Population Pharmacokinetics of Voriconazole in Patients With Invasive Aspergillosis: Serum Albumin Level as a Novel Marker for Clearance and Dosage Optimization

Therapeutic drug monitoring of azole antifungal agents

Deep-seated mycoses are generally opportunistic infections that are difficult to diagnose and treat. They are expected to increase with the spread of advanced medical care and aging populations, thus highlighting the need for safe, effective, and rapid drug-based treatments. Depending on a patient's age, sex, underlying diseases, and immune system status, therapeutic drug monitoring (TDM) may be important for assessing variable pharmacokinetic parameters, as well as preventing drug-drug interactions, adverse events, and breakthrough infections caused by fungal resistance. Azole antifungal agents play an important role in the prevention and treatment of deep-seated fungal infections, with each azoles having its own unique pharmacokinetic properties and specific adverse events. Therefore, it is necessary to use national and international guidelines to build evidence for the expansion of TDM indications. This review focuses on the clinical utility and future perspectives of TDM using azole antifungal agents, in the context of recent evidence in the literature.

Association of procalcitonin with voriconazole concentrations: a retrospective cohort study

Inflammation is a potential risk factor of voriconazole (VCZ) overdose, procalcitonin (PCT) is reported to act as a diagnostic marker for bacterial infections. However, the association of PCT with VCZ trough serum concentrations (VCZ-Cmin) is not fully clear. Our study aims to investigate the associations between PCT and VCZ-Cmin. In this retrospective cohort study, we collected the clinical data of 147 patients who received VCZ and monitored the VCZ concentration of them in our hospital from August 2017 to August 2021. All patients underwent routine clinical examinations on the day or the day before VCZ administration. General information and clinical symptoms of these patients were recorded. Multivariate liner analysis showed that PCT was significantly associated with VCZ-Cmin (p < 0.001). Overall, it was shown that VCZ-Cmin was significantly increased by 0.32 µg/mL for each fold increment in PCT in crude model. In the minor adjusted model (Model 1, adjustment for sex, age, albumin, direct bi1irubin, WBC) and fully adjusted model (Model 2, adjustment for sex, age, albumin, direct bilirubin, WBC, AST and ALT), VCZ-Cmin was significantly increased by 0.23 µg/mL and 0.21 µg/mL, respectively, for each fold increment in PCT. In conclusion, this research reveals the correlation between PCT and VCZ-Cmin, indicating that PCT has the potential to serve as a valuable biomarker for drug monitoring in the treatment of VCZ.

Associated factors with voriconazole plasma concentration: a systematic review and meta-analysis

Background: Voriconazole plasma concentration exhibits significant variability and maintaining it within the therapeutic range is the key to enhancing its efficacy. We conducted a systematic review and meta-analysis to estimate the prevalence of patients achieving the therapeutic range of plasma voriconazole concentration and identify associated factors. Methods: Eligible studies were identified through the PubMed, Embase, Cochrane Library, and Web of Science databases from their inception until 18 November 2023. We conducted a meta-analysis using a random-effects model to determine the prevalence of patients who reached the therapeutic plasma voriconazole concentration range. Factors associated with plasma voriconazole concentration were summarized from the included studies. Results: Of the 60 eligible studies, 52 reported the prevalence of patients reaching the therapeutic range, while 20 performed multiple linear regression analyses. The pooled prevalence who achieved the therapeutic range was 56% (95% CI: 50%-63%) in studies without dose adjustment patients. The pooled prevalence of adult patients was 61% (95% CI: 56%-65%), and the pooled prevalence of children patients was 55% (95% CI: 50%-60%) The study identified, in the children population, several factors associated with plasma voriconazole concentration, including age (coefficient 0.08, 95% CI: 0.01 to 0.14), albumin (-0.05 95% CI: -0.09 to -0.01), in the adult population, some factors related to voriconazole plasma concentration, including omeprazole (1.37, 95% CI 0.82 to 1.92), pantoprazole (1.11, 95% CI: 0.17-2.04), methylprednisolone (-1.75, 95% CI: -2.21 to -1.30), and dexamethasone (-1.45, 95% CI: -2.07 to -0.83). Conclusion: The analysis revealed that only approximately half of the patients reached the plasma voriconazole concentration therapeutic range without dose adjustments and the pooled prevalence of adult patients reaching the therapeutic range is higher than that of children. Therapeutic drug monitoring is crucial in the administration of voriconazole, especially in the children population. Particular attention may be paid to age, albumin levels in children, and the use of omeprazole, pantoprazole, dexamethasone and methylprednisolone in adults. Systematic Review Registration: https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42023483728.

Development and validation of a clinical risk score nomogram for predicting voriconazole trough concentration above 5 mg/L: a retrospective cohort study

The therapeutic range of voriconazole (VRC) is narrow, this study aimed to explore factors influencing VRC plasma concentrations > 5 mg/L and to construct a clinical risk score nomogram prediction model. Clinical data from 221 patients with VRC prophylaxis and treatment were retrospectively analyzed. The patients were randomly divided into a training cohort and a validation cohort at a 7:3 ratio. Univariate and binary logistic regression analysis was used to select independent risk factors for VRC plasma concentration above the high limit (5 mg/L). Four indicators including age, weight, CYP2C19 genotype, and albumin were selected to construct the nomogram prediction model. The area under the curve values of the training cohort and the validation cohort were 0.841 and 0.802, respectively. The decision curve analysis suggests that the nomogram model had good clinical applicability. In conclusion, the nomogram provides a reference for early screening and intervention in a high-risk population.

Effectiveness and safety of the simulation-based first-dose design of voriconazole

BACKGROUND

We investigated whether the initial voriconazole (VRCZ) dosing design, as determined using simulation software with a population pharmacokinetic model of Japanese patients, impacts the effectiveness and safety when compared with VRCZ initiation according to the package insert.

METHODS

In this single-center retrospective observational study, we employed records from Tosei General Hospital (a 633-bed hospital), dated April 2017 to September 2023. Eligible patients were divided into the software-based simulation group, comprising patients administered initial VRCZ dosage adjustment by pharmacists using software-based simulation, and the standard therapy group, whose dosage was administered by a physician following the package insert recommendations without simulation. The primary objective of this study was to determine the efficacy of VRCZ first-dose design in reducing the incidence of hepatotoxicity and visual symptoms.

RESULTS

The median ages of enrolled participants (n = 93) were 75 (68-79) and 72 (65-78) years in the software-based simulation and standard therapy groups, respectively. Regardless of formulation, initial trough concentrations were lower in the VRCZ software-based first dosage adjustment group and higher rate within the appropriate range (1-4 μg/mL). The incidence of all-grade hepatotoxicity or visual symptoms was significantly lower in the software-based simulation group. The log-rank test revealed a significant impact on the occurrence of ≥grade 2 hepatotoxicity in the software-based first dosage adjustment group compared to that in the standard therapy group.

CONCLUSIONS

The initial VRCZ dosing design using simulation software improved the achievement of appropriate initial trough concentrations and resulted in fewer occurrences of hepatotoxicity (≥grade 2) when compared with the standard therapy.

Combined impact of hypoalbuminemia and pharmacogenomic variants on voriconazole trough concentration: data from a real-life clinical setting in the Chinese population

Voriconazole (VRC) displays highly variable pharmacokinetics impacting treatment efficacy and safety. To provide evidence for optimizing VRC therapy regimens, the authors set out to determine the factors impacting VRC steady-state trough concentration (Cmin) in patients with various albumin (Alb) level. A total of 275 blood samples of 120 patients and their clinical characteristics and genotypes of CYP2C19, CYP3A4, CYP3A5, CYP2C9, FMO3, ABCB1, POR, NR1I2 and NR1I3 were included in this study. Results of multivariate linear regression analysis demonstrated that C-reactive protein (CRP) and total bilirubin (T-Bil) were predictors of the VRC Cmin adjusted for dose in patients with hypoalbuminemia (Alb < 35 g/L) (R2 = 0.16, P < 0.001). Additionally, in patients with normal albumin level (Alb ≥ 35 g/L), it resulted in a significant model containing factors of the poor metabolizer (PM) CYP2C19 genotype and CRP level (R2 = 0.26, P < 0.001). Therefore, CRP and T-Bil levels ought to receive greater consideration than genetic factors in patients with hypoalbuminemia.

Population Pharmacokinetics of Voriconazole and Dose Optimization in Elderly Chinese Patients

Voriconazole is commonly recommended as a first-line therapy for invasive aspergillosis infections. Elderly patients are susceptible to infectious diseases owing to their decreased physical function and immune system. Our study aims to establish a population pharmacokinetics model for elderly patients receiving intravenous voriconazole, and to optimize dosing protocols through a simulated approach. An accurate fit to the concentration-time profile of voriconazole was achieved by employing a 1-compartment model featuring first-order elimination. The typical clearance rate of voriconazole was found to be 3.22 L/h, with a typical volume of distribution of 194 L. The covariate analysis revealed that albumin (ALB), gamma-glutamyl transpeptidase, and direct bilirubin had significant impacts on voriconazole clearance. Additionally, body weight was found to be associated with the volume of distribution. Individualized dosing regimens were recommended for different ALB levels based on population pharmacokinetics model prediction. The proposed dosing regimens could provide a rationale for dosage individualization, improve the clinical outcomes, and minimize drug-related toxicities.

Individualized treatment with voriconazole in the Chinese population: Inflammation level as a novel marker for dose optimization

AIMS

The aim of this study was to explore the influence and possible mechanisms of pharmacokinetics-related gene polymorphisms, especially CYP2C19 polymorphisms, and non-genetic factors combined with the inflammatory status on the voriconazole (VRC) metabolism of the Chinese population.

METHODS

Clinical studies were performed by collecting more than one VRC trough concentration and C-reactive protein (CRP) level. A total of 265 blood samples were collected from 120 patients.

RESULTS

Results of multiple regression analyses demonstrated that CYP2C19 genotypes and albumin (Alb) level remained predictors of Cmin ss/D in patients with no to mild inflammation (R2  = 0.12, P < .001). In addition, in patients with moderate to severe inflammation, it resulted in a significant model containing factors of CRP and total bilirubin (T-Bil) levels (R2  = 0.19, P < .001). In non-clinical studies, 32 rats were divided into control and inflammatory groups, and it was found that the mean residence time (MRT(0-t) ) of VRC in the inflammatory group was significantly longer than that in the control group (P < .001), which may be due to down-regulation of mRNA and protein expression of CYP2C19 (CYP2C6 in rats) through interleukin (IL)-6/signal transducer and activator of transcription (STAT) 3 pathway.

CONCLUSIONS

Therefore, the effect of CYP2C19 polymorphisms on VRC metabolism may be masked by inflammatory status, which should be of more concern than CYP2C19 polymorphisms in patients with moderate to severe inflammation. Additionally, the impact of Alb and T-Bil on VRC metabolism should not be disregarded.

Prediction of plasma trough concentration of voriconazole in adult patients using machine learning

OBJECTIVE

Plasma trough concentration of voriconazole (VCZ) was associated with its toxicity and efficacy. However, the nonlinear pharmacokinetic characteristics of VCZ make it difficult to determine the relationship between clinical characteristics and its concentration. We intended to present a machine learning (ML)-based method to predict toxic plasma trough concentration of VCZ (>5 μg/mL).

METHODS

A single center retrospective study was conducted. Three ML algorithms were used to estimate the concentration in adult patients, including random forest (RF), gradient boosting (GB), and extreme gradient boosting (XGBoost). The importance of variables was recognized by the SHapley Additive exPlanations (SHAP) method. In addition, an external validation set was used to validate the robustness of models.

RESULTS

A total of 1318 VCZ plasma concentration were included, with 33 variables enrolled in the model. Nine classification models were developed using the RF, GB, and XGBoost algorithms. Most models performed well for both the training set and test set, with an average balanced accuracy (BA) of 0.704 and an average accuracy (ACC) of 0.788. In addition, the average Matthews correlation coefficient value reached 0.484, which indicated the predicted values are meaningful. Based on the average BA and ACC values, the predictive ability of the models can be ranked from best to worst as follows: younger adult models > mixed models > elderly models, and XGBoost models > GBT models > RF models. The SHAP results showed that the top five influencing factors in younger adult patients (<60 years) were albumin, total bile acid (TBA), platelets count, age, and inflammation, while the top five influencing factors in elderly patients were albumin, TBA, aspartate aminotransferase, creatinine, and alanine aminotransferase. Furthermore, the prediction of external validation set for VCZ concentrations verified the high reliability of the models, for the ACC value of 0.822 by the best model.

CONCLUSIONS

The ML models can be reliable tools for predicting toxic concentration exposure of VCZ. The SHAP results may provide useful guidelines for dosage adjustment of VCZ.

Voriconazole Pharmacokinetics in Critically Ill Patients and Extracorporeal Membrane Oxygenation Support: A Retrospective Comparative Case-Control Study

Voriconazole, an antifungal agent, displays high intra- and inter-individual variability. The predictive pharmacokinetic (PK) index requires a minimum plasma concentration (C_min) in patient serum of between 1-5.5 mg/L. It is common to encounter fungal infections in patients undergoing extracorporeal membrane oxygenation (ECMO) support, and data regarding voriconazole PK changes during ECMO are scarce. Our study compared voriconazole PKs in patients with and without ECMO support in a retrospective cohort of critically-ill patients. Fifteen patients with 26 voriconazole C_min determinations in the non-ECMO group and nine patients with 27 voriconazole C_min determinations in the ECMO group were recruited. The ECMO group had lower C_min (0.38 ± 2.98 vs. 3.62 ± 3.88, p < 0.001) and higher infratherapeutic C_min values (16 vs. 1, p < 0.001) than the non-ECMO group. Multivariate analysis identified ECMO support (-0.668, CI₉₅ -0.978--0.358) and plasma albumin levels (-0.023, CI₉₅ -0.046--0.001) as risk factors for low C_min values. When comparing pre- and post-therapeutic drug optimisation samples from the ECMO group, the dose required to achieve therapeutic C_min was 6.44 mg/kg twice a day. Therapeutic drug optimisation is essential to improve target attainment.

Factors influencing plasma concentration of voriconazole and voriconazole- N-oxide in younger adult and elderly patients

Background: Voriconazole (VCZ) metabolism is influenced by many factors. Identifying independent influencing factors helps optimize VCZ dosing regimens and maintain its trough concentration (C₀) in the therapeutic window. Methods: We conducted a prospective study investigating independent factors influencing VCZ C₀ and the VCZ C₀ to VCZ N-oxide concentration ratio (C₀/C_N) in younger adults and elderly patients. A stepwise multivariate linear regression model, including the IL-6 inflammatory marker, was used. The receiver operating characteristic (ROC) curve analysis was used to evaluate the predictive effect of the indicator. Results: A total of 463 VCZ C₀ were analyzed from 304 patients. In younger adult patients, the independent factors that influenced VCZ C₀ were the levels of total bile acid (TBA) and glutamic-pyruvic transaminase (ALT) and the use of proton-pump inhibitors. The independent factors influencing VCZ C₀/C_N were IL-6, age, direct bilirubin, and TBA. The TBA level was positively associated with VCZ C₀ (ρ = 0.176, p = 0.019). VCZ C₀ increased significantly when the TBA levels were higher than 10 μmol/L (p = 0.027). ROC curve analysis indicated that when the TBA level ≥4.05 μmol/L, the incidence of a VCZ C₀ greater than 5 μg/ml (95% CI = 0.54-0.74) (p = 0.007) increased. In elderly patients, the influencing factors of VCZ C₀ were DBIL, albumin, and estimated glomerular filtration rate (eGFR). The independent factors that affected VCZ C₀/C_N were eGFR, ALT, γ-glutamyl transferase, TBA, and platelet count. TBA levels showed a positive association with VCZ C₀ (ρ = 0.204, p = 0.006) and C₀/C_N (ρ = 0.342, p < 0.001). VCZ C₀/C_N increased significantly when TBA levels were greater than 10 μmol/L (p = 0.025). ROC curve analysis indicated that when the TBA level ≥14.55 μmol/L, the incidence of a VCZ C₀ greater than 5 μg/ml (95% CI = 0.52-0.71) (p = 0.048) increased. Conclusion: TBA level may serve as a novel marker for VCZ metabolism. eGFR and platelet count should also be considered when using VCZ, especially in elderly patients.

Voriconazole exposure is influenced by inflammation: A population pharmacokinetic model

BACKGROUND

Voriconazole is an antifungal drug used for the treatment of invasive fungal infections. Due to highly variable drug exposure, therapeutic drug monitoring (TDM) has been recommended. TDM may be helpful to predict exposure accurately, but covariates, such as severe inflammation, that influence the metabolism of voriconazole have not been included in the population pharmacokinetic (popPK) models suitable for routine TDM.

OBJECTIVES

To investigate whether the effect of inflammation, reflected by C-reactive protein (CRP), could improve a popPK model that can be applied in clinical care.

PATIENTS AND METHODS

Data from two previous studies were included in the popPK modelling. PopPK modelling was performed using Edsim++. Different popPK models were compared using Akaike Information Criterion and goodness-of-fit plots.

RESULTS

In total, 1060 voriconazole serum concentrations from 54 patients were included in this study. The final model was a one-compartment model with non-linear elimination. Only CRP was a significant covariate, and was included in the final model and found to affect the maximum rate of enzyme activity (V_max). For the final popPK model, the mean volume of distribution was 145 L [coefficient of variation percentage (CV%)=61%], mean Michaelis-Menten constant was 5.7 mg/L (CV%=119%), mean V_max was 86.4 mg/h (CV%=99%) and mean bioavailability was 0.83 (CV%=143%). Internal validation using bootstrapping resulted in median values close to the population parameter estimates.

CONCLUSIONS

This one-compartment model with non-linear elimination and CRP as a covariate described the pharmacokinetics of voriconazole adequately.

Dosage optimization of voriconazole in children with haematological malignancies based on population pharmacokinetics

WHAT IS KNOWN AND OBJECTIVES

Voriconazole has a complex pharmacokinetic profile and exhibits different pharmacokinetic characteristics in adults and children. Nevertheless, few studies have been conducted on the population pharmacokinetics (PPK) of voriconazole in children with haematological malignancies. This study aims to build a PPK model and propose a suitable voriconazole treatment scheme for children with haematological malignancies.

METHODS

We retrospectively collected 146 samples from 67 children aged from 1.08 to 17.92 years. The PPK model was established using nonlinear mixed effects modelling (NONMEM). Dosage simulations were conducted on the basis of the final model's covariates.

RESULTS AND DISCUSSION

Data were fully characterized by a one-compartment model with first-order absorption and elimination. The weight (WT), CYP2C19 phenotype, and Albumin (ALB) were notable covariates for clearance (CL). The typical values of CL, the volume of distribution (V), and oral bioavailability (F) were 2.29 L/h, 76 L, and 0.902, respectively. The proposed doses for different CYP2C19 genotypes were presented in this ranking: EM (extensive metabolizer) > IM (intermediate metabolizer) > PM (poor metabolizer). Furthermore, higher dosages for light WT patients were recommended while lower ALB levels required lower doses. The probability of achieving the target (PTA) for the recommended doses ranged from 72.2% to 99%.

WHAT IS NEW AND CONCLUSION

We successfully built a voriconazole PPK model for children with hematologic malignancies. Dosing regimens were developed for different patients based on the final model, which could enhance the rational use of voriconazole in children with haematological malignancies.

Population pharmacokinetics of voriconazole and initial dosage optimization in patients with talaromycosis

The high variability and unpredictability of the plasma concentration of voriconazole (VRC) pose a major challenge for clinical administration. The aim of this study was to develop a population pharmacokinetics (PPK) model of VRC and identify the factors influencing VRC PPK in patients with talaromycosis. Medical records and VRC medication history of patients with talaromycosis who were treated with VRC as initial therapy were collected. A total of 233 blood samples from 69 patients were included in the study. A PPK model was developed using the nonlinear mixed-effects models (NONMEM). Monte Carlo simulation was applied to optimize the initial dosage regimens with a therapeutic range of 1.0–5.5 mg/L as the target plasma trough concentration. A one-compartment model with first-order absorption and elimination adequately described the data. The typical voriconazole clearance was 4.34 L/h, the volume of distribution was 97.4 L, the absorption rate constant was set at 1.1 h^-1, and the bioavailability was 95.1%. Clearance was found to be significantly associated with C-reactive protein (CRP). CYP2C19 polymorphisms had no effect on voriconazole pharmacokinetic parameters. ‏Monte Carlo simulation based on CRP levels showed that a loading dose of 250 mg/12 h and a maintenance dose of 100 mg/12 h are recommended for patients with CRP ≤ 96 mg/L, whereas a loading dose of 200 mg/12 h and a maintenance dose of 75 mg/12 h are recommended for patients with CRP > 96 mg/L. The average probability of target attainment of the optimal dosage regimen in CRP ≤ 96 mg/L and CRP > 96 mg/L groups were 61.3% and 13.6% higher than with empirical medication, and the proportion of C_min > 5.5 mg/L decreased by 28.9%. In conclusion, the VRC PPK model for talaromycosis patients shows good robustness and predictive performance, which can provide a reference for the clinical individualization of VRC. Adjusting initial dosage regimens based on CRP may promote the rational use of VRC.

Systematic Evaluation of Voriconazole Pharmacokinetic Models without Pharmacogenetic Information for Bayesian Forecasting in Critically Ill Patients

Voriconazole (VRC) is used as first line antifungal agent against invasive aspergillosis. Model-based approaches might optimize VRC therapy. This study aimed to investigate the predictive performance of pharmacokinetic models of VRC without pharmacogenetic information for their suitability for model-informed precision dosing. Seven PopPK models were selected from a systematic literature review. A total of 66 measured VRC plasma concentrations from 33 critically ill patients was employed for analysis. The second measurement per patient was used to calculate relative Bias (rBias), mean error (ME), relative root mean squared error (rRMSE) and mean absolute error (MAE) (i) only based on patient characteristics and dosing history (a priori) and (ii) integrating the first measured concentration to predict the second concentration (Bayesian forecasting). The a priori rBias/ME and rRMSE/MAE varied substantially between the models, ranging from −15.4 to 124.6%/−0.70 to 8.01 mg/L and from 89.3 to 139.1%/1.45 to 8.11 mg/L, respectively. The integration of the first TDM sample improved the predictive performance of all models, with the model by Chen (85.0%) showing the best predictive performance (rRMSE: 85.0%; rBias: 4.0%). Our study revealed a certain degree of imprecision for all investigated models, so their sole use is not recommendable. Models with a higher performance would be necessary for clinical use.

Inflammation Affects Liver Function and the Metabolism of Voriconazole to Voriconazole-N-Oxide in Adult and Elderly Patients

Background: The inner association of inflammation with voriconazole (VCZ) metabolism has not been fully investigated. We intend to investigate the effects of inflammation on liver function, VCZ trough concentration (C₀), C₀/dose ratio and the ratio of VCZ to VCZ-N-oxide concentration (C₀/C_N) in adult and elderly patients.

Methods: A single-center retrospective study was conducted among patients who were treated in our hospital between January 2018 and December 2021. For each eligible patient, demographic details, medical history, laboratory parameters, procalcitonin (PCT), C reactive protein (CRP), and interleukin-6 (IL-6) were collected from the medical chart. VCZ C_N, TNF-α, IL-1β, IL-8, and IL-10 concentrations were detected in blood samples.

Results: A total of 356 patients were included in our study, with 195 patients in the adult cohort (<60 years) and 161 patients in the elderly cohort (≥60 years). In adult patients, CRP and IL-8 levels showed moderate association with VCZ C₀/C_N ratio (CRP: r = 0.512, p < 0.001; IL-8: r = 0.476, p = 0.002). IL-6 level shallowly associated with VCZ C₀/C_N ratio both in adult and elderly patients (r = 0.355, p = 0.003; r = 0.386, p = 0.001). A significantly higher VCZ C₀, C₀/dose ratio and C₀/C_N ratio was observed in adult patients with severe inflammation compared with patients with moderate inflammation and no to mild inflammation, as reflected by PCT levels (p < 0.05). However, there was no significant difference observed among different inflammation degrees in elderly patients. Lower albumin (AL) and higher total bilirubin (TBIL) were observed along with the degree of inflammation in both adult and elderly patients, as reflected by CRP and PCT levels (p < 0.05).

Conclusion: Inflammation may affect the metabolism of VCZ to VCZ-N-oxide both in adult and elderly patients, and decreased plasma AL levels and increased TBIL levels under inflammatory conditions may also alter VCZ metabolism.

Predictive Value of C-Reactive Protein and Albumin for Temporal Within-Individual Pharmacokinetic Variability of Voriconazole in Pediatric Hematopoietic Cell Transplant Patients

Voriconazole is a widely used antifungal agent in immunocompromised patients, but its utility is limited by its variable exposure and narrow therapeutic index. Population pharmacokinetic (PK) models have been used to characterize voriconazole PK and derive individualized dosing regimens. However, determinants of temporal within-patient variability of voriconazole PK were not well-established. We aimed to characterize temporal variability of voriconazole PK within individuals and identify predictive clinical factors. This study was conducted as a part of a single-institution, phase I study of intravenous voriconazole in children undergoing HCT (NCT02227797). We analyzed voriconazole PK study data collected at week 1 and again at week 2 after the start of voriconazole therapy in 59 pediatric HCT patients (age <21 years). Population PK analysis using nonlinear mixed effect modeling was performed to analyze temporal within-individual variability of voriconazole PK by incorporating a between-occasion variability term in the model. A two-compartment linear elimination model incorporating body weight and CYP2C19 phenotype described the data. Ratio of individual voriconazole clearance between weeks 1 to 2 ranged from 0.11 to 3.3 (-9.1 to +3.3-fold change). Incorporation of covariate effects by serum C-reactive protein (CRP) and albumin levels decreased between-occasion variability of clearance (coefficient of variation: from 59.5% to 41.2%) and improved the model fit (p<0.05). As significant covariates on voriconazole PK, CRP and albumin concentrations may potentially serve as useful biomarkers as part of therapeutic drug monitoring. This article is protected by copyright. All rights reserved.