|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2001 American Society for Clinical Oncology Mechanism-Based Pharmacokinetic Model for PaclitaxelByFrom the Division of Pharmacokinetics and Drug Therapy, Department of Pharmaceutical Biosciences, Faculty of Pharmacy, Uppsala University, Division of Medical Oncology, Istituto Nazionale Tumori, Milan, Italy; and Department of Medical Oncology, Rotterdam Cancer Institute, Rotterdam, the Netherlands. Address reprint requests to Mats O. Karlsson, PhD, Division of Pharmacokinetics and Drug Therapy, Department of Pharmaceutical Biosciences, Faculty of Pharmacy, Uppsala University, Box 591, SE-751 24 Uppsala, Sweden; email: mats.karlsson{at}farmbio.uu.se
PURPOSE: To create a model based on known mechanisms of paclitaxel distribution that could describe the pharmacokinetics (PK) of total and unbound plasma concentrations, as well as blood concentrations. In addition, to investigate the relationship between exposure, based on unbound and total concentrations, and neutropenia. PATIENTS AND METHODS: Paclitaxel and Cremophor EL (CrEL) concentrations were obtained from 23 female and three male patients (50 courses in total) with different cancer types that received paclitaxel (Taxol; Bristol-Myers Squibb Co, Princeton, NJ) (135 to 225 mg/m2) as 3- or 24-hour intravenous infusions. Seven of the patients received combination therapy with doxorubicin or cisplatin. The population PK model was built to fit three types of data simultaneously: unbound, total plasma, and blood concentrations. The area under the curve, threshold, and general models were used to relate neutrophil survival fraction from 19 patients (29 courses in total) to exposure based on unbound and total plasma concentration, respectively. RESULTS: The PK model included a linear three-compartment model for unbound concentration, binding directly proportional to CrEL, linear and nonlinear binding to plasma proteins, and linear and nonlinear binding to blood cells. The threshold model best described the PK/pharmacodynamic (PD) relationship for total concentration. No distinction could be made between the models for unbound drug. CONCLUSION: Earlier PK models for paclitaxel have been empirical. This study shows that a mechanistic model can be used to describe the nonlinear PK of paclitaxel. There is an indication that the PK/PD relationship is not the same for unbound and total plasma concentrations.
DUE TO ITS poor solubility, the anticancer drug paclitaxel is administered with the micelle-forming vehicle Cremophor EL (CrEL). CrEL has recently been proposed to cause the nonlinear pharmacokinetic behavior of paclitaxel in plasma by trapping the drug in micelles and thereby making it less available for distribution to tissues, metabolism, and biliary excretion.1 Earlier, the nonlinearity has been described as both saturable elimination and saturable distribution, where the saturable distribution has been described as saturable transport2 or saturable binding.3 The free fraction of paclitaxel has been shown to be inversely related to CrEL concentrations in vitro,1,4 and CrEL has also been shown to alter the blood-to-plasma ratio in vitro and in vivo by reducing the uptake in RBCs.4 When paclitaxel, dissolved in another vehicle, was administered to mice, no pharmacokinetic (PK) nonlinearity in plasma concentration profiles was evident, and the concentrations in tissues also increased linearly with increasing dose even when dissolved in CrEL, suggesting linear kinetics for the unbound drug.5 No nonlinearity is present in the kinetics of docetaxel (Taxotere; Aventis, Antony, France) where Tween-80 is used as solvent for docetaxel. Tween-80 is rapidly degraded in plasma and does not interfere with the kinetics of docetaxel.6 The exposure-toxicity relationship of paclitaxel has mostly been described with a threshold model,7,8 according to which the toxicity is related to the duration of exposure above a certain threshold concentration, but other models have been proposed.9,10 The action exerted by drugs is generally more closely related to unbound than total concentration of drugs. Any PK/pharmacodynamic (PD) relationship based on total concentration, therefore, needs to be re-evaluated if the ratio of unbound to total drug is found to vary. For paclitaxel, a correct PK/PD relationship is particularly important if extrapolations are to be made to other conditions (schedules, vehicles, or routes of administration) than those from which it has been derived. The purpose of the present study was to develop a mechanism-based PK model that could explain the concentration-time profile on unbound concentration as well as those of total plasma and blood concentrations. In addition, we wanted to reinvestigate the PK/PD relationship between exposure and toxicity in the light of the new PK information.
Patients and Sampling Twenty-three female and three male patients, 32 to 69 years of age, with different cancer types (ovary, breast, bladder, lung, and adenocarcinoma with unknown primary) were included in this study. Total plasma concentrations were measured in 26 patients, 50 courses in total. Twenty patients (41 courses) received 3-hour intravenous (IV) infusions (135, 150, 175, 200, and 225 mg/m2) of paclitaxel (Taxol; Bristol Myers Squibb Co, Princeton, NJ). Five of these patients received concurrent therapy with doxorubicin, and two received concurrent therapy with cisplatin. Doxorubicin 60 mg/m2 was administered as a 5-minute IV infusion 24 hours (first course) or 15 minutes (subsequent courses) before paclitaxel (150 or 200 mg/m2). Cisplatin 75 mg/m2 was administered as a 3-hour IV infusion immediately after paclitaxel (225 mg/m2). Six patients (nine courses) received 24-hour IV infusions (135 and 175 mg/m2). None of these had concurrent therapy with doxorubicin or cisplatin. Unbound plasma concentrations were measured in seven patients (three males and four females) (19 courses) that received 3-hour infusions at dose levels of 135, 175, and 225 mg/m2. Five of the patients received all three dose levels, and two of the patients received 175 and 225 mg/m2. The patient characteristics of these seven had similar median values as the total population and all cancer types were represented. Total blood concentrations where obtained from the same patients as for unbound concentrations with the addition of one course from another female that received 157.5 mg/m2 as a 3-hour infusion. CrEL concentrations were measured in all patients. For CrEL and total plasma paclitaxel concentrations, on average, 12 samples (range, five to 18 samples) and 10 samples (range, six to 13 samples) per course were obtained for the 3- and 24-hour infusions, respectively. The number of samples per course were, on average, 10 (range, seven to 12 samples) and 11 (range, eight to 16 samples) for unbound and blood concentrations, respectively. The exact sampling time points can be seen in Fig 1.
Baseline and nadir values of absolute neutrophil count (ANC) were obtained from 24 individuals (40 courses). Five of these patients (seven courses) received paclitaxel as 24-hour infusions and the remaining 19 patients as 3-hour infusions. Five patients (11 courses) received combination therapy with doxorubicin, and two patients (two courses) received paclitaxel in combination with cisplatin. Blood samples for toxicity were collected once or twice a week according to the clinical trial protocols. The sampling frequency was not dependent on the value of the previous measurement. The lowest observed value was used as nadir. The baseline value was obtained 1 day before or on the day of the infusion. Patient characteristics are listed in Table 1. The trials from which the data was obtained were approved by the Ethical Committee of the Istituto Nazionale Tumori or Rotterdam Cancer Institute Ethics Board, and all patients had given their informed consent.
Assays Concentrations of paclitaxel were measured using an isocratic reversed-phase high-performance liquid chromatographic method with UV detection at 230 nm, as described previously.11 This assay has a lower limit of quantitation (LOQ) of 10 ng/mL (0.0117 µmol/L), with an accuracy (ie, percentage deviation from nominal concentrations) of ± 2.77% and within-run and between-run precision of less than 2.99% and less than 2.76%, respectively. Observations between limit of detection and LOQ were also used in the analysis. The same assay methodology was also used for determination of paclitaxel in whole blood, and was validated as described11 using spiked quality control samples containing 40, 200, 400, and 15,000 ng of paclitaxel/mL. The within-run precision and between-run precision ranged from 2.23% to 8.74% and 1.81% to 8.07% (n = 14 at each of the concentrations), respectively, with a mean percentage deviation from nominal values less than ± 3.30%. A potential CrEL-concentration dependency in extraction recovery of paclitaxel from whole blood was tested by repeated analysis of spiked whole blood (nominal concentration: 1,000 ng/mL) containing 0, 0.1, 0.5, 1.0, 5.0, or 10 µL of CrEL/mL. The mean (± SD) recovered concentrations in these samples were 1,037 ± 28.4, 1,067 ± 33.7, 1,064 ± 7.91, 1,074 ± 1.90, 1,041 ± 35.1, and 1,057 ± 6.71 ng/mL, respectively, with no trend in significant deviations from the control value (P = .194; n = 16 at each CrEL concentration). The detector response of the internal standard docetaxel, as measured by the peak height, was also not significantly different during analysis of either plasma or whole blood samples with mean values of 44,019 ± 4,739 µV and 43,728 ± 5,410 µV, respectively (P = .525; n = 164 for each matrix). These data were considered acceptable for analysis of paclitaxel in whole blood samples obtained from patients for the conduct of the present study. The analytic procedure for CrEL was based on a colorimetric dye–binding assay using Coomassie brilliant blue G-250.12,13 The lower limit of quantitation of this procedure was 0.50 µL/ml, with an accuracy of less than 6.33% and within-run and between-run precision of less than 9.47% and less than 1.83%, respectively, in the tested concentration range (ie, 0.50 to 10.0 µL/mL). Observations between limit of detection and LOQ were also used in the analysis. Unbound concentrations were obtained from an equilibrium dialysis method.14 In this method, free fraction of paclitaxel is measured at 37°C in a humidified atmosphere of 5% CO2 using test cells made from 2.0-mL polypropylene Safe-lock vials (Eppendorf, Hamburg, Germany), carrying a 260-µL inside recess in the lids. Before incubation, 2 µL of a [3H] paclitaxel solution in ethanol was added to 300 µL of plasma, followed by mixing for 10 seconds. Dialysis was carried out with 260 µL of this sample against an equal volume of phosphate-buffered saline (PBS) for 24 hours in a moist chamber, which is sufficiently long to attain equilibrium. Spectra/Por 3 (Spectrum Medical, Kitchener, Canada) dialysis tubing with a molecular weight cutoff of 12,500 was soaked in PBS before use. After establishment of equilibrium, 150 µL of buffer solution, containing only unbound paclitaxel, and 150 µL of the plasma fraction, containing both bound and unbound drug, were transferred to separate 2-mL vials and 1.9 mL of Emulsifier-Safe (Packard, Groningen, the Netherlands) scintillation cocktail were added. After manual mixing for 30 seconds, the tritiated-paclitaxel was quantified by liquid scintillation counting. All samples were counted until a preset time of 20 minutes was reached, with quench correction by external standardization. The ratio of drug concentrations measured in the buffer and plasma after dialysis was taken as an estimate of paclitaxel free drug fraction (fu). Because the volume shift during dialysis was negligible (< 10%), the results were used directly without applying a correction factor. The addition of tritiated paclitaxel was performed with all samples, including the blank samples that were taken from the patients before drug administration. Experiments demonstrated that the radioactivity seen in the PBS fraction was not derived from any radioactivity that might have been associated with the polypropylene tube wall or the membrane. However, very small amounts of tritiated water, which accounted for less than 4.7% of total radioactivity, were formed during the 24-hour incubation experiments and were independent of the CrEL concentration tested. The PBS solution collected after dialysis of paclitaxel was also analyzed by a specific high-performance liquid chromatographic method to exclude an artefact due to alteration of the fraction of total radioactivity associated with unchanged compound. During the establishment of the method, duplicate or triplicate plasma samples with differing paclitaxel fu values, depending on the spiked CrEL concentration, were subject to repeat analysis on 6 consecutive days to enable assessment of reproducibility. The mean relative SD of these samples was 9.2% (n-82), assuring high discriminatory power in the detection of changes in paclitaxel fu in the presence of CrEL in the patients. With the final method, the within-run and between-run variability were less than 9.20% and less than 3.30%, respectively, and the lower limit of quantitation of the free fraction ratio was 1%. These values are comparable with that observed for a variety of other compounds using equilibrium dialysis or ultrafiltration.
PK Analysis For the model relating unbound to total plasma concentrations, different combinations of linear binding to plasma constituents, nonlinear binding to plasma constituents, and binding related to CrEL were tried. In this relation, the observed concentrations of CrEL were used. The blood concentration model part was modelled with a linear and/or a nonlinear binding to blood cells. In the relation between blood and plasma concentration, the hematocrit (H) was used according to equation 1:
where Cb, Cbc, and Cp are blood, blood cell, and total plasma concentration of paclitaxel, respectively.
The model parameters were estimated in a nonlinear mixed effects (population) analysis, where data from all patients are analyzed simultaneously. This requires models for interindividual (IIV) and interoccasion (IOV) variability in parameters. For both types of variability, log-normal parameter distribution was assumed and the IOV (or course-to-course variability) was implemented as previously described.15 Residual error was modelled with an additive ( All analyses were performed with the first-order conditional estimation method with interaction in the NONMEM program, version V/VI beta (S.L. Beal and L.B. Sheiner, University of California, San Francisco, CA). The models can be implemented in NONMEM version V. NONMEM version VI was used to shorten run-times. Graphical diagnostics, using the Xpose program,16 and comparison of competing models, using the objective function values (OFV) in the likelihood ratio test, guided the model development. A difference in OFV of more than 10.83, corresponding to a significance level P < .001, was used for discrimination between two hierarchical models differing in one parameter.
PK/PD Analysis The basis for the general model is that the transformation of the plasma concentration most closely related to the toxicity is estimated as part of the model.9 The result of this transformation can be thought of as a direct effect, which is not observable, but ultimately leads to the observed effect. The model is not relying on such a direct effect to exist, but it is a useful hypothesis for describing the model. The direct effect, Edir is a function of the drug concentration. The observed effect, Eobs is related to the area under the Edir-time curve, AUCEdir. Sigmoidal Emax models are used to describe the relationships as follows in equations 2 and 3:
where ß and
The threshold and AUC models are special cases of the general model. For the threshold model, ß is infinitely high and EC50 is equivalent to the threshold concentration, and for the AUC model, ß is 1 and EC50>>C. For the threshold modelling, ß was fixed at 20, and EC50 (the threshold concentration) was estimated. When developing the AUC model, AUCEdir was set to the area under the plasma concentration time curve.
Log-normal distribution was used when modelling IIV. Residual error (which in this case includes IOV) was modelled with an additive (
PK Analysis The observed unbound plasma, total plasma, and blood paclitaxel concentrations as well as CrEL concentrations are shown in Fig 1. Although a two-compartment model could approximate the unbound concentration-time profile without marked misfit, there was a preference for the three-compartment model (  OFV = 85.1), and the parameters were in general well estimated (Table 2) and the observed data well described (Fig 2). The mean half-lives of the unbound concentrations were 0.21 hours (range, 0.073 to 0.34 hours), 1.4 hours (range, 1.2 to 2.1 hours) and 7.0 hours (range, 3 to 20 hours) for the alpha, beta, and gamma phase, respectively. No dose-dependency was evident. One way of investigating that was by treating each study occasion as a separate individual and inspecting the individual estimates of clearance. These showed no trend with dose, regardless of only the individuals with measured unbound concentration or all individuals were considered (r2 values in linear regressions were  0.01). Volume terms showed a similar lack of dose dependence.
Bound plasma concentration was modelled with the following three components: (1) a binding component directly proportional to CrEL concentrations, (2) a linear binding, and (3) a nonlinear binding to other plasma components. Each of these components was significant because when each of component 1, 2, or 3 was omitted from the final model, the OFV increased with 259, 58, and 68 units, respectively. The requirement to include multiple plasma binding components is also evident from Fig 3. The free fraction of paclitaxel is higher at early time points than at later points, even though the concentration of CrEL is the same. This would imply that at higher concentration the plasma protein binding is saturated.
With the different binding components, the total plasma concentration could be described with the equation 5:
where the definition and estimated values of BCrEL, Blin, p, Bmax, p, and Km, p are given in Table 2. The model predicts that in the absence of CrEL, plasma protein binding would be 85%. The observed ratio Cu/Cbc is increasing with increasing Cu (Fig 4), suggesting some saturable process in the binding to blood cells why a nonlinear binding component is needed in the model. The blood cell binding was then modelled using two components, a linear and a nonlinear component. Each of these components was significant, because when omitted, the OFV increased with 28 and 64 units, respectively. The binding to blood cells could thus be described with equation 6, from which the total blood concentration could be obtained from equation 1:
equation
where the definition and estimated values of Blin, bc, Bmax, bc, and Km, bc are given in Table 2. The value of Bmax, p is very low in view of concentrations of plasma proteins and that could be because of the difficulty in estimating the different parameters when combining linear and nonlinear parts because the observed will be a combination of the two. The value of Bmax, bc is in agreement with previously reported Bmax to platelets in vitro. Km values for nonlinear binding components to both plasma and blood cells are very low and relatively poorly estimated. This is probably because of the Km values being in a region at the low range of the observed concentrations and, therefore, difficult to estimate properly. The only patient characteristic that is included in the final model is the hematocrit value. It was statistically significant (P < .001) to use individual values rather than the median value for hematocrit when describing total blood concentration. IIV in CL was relatively low, 25%, but higher than the IOV (course-to-course), which was estimated to be 20%. Both IIV and IOV in distribution parameters were higher (Table 2). Notable is that no IIV or IOV is included in any of the parameters that determine the relationship between unbound concentration and total plasma concentration. Thus, this seems to be showing little IIV, once the CrEL concentration is known. The same is true for the relation between unbound concentration and total concentration in blood where no IIV is apparent once the hematocrit is determined. This should mean that once total plasma paclitaxel concentration and CrEL concentrations are known, unbound concentrations can be predicted with good precision by the model. To test this, observed unbound concentrations were deleted from the individuals for which such existed. Thereafter, the deleted unbound concentrations were predicted based on total plasma paclitaxel concentrations, observed CrEL concentrations, and the model. The observations of the unbound concentrations were in this situation well predicted. A linear regression of observed concentrations versus these predicted concentrations had the following slope, intercept, and r2 value: 1.004, 0.001, and 0.927, respectively. The same parameters for the situation when unbound concentrations were part of the data were (ie, lower left panel of Fig 2) as follows: 1.016, 0.001, and 0.929. Also, the individual parameter estimates of CL and other parameters showed equally good correspondence between the situation when all paclitaxel data were used and when only total paclitaxel concentrations were used. The relative contribution to the total concentration in blood and plasma will vary over time administration of paclitaxel formulated in CrEL. How the different components may vary in individuals receiving 3- and 24-hour infusions are shown in Fig 5, where individual-predicted CrEL concentrations were used to simulate the concentrations during infusion.
PK/PD Analysis Preliminary results showed that the patients who received combination therapy with doxorubicin had an increased toxicity and the toxicity could not be related to the paclitaxel exposure, eg, AUC. Because both doxorubicin and paclitaxel cause neutropenia, the toxicity does not depend solely on one drug, and therefore, patients receiving this combination were excluded when modelling the paclitaxel exposure-toxicity relationship (the modelling was thus based on 29 courses from 19 patients).
When total plasma concentration was used as exposure measure, the threshold model was significantly better than the AUC model
When unbound plasma concentration was used as exposure measure, the general, threshold, and AUC models describe the data equally well. The parameter estimates and relative SE (%) for the different models are listed in Table 3.
Estimating Eobs, max and Including the value of neutrophils at baseline as covariate on the time needed of direct effect to produce half maximal observed effect (AUCE50) resulted in a significant (P < .05) improvement of the fit in the general and AUC model for unbound concentration and borderline significant (P < .1) improvement in the threshold models for both total and unbound concentration. AUCE50 was estimated to increases with 15%, 13%, 9%, and 7% per unit higher baseline value, respectively.
Previous models of paclitaxel PK have been empirical and included components such as saturable tissue binding or saturable transport to tissues, which have little or no mechanistic basis. The present model on the other hand have a foundation in the known properties of paclitaxel distribution as it have been determined in vitro with micelle trapping with CrEL,4 distribution to RBCs,4 and binding to albumin, alpha acid glycoprotein,17 and platelets.18 The results show that the nonlinear PK of paclitaxel could be explained by these nonlinear binding components, mainly CrEL binding and that unbound drug displayed linear PK. No additional saturable transport or metabolism, as in previous models, were required to appropriately describe paclitaxel PK. Neither does there seem to be support in these data for other suggested nonlinear mechanisms, such as P-glycoprotein–mediated inhibition of biliary drug excretion.19 However, unbound concentrations were only available from 3-hour infusions, so it should be noted that nonlinearity in the unbound concentration cannot be ruled out entirely. The fraction not bound to either plasma proteins or CrEL is a rather small fraction of the total. At high CrEL concentrations, paclitaxel is mainly bound to CrEL. From the simulated concentration components in the 24-hour infusion, we can see that because CrEL concentrations are rather low, the linear binding to plasma proteins and binding to blood cells are of greater importance than the CrEL binding. It has been reported that the volume of distribution for CrEL is small,20 thereby indicating that its distribution is more or less limited to plasma. Therefore, the nonlinear binding as affecting the distribution in the central compartment only can be justified. However, an additional complexity besides the fact that the CrEL concentration changes with time is the fact that CrEL itself shows nonlinear kinetics. CrEL has schedule-dependant PK in that it has a shorter half-life in patients that received 24-hour infusion than patient that received 3-hour infusion. Not much is known about the elimination pathways of CrEL. We have found that a rather slow degradation occurs in plasma (unpublished data). If plasma is the only site of elimination, esterases (or whatever enzymes are involved) might be saturated. One could also speculate that if CrEL is in a micellar form, the degradation enzymes could not be very efficient; whereas if CrEL is in a free form, it is available for metabolism. Further, if the concentration of CrEL were lower than the critical micelle concentration (CMC) in plasma, paclitaxel would not be trapped. No alteration in free fraction was seen below CMC in vitro (water solution).4 The CMC in vivo has not been determined, and it could be rather different from the in vitro situation because of interference by macromolecules present in the body. Thus, the model that uses total CrEL concentration to describe trapping is a simplification (no CMC, homogeneous CrEL, and no competing agents for micellar trapping), and further studies may be able to refine this part of the model. Covariate-parameter relationships are usually included in population PK and PK/PD models. However, because of the small size of the present study, such relationships were not evident when investigated with standard procedures21,22 apart from those presented for hematocrit and ANC baseline. When total concentrations are used, the threshold model is significantly better. The estimated threshold concentration of 0.197 µmol/L and time above threshold to get half Eobs, max of 14.4 hours are in agreement with previous studies where 11.16 hours above 0.1 µmol/L8 and 17.4 hours above 0.05 µmol/L7 were predicted to yield a 50% decrease in ANC. Thus, despite its being sparse, the data seem to be sufficient to identify important components of the PK/PD relationship for total concentration in agreement with previous reports.7,8 However, no distinction could be made between the various models (threshold, AUC, or general model) when exposure is based on unbound concentrations. This indicates that the exposure-toxicity may be different for unbound compared with total concentration. The present data did not support, possibly because of lack of data, the notion that unbound concentration is more closely related to drug effects compared with total concentrations. This is, however, a generally accepted hypothesis in clinical pharmacology, and if it is true for paclitaxel, it may have profound consequences for the optimal use of the drug whenever new therapies (doses, infusion duration, routes of administration, or formulation vehicle) with paclitaxel are considered. The present model allows, to some extent, the consequence with respect to unbound concentration and toxicity to be simulated for such situations. However, to act as a more powerful simulation tool, a more precise relationship between unbound paclitaxel concentration and toxicity needs to be defined. With extensions, such a model may also serve the purpose of being able to better predict or explain PK and PD drug interactions. In conclusion, the nonlinear PK of paclitaxel can be described mechanistically, without any need for the empirical components as has been used in previous models. The clinical implications for this is that in the absence of CrEL, paclitaxel can be expected to have essentially linear PK, in particular the unbound concentration displays linear PK. Also, it suggests that the unbound concentration-time profile can be predicted based on total concentration and CrEL concentrations. PK/PD relationships for paclitaxel hematologic toxicity established based on total concentration are not likely to be valid in the absence of CrEL, which should be considered when new routes of administration or new formulations of paclitaxel are planned. Failure to do so is likely to result in excessive hematologic toxicity.
Supported by Swedish Cancer Society, Bristol Myers Squibb Pharmaceutical Research Institute, and Associazione Italiana Ricerca sul Cancro (AIRC).
1. van Tellingen O, Huizing MT, Panday VR, et al: Cremophor EL causes (pseudo-) non-linear pharmacokinetics of paclitaxel in patients. Br J Cancer 81: 330-335, 1999[Medline] 2. Sonnichsen DS, Hurwitz CA, Pratt CB, et al: Saturable pharmacokinetics and paclitaxel pharmacodynamics in children with solid tumors. J Clin Oncol 12: 532-538, 1994[Abstract]
3.
Karlsson MO, Molnar V, Freijs A, et al: Pharmacokinetic models for the saturable distribution of paclitaxel. Drug Metab Dispos 27: 1220-1223, 1999
4.
Sparreboom A, van Zuylen L, Brouwer E, et al: Cremophor EL-mediated alteration of paclitaxel distribution in human blood: clinical pharmacokinetic implications. Cancer Res 59: 1454-1457, 1999
5.
Sparreboom A, van Tellingen O, Nooijen WJ, et al: Nonlinear pharmacokinetics of paclitaxel in mice results from the pharmaceutical vehicle Cremophor EL. Cancer Res 56: 2112-2115, 1996
6.
van Tellingen O, Beijnen JH, Verweij J, et al: Rapid esterase-sensitive breakdown of polysorbate 80 and its impact on the plasma pharmacokinetics of docetaxel and metabolites in mice. Clin Cancer Res 5: 2918-2924, 1999
7.
Gianni L, Kearns CM, Giani A, et al: Nonlinear pharmacokinetics and metabolism of paclitaxel and its pharmacokinetic/pharmacodynamic relationships in humans. J Clin Oncol 13: 180-190, 1995
8.
Huizing MT, Keung AC, Rosing H, et al: Pharmacokinetics of paclitaxel and metabolites in a randomized comparative study in platinum-pretreated ovarian cancer patients. J Clin Oncol 11: 2127-2135, 1993 9. Karlsson MO, Molnar V, Bergh J, et al: A general model for time-dissociated pharmacokinetic-pharmacodynamic relationship exemplified by paclitaxel myelosuppression. Clin Pharmacol Ther 63: 11-25, 1998[Medline] 10. Minami H, Sasaki Y, Saijo N, et al: Indirect-response model for the time course of leukopenia with anticancer drugs. Clin Pharmacol Ther 64: 511-521, 1998[Medline] 11. Sparreboom A, de Bruijn P, Nooter K, et al: Determination of paclitaxel in human plasma using single solvent extraction prior to isocratic reversed-phase high-performance liquid chromatography with ultraviolet detection. J Chromatogr B Biomed Sci Appl 705: 159-164, 1998[Medline] 12. Brouwer E, Verweij J, Hauns B, et al: Linearized colorimetric assay for cremophor EL: application to pharmacokinetics after 1-hour paclitaxel infusions. Anal Biochem 261: 198-202, 1998[Medline] 13. Sparreboom A, Loos WJ, Verweij J, et al: Quantitation of Cremophor EL in human plasma samples using a colorimetric dye-binding microassay. Anal Biochem 255: 171-175, 1998[Medline]
14.
Brouwer E, Verweij J, De Bruijn P, et al: Measurement of fraction unbound paclitaxel in human plasma. Drug Metab Dispos 28: 1141-1145, 2000 15. Karlsson MO, Sheiner LB: The importance of modeling interoccasion variability in population pharmacokinetic analyses. J Pharmacokinet Biopharm 21: 735-750, 1993[Medline] 16. Jonsson EN, Karlsson MO: Xpose: An S-PLUS based population pharmacokinetic/pharmacodynamic model building aid for NONMEM. Comput Methods Programs Biomed 58: 51-64, 1999[Medline] 17. Kumar GN, Walle UK, Bhalla KN, et al: Binding of taxol to human plasma, albumin and alpha 1-acid glycoprotein. Res Commun Chem Pathol Pharmacol 80: 337-344, 1993[Medline] 18. Wild MD, Walle UK, Walle T: Extensive and saturable accumulation of paclitaxel by the human platelet. Cancer Chemother Pharmacol 36: 41-44, 1995[Medline]
19.
Gianni L, Vigano L, Locatelli A, et al: Human pharmacokinetic characterization and in vitro study of the interaction between doxorubicin and paclitaxel in patients with breast cancer. J Clin Oncol 15: 1906-1915, 1997 20. Sparreboom A, Verweij J, van der Burg ME, et al: Disposition of Cremophor EL in humans limits the potential for modulation of the multidrug resistance phenotype in vivo. Clin Cancer Res 4: 1937-1942, 1998[Abstract] 21. Jonsson EN, Karlsson MO: Automated covariate model building within NONMEM. Pharm Res 15: 1463-1468, 1998[Medline] 22. Mandema JW, Verotta D, Sheiner LB: Building population pharmacokinetic–pharmacodynamic models: I. Models for covariate effects. J Pharmacokinet Biopharm 20: 511-528, 1992[Medline] Submitted November 27, 2000; accepted June 4, 2001.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2001 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
TOP
ABSTRACT
INTRODUCTION
1) and a proportional part (
are sigmoidicity factors. Because Edir is not observed, Edir, max only scales AUCEdir, 50 and can be set to 1. The parameter AUCEdir, 50 is the duration of maximal direct effect needed to produce half-maximal observed effect. Survival fraction of neutrophils at nadir (SFN) was used as observed effect and was modelled as follows in 
