药代动力学建模的人工神经网络新方法

人工神经网络(ANN)在药代动力学领域主要用于血药浓度预测、药物结构和药代动力学定量关系、体内体外相关关系研究,群体药物动力学数据分析、药代动力学-药效动力学统一模型研究等。本文就ANN的基本理论及其在药代动力学研究的应用作简要综述。     

Toward a new age of cellular pharmacokinetics in drug discovery

Pharmacokinetics, pharmacology, and toxicology are the major determinants of the success or failure of candidates during drug development. Because inappropriate pharmacokinetics often leads to inefficacy, even toxicity, pharmacokinetics studies have been regarded as crucial components in drug preclinical and clinical research. However, new data increasingly reveal that drug concentrations in plasma or tissues cannot totally explain the efficacy of drug on the target organ. For most drugs that interact with targets localized in cells, intracellular penetration, accumulation, distribution, and elimination are important parameters governing the efficacy in the target cells. So, there is a pressing need to clarify the cellular pharmacokinetics and thus evaluate the efficacy of drugs in the target cells. This review provides a general overview regarding current knowledge about cellular pharmacokinetics in some specific cells and also summarizes the factors that can influence cellular pharmacokinetics. It concludes by discussing potential strategies for optimizing cellular pharmacokinetics and advocating that global cellular pharmacokinetics studies be conducted in future research toward improving drug efficacy.     

Pharmacokinetics of new calcium channel antagonist clevidipine in the rat, rabbit, and dog and pharmacokinetic/pharmacodynamic relationship in anesthetized dogs.

Clevidipine is a new vascular selective calcium channel antagonist of the dihydropyridine type, structurally related to felodipine. Clinical trials have shown that the drug can be used to effectively control the blood pressure in connection with cardiac surgical procedures. The compound is tailored to be a short-acting drug and, due to incorporation of an ester linkage into the drug molecule, clevidipine is rapidly metabolized by ester hydrolysis. The pharmacokinetics of clevidipine and its primary metabolite, H 152/81, were studied in rats, rabbits, and dogs. In addition, the influence of the pharmacokinetics on the effect on mean arterial blood pressure was evaluated in anesthetized dogs. Compartmental nonlinear mixed effect regression analysis was used to calculate the population mean and individual pharmacokinetics of clevidipine, whereas nonlinear regression analysis of individual data was used to determine the pharmacokinetics of the primary metabolite. A linked Emax model was fitted to the individual pharmacodynamic/pharmacokinetic data in dogs. According to the results, clevidipine is a high-clearance drug with a relatively small volume of distribution, resulting in an extremely short half-life in all species studied. The median initial half-life of the individual value (Bayesian estimates) is 12, 20, and 22 s in the rabbit, rat, and dog, respectively. The primary metabolite is a high-clearance compound in the dog, whereas it is a low-clearance compound in the rat. A significant gender difference in the clearance of the metabolite was observed in the rat. The mean maximum reduction in arterial blood pressure is 38 +/- 12% (Emax) and is achieved at 85 +/- 46 nM (EC50). The half-life for reaching equilibrium between the central and the effect compartment (T1/2ke0) is 47 +/- 49 s.     

Population approaches/sparse data analysis for human variability in kinetics and dynamics

The application of sparse data analysis techniques, otherwise known as the population approach, to the assessment of pharmacokinetic and pharmacodynamic variability is reviewed. Using software that has recently become available it is now possible to analyse heterogeneous, sparse data from a variety of clinical trials, particularly during drug development. Therefore the pharmacokinetics and pharmacodynamics of a drug can be studied in the target population and this information can be used to guide dosage for optimal therapeutic benefit.     

Statin response and pharmacokinetics variants

Despite statin therapy being effective in the primary and secondary prevention of coronary heart disease, the benefit of treatment varies between individuals. Interindividual variations in pharmacokinetics play a central role in the cause of variability of drug disposition, and, in turn, the drug's clinical efficacy. Exploring genetic variations that influence pharmacokinetics may lead clinicians to apply the most efficient and safe drug therapy. So far, variants in eight candidate genes related to pharmacokinetics of statins have been investigated as the potential determinant of drug responsiveness or adverse event risk. All reported data remains inconclusive, but it has been suggested that combined analysis of more than two different polymorphisms, or a combination of genetic association and studies using in vitro recombinant expression techniques, may be more informative in predicting the specific phenotype of a genetic variant. Future studies using these approaches could provide more striking evidence, which may be sufficient to justify genetic analysis regarding pharmacokinetic variants in the clinical practice of statin therapy.     

Pharmacokinetics of dipyroxime in the body of rats and dogs

Pharmacokinetics of dipyroxime was studied following intravenous injection to noninbred albino male rats. Dipyroxime is a cholinesterase reactivating drug now widely used as an antidote in poisonings caused by organophosphorus pesticides. Dipyroxime pharmacokinetics analysis made it possible to propose a formula for approximation of the data on another animal species (dogs). The drug concentration in the dog blood plasma is in a good correlation with the predicted level.     

Linear mixed-effect multivariate adaptive regression splines applied to nonlinear pharmacokinetics data

In a frequently performed pharmacokinetics study, different subjects are given different doses of a drug. After each dose is given, drug concentrations are observed according to the same sampling design. The goal of the experiment is to obtain a representation for the pharmacokinetics of the drug, and to determine if drug concentrations observed at different times after a dose are linear in respect to dose. The goal of this paper is to obtain a representation for concentration as a function of time and dose, which (a) makes no assumptions on the underlying pharmacokinetics of the drug; (b) takes into account the repeated measure structure of the data; and (c) detects nonlinearities in respect to dose. To address (a) we use a multivariate adaptive regression splines representation (MARS), which we recast into a linear mixed-effects model, addressing (b). To detect nonlinearity we describe a general algorithm that obtains nested (mixed-effect) MARS representations. In the pharmacokinetics application, the algorithm obtains representations containing time, and time and dose, respectively, with the property that the bases functions of the first representation are a subset of the second. Standard statistical model selection criteria are used to select representations linear or nonlinear in respect to dose. The method can be applied to a variety of pharmacokinetics (and pharmacodynamic) preclinical and phase I-III trials. Examples of applications of the methodology to real and simulated data are reported.     

Effects of age on the elimination of labetalol

Labetalol is an alpha 1- and beta-adrenergic antagonist currently used in the treatment of hypertension. Studies which have evaluated the effects of age on its pharmacokinetics have yielded conflicting results. The purpose of this study is to comprehensively re-evaluate the effect of age on the elimination of labetalol. Data were obtained from 4 single-dose and 3 multiple-dose studies of the pharmacokinetics of the drug. An analysis of covariance was performed on the single-dose data to determine whether the type of subject evaluated (normotensive vs hypertensive), type of assay methodology used and/or age were significant factors affecting labetalol clearance estimates. A similar covariance procedure was used for the multiple-dose data, to assess whether the type of subject, duration of treatment and/or age were significant variables affecting labetalol elimination. Subsequent to the analysis of covariance, linear regression and correlation analysis was used to evaluate the effects of age on labetalol clearance. A modest though significant relationship was observed between the apparent oral clearance of the drug and age; it appeared slightly stronger when clearance was normalised for bodyweight. No relationship was found following multiple doses of the drug. Hence, age does not appear to be a significant factor affecting the oral clearance of labetalol, particularly in individuals receiving the drug in the long term.     

Pharmacokinetics and pharmacodynamics of methotrexate in non-neoplastic diseases

Low dose pulse methotrexate (LDMTX) therapy has become effective in the treatment of autoimmune and lymphoproliferative diseases. The pharmacokinetics of LDMTX is individually highly variable, resulting in a different systemic exposure to the drug and a variable therapeutic/toxic effect in patients. The improvements and exacerbations of disease activity in relation to the introductions and discontinuations of LDMTX therapy suggest the possible immunosuppresive and anti-inflammatory properties of the drug. Because of a strong correlation between the drug pharmacokinetics and the therapeutic outcomes (pharmacodynamics), it seems to be possible to individualise the LDMTX therapy according to the results of pharmacokinetic/pharmacodynamic analysis. In the case of psoriasis, pharmacokinetic/pharmacodynamic analysis in our local study revealed a highly significant inverse relationship between PASI (expressed as a percent of the initial value) and a steady-state AUC(MTX) (area under the curve of methotrexate plasma concentrations; r(8) = -0.65, p < 0.001). The considerable inter-individual variability and low intra-individual variability in MTX pharmacokinetics, supports a role for therapeutic monitoring and dose individualisation at the start of pharmacotherapy. The results of this study suggest that a steady-state AUC(MTX) value of 700 nmol x h/L and higher are associated with a significantly better success rate of antipsoriatic therapy than lower values. The preliminary results in our follow-up study suggest the statistically higher incidence of unwanted effects depending on maximum plasma concentration of the drug. Moreover, statistically significant correlation was found between the toxic effects and exposure to the drug regarding methotrexate plasma concentrations and intracellular storage in erythrocytes. However, the data are still in the process of being completed and are not yet published.     

Impact of genetic polymorphisms of transporters on the pharmacokinetic, pharmacodynamic and toxicological properties of anionic drugs

As the importance of drug transporters in the clinical pharmacokinetics of drugs is recognized, genetic polymorphisms of drug transporters have emerged as one of the determinant factors to produce the inter-individual variability of pharmacokinetics. Many clinical studies have shown the influence of genetic polymorphisms of drug transporters on the pharmacokinetics and subsequent pharmacological and toxicological effects of drugs. The functional change in a transporter in clearance organs such as liver and kidney affects the drug concentration in the blood circulation, while that in the pharmacological or toxicological target can alter the local concentration at the target sites without changing its plasma concentration. As for the transporters for organic anions, some single nucleotide polymorphisms (SNPs) or haplotypes occurring with high frequency in organic anion transporting polypeptide (OATP) 1B1, multidrug resistance 1 (MDR1), and breast cancer resistance protein (BCRP) have been extensively investigated in both human clinical studies and in vitro functional assays. We introduce some examples showing the relationship between haplotypes in transporters and pharmacokinetics and pharmacological effects of drugs. We also discuss how to predict the effect of functional changes in drug transporters caused by genetic polymorphisms on the pharmacokinetics of drugs from in vitro data.     

Genetic polymorphism of human cytochrome p450 involved in drug metabolism

Recent advances in human gene analysis promoted by the human genome project have brought us a massive amount of information. These data can be seen and analyzed by personal computer through individual Web sites. As a result, the best use of bioinformatic is essential for recent molecular biology research. Genetic polymorphism of drug-metabolizing enzymes influences individual drug efficacy and safety through the alteration of pharmacokinetics and disposition of drugs. Considerable amounts of data have now accumulated as allelic differences of various drug metabolizing enzymes. Current understanding of genotype information on cytochrome P450 is hereby summarized, based on the Web site for their use in individual optimization of drug therapy.     

Modeling of PET data in CNS drug discovery and development

Positron emission tomography (PET) is increasingly used in drug discovery and development for evaluation of CNS drug disposition and for studies of disease biomarkers to monitor drug effects on brain pathology. The quantitative analysis of PET data is based on kinetic modeling of radioactivity concentrations in plasma and brain tissue compartments. A number of quantitative methods of analysis have been developed that allow the determination of parameters describing drug pharmacokinetics and interaction with target binding sites in the brain. The optimal method of quantification depends on the properties of the radiolabeled drug or radioligand and the binding site studied. We here review the most frequently used methods for quantification of PET data in relation to CNS drug discovery and development. The utility of PET kinetic modeling in the development of novel CNS drugs is illustrated by examples from studies of the brain kinetic properties of radiolabeled drug molecules.     

药物代谢动力学研究(Ⅰ)——硫脒头孢菌素

本文对硫脒头孢菌素在7例健康受试者体内的药物代谢动力学过程,以及合用丙磺舒的影响,进行了初步的分析。静脉注射后的血药浓度-时间曲线符合开放型二室模型;但肌内注射时可以简化为开放型单室模型。合用丙磺舒对硫脒头孢菌素药代动力学过程的影响比较复杂。对计算的主要结果进行分析讨论后认为:丙磺舒能够抑制硫脒头孢菌素的经尿排泄,表现在尿药排泄量减少,血药峰浓度及血药浓度-时间曲线下总面积增加。对合用丙磺舒的临床意义也进行了讨论。与常用头孢菌素的药代动力学特性,进行了讨论比较。     

Role of pharmacokinetics in the discovery and development of indinavir

The discovery of indinavir is a successful example in which pharmacokinetic and metabolic information were incorporated into drug design. The use of animal and in vitro human metabolic data in predicting the oral bioavailability and hepatic clearance in humans was critical in selecting indinavir as a drug candidate for development. In its development stage, pharmacokinetics continued to play an important role in identifying the key properties of indinavir in vivo, which allowed the characterization and prediction of the time course of drug action under physiological and pathological conditions. This review describes the role of pharmacokinetics and drug metabolism in the discovery and development of indinavir.     

Pharmacokinetics of tenoxicam after oral administration in healthy human subjects of various single doses

Single oral dose pharmacokinetics of tenoxicam in the dose range intended for therapeutic application of the drug have been defined in healthy human subjects. Twelve male volunteers were given, in random fashion, oral administration of 10, 20 or 40 mg of tenoxicam. Plasma drug levels were determined by a standard HPLC method with U.V. detection. Model-independent analysis showed the clearance, elimination half-life and apparent volume of distribution to be independent of dose with mean values of 0.096 L/h, 76 h and 9.2 L respectively. Interindividual variation was six-fold and individual variation less than two-fold. Similar values for kinetic parameters were obtained by model-dependent methods. The pharmacokinetics of oral tenoxicam were linear in the 10-40 mg dose range and the single dose data predict five-fold accumulation for once daily administration.     

An introduction to mixed effect modeling: Concepts,definitions, and justification

Summary In spite of the vast amount of data generated during the new drug development process, much remains to be learned about the pharmacokinetics of a drug once it is released on the market. The ethical and logistical problems which arise during an experimental pharmacokinetic study frequently prevent the study of elderly, pediatric, or critically ill patients. The recognition of these limitations by scientists and regulators have prompted a desire to extract the maximum amount of information from data available during Phase III and Phase IV clinical trials and, in addition, to use information generated during the routine clinical care of patients. Mixed effect model analysis allows one to overcome the problems of analyzing observational data to obtain accurate and precise estimates of population pharmacokinetic parameters. Use of this approach expands the methodologies available to the data analyst and opens up a number of data sources which can now be considered for analysis.     

Implementing and maintaining a private pharmacokinetics practice

The development and maintenance of a private-practice pharmacokinetics service is described. A contracted pharmacokinetics service has been in place at Georgia Baptist Medical Center, a 525-bed hospital, for the past eight years. Physician support, inhouse study results, and literature documentation of the benefits of scheduling drug concentration determinations were used to convince hospital administrators of the value of a pharmacokinetics service. Services were reimbursed by increasing the fee charged for each drug concentration determination, with payment made to the pharmacokinetics service on a monthly basis. The service is associated with the pathology department for administrative purposes. Each member of the pharmacokinetics service is credentialed by the medical staff. The group trains part-time personnel to provide services when needed. Services provided by the group but not required by the contract include collecting quality assurance data, conducting research, serving on hospital committees, writing newsletter articles, conducting inservice-education programs, and providing clinical interventions for nonconsultation patients. This private-practice pharmacokinetics service provides high-quality services and is well accepted within the institution.     

Development of gene drug delivery systems based on pharmacokinetic studies.

A series of pharmacokinetic studies following systemic or local administration for the development of delivery systems for gene drugs, such as plasmid DNA and oligonucleotides, are reviewed. The pharmacokinetics of gene drugs after intravenous injection into mice was evaluated based on clearance concepts. Pharmacokinetic analysis revealed that the overall disposition characteristics of the gene drug itself were determined by the physicochemical properties of its polyanionic DNA. Based on these findings, liver cell-specific carrier systems via receptor-mediated endocytosis were successfully developed by optimizing physicochemical characteristics. On the other hand, the pharmacokinetics of gene drugs after intratumoral injection were assessed in a tissue-isolated tumor perfusion system. The relationship between the physicochemical properties of gene drug delivery systems and intratumoral pharmacokinetics was determined and the therapeutic effect was also discussed in relation to pharmacokinetics. Collectively, it was demonstrated that a rational design of gene drug delivery systems that can control their in vivo disposition is possible by means of pharmacokinetic studies at whole body, organ and cellular levels.     

Effect of diabetes mellitus on pharmacokinetic and pharmacodynamic properties of drugs

The effects of diabetes mellitus on the pharmacokinetics and pharmacodynamics of drugs have been well described in experimental animal models; however, only minimal data exist for humans and the current knowledge regarding the effects of diabetes on these properties remains unclear. Nevertheless, it has been observed that the pharmacokinetics and pharmacodynamics of drugs are changed in subjects with diabetes. It has been reported that diabetes may affect the pharmacokinetics of various drugs by affecting (i) absorption, due to changes in subcutaneous adipose blood flow, muscle blood flow and gastric emptying; (ii) distribution, due to non-enzymatic glycation of albumin; (iii) biotransformation, due to regulation of enzymes/transporters involved in drug biotransformation; and (iv) excretion, due to nephropathy. Previously published data also suggest that diabetes-mediated changes in the pharmacokinetics of a particular drug cannot be translated to others. Although clinical studies exploring the effect of diabetes on pharmacodynamics are still very limited, there is evidence that disease-mediated effects are not limited only to pharmacokinetics but also alter pharmacodynamics. However, for many drugs it remains unclear whether these influences reflect diabetes-mediated changes in pharmacokinetics rather than pharmacodynamics. In addition, even though diabetes-mediated pharmacokinetics and pharmacodynamics might be anticipated, it is important to study the effect on each drug and not generalize from observed data. The available data indicate that there is a significant variability in drug response in diabetic subjects. The discrepancies between individual clinical studies as well as between ex vivo and clinical studies are probably due to (i) the restricted and focused population of subjects in clinical studies; (ii) failure to consider type, severity and duration of the disease; (iii) histopathological characteristics generally being missing; and (iv) other factors such as varying medication use, dietary protein intake, age, sex and obesity. The obesity epidemic in the developed world has also inadvertently influenced the directions of pharmacological research. This review attempts to map new information gained since Gwilt published his paper in Clinical Pharmacokinetics in 1991. Although a large body of research has been conducted and significant progress has been made, we still have to conclude that the available information regarding the effect of diabetes on pharmacokinetics and pharmacodynamics remains unclear and further clinical studies are required before we can understand the clinical significance of the effect. An understanding of diabetes-mediated changes as well as of the source of the variability should lead to the improvement of the medical management and clinical outcomes in patients with this widespread disease.