The serum level approach to individualization of drug dosage

Summary The importance of individualizing the dosage of potent drugs in order to maximize their therapeutic effectiveness and safety is generally accepted. Whenever possible the dosage of a drug should be titrated directly in each patient against the intensity of its therapeutic or toxic actions. Unfortunately, for many drugs convenient clinical yardsticks of the intensity of their pharmacologic effects are lacking. Determination of the serum concentration of such compounds can help to guide adjustment of dosage during their therapeutic use. By measuring the serum level of drugs one bypasses the largest source of individual differences in doseeffect relationships — the pharmacokinetic variation between subjects. However, the relationship between the serum concentration of a drug and the intensity of its pharmacodynamic action is influenced by many other factors, which must always be considered in interpretation of serum levels. Therapeutic decisions should never be based solely on the serum concentration of a compound, nor can such measurements ever substitute for careful medical observation and judgement.     

基于群体药动学的奥卡西平儿童个体化给药模式的建立

目的基于群体药动学(PPK)和贝叶斯原理,建立奥卡西平的癫痫患儿个体化给药工作模式,促进临床合理用药。方法利用已发表的中国癫痫患儿口服奥卡西平后的PPK模型和JPKD-Bayesian软件建立奥卡西平个体化给药的预测模型。运用该模型对100例癫痫患儿进行个体化给药方案设计。患儿按设计方案规律服药2~4周后测定10-羟基卡马西平的稳态血清谷浓度并与模型预测值相比较,计算平均预测误差、平均绝对预测误差、平均相对预测误差、相对预测误差在±20%和±30%内的比例来验证模型的预测性能。结果个体化给药预测模型的平均预测误差为(0.54±2.00)mg·L~(-1),平均绝对预测误差为(1.75±1.09)mg·L~(-1),平均相对预测误差为(3.86±14.56)%,其中分别有78%和96%的血药浓度数据相对预测误差在±20%和±30%以内。血药浓度预测值对实测值的决定系数R2=87.8%。上述验证结果说明模型的预测准确度和精密度均较高。结论本研究成功建立了可用于儿科患者实施奥卡西平个体化给药的预测模型和完整的工作模式,有助于临床合理用药。     

Comparison of non-kinetic and kinetic approaches to individualization of gentamicin dosage

Summary A prospective study was carried out in 40 acutely ill patients to compare the non-kinetic and kinetic approaches to individualization of the dosage regimen of gentamicin. The patients were divided into two equal groups. For the non-kinetic group, the doses were derived from the physician's personal experience, on a mg/kg basis, and by use of nomograms. The total daily dose ranged from 1.43 to 4.5 mg/kg. Based on serum concentration measurements, the dosage regimen for individual patient was calculated by Sawchuk-Zaske's method. The calculated doses were compared to the prescribed doses in each patient. Of the patients on empirically prescribed doses 65% received 36% more drug than the calculated dose and 20% received 36% less than the calculated dose. The calculated dosing intervals were greater than the recommended intervals in 60% of the patients. The gentamicin trough concentration was > 2 g/ml in 70% of the patients. There was a significant tendency to overdosage of the patients.For the kinetic group, following administration of the calculated dose, the steady-state peak and trough concentrations in each patient were measured. The correlation of measured to predicted steady-state serum concentrations was excellent (r=0.9968, p<0.05). About 85% of the served trough concentrations and 90% of the peak values fell within the therapeutic range. The mean of the prediction error (ME), mean absolute error (MAE), mean squared error (MSE), and root mean squared error (RMSE) of the trough and peak concentrations were calculated. The 95% confidence interval of the ME for the trough and peak concentrations included zero, which shows that the prediction was not significantly biased.A significant relationship between gentamicin clearance and the ratio of the peak and trough concentrations achieved to the administered dose (r=0.873, 0.916 for trough and peak, respectively) was found.The findings suggest that the individualized approach to dosage determination using pharmacokinetic principles, in conjunction with daily monitoring of serum gentamicin concentrations, may provide safe and effective therapy.     

The random-effects model applied to refractory ceramic fiber data

Refractory ceramic fiber (RCF) is a valuable, high-temperature, insulating material with a variety of industrial uses. Because some fibers are respirable by humans and RCF is relatively durable in simulated lung fluids, RCF may pose a health hazard in the workplace. The RCF industry has established a comprehensive product stewardship program (PSP) to identify, quantify, and manage risks. One key element of this PSP is a workplace monitoring program. This paper analyzes monitoring data collected as part of a Consent Agreement with the U.S. Environmental Protection Agency over the period from 1993 to 1998. More specifically, this paper applies the random-effects model (REM) to data collected at several Unifrax plants and applicable to several groups of workers. The REM fits the RCF data well. Depending upon the plant and the functional job category values of the variance of the log-transformed time-weighted average workplace concentrations range from slightly less than 0.5 to 1.0. The estimated intraclass correlations (ratio of the between-worker variance to the total variance) were less than 0.4, and most were less than 0.2. Implications of these findings are examined. Use of the REM in the development of a workplace respiratory policy is described. Finally, two possible criteria for measuring compliance with an occupational exposure limit are reviewed: an "overexposure" criterion developed by Rappaport and co-workers and a conventional "no exceedance" criterion reportedly used by regulatory agencies. The overexposure criterion is logically correct for potential toxicants with chronic effects. For representative values of statistical parameters for RCF from the plants considered, the overexposure criterion is less stringent.     

CYP3A酶活性测定方法与依托泊苷个体化给药方案

目的：探索白血病患者体内依托泊苷(VP16)浓度与CYP3A酶活性(CA)的相关性。方法：采用HPLC法分别测定20例白血病患者血浆中VP16浓度及尿液CA，进行相关性分析。结果：白血病患者体内VP16浓度与CA无相关性，而与CA的常用对数(1gCA)线性相关，相关系数r=0．969。结论：通过测定CA制定VP16给药方案是可行的。     

Evaluation of population pharmacokinetic models for amikacin dosage individualization in critically ill patients

Objectives The aim of this study was to evaluate the reliability for dosage individualization and Bayesian adaptive control of several literature‐retrieved amikacin population pharmacokinetic models in patients who were critically ill. Methods Four population pharmacokinetic models, three of them customized for critically‐ill patients, were applied using pharmacokinetic software to fifty‐one adult patients on conventional amikacin therapy admitted to the intensive care unit. An estimation of patient‐specific pharmacokinetic parameters for each model was obtained by retrospective analysis of the amikacin serum concentrations measured (n = 162) and different clinical covariates. The model performance for a priori estimation of the area under the serum concentration‐time curve (AUC) and maximum serum drug concentration (C_max) targets was obtained. Key findings Our results provided valuable confirmation of the clinical importance of the choice of population pharmacokinetic models when selecting amikacin dosages for patients who are critically ill. Significant differences in model performance were especially evident when only information concerning clinical covariates was used for dosage individualization and over the two most critical determinants of clinical efficacy of amikacin i.e. the AUC and C_max values. Conclusions Only a single amikacin serum level seemed necessary to diminish the influence of population model on dosage individualization.     

Confounding issues in estimation of patient-specific pharmacokinetic parameters and dosage individualization of aminoglycosides

Aminoglycoside antibiotics are usually administered by multiple short intravenous infusions at fixed intervals. Today, equations reported 35 years ago by Sawchuk and Zaske are still the cornerstone of methods used for determination of patient-specific pharmacokinetic parameters of aminoglycosides and individualization of drug dosage regimens in many clinical settings. Additionally, these methods are included in many clinical pharmacology curricula in pharmacy and other related fields. However, there are a few issues with regard to the application and/or modification of this method in clinical settings, which may result in some confusion among novice clinicians. For example, serum samples collected from different intervals at steady state, instead of samples obtained during the same interval, require special manipulation of sampling time before they can be used for estimation of pharmacokinetic parameters. Furthermore, there are various ways that the original equations are modified or simplified, which can result in some degree of error in the estimates of pharmacokinetic parameters and ensuing dosage regimen calculations. Simulation data presented here indicate that in some cases, these errors may be substantial, depending on the length of short infusion, half life of the drug, and the dosage interval. For instance, using equations developed for intravenous bolus mode of administration, ignoring the short infusion, may result in ≥ 25% error for a typical patient and dosing scenario. Although experts may use modified equations, understanding their error ramifications, these modifications may be confusing to the novice clinician. Therefore, it is recommended that exact equations developed specifically for multiple intravenous infusions be used without any modification, particularly in settings where clinicians are being trained.     

Drug polymorphism and dosage form design: a practical perspective

Formulators are charged with the responsibility to formulate a product which is physically and chemically stable, manufacturable, and bioavailable. Most drugs exhibit structural polymorphism, and it is preferable to develop the most thermodynamically stable polymorph of the drug to assure reproducible bioavailability of the product over its shelf life under a variety of real-world storage conditions. There are occasional situations in which the development of a metastable crystalline or amorphous form is justified because a medical benefit is achieved. Such situations include those in which a faster dissolution rate or higher concentration are desired, in order to achieve rapid absorption and efficacy, or to achieve acceptable systemic exposure for a low-solubility drug. Another such situation is one in which the drug remains amorphous despite extensive efforts to crystallize it. If there is no particular medical benefit, there is less justification for accepting the risks of intentional development of a metastable crystalline or amorphous form. Whether or not there is medical benefit, the risks associated with development of a metastable form must be mitigated by laboratory work which provides assurance that (a) the largest possible form change will have no substantive effect on product quality or bioavailability, and/or (b) a change will not occur under all reasonable real-world storage conditions, and/or (c) analytical methodology and sampling procedures are in place which assure that a problem will be detected before dosage forms which have compromised quality or bioavailability can reach patients.     

Amoxycillin release from a floating dosage form based on alginates

Floating alginate beads have been prepared from alginate solutions containing either dissolved or suspended amoxycillin. The beads were produced by the dropwise addition of the alginate into calcium chloride solution, followed by removal of the gel beads and freeze drying. Drug release studies showed that beads prepared with the drug in solution provided some sustained release characteristics and that these could be improved by the addition of amylose. In all cases, the drug release was consistent with release of a dissolved solute from a granular or porous matrix. The beads retained their buoyancy when amylose and amoxycillin were incorporated, exhibiting resultant weight values greater than zero after 20 h. Preparation of the beads from alginate solutions containing the drug in suspension allowed higher drug loadings, at the expense of faster release and lower buoyancy.     

Adolescent alcohol use and adult alcohol disorders: a two-part random-effects model with diagnostic outcomes

Alcohol use is often analyzed by treating the behavior as a single dimension, such as focusing on frequency of use. Based on data from a longitudinal study, this report considers two distinct aspects of semi-continuous alcohol use data. A two-part random-effects model was used to evaluate change in the log-odds and frequency of use from about age 13 to about age 18 years. Change features were then related to the log-odds of later alcohol disorders. Results suggested differences in the two aspects of use over time and their relationships with later disorders. Most important for the purposes of this study, different methods of analyzing antecedents and consequences of alcohol use trajectories were shown to generate both similar and disparate findings.     

Drug interactions between HIV protease inhibitors based on physiologically-based pharmacokinetic model

A Physiologically-based pharmacokinetic (PB-PK) model was developed to describe the aspects of pharmacokinetic interactions between five HIV protease inhibitors (ritonavir, amprenavir, nelfinavir, saquinavir, indinavir) in rats. To increase usefulness of this BP-PK model, liver, intestinal tissue and other organ were assumed as compartments in this model. Each compartment was linked with the blood flow and the blood-to-plasma concentration ratios of those drugs, and the absorption process in the intestinal tract was presumed as a first-order kinetics. In addition, this PB-PK model incorporates two elimination processes due to hepatic and intestinal metabolism constructed by in vitro metabolic clearance rates and inhibition constants between HIV protease inhibitors. Excellent agreements were obtained between the predicted and observed concentrations of HIV protease inhibitors in rat plasma after a 20 mg/kg oral dose or co-administration of two kinds of HIV protease inhibitors (amprenavir/indinavir, nelfinavir/amprenavir, saquinavir/amprenavir, amprenavir/ritonavir, indinavir/ritonavir, nelfinavir/ritonavir, and saquinavir/ritonavir) with each 20 mg/kg oral dose. However, underestimates in the predicted plasma concentrations of saquinavir, indinavir and amprenavir were observed during the terminal phase after co-administration with ritonavir or amprenavir, suggesting that a term of other inhibitory process, such as a mechanism-based inhibition, might be incorporated into this PB-PK model. This BP-PK model enables us to know useful information about pharmacokinetic interaction when HIV infected patients would receive double protease therapy.     

Outcomes of weight-based heparin dosing based on literature guidelines and institution individualization

STUDY OBJECTIVE: To determine whether unfractionated heparin is optimally dosed using published weight-based guidelines. DESIGN: Six-month, prospective study. SETTING: University hospital. PATIENTS: Ninety-six patients in the weight-based unfractionated heparin-dosing group 1 (WBHD1; 37 men; mean age 66.9 +/- 15.1 years; mean weight 80.1 +/- 20.6 kg) and 68 patients in the WBHD2 (25 men; mean age 68.2 +/- 15.6 years; mean weight 82.0 +/- 19.6 kg). INTERVENTIONS: The WBHD1 received a 100-U/kg intravenous bolus followed by an 18-U/kg/hour continuous intravenous infusion. After 3 months, the protocol was modified, and the WBHD2 received a 90-U/kg bolus followed by a 16-U/kg/hour continuous infusion for 3 months. MEASUREMENTS AND MAIN RESULTS: Activated partial thromboplastin times (aPTTs), frequency of bleeding episodes that required blood transfusions, and the number of recurrent thromboembolic events were collected from both groups after 3 months on the study. In the WBHD1, 24 hours after starting heparin, 38.5% of patients had therapeutic aPTTs, and at 48 hours, 54.3% were therapeutic. In the WBHD2, 42.6% and 51.4% of patients had therapeutic aPTTs at 24 and 48 hours, respectively. There was no statistical difference between the WBHD1 and WBHD2 in the percentage of patients with therapeutic aPTTs. CONCLUSIONS: Weight-based heparin dosing resulted in low percentages of patients with therapeutic aPTTs. The use of weight alone to dose heparin may not be adequate to optimize therapy.     

Risk-adjusted monitoring of veno-occlusive disease following Bayesian individualization of busulfan dosage for bone marrow transplantation in paediatrics

In order to assess the performance of Bayesian individualization of busulfan (BU) dosage regimens, veno-occlusive disease (VOD) rate was monitored for paediatric patients undergoing allogeneic bone marrow transplantation (BMT). Consecutive patients undergoing allogeneic BMT with BU as conditioning regimen during 5-years period (January 2000-February 2006) were reviewed. VOD was a major outcome variable. Preconditioning risk of VOD was estimated for each patient using a scoring system that included type of transplant, recipient CMV-positive status and total parenteral nutrition (TPN) provided pretransplantation. A risk-adjusted cumulative sum method was used to compare observed versus predicted outcome by assigning a risk score, based on log-likelihood ratios, to each patient. These cumulative scores were sequentially plotted with preset control limits for 'signalling' where results were substantially different than expected (doubling or halving of odds ratio). Sixty-six children received BMT after oral busulfan-based conditioning regimen with median age 3.9 years, 63.6% of male. Median preconditioning risk of VOD was 0.34 ranging from 0.23 to 0.84. Observed VOD rate was 16.7% (n = 11) which was 60.7% (17) fewer than the expected number estimated by the risk score. The resulting risk-adjusted score for each patient was plotted sequentially. This plot adopted early a negative slope, crossing the lower control limit twice, after 27 and 66 patients, indicating improved results compared to those expected. Bayesian individualization of oral busulfan dosage regimens is useful to reduce the VOD rate in children undergoing allogeneic BMT.     

Drug dosage in the elderly: dermatological drugs

Drug pharmacokinetics and pharmacodynamics may be altered in the elderly. An important contribution is made by decreased renal function, but biotransformation in the liver may also play a role. Commonly prescribed dermatological drugs such as methotrexate and cetirizine are likely to be eliminated more slowly in the elderly and potentially hepatotoxic drugs such as itraconazole and acitretin should be used with caution. Altered drug distribution as a result of body composition changes can lead to prolonged half-life or higher plasma concentrations of many drugs. Higher prevalence of adverse drug reactions and multidrug regimens, and large interindividual variability in drug response make drug dosage and administration in the elderly challenging. New immunobiological agents such as alefacept, efalizumab and etanercept, which are approved for treatment of psoriasis, seem to be as well tolerated in the elderly as in younger patients. A recommended approach when prescribing drugs to the elderly would be to start with a small initial dose and to reduce the number of drugs administered simultaneously. It is crucial to simplify the drug regimen as much as possible in order to enhance drug management in the elderly. To improve pharmacotherapy in the elderly, we review age-related changes in pharmacokinetics that are likely to play a role in dermatological practice.     

A random-effects mixture model for classifying treatment response in longitudinal clinical trials

A random-effects regression model that allows the random coefficients to have a multivariate normal mixture distribution is described for classifying treatment response in longitudinal clinical trials. The proposed model is capable of dealing with longitudinal data from unknown heterogeneous populations. As applied to longitudinal clinical trials, for example, the model can distinguish subgroups of treatment response. Use of the proposed model is illustrated by analyzing data from two psychiatric clinical trials. The first includes depressed patients assigned to drug treatment who are repeatedly measured in terms of their level of depression. The second trial examined schizophrenic patients longitudinally who were assigned to either a drug or placebo condition. For both, the random-effects mixture model allows an assessment of whether patients comprise distinct populations in terms of their treatment response. Based on parameter estimates of the mixture model, ample evidence for a mixture of response to treatment is observed for both datasets.     

Comparisons of thresholds for carcinogenicity on linear and logarithmic dosage scales

Comparisons on a linear and the Rozman logarithmic scale for dosage versus carcinogenicity in rodents are presented for methyl eugenol (ME), nitrosodiethylamine (NDEA), ethyl carbamate (EC) and 2-acetylaminofluorene (AAF). Each of these chemicals has been shown to be carcinogenic in experimental animals and, in addition, humans are regularly exposed to at least three of these compounds (ME, NDEA, EC) in foods. Although the source of adducts from AAF is not known, the aminofluorene (AF) adduct is present in humans. Plotted on the same graphs are either some doses from common foods (ME, NDEA, EC) or adducts (AF) on human haemoglobin, for perspective, with their thresholds for carcinogenesis in animals. Use of a linear scale when comparing doses administered to animals in studies of carcinogenicity with doses of those same chemicals to which humans are exposed does not provide useful, comparative information. On the other hand, the Rozman logarithmic scale for dose allows one to put these relative doses in perspective. It is also evident that forcing a linear extrapolation through the zero, zero origin does not agree with the experimental data. Further analyses for goodness of fit for these dose responses reveal that the dose response for three of these compounds (ME, NDEA, EC) appears to be linear with the logarithm of the dose. However, AAF appears to be linear with the logarithm of the dose for bladder, but not for liver. It is suggested that the high background incidence of tumours in the BALB/c StCrlfC3Hf/Nctr mouse liver may confound the interpretation of dose response from AAF carcinogenesis in mouse liver.     

中国汉族人群基于临床特征和基因型华法林个体化给药模型的研究

目的：评价中国汉族人群临床特征及基因型对华法林剂量的影响,并构建华法林给药模型,为临床华法林个体化给药提供参考。方法：按照设定标准选取某医院2011年1月至2013年10月行心脏瓣膜手术后接受华法林抗凝治疗并达到华法林稳定剂量的中国汉族人群203例,对纳入人群进行CYP2C9＊3、VKORC1-1639G/A基因多态性检测,结合基因型及患者临床特征,分析对华法林稳定剂量的影响,并采用多元逐步线性回归分析建立数学模型。结果：性别、吸烟、饮酒及高血压病史对华法林剂量无明显影响（P＞0.05）。华法林剂量与年龄呈负相关（r=-0.155,P=0.027）;与身高、体质量呈正相关（r=0.166、0.190,P=0.009、0.003）。CYP2C9＊3、VKORC1-1639G/A基因多态性对华法林剂量的影响在统计学上有统计学差异（P＜0.01）。拟合得到华法林给药模型D=2.855－1.173×CYP2C9（AC）＋0.020×W－0.024×A＋4.064×VKORC1（GG）＋1.486×VKORC1（GA）。结论：华法林的剂量受到年龄、身高、体质量及CYP2C9＊3、VKORC1-1639G/A基因多态性的影响,依据华法林给药模型可优化华法林个体化给药方案,但仍有待于临床进一步验证。