Michel Tod

Facilitation of Drug Evaluation in Children by Population Methods and Modelling

The pharmacokinetics and pharmacodynamics of drugs are different in adult and paediatric populations, the latter being particularly heterogeneous. These differences in pharmacokinetics and pharmacodynamics justify specific studies but raise a number of ethical and practical issues. The main practical difficulties to circumvent while performing clinical studies in children are the invasiveness of the procedures and the obstacles to patient recruitment. The invasiveness related to pain/anxiety and blood loss precludes the performance of classical pharmacokinetic studies in children in many instances, particularly in neonates and infants. Population approaches, which rely on pharmacokinetic-pharmacodynamic modelling, are particularly appealing in paediatric populations because these models can cope with sparse data. The relevance of population approaches to investigation of the dose-concentration-effect relationships and to qualitative/quantitative assessment of factors that may explain interindividual variability has already been emphasized.The aims of this review are to summarize the currently available literature on population pharmacokineticpharmacodynamic studies in children and to discuss a number of recent methodological developments that may facilitate the evaluation of drugs in this population by alleviating invasiveness and, subsequently, facilitating recruitment of patients. The present survey confirms that population approaches in paediatrics have already reached a large audience and that they are mostly used for analysis of sparse data. However, pharmacokineticpharmacodynamic studies in children are still scarce. New classes of models may extend the scope of the use of population models in paediatrics. Kinetic-pharmacodynamic models, where use of the term ‘kinetic’ rather than ‘pharmacokinetic’ emphasizes the absence of pharmacokinetic data, are indirect models where the (unobserved) drug kinetics are described by a single compartment involving a single rate constant. These models, which alleviate the need for blood samples used for the measurement of drug concentration, may be very useful in paediatric studies. Physiological and physiopathological models also have potential applications but require further development. Because the number of measurements in a single individual needs to be limited in children, it is crucial to optimize the design of the experiment in order to avoid inaccurate and unreliable results. In this review, formal optimization and simulation to evaluate a design are presented, and specific problems raised by the application of these techniques in paediatrics are addressed. Finally, the related technique of clinical trial simulation and its applications are presented and discussed.

Influence of burns on pharmacokinetics and pharmacodynamics of drugs used in the care of burn patients.

The pharmacokinetics and pharmacodynamics of drugs are significantly altered in the burn patient, and the burn patient population shows wide inter- and intraindividual variation in drug handling. Burn injury evolves in two phases. The first phase corresponds to the burn shock, which occurs during the first 48 hours after thermal injury. In this phase, hypovolaemia, oedema, hypoalbuminaemia and a low glomerular filtration rate are observed, which result in a slower rate of drug distribution and lower renal clearance. The second phase (beyond 48 hours after injury) is a hyperdynamic state with high blood flow in the kidneys and liver, an increased α1-acid-glycoprotein level and loss of the drug with exudate leakage. As a result, protein binding, drug distribution and clearance may be altered.Because of the alteration in these variables, wide intraindividual variation of pharmacokinetic parameters occurs depending upon the time since thermal injury and fluid resuscitation. Interindividual variations may be correlated with the percentage of the body surface area that is burnt, creatinine clearance, albuminaemia or the α1-acid-glycoprotein level. A number of important variations in pharmacodynamic parameters have been described, but their mechanisms are poorly understood.From a practical point of view, for the subpopulation of burn patients who eliminate drugs extremely rapidly, higher doses and/or shorter dosing intervals are required to avoid treatment inefficacy. Drug concentration measurements help to take into account interindividual variability. However, adaptation of doses based on Bayesian methods is frequently not possible because the distribution of pharmacokinetic parameters is poorly characterized in this population. Methods based only on individual data or on a surrogate marker for efficacy may be used to optimize the dosing regimen in this population.

Validation of an HPLC-UV method for sorafenib determination in human plasma and application to cancer patients in routine clinical practice.

Sorafenib, a new oral multikinase inhibitor with antiangiogenic properties, has demonstrated preclinical and clinical activity against several tumor types. The aims of this study were to validate a method for the measurement of sorafenib in plasma from cancer patients, then to test this method in clinical practice. Following liquid-liquid extraction, the compounds were separated with gradient elution (on a C18 ultrasphere ODS column using a mobile phase of acetonitrile/20 mM ammonium acetate), then detected at 255 nm. The calibration was linear in the range 0.5-20 mg/L. Intra- and inter-assay precision was lower than 7 and 10%, respectively, at 0.5, 3 and 20 mg/L. Plasma sorafenib concentrations were measured in 22 cancer patients (99 samples). The mean trough sorafenib concentration (C(min)) and concentration at peak were 4.3+/-2.5 mg/L (n=68, CV=57.5%) and 6.2+/-3.0 mg/L (n=31, CV=47.5%), respectively. Mean sorafenib C(min) in eight patients who experienced grade 3 drug-related adverse events was approximately 1.5-fold greater than that observed in the remaining patients (7.7+/-3.6 mg/L vs. 4.4+/-2.4 mg/L, P=0.0083). In conclusion, the method was successfully used in routine practice to monitor plasma concentrations of sorafenib in cancer patients. Finally, large interindividual variability and higher exposure in patients experiencing severe toxicity support the need for therapeutic drug monitoring to ensure an optimal exposure to sorafenib.

Early Sorafenib-Induced Toxicity Is Associated with Drug Exposure and UGTIA9 Genetic Polymorphism in Patients with Solid Tumors: A Preliminary Study

Background Identifying predictive biomarkers of drug response is of key importance to improve therapy management and drug selection in cancer therapy. To date, the influence of drug exposure and pharmacogenetic variants on sorafenib-induced toxicity remains poorly documented. The aim of this pharmacokinetic/pharmacodynamic (PK/PD) study was to investigate the relationship between early toxicity and drug exposure or pharmacogenetic variants in unselected adult outpatients treated with single-agent sorafenib for advanced solid tumors. Methods Toxicity was recorded in 54 patients on days 15 and 30 after treatment initiation and sorafenib exposure was assessed in 51 patients. The influence of polymorphisms in CYP3A5, UGT1A9, ABCB1 and ABCG2 was examined in relation to sorafenib exposure and toxicity. Clinical characteristics, drug exposure and pharmacogenetic variants were tested univariately for association with toxicities. Candidate variables with p<0.1 were analyzed in a multivariate analysis. Results Gender was the sole parameter independently associated with sorafenib exposure (p = 0.0008). Multivariate analysis showed that increased cumulated sorafenib (AUCcum) was independently associated with any grade ≥3 toxicity (p = 0.037); UGT1A9 polymorphism (rs17868320) with grade ≥2 diarrhea (p = 0.015) and female gender with grade ≥2 hand-foot skin reaction (p = 0.018). Using ROC curve, the threshold AUCcum value of 3,161 mg/L.h was associated with the highest risk to develop any grade ≥3 toxicity (p = 0.018). Conclusion In this preliminary study, increased cumulated drug exposure and UGT1A9 polymorphism (rs17868320) identified patients at high risk for early sorafenib-induced severe toxicity. Further PK/PD studies on larger population are warranted to confirm these preliminary results.

Robust Optimal Design for the Estimation of Hyperparameters in Population Pharmacokinetics

The expectation of the determinant of the inverse of the population Fisher information matrix is proposed as a criterion to evaluate and optimize designs for the estimation of population pharmacokinetic (PK) parameters. Given a PK model, a measurement error model, a parametric distribution of the parameters and a prior distribution representing the belief about the hyperparameters to be estimated, the EID criterion is minimized in order to find the optimal population design. In this approach, a group is defined as a number of subjects to whom the same sampling schedule (i.e., the number of samples and their timing) is applied. The constraints, which are defined a priori, are the number of groups, the size of each group and the number of samples per subject in each group. The goal of the optimization is to determine the optimal sampling times in each group. This criterion is applied to a one-compartment open model with first-order absorption. The error model is either homoscedastic or heteroscedastic with constant coefficient of variation. Individual parameters are assumed to arise from a lognormal distribution with mean vectorMand covariance matrixC. Uncertainties about theMandCare accounted for by a prior distribution which is normal forMand Wishart forC. Sampling times are optimized by using a stochastic gradient algorithm. Influence of the number of different sampling schemes, the number of subjects per sampling schedule, the number of samples per subject in each sampling scheme, the uncertainties onMandCand the assumption about the error model and the dose have been investigated.

A Tumor Growth Inhibition Model For Low-Grade Glioma Treated With Chemotherapy or Radiotherapy

Purpose: To develop a tumor growth inhibition model for adult diffuse low-grade gliomas (LGG) able to describe tumor size evolution in patients treated with chemotherapy or radiotherapy. Experimental Design: Using longitudinal mean tumor diameter (MTD) data from 21 patients treated with first-line procarbazine, 1-(2-chloroethyl)-3-cyclohexyl-l-nitrosourea, and vincristine (PCV) chemotherapy, we formulated a model consisting of a system of differential equations, incorporating tumor-specific and treatment-related parameters that reflect the response of proliferative and quiescent tumor tissue to treatment. The model was then applied to the analysis of longitudinal tumor size data in 24 patients treated with first-line temozolomide (TMZ) chemotherapy and in 25 patients treated with first-line radiotherapy. Results: The model successfully described the MTD dynamics of LGG before, during, and after PCV chemotherapy. Using the same model structure, we were also able to successfully describe the MTD dynamics in LGG patients treated with TMZ chemotherapy or radiotherapy. Tumor-specific parameters were found to be consistent across the three treatment modalities. The model is robust to sensitivity analysis, and preliminary results suggest that it can predict treatment response on the basis of pretreatment tumor size data. Conclusions: Using MTD data, we propose a tumor growth inhibition model able to describe LGG tumor size evolution in patients treated with chemotherapy or radiotherapy. In the future, this model might be used to predict treatment efficacy in LGG patients and could constitute a rational tool to conceive more effective chemotherapy schedules. Clin Cancer Res; 18(18); 5071–80. ©2012 AACR.

Pharmacodynamics and pharmacokinetics of antibacterial drugs in the management of febrile neutropenia

We review experimental and clinical data on the pharmacokinetics and pharmacodynamics of antibacterial drugs in febrile neutropenic hosts. Since major pharmacokinetic changes have been reported for various classes of antibiotics in these patients, we advocate the need for adequate initial dosing regimens in all cases. Monitoring drug serum concentrations is mandatory for aminoglycosides and glycopeptides, and special attention should be paid to the dosing frequency of the short half-life beta-lactams to optimise the management of febrile neutropenia, especially in patients with severe sepsis.

Individualising aminoglycoside dosage regimens after therapeutic drug monitoring: simple or complex pharmacokinetic methods?

Measurements of aminoglycoside concentration in serum are used to individualise dosage regimens (dose per administration and/or administration interval) with the goal of attaining the desired therapeutic range as quickly as possible. Therapeutic range is defined in terms of peak concentration (to monitor effectiveness) and trough concentration (to avoid toxicity). This article focuses on methods to individualise aminoglycoside dosage regimens in the context of extended dosage intervals.Simple pharmacokinetic methods involve linear dosage adjustment based on peak or trough concentrations or area under the concentration-time curve, or nomograms. The once daily aminoglycoside nomogram determines the dosage interval for aminoglycosides given as a fixed dose per administration, based on a single concentration measurement drawn 6 to 14 hours after the start of the first infusion. This is a preferred method because of its simplicity, strong pharmacodynamic rationale and prospective validation in a large population. However, it does not work when the fixed dose assumed is not relevant, for example for patients with burns, cystic fibrosis, ascites or pregnancy. Furthermore, it has not been validated in children. In these cases, a more sophisticated method is required.Complex pharmacokinetic methods require dedicated software. Non-Bayesian least-squares methods allow the optimisation of both the dose and the dosage interval, but require aminoglycoside concentrations from two or more samples taken in the post-distributive phase during a single dosage interval. With Bayesian least-squares methods, only one concentration measurement is required, although any number of samples can be taken into account. In the Bayesian maximum a posteriori (MAP) method, the parameter estimates are taken as the values corresponding to the maximum of the posterior density. In ‘full’ Bayesian approaches (also called stochastic control), all the information about the parameters revealed by the posterior distribution is taken into account, and the optimal regimen is found by minimising the expected value of the weighted sum of squared deviations between predicted and target concentrations.If the population model is reasonably well known, Bayesian methods (MAP or stochastic control) should be used because of their good predictive performance. Although only one concentration measurement is required, better precision is afforded by a two-sample strategy, preferably drawn 1 and 6 hours after the start of the first infusion. If the population model is not known, then the non-Bayesian least-squares method is the method of choice, because of its robustness and lack of requirement for prior information about the distribution of parameters in the population.

Development and validation of an HPLC-UV-visible method for sunitinib quantification in human plasma.

BACKGROUND Sunitinib malate is a novel oral multitargeted tyrosine kinase inhibitor with antitumor and antiangiogenic activities. Only mass spectrometry detection is currently available to determine sunitinib in human plasma. The purpose of this study was to develop a simple and sensitive high-performance liquid chromatographic method with UV-Visible detection for quantification of sunitinib concentrations in human plasma. METHODS After a liquid-liquid extraction with ethyl acetate, sunitinib and ranitidine (internal standard) are separated on cyanopropyl column using a simple binary mobile phase of ammonium acetate buffer (20 mM; pH 6.8):acetonitrile (55:45,v/v). Samples were eluted isocratically at a flow rate of 1 mL/min throughout the 10 min run. Dual wavelength mode was used, with ranitidine monitored at 255 nm, and sunitinib at 431 nm. RESULTS The calibration was linear in the range 20-200 ng/mL. Inter- and intra-day coefficients of variation were less than 7%. This method is sensitive, accurate and selective. It has been successfully implemented to monitor trough sunitinib concentrations in plasma samples (n = 39) from 14 unselected cancer patients treated with the recommended once daily dose of 50 mg or less. CONCLUSION This method can be used in routine clinical practice to monitor plasma sunitinib concentrations in cancer patients treated with once daily administration.

Implementation of OSPOP, an algorithm for the estimation of optimal sampling times in pharmacokinetics by the ED, EID and API criteria

The most common approach to optimize the sampling schedule in parameter estimation experiments is the D-optimality criterion, which consists in maximizing the determinant of the Fisher information matrix (max det F). In order to incorporate prior parameter uncertainty in the optimal design, other criteria have been proposed: The ED = max E (det F), EID = min E (l/det F) and API = max E (log det F) criteria, where the expectation is with respect to the given prior distribution of the parameters. Previously described algorithm for the estimation of optimal sampling times according to these criteria are adaptive random search (ARS), a robust and global but slow optimizer for API, and stochastic gradient (SG), a fast but local optimizer for ED and EID. We implemented an algorithm named OSPOP 1.0, based on non-adaptive random search (RS) followed by stochastic gradient to determine optimal sampling times for parameter estimation in various pharmacokinetic models according to ED, EID and API criteria. Prior distributions are allowed to be uniform, normal or lognormal. This algorithm combines the robustness of RS and the speediness of SG (convergence is obtained in a few minutes on a microcomputer). The results of the SG algorithm have been compared to those described in the literature using the ARS algorithm on a one compartment model with first- order absorption and were very similar. Also, the CPU time needed by SG and ARS algorithms were compared and the former proved to be much faster. Then, it has been applied to a five parameters stochastic model with zero-order absorption rate and Weibull-distributed residence times which was shown to describe adequately the kinetics of metacycline in humans. Population pharmacokinetic parameters of metacycline were estimated from a six subject pilot study, by the iterative two-staged method, using ADAPT II repeatedly. Optimal sampling times were determined with each criterion (ED, EID, API) with a multivariate normal prior parameter distribution. Six to seven distinct sampling times could be estimated. Higher numbers of samples revealed coalescing of design points.