Effects of multiple-trough sampling design and algorithm on the estimation of population and individual pharmacokinetic parameters / 药学学报

The purpose of this study is to investigate the effects of multiple-trough sampling design and nonlinear mixed effect modeling (NONMEM) algorithm on the estimation of population and individual pharmacokinetic parameters. Oxcarbazepine and tacrolimus were used as one-compartment and two-compartment model drugs, respectively. Seven sampling designs were investigated using various number of trough concentrations per individual ranging from 1-4. Monte Carlo simulations were performed to produce state-steady trough concentrations. One-compartment model was used to fit simulated data from oxcarbazepine and tacrolimus. The accuracy and precision of the estimated parameters were evaluated using the median prediction error (PE), the median absolute PE and boxplot. The results indicated that trough concentrations could yield reliable estimates of apparent clearance (CL/F). For oxcarbazepine, as the number of trough concentrations per subject increased, the accuracy and precision of CL/F, between-subject variability (BSV) of CL/F and residual variability (RUV) tended to be improved. For tacrolimus, however, although no improvement were observed in the accuracy of CL/F and BSV of CL/F, the PE distribution ranges were significantly narrowed and the RUV estimates were less bias and imprecise. In terms of algorithm, Monte Carlo importance sampling (IMP) and IMP assisted by mode a posteriori estimation (IMPMAP) were consistently better than other methods. Additionally, the sampling design had no significant effects on the individual parameter estimates, which were only depended on the interaction between BSV and RUV in various algorithms. Decreased in BSV and RUV levels can improve the accuracy and precision of the estimation for both population and individual pharmacokinetic parameter estimates.

Estimation of individual pharmacokinetic parameters using maximum a posteriori Bayesian method with D-optimal sampling strategy / 药学学报

This study was aimed to develop a maximum a posteriori Bayesian (MAPB) estimation method to estimate individual pharmacokinetic parameters based on D-optimal sampling strategy. Meanwhile, the performance of MAPB was compared with the multiple linear regression (MLR) method in terms of accuracy and precision. Pharmacokinetic study of pioglitazone was employed as the example case. The population pharmacokinetics was characterized by nonlinear mixed effects model (NONMEM). The sparse sampling strategy (1-4 points) was identified by D-optimal algorithm using WinPOPT software. The simulated data generated by Monte Carlo method were used to access the performance of MAPB and MLR. As the number of samples per subject decreased, the accuracy and precision of MAPB method tended to get worse. The estimation for CL and Vby MAPB using D-optimal two-point design had less bias with low inter-individual variability, and had more bias and imprecision with high residue variability. The estimation of AUC by MAPB using D-optimal 2 points design had similar accuracy and precision to MLR. However, MAPB estimation was better than MLR while adjusting the sampling time to one hour. Overall, the MAPB method had similar predictive performance as MLR, but MAPB could provide more pharmacokinetic information with higher sampling flexibility.

Development and identifiability analysis of parent-metabolite pharmacokinetic model for risperidone and its main active metabolite 9-hydroxyrisperidone / 药学学报

To develop a parent-metabolite pharmacokinetic model for risperidone (RIP) and its major active metabolite (9-hydroxyrisperidone) and investigate their pharmacokinetics characteristics in healthy male volunteers, twenty-two healthy volunteers were orally given a single dose of 2 mg RIP. Plasma samples were collected in the period of 96 hours and concentrations of RIP and 9-hydroxyrisperidone were measured by a validated HPLC/MS method. CYP2D6 phenotypes were identified by the T1/2 of RIP and 9-hydroxyrisperidone according to the literature. Model structure identifiability analysis was performed by the similarity transformation approach to investigate whether the unknown parameters of the proposed model could be estimated from the designed experiment. Pharmacokinetics parameters were estimated using weighted least squares method, and the final pharmacokinetics model were tested and evaluated by Monte Carlo simulation. Eighteen volunteers were phenotyped as extensive metabolizers (EM) and four volunteers were identified as intermediate metabolizers (IM). The final model included central and peripheral compartment for both parent (RIP) and metabolite (9-hydroxyrisperidone) respectively. Model structure identifiability analysis indicated that the proposed model was local identifiable. However, if the ratio of RIP converted to 9-hydroxyrisperidone was assumed to be 32% in EM, and 22% in IM, the model could be globally identifiable. The predicted time-concentration curve and AUC(0-t), C(max), T(max) of RIP and 9-hydroxyrisperidone estimated by the established model were in agreement with the observations and noncompartment analysis. Rate constant of RIP conversion to 9-hydroxyrisperidone was (0.12 +/- 0.08) h(-1) and (0.014 +/- 0.007) h(-1) for EM and IM, respectively. Elimination rate constants of RIP were (0.25 +/- 0.18) and (0.05 +/- 0.23) h(-1) for EM and IM, respectively. Model validation result showed that all parameters derived from the concentration data fitted well with the theoretical value, with mean prediction error of most PK parameter within +/- 15%. The established model well defined the disposition of RIP and 9-hydroxyrisperidone simultaneously and showed large inter-individual pharmacokinetics variation in different CYP2D6 phenotype. The model also provide a useful approach to characterize pharmacokinetics of other parent-metabolite drugs.

Bioequivalence assessment of pioglitazone hydrochloride oral preparation by limited sampling strategy / 药学学报

<p><b>AIM</b>To develop limited sampling strategy (LSS) for estimation of C(max) and AUC(0-t) and assessing the bioequivalence of two pioglitazone hydrochloride (PGT) preparations.</p><p><b>METHODS</b>Healthy subjects (n = 20), enrolled in a bioequivalence study, were received 30 mg PGT po of reference or test formulation. The plasma concentration of PGT was determined by the validated HPLC method. A multiple linear regression analysis of the Cmax and AUC(0-t) against the PGT concentration for the reference formulation was carried out to develop LSS models to estimate these parameters. The models were internally validated by the Jackknife method and externally validated using simulated sets generated by Monte Carlo method. The best model was employed to assess bioequivalence of the two PGT formulations.</p><p><b>RESULTS</b>The linear relationship between pharmacokinetics parameters and single concentration point was poor. Several models for these parameters estimation met the predefined criteria (r2 > 0.9). The Jackknife validation procedure revealed that LSS models based on two sampling times (C1, C2.5 and C1.5, C2.5 for C(max); C1.5, C9 and C2.5, C9 for AUC(0-t) predict accurately. Mean prediction errors (MPE) were less than 3%, and mean absolute prediction error (MAE) were less than 9%. The prediction error (PE) beyond 20% was less than 5% of total samples. Model external validation by Monte Carlo simulated data indicated that the most informative sampling combinations were C1.5, C2.5 for C(max), and C1.5, C9 for AUC(0-t), respectively. MPE and MAE of the proposed models were less than 5% , and 9% respectively. The PE beyond 20% was less than 5% of the total. Bioequivalence assessment of the two PGT formulations, based on the best LSS models, provided results similar to those obtained using all the observed concentration-time data points, and indicated that the two PGT formulations were bioequivalent.</p><p><b>CONCLUSION</b>The LSS method for bioequivalence assessment of PGT formulations was established and proved to be applicable and accurate. Thus, it could be considered appropriate for PGT bioequivalence study with inexpensive cost of sampling acquisition and analysis. Key words: pioglitazone hydrochloride; limited sampling strategy; Monte Carlo simulation; bioequivalence</p>

New 3D amino acid structure descriptors and its application to the polypeptide QSAR / 药学学报

<p><b>AIM</b>To establish a new amino acid structure descriptor that can be applied to polypeptide QSAR studies.</p><p><b>METHODS</b>The new amino acid structure descriptor c-scales were derived from a principal components analysis of 167 amino acid structure descriptor indexes by theoretic calculation. The c1,c2,c3-scales were related to 3D structural features of amino acid such as steric, electronic and conformation properties etc. G/PLS regression method was used to find out the relationship between the c-scales and the biological activity and developed QSAR models of the polypeptides.</p><p><b>RESULTS</b>Using the established method, we developed accordingly QSAR models of Bitter tasting dipeptide, ACE inhibitors and bradykinin-potentiating pentapeptides and their r2 and XV-r2 were more than 0.70.</p><p><b>CONCLUSION</b>The c-scales can quantitatively describe the 3D structural features of any coded and non-coded amino acid and can be used to establish a QSAR model of good predictability.</p>

Effect of MDR1 polymorphic expression on oral disposition of cyclosporine A / 药学学报

<p><b>AIM</b>To determine the relationship between C3435T mutation in exon 26 of the human multidrug resistant 1 gene and cyclosporine (CsA) pharmacokinetic (PK) parameters among healthy Chinese volunteers by nonlinear mixed effect model (NONMEM).</p><p><b>METHODS</b>Twenty healthy subjects were given orally a single dose of 500 mg CsA in microemulsion solution. Blood CsA concentrations were measured with HPLC and the genotype for the C3435T polymorphism of MDR1 gene was determined with the PCR and restriction fragment length polymorphism. The results were further confirmed by sequencing. NONMEM was performed to assess the effect of genotype on CsA PK profile.</p><p><b>RESULTS</b>MDR1 C3435T genotype was identified as the best predictor of CsA systemic exposure. The relative bioavailability of CsA was 40% higher in subjects who carried at least one 3435C allele compared to that of TT type individuals in the study population.</p><p><b>CONCLUSION</b>The MDR1 C3435T genotype offers a potential basis of mechanism to explain inter-subject differences in CsA oral bioavailability.</p>