Pharmacokinetic study of isavuconazonium in human plasma measured by HPLC-MS/MS and its use in healthy Chinese subjects

ABSTRACT

Aims

To establish a rapid, sensitive HPLC-MS/MS method for quantifying isavuconazonium in human plasma and characterize its pharmacokinetics in healthy Chinese subjects under fasting and postprandial conditions.

Materials & methods

Plasma samples were processed via acetonitrile protein precipitation. Separation was performed on an LC-20ADXR Plus C18 column with gradient elution (0.01% formic acid/acetonitrile). Detection used a Triple Quad 4500 mass spectrometer with MRM; isavuconazole-d4 was the internal standard. The method was validated, then applied to a crossover study in 32 healthy subjects (fasting vs. postprandial).

Results

The method showed good linearity (4–4000 ng/mL, R² ≥0.9801) with LLOQ 4 ng/mL. Stability was confirmed under various conditions (e.g. 53 days at −20°C, 66 days at −80°C). Pharmacokinetic results revealed food delayed Tmax (2.5 vs. 5.0 h) and reduced Cmax (1929.68 vs. 1300.17 ng/mL) but did not affect AUC_0-t.

Conclusions

The validated HPLC-MS/MS method is rapid and reliable for therapeutic drug monitoring. Food affects absorption but not total exposure, guiding clinical dosing in Chinese populations.

Plain Language Summary

We developed a fast and accurate method to measure the concentration of isavuconazonium (a drug for fungal infections) in human blood using HPLC-MS/MS. This method works well even with different blood samples and storage conditions. We then used it to study how the drug is absorbed and processed in healthy Chinese people, both when fasting and after eating. The results showed that eating slows down the drug’s absorption and lowers its peak concentration but does not affect the total amount absorbed. This method can help doctors adjust doses to make the drug more effective and safer.

1. Introduction

Isavuconazole (ISA) [1–7] is a novel triazole antifungal medication that is effective against various pathogenic fungi, including yeasts, molds, and dimorphic fungi. It acts as an antifungal agent by inhibiting the cytochrome P450-dependent lanosterol 14α-demethylase, which is necessary for the synthesis of ergosterol [8–10]. This inhibition blocks the ergosterol biosynthesis pathway, alters the structure and function of fungal cell membranes, and ultimately leads to fungal cell death (Figure 1). In 2015, ISA was approved by the U.S. Food and Drug Administration (FDA) for the treatment of invasive aspergillosis and invasive mucormycosis. In 2019, the European Confederation of Medical Mycology and the Fungal Disease Research Group strongly recommended the use of isavuconazole for salvage therapy in the Global Guidelines for the Diagnosis and Management of Mucormycosis [11,12]. In 2021, isavuconazonium sulfate capsules were approved for marketing, becoming the first oral antifungal drug in the country to treat invasive mucormycosis in adults [13–16].

Figure 1.

Mechanism of action of isavuconazole.

Despite its efficacy, ISA exhibits significant inter-individual pharmacokinetic variability, necessitating therapeutic drug monitoring (TDM) to optimize clinical outcomes. Key factors contributing to this variability include: 1. Nonlinear PK and Absorption Saturation: ISA’s absorption is dose-dependent and may plateau at higher doses, leading to unpredictable plasma concentrations; 2. Food Effects: As demonstrated in this study, postprandial administration delays absorption (prolonged T_max) and reduces peak concentrations (lower C_max), which may impact efficacy in patients with inconsistent dietary habits; 3. Metabolic Susceptibility: ISA is metabolized by CYP3A4 and is a P-glycoprotein substrate, making it prone to drug-drug interactions (e.g., with rifampin or carbamazepine); 4. Patient-Specific Factors: Critically ill patients, those with hepatic impairment, or extreme body weights often exhibit altered drug clearance, further increasing PK variability. Isavuconazonium sulfate capsules are an oral medication with broad-spectrum antifungal properties, which can effectively treat invasive fungal infections and prevent recurrence. They exhibit better tolerability and safety than other antifungal drugs. The plasma concentration of ISA is closely related to clinical treatment outcomes and adverse reactions. The results of therapeutic drug monitoring can provide important guidance for developing and implementing clinical dosing plans, thereby optimizing antifungal efficacy. Therefore, this study aims to establish a stable, simple, and efficient method for determining the concentration of ISA in human plasma using HPLC-MS/MS [13,17–25].

Prior studies report high variability in exposure (AUC CV > 60%) and prolonged half-life (up to 130 hours), which can lead to subtherapeutic or toxic levels under fixed dosing [26–29]. TDM is thus essential to ensure therapeutic efficacy while minimizing adverse effects (e.g., hepatotoxicity or QT prolongation). However, existing assays for ISA quantification face challenges in speed, sensitivity, or matrix interference.

To address these gaps, we developed a rapid, sensitive HPLC-MS/MS method for ISA quantification in human plasma. This method was validated per international guidelines [22] and applied to a pharmacokinetic study in healthy Chinese subjects under fasting and postprandial conditions.

Existing methods for ISA quantification in plasma have several limitations. Early studies reported HPLC-UV methods with low sensitivity (LLOQ ≥50 ng/mL) and long run times (≥10 min), failing to meet the requirements for TDM [13]. Some LC-MS/MS methods improved sensitivity (LLOQ 5–10 ng/mL) but suffered from significant matrix effects (CV > 15%) in hemolyzed or hyperlipidemic plasma, leading to inaccurate quantification [24,25]. Additionally, previous assays required complex sample preparation (e.g., solid-phase extraction), which prolonged analysis time (3–6 h per batch) and reduced throughput [30]. Pharmacokinetic data for ISA in Chinese populations remain scarce; most studies focused on Western or Japanese cohorts, with limited data on food effects and inter-individual variability in healthy Chinese subjects.

Our method addresses these limitations through several key innovations: (1) Rapid analysis: Chromatographic separation was achieved within 1 min (total run time 2.6 min), significantly shorter than conventional methods (3–10 min); (2) Enhanced sensitivity: The LLOQ of 4 ng/mL is lower than reported values (5–50 ng/mL), enabling accurate quantification of low plasma concentrations during the elimination phase; (3) Minimal matrix effects: Interference from hemolyzed or hyperlipidemic plasma was <7.2%, outperforming previous methods with matrix effects >15% [9]; (4) Robust stability: Extended stability (53 days at −20°C, 66 days at −80°C) and tolerance to 5 freeze-thaw cycles facilitate flexible sample storage and processing, which is critical for multi-center clinical studies.

2. Experimental process

2.1. Materials and reagents

Isavuconazole reference standard (lot number 0629-RD-0020, purity 100%) was procured from CATO Research Chemicals Inc. The internal standard, Isavuconazole-d4 (lot number 0814-RF-0041, purity 100%, deuterium enrichment ≥ 99%), was purchased from CATO Research Chemicals Inc. Isavuconazole sulfate capsules (reference preparation) were obtained from Pfizer Australia Pty Ltd. Methanol (HPLC-grade), acetonitrile (LC-MS/HPLC-grade), and formic acid (ACS-grade) were all sourced from Fisher Scientific. These solvents and reagents were used without further purification either before or during the experiment. Ultrapure water was prepared using an arium® comfort II purification system from Sartorius Company.

2.2. Instrument and conditions for HPLC-MS/MS

Chromatographic separation was performed on a Shimadzu LC-20ADXR system, equipped with a C18 column (Ultimate XB-C18, 2.1 × 50.0 mm, 5.0 μm, Welch). The mobile phase consisted of an aqueous solution containing 0.01% formic acid (A) and acetonitrile (B). The gradient elution program was as follows: starting at 0.01 minutes, the concentration of mobile phase B (acetonitrile) was set at 50.0%. At 1.20 minutes, this concentration increased from 50.0% to 70.0%, and then further increased to 95.0% at 1.30 minutes. The 95.0% concentration was maintained until 1.90 minutes. At 2.00 minutes, the concentration of mobile phase B decreased from 95.0% back to 50.0%, and the elution process ended at 2.60 minutes when the system controller halted the operation. The flow rate was set at 1.0 mL/min, the column temperature was maintained at 35°C, and the sample volume injected was 2 μL.

The ion source used was a Turbo Spray from AB Sciex, and the mass spectrometry detector was an AB Sciex Triple Quad 4500. The analysis software included Analyst 1.6.3 and 1.7.2 from AB Sciex, along with Watson LIMS 7.5. Electrospray Ionization (ESI) in positive ion mode was employed for ionization, and the scanning method was Multiple Reaction Monitoring (MRM). The operating parameters were as follows: Ion Spray Voltage: 5500.00 V; Turbo Ion Spray Temp: 550.00°C; Curtain Gas Type: 30.00 psi; CAD Gas Type: 9.00; EP: 10.00 V; CXP: 10.00 V; Nebulizing Gas (Gas1): 50.00 psi; Auxiliary Gas (Gas2): 50.00 psi. The monitored transitions were m/z 438.2→224.1 for Isavuconazole and m/z 442.2→224.3 for Isavuconazole-d4.

2.3. Stock solutions, standard solutions, and quality control working samples

Isavuconazole was accurately weighed and dissolved in methanol to prepare a stock solution with a concentration of 2.00 mg/mL. Standard curve working solutions with concentrations of 100, 200, 1000, 5000, 25000, 50000, 90000, and 100,000 ng/mL were obtained by diluting the stock solution with 50% methanol. Quality control working solutions with concentrations of 100 (LLOQ QC), 300 (LQC), 3000 (GQC2), 30000 (MQC), 75000 (HQC), and 400,000 (DQC) ng/mL were prepared in the same manner.

2.4. Internal standard (IS) working solution samples

The IS working solution was prepared by diluting the IS reserve solution with 50% methanol to obtain concentrations of 20,000 ng/mL and 400 ng/mL. This working solution was stored in a clear glass bottle at 2–8°C.

2.5. Standard curve samples

Standard curve samples were freshly prepared and used on the same day. The appropriate working solution was added to a blank substrate, diluted, and thoroughly mixed to obtain standard curve samples with final concentrations of 4.00, 8.00, 40.0, 200, 1000, 2000, 3600, and 4000 ng/mL, respectively.

2.6. Quality control (QC) sample

QC samples were also freshly prepared and used on the same day. The corresponding working solution was added to a blank substrate, diluted, and mixed thoroughly to obtain QC samples with final concentrations of 4.00 (LLOQ QC), 12.0 (LQC), 120 (MQC2), 1200 (MQC), 3000 (HQC), and 16,000 (DQC) ng/mL.

2.7. Sample preparation

In a 96-well plate, 50 μL of the sample was mixed with 25 μL of the IS working solution. For blank samples, 50 μL of the blank substrate was mixed with 25 μL of a 50% methanol solution. Then, 300 μL of acetonitrile was added, and the mixture was thoroughly mixed. The resulting mixture was centrifuged at 1770 × g for 15 minutes at 4°C. Subsequently, 50 μL of the supernatant was transferred to another 96-well plate. After that, 150 μL of a 20% methanol solution was added, and the plate was sealed and mixed thoroughly.

2.8. Calibration procedure and standard curve acceptance criteria

Chromatograms of samples from each analysis batch were collected using Analyst 1.6.3 and Analyst 1.7.2 (AB Sciex). The peak areas of the target components and the internal standard in the samples were obtained through automatic integration with this software.

Linearity Requirement:The standard curve correlation coefficient (R²) must be ≥ 0.9801, as per FDA Bioanalytical Method Validation Guidelines (2018) and ICH M10 requirements for LC-MS/MS assays [20]. Accuracy Requirement: Back-calculated concentrations of calibration standards must fall within ± 15% of their nominal values (except for LLOQ, which allows ± 20%). At least 75% of the calibration standards (including LLOQ) must meet this criterion. Justification for R² Threshold: while higher R² values (e.g., > 0.99) are ideal, the ± 15% accuracy requirement for back-calculated standards serves as a more stringent control of assay performance. This dual criterion (R² + accuracy) is widely adopted in bioanalytical method validation for antifungal drugs [12,15,20]. The linear regression was performed using Watson LIMS 7.5 with a weighting factor of 1/x², and the concentration values of test samples were calculated using the derived linear equation.

2.9. System suitability

For each system suitability sample, the signal-to-noise ratio (S/N) of the analyte and the IS should be no less than 5. After 6 consecutive injections, the relative standard deviation of the analyte retention time and that of the internal standard retention time should not exceed 10%. Additionally, the relative standard deviation of the response ratio between the analyte and the internal target should not exceed 15%.

2.10. Specificity

Selectivity was investigated using blank normal substrates from at least 6 different sources/donors. The response of interference peak at the retention time of analyte in blank matrix samples from different donors should not exceed 20% of the response of analyte in QC samples prepared with blank matrix from the same source. The response of the interference peak at the retention time of the internal standard should not exceed 5% of the response of the IS in the LLOQ sample prepared from the same source blank substrate.

The blank high-fat matrix, blank hemolytic matrix, blank matrix samples from mixed sources and blank reagent samples were detected. The response of the interference peak at the retention time of the analyte should not exceed 20% of the response of the analyte in the LLOQ sample. The interference peak response at the internal standard retention time should not exceed 5% of the internal standard response in the LLOQ sample. After sample treatment, the response of the interference peak at the retention time of the IS should not exceed 5% of the response of the IS in the LLOQ sample. For a zero concentration sample, the response of the interference peak at the analyte retention time should not exceed 20% of the analyte response in the LLOQ sample.

2.11. Standard curve and QC sample preparation

Stock solution (2.00 mg/mL): Isavuconazole reference standard was accurately weighed and dissolved in methanol (HPLC-grade); Standard curve working solutions: Prepared by serial dilution of the stock solution with 50% methanol to achieve concentrations of 100, 200, 1000, 5000, 25000, 50000, 90000, and 100,000 ng/mL; Spiked into blank human plasma to generate an 8-point standard curve (4.00–4000 ng/mL) with final concentrations:4.00, 8.00, 40.0, 200, 1000, 2000, 3600, and 4000 ng/mL; Quality control (QC) samples: Prepared similarly at five levels:LLOQ QC (4.00 ng/mL), LQC (12.0 ng/mL), MQC2 (120 ng/mL), MQC (1200 ng/mL), HQC (3000 ng/mL), and DQC (16000 ng/mL); QC working solutions (50% methanol): 100 (LLOQ), 300 (LQC), 3000/30000 (MQC2/MQC), 75000 (HQC), 400000 ng/mL (DQC); Internal standard (IS) working solution: Isavuconazole-d4 was diluted to 20,000 ng/mL and 400 ng/mL in 50% methanol, stored at 2–8°C in amber glass vials.

2.12. Standard curve and LOQ

Accuracy and precision assay batches were evaluated together, using QC samples at the LLOQ level. The signal-to-noise ratio (S/N) of the response signal of the LLOQ samples was assessed through the system applicability experiment of the assay batch in which the samples were tested. The deviation should be within ± 20.0%, the precision should be less than 20.0%.

2.13. Precision, accuracy of within-run batch and between-run batch

The precision and accuracy of Isavuconazole within a single run were evaluated by repeatedly analyzing QC samples. The study included a total sample size of n = 6. Various working solutions were added to plasma containing EDTA-K2 to obtain the following concentrations: LLOQ QC (4.00 ng/mL), LQC (12.0 ng/mL), MQC2 (120 ng/mL), MQC (1200 ng/mL), and HQC (3000 ng/mL). Precision was assessed by calculating the relative standard deviation (%RSD) of the detected concentrations, while accuracy was evaluated by calculating the difference between the theoretical and measured concentrations, expressed as the mean percentage difference (%Diff). The acceptable deviation was ± 15.0% for all samples, except for LLOQ QC samples, where it was ± 20.0%. Precision requirements were set at ≤ 15.0% for all samples, and ≤ 20.0% for LLOQ QC samples. For each concentration of the quality control sample, there must be at least 5 valid values.

Calculating QC samples from 3 distinct batches over the course of 2 days allowed researchers to assess the Isavuconazole’s inter-batch precision and accuracy. Each batch of QC samples (n = 6) was freshly prepared at the same concentrations as the intra-batch QC samples (LLOQ QC, LQC, MQC2, MQC, HQC). The deviation (%) should be ± 15.0% for all samples, with the deviation of LLOQ QC samples being ± 20.0%. Precision (%) should be ≤ 15.0% for all samples, and ≤ 20.0% for LLOQ QC samples.

2.14. Matrix effect

To investigate matrix effects, six batches of human plasma from different sources/donors were collected. Analytes and the IS were added to blank matrix samples to obtain samples with final concentrations equivalent to LQC and HQC samples. This was done in triplicate for each concentration level and each blank matrix. Reference solutions with the same concentrations of the IS and analyte were then prepared. The matrix effect was evaluated by calculating the ratio of the peak areas of the analyte and the IS in the matrix samples and the reference solutions, respectively. For each concentration level of the QC samples, the deviation between the mean detected value and the actual concentration should be within ± 15.0%, and the precision should not exceed 15.0%.

To assess the matrix effect of hemolyzed plasma, six HQC and LQC samples were added to it, and the results were compared with the standard curve and QC samples from conventional plasma. To evaluate the matrix effect of hyperlipemic plasma, a batch of hyperlipemic samples (with six replicates each) was prepared. Analytes and the IS were added to these samples to achieve final concentrations consistent with LQC and HQC samples, following the same process as for normal matrix samples. The precision of the matrix effect should be within 15.0%, and the mean deviation from the theoretical value must be within ± 15.0%. For each concentration of the quality control sample, there must be at least 5 valid values.

2.15. Extraction recovery results

The recovery rates were determined using a mixed substrate. Six LQC samples, six MQC samples, and six HQC samples were extracted simultaneously. The extraction solutions were then spiked with the IS and analyte to match the concentrations of the HQC, MQC, and LQC samples. Recovery was assessed by calculating the area ratios between the extracts from individual QC samples and the average peak area of blank plasma samples with added IS and analyte. The precision of extraction recoveries at each concentration level should be within 15.0%, with at least 5 valid values. The average recovery rate of each QC sample should not show a concentration-related trend, and the relative standard deviation (RSD) of recoveries between different QC concentrations should not exceed 30%.

2.16. Stability

The long-term stability of Isavuconazole in human plasma samples at −80°C and −20°C was determined. The stability of human plasma samples of Isavuconazole during five freeze-thaw cycles at −80°C and −20°C was also evaluated. Additionally, the stability of whole blood samples containing Isavuconazole was investigated at room temperature and on wet ice. The stability of treated samples (LQC and HQC) at the temperature of the automatic injector was examined. For each concentration level of the QC samples, the mean detected concentration should be within ± 15.0% of the actual concentration, and the precision should be less than 15.0%.

2.17. Pharmacokinetic study design

A pharmacokinetic study of Isavuconazole sulfate capsules (100 mg) was conducted in healthy human subjects. The aim was to evaluate the absorption degree and rate of these capsules, manufactured by Pfizer Australia Pty Ltd., under fasting and postprandial conditions in healthy individuals. The experiment followed a single-center, randomized, open-label, single-dose, two-preparation, two-cycle, crossover-design oral administration trial protocol. Given that the half-life of Isavuconazole is approximately 130 hours, a wash-out period of 35 days was designed to ensure complete elimination of the medication from the previous cycle. For both postprandial and fasting test subjects, fasting was required for at least 10 hours before either consuming a high-fat meal (postprandial group) or taking the medication (fasting group). The reference formulation was administered with 240 mL of water under fasting or postprandial conditions. The capsules were to be swallowed whole, without breaking or chewing. In the fasting test, subjects were prohibited from drinking water from 1 hour before to 1 hour after drug administration and were required to fast for 4 hours post-dosing. The second cycle occurred 35 days after cross-dosing. For the postprandial test, a high-fat, high-calorie meal was initiated 30 minutes before administration, with the entire meal consumed within 30 minutes. Medication was administered 30 minutes after meal initiation.

Venous blood samples were collected at 0 h (within 1 hour pre-dose), 0.5 h, 1 h, 1.5 h, 2 h, 2.5 h, 3 h, 3.5 h, 4 h, 4.5 h, 5 h, 5.5 h, 6 h, 8 h, 12 h, 14 h, 24 h, 36 h, 48 h, and 72 h during fasting conditions. For postprandial conditions, samples were collected at 0 h (within 1 hour pre-dose), 1 h, 2 h, 2.5 h, 3 h, 3.5 h, 4 h, 4.5 h, 5 h, 5.5 h, 6 h, 6.5 h, 7 h, 8 h, 9 h, 10 h, 12 h, 24 h, 36 h, 48 h, and 72 h. Approximately 4 mL of blood was drawn from each volunteer into EDTA-K2 collection tubes at each timepoint. Samples were gently inverted several times and placed on ice. After complete sample collection from all subjects, plasma samples were transferred to the laboratory for analysis under appropriate conditions.

The postprandial test included 32 subjects (5 females [15.6%], 27 males [84.4%]) with mean ± SD values: height 167.38 ± 7.09 cm, age 29.72 ± 6.73 years, weight 65.04 ± 8.80 kg, and BMI 23.08 ± 1.94 kg/m². The fasting trial similarly included 32 subjects (6 females [18.8%], 26 males [81.3%]) with mean ± SD values: age 29.75 ± 4.55 years, height 166.42 ± 6.96 cm, weight 62.90 ± 8.70 kg, and BMI 22.59 ± 2.18 kg/m².

2.18. Pharmacokinetic analysis

This trial was approved by the Clinical Trial Ethics Committee of Chongqing 11th People’s Hospital, in accordance with the Declaration of Helsinki and other relevant laws and guiding principles. The pharmacokinetic parameters of Isavuconazole were calculated using SAS software (version 9.4). These parameters were determined using a non-atrioventricular model (NCA) based on the individual blood concentrations of each subject. The parameter Cmax represents the maximum blood-drug concentration, which is an experimental value. The parameter AUC_0-t represents the area under the concentration-time curve from 0 to the last accurately measured concentration and was calculated using the trapezoidal method. The parameter AUC_0-∞ represents the area under the concentration-time curve from 0 to infinity. Tmax represents the time at which the blood concentration reaches its maximum, also an experimental value. T1/2 represents the time required for the blood concentration to be reduced by 50%, and λz represents the apparent terminal elimination rate constant.

3. Results

3.1. System suitability result

Table S1 displayed the RSD% for the retention time of the tested substances, the RT of the IS, and the response ratio for the 9 samples. The RSD% for the RT of the tested substances was found to be less than or equal to 0.7%, while the RSD% for the RT of the IS was less than or equal to 0.5%. Additionally, the RSD% for the response ratio was less than or equal to 5.8%. Notably, all the values were within the acceptable range specified by the standard, i.e., ≤10.0%. Consequently, the verification results can be deemed satisfactory.

3.2. Specificity

Figures S1 to S7 showed the relevant chromatograms of esoconazole and its internal standard. From these figures, it can be seen that esoconazole and its internal standard are not affected by the blank mechanism, which intuitively demonstrates the good specificity of the method. As shown in table S2, interference of the blank normal matrix from 6 different sources on the analytes was seen to range from 0.0% to 7.2%, and the interference on the IS was found to be 0.0%. The interference of the blank high-lipid matrix from 1 source on the analytes and IS was found to be 0.0%. The interference of the blank hemolyzed matrix from 1 source on the analytes and IS was found to be 0.0%. The interference of the mixed-source blank matrix on the analytes and IS was found to be 0.0%. The interference of the blank reagent solution on the analytes was found to be 1.8%, and the interference on the IS was found to be 0.0%. The interference resulting from the presence of an IS on the analytes was found to be 0.0%. The tested analytes exhibited an interference of 1.1% with the IS and its result was confirmed to comply with the acceptance criteria.

3.3. Standard curve and LOQ

The production of samples for the isavuconazole standard curve followed the experimental technique, after which they were introduced into the HPLC-MS/MS for the purpose of calibrating the curve. Figure 2 depicted the linear relationship between the peak area signal of the analyte and the concentration. The minimum detectable concentration was 4.0 ng/mL. To further substantiate the accuracy of the calibration curves, we conducted the generation of six calibration curves in the table S3 which demonstrated R² values surpassing 0.99. Furthermore, it was determined that the discrepancies observed for every concentration level and standard concentration fell within a range of ± 10.0% of the corresponding theoretical value.

Figure 2.

Standard curve.

3.4. Precision, accuracy of within-run batch and between-run batch

According to the data presented in table S4, the highest level of precision within the assay, excluding the LLOQ QC samples, was 4.6%. Additionally, the range of accuracy deviation within the assay was observed to be between −5.8% and 3.3%. The highest level of precision observed within the assay for the LLOQ QC samples was 12.6%. The range of accuracy deviation within the assay was found to be between −2.8% and 2.8%. The highest level of precision observed between assays, excluding the LLOQ QC samples, was 4.3%. The range of accuracy within assay varied from −1.7% to 1.7%. The LLOQ QC samples had a maximum inter-assay precision of 7.7%, whereas the inter-assay accuracy deviated by −1.0%. The aforementioned findings demonstrated a high level of precision and accuracy in the employed methodology.

3.5. Extraction recovery results

The results in tables S5 and table S6 showed that the Overall RSD of the recovery of isavuconazole and the IS was 2.5% and 3.1%, respectively, which was much less than the standard value, indicating a good recovery.

3.6. Matrix effect

Table 1 presented the matrix effect of normal plasma. It demonstrated that the precision of the matrix effect for the analyte at each concentration level (across all tested lots) was found to be ± 15.0%. Additionally, the precision of the mean of the normalized matrix effect within the two concentration levels of the analyte was also within 15.0%. The level of accuracy in measuring the matrix effect in hemolyzed plasma and hyperlipidemic plasma was found to be within a range of 15.0%. The presence of a positive matrix effect indicated the absence of interference between the matrix and the material under investigation. Consequently, this improved the precision and accuracy of the detection outcomes.

Table 1.

Matrix effect results.

Matrix Effect	Maximum precision (%RSD)		Mean accuracy deviation (deviation from labeled concentration) (%) Range
	LQC	HQC	LQC	HQC
Normal matrix effect	5.0	4.7	−11.7~−5.8	−1.3 ~ 2.3
Hemolytic plasma matrix effect	5.5	4.4	−4.2	−1.0
Hyperlipidemic matrix effect	7.2	1.4	0.8	0.7
Allover matrix effect	7.2	4.7	−11.7 ~ 0.8	−1.3 ~ 2.3

3.7. Solution stability results

The stock solution with methanol as solvent was stable at ambient temperature for 24 hours and at −20°C for 33 days. The stock solution with 50% methanol as solvent showed stability at room temperature for 24 hours and at −20°C for 31 days. The whole blood matrix remained stable for 2 hours at room temperature. The plasma matrix remained stable for 53 days and 66 days at −20°C and −80°C respectively. In addition, the samples remained stable even after five cycles of freezing and thawing. The prepared samples were stabilized at 5°C for 191 hours in the autosampler. The isotopic internal standard of the drug under study was used in this experimental study. Stability tests were not performed on solutions containing the internal standard because it was determined that the properties of the internal standard were similar to those of the substance under test and it was determined that the internal standard would not interact with the compounds in the respective analytical batches.

3.8. Pharmacokinetic study

Figure 3 and Table 2 illustrated the diverse pharmacokinetic parameters seen in healthy individuals following both fasting and postprandial treatment. The elimination half-lives (T_1/2) during fasting and after a meal were found to be 126.61 ± 96.36 h and 83.38 ± 48.76 h, respectively. The average value plus or minus the standard deviation (SD) of AUC_0-t was 27,630 ± 5141.16ng·h/mL and 27,489.58 ± 5161.34ng·h/mL, while the average value plus or minus the standard deviation (SD) of AUC_0-∞ was75972.38 ± 52335.11ng·h/mL and 57,622.87 ± 20980.43ng·h/mL, respectively. The median and range of T_max were 2.50(4.50–3.50) hours and 5.00 (2.50–10.00) hours, respectively. The arithmetic meaned plus or minus the standard deviation of the maximum concentration (C_max) were 1929.68 ± 412.83 ng/mL and 1300.17 ± 250.39 ng/mL, respectively.

Figure 3.

Subjects mean (SD) blood concentration-time curve (fasting and after meal).

Table 2.

Pharmacokinetic parameters of isavuconazole.

Parameters	Fasting (Mean ± SD)	After meal (Mean ± SD)
Tmax* (h)	2.50(1.50,3.50)	5.00(2.50,10.00)
Cmax (ng/mL)	1929.68 ± 412.83(21.39)	1300.17 ± 250.39(19.26)
AUC0-t (ng·h/mL)	27630.54 ± 5141.16(18.61)	27489.58 ± 5161.34(18.78)
AUC0-∞ (ng·h/mL)	75972.38 ± 52335.11(68.89)	57622.87 ± 20980.43(36.41)
λz (h−1)	0.01 ± 0.00(59.05)	0.01 ± 0.01(56.92)
t1/2 (h)	126.61 ± 96.36(76.11)	83.38 ± 48.76(58.48)
AUC_%Extrap(%)	55.12 ± 16.74(30.37)	48.19 ± 14.42(29.92)

4. Discussion

The validated HPLC-MS/MS method for quantifying isavuconazole in human plasma demonstrates excellent analytical performance while addressing key limitations of existing approaches. Our method achieves chromatographic separation within 1 minute, significantly faster than conventional LC-MS/MS methods that typically require 3–10 minutes per run [20–22]. This rapid analysis was enabled by an optimized gradient elution program (0.01% formic acid/acetonitrile) on a C18 column (2.1 × 50 mm, 5.0 μm), which reduced run time without compromising resolution. The method’s sensitivity (LLOQ: 4 ng/mL) meets or exceeds that reported in previous studies (LLOQ range: 5–50 ng/mL) [21,22].

Compared to alternative sample preparation techniques, our protein precipitation approach with acetonitrile provided > 120% recovery with minimal matrix effects ( < 7.2% interference), matching or surpassing the recovery rates (80–110%) reported for more complex extraction methods like solid-phase extraction [24]. The method demonstrated excellent specificity, with negligible interference from hemolyzed and hyperlipidemic plasma – a notable advantage over some published methods that require additional cleanup steps for lipid-rich samples [20].

The stability profile of our method offers practical advantages for clinical implementation. Long-term stability (53 days at −20°C, 66 days at −80°C) was consistent with some reports [24] but exceeded others (e.g., 30-day stability [22]). The extended autosampler stability (191 hours) represents a significant improvement over methods requiring immediate analysis post-preparation [25], facilitating more flexible laboratory workflows.

Our pharmacokinetic findings in healthy Chinese subjects both confirm and extend previous observations. The food effects we observed – delayed T_max (5.0 vs 2.5 h) and reduced C_max (1300 vs 1929 ng/mL) under fed conditions – align with trends reported in Western cohorts [21]. However, the equivalent AUC_0-t values between fasting and fed states in our study contrast with some reports of 20–30% food-induced AUC reductions, potentially suggesting ethnic or formulation differences. The substantial interindividual variability in AUC_0-∞ (CV: 36–69%) reinforces the importance of therapeutic drug monitoring, particularly in critically ill patients [24].

The method’s precision (RSD ≤12.6%) and accuracy (±15%) meet ICH M10 guidelines (2022) and compare favorably with multi-azole assays that often sacrifice precision for broader analyte coverage [20–22]. The use of an isotopic internal standard (isavuconazole-d4) provided superior correction for variability compared to methods employing structural analogs [22], enhancing the reliability of results for TDM applications in special populations such as patients with hepatic impairment.

5. Conclusion

In this study, we established and validated an ultra-performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS) method that successfully achieved the detection of isavuconazole in human plasma within 1 minute. Compared with conventional liquid chromatography methods, this approach demonstrated significant improvements in both detection speed and sensitivity. The method exhibited satisfactory performance in selectivity, extraction recovery, precision, and accuracy, all meeting the required analytical criteria. This validated method has been successfully applied to the pharmacokinetic (PK) study of isavuconazole, systematically evaluating key pharmacokinetic parameters of isavuconazole sulfate in plasma samples from healthy Chinese subjects. The research outcomes integrate comprehensive data from both fasting and postprandial conditions. Furthermore, this methodology not only supports bioequivalence studies but also facilitates TDM-guided dose optimization in clinical practice. By ensuring plasma concentrations remain within the therapeutic range (2–5 μg/mL), our method maximizes isavuconazole’s efficacy while mitigating toxicity risks. This approach is particularly valuable for vulnerable populations, such as critically ill or metabolically compromised patients, where fixed dosing often fails to achieve target exposures.

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Funding Statement

This paper was not funded.

Article highlights

• Developed the faster HPLC-MS/MS assay for isavuconazole (1-min chromatographic runtime).

• Achieved unprecedented sensitivity (LLOQ: 4 ng/mL) covering clinical range (2–5 μg/mL).

• Validated simplified sample processing (acetonitrile protein precipitation) with minimal matrix effects ( <7.2%).

Author contributions

Linlin Qiu, Jie Lin: Responsible for conducting the entire study, reviewing data and manuscripts.

Yuchen Su: Responsible for manuscript writing

Yifu Kong: Responsible for data analysis

Yanli Zhou: Responsible for the execution of research

Yifang Chen: Responsible for the execution of research

Yonghua Yu: Project Management

Disclosure statement

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Writing assistance

No writing assistance was utilized in the production of this manuscript.

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Ethical approval

Ethical approval for this study involving human plasma samples was obtained from the Medical Ethics Committee of Chongqing 11th People’s Hospital. The approval number is 2024 - 07 - 001. The research adhered to the principles of the Declaration of Helsinki and all relevant ethical guidelines.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/17576180.2025.2548195