Effects of SLC22A1 Polymorphisms on Metformin-Induced Reductions in Adiposity and Metformin Pharmacokinetics in Obese Children With Insulin Resistance.

Steady-state population pharmacokinetics of a noncommercial immediate-release metformin (hydrochloride) drug product were characterized in 28 severely obese children with insulin resistance. The concentration-time profiles with double peaks were well described by a 1-compartment model with 2 absorption sites. Mean population apparent clearance (CL/F) was 68.1 L/h, and mean apparent volume of distribution (V/F) was 28.8 L. Body weight was a covariate of CL/F and V/F. Estimated glomerular filtration rate was a significant covariate of CL/F (P < .001). SLC22A1 genotype did not significantly affect metformin pharmacokinetics. The response to 6 months of metformin treatment (HbA1c , homeostasis model assessment for insulin resistance, fasting insulin, and glucose changes) did not differ between SLC22A1 wild-type subjects and carriers of presumably low-activity SLC22A1 alleles. However, SLC22A1 variant carriers had smaller reductions in percentage of total trunk fat after metformin therapy, although the percentage reduction in trunk fat was small. The median % change in trunk fat was -2.20% (-9.00% to 0.900%) and -1.20% (-2.40% to 7.30%) for the SLC22A1 wild-type subjects and variant carriers, respectively. Future study is needed to evaluate the effects of SLC22A1 polymorphisms on metformin-mediated weight reduction in obese children.

PARP activity in peripheral blood lymphocytes as a predictive biomarker for PARP inhibition in tumor tissues - A population pharmacokinetic/pharmacodynamic analysis of rucaparib.

PURPOSE: Rucaparib is a potent Poly (ADP-ribose) Polymerase (PARP) inhibitor currently under clinical development. The objectives of this analysis were to establish population PK and PK/PD models for rucaparib, and to evaluate the predictability of PARP activity in PBL for PARP activity in tumor tissues. EXPERIMENTAL DESIGN: Rucaparib concentrations and PARP activity in human PBLs and tumor issues were obtained from 32 patients with solid tumors in a Phase 1 First-in-Patient study. Simulations were conducted to evaluate different dosing regimens. RESULTS: A 3-compartment PK model best described the PK of rucaparib. An Emax model best described the exposure and PARP inhibition relationship. The maximum PARP inhibition (Imax) achieved in PBLs and in tumors were 90.9% and 90.0% of the baseline PARP activity, and the IC50 values were 1.05 ng/mL and 1.10 ng/mL, respectively. PAR polymer baseline value was found to be a covariate of Emin. CONCLUSION: Population PK and PK/PD models have been established to describe population PK of rucaparib and the relationship between rucaparib plasma concentration and PARP inhibition in both PBLs and tumor issues. Results from this trial indicated that PARP inhibition in PBLs can be used as a substitute for PARP inhibition in melanoma tumor tissues.

The role of mitochondrial bioenergetics and reactive oxygen species in coronary collateral growth.

Coronary collateral growth is a process involving coordination between growth factors expressed in response to ischemia and mechanical forces. Underlying this response is proliferation of vascular smooth muscle and endothelial cells, resulting in an enlargement in the caliber of arterial-arterial anastomoses, i.e., a collateral vessel, sometimes as much as an order of magnitude. An integral element of this cell proliferation is the process known as phenotypic switching in which cells of a particular phenotype, e.g., contractile vascular smooth muscle, must change their phenotype to proliferate. Phenotypic switching requires that protein synthesis occurs and different kinase signaling pathways become activated, necessitating energy to make the switch. Moreover, kinases, using ATP to phosphorylate their targets, have an energy requirement themselves. Mitochondria play a key role in the energy production that enables phenotypic switching, but under conditions where mitochondrial energy production is constrained, e.g., mitochondrial oxidative stress, this switch is impaired. In addition, we discuss the potential importance of uncoupling proteins as modulators of mitochondrial reactive oxygen species production and bioenergetics, as well as the role of AMP kinase as an energy sensor upstream of mammalian target of rapamycin, the master regulator of protein synthesis.

Mitochondrial oxidative stress corrupts coronary collateral growth by activating adenosine monophosphate activated kinase-α signaling.

OBJECTIVE: Our goal was to determine the mechanism by which mitochondrial oxidative stress impairs collateral growth in the heart. APPROACH AND RESULTS: Rats were treated with rotenone (mitochondrial complex I inhibitor that increases reactive oxygen species production) or sham-treated with vehicle and subjected to repetitive ischemia protocol for 10 days to induce coronary collateral growth. In control rats, repetitive ischemia increased flow to the collateral-dependent zone; however, rotenone treatment prevented this increase suggesting that mitochondrial oxidative stress compromises coronary collateral growth. In addition, rotenone also attenuated mitochondrial complex I activity and led to excessive mitochondrial aggregation. To further understand the mechanistic pathway(s) involved, human coronary artery endothelial cells were treated with 50 ng/mL vascular endothelial growth factor, 1 µmol/L rotenone, and rotenone/vascular endothelial growth factor for 48 hours. Vascular endothelial growth factor induced robust tube formation; however, rotenone completely inhibited this effect (P<0.05 rotenone versus vascular endothelial growth factor treatment). Inhibition of tube formation by rotenone was also associated with significant increase in mitochondrial superoxide generation. Immunoblot analyses of human coronary artery endothelial cells with rotenone treatment showed significant activation of adenosine monophosphate activated kinase (AMPK)-α and inhibition of mammalian target of rapamycin and p70 ribosomal S6 kinase. Activation of AMPK-α suggested impairments in energy production, which was reflected by decrease in O2 consumption and bioenergetic reserve capacity of cultured cells. Knockdown of AMPK-α (siRNA) also preserved tube formation during rotenone, suggesting the negative effects were mediated by the activation of AMPK-α. Conversely, expression of a constitutively active AMPK-α blocked tube formation. CONCLUSIONS: We conclude that activation of AMPK-α during mitochondrial oxidative stress inhibits mammalian target of rapamycin signaling, which impairs phenotypic switching necessary for the growth of blood vessels.

Associations of ABCB1 and IL-10 genetic polymorphisms with sirolimus-induced dyslipidemia in renal transplant recipients.

BACKGROUND: Hyperlipidemia is a common adverse effect of sirolimus (SRL). We previously showed significant associations of ABCB1 3435C>T and IL-10 -1082G>A with log-transformed SRL dose-adjusted weighted-normalized trough. We further examined to see whether these polymorphisms were also associated with SRL-induced dyslipidemia. METHODS: Genotyping was performed for ABCB1 1236C>T, 2677 G>T/A, and 3435C>T; CYP3A4 -392A>G; CYP3A5 6986A>G and 14690G>A; IL-10 -1082G>A; TNF -308G>A; and ApoE ε2, ε3, and ε4 alleles. The longitudinal changes of total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), and triglyceride (TG) levels after SRL treatment before statin therapy were analyzed by a linear mixed-effects model, with adjustments for selected covariates for each lipid. RESULTS: Under the dominant genetic model, ABCB1 3435C>T was associated with TC (P=0.0001) and LDL-C (P<0.0001) values after SRL administration. Mean TC and LDL-C levels were 26.9 and 24.9 mg/dL higher, respectively, in ABCB1 3435T carriers than 3435CC homozygotes at an average SRL trough concentration of 4 ng/mL without concomitant medication. ABCB1 1236C>T under the recessive model and IL-10 -1082G>A under the dominant model were associated with log-transformed TG values (P=0.0051 and 0.0436, respectively). Mean TG value was 25.1% higher in ABCB1 1236TT homozygotes compared with ABCB1 1236C carriers and was 12.4% higher in IL-10 -1082AA homozygotes than -1082G carriers. CONCLUSIONS: ABCB1 polymorphisms were found to be associated with lipid responses to SRL treatment, confirming the role of ABCB1 gene in SRL pharmacokinetics and pharmacodynamics. Further studies are necessary to define the role of ABCB1 and IL-10 polymorphisms on SRL-induced dyslipidemia in renal transplantation.

Diabetes mellitus reduces the clearance of atorvastatin lactone: results of a population pharmacokinetic analysis in renal transplant recipients and in vitro studies using human liver microsomes.

BACKGROUND AND OBJECTIVE: Patients with diabetes mellitus might be at a higher risk of HMG-CoA reductase inhibitor (statin)-induced myotoxicity, possibly because of reduced clearance of the statin lactone. The present study was designed to investigate the effect of diabetes on the biotransformation of atorvastatin acid, both in vivo in nondiabetic and diabetic renal transplant recipients, and in vitro in human liver samples from nondiabetic and diabetic donors. SUBJECTS AND METHODS: A total of 312 plasma concentrations of atorvastatin acid and atorvastatin lactone, from 20 nondiabetic and 32 diabetic renal transplant recipients, were included in the analysis. Nonlinear mixed-effects modelling was employed to determine the population pharmacokinetic estimates for atorvastatin acid and atorvastatin lactone. In addition, the biotransformation of these compounds was studied using human liver microsomal fractions obtained from 12 nondiabetic and 12 diabetic donors. RESULTS: In diabetic patients, the plasma concentration of atorvastatin lactone was significantly higher than that of atorvastatin acid throughout the 24-hour sampling period. The optimal population pharmacokinetic model for atorvastatin acid and atorvastatin lactone consisted of a two- and one-compartment model, respectively, with interconversion between atorvastatin acid and atorvastatin lactone. Parent drug was absorbed orally with a population estimate first-order absorption rate constant of 0.457 h(-1). The population estimates of apparent oral clearance (CL/F) of atorvastatin acid to atorvastatin lactone, intercompartmental clearance (Q/F), apparent central compartment volume of distribution after oral administration (V(1)/F) and apparent peripheral compartment volume of distribution after oral administration (V(2)/F) for atorvastatin acid were 231 L/h, 315 L/h, 325 L and 4910 L, respectively. The population estimates of apparent total clearance of atorvastatin lactone (CL(M)/F), apparent intercompartmental clearance of atorvastatin lactone (Q(M)/F) and apparent volume of distribution of atorvastatin lactone after oral administration (V(M)/F) were 85.4 L/h, 166 L/h and 249 L, respectively. The final covariate model indicated that the liver enzyme lactate dehydrogenase was related to CL/F and alanine aminotransferase (ALT) was related to Q/F. Importantly, diabetic patients have 3.56 times lower CL(M)/F than nondiabetic patients, indicating significantly lower clearance of atorvastatin lactone in these patients. Moreover, in a multivariate population pharmacokinetics model, diabetes status was the only significant covariate predicting the values of the CL(M)/F. Correspondingly, the concentration of atorvastatin acid remaining in the microsomal incubation was not significantly different between nondiabetic and diabetic liver samples, whereas the concentration of atorvastatin lactone was significantly higher in the samples from diabetic donors. In vitro studies, using recombinant enzymes, revealed that cytochrome P450 (CYP) 3A4 is the major CYP enzyme responsible for the biotransformation of atorvastatin lactone. CONCLUSIONS: These studies provide compelling evidence that the clearance of atorvastatin lactone is significantly reduced by diabetes, which leads to an increased concentration of this metabolite. This finding can be clinically valuable for diabetic transplant recipients who have additional co-morbidities and are on multiple medications.

Associations of ABCB1 3435C>T and IL-10-1082G>A polymorphisms with long-term sirolimus dose requirements in renal transplant patients.

BACKGROUNDS: Sirolimus (SRL) absorption and metabolism are affected by p-glycoprotein-mediated transport and CYP3A enzyme activity, which are further under the influences of cytokine concentrations. This retrospective study determined the associations of adenosine triphosphate-binding cassette, subfamily B, member 1 (ABCB1) 1236C>T, 2677 G>T/A, and 3435C>T, cytochrome P450, family 3, subfamily A, polypeptide 4 (CYP3A4) -392A>G, cytochrome P450, family 3, subfamily A, polypeptide 5 (CYP3A5) 6986A>G and 14690G>A, interleukin (IL)-10 -1082G>A, and tumor necrosis factor (TNF) -308G>A polymorphisms with SRL dose-adjusted, weight-normalized trough concentrations (C/D) at 7 days, and at 1, 3, 6, and 12 months after initiation of SRL. METHODS: Genotypes for 86 renal transplant patients who received SRL-based maintenance immunosuppressive therapy were determined using polymerase chain reaction followed by chip-based mass spectrometry. The changes of log-transformed C/D over the days posttransplantation were analyzed using a linear mixed-effects model, with adjustments for body mass index and weight-normalized doses of tacrolimus, prednisone, clotrimazole, and statins. RESULTS: ABCB1 3435C>T and IL-10 -1082G>A were significantly associated with log C/D (P=0.0016 and 0.0394, respectively). Mean SRL C/D was 48% higher in patients with ABCB1 3435CT/TT genotype than those with 3435CC genotype, and was 24% higher in IL-10 -1082GG compared with -1082AG/AA. CONCLUSIONS: ABCB1 3435C>T and IL-10 -1082G>A were significantly associated with long-term SRL dose requirements. Genetics can play a significant role in SRL dosing and may be useful in therapeutic monitoring of SRL in renal transplantation. Future replication studies are needed to confirm these associations.

Morphine and its metabolites after patient-controlled analgesia: considerations for respiratory depression.

STUDY OBJECTIVE: To assess concentrations of morphine and its metabolites after patient-controlled analgesia (PCA). DESIGN: Pilot pharmacokinetic study of morphine and pharmacokinetic simulation. SETTING: Post-anesthesia care room and ward of an academic teaching hospital. PATIENTS: 10 ASA physical status I, II, and III postoperative surgical patients. INTERVENTIONS: Patients received morphine via PCA by routine hospital protocols. MEASUREMENTS: The population mean plasma and effect-site concentrations of morphine, morphine-6-glucuronide (M6G), and morphine-3-glucuronide (M3G) was simulated in 4 patient group scenarios: morphine PCA used alone, morphine PCA used with continuous background morphine infusion of 0.5 mg/hr, morphine PCA used with continuous background morphine infusion of 1.0 mg/hr, and morphine PCA used with continuous background morphine infusion of 2.0 mg/hr. MAIN RESULTS: The 4 groups exhibited simulated peak morphine, M6G, and M3G effect-site concentrations at 8 to 24 hours post-infusion. The highest peak morphine, M6G, and M3G effect-site concentrations decreased in the following order by group: 2.0 mg/hr morphine infusion + PCA group, 1.0 mg/hr morphine infusion + PCA group, and 0.5. mg/hr morphine infusion + PCA group. CONCLUSIONS: Patients receiving morphine PCA should be monitored closely from 8 to 24 hours postoperatively. Morphine PCA given with background infusion rates up to 1.0 mg/hr does not offer distinct pharmacokinetic advantages over morphine PCA alone. Morphine PCA with background infusion rate of 2.0 mg/hr is associated with the greatest risk of respiratory depression.

Population pharmacokinetics of mycophenolic acid and metabolites in patients with glomerulonephritis.

Mycophenolic acid (MPA) is an inosine monophosphate dehydrogenase inhibitor used for glomerulonephritis treatment. The objective of the current study was to develop a population pharmacokinetic model for MPA and metabolites in glomerulonephritis to enable appropriate design of MPA regimens in these patients with alterations in kidney structure and function. Thirty-nine patients with glomerulonephritis and receiving mycophenolate mofetil were recruited to participate in a 24-hour pharmacokinetic study. Blood was collected at times 0, 0.5, 1.0, 1.5, 2, 3, 4, 6, 8, 12, and 24 hours and urine was collected over the intervals of 0 to 6, 6 to 12, and 12 to 24 hours. Plasma and urine samples were assayed for MPA and MPA glucuronide (MPAG) by high-performance liquid chromatography and for acyl-MPA glucuronide (AcMPAG) by liquid chromatography/mass spectrometry. Population pharmacokinetic analysis and covariate model building were evaluated using Non-linear Mixed Effect Modeling software (NONMEM, Version 6.2.0; ICON Development Solutions, Ellicott City, MD). The final model for MPA and its metabolites consisted of nine discrete compartments; 1) depot gastrointestinal; 2) central MPA; 3) peripheral MPA; 4) gallbladder; 5) MPA urine; 6) MPAG central; 7) MPAG urine; 8) AcMPAG central; and 9) AcMPAG urine compartment. The MPA population mean estimates for apparent nonrenal clearance (ClNR/F) and apparent central volume of distribution were 14.3 L/hr and 21.1 L, respectively. The mean population estimate for apparent renal clearance (ClR/F) was dependent on estimated creatinine clearances (eClcr); 0.0975 L/hr for eClcr 80 mL/min or less and 0.157 L/hr for eClcr greater than 80 mL/min. Covariate analyses identified: eClcr on CLNR,MPA/F (P < 0.001), eClcr (with a cutoff value at 80 mL/min) on CLR,MPA/F (P < 0.025), serum albumin on CLNR,MPA/F (P < 0.01), eClcr on CLR,MPAG/F (P < 0.001), and eClcr on CLR,AcMPAG/F (P < 0.001). Evaluation of the final model by visual predictive check showed that most of the observed values were within the 95th percent prediction interval generated from 100 simulations of the final model. The current population pharmacokinetic model demonstrated eClcr and serum albumin influenced the renal and nonrenal components of Cl/F, suggesting patients with glomerulonephritis would have highly altered MPA exposures.

Pharmacokinetic-pharmacodynamic modeling of the hypotensive effect of remifentanil in infants undergoing cranioplasty.

OBJECTIVES: Although remifentanil has been used to induce hypotension during surgery in infants, no pharmacokinetic-pharmacodynamic (PKPD) model exists for its quantitative analysis. Our aim was to determine the quantitative relationship between whole blood remifentanil concentration and its hypotensive effect during surgery in infants. METHODS/MATERIALS: We studied seven infants (age 0.3-1 year) who underwent cranioplasty surgery and received remifentanil delivered by a computer-controlled infusion pump during the maintenance of anesthesia. Arterial blood samples to determine remifentanil concentration and mean arterial blood pressure (MAP) measurements were collected. A simultaneous PKPD mixed-effects model was built in NONMEM. RESULTS: A total of 77 remifentanil concentrations and 185 MAP measurements were collected. Remifentanil pharmacokinetics was described with a two-compartment model, parameter estimates were 2.99 l x min(-1) x 70 kg(-1) for clearance and 16.23 l x 70 kg(-1) for steady state volume of distribution. Mean baseline MAP was 69.7 mmHg and was decreased as per clinical requirements. A sigmoidal E(max) model driven by an effect compartment described the decrease in MAP, with an estimated concentration to decrease MAP by half (EC(50)) being 17.1 ng x ml(-1). CONCLUSIONS: Remifentanil is effective in causing hypotension. The final model predicts that a steady state remifentanil concentration of 14 ng.ml(-1) would typically achieve a 30% decrease in MAP.

Population pharmacokinetics of remifentanil in infants and children undergoing cardiac surgery.

BACKGROUND: The aim of this study was to provide a model-based analysis of the pharmacokinetics of remifentanil in infants and children undergoing cardiac surgery with cardiopulmonary bypass (CPB). METHODS: We studied nine patients aged 0.5 to 4 years who received a continuous remifentanil infusion via a computer-controlled infusion pump during cardiac surgery with mildly hypothermic CPB were studied. Arterial blood samples taken prior to, during and after CPB were analyzed for remifentanil concentrations using a validated gas-chromatographic mass-spectrophotometric assay. We used population mixed-effects modeling to characterize remifentanil pharmacokinetics. The final model was evaluated by its predictive performance. RESULTS: The pharmacokinetics of remifentanil was described by a 1-compartment model with adjustments for CPB. Population mean parameter estimates were 1.41 L for volume of distribution (V) and 0.244 L/min for clearance. V was increased during CPB and post-CPB to 2.41 times the pre-CPB value. The median prediction error and the median of individual median absolute prediction error were 2.44% and 21.6%, respectively. CONCLUSION: Remifentanil dosage adjustments are required during and after CPB due to marked changes in the V of the drug. Simulations indicate that a targeted blood concentration of 14 ng/mL is achieved and maintained in 50% of typical patients by administration of an initial dose of 18 mug remifentanil followed by an infusion of 3.7 mug/min before, during and post-CPB, supplemented with a bolus dose of 25 mug given at the start of CPB.

Determination of the pharmacodynamic interaction of propofol and dexmedetomidine during esophagogastroduodenoscopy in children.

OBJECTIVES: Propofol is a sedative-hypnotic drug commonly used to anesthetize children undergoing esophagogastroduodenoscopy (EGD). Dexmedetomidine is a highly selective alpha-2 adrenergic receptor agonist that has been utilized in combination with propofol to provide anesthesia. There is currently no information regarding the effect of intravenous dexmedetomidine on the propofol plasma concentration-response relationship during EGD in children. This study aimed to investigate the pharmacodynamic interaction of propofol and dexmedetomidine when used in combination for children undergoing EGD. METHODS: A total of 24 children undergoing EGD, ages 3-10 years, were enrolled in this study. Twelve children received dexmedetomidine 1 microg x kg(-1) given over 10 min as well as a continuous infusion of propofol delivered by a computer-assisted target-controlled infusion (TCI) system with target plasma concentrations ranging from 2.8 to 4.0 microg x ml(-1) (DEX group). Another group of 12 children undergoing EGD also received propofol administered by TCI targeting comparable plasma concentrations without dexmedetomidine (control group). We used logistic regression to predict plasma propofol concentrations at which 50% of the patients exhibited minimal response to stimuli (EC50 for anesthesia). RESULTS: The EC50 +/- SE values in the control and DEX groups were 3.7 +/- 0.4 microg x ml(-1) and 3.5 +/- 0.2 microg x ml(-1), respectively. There was no significant shift in the propofol concentration-response curve in the presence of dexmedetomidine. CONCLUSION: The EC50 of propofol required to produce adequate anesthesia for EGD in children was unaffected by a concomitant infusion of dexmedetomidine 1 microg x kg(-1) given over 10 min.

Population pharmacokinetics of mycophenolic acid and its 2 glucuronidated metabolites in kidney transplant recipients.

The population pharmacokinetics of mycophenolic acid (MPA) and its phenolic (MPAG) and acyl (AcMPAG) glucuronide metabolites were studied in patients taking enteric-coated mycophenolate sodium. Plasma samples (n = 232), obtained from 18 renal transplant recipients, were analyzed for MPA, MPAG, and AcMPAG using a validated high-performance liquid chromatography/ultraviolet assay. Population pharmacokinetic analysis was performed using NONMEM. The pharmacokinetics of MPA were best described by a 2-compartment model, with MPAG and AcMPAG produced from the central compartment and with enterohepatic recirculation of MPA via these 2 metabolites. Population mean estimates for MPA were apparent clearance (CL/F) of 10.6 L/h (interindividual variability [IIV] = 21.4%) and apparent volume of distribution of the central compartment (V(1)/F) of 25.9 L (IIV = 87.8%). Mean elimination rate constants of MPAG and AcMPAG were 0.323 h(-1) (IIV = 29.1%) and 0.206 h(-1) (IIV = 48.8%), respectively. The mean fraction of MPA converted to MPAG and AcMPAG, normalized by their volumes of distribution (FM(AG) and FM(AC), respectively), was also estimated. The elimination rate constant for MPAG and FM(AC) was influenced by glomerular filtration rate in patients with renal impairment. The visual predictive check, based on 100 simulated data sets each for MPA, MPAG, and AcMPAG, found that the final pharmacokinetic model adequately predicts the observed concentrations of all 3 species.

Population pharmacokinetics of tacrolimus in whole blood and plasma in asian liver transplant patients.

OBJECTIVES: The objectives of this study were to develop population pharmacokinetic models of tacrolimus in an Asian population with whole blood and plasma drug concentration data, to compare the variability of the pharmacokinetic parameters in these two matrices and to search for the main patient characteristics that explain the variability in pharmacokinetic parameters. STUDY DESIGN: Prospective pharmacokinetic assessment followed by model fitting. PATIENTS: Whole blood samples from 31 liver transplant patients in a local hospital receiving oral tacrolimus as part of their immunosuppressive therapy were assessed. Plasma samples from 29 of the 31 patients were also evaluated. Concentrations of tacrolimus in whole blood and plasma were determined by an electrospray high-performance liquid chromatography with tandem mass spectrometry. Two hundred and thirteen whole blood and 157 plasma tacrolimus concentrations were used for building two nonlinear mixed-effects population models to describe the disposition of tacrolimus in whole blood and plasma, respectively. Covariates that were investigated included demographic characteristics, biological markers of liver and renal functions, corticosteroid dose and haematological parameter. RESULTS: A one-compartment model was used to describe the whole blood and plasma concentration-time data of tacrolimus after oral administration. For the whole blood population model, the population estimates of the first-order absorption rate constant (k(a)), apparent clearance based on whole blood concentration after oral administration (CL(B)/F) and apparent volume of distribution based on whole blood concentrations after oral administration (V(d,B)/F) were 2.08h(-1), 14.1 L/h and 217L, respectively. The coefficient of variations (CVs) of interpatient variabilities in CL(B)/F and V(d,B)/F were 65.7% and 63.8%, respectively. Bodyweight, liver and renal function influenced CL(B)/F, while height and haematocrit influenced V(d,B)/F. The residual (unexplained) variability was 34.8%. For the plasma population model, the population estimates of the k(a), apparent clearance based on plasma concentrations after oral administration (CL(P)/F) and apparent volume of distribution based on plasma concentrations after oral administration (V(d,P)/F) were 5.21h(-1), 537 L/h and 563L, respectively. The CVs of interpatient variabilities in CL(P)/F and V(d,P)/F were 96.0% and 105.4%, respectively. Bodyweight was found to influence CL(P)/F, while the erythrocyte-to-plasma concentration ratio influenced V(d,P)/F. The residual (unexplained) variability was 49.8% at the mean plasma concentration of 1.1 ng/mL. CONCLUSIONS: Whole blood and plasma population pharmacokinetic models of tacrolimus in Asian adult and paediatric liver transplant patients were developed using prospective data in a clinical setting. This has identified and quantified sources of interindividual variability in CL(B)/F, V(d,B)/F, CL(P)/F and V(d,P)/F of tacrolimus in Asian liver transplant patients. Information derived from the whole blood population model may subsequently be used by clinicians for dosage individualisation through Bayesian forecasting.