|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2002 American Society for Clinical Oncology Population Pharmacokinetics of the Novel Anticancer Agent E7070 During Four Phase I Studies: Model Building and ValidationFrom the Department of Pharmacy and Pharmacology and Department of Medical Oncology, the Netherlands Cancer Institute/Slotervaart Hospital; and NDDO Oncology/Free University Hospital, Amsterdam; UMC St Radboud, Nijmegen; and Faculty of Pharmacy, Utrecht University, Utrecht, the Netherlands; Institut Gustave Roussy, Villejuif, Centre Claudius Regaud, Toulouse; Centre Regional Leon Berard, Lyon; and Centre René Gauducheau, Nantes, France; LBI-ACR and KFJ-Spital, Vienna, Austria; University Hospital Gasthuisberg, Leuven, Belgium; and Eisai Ltd, London, United Kingdom. Address correspondence to Ch. van Kesteren, MD, Department of Pharmacy and Pharmacology, the Netherlands Cancer Institute/Slotervaart Hospital, Louwesweg 6, 1066 EC Amsterdam, the Netherlands; email: apcks{at}slz.nl
PURPOSE: N-(3-Chloro-7-indolyl)-1,4-benzenedisulfonamide (E7070) is a novel sulfonamide anticancer agent currently in phase II clinical development for the treatment of solid tumors. Four phase I studies have been finalized, with E7070 administered at four different treatment schedules to identify the maximum-tolerated dose and the dose-limiting toxicities. Pharmacokinetic analyses of all studies revealed E7070 to have nonlinear pharmacokinetics. A population pharmacokinetic model was designed and validated to describe the pharmacokinetics of E7070 at all four treatment schedules and to identify the possible influences of patient characteristics on the pharmacokinetic parameters. PATIENTS AND METHODS: Plasma concentration-time data of all patients (n = 143) were fitted to several pharmacokinetic models using NONMEM. Seventeen covariables were investigated for their relation with individual pharmacokinetic parameters. A bootstrap procedure was performed to check the validity of the model. RESULTS: The data were best described using a three-compartment model with nonlinear distribution to a peripheral compartment and two parallel pathways of elimination from the central compartment: a linear and a saturable pathway. Body-surface area (BSA) was significantly correlated to both the volume of distribution of the central compartment and to the maximal elimination capacity. The fits of 500 bootstrap replicates of the data set demonstrated the robustness of the developed population pharmacokinetic model. CONCLUSION: A population pharmacokinetic model has been designed and validated that accurately describes the data of four phase I studies with E7070. Furthermore, it has been demonstrated that BSA-guided dosing for E7070 is important.
SULFONAMIDES HAVE shown a wide variety of pharmacologic activity, such as antibacterial, antidiabetic, carbonic anhydrase inhibitory, and antithyroid activity.1 In the search for potent anticancer agents, a novel class of sulfonamides have been synthesized that block cell cycle progression in the G1 phase.2 One of those compounds, N-(3-chloro-7-indolyl)-1,4-benzenedisulfonamide (E7070) (Fig 1), showed potent cytotoxic activity both in vitro (using human cell lines) and in vivo with human tumor xenografts (eg, HCT116 colon carcinoma and LX-1 non–small-cell lung carcinoma). E7070 exerts its antitumor activity by inhibiting the activation of cyclin-dependent kinase 2 and cyclin E, which are required for the transition of the G1 to S phase in the cell cycle.2,3
On the basis of this unique mechanism of action and the promising results from the preclinical studies, a phase I clinical program was conducted with E7070 administered by means of four different treatment schedules.4-8 The aim of these phase I studies was to identify the maximum-tolerated dose and the dose-limiting toxicities for all treatment schedules. Currently, all studies are being finalized and recommended doses for further phase II studies have been defined. The dose-limiting toxicities were neutropenia, thrombocytopenia, and anemia. Other toxicities were stomatitis, fatigue, alopecia, and local reactions at the site of drug administration. E7070 showed antitumor activity in patients with breast, endometrial, renal, and ovarian carcinoma. Furthermore, several patients experienced prolonged stabilization of their disease. Pharmacokinetic research during the phase I studies demonstrated that with an increasing dose, the increase in exposure to E7070, expressed as the area under the plasma concentration-time curve (AUC), was disproportionate.4-8 These results were obtained in all four phase I studies, suggesting that saturable processes occur in the disposition of E7070 at higher dose levels, independent of the administration schedule. The population pharmacokinetic approach provides a valuable tool for obtaining further insight into the pharmacokinetic behavior of new compounds, and the use of this technique in new drug development has been advocated.9-11 At an early stage of clinical development, the influence of patient characteristics and biochemical markers on pharmacokinetic parameters can be identified. The aim of the present study was to develop a population pharmacokinetic model that provides an accurate description of the full pharmacokinetic profile of E7070 at all dose levels for all four treatment schedules. Furthermore, patient characteristics were investigated for their influence on the pharmacokinetics. As the developed model will serve as a basis for further investigation of, for example, the pharmacokinetic-pharmacodynamic relationships, the validity of the developed model was checked using a bootstrap procedure.
Patient Population and Data Collection The data analyzed were obtained from four phase I dose-escalation studies performed within the framework of the European Organization for Research and Treatment of Cancer. The studies evaluated four different intravenous treatment schedules: a 1-hour infusion, every 3 weeks; a daily times 5, 1-hour infusion, every 3 weeks; a weekly times 4, 1-hour infusion, every 6 weeks; and a continuous infusion over 5 days (CIV), every 3 weeks. Table 1 lists the number of patients and the evaluated dose range for each study.
Briefly, eligibility criteria included the diagnosis of a solid tumor not amenable to established forms of treatment; minimum age of 18 years; World Health Organization performance status of  2; a life expectancy of at least 3 months; and adequate renal, hepatic, and bone marrow function. All patients gave written informed consent and the study protocols were approved by the medical ethics committee in all study centers. Patient characteristics are listed in Table 2.
Serial blood sampling was performed during the first treatment cycle in order to characterize the pharmacokinetic profile of E7070 with each treatment regimen. In the 1-hour infusion every 3 weeks study, samples were taken up to 120 hours after the end of infusion. Both day 1 and day 5 were obtained for the daily times 5, 1-hour infusion study, and intervening the peak and trough levels. In the weekly times 4, 1-hour infusion, every 6 weeks schedule, a pharmacokinetic profile was obtained up to 96 hours after the administration of E7070 on week 1 and week 4. For the CIV schedule, samples were taken regularly during the infusion and up to 48 hours after the end of infusion. High-performance liquid chromatography with ultraviolet detection was used for the quantification of E7070 in plasma (NOTOX, ’s-Hertogenbosch, the Netherlands). The method was linear between 0.02 and 0.50 µg/mL, with a lower limit of quantification of 25 ng/mL. The within-batch and between-batch accuracy and precision were less than 18.8%. Briefly, 0.5 mL plasma was supplemented with 1.0 mL of 0.1 mol/L phosphate buffer (pH 5.0) and vortex mixed for 5 seconds. After the addition of 3 mL diethyl ether, the container was shaken for 10 minutes before centrifugation at 1,500 x g for 5 minutes. The upper layer was pipetted off and added to 200 µL 99:1 diethyl ether/glycerol. The organic phase was evaporated to dryness under a nitrogen stream at 30°C over approximately 20 minutes. This residue was dissolved in 200 µL mobile phase. A 100-µL aliquot of this solution was injected onto the chromatographic system.
Population Pharmacokinetic Analyses
Interindividual variability for the pharmacokinetic parameters was modeled using a proportional error model. For example, variability in V (volume of distribution) was estimated using Vi = Vpop x (1 +  i), where Vi represents the V of the ith individual, Vpop is the population value, and is the interindividual random effect with mean 0 and variance  2. The difference between the jth observed concentration in the ith individual (Cobsij) and its respective prediction (Cpredij) (ie, the residual variability) was modeled with a combined additive-proportional error model: Cobsij= Cpredij(1 +  2) + 1, where 1 (additive component) and 2 (proportional component) are random effects with mean 0 and variance  2. Because of the computational intensity of the model using the first-order conditional estimation method, the first-order method12 was used throughout the analyses. The adequacy of the developed structural models was evaluated using goodness-of-fit plots15,16 and precision of parameter estimates. Xpose (Version 2.0), an S-PLUS–based (Version 2000 Professional Release I, MathSoft, Inc, Cambridge, MA) model building aid,17 was used for the graphical goodness-of-fit analyses. The objective function value (OFV) provided by NONMEM was used for the comparison of the models. Discrimination between hierarchical models was based on the OFV using the log-likelihood ratio test.12 A value of P = .001, representing a decrease in OFV of 10.8, was considered statistically significant (df = 1).
Individual Bayesian estimates of the pharmacokinetic parameters were obtained using the POSTHOC option in NONMEM; for each subject, individual pharmacokinetic parameters were calculated taking both individual observations and population effects into account. The relationships between the individual pharmacokinetic parameter estimates and the covariates were visually inspected and investigated in NONMEM using a stepwise procedure.18,19 The covariates tested with their range in the population are listed in Table 2. For 28 patients, values of certain covariables were not available. Missing values of continuous variables (in 11 patients) were replaced by the corresponding median value of the total population. The missing dichotomous variables (in 18 patients) were consecutively substituted by 0 and by 1, and the covariate selection procedure was performed for both situations. It appeared that there was no difference in the inclusion of covariates between the two situations. A generalized additive modeling procedure (GAM) was applied to select explanatory variables. Calculations were performed using Xpose.17,18 Covariates that correlated significantly with the pharmacokinetic parameters, as indicated by the Akaike information criterion, were selected for testing in NONMEM. Continuous variables were centered to their median values. For example, the relationship between V and body weight (WT) was modeled as follows: Vpop =
Dichotomous variables were modeled using: Vpop =
Statistical Refinement
Model Validation
In total, 150 patients were treated in the four studies, and complete pharmacokinetic profiles from 144 patients were obtained (Table 1). The data from one patient receiving the CIV schedule were excluded, as the infusion was not administered continuously and the duration of the infusion could therefore not be reliably determined. In total, 3,314 plasma levels were available for the analyses. On average, approximately 23 samples were available per patient (range, eight to 33 samples). Figure 2 shows typical plasma concentration-time profiles of the four treatment schedules at doses where the nonlinear pharmacokinetics were observed.
Several pharmacokinetic models were tested. The first model comprised three compartments with saturable elimination from the central compartment. With this model, however, plasma concentrations in the first distribution phase of the curve (approximately 50 mg/L) were underestimated. Therefore, a saturable distribution pathway to one of the peripheral compartments was added to the model (Fig 3). With this model, a marked decrease in OFV was obtained as compared to the former model (
The GAM analyses and the visual inspection of the relations between the individual Bayesian pharmacokinetic parameters and the covariates indicated several possible explanatory covariables: alkaline phosphatase, total protein, and WT for V; age for Tm; body-surface area (BSA), total protein, LDH, sex, and pleural effusion for Vmax; and total protein, BSA, and pleural effusion for K10 (elimination rate constant). These covariates were separately tested for significance in NONMEM, using a univariate procedure. Three relationships were significant: WT versus V, BSA versus Vmax, and BSA versus K10, and these were included simultaneously in the model. However, it appeared that the relationship between BSA and K10 could no longer be estimated (parameter approximated zero) and this was excluded from the model. The other two relations (WT v V and BSA v Vmax) remained in the model and this was considered the intermediate model. In the backward step, these two relationships were excluded from the model, one by one. However, the exclusion of both relationships produced a significant increase in OFV (105 and 38 points, respectively); therefore, these covariates remained in the model. As BSA is a function of WT, it was tested whether WT and BSA could be interchanged as covariables for V and Vmax. It appeared that replacement of WT by BSA as an explanatory variable for V resulted in a comparable significance. Therefore, the final model comprised relations between V and BSA and between Vmax and BSA. The relations were described according to the following: V = 6.51 * (1 + 0.646 * [BSA - 1.76]), and Vmax = 2.55 * (1 + 0.528 * [BSA - 1.76]). These equations indicate that, for example, a patient with a BSA of 2.0 m2 would have a V of approximately 1 (L) higher and a Vmax of 0.3 (mg/h) higher than an individual with median BSA (1.76 m2). The inclusion of the covariates in the model resulted in a reduction of the interindividual variabilities in V and Vmax, both with 5%. The relationships between the individual Bayesian estimates of both V and Vmax and BSA are shown in Fig 4.
Statistical Refinement The correlations between the values of interindividual variability of V, Tmax, Tm, Vmax, K10, and the intercompartmental rate constants K21, K13, and K31 were evaluated. A correlation was observed between Tm and K21 and therefore the covariance of those parameters was added to the final model. The correlation coefficient between the two parameters was -0.87. The introduction of this parameter to the model resulted in a decreased OFV ( = -59, P < .001, df = 1) and therefore the parameter remained in the model. Inclusion of the correlation between Tm and K21 did not affect the significance of the covariate relationships. The final model is shown in Fig 3. Parameter estimates are listed in Table 3. All parameters were estimated with an acceptable coefficient of variation (3.4% to 55%). Interindividual variability could be quantified for V, Tmax, Tm, K21, K13, K31, Vmax, and K10. The interindividual variability was moderate for most parameters (26% to 55%), but considerable for both K21 (70%) and Tm (120%). For Km, interindividual variability could not be estimated. This should not be interpreted as an absence of variability for Km but simply that the data did not contain enough information to quantify this parameter. Residual variability was rather small, being 0.025 (mg/L) and 15.1%. This would result in a prediction error for a low concentration in the data set of 0.1 ± 0.04 mg/L and for the highest measured concentration 196 ± 30 (mg/L). In Fig 5, the model-predicted and the Bayesian individual–predicted plasma concentrations, on the basis of the final model, are plotted versus the observed plasma concentrations. The model-based predictions are symmetrically distributed around the line of identity (Fig 5A), indicating that the model adequately describes the pharmacokinetic profile of E7070 in the four treatment schedules. Furthermore, the current model enables an accurate prediction of the individual pharmacokinetic profile (Fig 5B).
Model Validation From the original data set, 500 replicate data sets were generated and used for the evaluation of the stability of the final model. Table 3 lists the results of the bootstrap procedure, presented as means and percent coefficient of variation, and the parameter estimates of the final model with the corresponding percent coefficient of variation. Mean values of the bootstrap procedure were similar to the parameter estimates of the original data set, indicating that the developed model is stable. Using the population pharmacokinetic parameters, the dependence of the total E7070 clearance (including linear and nonlinear processes) on the plasma concentration was evaluated. The results are shown in Fig 6. At concentrations below 0.1 mg/L, total clearance is independent of the plasma concentration and almost equals the nonlinear clearance. Here, the nonlinear pathway is not saturated and is faster than the linear pathway. However, at concentrations above 0.1 mg/L, the nonlinear component becomes saturated and clearance is highly dependent on the plasma concentration. The slower, linear clearance dominates at concentrations above 10 mg/L, where the nonlinear clearance is reduced to a minimal value.
The covariate analysis in this study revealed that BSA partially explained the interpatient variability in both V and Vmax, two pharmacokinetic parameters that greatly influence the AUC of E7070. Therefore, the potential benefit of BSA-guided dosing for E7070 was evaluated for two patients treated with a 1-hour infusion every 3 weeks at a dose level of 700 mg/m2, the recommended phase II dose for this schedule. These two patients had the smallest and the largest value for BSA within the current data set (ie, 1.34 and 2.36 m2). Both patients were given a fixed dose and a dose determined on the basis of their BSA value, and AUC values were predicted for both treatments using the developed model. The fixed dose was calculated by multiplying the dose level (700 mg/m2) by the median BSA value in the data set (1.76 m2). The results (Table 4) show that when the patient with a small BSA value is treated with a fixed, median dose, this results in an AUC that is 1.6 times higher than the AUC obtained after the dose derived from BSA. For the patient with the larger BSA value, the total exposure to E7070 would be reduced to approximately 60% with the fixed dose as compared with the BSA-based dose. The difference in achieved AUC values between the two patients was markedly reduced in the case of the BSA-guided dosing.
The population pharmacokinetic approach is increasingly recognized as a valuable tool in drug development.9-11 The major advantages of this technique include the ability to describe more complex pharmacokinetic models, to quantify interindividual variability, and to identify the influence of patient characteristics on this variability. Furthermore, sparse and dense data, multiple dose levels, and different treatment schedules can be analyzed simultaneously. In oncology, population pharmacokinetic methods are increasingly applied. Launay-Iliadis et al,22 Bruno et al,23,24 and Baille et al25 clearly demonstrated the benefits of the technique during the clinical development of docetaxel. Furthermore, it has been applied for multiple other cytotoxic drugs such as topotecan26 and etoposide27 and for the analyses of phase I data.28,29 In addition, mechanism-based pharmacokinetic models have been developed (eg, for ifosfamide,30 cyclophosphamide, and thiotepa31). For the novel sulfonamide anticancer agent E7070, it is of considerable importance that the pharmacokinetic profile is well understood, as the compound displays nonlinear pharmacokinetics at clinically relevant doses. In previously published reports, the use of noncompartmental methods resulted in the estimation of apparent values for clearance and half-lives,4-8 and these methods provided an inaccurate description of the complex concentration-time profile. Therefore, the more flexible population pharmacokinetic approach was applied. The data from the four phase I studies conducted with E7070 were combined for the present analysis, providing a data set encompassing a wide range of doses and treatment schedules. The availability of such a rich data set is likely to yield a robust model. The developed model comprises three compartments with both saturable and linear elimination from the central compartment and saturable transport to one of the peripheral compartments. For all four treatment schedules, the model provided accurate predictions of the plasma concentrations. Furthermore, all parameters were estimated with acceptable precision, and residual variability was small. The elimination of E7070 was found to be readily saturable at clinically relevant concentrations. Nevertheless, the presence of a linear elimination pathway prevents the occurrence of an unlimited accumulation of E7070. The inclusion of relevant covariables in the model has been performed according to a stepwise procedure.18,19 Seventeen covariables were investigated for their influence on the variability of eight pharmacokinetic parameters, which results in a large number of possible relationships to be investigated. To facilitate the selection of relevant covariables, a GAM analysis was used, allowing the number of models tested in NONMEM to be markedly reduced. A disadvantage of this approach is that the evaluated relationships are dependent on the quality of the individual Bayesian parameter estimates. However, in the present analysis, these parameters seemed to be accurate, as the individual concentrations were well predicted as indicated by the low residual error (15%). The final model included correlations between BSA and both V and Vmax. As E7070 displays nonlinear pharmacokinetics, both V and Vmax strongly influence the AUC of E7070. Further evaluations demonstrated that individual adaptation of the dose derived from a patient’s BSA markedly reduced interpatient variability for AUC. Therefore, E7070 appears to be one of the few anticancer agents for which the commonly used BSA-guided dosing is actually justified.32,33 The presented model accurately described the complicated pharmacokinetics of E7070 and elucidated possible mechanisms regarding the observed nonlinear pharmacokinetic behavior. The physiologic interpretation of both the saturable distribution and the two elimination pathways are currently unknown, although research has been performed on the metabolic fate and tissue binding of E7070. In a recent mass balance study with E7070,34 the metabolism and the RBC binding of the drug were studied using radiolabeled E7070. The study revealed that less than 2% of the total dose of E7070 was detected in both urine and feces, although the total recovery of radioactivity in urine and feces was over 60% and 20%, respectively. Therefore, it was concluded that E7070 is extensively metabolized and that multiple metabolic products are formed. It could be speculated that several metabolites are formed through saturable enzymatic processes whereas others are formed linearly, indicating that the developed model is plausible. Nevertheless, no correlation could be detected between liver function tests (eg, ALT and AST) and model parameters associated with elimination. During this study,34 the concentration profile of radioactivity was also quantified in RBCs. Concentrations reached in RBCs were lower than in plasma, especially the maximal radioactivity concentration and the time to reach maximal radioactivity concentration was longer than in plasma. Furthermore, the E7070 concentration in RBCs decreased more slowly than in plasma. These findings suggest that the distribution of E7070 to RBCs might be saturable and it may therefore be a physiologic explanation for the saturable distribution pathway of the proposed population pharmacokinetic model. The development of the current population pharmacokinetic model was undertaken for subsequent clinical application in, for example, the development of limited sampling models for use in phase II studies and to elucidate possible pharmacokinetic-pharmacodynamic relationships. Therefore, model validation is of great importance. Several methods to perform a model validation have been described,9,21 and in the current study the bootstrap resampling technique was applied as an internal validation. The 500 replicate data sets yielded mean model parameters that were comparable to the estimates of the original data set, indicating the stability of the developed model. Furthermore, estimates of interindividual variability were considered stable, as these were essentially equal to those generated with the original data (0% to 15% deviation). In conclusion, the complex pharmacokinetics of E7070 were accurately described by the model presented, which also provided stable parameter estimates. BSA-guided dosing appears to be important for E7070. This model may form the basis for the optimization of the therapeutic window of E7070. Further studies will focus on the development of limited sampling strategies for use in phase II clinical studies and on pharmacokinetic-pharmacodynamic relationships.
1. Owa T, Okachi T, Yoshimatsu K, et al: A focused compound library of novel N-(7-indolyl)benzenesulfonamides for the discovery of potent cell cycle inhibitors. Bioorg Med Chem Lett 10: 1223-1226, 2000[CrossRef][Medline] 2. Owa T, Yoshino H, Okauchi T, et al: Discovery of novel anti-tumor sulfonamides targeting G1 phase of the cell cycle. J Med Chem 42: 3789-3799, 1999[CrossRef][Medline] 3. Fukuoka K, Usuda J, Fukumoto H, et al: Mechanism of action of a novel sulfonamide antitumor agent, E7070. Proc Am Assoc Cancer Res 41: 59, 2000 (abstr) 4. Raymond E, ten Bokkel Huinink WW, Taieb J, et al: Phase I and pharmacokinetic study of E7070, a new chloro-indolysulfonamide, given as a 1-hour infusion every 3 weeks in patients with advanced solid tumors. Proc Am Assoc Cancer Res 41: 611, 2000 (abstr)
5. Punt CJA, Fumoleau P, van de Walle B, et al: Phase I and pharmacokinetic study of E7070, a novel sulfonamide, given at a daily times 5 schedule in patients with solid tumors: A study by the EORTC-Early Clinical Studies Group (ECSG). Ann Oncol 12: 1289-1293, 2001 6. Droz JP, Roch H, Zanetta S, et al: Phase I trial of five-days continuous infusion E7070 [N(3-chloro-7-indoly)-1,4-bezene-disulfonamide] in patients with solid tumors. Proc Am Assoc Cancer Res 41: 609, 2000 (abstr) 7. Dittrich C, Dumez H, Calvert H, et al: Phase I and pharmacokinetic study of E7070 in patients with solid tumors as single IV infusion, weekly x 4, q 6 weeks. Proc Am Assoc Cancer Res 41: 609, 2000 (abstr) 8. Raymond E, Fumoleau P, Roché H, et al: Combined results of 4 phase I and pharmacokinetic studies of E7070, a novel chloroindoly-sulfonamide inhibiting the activation of cdk2 and cyclin E. Clin Cancer Res 6: 4529s, 2000 (abstr) 9. U.S. Department of Health and Human Services, Food and Drug Administration: Guidance for Industry: Population Pharmacokinetics. Washington DC, U.S. Department of Health and Human Services, 1999 10. Williams PJ, Ette EI: The role of population pharmacokinetics in drug development in light of the Food and Drug Administration’s ‘Guidance for Industry: Population pharmacokinetics’. Clin Pharmacokinet 39: 385-395, 2000[CrossRef][Medline] 11. Peck CC, Barr WH, Benet LZ, et al: Opportunities for integration of pharmacokinetics, pharmacodynamics and toxicokinetics in rational drug development. Clin Pharmacol Ther 51: 465-473, 1992[Medline] 12. Beal SL, Boeckman AJ, Sheiner LB: NONMEM: User’s Guides. San Francisco CA, University of California at San Francisco, 1988 -1992
13. Karlsson MO, Molnar V, Freijs A, et al: Pharmacokinetic models for the saturable distribution of paclitaxel. Drug Metab Dispos 27: 1220-1223, 1999 14. van Kesteren CH, Mathôt RAA, Raymond E, et al: Population pharmacokinetic analysis of the novel anti-cancer agent E7070 in three phase I clinical studies using non-linear mixed effects modeling (NONMEM). Proc Am Soc Clin Oncol 20: 93a, 2001 (abstr) 15. Ette E, Ludden T: Population pharmacokinetic modeling: The importance of informative graphics. Pharm Res 12: 1845-1855, 1995[CrossRef][Medline] 16. Ette E: Statistical graphics in pharmacokinetics and pharmacodynamics: A tutorial. Ann Pharmacother 32: 818-828, 1998[Abstract] 17. Jonsson EN, Karlsson MO: Xpose: An S-PLUS based model building aid for population analysis with NONMEM. Comp Meth Prog Biomed 58: 51-64, 1999[CrossRef][Medline] 18. Mandema JW, Verotta D, Sheiner LB: Building population pharmacokinetic models: I. Models for covariate effects. J Pharmacokinet Biopharm 20: 511-528, 1992[CrossRef][Medline] 19. Maitre PO, Bührer M, Thomson D, et al: A three-step approach combining Bayesian regression and NONMEM population analysis: Application to midazolam. J Pharmacokinet Biopharm 19: 377-384, 1991[CrossRef][Medline] 20. Karlsson MO, Jonsson EN, Wiltse CG, et al: Assumption testing in population pharmacokinetic models: Illustrated with an analysis of moxonidine data from congestive heart failure patients. J Pharmacokinet Biopharm 26: 207-246, 1998[CrossRef][Medline] 21. Ette E: Stability and predictive performance of a population pharmacokinetic model. J Clin Pharmacol 37: 486-495, 1997[Abstract] 22. Launay-Iliadis MC, Bruno R, Cosson V, et al: Population pharmacokinetics of docetaxel during phase I studies using non-linear mixed-effect modeling and nonparametric maximum-likelihood estimation. Cancer Chemother Pharmacol 37: 47-54, 1995[Medline] 23. Bruno R, Vivier N, Vergniol JC, et al: A population pharmacokinetic model for docetaxel (Taxotere): Model building and validation. J Pharmacokinet Biopharm 24: 153-172, 1996[CrossRef][Medline]
24. Bruno R, Hille D, Riva A, et al: Population pharmacokinetics/pharmacodynamics of docetaxel in phase II studies in patients with cancer. J Clin Oncol 16: 187-196, 1998 25. Baille P, Bruno R, Schellens JHM, et al: Optimal sampling strategies for Bayesian estimation of docetaxel (Taxotere) clearance. Clin Cancer Res 3: 1535-1538, 1997[Abstract]
26. Gallo JM, Laub PB, Rowinsky EK, et al: Population pharmacokinetic model for topotecan derived from phase I clinical trials. J Clin Oncol 18: 2459-2467, 2000 27. Nguyen L, Chatelut E, Chevreau C, et al: Population pharmacokinetics of total and unbound etoposide. Cancer Chemother Pharmacol 41: 125-132, 1998[CrossRef][Medline] 28. Thomson AH, Vasey PA, Murray LS, et al: Population pharmacokinetics in phase I development: A phase I study of PK1 in patients with solid tumours. Br J Cancer 81: 99-107, 1999[CrossRef][Medline] 29. Takama H, Tanaka H, Sudo T, et al: Population pharmacokinetic modeling and model validation of a spicamycin derivative, KRN5500, in a phase I study. Cancer Chemother Pharmacol 47: 404-410, 2001[CrossRef][Medline] 30. Huitema ADR, Mathôt RAA, Tibben MM, et al: A mechanism-based pharmacokinetic model for the cytochrome P450 drug-drug interaction between cyclophosphamide and thioTEPA and the autoinduction of cyclophosphamide. J Pharmacokinet Pharmacodyn 28: 211-230, 2001[CrossRef][Medline] 31. Kerbusch T, Huitema ADR, Ouwerkerk J, et al: Evaluation of the autoinduction of ifosfamide metabolism by a population pharmacokinetic approach using NONMEM. Br J Clin Pharmacol 49: 555-61, 2000[CrossRef][Medline] 32. Gurney H: Dose calculation of anticancer drugs: A review of the current practice and introduction of an alternative. J Clin Oncol 14: 2590-2611, 1996[Abstract] 33. Van den Bongard HJGD, Mathôt RAA, Beijnen JH, et al: Pharmacokinetically guided administration of chemotherapeutic agents. Clin Pharmacokinet 39: 345-367, 2000[CrossRef][Medline] 34. Van den Bongard HJGD, Pluim D, Rosing H, et al: Metabolism and excretion of E7070 after intravenous administration to cancer patients. Anticancer Drugs (in press) Submitted January 2, 2002; accepted June 17, 2002.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2002 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
1 * (1 + 
