained from the Waters Corporation (Milford, MA, USA). A column designed for hydrophilic interaction chromatography (HILIC) (4.6 mm × 50 mm, 3 μm) was purchased from the OSAKA SODA CO., LTD. (Osaka, Japan).

2.2 Instruments and conditions

Chromatographic separation and analyte quantification were performed using a Jasper™ LC system (AB Sciex Pte. Ltd., Singapore) coupled to a Triple Quad™ 4500MD mass spectrometer (AB Sciex Pte. Ltd., Singapore). The LC system comprised a binary pump (Sciex Dx™), an online degasser (Sciex Dx™), an autosampler (Sciex Dx™), and a thermostated column compartment (Sciex Dx™).

Chromatographic separation was performed on a Kinetex C18 (2.1 mm × 75 mm, 2.6 μm; Phenomenex) protected by a security Guard-C18 column (4 × 2.0 mm; Phenomenex). A gradient elution program was employed using: (A) 0.1% formic acid in H₂O and (B) 0.1% FA in ACN at a mixed flow rate of 0.25 mL/min and 0.5 mL/min, respectively. The gradient profile was:

0–0.5 min: 8% B,

0.5–1 min: 8→50% B,

1–3.5 min: 50→85% B,

3.5–4.2 min: 85→90% B,

4.2–4.7 min: 90% B,

4.7–4.8 min: 90→8% B,

4.8–6.5 min: 8% B.

The column was maintained at 45 °C and autosampler at 4 °C, with an injection volume was 5 μL.

Quantification was conducted on a Triple Quad™ 4500MD mass spectrometer (AB Sciex Pte. Ltd., Singapore) equipped with an ESI source operated in positive/negative ion switching mode. Optimized multiple reaction monitoring (MRM) parameters for each analyte are provided in Table 2. Key source parameters including curtain gas (CUR), collision-activated dissociation (CAD), temperature (TEM), nebulizer gas (GS1), and heater gas (GS2) were set as 20 psi, 6 units, 450 °C, 50 psi, and 50 psi, respectively. The ion spray voltage was maintained at 5500 V in positive ion mode. Data acquisition and peak processing were performed using Analyst MD software (version 1.6.3; AB Sciex Pte. Ltd., Singapore).

Table 2

Table 2. Multiple reaction monitoring parameters, linear, IS, and chromatographic retention times for the 11 ATBs tested in this study.

2.3 Preparation of stock solutions, calibration standards, and quality control (QC) samples

Stock solutions of CAS, CFZ, CMZ, MOX, PRL, VCZ, and MEM were prepared in MeOH at 10.00 mg/mL, except for POS and LZD prepared in 5.00 mg/mL. Stock solutions of IPM were dissolved in H₂O at 5.00 mg/mL, and 8.00 mg/mL for TIG in MeOH, respectively. The working solutions of antibiotics were further prepared by diluting the stock solution in 50% MeOH to obtain the target concentrations. All the stock solutions and working solutions were stored at −80 °C. Internal standard (IS) stock solutions of PRL-d5, CAS-d4, MEM-d6, LZD-d3, and VCZ-d3 were prepared in MeOH to give a concentration of 1.00 mg/mL. A mixed IS working solution was prepared in MeOH containing CAS-d4 and MEM-d6 at 1,000 ng/mL, LZD-d3 and VCZ-d3 at 50.0 ng/mL, and PRL-d5 at 100 ng/mL, respectively.

Calibration standards were prepared in duplicate by spiking 10.0 μL of the standard working solutions into 10.0 μL blank plasma. Final concentrations covered the following ranges:

• IPM, MEM, MOX, PRL, CMZ, CFZ: 1.00, 2.00, 4.00, 10.0, 20.0, 60.0, 100.0 μg/mL

• CAS, POS, TIG: 0.500, 1.00, 2.00, 5.00, 10.0, 30.0, 50.0 μg/mL

• LZD: 0.25, 0.50, 1.00, 2.50, 5.00, 15.0, 25.0 μg/mL

• VCZ: 0.100, 0.200, 0.400, 1.00, 2.00, 6.00, 10.0 μg/mL

Calibration curves were generated by plotting the analyte-to-internal standard peak area ratio versus the nominal standard concentrations.

QC samples were prepared in the same way as calibration standards at 4 concentration levels:

• •1.00 μg/mL (the lower limit of quantification QC, LLOQ QC), 3.00 μg/mL (low QC, LQC), 16.0 μg/mL (medium QC, MQC), 80.0 μg/mL (high QC, HQC) for IPM, MEM, MOX, PRL, CMZ, and CFZ;

• 0.500, 1.50, 8.00, and 40.0 μg/mL for CAS, POS, and TIG;

• 0.250, 0.750, 4.00, and 20.0 μg/mL for LZD;

• 0.100, 0.300, 1.60, and 8.00 μg/mL for VCZ.

2.4 Sample preparation

Calibration standards, QC samples, and plasma samples were processed via protein precipitation: for calibrators/QCs, a mix of 10.0 μL working solutions and 10.0 μL blank plasma was added to 80.0 μL of a MeOH/FA mixture (99.9/0.1; v/v), while plasma samples used a mix of 10.0 μL plasma sample and 10.0 μL MeOH added to 80.0 μL identical precipitant. Then, all mixtures were vortexed for 5 min and centrifuged for 5 min at 4,000 rpm at 4 °C. An aliquot of 20.0 μL of the supernatant was transferred to another clean 1.5 mL Eppendorf tube and diluted with 180 μL of H₂O:ACN (95:5; v/v). Next, the mixtures were vortexed well for 3 min and a 5 μL aliquot of the diluted supernatant was injected for LC-ESI-MS/MS analysis. Drug-free human plasma was obtained from the Department of Blood Transfusion, Children’s Hospital of Nanjing Medical University.

2.5 Method validation

Method validation was conducted in accordance with the US Food and Drug Administration Bioanalytical Method Validation Guidance for Industry (2018) (U.S. Food and Drug Administration, 2018).

2.5.1 Selectivity and carryover

The double blank plasmas from six sources were tested to investigate the selectivity of the method. Furthermore, the interference between ATBs and IS was assessed using the upper limit of quantification (ULOQ) without IS and control blank samples (only IS working solution was spiked into blank plasma). The interference between each analyte was also evaluated by analyzing blank plasma samples spiked with single ATB at ULOQ.

In addition, carryover was assessed by injecting blank plasma samples immediately after the ULOQ calibration of standard samples.

2.5.2 Linearity and LLOQ

Linearity was constructed by the peak area ratio of ATBs to IS versus the concentrations of the calibration standards, and the calibration curves in the form of y = a + bx were processed by weighted (1/x²) least-squares linear regression. The LLOQs are considered the lowest concentrations on the calibration curves and were based on a signal-to-noise ratio (S/N) of 5, at least. The intra- and inter-day accuracy and precision of LLOQs should be less than 20% each.

2.5.3 Accuracy and precision

Method accuracy and precision were validated through six replicates at four QC levels (LLOQ, LQC, MQC, and HQC). Intra-day precision was evaluated in a single analytical batch, while inter-day precision was evaluated in three consecutive batches. Dilution integrity was additionally demonstrated using diluted QC (DQC). The accuracy of the method was expressed as relative error (RE%) and the precision was calculated as the relative standard deviation (RSD%).

2.5.4 Recovery and matrix effect

Recovery of the analyte at the three QC levels by comparing the ratio of peak areas from the regularly prepared samples with that from the spike-after-extraction samples with six replicates per level. To prepare the spike-after-extraction samples, a series of intermediate concentration (named “INT” concentration) points was formulated. Briefly, the blank sample was processed according to the sample preparation procedure described above. The supernatant was mixed with an “INT” concentration of ATBs at concentrations corresponding to the final concentration of the pretreated plasma samples. The extraction recovery of the internal standard was determined similarly. The matrix effect was evaluated by calculating the ratio of peak areas of analyte spike-after-extraction (six different sources of blank sample) to an equivalent concentration of the same analyte standard in H₂O:ACN (95:5; v/v). IS-normalized matrix factors were used to evaluate the matrix effect.

2.5.5 Stability

The stability in plasma was performed using triplicate QC samples at two concentration levels (LQC and HQC) of each analyte. Samples were exposed to different storage conditions: (1) bench-top stability (at room temperature), (2) freeze and thaw stability (five freeze-thaw (−20 °C) cycles), (3) long-term stability (at −20 °C), and (4) autosampler stability (in autosampler at 4 °C).

2.6 Application to clinical samples for routine TDM

To demonstrate the applicability of the method, the validated method was applied to pediatric patients at the Children’s Hospital of Nanjing Medical University, who were treated with one or more antibiotics included in this assay. Blood samples were collected in EDTA anticoagulant tubes during steady-state concentrations (24–48 h post-treatment initiation). The study complied with the Declaration of Helsinki and received ethical approval from the Children’s Hospital of Nanjing Medical University ethics committee.

3 Results

3.1 Method development

3.1.1 Optimization of mass spectrometry

Using a syringe pump, ESI and atmospheric pressure chemical ionization (APCI) were compared to obtain a more suitable ionization mode. All ATBs generated the protonated molecular ion in positive-ion mode in full scan with a m/z range from 100 to 2000. TIG and CAS were monitored with [M + 2H]²⁺ species, and other ATBs were monitored with [M + H]⁺ species. The MRM transitions with the best S/N for ATBs and internal standard were presented in Table 2, and the MS2 spectra are shown in Figure 1. The sensitivity of the signal was largely dependent on the mass spectrometric parameters for the ESI process. The CUR (from 25 to 40 psi), CAD (from 4 to 10 units), IS (from 4,500 to 5,500 V), GS1 (from 40 to 55 psi), GS2 (from 40 to 55 psi), and TEM (from 450 °C to 550 °C) were optimized to obtain the highest intensity of protonated molecules. Similarly, parameters like dwell time, declustering potential (DP), entrance potential (EP), collision energy (CE), and collision cell exit potential (CXP) were also optimized. The MS operating parameters are summarized in Table 2.

Figure 1

Figure 1. The MS2 chromatograms for 11 ATBs and ISs.

3.1.2 Liquid chromatographic conditions optimization

In this study, the chromatographic columns of different particle sizes (from 1.7 to 5 μm), lengths (from 50 to 100 mm), and packing materials (core-shell particles and porous particles) were compared to obtain better separation, peak shapes, and resolution. Finally, the Kinetex C18 (2.1 × 75 mm, 2.6 μm) column provided better chromatographic separation and peak shapes. The effects of column temperature and flow rate are also taken into account. The high abundance and peak shape were observed at a column temperature of 45 °C and a mixed flow rate of 0.25/0.5 mL/min. A methanol formic acid mixture (99.9/0.1; v/v) as precipitant could achieve a good resolution and lower background noise.

Considering the wide polarity distribution of antimicrobial agents, a gradient elution instead of isocratic elution was utilized to obtain narrow and symmetrical peak shapes. A “corner-folded cleaver-shaped” gradient elution was established, with an initial B-phase set at 8% so that the carry-over of TIG and MOX was reduced (Figure 2). However, significant carry-over was still observed. As a result, a complex needle wash solution containing ACN: MeOH: IPA: H₂O (4:4:1:1, v/v/v/v) was used to decrease the carry-over.

Figure 2

Figure 2. The “corner-folded cleaver-shaped” gradient elution program for 11 ATBs. Under this gradient condition, the chromatographic peak pattern of each antibiotic.

Different concentrations of FA and NH₄AC were added to the mobile phase. Results showed that the carry-over of TIG and MOX significantly decreased when the FA concentrations increased. In addition, different concentrations of NH₄AC showed minimal impact on the chromatographic performance of the analytes. Thus, the addition of 0.1% FA to H₂O/ACN was ultimately selected as the mobile phase.

3.2 Method validation

3.2.1 Selectivity and carryover

The representative MRM chromatograms of the double blank sample (A) and the QC sample of 11 ATBs (B) are shown in Figure 3. No significant interference was found at the retention times of each analyte. Moreover, no interference was observed between each analyte.

Figure 3

Figure 3. Representative chromatograms of double blank sample, IS of each ATB, LLOQ of 11 ATBs (Caspofungin, CAS; Cefazolin, CFZ; Cefmetazole, CMZ; Imipenem, IPM; Linezolid, LZD; Latamoxef, MOX; Meropenem, MEM; Posaconazole, POS; Tigecycline, TIG; Piperacillin, PRL; Voriconazole, VCZ; Caspofungin Acetate-d4, CAS-d4; Linezolid-d3, LZD-d3; Meropenem-d6, MEM-d6; Piperacillin-d5, PER-d5; Voriconazole-d3, VCZ-d3).

Additionally, the carry-over was defined as the percentage of responses from the blank sample to the respective LLOQ sample. No significant carry-over was observed for the tested ATBs, except for MOX in the system (Table 3).

Table 3

Table 3. The selectivity and carry-over of 11 ATBS and ISs.

3.2.2 Linearity and LLOQ

The typical regression equations of the calibration curves are listed in Table 2. The linearity of the calibration was satisfactory over the concentration range for all the compounds, with correlation coefficients higher than 0.99. The representative chromatograms of LLOQ (S/N > 5) are shown in Figure 3, and the accuracy in plasma ranged from 88.5% to 108.0%, and the RSD of precision was less than 13.3%. The results indicated that this method was accurate and reliable.

Figure 4

Figure 4. The stability of 11 antimicrobial agents under different storage conditions. A at room temperature. B in a 4 °C refrigerator. C in the autosampler at 4 °C after extraction.

3.2.3 Accuracy and precision

The intra- and inter-day accuracy and precision data are listed in Table 4. The intra-day precision ranged from 2.9% to 10.2%, and the mean accuracy values ranged from 90.7% to 111.8%. The inter-day precision and mean accuracy ranged from 4.0% to 10.3% and from 92.5% to 108.3%. The results demonstrated that the method was accurate and precise (RSD <15%, RE within ±15%).

Table 4

Table 4. Intra- and inter-day accuracy and precision of the LC-MS/MS method for the 15 ATBs used in this study (n = 6).

3.2.4 Recovery and matrix effect

Table 5 presents the recovery and matrix effects of analytes. The RSD of matrix effects derived from QC samples was below 10.9%. The RSD of recovery was between 1.5% and 12.1%, which met the criteria.

Table 5

Table 5. Matrix effects and recovery for the 15 ATBs.

3.2.5 Stability

The stability results of all ATBs are summarized in Table 6. In plasma stored at room temperature, IPM resulted in the most unstable drug, with a loss percentage at 4 h > 15%. After 8 h, a degradation trend became evident for MEM, MOX, and PRL. The autosampler stability is limited for IPM, where significant degradation occurs after 24 h. All the other compounds remained stable in the autosampler at 4 °C up to 5 days. In a 4 °C refrigerator, the compounds in the plasma samples tested were stable for at least 24 h, except for IPM, which was only stable for 8 h. At −80 °C all drugs were stable for 37 days. Freeze and thaw stability results were within the acceptability ranges for all the drugs, except for IPM.

Table 6

Table 6. Stability of 15 ATBs in human plasma.

3.3 Application to clinical samples for routine TDM

The validated method was routinely applied to the concentration monitoring of ATBs in 26 plasma samples from 16 patients treated at the Children’s Hospital of Nanjing Medical University (Table 7). Some patients were under a combined therapy, and their samples contained two antibiotics simultaneously. Among the 25 samples, LZD was present in 24 of them; the concentration ranged from 0.03 to 11.7 μg/mL for LZD. The lowest value for LZD was actually below the LLOQ (0.25 μg/mL). Nevertheless, we consider that the clinical significance of these low levels was essentially the same. At the same time, the Internal Quality Control practice of MEM, IPM, LZD, and TIG has also been qualified, which was organized by the National Center for Clinical Laboratories (NCCL).

Table 7

Table 7. Details of patients with infection and drug through concentration (C_min) in plasma.

4 Discussion

Indeed, blood-based TDM has the potential to optimize antimicrobial concentrations through personalized dosing, maximizing therapeutic efficacy, minimizing drug-related toxicity, and mitigating the development of antimicrobial resistance (Brasier et al., 2023). Of course, the optimal analytical method is a prerequisite for achieving these clinical objectives. Impressively, in the current study, we have developed and validated a LC-ESI-MS/MS method for the simultaneous determination of 11 antibiotics. Although several reports exist (Lu et al., 2022; Barco et al., 2020), our method still holds advantages in terms of sample consumption, sample pretreatment, analysis time, and linear range. Of note, our method’s development involved resolving critical analytical hurdles, thereby facilitating accurate multi-antimicrobial quantification.

Firstly, for column selection, we decided to use reverse-phase chromatography based on C18 chemistry in accordance with many other methods (Table 1). Given the relatively high polarity of β-lactam antibiotics, we initially attempted to separate these agents using a HILIC column (4.6 mm × 50 mm, 3 μm) and Waters (4.6 mm × 75 mm, 3.5 μm) (Kahsay et al., 2014; Rehm and Rentsch, 2020). However, compounds such as CAS and TIG, which exhibit extremely high hydrophobicity (logP >3), showed negligible retention under HILIC conditions. This phenomenon was likely attributed to the strong hydrophobic interactions or hydrogen bonding between the hydrophobic side chains of CAS (e.g., the myristoyl group) and residual silanol or amino groups on the HILIC stationary phase (Jandera, 2011), leading to irreversible adsorption and failure to elute detectable peaks.

To address this challenge, we tested various chromatographic columns packed with core-shell stationary phases for the analysis. The core-shell design minimizes longitudinal diffusion (B term in the Van Deemter equation) to reduce peak broadening, while its reduced eddy diffusion (A term) and enhanced mass transfer kinetics (C term) collectively yield flatter Van Deemter curves to enhance column efficiency (theoretical plates increased by ∼30%). This is quantitatively described by: H = A + B/u + C·u, allowing for operation at higher flow rates (0.5 mL/min) without significant loss of efficiency (Hayes et al., 2014; Gritti et al., 2010). By optimizing a gradient elution program starting with a high aqueous phase (e.g., 92% H₂O containing 0.1% FA) and gradually increasing the organic phase (ACN), we achieved stable separation of antibiotics with wide polarity ranges. For instance, polar β-lactams (e.g., IPM, logP ∼1.82) were retained early in the run, while highly hydrophobic compounds like CAS eluted later with sharp peaks (asymmetry factor <1.2), demonstrating robust resolution and reproducibility.

In addition, chromatographic performance depends critically on column geometry and particle characteristics. While longer columns (e.g., 100 or 150 mm) enhance resolution through increased phase interactions, they proportionally extend analysis time. Similarly, smaller particles (e.g., <3 μm) improved efficiency but required higher operating pressures (Nguyen et al., 2006). Following parameter evaluation, a Kinetex C18 column (2.1 × 75 mm; 2.6 μm) was selected to optimally balance chromatographic separation efficiency with practical analytical constraints.

Secondly, carryover is a prevalent issue in LC-ESI-MS/MS analysis, which can significantly impact the accuracy and reliability of analytical results. Carryover can originate from the autosampler, switching system, and the liquid chromatography column (Hughes et al., 2007). Indeed, we encountered significant carryover challenges during the development of this assay. Impressively, the carryover contamination from the autosampler and separation column was systematically identified and eliminated. A more thorough auto-sampler flushing (before and after rising) and washing procedures using different solvents (MeOH, ACN, 50%MeOH, and ACN:MeOH:IPA:H₂O (4:4:1:1, v/v/v/v)) were tested. Following unsuccessful autosampler washing procedures, column-derived carryover was investigated via duplicate solvent gradient analysis.

Interestingly, we observed significant column memory effects for MOX. To address this issue, we modified the elution gradient by increasing the initial proportion of the organic phase and extending the elution time. Despite these efforts, chromatographic memory effects persisted, yielding MOX carryover peaks at approximately 70% of corresponding analyte signals. Such high carryover rates posed a significant challenge, especially when analyzing low-concentration samples following high-concentration ones, as it could introduce substantial deviations in the analytical results.

One approach to minimize carry-over involves sequential blank injections interspersed between sample analyses. While sequential blank injections progressively reduce carry-over peaks, they concomitantly increase analytical run times, rendering this approach suboptimal for high-throughput workflows. To achieve further carryover mitigation, a CE defect strategy was implemented (Shen et al., 2021). During method optimization, MOX intensity exceeding 10⁵ cps prompted incremental CE reduction (5 eV steps from 20 to 10 eV). At 50% CE reduction (10 eV), the maximum MOX response decreased to 10⁴ cps, with residual carryover reduced to 18% – satisfying predefined acceptance criteria.

Finally, given the well-documented instability of β-lactam antibiotics, particularly in biological matrix, stringent sample handling protocols were implemented throughout the investigation. This included sample transportation, pre-analytical, and analytical process. Consistent with previous reports (Dailly et al., 2011), our stability studies revealed that significant degradation of IPM at ambient temperature, with observable analyte loss occurring within 2 h. Other β-lactams (like MOX, MEM, and PRL) showed a degradation trend after 4 and 6 h, respectively (Figure 4). To ensure analyte integrity, the following measures were strictly enforced: (1) all specimens were maintained at 0 °C–4 °C during transportation using ice-packed containers; (2) time from collection to laboratory processing was constrained to <2 h; (3) immediate centrifugation was conducted at 4 °C (3,000 × g, 10 min) upon arrival; and (4) the separated plasma was then immediately frozen at −80 °C for further analysis. Indeed, these steps maintained >95% analyte stability throughout the sample handling workflow. Thus, the 2-h processing window was selected based on degradation kinetics while accommodating typical clinical logistics.

Compared to existing assays, our study demonstrates obvious advantages in three key aspects: (1) By overcoming the inherent limitation of HILIC in retaining hydrophobic analytes while fully leveraging the high efficiency and versatility of core-shell columns, we developed a novel one-step precipitation method enabling simultaneous detection of 11 structurally diverse antimicrobial agents. (2) We propose comprehensive implementation of stringent sample-handling protocols throughout the analytical workflow to minimize pre-analytical degradation during transport and processing, thereby enhancing measurement accuracy in real-world clinical settings. (3) Crucially, our established synchronous quantification method for these 11 antimicrobials meets all critical internal validation criteria—including precision, accuracy, linearity, and stability. Furthermore, external validation conducted through the NCCL proficiency testing program confirmed the method’s reliability for key antibiotics, including MEM, IPM, LZD, and TIG.

Certainly, our study also has its limitations. We have refined the extraction process of the samples and conducted a comparison of the extraction effects using different precipitants (MeOH, ACN, MeOH: ACN, ACN containing 0.1% FA, and MeOH containing 0.1% FA). Our results did not show significant differences in extraction recovery rates. In our study, the recovery of IPM was only about 40% (Table 4). As reported by Ohmori et al., the recovery for doripenem was found to be approximately 50%, whereas the extraction rates for other compounds were consistently above 80% (Colin et al., 2013). Given these findings, it was evident that alternative sample preparation techniques should be explored to enhance the recovery efficiency of individual compounds in the samples. Despite the observed low recovery IPM in this work, these limitations did not affect its quantitative analysis, owing to the high specificity and sensitivity of LC-ESI-MS/MS. The simple sample pretreatment process further improves its compatibility with clinical sample analysis.

In the clinical application, due to the nature of the available clinical samples, we only tested the plasma concentrations of LZD, POS, MEM, and MOX. Therefore, it was impossible to comprehensively evaluate the clinical performance of the newly established method, which might affect the general applicability of the validation results across all analytes. To address this issue, we will implement this method in routine clinical practice and optimize the assay with larger size samples.

5 Conclusion

A rapid and sensitive LC-ESI-MS/MS method for the simultaneous determination of 11 ATBs (CAS, IPM, LZD, MEM, MOX, PRL, VCZ, POS, CMZ, TIG, and CFZ) has been presented. The validated method has been proven to be accurate and precise with a simple protein precipitation procedure. Moreover, the reproducible method was successfully used for monitoring 6 ATBs (LZD, MEM, MOX, POS, IPM, TIG) in clinical practice. The concentration of ATBs could be quantified using small volumes of human plasma (10 μL), and it is significantly important for pediatric patients undergoing polytherapy. Therefore, the determination of 11 ATBs levels can provide valuable information for pediatricians to implement individualized treatment for children and reduce the occurrence of antibiotic resistance.

Data availability statement

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding authors.

Ethics statement

The studies involving humans were approved by Children’s Hospital of Nanjing Medical University ethics committee. The studies were conducted in accordance with the local legislation and institutional requirements. The human samples used in this study were acquired from primarily isolated as part of your previous study for which ethical approval was obtained. Written informed consent for participation was not required from the participants or the participants’ legal guardians/next of kin in accordance with the national legislation and institutional requirements.

Author contributions

Y-YZ: Methodology, Validation, Writing – original draft, Writing – review and editing. JW: Data curation, Software, Validation, Writing – original draft, Writing – review and editing. H-LG: Conceptualization, Formal Analysis, Funding acquisition, Writing – review and editing. S-YW: Formal Analysis, Methodology, Writing – review and editing. H-FZ: Data curation, Formal Analysis, Writing – review and editing. P-FZ: Conceptualization, Resources, Supervision, Writing – review and editing. H-JM: Formal Analysis, Investigation, Resources, Writing – review and editing. FC: Conceptualization, Data curation, Funding acquisition, Investigation, Project administration, Writing – review and editing, Writing – original draft.

Funding

The author(s) declared that financial support was received for this work and/or its publication. This study was supported by the Specially-Appointed Medical Expert Project of Jiangsu Commission of Health (2019), the New Medical Technology Project from the Children’s Hospital of Nanjing Medical University (XJS202269 and XJS2024EJ05). This study was also supported by Jiangsu Health International Exchange Program.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

References

Abdul-Aziz, M. H., Alffenaar, J. C., Bassetti, M., Bracht, H., Dimopoulos, G., Marriott, D., et al. (2020). Antimicrobial therapeutic drug monitoring in critically ill adult patients: a position paper. Intensive Care Med. 46 (6), 1127–1153. doi:10.1007/s00134-020-06050-1

PubMed Abstract | CrossRef Full Text | Google Scholar

Brasier, N., Ates, H. C., Sempionatto, J. R., Cotta, M. O., Widmer, A. F., Eckstein, J., et al. (2023). A three-level model for therapeutic drug monitoring of antimicrobials at the site of infection. Lancet Infect. Dis. 23 (10), e445–e453. doi:10.1016/s1473-3099(23)00215-3

PubMed Abstract | CrossRef Full Text | Google Scholar

Barco, S., Mesini, A., Barbagallo, L., Maffia, A., Tripodi, G., Pea, F., et al. (2020). A liquid chromatography-tandem mass spectrometry platform for the routine therapeutic drug monitoring of 14 antibiotics: application to critically ill pediatric patients. J. Pharm. Biomed. Anal. 186, 113273. doi:10.1016/j.jpba.2020.113273

PubMed Abstract | CrossRef Full Text | Google Scholar

Bruns, N., and Dohna-Schwake, C. (2022). Antibiotics in critically ill children-a narrative review on different aspects of a rational approach. Pediatr. Res. 91 (2), 440–446. doi:10.1038/s41390-021-01878-9

PubMed Abstract | CrossRef Full Text | Google Scholar

Colin, P., De Bock, L., T'Jollyn, H., Boussery, K., and Van Bocxlaer, J. (2013). Development and validation of a fast and uniform approach to quantify beta-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

PubMed Abstract | CrossRef Full Text | Google Scholar

Collaborators, G. A. R. (2022). Global mortality associated with 33 bacterial pathogens in 2019: a systematic analysis for the global burden of disease study 2019. Lancet 400 (10369), 2221–2248. doi:10.1016/s0140-6736(22)02185-7

PubMed Abstract | CrossRef Full Text | Google Scholar

Crossley, K. B., Rotschafer, J. C., Chern, M. M., Mead, K. E., and Zaske, D. E. (1980). Comparison of a radioimmunoassay and a microbiological assay for measurement of serum vancomycin concentrations. Antimicrob. Agents Chemother. 17 (4), 654–657. doi:10.1128/AAC.17.4.654

PubMed Abstract | CrossRef Full Text | Google Scholar

Dailly, E., Bouquie, R., Deslandes, G., Jolliet, P., and Le Floch, R. (2011). A liquid chromatography assay for a quantification of doripenem, ertapenem, imipenem, meropenem concentrations in human plasma: application to a clinical pharmacokinetic study. J. Chromatogr. B Anal. Technol. Biomed. Life Sci. 879 (15-16), 1137–1142. doi:10.1016/j.jchromb.2011.03.038

PubMed Abstract | CrossRef Full Text | Google Scholar

Ferrari, D., Ripa, M., Premaschi, S., Banfi, G., Castagna, A., and Locatelli, M. (2019). LC-MS/MS method for simultaneous determination of linezolid, meropenem, piperacillin and teicoplanin in human plasma samples. J. Pharm. Biomed. Anal. 169, 11–18. doi:10.1016/j.jpba.2019.02.037

PubMed Abstract | CrossRef Full Text | Google Scholar

Gritti, F., Leonardis, I., Abia, J., and Guiochon, G. (2010). Physical properties and structure of fine core-shell particles used as packing materials for chromatography relationships between particle characteristics and column performance. J. Chromatogr. A 1217 (24), 3819–3843. doi:10.1016/j.chroma.2010.04.026

PubMed Abstract | CrossRef Full Text | Google Scholar

Hayes, R., Ahmed, A., Edge, T., and Zhang, H. (2014). Core-shell particles: preparation, fundamentals and applications in high performance liquid chromatography. J. Chromatogr. A 1357, 36–52. doi:10.1016/j.chroma.2014.05.010

PubMed Abstract | CrossRef Full Text | Google Scholar

Hu, M. W., Anne, L., Forni, T., and Gottwald, K. (1990). Measurement of vancomycin in renally impaired patient samples using a new high-performance liquid chromatography method with vitamin B12 internal standard: comparison of high-performance liquid chromatography, emit, and fluorescence polarization immunoassay methods. Ther. Drug Monit. 12 (6), 562–569. doi:10.1097/00007691-199011000-00009

PubMed Abstract | CrossRef Full Text | Google Scholar

Hughes, N. C., Wong, E. Y., Fan, J., and Bajaj, N. (2007). Determination of carryover and contamination for mass spectrometry-based chromatographic assays. AAPS J. 9 (3), E353–E360. doi:10.1208/aapsj0903042

PubMed Abstract | CrossRef Full Text | Google Scholar

Jandera, P. (2011). Stationary and Mobile phases in hydrophilic interaction chromatography: a review. Anal. Chim. Acta 692 (1-2), 1–25. doi:10.1016/j.aca.2011.02.047

PubMed Abstract | CrossRef Full Text | Google Scholar

Kahsay, G., Song, H., Van Schepdael, A., Cabooter, D., and Adams, E. (2014). Hydrophilic interaction chromatography (HILIC) in the analysis of antibiotics. J. Pharm. Biomed. Anal. 87, 142–154. doi:10.1016/j.jpba.2013.04.015

PubMed Abstract | CrossRef Full Text | Google Scholar

Kearns, G. L., Abdel-Rahman, S. M., Alander, S. W., Blowey, D. L., Leeder, J. S., and Kauffman, R. E. (2003). Developmental pharmacology--drug disposition, action, and therapy in infants and children. N. Engl. J. Med. 349 (12), 1157–1167. doi:10.1056/NEJMra035092

PubMed Abstract | CrossRef Full Text | Google Scholar

Lefeuvre, S., Bois-Maublanc, J., Hocqueloux, L., Bret, L., Francia, T., Eleout-Da Violante, C., et al. (2017). A simple ultra-high-performance liquid chromatography-high resolution mass spectrometry assay for the simultaneous quantification of 15 antibiotics in plasma. J. Chromatogr. B Anal. Technol. Biomed. Life Sci. 1065-1066, 50–58. doi:10.1016/j.jchromb.2017.09.014

PubMed Abstract | CrossRef Full Text | Google Scholar

Legrand, T., Vodovar, D., Tournier, N., Khoudour, N., and Hulin, A. (2016). Simultaneous determination of eight beta-Lactam antibiotics, amoxicillin, cefazolin, cefepime, cefotaxime, ceftazidime, cloxacillin, oxacillin, and piperacillin, in human plasma by using ultra-high-performance liquid chromatography with ultraviolet detection. Antimicrob. Agents Chemother. 60 (8), 4734–4742. doi:10.1128/AAC.00176-16

PubMed Abstract | CrossRef Full Text | Google Scholar

Li, W., Li, R., Liu, H., Guo, X., Shaikh, A. S., Li, P., et al. (2017). A comparison of liquid chromatography-tandem mass spectrometry (LC-MS/MS) and enzyme-multiplied immunoassay technique (EMIT) for the determination of the cyclosporin A concentration in whole blood from Chinese patients. Biosci. Trends 11 (4), 475–482. doi:10.5582/bst.2017.01121

PubMed Abstract | CrossRef Full Text | Google Scholar

Lu, W., Pan, M., Ke, H., Liang, J., Liang, W., Yu, P., et al. (2022). An LC-MS/MS method for the simultaneous determination of 18 antibacterial drugs in human plasma and its application in therapeutic drug monitoring. Front. Pharmacol. 13, 1044234. doi:10.3389/fphar.2022.1044234

PubMed Abstract | CrossRef Full Text | Google Scholar

Martson, A. G., Sturkenboom, M. G. G., Stojanova, J., Cattaneo, D., Hope, W., Marriott, D., et al. (2020). How to design a study to evaluate therapeutic drug monitoring in infectious diseases? Clin. Microbiol. Infect. 26 (8), 1008–1016. doi:10.1016/j.cmi.2020.03.008

PubMed Abstract | CrossRef Full Text | Google Scholar

Neef, S. K., Hinderer, A. D., Arbash, W., Kinzig, M., Sorgel, F., Wunder, C., et al. (2024). A high performance liquid chromatography-tandem mass spectrometry assay for therapeutic drug monitoring of 10 drug compounds commonly used for antimicrobial therapy in plasma and serum of critically ill patients: method optimization, validation, cross-validation and clinical application. Clin. Chim. Acta 559, 119690. doi:10.1016/j.cca.2024.119690

PubMed Abstract | CrossRef Full Text | Google Scholar

Nguyen, D. T., Guillarme, D., Rudaz, S., and Veuthey, J. L. (2006). Fast analysis in liquid chromatography using small particle size and high pressure. J. Sep. Sci. 29 (12), 1836–1848. doi:10.1002/jssc.200600189

PubMed Abstract | CrossRef Full Text | Google Scholar

Ohmori, T., Suzuki, A., Niwa, T., Ushikoshi, H., Shirai, K., Yoshida, S., et al. (2011). Simultaneous determination of eight beta-lactam antibiotics in human serum by liquid chromatography-tandem mass spectrometry. J. Chromatogr. B Anal. Technol. Biomed. Life Sci. 879 (15-16), 1038–1042. doi:10.1016/j.jchromb.2011.03.001

PubMed Abstract | CrossRef Full Text | Google Scholar

Radovanovic, M., Day, R. O., Jones, G. D. R., Galettis, P., and Norris, R. L. G. (2022). LC-MS/MS method for simultaneous quantification of ten antibiotics in human plasma for routine therapeutic drug monitoring. J. Mass Spectrom. Adv. Clin. Lab. 26, 48–59. doi:10.1016/j.jmsacl.2022.11.001

PubMed Abstract | CrossRef Full Text | Google Scholar

Rashed, A. N., Jackson, C., Gastine, S., Hsia, Y., Bielicki, J., Standing, J. F., et al. (2019). Pediatric pharmacokinetics of the antibiotics in the access and watch groups of the 2019 WHO model list of essential medicines for children: a systematic review. Expert Rev. Clin. Pharmacol. 12 (12), 1099–1106. doi:10.1080/17512433.2019.1693257

PubMed Abstract | CrossRef Full Text | Google Scholar

Rawson, T. M., Wilson, R. C., O'Hare, D., Herrero, P., Kambugu, A., Lamorde, M., et al. (2021). Optimizing antimicrobial use: challenges, advances and opportunities. Nat. Rev. Microbiol. 19 (12), 747–758. doi:10.1038/s41579-021-00578-9

PubMed Abstract | CrossRef Full Text | Google Scholar

Rehm, S., and Rentsch, K. M. (2020). HILIC LC-MS/MS method for the quantification of cefepime, imipenem and meropenem. J. Pharm. Biomed. Anal. 186, 113289. doi:10.1016/j.jpba.2020.113289

PubMed Abstract | CrossRef Full Text | Google Scholar

Roberts, J. A., Abdul-Aziz, M. H., Lipman, J., Mouton, J. W., Vinks, A. A., Felton, T. W., et al. (2014). Individualised antibiotic dosing for patients who are critically ill: challenges and potential solutions. Lancet Infect. Dis. 14 (6), 498–509. doi:10.1016/S1473-3099(14)70036-2

PubMed Abstract | CrossRef Full Text | Google Scholar

Schwenzer, K. S., Wang, C. H., and Anhalt, J. P. (1983). Automated fluorescence polarization immunoassay for monitoring vancomycin. Ther. Drug Monit. 5 (3), 341–345. doi:10.1097/00007691-198309000-00017

PubMed Abstract | CrossRef Full Text | Google Scholar

Shen, J., Wang, H., Huang, H., Li, H., Li, C., Yan, C., et al. (2021). Absolute quantitative analysis of endogenous neurotransmitters and amino acids by liquid chromatography-tandem mass spectrometry combined with multidimensional adsorption and collision energy defect. J. Chromatogr. A 1638, 461867. doi:10.1016/j.chroma.2020.461867

PubMed Abstract | CrossRef Full Text | Google Scholar

Shipkova, M., and Svinarov, D. (2016). LC-MS/MS as a tool for TDM services: where are we? Clin. Biochem. 49 (13-14), 1009–1023. doi:10.1016/j.clinbiochem.2016.05.001

PubMed Abstract | CrossRef Full Text | Google Scholar

Singer, B., Stevens, R. W., Westley, B. P., and Nicolau, D. P. (2020). Falsely elevated vancomycin-concentration values from enzyme immunoassay leading to treatment failure. Am. J. Health Syst. Pharm. 77 (1), 9–13. doi:10.1093/ajhp/zxz258

PubMed Abstract | CrossRef Full Text | Google Scholar

Sumi, C. D., Heffernan, A. J., Lipman, J., Roberts, J. A., and Sime, F. B. (2019). What antibiotic exposures are required to suppress the emergence of resistance for gram-negative bacteria? A systematic review. Clin. Pharmacokinet. 58 (11), 1407–1443. doi:10.1007/s40262-019-00791-z

PubMed Abstract | CrossRef Full Text | Google Scholar

Sun, H., Xing, H., Tian, X., Zhang, X., Yang, J., and Wang, P. (2022). UPLC-MS/MS method for simultaneous determination of 14 antimicrobials in human plasma and cerebrospinal fluid: application to therapeutic drug monitoring. J. Anal. Methods Chem. 2022, 7048605. doi:10.1155/2022/7048605

PubMed Abstract | CrossRef Full Text | Google Scholar

Tangden, T., Ramos Martin, V., Felton, T. W., Nielsen, E. I., Marchand, S., Bruggemann, R. J., et al. (2017). The role of infection models and PK/PD modelling for optimising care of critically ill patients with severe infections. Intensive Care Med. 43 (7), 1021–1032. doi:10.1007/s00134-017-4780-6

PubMed Abstract | CrossRef Full Text | Google Scholar

U.S. Food and Drug Administration (2018). Bioanalytical method validation guidance for industry. Available online at: https://www.fda.gov/regulatory-information/search-fda-guidance-documents/bioanalytical-method-validation-guidance-industry.

Google Scholar

Verdier, M. C., Tribut, O., Tattevin, P., Le Tulzo, Y., Michelet, C., and Bentue-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 (10), 4873–4879. doi:10.1128/AAC.00533-11

PubMed Abstract | CrossRef Full Text | Google Scholar