European Journal of Drug Metabolism and Pharmacokinetics最新文献

Background and objective: Tramadol is a widely used opioid, and pharmacogenetics is partially responsible for the variability in its response. The objective was to develop a tramadol population pharmacokinetic (popPK) model including pharmacogenetic information.

Methods: A single oral dose of 37.5 mg tramadol + 400 mg ibuprofen arginine was administered to 24 European healthy volunteers, and 18 blood samples were obtained from 0.25 h to 24 h after drug intake. Subjects were genotyped for the main pharmacogenes and a popPK model was built using NONMEM^® (version 7.3). The role of CYP2D6, CYP2B6, and CYP3A4 phenotypes was analyzed. The final parameter estimates were compared with results obtained by noncompartmental analysis (NCA) in the same cohort. Simulations depending on CYP2D6 phenotype were performed in single dose and steady state conditions.

Results: A two-compartment model, with transit absorption into the depot compartment and first-order elimination, was the best fit to the data. In total, 8 volunteers were CYP2D6 intermediate metabolizers (IMs) and 14 were normal metabolizers (NMs), merged for the analysis with the two ultrarapid metabolizers (UMs), whereas no poor metabolizers (PMs) were present (European frequencies: 38.3%, 49.2%, 2.3%, and 6.5%, respectively). CYP2D6 phenotype affected clearance, which was 20% reduced in IMs compared with NMs + UMs. CYP2B6 (NMs = 11, IMs = 11, and PMs = 2) and CYP3A4 (NMs = 20, IMs = 3) phenotypes did not affect clearance. PopPK and NCA estimates were in close agreement. Simulations indicated a 20% and even 40% higher area under the curve after single dose and steady state conditions, respectively, in CYP2D6 IMs compared with NMs and UMs.

Conclusions: A popPK model including CYP2D6 phenotype well described the data. Further research with increased sample size is needed to analyze the clinical impact and effect of CYP2B6 and CYP3A4 phenotype on tramadol pharmacokinetics.

Background and objective: Infections in the central nervous system (CNS) are serious and carry a significant risk of morbidity and mortality. Though commonly used as prophylaxis for neurosurgical interventions, cefazolin as a treatment for CNS infections due to methicillin-susceptible Staphylococcus aureus (MSSA) has been debated owing to the perceived inability to achieve adequate concentrations at the site of infection. The objective of the current study was to evaluate the dose-exposure-response relationship of cefazolin in the CNS.

Methods: To leverage sparse data of cefazolin in the cerebrospinal fluid (CSF) and derive an understanding of the dose-exposure-response profile in the CNS, a physiologically-based pharmacokinetic (PBPK) model was created in PK-Sim. Simulations were performed using standard cefazolin dosing of 2000 mg every 8 h and alternative regimens to maximize the probability of target attainment (PTA). The pharmacodynamic target used was 100% fT > MIC (100% of free drug concentrations above the minimum inhibitory concentration). Furthermore, a neurotoxicity threshold of ≥ 300 mg/L and ≥ 30 mg/L for trough concentrations was set as the safety indicator in plasma and CSF, respectively.

Results: The cefazolin CSF-PBPK model was successfully validated such that predicted CSF:plasma ratios were within a 1.5-fold error compared with the observed values. In addition, the median predicted CSF:epidemiological cut-off (ECOFF) concentration ratio was 2.52, compared with an observed value of 2.8. In silico simulations demonstrate that intermittent doses of 2000 mg every 6 h or a continuous infusion of 8-10 g/day may be required to ensure 90% PTA for MSSA to a MIC ≤ 2 mg/L. Predicted plasma and CSF concentrations were well below concentrations associated with neurotoxicity.

Conclusions: This study is the first to use sparse observed CNS data to develop a mechanistic model to describe the pharmacokinetics of cefazolin in the CSF. This work supports existing research on the viability of cefazolin as a therapeutic alternative for CNS infections attributed to MSSA and can be used for future clinical trial planning.

Background: This study investigated the effects of temperature and concentration on the short-term stability of colistin and colistin methanesulphonate in three biological matrices; whole blood, plasma and serum.

Methods: Human whole blood, plasma or serum containing colistin (2; 10; 20 mg/L) or colistin methanesulphonate (2; 10; 20 mg/L) was stored at room temperatrure or at 37 °C for a predetermined time interval. The concentrations of colistin in colistin spiked samples, and formed colistin in colistin methanesulphonate spiked samples, were analysed by LC-MS method.

Results: Findings indicate that colistin A and B were stable at the room temperature, but degradation accelerated at 37 °C, particularly at lower concentrations. The conversion of CMS to colistin was also temperature-dependent, where the concersion was significantly accelerated at elevated temperatures. Freeze-thaw stability were tested, COL A as well as COL B were stable for at least three freeze-thaw cycles.

Conclusion: These results underscore the importance of rapid sample processing to ensure reliable therapeutic drug monitoring and emphasize the importance of considering temperature and concentration factors in clinical practice to optimise dosing and minimise toxicity. Further research is required to investigate the full range of factors affecting the stability of colistin and colistin methanesulphonate in biological matrices.

Background and objective: Therapeutic drug monitoring (TDM) typically uses total drug concentration (C_t), but pharmacological effects depend on free concentration (C_f), especially for highly protein-bound drugs like rifapentine (RFPT) (96-99% bound). Monitoring C_f is critical for optimizing efficacy and minimizing hepatotoxicity in patients with tuberculosis (TB) with individual variability. Addressing limitations of existing assays, this study developed a sensitive liquid chromatography-tandem mass spectrometry (LC-MS/MS) method for quantifying free RFPT and evaluated C_f, C_t, and clinical factor correlations in adults and children.

Methods: A validated LC-MS/MS method used Centrifree ultrafiltration (CF-UF) for free RFPT isolation and isotope internal standard (rifapentine-d8) for quantification. Validation included specificity, linearity (3.00-299.60 ng/mL), accuracy, precision, matrix effects, and stability. Clinical samples from 58 patients with TB (adults and children) receiving RFPT were analyzed. Total and free RFPT, albumin, and biochemical parameters were compared.

Results: The method showed excellent linearity (R² = 0.9999), accuracy (93.11-102.67%), and precision (intra-/inter-day RSD ≤ 7.40%). The lower limit of quantification (LLOQ) was 3.00 ng/mL, suitable for clinical C_f detection. C_f correlated nonlinearly with C_t. C_t was significantly higher in adults than children (20.91 ± 14.08 versus 14.64 ± 8.47 μg/mL, P = 0.03), but C_f (0.085 ± 0.09 versus 0.054 ± 0.05 μg/mL, P = 0.06) and free fractions (0.38 ± 0.29% versus 0.35 ± 0.20%, P = 0.65) showed no significant difference.

Conclusions: The LC-MS/MS method is rapid, sensitive, and suitable for routine free RFPT TDM. Nonlinear C_t-C_f relationships highlight the necessity of direct C_f monitoring, particularly with altered protein binding. Similar free fractions in adults and children despite dose-related C_t differences suggest tailored dosing may mitigate toxicity. Adjustments based solely on C_t may not be universally applicable.

Background and objective: Obtaining pharmacokinetic curves is crucial for drug dosage selection, and for assessment of efficacy and adverse effects in clinical practice. The aim of this study was to utilize a minimal physiologically-based pharmacokinetic model to predict concentration-time profiles of cinacalcet hydrochloride, a poorly soluble drug, under fasting and fed conditions, and further to establish correlations between in vivo and in vitro dissolution profiles of cinacalcet hydrochloride.

Methods: The mPBPK model consisted of six tissue compartments, and an additional compartmental absorption and transit model, incorporating the stomach, seven small intestinal compartments, and the large intestine along with Johnson's equation, was integrated to enable precise calculations of in vivo dissolution.

Results: External validations for three dosages showed that the average fold error and the absolute average fold error were all within a two-fold error range, indicating the accuracy and reliability of the established model. Subsequently, the model was used to calculate the in vivo dissolution profile and to establish a correlation with the in vitro dissolution profile (R² = 0.991 in fasting conditions and R² = 0.991 in fed conditions, both in water medium). Notably, under fasting conditions, this correlation exhibited superior performance compared to convolution, deconvolution, and Wagner-Nelson methods. However, under fed conditions, all four methods demonstrated satisfactory correlations.

Conclusion: The mPBPK model can accurately predict the plasma concentration-time curves under both fasted and fed conditions, and provides a new perspective for establishing in vivo-in vitro correlations of drug products such as incomplete in vivo release, sustained/controlled release, and poor absorption.

Background and objective: Fluvoxamine, a potent CYP1A2 inhibitor, increases clozapine serum concentrations by inhibiting its N-dealkylation to norclozapine, thus reducing norclozapine formation and increasing the clozapine/norclozapine ratio. Clinically, fluvoxamine may reduce clozapine tablet burden and mitigate norclozapine-related metabolic side-effects. However, clinical guidance on this co-administration is limited. The current study evaluates the effect of fluvoxamine and its dose on clozapine and norclozapine concentrations and their ratio, while exploring potential influencing factors.

Methods: Patients from long-stay psychiatric wards were included if they had been or were currently receiving clozapine and fluvoxamine, with available steady-state concentrations with and without fluvoxamine coadministraton. Dose-adjusted clozapine and norclozapine concentrations (C/D, ng/mL per mg/day) and their ratios were compared among the two conditions using the Wilcoxon signed-rank test. Median fold increases and associations with clinical variables (e.g. gender, smoking, dosing frequency, baseline clozapine level, fluvoxamine dose), were analysed using Mann-Whitney U tests.

Results: Sixty-seven patients were included. During fluvoxamine co-administration, median C/D clozapine increased from 0.70 to 1.72, C/D norclozapine from 0.43 to 0.80 and the clozapine/norclozapine ratio from 1.66 to 2.16 (all p < 0.001). Median fold increases were 2.51 (clozapine), 1.92 (norclozapine) and 1.27 (clozapine/norclozapine ratio). Greater increases were observed with fluvoxamine doses > 25 mg and baseline clozapine levels < 350 ng/mL, without a significant increase in clozapine/norclozapine ratio. Other factors showed no significant association.

Conclusion: Fluvoxamine significantly increases clozapine and norclozapine concentrations, and their ratio. Doses > 25 mg lead to greater fold increases and more variability. Initiation at 25 mg with a 50% clozapine dose reduction is recommended.

Introduction: Despite dosing protocols and tight therapeutic drug monitoring (TDM), tacrolimus concentrations remain highly variable in pediatric liver transplant (LTx) recipients during the first month post-transplantation. The objective of this study was to describe weight-adjusted tacrolimus concentration-to-dose (C/D/kg) ratios and to identify physical, clinical, and laboratory parameters associated with interpatient pharmacokinetic (PK) variability in hospitalized children during the first month post-LTx.

Methods: In this single-center retrospective cohort study (January 2018-October 2021), we calculated C/D/kg ratios for 36 LTx recipients aged 0-2 years. Descriptive statistics and linear mixed models characterized changes in tacrolimus C/D/kg ratios over time, and we determined the percentage of concentrations within six predefined ranges (0-4, 4-6, 6-8, 8-10, 10-15, and > 15 μg/L).

Results: In total, 524 trough concentrations of orally administered tacrolimus were analyzed. Tacrolimus C/D/kg ratios ranged from 0.19 to 0.75, demonstrating substantial interpatient variability. Time post-transplantation, alanine aminotransferase, aspartate aminotransferase, total bilirubin, coadministration of corticosteroids, spironolactone, fluconazole, fentanyl, amlodipine, flucloxacillin, and ciprofloxacin were significantly associated with interpatient variability (P < 0.05 for all). In the first week, 40.0% tacrolimus trough concentrations were below 4 μg/L, and using TDM the distribution shifted towards the therapeutic mid-range (6-10 μg/L).

Conclusion: TDM of tacrolimus is often not enough to obtain the concentrations in the therapeutic range. Identifying cofounders for variability a priori is essential for guiding efficient and accurate dosing, shifting the focus from reactive TDM towards better dosing strategies that improve PK predictions and ultimately improve therapy for pediatric LTx recipients.

Background and objective: Voriconazole is a broad-spectrum antifungal agent whose efficacy and toxicity are closely related to plasma concentrations, which are highly variable between individuals. Therapeutic drug monitoring (TDM) helps optimize its use but is not always available. In this context, machine learning may help predict subtherapeutic or supratherapeutic levels before TDM results are obtained.

Methods: This was a single-center retrospective study conducted between May 2021 and June 2024 in a tertiary hospital in northern Spain. Adult patients treated with voriconazole for at least 3 days and with a steady-state plasma level measurement were included. Clinical, laboratory, and treatment-related variables were collected. Supervised machine learning models (random forest, support vector machines (SVM), XGBoost, etc.) were trained to classify plasma levels as subtherapeutic, therapeutic, or supratherapeutic.

Results: A total of 147 patients were included (65% male; median age 65 years). Therapeutic concentrations were found in 71% of patients, supratherapeutic in 15%, and subtherapeutic in 14%. Significant differences were observed on the basis of route of administration, dosage form, age, liver function, and certain comorbidities. Aspartate aminotransferase (AST), glomerular filtration rate, and administration route were the most relevant predictors in the models. Random forest achieved the best performance (area under the curve (AUC) 0.675), though still below the threshold for clinical applicability.

Conclusions: Although machine learning models identified relevant predictors of voriconazole exposure, their predictive accuracy was limited and insufficient to replace therapeutic drug monitoring. TDM remains essential for individualized and safe dosing. Integrating pharmacogenetic data and hybrid models combining TDM and computational tools may improve predictive performance and clinical applicability.